Post on 11-Aug-2020
Special Article
Nut consumption and incidence of cardiovascular diseases andcardiovascular disease mortality a meta-analysis of prospectivecohort studies
Nerea Becerra-Tomas Indira Paz-Graniel Cyril WC Kendall Hana Kahleova Dario RahelicJohn L Sievenpiper and Jordi Salas-Salvado
Context Previous meta-analyses evaluating the association between nut consump-tion and the risk of cardiovascular disease (CVD) had substantial methodological lim-itations and lacked recently published large prospective studies hence making anupdated meta-analysis highly desirable Objective To update the clinical guidelinesfor nutrition therapy in relation to the European Association for the Study ofDiabetes (EASD) a systematic review and meta-analysis of prospective studies wasconducted using the Grading of Recommendations Assessment Development andEvaluation (GRADE) system to summarize the evidence of the association betweentotal nuts specific types of nuts and the incidence of and mortality from CVD out-comes Data sources Relevant articles were identified by searching the PubMedand Cochrane databases Data extraction Two independent researchers screenedthe articles to identify those that met the inclusion criteria Data analysis The in-verse variance method with fixed-effect or random-effects models was used to pooldata across studies (expressed as risk ratio [RR] and 95 confidence interval [CI])Heterogeneity was tested and quantified using the Cochrane Q test and I2-statisticrespectively The GRADE system was used to assess the quality of the evidenceResults Nineteen studies were included in the analyses The results revealed an in-verse association between total nut consumption (comparing highest vs lowest cat-egories) and CVD incidence (RR 085 95CI 080091 I2 0) CVD mortality (RR077 95CI 072ndash082 I2 3) coronary heart disease (CHD) incidence (RR 08295CI 069ndash096 I2 74) CHD mortality (RR 076 95CI 067ndash086 I2 46)stroke mortality (RR 083 95CI 075ndash093 I2 0) and atrial fibrillation (RR 08595CI 073ndash099 I2 0) No association was observed with stroke incidence and
Affiliation Nerea Becerra-Tomas Indira Paz-Graniel and Jordi Salas-Salvado are with the Human Nutrition Unit Department of Biochemistryand Biotechnology Universitat Rovira i Virgili Reus Spain the Nutrition Unit University Hospital of Sant Joan de Reus Reus Spain and theInstitut drsquoInvestigacio Sanitaria Pere Virgili (IISPV) Reus Spain Nerea Becerra-Tomas and Jordi Salas-Salvado are with the Centro deInvestigacion Biomedica en Red Fisiopatologıa de la Obesidad y la Nutricion (CIBEROBN) Institute of Health Carlos III Madrid Spain Cyril WCKendall and John L Sievenpiper are with the Department of Nutritional Sciences Faculty of Medicine University of Toronto Toronto OntarioCanada and the Toronto 3D Knowledge Synthesis and Clinical Trials Unit Clinical Nutrition and Risk Factor Modification Centre St MichaelrsquosHospital Toronto Ontario Canada Cyril WC Kendall is with the College of Pharmacy and Nutrition University of Saskatchewan SaskatoonSaskatchewan Canada Hana Kahleova is with the Institute for Clinical and Experimental Medicine Prague Czech Republic and the PhysiciansCommittee for Responsible Medicine Washington District of Columbia USA Dario Rahelic is with Vuk Vrhovac University Clinic for DiabetesEndocrinology and Metabolic Diseases Merkur University Hospital Zagreb Croatia and the School of Medicine University of Zagreb ZagrebCroatia John L Sievenpiper is with the Li Ka Shing Knowledge Institute St Michaelrsquos Hospital Toronto Ontario Canada John L Sievenpiper iswith the Division of Endocrinology and Metabolism Department of Medicine St Michaelrsquos Hospital Toronto Ontario Canada
Correspondence Jordi Salas-Salvado and Nerea Becerra-Tomas Human Nutrition Unit Faculty of Medicine and Health Sciences UniversitatRovira i Virgili CSant Llorenc 21 43201 Reus Spain E-mails jordisalasurvcat and nereabecerraurvcat
Key words cardiovascular disease meta-analysis nuts peanut butter peanuts tree nuts walnuts
VC The Author(s) 2019 Published by Oxford University Press on behalf of the International Life Sciences InstituteAll rights reserved For permissions please e-mail journalspermissionsoupcom
doi 101093nutritnuz042Nutrition ReviewsVR Vol 77(10)691ndash709 691
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
heart failure The certainty of the evidence ranged from moderate to very lowConclusions This systematic review and meta-analysis revealed a beneficial roleof nut consumption in reducing the incidence of and mortality from different CVDoutcomes
INTRODUCTION
Cardiovascular disease (CVD) is an important publichealth issue According to the World HealthOrganization (WHO) it is the leading cause of death
worldwide affecting approximately 177 million peoplein 20151 Importantly CVDs are susceptible to behavior
modifications In this sense a healthy diet is one of thelifestyle components that could be promoted to help ad-
dress this global health concern2 Different healthy die-tary patterns such as the Mediterranean diet
(MedDiet)3 the DASH diet4 or a vegetarian diet5 havenuts as a key food component Nuts despite their high
fat content (mainly unsaturated fatty acids)6 are alsorich in minerals vitamins fiber and bioactive com-
pounds7 Given this exceptional nutritional profile fre-quent nut consumption has been inversely associated
with a lower risk of CVD in large prospective cohortstudies which have been summarized in several meta-
analyses8ndash20 Nonetheless some of the previous meta-analyses had methodological limitations such as the in-
clusion of studies with nuts plus seeds or legumes as ex-posure the inclusion of studies combining different
outcomes across analyses (eg inclusion of studies withonly fatal CHD [coronary heart disease] outcome in the
CVD mortality analysis) and the inclusion of studieswithout the first category of exposure as reference
Moreover since publication of the last meta-analysesthe results of two new large prospective cohort studies
evaluating the association between nut consumptionand CVD outcomes have been published2122 One study
reported updated results from the Nursesrsquo Health StudyI (NHSI) Nursesrsquo Health Study II (NSHII) and Health
Professionals Follow-up Study (HPFS) comprising upto 32 years of follow-up and a large number of cases21
Another reported the association between nut con-sumption and the incidence of 7 CVD outcomes in apopulation of 32 911 males22 Importantly most of the
previous meta-analyses have focused on total nut in-take and only a few have taken into account the poten-
tial associations between specific types of nutconsumption and the risk of CVD outcomes which
may vary considerablyTherefore taking into consideration the aforemen-
tioned issues and in order to develop evidence-basedrecommendations the Diabetes and Nutrition Study
group (DNSG) of the EASD (European Association forthe Study of Diabetes) commissioned a systematic re-
view and meta-analysis (SRMA) of prospective cohortstudies using the Grading of Recommendations
Assessment Development and Evaluation (GRADE)system to summarize the evidence of the association be-
tween the consumption of total nuts or specific types ofnuts and the incidence of and mortality from certain
CVD outcomes The shape of the associations with lin-ear and non-linear dose-response analysis was also
evaluated
METHODS
The current systematic review and meta-analysis fol-
lowed the methodological guidelines of the CochraneHandbook for Systematic Reviews of Interventions23
Results are reported according to Meta-analysis ofObservational Studies in Epidemiology (MOOSE)
guidelines24 The protocol is available at httpwwwcrdyorkacukPROSPERO (identifier PROSPERO
2018 CRD42018103360)
Search strategy
A systematic search limited to human studies with no lan-
guage restrictions was conducted of the MEDLINE(PubMed) and Cochrane Library databases through 5
June 2018 An updated search was then performed onMarch 19 2019 Table 1 shows the PICOS (participants
interventionsexposures comparators outcomes andstudy design) criteria used to identify studies eligible for
inclusion The electronic search was supplemented with amanual review of the reference lists of the retrieved
articles Figure 1 and Table S1 in the SupportingInformation online summarize the search and selection
process
Study selection
An initial screening of all titles and abstracts of the re-
trieved articles was performed to evaluate compliancewith the eligibility criteria Inclusion criteria were
prospective cohort studies with at least 1 year of follow-up conducted in an adult population with total nuts or
specific types of nuts as exposure with the incidence of
692 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
or mortality from CVD CHD stroke heart failure(HF) or atrial fibrillation (AF) as the outcome and
reporting effect estimators as odds ratios (ORs) riskratios (RRs) or hazard ratios (HRs) and their corre-
sponding 95 confidence intervals (CIs) When morethan 1 article from the same study was revealed both
papers were included if the end points were different(ie AF in one article and stroke in the other) When
multiple publications from the same study reported thesame outcome the study with the longest follow-up was
selected for inclusion Proceedings or publishedabstracts were not included in the present systematic re-
view and meta-analysisTwo publications were identified for the Nursesrsquo
Health Study I and Health Professionals Follow-up
Study which evaluated the association between nutconsumption and CVD outcomes in a whole population
cohort study21 and in individuals with diabetes only25
Therefore for the present meta-analysis the study con-
ducted in the whole cohort where participants with dia-betes where also included was selected Similarly for
The Netherlands Cohort Study two different articleswere also identified one published in 201526 and an-
other one in 201927 However in both cases the totalsample size and the period of follow-up were the same
Therefore the 2015 publication26 was included becauseits main aim fitted better with the objective of the pre-
sent systematic review and meta-analysis
Data extraction
Two independent reviewers (NB-T and IP-G) reviewed
the full text of the articles that were selected in the firstphase of the screening process Using a standardized pro-
forma the following relevant information was extractedfrom those studies that met all of the inclusion criteria
authors journal and year of publication study design co-hort name country or study location total sample size
characteristics of subjects follow-up duration sources offindings type of exposure and method used for its assess-
ment outcome and method assessment effect estimators
(OR RR or HR and 95 confidence intervals) and statis-tical analyses When necessary authors were contacted by
email to obtain additional information relevant to theanalyses Any disagreement was resolved by discussion or
if necessary by a third author (JS-S)
Quality of the included studies
The Newcastle-Ottawa scale (NOS) was utilized to as-
sess the quality of the included studies28 It is a ratingscale from 0 to 9 where points are allocated according
to 3 different domains population selection outcomeassessment and comparability of the groups A maxi-
mum of 4 3 and 2 points were allotted to each study af-ter evaluation of the aforementioned domains High-
quality studies were considered those studies with a to-tal score of at least 7 points Disagreements in grading
the quality of the studies were resolved through consen-sus between the reviewers
Outcomes
The primary outcomes were CVD incidence (including
only nonfatal or a combination of nonfatal and fataloutcomes of a composite of different CVD outcomes)and CVD mortality which only included a composite
of different fatal CVD end points Secondary outcomesincluded incidence of nonfatal or a combination of
nonfatal and fatal outcomes and mortality from fataloutcomes ie CHD stroke AF and HF
Studies that reported fatal CHD and nonfatal myo-cardial infarction separately were combined using a
fixed-effects model to generate an overall estimate forCHD incidence2930 In the same way following the
same procedure ischemic stroke and intracerebralhemorrhage outcomes22 and fatal ischemic stroke and
fatal hemorrhagic stroke31 end points were combinedto obtain an overall estimate for stroke incidence and
stroke mortality respectively
Table 1 PICOS criteria for inclusion and exclusion of studiesParameter Inclusion criteria Exclusion criteria
Participants General population of adults Aged lt18 yearsInterventionexposure Nut consumption (including total nuts or
subtypes of nuts eg walnuts almondspeanuts peanut butter hazelnuts)
Dietary intakes do not include total nut consumptionor different subtypes of nut consumption
Comparison Extreme quantiles Risk estimate on continuous scaleOutcome Incidence of or mortality from cardiovas-
cular disease coronary heart diseasestroke heart failure atrial fibrillation
Other cardiovascular disease outcomes
Study design Prospective cohort studies Cross-sectional case-control ecological retrospectiveobservational studies clinical trials and non-hu-man studies
Nutrition ReviewsVR Vol 77(10)691ndash709 693
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Statistical analyses
The generic inverse variance method with a random-
effects model (5 comparisons) or fixed-effects model(lt5 comparisons) was used to pool the natural log-
transformed RRs for CVD incidence and mortality out-comes to compare highest vs lowest categories of nut
consumption For one study32 that reported results us-ing the second category of nut consumption rather
than the lowest one as the reference the RR and its cor-responding 95CI were recalculated following the
Hamling et al33 method using the first category as thereference
Heterogeneity among studies was estimated usingCochranrsquos Q test and quantified by the I2 statistic
Statistical significance was set at Plt 010 and an I2
value 50 was considered to reflect substantial
heterogeneityMeta-regression analysis was performed in order
to assess whether a priori specified study characteris-tics (ie sex follow-up geographical area NOS scale
and its individual domains) may have affected theoverall effect estimates This subgroup analysis was
only conducted if at least 10 study comparisons wereavailable23
Sensitivity analysis excluding 1 study at a time and
recalculating the summary estimates was performed toascertain the influence of individual studies on the sum-
mary estimates If the removal of a study yielded achange in the level of significance magnitude (by
gt10) or direction of the pooled risk estimates orchanged the evidence of heterogeneity then it was con-
sidered as influentialLinear dose-response analysis for total nut con-
sumption and different CVD outcomes was conductedfollowing the 2-stage generalized least-squares trend
(GLST) estimation method developed by Greenlandand Longnecker34 and Orsini et al35 In the first stage
the method fits the dose-response model within eachstudy and in the second stage it combines study-
specific trends Data on RRs and the corresponding95CIs total number of participants cases and doses
for at least 3 categories of nut consumption wereneeded to carry out this method The mean or median
of nut consumption from each exposure category wasused if it was directly reported For those studies that
did not report this information the midpoint betweenthe upper and lower boundaries was assigned when
ranges of nut consumption were available For studiesthat reported open-ended extreme categories a width
equal to the adjacent category was assumed in order toestimate the upper or lower cutoff value Some studies
reported the information on nut consumption in gramsand others in servings Therefore servings were
converted to grams where 1 serving equated to 28 g
unless authors specified other serving sizesPotential nonlinear association between nut con-
sumption and CVD outcomes was assessed using re-stricted cubic splines (MKSPLINE procedure) which
were combined using multivariate meta-analysis Thedeparture from linearity was assessed by the Wald testconstraining the regression coefficient for the second
spline equal to zero36
Publication bias was tested by the visual inspection
of the funnel plots for asymmetry and statisticallyBeggrsquos test and Eggerrsquos test When few studies are in-
cluded in the analysis the power of the tests is too lowtherefore publication bias was only examined if more
than 10 study comparisons were included in the analy-sis23 Statistical significance was set at Plt 005
Data analysis was performed using ReviewManager (RevMan) software version 53 Copenhagen
The Nordic Cochrane Centre The CochraneCollaboration 2014 and STATA version 15 software
(StataCorp LP College Station Texas)
Grading the evidence
The GRADE system was used to rate the overall qualityand the strength of the evidence The quality of evi-
dence for each outcome was categorized as high mod-erate low or very low This system regards
observational studies as low-quality evidence37 Thelevel of evidence can be upgraded or downgraded
according to different specified criteria Determinantsto downgrade included study design and execution lim-
itations38 inconsistency39 indirectness40 impreci-sion41 and publication bias42 Determinants to upgradeincluded large magnitude effect dose-response gradi-
ent and attenuation by plausible confounding effects43
Discrepancies in ratings of the evidence quality were
resolved by consensus between NB-T and IP-G
RESULTS
Study selection process
The present systematic review and meta-analysis in-cluded 19 prospective studies from the 2992 identified
articles (Figure 1) Three study comparisons (1 report)were included in the meta-analysis for total CVD21 14
study comparisons (9 reports) for CVD mortal-ity21263144ndash49 7 study comparisons (5 reports) for
CHD2122293050 12 study comparisons (8 reports) forCHD mortality2122263031444851 7 study comparisons
(5 reports) for stroke2122325253 11 study comparisons(7 reports) for stroke mortality21263132474851 5 study
comparisons (4 reports) for hemorrhagic stroke2252ndash54
694 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
7 study comparisons (5 reports) for ischemic
stroke2122325253 2 study comparisons for AF2255 and2 study comparisons for HF2256
Four studies57ndash60 that reported the risk estimate oncontinuous scale instead of categories of nut consump-
tion were identified and therefore were not included inthe high vs low categories of consumption analyses and
the dose-response analyses
Study characteristics
The characteristics of the included studies are presentedin Table 221222629ndash324446ndash60 Publication date ranged
from 1992 to 2018 Six of the studies originated fromEurope 13 from America 1 from Asia 1 from
Australia 1 from both China and the USA and 1 fromboth China and Germany The duration of follow-up
ranged from 43 to 287 years All the studies assessed
nut intake via a food frequency questionnaire The vast
majority of studies (709) were of high quality accord-ing to the NOS scale Tables S1ndashS5 in the Supporting
Information online describe the characteristics of theincluded studies by type of nut consumption
High vs low categories of consumption analyses
Nuts and cardiovascular disease incidence Three cohort
comparisons involving 210 836 participants and 14 136cases analyzed the association between nut consump-
tion and the risk of CVD incidence The summary RR(95CI) for high vs low categories of nut consumption
was 085 (080ndash091) with no evidence of interstudy het-erogeneity (I2frac14 0 Pheterogeneity 081) (Figure 2 and
Figure S1 in the Supporting Information online)Regarding specific types of nuts consumption of tree
nuts ([RR 085 95CI 079ndash091] I2 0
Figure 1 Flow diagram of the literature search and selection process
Nutrition ReviewsVR Vol 77(10)691ndash709 695
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Char
acte
rist
ics
ofth
ein
clud
edst
udie
sev
alua
ting
the
asso
ciat
ion
betw
een
nuts
and
risk
ofca
rdio
vasc
ular
dise
ase
outc
omes
Stud
yCo
untr
ySt
udy
nam
ePo
pula
tion
Nut
cons
umpt
ion
asse
ssm
ent
met
hod
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Fras
eret
al(1
992)
29U
SASe
vent
h-da
yAd
vent
ists
2647
3M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
51
serv
ings
wk
vslt
1se
rvin
gw
k
25
6N
onfa
talM
I13
4Ag
ency
7CH
Dm
orta
lity
463
Albe
rtet
al(2
002)
30U
SAPh
ysic
ians
rsquoH
ealth
Stud
y21
454
MSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gm
o
40ndash8
417
CHD
mor
talit
y56
6Ag
ency
7N
onfa
talM
I10
37
Blom
hoff
etal
(200
6)44
USA
Iow
aW
omen
rsquosH
ealth
Stud
y31
778
FSe
lf-ad
min
is-
tere
dFF
QN
uts
plus
pean
utbu
tter
7se
rvin
gsw
kvs 0
5se
rvin
gsw
k
55ndash6
915
CVD
mor
talit
y16
75Ag
ency
6CH
Dm
orta
lity
948
Djo
uss e
etal
(200
8)56
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2097
6M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gw
k
40ndash8
419
6H
eart
failu
re10
93N
R6
Net
tleto
net
al(2
008)
57U
SAAt
hero
scle
rosi
sRi
skin
Com
mun
ities
Stud
y
1415
3M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
erCo
ntin
uous
1se
rvin
gd
45ndash6
413
Hea
rtfa
ilure
1140
Agen
cy8
Djo
uss e
etal
(201
0)52
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2107
8M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
7se
rvin
gsw
kvs lt
1se
rvin
gw
k
40ndash8
421
1St
roke
1424
Agen
cy6
Isch
emic
stro
ke11
89H
emor
rhag
icst
roke
219
Bern
stei
net
al(2
012)
54U
SAN
urse
srsquoH
ealth
Stud
yI
8001
0F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
034
serv
ings
dvs
0se
rvin
gsd
30ndash5
526
Hem
orrh
agic
stro
ke13
83Ag
ency
6
Hea
lthPr
ofes
sion
als
Follo
w-U
pSt
udy
4315
0M
060
serv
ings
dvs
0se
rvin
gsd
40ndash7
522
Hem
orrh
agic
stro
ke82
9
Khaw
aja
etal
(201
2)55
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2105
4M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
6
serv
ings
wk
vs0
serv
ings
wk
40ndash8
424
Atria
lfib
rilla
tion
3317
Agen
cy7
Yaem
siri
etal
(201
2)58
USA
Wom
enrsquos
Hea
lthIn
itiat
ive
Obs
erva
tiona
lSt
udy
8702
5F
Self-
adm
inis
-te
red
FFQ
Not
spec
ified
Cont
inuo
us1
serv
ing
d50
ndash79
76
Isch
emic
stro
ke10
49Ag
ency
8
von
Rues
ten
etal
(201
3)59
Ger
man
yEP
IC-P
otsd
amst
udy
2353
1M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Pean
uts
wal
nuts
br
azil
nuts
Cont
inuo
us1
serv
ing
d35
ndash65
8CV
D36
3Ag
ency
8
(con
tinue
d)
696 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Gua
sch-
Ferr
eet
al(2
013)
46Sp
ain
PRED
IMED
stud
y72
16M
FIn
terv
iew
-adm
in-
iste
red
vali-
date
dse
mi-
quan
titat
ive
FFQ
Alm
onds
pea
-nu
tsh
azel
-nu
tsp
ista
-ch
ios
pine
nuts
gt3
serv
ings
wk
vs neve
r
55ndash8
04
8CV
Dm
orta
lity
81Ag
ency
-In
dust
ry9
Bona
ccio
etal
(201
5)48
Italy
Mol
i-san
istu
dy10
509
F88
77M
Valid
ated
sel
f-ad
min
iste
red
FFQ
Wal
nuts
haz
el-
nuts
alm
onds
pe
anut
s
Inta
kevs no
inta
ke
gt35
43
CVD
mor
talit
y10
4Ag
ency
6CH
Dm
orta
lity
39St
roke
mor
talit
y19
diG
iuse
ppe
etal
(201
5)32
Ger
man
yEP
IC-P
otsd
amst
udy
2599
7M
FVa
lidat
eds
elf-
adm
inis
tere
dFF
Q
Pean
uts
wal
nuts
br
azil
nuts
142
gd
vs 0g
d
F49
2M
52
58
3St
roke
288
Agen
cy8
Isch
emic
stro
ke23
5St
roke
mor
talit
y36
Gop
inat
het
al(2
015)
51Au
stra
liaBl
ueM
ount
ains
Eye
Stud
y13
12F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
490
to10
0g
dvs 0
to0
50g
d
49
15CV
Dm
orta
lity
258
Agen
cy7
IHD
mor
talit
y18
8St
roke
mor
talit
y10
1
1581
MCV
Dm
orta
lity
288
IHD
mor
talit
y24
2St
roke
mor
talit
y75
Har
ing
etal
(201
4)50
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1206
6M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Not
spec
ified
06
serv
ings
dvs
01
serv
ings
d45
ndash64
22CH
D11
47Ag
ency
9
Har
ing
etal
(201
5)53
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1160
1M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
er1
serv
ings
dvs 0
serv
ings
d
45ndash6
422
7St
roke
699
Agen
cy8
Hem
orrh
agic
stro
ke11
4
Isch
emic
stro
ke59
8H
shie
het
al(2
015)
47U
SAPh
ysic
ians
rsquoH
ealth
Stud
y20
742
MU
nval
idat
eds
elf-
adm
inis
tere
dFF
Q
Not
spec
ified
5
serv
ings
wk
vs lt1
serv
ing
wk
40ndash8
49
6CV
Dm
orta
lity
760
Agen
cy6
Stro
kem
orta
lity
143
Luu
etal
(201
5)31
USA
Sout
hern
Com
mun
ityCo
hort
Stud
y
7176
4M
FSe
mi-q
uant
ita-
tive
FFQ
Tota
lnut
san
dpe
anut
butt
er
184
5g
dvs lt
095
gd
40ndash7
95
4CV
Dm
orta
lity
1857
Agen
cy8
IHD
mor
talit
y79
3Is
chem
icst
roke
mor
talit
y12
1
Hem
orrh
agic
stro
kem
orta
lity
96
(con
tinue
d)
Nutrition ReviewsVR Vol 77(10)691ndash709 697
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Chin
aSh
angh
aiW
omen
rsquosH
ealth
Stud
yan
dSh
angh
aiM
enrsquos
Hea
lthSt
udy
134
265
MF
Pean
ut
254
gd
vs lt0
14gr
amd
ay
40ndash7
0an
d40
ndash74
122
and
65
CVD
mor
talit
y25
87IH
Dm
orta
lity
631
Isch
emic
stro
kem
orta
lity
588
Hem
orrh
agic
stro
kem
orta
lity
597
van
den
Bran
dtan
dSc
hout
en(2
015)
26
Net
herla
nds
Net
herla
nds
Coho
rtSt
udy
3202
MF
(sub
coho
rt)
Self-
adm
inis
-te
red
valid
ated
FFQ
Tree
nuts
pe
anut
s19
6g
dvs 0
gd
55ndash6
99
CVD
mor
talit
y29
85N
R7
IHD
mor
talit
y14
88St
roke
mor
talit
y56
5
Wan
get
al(2
016)
60Ch
ina
Linx
ian
NIT
coho
rt24
45M
FFF
QPe
anut
sch
est-
nuts
wal
nuts
Cont
inuo
us3
serv
ings
mo
40ndash6
926
Hea
rtdi
seas
em
orta
lity
355
Agen
cy6
Ger
man
ySt
roke
mor
talit
y45
2
Esla
mpa
rast
etal
(201
7)49
Iran
Gol
esta
nCo
hort
Stud
y28
257
FVa
lidat
eds
elf-
adm
inis
tere
dse
miq
uant
ita-
tive
FFQ
Pean
uts
tree
nuts
3
serv
ings
wk
vs neve
r
40ndash8
77
CVD
mor
talit
y91
1Ag
ency
720
855
M11
05
Gua
sch-
Ferr
eet
al(2
017)
21U
SAN
urse
srsquoH
ealth
Stud
yI
7636
4F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Pean
uts
othe
rnu
tsa
ndw
al-
nuts
(ifav
aila
ble)
5
serv
ings
wk
vs neve
rora
lmos
tne
ver
30ndash5
528
7CV
D67
27Ag
ency
7CV
Dm
orta
lity
1770
CHD
3552
CHD
mor
talit
y99
6St
roke
3322
Stro
kem
orta
lity
773
Isch
emic
stro
ke16
35N
urse
srsquoH
ealth
Stud
yII
9294
6F
25ndash4
221
5CV
D19
15CV
Dm
orta
lity
82CH
D67
0CH
Dm
orta
lity
46St
roke
1262
Stro
kem
orta
lity
36
Isch
emic
stro
ke22
0H
ealth
Prof
essi
onal
sFo
llow
-Up
Stud
y
4152
6M
40ndash7
522
5CV
D54
94CV
Dm
orta
lity
2599
CHD
4168
CHD
mor
talit
y19
21St
roke
1326
Stro
kem
orta
lity
367
Isch
emic
stro
ke74
2
(con
tinue
d)
698 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Pheterogeneity 070) peanuts ([RR 087 95CI
081ndash093] I2 0 Pheterogeneity 067) and walnuts ([RR081 95CI 071ndash092] I2 73 Pheterogeneity 003) was
associated with a lower risk of CVD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S2ndashS4 in the SupportingInformation online) No association was reported be-tween peanut butter consumption and the risk of CVD
incidence ([RR 098 95CI 093ndash103] I2 89Pheterogeneitylt001) (Table 3 and Figure S5 in the
Supporting Information online)
Nuts and cardiovascular disease mortality Fifteen co-hort comparisons analyzed the association between
nut consumption and CVD mortality including413 727 participants and 14 475 cases The summary
RR (95CI) for high vs low categories of nut con-sumption was 077 (072ndash082) with no evidence of
interstudy heterogeneity (I2 3 Pheterogeneity 042)(Figure 2 and Figure S6 in the Supporting Information
online) Regarding specific types of nuts only peanutshave been studied in relation to CVD mortality show-
ing a summary RR of 077 (95CI 070ndash085) for highvs low categories of consumption with no evidence of
interstudy heterogeneity (I2 0 Pheterogeneity 081)(Table 3 and Figure S7 in the Supporting Information
online)
Nuts and coronary heart disease incidence Seven cohortcomparisons analyzed the association between nut con-
sumption and CHD incidence including 275 812 par-ticipants and 12 654 cases The summary RR (95CI)
for high vs low categories of nut consumption was 082(069ndash096) with evidence of substantial interstudy het-
erogeneity (I2 74 Pheterogeneitylt001) (Figure 2 andFigure S8 in the Supporting Information online)
Regarding specific types of nuts consumption of treenuts ([RR 077 95CI 070ndash084] I2 61
Pheterogeneity 008) peanuts ([RR 085 95CI 079ndash092] I2 0 Pheterogeneity 055) and walnuts ([RR0791 95CI 066ndash094] I2 69 Pheterogeneity 004) was
associated with a lower risk of CHD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S9ndashS11 in the SupportingInformation online) No association was reported be-
tween peanut butter consumption and the risk of CHDincidence ([RR 100 95CI 094ndash107] I2 43
Pheterogeneity 017) (Table 3 and Figure S12 in theSupporting Information online)
Nuts and coronary heart disease mortality Thirteen co-
hort comparisons analyzed the association between nutconsumption and CHD mortality including 396 041
participants and 7877 cases The summary RR (95CI)Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Lars
son
etal
(201
8)22
Swed
enCo
hort
ofSw
edis
hM
en32
911
MFF
QN
uts
(not
incl
ud-
ing
coco
nuto
rch
estn
uts)
3
serv
ings
wk
vs neve
r
45ndash8
317
MI
4983
Agen
cy8
MIm
orta
lity
917
Hea
rtfa
ilure
3160
Atria
lfib
rilla
tion
7550
Isch
emic
stro
ke37
82In
trac
ereb
ral
hem
orrh
age
543
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CVD
car
diov
ascu
lard
isea
seE
PIC
Euro
pean
Pros
pect
ive
Inve
stig
atio
nin
toCa
ncer
and
Nut
ritio
nF
fem
ales
FFQ
foo
dfr
eque
ncy
ques
tionn
aire
IH
Di
sche
mic
hear
tdis
ease
Mm
ales
MI
myo
card
iali
nfar
ctio
nN
ITN
utrit
ion
Inte
rven
tion
Tria
lsN
OS
New
cast
le-O
ttaw
asc
ale
NR
Not
repo
rted
PRE
DIM
EDP
reve
nci o
nco
nD
ieta
MED
iterr
anea
Nutrition ReviewsVR Vol 77(10)691ndash709 699
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
for high vs low categories of nut consumption was 076(067ndash086) with evidence of moderate interstudy het-
erogeneity (I2 46 Pheterogeneity 004) (Figure 2 andFigure S13 in the Supporting Information online)
Regarding specific types of nuts peanut consumption
was inversely associated with the risk of CHD mortalityafter comparing high vs low categories of consumption
([RR 075 95CI 064ndash088] I2 0 Pheterogeneity 046)(Table 3 and Figure S14 in the Supporting Information
online)
Table 3 Summary RR of cardiovascular disease outcomes by different types of nut consumption (comparing highest vslowest categories)
No of cohorts No of participants No of cases RR (95CI) P for heterogeneity I2 ()
Tree nutsCVD 3 210 836 14 136 085 (079 091) 070 0CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 100 (089 111) 093 0Stroke mortality 3 118 962 1851 093 (077 113) 044 0CHD 3 210 836 8390 077 (070 084) 008 61CHD mortality ndash ndash ndash ndash ndash ndash
PeanutsCVD 3 210 836 14 136 087 (081 093) 067 0CVD mortality 2 134 265 5572 077 (070 085) 081 0Stroke 3 210 836 5910 090 (081 099) 032 13Stroke mortality 4 253 227 3036 083 (073 095) 007 57CHD 3 210 836 8390 085 (079 092) 055 0CHD mortality 2 134 265 2119 075 (064 088) 046 0
WalnutsCVD 3 144 021 5255 081 (071 092) 003 73CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 144 021 5910 085 (071 102) 019 39Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 144 021 2685 079 (066 094) 004 69CHD mortality ndash ndash ndash ndash ndash ndash
Peanut butterCVD 3 210 836 14 136 098 (093 103) lt001 89CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 094 (087 102) lt001 86Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 210 836 8390 100 (094 107) 017 43CHD mortality ndash ndash ndash ndash ndash ndash
Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovascular disease RR risk ratio
Figure 2 Summary plots of effect estimates from prospective cohort studies evaluating the association between nut consumptionand the risk of different cardiovascular outcomes Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovasculardisease RR risk ratio
700 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Nuts and stroke incidence Seven cohort comparisonsanalyzed the association between nut consumption and
stroke incidence including 302 888 participants and12 646 cases The summary RR (95CI) for high vs low
categories of nut consumption was 100 (092ndash109)with no evidence of interstudy heterogeneity (I2 0
Pheterogeneity 097) (Figure 2 and Figure S15 in theSupporting Information online) Regarding specific
types of nuts peanut consumption was associated witha lower risk of stroke incidence after comparing highest
vs lowest categories of consumption ([RR 090 95CI081ndash099] I2 13 Pheterogeneity 032) (Table 3 andFigure S16 in the Supporting Information online) No
association was observed between tree nut walnut andpeanut butter consumption and the risk of stroke inci-
dence (Table 3 and Figures S17ndashS19 in the SupportingInformation online)
Nuts and stroke mortality Twelve cohort comparisons
analyzed the association between nut consumption andstroke mortality including 351 618 participants and
2332 cases The summary RR (95CI) for high vs lowcategories of nut consumption was 083 (075ndash093)
with no evidence of interstudy heterogeneity (I2 0Pheterogeneity 045) (Figure 2 and Figure S20 in the
Supporting Information online) Regarding specifictypes of nuts peanut consumption was associated with
a lower risk of stroke mortality after comparing highestvs lowest categories of consumption ([RR 085 95CI
079ndash092] I2 0 Pheterogeneity 055) (Table 3 andFigure S21 in the Supporting Information online) No
association was reported between tree nut consumptionand the risk of stroke death (Table 3 and Figure S22 in
the Supporting Information online)
Nuts and hemorrhagic stroke Five cohort comparisonsanalyzed the association between nut consumption and
hemorrhagic stroke incidence involving 188 750 partic-ipants and 3088 cases The summary RR (95CI) for
high vs low categories of nut consumption was 102(077ndash134) with no evidence of interstudy heterogene-ity (I2 15 Pheterogeneity 032) (Figure 2 and Figure S23
in the Supporting Information online) No studyanalyzed the association between different types of nuts
and the risk of hemorrhagic stroke
Nuts and ischemic stroke Seven cohort comparisonsanalyzed the association between nut consumption and
ischemic stroke incidence involving 302 423 partici-pants and 8401 cases The summary RR (95CI) for
high vs low categories of nut consumption was 099(089ndash110) with no evidence of interstudy heterogene-
ity (I2 0 Pheterogeneity 062) (Figure 2 and Figure S24 in
the Supporting Information online) No study analyzed
the association between different types of nuts and therisk of hemorrhagic stroke
Nuts and atrial fibrillation Two cohort comparisonsanalyzed the association between nut consumption and
AF involving 53 965 participants and 10 867 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 085 (073ndash099) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 052)
(Figure 2 and Figure S25 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of AF
Nuts and heart failure Two cohort comparisonsanalyzed the association between nut consumption and
HF involving 53 887 participants and 4253 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 100 (086ndash116) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 085)
(Figure 2 and Figure S26 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of HF
Dose-response analyses
Figures S27ndashS34 in the Supporting Information onlineshow the linear and non-linear dose-response analyses
between total nut consumption and CVD outcomesThe summary RR (95CI) for a 28-gd increment was
087 (081ndash093) for CVD incidence 071 (061ndash084)for CVD mortality 075 (064ndash088) for CHD incidence
067 (052ndash087) for CHD mortality 106 (097ndash115) forstroke incidence 101 (088ndash118) for stroke mortality
105 (077ndash143) for hemorrhagic stroke and 106(086ndash131) for ischemic stroke
Total nut consumption and the risk of CVD inci-dence (Figure S27 in the Supporting Information on-
line) CVD mortality (Figure S28 in the SupportingInformation online) stroke mortality (Figure S32 in theSupporting Information online) and hemorrhagic
stroke (Figure S33 in the Supporting Information on-line) showed a non-linear association (Pnon-line-
aritylt001) The reduction in the risk of CVD incidencewas observed up to a consumption of 10 gd with no
further reduction with higher consumptions (Table S6in the Supporting Information online) For CVD mor-
tality and CHD mortality there was a steeper reductionin the risk at approximately 15ndash20 gd with no further
reduction with a higher consumption (Table S6 in theSupporting Information online) The reduction in the
risk of stroke mortality was observed up to a consump-tion of 5 gd with no significant reductions above this
amount (Table S6 in the Supporting Information
Nutrition ReviewsVR Vol 77(10)691ndash709 701
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
heart failure The certainty of the evidence ranged from moderate to very lowConclusions This systematic review and meta-analysis revealed a beneficial roleof nut consumption in reducing the incidence of and mortality from different CVDoutcomes
INTRODUCTION
Cardiovascular disease (CVD) is an important publichealth issue According to the World HealthOrganization (WHO) it is the leading cause of death
worldwide affecting approximately 177 million peoplein 20151 Importantly CVDs are susceptible to behavior
modifications In this sense a healthy diet is one of thelifestyle components that could be promoted to help ad-
dress this global health concern2 Different healthy die-tary patterns such as the Mediterranean diet
(MedDiet)3 the DASH diet4 or a vegetarian diet5 havenuts as a key food component Nuts despite their high
fat content (mainly unsaturated fatty acids)6 are alsorich in minerals vitamins fiber and bioactive com-
pounds7 Given this exceptional nutritional profile fre-quent nut consumption has been inversely associated
with a lower risk of CVD in large prospective cohortstudies which have been summarized in several meta-
analyses8ndash20 Nonetheless some of the previous meta-analyses had methodological limitations such as the in-
clusion of studies with nuts plus seeds or legumes as ex-posure the inclusion of studies combining different
outcomes across analyses (eg inclusion of studies withonly fatal CHD [coronary heart disease] outcome in the
CVD mortality analysis) and the inclusion of studieswithout the first category of exposure as reference
Moreover since publication of the last meta-analysesthe results of two new large prospective cohort studies
evaluating the association between nut consumptionand CVD outcomes have been published2122 One study
reported updated results from the Nursesrsquo Health StudyI (NHSI) Nursesrsquo Health Study II (NSHII) and Health
Professionals Follow-up Study (HPFS) comprising upto 32 years of follow-up and a large number of cases21
Another reported the association between nut con-sumption and the incidence of 7 CVD outcomes in apopulation of 32 911 males22 Importantly most of the
previous meta-analyses have focused on total nut in-take and only a few have taken into account the poten-
tial associations between specific types of nutconsumption and the risk of CVD outcomes which
may vary considerablyTherefore taking into consideration the aforemen-
tioned issues and in order to develop evidence-basedrecommendations the Diabetes and Nutrition Study
group (DNSG) of the EASD (European Association forthe Study of Diabetes) commissioned a systematic re-
view and meta-analysis (SRMA) of prospective cohortstudies using the Grading of Recommendations
Assessment Development and Evaluation (GRADE)system to summarize the evidence of the association be-
tween the consumption of total nuts or specific types ofnuts and the incidence of and mortality from certain
CVD outcomes The shape of the associations with lin-ear and non-linear dose-response analysis was also
evaluated
METHODS
The current systematic review and meta-analysis fol-
lowed the methodological guidelines of the CochraneHandbook for Systematic Reviews of Interventions23
Results are reported according to Meta-analysis ofObservational Studies in Epidemiology (MOOSE)
guidelines24 The protocol is available at httpwwwcrdyorkacukPROSPERO (identifier PROSPERO
2018 CRD42018103360)
Search strategy
A systematic search limited to human studies with no lan-
guage restrictions was conducted of the MEDLINE(PubMed) and Cochrane Library databases through 5
June 2018 An updated search was then performed onMarch 19 2019 Table 1 shows the PICOS (participants
interventionsexposures comparators outcomes andstudy design) criteria used to identify studies eligible for
inclusion The electronic search was supplemented with amanual review of the reference lists of the retrieved
articles Figure 1 and Table S1 in the SupportingInformation online summarize the search and selection
process
Study selection
An initial screening of all titles and abstracts of the re-
trieved articles was performed to evaluate compliancewith the eligibility criteria Inclusion criteria were
prospective cohort studies with at least 1 year of follow-up conducted in an adult population with total nuts or
specific types of nuts as exposure with the incidence of
692 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
or mortality from CVD CHD stroke heart failure(HF) or atrial fibrillation (AF) as the outcome and
reporting effect estimators as odds ratios (ORs) riskratios (RRs) or hazard ratios (HRs) and their corre-
sponding 95 confidence intervals (CIs) When morethan 1 article from the same study was revealed both
papers were included if the end points were different(ie AF in one article and stroke in the other) When
multiple publications from the same study reported thesame outcome the study with the longest follow-up was
selected for inclusion Proceedings or publishedabstracts were not included in the present systematic re-
view and meta-analysisTwo publications were identified for the Nursesrsquo
Health Study I and Health Professionals Follow-up
Study which evaluated the association between nutconsumption and CVD outcomes in a whole population
cohort study21 and in individuals with diabetes only25
Therefore for the present meta-analysis the study con-
ducted in the whole cohort where participants with dia-betes where also included was selected Similarly for
The Netherlands Cohort Study two different articleswere also identified one published in 201526 and an-
other one in 201927 However in both cases the totalsample size and the period of follow-up were the same
Therefore the 2015 publication26 was included becauseits main aim fitted better with the objective of the pre-
sent systematic review and meta-analysis
Data extraction
Two independent reviewers (NB-T and IP-G) reviewed
the full text of the articles that were selected in the firstphase of the screening process Using a standardized pro-
forma the following relevant information was extractedfrom those studies that met all of the inclusion criteria
authors journal and year of publication study design co-hort name country or study location total sample size
characteristics of subjects follow-up duration sources offindings type of exposure and method used for its assess-
ment outcome and method assessment effect estimators
(OR RR or HR and 95 confidence intervals) and statis-tical analyses When necessary authors were contacted by
email to obtain additional information relevant to theanalyses Any disagreement was resolved by discussion or
if necessary by a third author (JS-S)
Quality of the included studies
The Newcastle-Ottawa scale (NOS) was utilized to as-
sess the quality of the included studies28 It is a ratingscale from 0 to 9 where points are allocated according
to 3 different domains population selection outcomeassessment and comparability of the groups A maxi-
mum of 4 3 and 2 points were allotted to each study af-ter evaluation of the aforementioned domains High-
quality studies were considered those studies with a to-tal score of at least 7 points Disagreements in grading
the quality of the studies were resolved through consen-sus between the reviewers
Outcomes
The primary outcomes were CVD incidence (including
only nonfatal or a combination of nonfatal and fataloutcomes of a composite of different CVD outcomes)and CVD mortality which only included a composite
of different fatal CVD end points Secondary outcomesincluded incidence of nonfatal or a combination of
nonfatal and fatal outcomes and mortality from fataloutcomes ie CHD stroke AF and HF
Studies that reported fatal CHD and nonfatal myo-cardial infarction separately were combined using a
fixed-effects model to generate an overall estimate forCHD incidence2930 In the same way following the
same procedure ischemic stroke and intracerebralhemorrhage outcomes22 and fatal ischemic stroke and
fatal hemorrhagic stroke31 end points were combinedto obtain an overall estimate for stroke incidence and
stroke mortality respectively
Table 1 PICOS criteria for inclusion and exclusion of studiesParameter Inclusion criteria Exclusion criteria
Participants General population of adults Aged lt18 yearsInterventionexposure Nut consumption (including total nuts or
subtypes of nuts eg walnuts almondspeanuts peanut butter hazelnuts)
Dietary intakes do not include total nut consumptionor different subtypes of nut consumption
Comparison Extreme quantiles Risk estimate on continuous scaleOutcome Incidence of or mortality from cardiovas-
cular disease coronary heart diseasestroke heart failure atrial fibrillation
Other cardiovascular disease outcomes
Study design Prospective cohort studies Cross-sectional case-control ecological retrospectiveobservational studies clinical trials and non-hu-man studies
Nutrition ReviewsVR Vol 77(10)691ndash709 693
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Statistical analyses
The generic inverse variance method with a random-
effects model (5 comparisons) or fixed-effects model(lt5 comparisons) was used to pool the natural log-
transformed RRs for CVD incidence and mortality out-comes to compare highest vs lowest categories of nut
consumption For one study32 that reported results us-ing the second category of nut consumption rather
than the lowest one as the reference the RR and its cor-responding 95CI were recalculated following the
Hamling et al33 method using the first category as thereference
Heterogeneity among studies was estimated usingCochranrsquos Q test and quantified by the I2 statistic
Statistical significance was set at Plt 010 and an I2
value 50 was considered to reflect substantial
heterogeneityMeta-regression analysis was performed in order
to assess whether a priori specified study characteris-tics (ie sex follow-up geographical area NOS scale
and its individual domains) may have affected theoverall effect estimates This subgroup analysis was
only conducted if at least 10 study comparisons wereavailable23
Sensitivity analysis excluding 1 study at a time and
recalculating the summary estimates was performed toascertain the influence of individual studies on the sum-
mary estimates If the removal of a study yielded achange in the level of significance magnitude (by
gt10) or direction of the pooled risk estimates orchanged the evidence of heterogeneity then it was con-
sidered as influentialLinear dose-response analysis for total nut con-
sumption and different CVD outcomes was conductedfollowing the 2-stage generalized least-squares trend
(GLST) estimation method developed by Greenlandand Longnecker34 and Orsini et al35 In the first stage
the method fits the dose-response model within eachstudy and in the second stage it combines study-
specific trends Data on RRs and the corresponding95CIs total number of participants cases and doses
for at least 3 categories of nut consumption wereneeded to carry out this method The mean or median
of nut consumption from each exposure category wasused if it was directly reported For those studies that
did not report this information the midpoint betweenthe upper and lower boundaries was assigned when
ranges of nut consumption were available For studiesthat reported open-ended extreme categories a width
equal to the adjacent category was assumed in order toestimate the upper or lower cutoff value Some studies
reported the information on nut consumption in gramsand others in servings Therefore servings were
converted to grams where 1 serving equated to 28 g
unless authors specified other serving sizesPotential nonlinear association between nut con-
sumption and CVD outcomes was assessed using re-stricted cubic splines (MKSPLINE procedure) which
were combined using multivariate meta-analysis Thedeparture from linearity was assessed by the Wald testconstraining the regression coefficient for the second
spline equal to zero36
Publication bias was tested by the visual inspection
of the funnel plots for asymmetry and statisticallyBeggrsquos test and Eggerrsquos test When few studies are in-
cluded in the analysis the power of the tests is too lowtherefore publication bias was only examined if more
than 10 study comparisons were included in the analy-sis23 Statistical significance was set at Plt 005
Data analysis was performed using ReviewManager (RevMan) software version 53 Copenhagen
The Nordic Cochrane Centre The CochraneCollaboration 2014 and STATA version 15 software
(StataCorp LP College Station Texas)
Grading the evidence
The GRADE system was used to rate the overall qualityand the strength of the evidence The quality of evi-
dence for each outcome was categorized as high mod-erate low or very low This system regards
observational studies as low-quality evidence37 Thelevel of evidence can be upgraded or downgraded
according to different specified criteria Determinantsto downgrade included study design and execution lim-
itations38 inconsistency39 indirectness40 impreci-sion41 and publication bias42 Determinants to upgradeincluded large magnitude effect dose-response gradi-
ent and attenuation by plausible confounding effects43
Discrepancies in ratings of the evidence quality were
resolved by consensus between NB-T and IP-G
RESULTS
Study selection process
The present systematic review and meta-analysis in-cluded 19 prospective studies from the 2992 identified
articles (Figure 1) Three study comparisons (1 report)were included in the meta-analysis for total CVD21 14
study comparisons (9 reports) for CVD mortal-ity21263144ndash49 7 study comparisons (5 reports) for
CHD2122293050 12 study comparisons (8 reports) forCHD mortality2122263031444851 7 study comparisons
(5 reports) for stroke2122325253 11 study comparisons(7 reports) for stroke mortality21263132474851 5 study
comparisons (4 reports) for hemorrhagic stroke2252ndash54
694 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
7 study comparisons (5 reports) for ischemic
stroke2122325253 2 study comparisons for AF2255 and2 study comparisons for HF2256
Four studies57ndash60 that reported the risk estimate oncontinuous scale instead of categories of nut consump-
tion were identified and therefore were not included inthe high vs low categories of consumption analyses and
the dose-response analyses
Study characteristics
The characteristics of the included studies are presentedin Table 221222629ndash324446ndash60 Publication date ranged
from 1992 to 2018 Six of the studies originated fromEurope 13 from America 1 from Asia 1 from
Australia 1 from both China and the USA and 1 fromboth China and Germany The duration of follow-up
ranged from 43 to 287 years All the studies assessed
nut intake via a food frequency questionnaire The vast
majority of studies (709) were of high quality accord-ing to the NOS scale Tables S1ndashS5 in the Supporting
Information online describe the characteristics of theincluded studies by type of nut consumption
High vs low categories of consumption analyses
Nuts and cardiovascular disease incidence Three cohort
comparisons involving 210 836 participants and 14 136cases analyzed the association between nut consump-
tion and the risk of CVD incidence The summary RR(95CI) for high vs low categories of nut consumption
was 085 (080ndash091) with no evidence of interstudy het-erogeneity (I2frac14 0 Pheterogeneity 081) (Figure 2 and
Figure S1 in the Supporting Information online)Regarding specific types of nuts consumption of tree
nuts ([RR 085 95CI 079ndash091] I2 0
Figure 1 Flow diagram of the literature search and selection process
Nutrition ReviewsVR Vol 77(10)691ndash709 695
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Char
acte
rist
ics
ofth
ein
clud
edst
udie
sev
alua
ting
the
asso
ciat
ion
betw
een
nuts
and
risk
ofca
rdio
vasc
ular
dise
ase
outc
omes
Stud
yCo
untr
ySt
udy
nam
ePo
pula
tion
Nut
cons
umpt
ion
asse
ssm
ent
met
hod
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Fras
eret
al(1
992)
29U
SASe
vent
h-da
yAd
vent
ists
2647
3M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
51
serv
ings
wk
vslt
1se
rvin
gw
k
25
6N
onfa
talM
I13
4Ag
ency
7CH
Dm
orta
lity
463
Albe
rtet
al(2
002)
30U
SAPh
ysic
ians
rsquoH
ealth
Stud
y21
454
MSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gm
o
40ndash8
417
CHD
mor
talit
y56
6Ag
ency
7N
onfa
talM
I10
37
Blom
hoff
etal
(200
6)44
USA
Iow
aW
omen
rsquosH
ealth
Stud
y31
778
FSe
lf-ad
min
is-
tere
dFF
QN
uts
plus
pean
utbu
tter
7se
rvin
gsw
kvs 0
5se
rvin
gsw
k
55ndash6
915
CVD
mor
talit
y16
75Ag
ency
6CH
Dm
orta
lity
948
Djo
uss e
etal
(200
8)56
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2097
6M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gw
k
40ndash8
419
6H
eart
failu
re10
93N
R6
Net
tleto
net
al(2
008)
57U
SAAt
hero
scle
rosi
sRi
skin
Com
mun
ities
Stud
y
1415
3M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
erCo
ntin
uous
1se
rvin
gd
45ndash6
413
Hea
rtfa
ilure
1140
Agen
cy8
Djo
uss e
etal
(201
0)52
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2107
8M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
7se
rvin
gsw
kvs lt
1se
rvin
gw
k
40ndash8
421
1St
roke
1424
Agen
cy6
Isch
emic
stro
ke11
89H
emor
rhag
icst
roke
219
Bern
stei
net
al(2
012)
54U
SAN
urse
srsquoH
ealth
Stud
yI
8001
0F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
034
serv
ings
dvs
0se
rvin
gsd
30ndash5
526
Hem
orrh
agic
stro
ke13
83Ag
ency
6
Hea
lthPr
ofes
sion
als
Follo
w-U
pSt
udy
4315
0M
060
serv
ings
dvs
0se
rvin
gsd
40ndash7
522
Hem
orrh
agic
stro
ke82
9
Khaw
aja
etal
(201
2)55
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2105
4M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
6
serv
ings
wk
vs0
serv
ings
wk
40ndash8
424
Atria
lfib
rilla
tion
3317
Agen
cy7
Yaem
siri
etal
(201
2)58
USA
Wom
enrsquos
Hea
lthIn
itiat
ive
Obs
erva
tiona
lSt
udy
8702
5F
Self-
adm
inis
-te
red
FFQ
Not
spec
ified
Cont
inuo
us1
serv
ing
d50
ndash79
76
Isch
emic
stro
ke10
49Ag
ency
8
von
Rues
ten
etal
(201
3)59
Ger
man
yEP
IC-P
otsd
amst
udy
2353
1M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Pean
uts
wal
nuts
br
azil
nuts
Cont
inuo
us1
serv
ing
d35
ndash65
8CV
D36
3Ag
ency
8
(con
tinue
d)
696 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Gua
sch-
Ferr
eet
al(2
013)
46Sp
ain
PRED
IMED
stud
y72
16M
FIn
terv
iew
-adm
in-
iste
red
vali-
date
dse
mi-
quan
titat
ive
FFQ
Alm
onds
pea
-nu
tsh
azel
-nu
tsp
ista
-ch
ios
pine
nuts
gt3
serv
ings
wk
vs neve
r
55ndash8
04
8CV
Dm
orta
lity
81Ag
ency
-In
dust
ry9
Bona
ccio
etal
(201
5)48
Italy
Mol
i-san
istu
dy10
509
F88
77M
Valid
ated
sel
f-ad
min
iste
red
FFQ
Wal
nuts
haz
el-
nuts
alm
onds
pe
anut
s
Inta
kevs no
inta
ke
gt35
43
CVD
mor
talit
y10
4Ag
ency
6CH
Dm
orta
lity
39St
roke
mor
talit
y19
diG
iuse
ppe
etal
(201
5)32
Ger
man
yEP
IC-P
otsd
amst
udy
2599
7M
FVa
lidat
eds
elf-
adm
inis
tere
dFF
Q
Pean
uts
wal
nuts
br
azil
nuts
142
gd
vs 0g
d
F49
2M
52
58
3St
roke
288
Agen
cy8
Isch
emic
stro
ke23
5St
roke
mor
talit
y36
Gop
inat
het
al(2
015)
51Au
stra
liaBl
ueM
ount
ains
Eye
Stud
y13
12F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
490
to10
0g
dvs 0
to0
50g
d
49
15CV
Dm
orta
lity
258
Agen
cy7
IHD
mor
talit
y18
8St
roke
mor
talit
y10
1
1581
MCV
Dm
orta
lity
288
IHD
mor
talit
y24
2St
roke
mor
talit
y75
Har
ing
etal
(201
4)50
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1206
6M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Not
spec
ified
06
serv
ings
dvs
01
serv
ings
d45
ndash64
22CH
D11
47Ag
ency
9
Har
ing
etal
(201
5)53
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1160
1M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
er1
serv
ings
dvs 0
serv
ings
d
45ndash6
422
7St
roke
699
Agen
cy8
Hem
orrh
agic
stro
ke11
4
Isch
emic
stro
ke59
8H
shie
het
al(2
015)
47U
SAPh
ysic
ians
rsquoH
ealth
Stud
y20
742
MU
nval
idat
eds
elf-
adm
inis
tere
dFF
Q
Not
spec
ified
5
serv
ings
wk
vs lt1
serv
ing
wk
40ndash8
49
6CV
Dm
orta
lity
760
Agen
cy6
Stro
kem
orta
lity
143
Luu
etal
(201
5)31
USA
Sout
hern
Com
mun
ityCo
hort
Stud
y
7176
4M
FSe
mi-q
uant
ita-
tive
FFQ
Tota
lnut
san
dpe
anut
butt
er
184
5g
dvs lt
095
gd
40ndash7
95
4CV
Dm
orta
lity
1857
Agen
cy8
IHD
mor
talit
y79
3Is
chem
icst
roke
mor
talit
y12
1
Hem
orrh
agic
stro
kem
orta
lity
96
(con
tinue
d)
Nutrition ReviewsVR Vol 77(10)691ndash709 697
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Chin
aSh
angh
aiW
omen
rsquosH
ealth
Stud
yan
dSh
angh
aiM
enrsquos
Hea
lthSt
udy
134
265
MF
Pean
ut
254
gd
vs lt0
14gr
amd
ay
40ndash7
0an
d40
ndash74
122
and
65
CVD
mor
talit
y25
87IH
Dm
orta
lity
631
Isch
emic
stro
kem
orta
lity
588
Hem
orrh
agic
stro
kem
orta
lity
597
van
den
Bran
dtan
dSc
hout
en(2
015)
26
Net
herla
nds
Net
herla
nds
Coho
rtSt
udy
3202
MF
(sub
coho
rt)
Self-
adm
inis
-te
red
valid
ated
FFQ
Tree
nuts
pe
anut
s19
6g
dvs 0
gd
55ndash6
99
CVD
mor
talit
y29
85N
R7
IHD
mor
talit
y14
88St
roke
mor
talit
y56
5
Wan
get
al(2
016)
60Ch
ina
Linx
ian
NIT
coho
rt24
45M
FFF
QPe
anut
sch
est-
nuts
wal
nuts
Cont
inuo
us3
serv
ings
mo
40ndash6
926
Hea
rtdi
seas
em
orta
lity
355
Agen
cy6
Ger
man
ySt
roke
mor
talit
y45
2
Esla
mpa
rast
etal
(201
7)49
Iran
Gol
esta
nCo
hort
Stud
y28
257
FVa
lidat
eds
elf-
adm
inis
tere
dse
miq
uant
ita-
tive
FFQ
Pean
uts
tree
nuts
3
serv
ings
wk
vs neve
r
40ndash8
77
CVD
mor
talit
y91
1Ag
ency
720
855
M11
05
Gua
sch-
Ferr
eet
al(2
017)
21U
SAN
urse
srsquoH
ealth
Stud
yI
7636
4F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Pean
uts
othe
rnu
tsa
ndw
al-
nuts
(ifav
aila
ble)
5
serv
ings
wk
vs neve
rora
lmos
tne
ver
30ndash5
528
7CV
D67
27Ag
ency
7CV
Dm
orta
lity
1770
CHD
3552
CHD
mor
talit
y99
6St
roke
3322
Stro
kem
orta
lity
773
Isch
emic
stro
ke16
35N
urse
srsquoH
ealth
Stud
yII
9294
6F
25ndash4
221
5CV
D19
15CV
Dm
orta
lity
82CH
D67
0CH
Dm
orta
lity
46St
roke
1262
Stro
kem
orta
lity
36
Isch
emic
stro
ke22
0H
ealth
Prof
essi
onal
sFo
llow
-Up
Stud
y
4152
6M
40ndash7
522
5CV
D54
94CV
Dm
orta
lity
2599
CHD
4168
CHD
mor
talit
y19
21St
roke
1326
Stro
kem
orta
lity
367
Isch
emic
stro
ke74
2
(con
tinue
d)
698 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Pheterogeneity 070) peanuts ([RR 087 95CI
081ndash093] I2 0 Pheterogeneity 067) and walnuts ([RR081 95CI 071ndash092] I2 73 Pheterogeneity 003) was
associated with a lower risk of CVD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S2ndashS4 in the SupportingInformation online) No association was reported be-tween peanut butter consumption and the risk of CVD
incidence ([RR 098 95CI 093ndash103] I2 89Pheterogeneitylt001) (Table 3 and Figure S5 in the
Supporting Information online)
Nuts and cardiovascular disease mortality Fifteen co-hort comparisons analyzed the association between
nut consumption and CVD mortality including413 727 participants and 14 475 cases The summary
RR (95CI) for high vs low categories of nut con-sumption was 077 (072ndash082) with no evidence of
interstudy heterogeneity (I2 3 Pheterogeneity 042)(Figure 2 and Figure S6 in the Supporting Information
online) Regarding specific types of nuts only peanutshave been studied in relation to CVD mortality show-
ing a summary RR of 077 (95CI 070ndash085) for highvs low categories of consumption with no evidence of
interstudy heterogeneity (I2 0 Pheterogeneity 081)(Table 3 and Figure S7 in the Supporting Information
online)
Nuts and coronary heart disease incidence Seven cohortcomparisons analyzed the association between nut con-
sumption and CHD incidence including 275 812 par-ticipants and 12 654 cases The summary RR (95CI)
for high vs low categories of nut consumption was 082(069ndash096) with evidence of substantial interstudy het-
erogeneity (I2 74 Pheterogeneitylt001) (Figure 2 andFigure S8 in the Supporting Information online)
Regarding specific types of nuts consumption of treenuts ([RR 077 95CI 070ndash084] I2 61
Pheterogeneity 008) peanuts ([RR 085 95CI 079ndash092] I2 0 Pheterogeneity 055) and walnuts ([RR0791 95CI 066ndash094] I2 69 Pheterogeneity 004) was
associated with a lower risk of CHD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S9ndashS11 in the SupportingInformation online) No association was reported be-
tween peanut butter consumption and the risk of CHDincidence ([RR 100 95CI 094ndash107] I2 43
Pheterogeneity 017) (Table 3 and Figure S12 in theSupporting Information online)
Nuts and coronary heart disease mortality Thirteen co-
hort comparisons analyzed the association between nutconsumption and CHD mortality including 396 041
participants and 7877 cases The summary RR (95CI)Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Lars
son
etal
(201
8)22
Swed
enCo
hort
ofSw
edis
hM
en32
911
MFF
QN
uts
(not
incl
ud-
ing
coco
nuto
rch
estn
uts)
3
serv
ings
wk
vs neve
r
45ndash8
317
MI
4983
Agen
cy8
MIm
orta
lity
917
Hea
rtfa
ilure
3160
Atria
lfib
rilla
tion
7550
Isch
emic
stro
ke37
82In
trac
ereb
ral
hem
orrh
age
543
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CVD
car
diov
ascu
lard
isea
seE
PIC
Euro
pean
Pros
pect
ive
Inve
stig
atio
nin
toCa
ncer
and
Nut
ritio
nF
fem
ales
FFQ
foo
dfr
eque
ncy
ques
tionn
aire
IH
Di
sche
mic
hear
tdis
ease
Mm
ales
MI
myo
card
iali
nfar
ctio
nN
ITN
utrit
ion
Inte
rven
tion
Tria
lsN
OS
New
cast
le-O
ttaw
asc
ale
NR
Not
repo
rted
PRE
DIM
EDP
reve
nci o
nco
nD
ieta
MED
iterr
anea
Nutrition ReviewsVR Vol 77(10)691ndash709 699
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
for high vs low categories of nut consumption was 076(067ndash086) with evidence of moderate interstudy het-
erogeneity (I2 46 Pheterogeneity 004) (Figure 2 andFigure S13 in the Supporting Information online)
Regarding specific types of nuts peanut consumption
was inversely associated with the risk of CHD mortalityafter comparing high vs low categories of consumption
([RR 075 95CI 064ndash088] I2 0 Pheterogeneity 046)(Table 3 and Figure S14 in the Supporting Information
online)
Table 3 Summary RR of cardiovascular disease outcomes by different types of nut consumption (comparing highest vslowest categories)
No of cohorts No of participants No of cases RR (95CI) P for heterogeneity I2 ()
Tree nutsCVD 3 210 836 14 136 085 (079 091) 070 0CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 100 (089 111) 093 0Stroke mortality 3 118 962 1851 093 (077 113) 044 0CHD 3 210 836 8390 077 (070 084) 008 61CHD mortality ndash ndash ndash ndash ndash ndash
PeanutsCVD 3 210 836 14 136 087 (081 093) 067 0CVD mortality 2 134 265 5572 077 (070 085) 081 0Stroke 3 210 836 5910 090 (081 099) 032 13Stroke mortality 4 253 227 3036 083 (073 095) 007 57CHD 3 210 836 8390 085 (079 092) 055 0CHD mortality 2 134 265 2119 075 (064 088) 046 0
WalnutsCVD 3 144 021 5255 081 (071 092) 003 73CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 144 021 5910 085 (071 102) 019 39Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 144 021 2685 079 (066 094) 004 69CHD mortality ndash ndash ndash ndash ndash ndash
Peanut butterCVD 3 210 836 14 136 098 (093 103) lt001 89CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 094 (087 102) lt001 86Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 210 836 8390 100 (094 107) 017 43CHD mortality ndash ndash ndash ndash ndash ndash
Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovascular disease RR risk ratio
Figure 2 Summary plots of effect estimates from prospective cohort studies evaluating the association between nut consumptionand the risk of different cardiovascular outcomes Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovasculardisease RR risk ratio
700 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Nuts and stroke incidence Seven cohort comparisonsanalyzed the association between nut consumption and
stroke incidence including 302 888 participants and12 646 cases The summary RR (95CI) for high vs low
categories of nut consumption was 100 (092ndash109)with no evidence of interstudy heterogeneity (I2 0
Pheterogeneity 097) (Figure 2 and Figure S15 in theSupporting Information online) Regarding specific
types of nuts peanut consumption was associated witha lower risk of stroke incidence after comparing highest
vs lowest categories of consumption ([RR 090 95CI081ndash099] I2 13 Pheterogeneity 032) (Table 3 andFigure S16 in the Supporting Information online) No
association was observed between tree nut walnut andpeanut butter consumption and the risk of stroke inci-
dence (Table 3 and Figures S17ndashS19 in the SupportingInformation online)
Nuts and stroke mortality Twelve cohort comparisons
analyzed the association between nut consumption andstroke mortality including 351 618 participants and
2332 cases The summary RR (95CI) for high vs lowcategories of nut consumption was 083 (075ndash093)
with no evidence of interstudy heterogeneity (I2 0Pheterogeneity 045) (Figure 2 and Figure S20 in the
Supporting Information online) Regarding specifictypes of nuts peanut consumption was associated with
a lower risk of stroke mortality after comparing highestvs lowest categories of consumption ([RR 085 95CI
079ndash092] I2 0 Pheterogeneity 055) (Table 3 andFigure S21 in the Supporting Information online) No
association was reported between tree nut consumptionand the risk of stroke death (Table 3 and Figure S22 in
the Supporting Information online)
Nuts and hemorrhagic stroke Five cohort comparisonsanalyzed the association between nut consumption and
hemorrhagic stroke incidence involving 188 750 partic-ipants and 3088 cases The summary RR (95CI) for
high vs low categories of nut consumption was 102(077ndash134) with no evidence of interstudy heterogene-ity (I2 15 Pheterogeneity 032) (Figure 2 and Figure S23
in the Supporting Information online) No studyanalyzed the association between different types of nuts
and the risk of hemorrhagic stroke
Nuts and ischemic stroke Seven cohort comparisonsanalyzed the association between nut consumption and
ischemic stroke incidence involving 302 423 partici-pants and 8401 cases The summary RR (95CI) for
high vs low categories of nut consumption was 099(089ndash110) with no evidence of interstudy heterogene-
ity (I2 0 Pheterogeneity 062) (Figure 2 and Figure S24 in
the Supporting Information online) No study analyzed
the association between different types of nuts and therisk of hemorrhagic stroke
Nuts and atrial fibrillation Two cohort comparisonsanalyzed the association between nut consumption and
AF involving 53 965 participants and 10 867 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 085 (073ndash099) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 052)
(Figure 2 and Figure S25 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of AF
Nuts and heart failure Two cohort comparisonsanalyzed the association between nut consumption and
HF involving 53 887 participants and 4253 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 100 (086ndash116) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 085)
(Figure 2 and Figure S26 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of HF
Dose-response analyses
Figures S27ndashS34 in the Supporting Information onlineshow the linear and non-linear dose-response analyses
between total nut consumption and CVD outcomesThe summary RR (95CI) for a 28-gd increment was
087 (081ndash093) for CVD incidence 071 (061ndash084)for CVD mortality 075 (064ndash088) for CHD incidence
067 (052ndash087) for CHD mortality 106 (097ndash115) forstroke incidence 101 (088ndash118) for stroke mortality
105 (077ndash143) for hemorrhagic stroke and 106(086ndash131) for ischemic stroke
Total nut consumption and the risk of CVD inci-dence (Figure S27 in the Supporting Information on-
line) CVD mortality (Figure S28 in the SupportingInformation online) stroke mortality (Figure S32 in theSupporting Information online) and hemorrhagic
stroke (Figure S33 in the Supporting Information on-line) showed a non-linear association (Pnon-line-
aritylt001) The reduction in the risk of CVD incidencewas observed up to a consumption of 10 gd with no
further reduction with higher consumptions (Table S6in the Supporting Information online) For CVD mor-
tality and CHD mortality there was a steeper reductionin the risk at approximately 15ndash20 gd with no further
reduction with a higher consumption (Table S6 in theSupporting Information online) The reduction in the
risk of stroke mortality was observed up to a consump-tion of 5 gd with no significant reductions above this
amount (Table S6 in the Supporting Information
Nutrition ReviewsVR Vol 77(10)691ndash709 701
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
or mortality from CVD CHD stroke heart failure(HF) or atrial fibrillation (AF) as the outcome and
reporting effect estimators as odds ratios (ORs) riskratios (RRs) or hazard ratios (HRs) and their corre-
sponding 95 confidence intervals (CIs) When morethan 1 article from the same study was revealed both
papers were included if the end points were different(ie AF in one article and stroke in the other) When
multiple publications from the same study reported thesame outcome the study with the longest follow-up was
selected for inclusion Proceedings or publishedabstracts were not included in the present systematic re-
view and meta-analysisTwo publications were identified for the Nursesrsquo
Health Study I and Health Professionals Follow-up
Study which evaluated the association between nutconsumption and CVD outcomes in a whole population
cohort study21 and in individuals with diabetes only25
Therefore for the present meta-analysis the study con-
ducted in the whole cohort where participants with dia-betes where also included was selected Similarly for
The Netherlands Cohort Study two different articleswere also identified one published in 201526 and an-
other one in 201927 However in both cases the totalsample size and the period of follow-up were the same
Therefore the 2015 publication26 was included becauseits main aim fitted better with the objective of the pre-
sent systematic review and meta-analysis
Data extraction
Two independent reviewers (NB-T and IP-G) reviewed
the full text of the articles that were selected in the firstphase of the screening process Using a standardized pro-
forma the following relevant information was extractedfrom those studies that met all of the inclusion criteria
authors journal and year of publication study design co-hort name country or study location total sample size
characteristics of subjects follow-up duration sources offindings type of exposure and method used for its assess-
ment outcome and method assessment effect estimators
(OR RR or HR and 95 confidence intervals) and statis-tical analyses When necessary authors were contacted by
email to obtain additional information relevant to theanalyses Any disagreement was resolved by discussion or
if necessary by a third author (JS-S)
Quality of the included studies
The Newcastle-Ottawa scale (NOS) was utilized to as-
sess the quality of the included studies28 It is a ratingscale from 0 to 9 where points are allocated according
to 3 different domains population selection outcomeassessment and comparability of the groups A maxi-
mum of 4 3 and 2 points were allotted to each study af-ter evaluation of the aforementioned domains High-
quality studies were considered those studies with a to-tal score of at least 7 points Disagreements in grading
the quality of the studies were resolved through consen-sus between the reviewers
Outcomes
The primary outcomes were CVD incidence (including
only nonfatal or a combination of nonfatal and fataloutcomes of a composite of different CVD outcomes)and CVD mortality which only included a composite
of different fatal CVD end points Secondary outcomesincluded incidence of nonfatal or a combination of
nonfatal and fatal outcomes and mortality from fataloutcomes ie CHD stroke AF and HF
Studies that reported fatal CHD and nonfatal myo-cardial infarction separately were combined using a
fixed-effects model to generate an overall estimate forCHD incidence2930 In the same way following the
same procedure ischemic stroke and intracerebralhemorrhage outcomes22 and fatal ischemic stroke and
fatal hemorrhagic stroke31 end points were combinedto obtain an overall estimate for stroke incidence and
stroke mortality respectively
Table 1 PICOS criteria for inclusion and exclusion of studiesParameter Inclusion criteria Exclusion criteria
Participants General population of adults Aged lt18 yearsInterventionexposure Nut consumption (including total nuts or
subtypes of nuts eg walnuts almondspeanuts peanut butter hazelnuts)
Dietary intakes do not include total nut consumptionor different subtypes of nut consumption
Comparison Extreme quantiles Risk estimate on continuous scaleOutcome Incidence of or mortality from cardiovas-
cular disease coronary heart diseasestroke heart failure atrial fibrillation
Other cardiovascular disease outcomes
Study design Prospective cohort studies Cross-sectional case-control ecological retrospectiveobservational studies clinical trials and non-hu-man studies
Nutrition ReviewsVR Vol 77(10)691ndash709 693
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Statistical analyses
The generic inverse variance method with a random-
effects model (5 comparisons) or fixed-effects model(lt5 comparisons) was used to pool the natural log-
transformed RRs for CVD incidence and mortality out-comes to compare highest vs lowest categories of nut
consumption For one study32 that reported results us-ing the second category of nut consumption rather
than the lowest one as the reference the RR and its cor-responding 95CI were recalculated following the
Hamling et al33 method using the first category as thereference
Heterogeneity among studies was estimated usingCochranrsquos Q test and quantified by the I2 statistic
Statistical significance was set at Plt 010 and an I2
value 50 was considered to reflect substantial
heterogeneityMeta-regression analysis was performed in order
to assess whether a priori specified study characteris-tics (ie sex follow-up geographical area NOS scale
and its individual domains) may have affected theoverall effect estimates This subgroup analysis was
only conducted if at least 10 study comparisons wereavailable23
Sensitivity analysis excluding 1 study at a time and
recalculating the summary estimates was performed toascertain the influence of individual studies on the sum-
mary estimates If the removal of a study yielded achange in the level of significance magnitude (by
gt10) or direction of the pooled risk estimates orchanged the evidence of heterogeneity then it was con-
sidered as influentialLinear dose-response analysis for total nut con-
sumption and different CVD outcomes was conductedfollowing the 2-stage generalized least-squares trend
(GLST) estimation method developed by Greenlandand Longnecker34 and Orsini et al35 In the first stage
the method fits the dose-response model within eachstudy and in the second stage it combines study-
specific trends Data on RRs and the corresponding95CIs total number of participants cases and doses
for at least 3 categories of nut consumption wereneeded to carry out this method The mean or median
of nut consumption from each exposure category wasused if it was directly reported For those studies that
did not report this information the midpoint betweenthe upper and lower boundaries was assigned when
ranges of nut consumption were available For studiesthat reported open-ended extreme categories a width
equal to the adjacent category was assumed in order toestimate the upper or lower cutoff value Some studies
reported the information on nut consumption in gramsand others in servings Therefore servings were
converted to grams where 1 serving equated to 28 g
unless authors specified other serving sizesPotential nonlinear association between nut con-
sumption and CVD outcomes was assessed using re-stricted cubic splines (MKSPLINE procedure) which
were combined using multivariate meta-analysis Thedeparture from linearity was assessed by the Wald testconstraining the regression coefficient for the second
spline equal to zero36
Publication bias was tested by the visual inspection
of the funnel plots for asymmetry and statisticallyBeggrsquos test and Eggerrsquos test When few studies are in-
cluded in the analysis the power of the tests is too lowtherefore publication bias was only examined if more
than 10 study comparisons were included in the analy-sis23 Statistical significance was set at Plt 005
Data analysis was performed using ReviewManager (RevMan) software version 53 Copenhagen
The Nordic Cochrane Centre The CochraneCollaboration 2014 and STATA version 15 software
(StataCorp LP College Station Texas)
Grading the evidence
The GRADE system was used to rate the overall qualityand the strength of the evidence The quality of evi-
dence for each outcome was categorized as high mod-erate low or very low This system regards
observational studies as low-quality evidence37 Thelevel of evidence can be upgraded or downgraded
according to different specified criteria Determinantsto downgrade included study design and execution lim-
itations38 inconsistency39 indirectness40 impreci-sion41 and publication bias42 Determinants to upgradeincluded large magnitude effect dose-response gradi-
ent and attenuation by plausible confounding effects43
Discrepancies in ratings of the evidence quality were
resolved by consensus between NB-T and IP-G
RESULTS
Study selection process
The present systematic review and meta-analysis in-cluded 19 prospective studies from the 2992 identified
articles (Figure 1) Three study comparisons (1 report)were included in the meta-analysis for total CVD21 14
study comparisons (9 reports) for CVD mortal-ity21263144ndash49 7 study comparisons (5 reports) for
CHD2122293050 12 study comparisons (8 reports) forCHD mortality2122263031444851 7 study comparisons
(5 reports) for stroke2122325253 11 study comparisons(7 reports) for stroke mortality21263132474851 5 study
comparisons (4 reports) for hemorrhagic stroke2252ndash54
694 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
7 study comparisons (5 reports) for ischemic
stroke2122325253 2 study comparisons for AF2255 and2 study comparisons for HF2256
Four studies57ndash60 that reported the risk estimate oncontinuous scale instead of categories of nut consump-
tion were identified and therefore were not included inthe high vs low categories of consumption analyses and
the dose-response analyses
Study characteristics
The characteristics of the included studies are presentedin Table 221222629ndash324446ndash60 Publication date ranged
from 1992 to 2018 Six of the studies originated fromEurope 13 from America 1 from Asia 1 from
Australia 1 from both China and the USA and 1 fromboth China and Germany The duration of follow-up
ranged from 43 to 287 years All the studies assessed
nut intake via a food frequency questionnaire The vast
majority of studies (709) were of high quality accord-ing to the NOS scale Tables S1ndashS5 in the Supporting
Information online describe the characteristics of theincluded studies by type of nut consumption
High vs low categories of consumption analyses
Nuts and cardiovascular disease incidence Three cohort
comparisons involving 210 836 participants and 14 136cases analyzed the association between nut consump-
tion and the risk of CVD incidence The summary RR(95CI) for high vs low categories of nut consumption
was 085 (080ndash091) with no evidence of interstudy het-erogeneity (I2frac14 0 Pheterogeneity 081) (Figure 2 and
Figure S1 in the Supporting Information online)Regarding specific types of nuts consumption of tree
nuts ([RR 085 95CI 079ndash091] I2 0
Figure 1 Flow diagram of the literature search and selection process
Nutrition ReviewsVR Vol 77(10)691ndash709 695
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Char
acte
rist
ics
ofth
ein
clud
edst
udie
sev
alua
ting
the
asso
ciat
ion
betw
een
nuts
and
risk
ofca
rdio
vasc
ular
dise
ase
outc
omes
Stud
yCo
untr
ySt
udy
nam
ePo
pula
tion
Nut
cons
umpt
ion
asse
ssm
ent
met
hod
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Fras
eret
al(1
992)
29U
SASe
vent
h-da
yAd
vent
ists
2647
3M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
51
serv
ings
wk
vslt
1se
rvin
gw
k
25
6N
onfa
talM
I13
4Ag
ency
7CH
Dm
orta
lity
463
Albe
rtet
al(2
002)
30U
SAPh
ysic
ians
rsquoH
ealth
Stud
y21
454
MSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gm
o
40ndash8
417
CHD
mor
talit
y56
6Ag
ency
7N
onfa
talM
I10
37
Blom
hoff
etal
(200
6)44
USA
Iow
aW
omen
rsquosH
ealth
Stud
y31
778
FSe
lf-ad
min
is-
tere
dFF
QN
uts
plus
pean
utbu
tter
7se
rvin
gsw
kvs 0
5se
rvin
gsw
k
55ndash6
915
CVD
mor
talit
y16
75Ag
ency
6CH
Dm
orta
lity
948
Djo
uss e
etal
(200
8)56
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2097
6M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gw
k
40ndash8
419
6H
eart
failu
re10
93N
R6
Net
tleto
net
al(2
008)
57U
SAAt
hero
scle
rosi
sRi
skin
Com
mun
ities
Stud
y
1415
3M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
erCo
ntin
uous
1se
rvin
gd
45ndash6
413
Hea
rtfa
ilure
1140
Agen
cy8
Djo
uss e
etal
(201
0)52
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2107
8M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
7se
rvin
gsw
kvs lt
1se
rvin
gw
k
40ndash8
421
1St
roke
1424
Agen
cy6
Isch
emic
stro
ke11
89H
emor
rhag
icst
roke
219
Bern
stei
net
al(2
012)
54U
SAN
urse
srsquoH
ealth
Stud
yI
8001
0F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
034
serv
ings
dvs
0se
rvin
gsd
30ndash5
526
Hem
orrh
agic
stro
ke13
83Ag
ency
6
Hea
lthPr
ofes
sion
als
Follo
w-U
pSt
udy
4315
0M
060
serv
ings
dvs
0se
rvin
gsd
40ndash7
522
Hem
orrh
agic
stro
ke82
9
Khaw
aja
etal
(201
2)55
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2105
4M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
6
serv
ings
wk
vs0
serv
ings
wk
40ndash8
424
Atria
lfib
rilla
tion
3317
Agen
cy7
Yaem
siri
etal
(201
2)58
USA
Wom
enrsquos
Hea
lthIn
itiat
ive
Obs
erva
tiona
lSt
udy
8702
5F
Self-
adm
inis
-te
red
FFQ
Not
spec
ified
Cont
inuo
us1
serv
ing
d50
ndash79
76
Isch
emic
stro
ke10
49Ag
ency
8
von
Rues
ten
etal
(201
3)59
Ger
man
yEP
IC-P
otsd
amst
udy
2353
1M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Pean
uts
wal
nuts
br
azil
nuts
Cont
inuo
us1
serv
ing
d35
ndash65
8CV
D36
3Ag
ency
8
(con
tinue
d)
696 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Gua
sch-
Ferr
eet
al(2
013)
46Sp
ain
PRED
IMED
stud
y72
16M
FIn
terv
iew
-adm
in-
iste
red
vali-
date
dse
mi-
quan
titat
ive
FFQ
Alm
onds
pea
-nu
tsh
azel
-nu
tsp
ista
-ch
ios
pine
nuts
gt3
serv
ings
wk
vs neve
r
55ndash8
04
8CV
Dm
orta
lity
81Ag
ency
-In
dust
ry9
Bona
ccio
etal
(201
5)48
Italy
Mol
i-san
istu
dy10
509
F88
77M
Valid
ated
sel
f-ad
min
iste
red
FFQ
Wal
nuts
haz
el-
nuts
alm
onds
pe
anut
s
Inta
kevs no
inta
ke
gt35
43
CVD
mor
talit
y10
4Ag
ency
6CH
Dm
orta
lity
39St
roke
mor
talit
y19
diG
iuse
ppe
etal
(201
5)32
Ger
man
yEP
IC-P
otsd
amst
udy
2599
7M
FVa
lidat
eds
elf-
adm
inis
tere
dFF
Q
Pean
uts
wal
nuts
br
azil
nuts
142
gd
vs 0g
d
F49
2M
52
58
3St
roke
288
Agen
cy8
Isch
emic
stro
ke23
5St
roke
mor
talit
y36
Gop
inat
het
al(2
015)
51Au
stra
liaBl
ueM
ount
ains
Eye
Stud
y13
12F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
490
to10
0g
dvs 0
to0
50g
d
49
15CV
Dm
orta
lity
258
Agen
cy7
IHD
mor
talit
y18
8St
roke
mor
talit
y10
1
1581
MCV
Dm
orta
lity
288
IHD
mor
talit
y24
2St
roke
mor
talit
y75
Har
ing
etal
(201
4)50
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1206
6M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Not
spec
ified
06
serv
ings
dvs
01
serv
ings
d45
ndash64
22CH
D11
47Ag
ency
9
Har
ing
etal
(201
5)53
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1160
1M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
er1
serv
ings
dvs 0
serv
ings
d
45ndash6
422
7St
roke
699
Agen
cy8
Hem
orrh
agic
stro
ke11
4
Isch
emic
stro
ke59
8H
shie
het
al(2
015)
47U
SAPh
ysic
ians
rsquoH
ealth
Stud
y20
742
MU
nval
idat
eds
elf-
adm
inis
tere
dFF
Q
Not
spec
ified
5
serv
ings
wk
vs lt1
serv
ing
wk
40ndash8
49
6CV
Dm
orta
lity
760
Agen
cy6
Stro
kem
orta
lity
143
Luu
etal
(201
5)31
USA
Sout
hern
Com
mun
ityCo
hort
Stud
y
7176
4M
FSe
mi-q
uant
ita-
tive
FFQ
Tota
lnut
san
dpe
anut
butt
er
184
5g
dvs lt
095
gd
40ndash7
95
4CV
Dm
orta
lity
1857
Agen
cy8
IHD
mor
talit
y79
3Is
chem
icst
roke
mor
talit
y12
1
Hem
orrh
agic
stro
kem
orta
lity
96
(con
tinue
d)
Nutrition ReviewsVR Vol 77(10)691ndash709 697
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Chin
aSh
angh
aiW
omen
rsquosH
ealth
Stud
yan
dSh
angh
aiM
enrsquos
Hea
lthSt
udy
134
265
MF
Pean
ut
254
gd
vs lt0
14gr
amd
ay
40ndash7
0an
d40
ndash74
122
and
65
CVD
mor
talit
y25
87IH
Dm
orta
lity
631
Isch
emic
stro
kem
orta
lity
588
Hem
orrh
agic
stro
kem
orta
lity
597
van
den
Bran
dtan
dSc
hout
en(2
015)
26
Net
herla
nds
Net
herla
nds
Coho
rtSt
udy
3202
MF
(sub
coho
rt)
Self-
adm
inis
-te
red
valid
ated
FFQ
Tree
nuts
pe
anut
s19
6g
dvs 0
gd
55ndash6
99
CVD
mor
talit
y29
85N
R7
IHD
mor
talit
y14
88St
roke
mor
talit
y56
5
Wan
get
al(2
016)
60Ch
ina
Linx
ian
NIT
coho
rt24
45M
FFF
QPe
anut
sch
est-
nuts
wal
nuts
Cont
inuo
us3
serv
ings
mo
40ndash6
926
Hea
rtdi
seas
em
orta
lity
355
Agen
cy6
Ger
man
ySt
roke
mor
talit
y45
2
Esla
mpa
rast
etal
(201
7)49
Iran
Gol
esta
nCo
hort
Stud
y28
257
FVa
lidat
eds
elf-
adm
inis
tere
dse
miq
uant
ita-
tive
FFQ
Pean
uts
tree
nuts
3
serv
ings
wk
vs neve
r
40ndash8
77
CVD
mor
talit
y91
1Ag
ency
720
855
M11
05
Gua
sch-
Ferr
eet
al(2
017)
21U
SAN
urse
srsquoH
ealth
Stud
yI
7636
4F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Pean
uts
othe
rnu
tsa
ndw
al-
nuts
(ifav
aila
ble)
5
serv
ings
wk
vs neve
rora
lmos
tne
ver
30ndash5
528
7CV
D67
27Ag
ency
7CV
Dm
orta
lity
1770
CHD
3552
CHD
mor
talit
y99
6St
roke
3322
Stro
kem
orta
lity
773
Isch
emic
stro
ke16
35N
urse
srsquoH
ealth
Stud
yII
9294
6F
25ndash4
221
5CV
D19
15CV
Dm
orta
lity
82CH
D67
0CH
Dm
orta
lity
46St
roke
1262
Stro
kem
orta
lity
36
Isch
emic
stro
ke22
0H
ealth
Prof
essi
onal
sFo
llow
-Up
Stud
y
4152
6M
40ndash7
522
5CV
D54
94CV
Dm
orta
lity
2599
CHD
4168
CHD
mor
talit
y19
21St
roke
1326
Stro
kem
orta
lity
367
Isch
emic
stro
ke74
2
(con
tinue
d)
698 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Pheterogeneity 070) peanuts ([RR 087 95CI
081ndash093] I2 0 Pheterogeneity 067) and walnuts ([RR081 95CI 071ndash092] I2 73 Pheterogeneity 003) was
associated with a lower risk of CVD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S2ndashS4 in the SupportingInformation online) No association was reported be-tween peanut butter consumption and the risk of CVD
incidence ([RR 098 95CI 093ndash103] I2 89Pheterogeneitylt001) (Table 3 and Figure S5 in the
Supporting Information online)
Nuts and cardiovascular disease mortality Fifteen co-hort comparisons analyzed the association between
nut consumption and CVD mortality including413 727 participants and 14 475 cases The summary
RR (95CI) for high vs low categories of nut con-sumption was 077 (072ndash082) with no evidence of
interstudy heterogeneity (I2 3 Pheterogeneity 042)(Figure 2 and Figure S6 in the Supporting Information
online) Regarding specific types of nuts only peanutshave been studied in relation to CVD mortality show-
ing a summary RR of 077 (95CI 070ndash085) for highvs low categories of consumption with no evidence of
interstudy heterogeneity (I2 0 Pheterogeneity 081)(Table 3 and Figure S7 in the Supporting Information
online)
Nuts and coronary heart disease incidence Seven cohortcomparisons analyzed the association between nut con-
sumption and CHD incidence including 275 812 par-ticipants and 12 654 cases The summary RR (95CI)
for high vs low categories of nut consumption was 082(069ndash096) with evidence of substantial interstudy het-
erogeneity (I2 74 Pheterogeneitylt001) (Figure 2 andFigure S8 in the Supporting Information online)
Regarding specific types of nuts consumption of treenuts ([RR 077 95CI 070ndash084] I2 61
Pheterogeneity 008) peanuts ([RR 085 95CI 079ndash092] I2 0 Pheterogeneity 055) and walnuts ([RR0791 95CI 066ndash094] I2 69 Pheterogeneity 004) was
associated with a lower risk of CHD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S9ndashS11 in the SupportingInformation online) No association was reported be-
tween peanut butter consumption and the risk of CHDincidence ([RR 100 95CI 094ndash107] I2 43
Pheterogeneity 017) (Table 3 and Figure S12 in theSupporting Information online)
Nuts and coronary heart disease mortality Thirteen co-
hort comparisons analyzed the association between nutconsumption and CHD mortality including 396 041
participants and 7877 cases The summary RR (95CI)Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Lars
son
etal
(201
8)22
Swed
enCo
hort
ofSw
edis
hM
en32
911
MFF
QN
uts
(not
incl
ud-
ing
coco
nuto
rch
estn
uts)
3
serv
ings
wk
vs neve
r
45ndash8
317
MI
4983
Agen
cy8
MIm
orta
lity
917
Hea
rtfa
ilure
3160
Atria
lfib
rilla
tion
7550
Isch
emic
stro
ke37
82In
trac
ereb
ral
hem
orrh
age
543
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CVD
car
diov
ascu
lard
isea
seE
PIC
Euro
pean
Pros
pect
ive
Inve
stig
atio
nin
toCa
ncer
and
Nut
ritio
nF
fem
ales
FFQ
foo
dfr
eque
ncy
ques
tionn
aire
IH
Di
sche
mic
hear
tdis
ease
Mm
ales
MI
myo
card
iali
nfar
ctio
nN
ITN
utrit
ion
Inte
rven
tion
Tria
lsN
OS
New
cast
le-O
ttaw
asc
ale
NR
Not
repo
rted
PRE
DIM
EDP
reve
nci o
nco
nD
ieta
MED
iterr
anea
Nutrition ReviewsVR Vol 77(10)691ndash709 699
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
for high vs low categories of nut consumption was 076(067ndash086) with evidence of moderate interstudy het-
erogeneity (I2 46 Pheterogeneity 004) (Figure 2 andFigure S13 in the Supporting Information online)
Regarding specific types of nuts peanut consumption
was inversely associated with the risk of CHD mortalityafter comparing high vs low categories of consumption
([RR 075 95CI 064ndash088] I2 0 Pheterogeneity 046)(Table 3 and Figure S14 in the Supporting Information
online)
Table 3 Summary RR of cardiovascular disease outcomes by different types of nut consumption (comparing highest vslowest categories)
No of cohorts No of participants No of cases RR (95CI) P for heterogeneity I2 ()
Tree nutsCVD 3 210 836 14 136 085 (079 091) 070 0CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 100 (089 111) 093 0Stroke mortality 3 118 962 1851 093 (077 113) 044 0CHD 3 210 836 8390 077 (070 084) 008 61CHD mortality ndash ndash ndash ndash ndash ndash
PeanutsCVD 3 210 836 14 136 087 (081 093) 067 0CVD mortality 2 134 265 5572 077 (070 085) 081 0Stroke 3 210 836 5910 090 (081 099) 032 13Stroke mortality 4 253 227 3036 083 (073 095) 007 57CHD 3 210 836 8390 085 (079 092) 055 0CHD mortality 2 134 265 2119 075 (064 088) 046 0
WalnutsCVD 3 144 021 5255 081 (071 092) 003 73CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 144 021 5910 085 (071 102) 019 39Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 144 021 2685 079 (066 094) 004 69CHD mortality ndash ndash ndash ndash ndash ndash
Peanut butterCVD 3 210 836 14 136 098 (093 103) lt001 89CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 094 (087 102) lt001 86Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 210 836 8390 100 (094 107) 017 43CHD mortality ndash ndash ndash ndash ndash ndash
Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovascular disease RR risk ratio
Figure 2 Summary plots of effect estimates from prospective cohort studies evaluating the association between nut consumptionand the risk of different cardiovascular outcomes Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovasculardisease RR risk ratio
700 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Nuts and stroke incidence Seven cohort comparisonsanalyzed the association between nut consumption and
stroke incidence including 302 888 participants and12 646 cases The summary RR (95CI) for high vs low
categories of nut consumption was 100 (092ndash109)with no evidence of interstudy heterogeneity (I2 0
Pheterogeneity 097) (Figure 2 and Figure S15 in theSupporting Information online) Regarding specific
types of nuts peanut consumption was associated witha lower risk of stroke incidence after comparing highest
vs lowest categories of consumption ([RR 090 95CI081ndash099] I2 13 Pheterogeneity 032) (Table 3 andFigure S16 in the Supporting Information online) No
association was observed between tree nut walnut andpeanut butter consumption and the risk of stroke inci-
dence (Table 3 and Figures S17ndashS19 in the SupportingInformation online)
Nuts and stroke mortality Twelve cohort comparisons
analyzed the association between nut consumption andstroke mortality including 351 618 participants and
2332 cases The summary RR (95CI) for high vs lowcategories of nut consumption was 083 (075ndash093)
with no evidence of interstudy heterogeneity (I2 0Pheterogeneity 045) (Figure 2 and Figure S20 in the
Supporting Information online) Regarding specifictypes of nuts peanut consumption was associated with
a lower risk of stroke mortality after comparing highestvs lowest categories of consumption ([RR 085 95CI
079ndash092] I2 0 Pheterogeneity 055) (Table 3 andFigure S21 in the Supporting Information online) No
association was reported between tree nut consumptionand the risk of stroke death (Table 3 and Figure S22 in
the Supporting Information online)
Nuts and hemorrhagic stroke Five cohort comparisonsanalyzed the association between nut consumption and
hemorrhagic stroke incidence involving 188 750 partic-ipants and 3088 cases The summary RR (95CI) for
high vs low categories of nut consumption was 102(077ndash134) with no evidence of interstudy heterogene-ity (I2 15 Pheterogeneity 032) (Figure 2 and Figure S23
in the Supporting Information online) No studyanalyzed the association between different types of nuts
and the risk of hemorrhagic stroke
Nuts and ischemic stroke Seven cohort comparisonsanalyzed the association between nut consumption and
ischemic stroke incidence involving 302 423 partici-pants and 8401 cases The summary RR (95CI) for
high vs low categories of nut consumption was 099(089ndash110) with no evidence of interstudy heterogene-
ity (I2 0 Pheterogeneity 062) (Figure 2 and Figure S24 in
the Supporting Information online) No study analyzed
the association between different types of nuts and therisk of hemorrhagic stroke
Nuts and atrial fibrillation Two cohort comparisonsanalyzed the association between nut consumption and
AF involving 53 965 participants and 10 867 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 085 (073ndash099) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 052)
(Figure 2 and Figure S25 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of AF
Nuts and heart failure Two cohort comparisonsanalyzed the association between nut consumption and
HF involving 53 887 participants and 4253 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 100 (086ndash116) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 085)
(Figure 2 and Figure S26 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of HF
Dose-response analyses
Figures S27ndashS34 in the Supporting Information onlineshow the linear and non-linear dose-response analyses
between total nut consumption and CVD outcomesThe summary RR (95CI) for a 28-gd increment was
087 (081ndash093) for CVD incidence 071 (061ndash084)for CVD mortality 075 (064ndash088) for CHD incidence
067 (052ndash087) for CHD mortality 106 (097ndash115) forstroke incidence 101 (088ndash118) for stroke mortality
105 (077ndash143) for hemorrhagic stroke and 106(086ndash131) for ischemic stroke
Total nut consumption and the risk of CVD inci-dence (Figure S27 in the Supporting Information on-
line) CVD mortality (Figure S28 in the SupportingInformation online) stroke mortality (Figure S32 in theSupporting Information online) and hemorrhagic
stroke (Figure S33 in the Supporting Information on-line) showed a non-linear association (Pnon-line-
aritylt001) The reduction in the risk of CVD incidencewas observed up to a consumption of 10 gd with no
further reduction with higher consumptions (Table S6in the Supporting Information online) For CVD mor-
tality and CHD mortality there was a steeper reductionin the risk at approximately 15ndash20 gd with no further
reduction with a higher consumption (Table S6 in theSupporting Information online) The reduction in the
risk of stroke mortality was observed up to a consump-tion of 5 gd with no significant reductions above this
amount (Table S6 in the Supporting Information
Nutrition ReviewsVR Vol 77(10)691ndash709 701
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Statistical analyses
The generic inverse variance method with a random-
effects model (5 comparisons) or fixed-effects model(lt5 comparisons) was used to pool the natural log-
transformed RRs for CVD incidence and mortality out-comes to compare highest vs lowest categories of nut
consumption For one study32 that reported results us-ing the second category of nut consumption rather
than the lowest one as the reference the RR and its cor-responding 95CI were recalculated following the
Hamling et al33 method using the first category as thereference
Heterogeneity among studies was estimated usingCochranrsquos Q test and quantified by the I2 statistic
Statistical significance was set at Plt 010 and an I2
value 50 was considered to reflect substantial
heterogeneityMeta-regression analysis was performed in order
to assess whether a priori specified study characteris-tics (ie sex follow-up geographical area NOS scale
and its individual domains) may have affected theoverall effect estimates This subgroup analysis was
only conducted if at least 10 study comparisons wereavailable23
Sensitivity analysis excluding 1 study at a time and
recalculating the summary estimates was performed toascertain the influence of individual studies on the sum-
mary estimates If the removal of a study yielded achange in the level of significance magnitude (by
gt10) or direction of the pooled risk estimates orchanged the evidence of heterogeneity then it was con-
sidered as influentialLinear dose-response analysis for total nut con-
sumption and different CVD outcomes was conductedfollowing the 2-stage generalized least-squares trend
(GLST) estimation method developed by Greenlandand Longnecker34 and Orsini et al35 In the first stage
the method fits the dose-response model within eachstudy and in the second stage it combines study-
specific trends Data on RRs and the corresponding95CIs total number of participants cases and doses
for at least 3 categories of nut consumption wereneeded to carry out this method The mean or median
of nut consumption from each exposure category wasused if it was directly reported For those studies that
did not report this information the midpoint betweenthe upper and lower boundaries was assigned when
ranges of nut consumption were available For studiesthat reported open-ended extreme categories a width
equal to the adjacent category was assumed in order toestimate the upper or lower cutoff value Some studies
reported the information on nut consumption in gramsand others in servings Therefore servings were
converted to grams where 1 serving equated to 28 g
unless authors specified other serving sizesPotential nonlinear association between nut con-
sumption and CVD outcomes was assessed using re-stricted cubic splines (MKSPLINE procedure) which
were combined using multivariate meta-analysis Thedeparture from linearity was assessed by the Wald testconstraining the regression coefficient for the second
spline equal to zero36
Publication bias was tested by the visual inspection
of the funnel plots for asymmetry and statisticallyBeggrsquos test and Eggerrsquos test When few studies are in-
cluded in the analysis the power of the tests is too lowtherefore publication bias was only examined if more
than 10 study comparisons were included in the analy-sis23 Statistical significance was set at Plt 005
Data analysis was performed using ReviewManager (RevMan) software version 53 Copenhagen
The Nordic Cochrane Centre The CochraneCollaboration 2014 and STATA version 15 software
(StataCorp LP College Station Texas)
Grading the evidence
The GRADE system was used to rate the overall qualityand the strength of the evidence The quality of evi-
dence for each outcome was categorized as high mod-erate low or very low This system regards
observational studies as low-quality evidence37 Thelevel of evidence can be upgraded or downgraded
according to different specified criteria Determinantsto downgrade included study design and execution lim-
itations38 inconsistency39 indirectness40 impreci-sion41 and publication bias42 Determinants to upgradeincluded large magnitude effect dose-response gradi-
ent and attenuation by plausible confounding effects43
Discrepancies in ratings of the evidence quality were
resolved by consensus between NB-T and IP-G
RESULTS
Study selection process
The present systematic review and meta-analysis in-cluded 19 prospective studies from the 2992 identified
articles (Figure 1) Three study comparisons (1 report)were included in the meta-analysis for total CVD21 14
study comparisons (9 reports) for CVD mortal-ity21263144ndash49 7 study comparisons (5 reports) for
CHD2122293050 12 study comparisons (8 reports) forCHD mortality2122263031444851 7 study comparisons
(5 reports) for stroke2122325253 11 study comparisons(7 reports) for stroke mortality21263132474851 5 study
comparisons (4 reports) for hemorrhagic stroke2252ndash54
694 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
7 study comparisons (5 reports) for ischemic
stroke2122325253 2 study comparisons for AF2255 and2 study comparisons for HF2256
Four studies57ndash60 that reported the risk estimate oncontinuous scale instead of categories of nut consump-
tion were identified and therefore were not included inthe high vs low categories of consumption analyses and
the dose-response analyses
Study characteristics
The characteristics of the included studies are presentedin Table 221222629ndash324446ndash60 Publication date ranged
from 1992 to 2018 Six of the studies originated fromEurope 13 from America 1 from Asia 1 from
Australia 1 from both China and the USA and 1 fromboth China and Germany The duration of follow-up
ranged from 43 to 287 years All the studies assessed
nut intake via a food frequency questionnaire The vast
majority of studies (709) were of high quality accord-ing to the NOS scale Tables S1ndashS5 in the Supporting
Information online describe the characteristics of theincluded studies by type of nut consumption
High vs low categories of consumption analyses
Nuts and cardiovascular disease incidence Three cohort
comparisons involving 210 836 participants and 14 136cases analyzed the association between nut consump-
tion and the risk of CVD incidence The summary RR(95CI) for high vs low categories of nut consumption
was 085 (080ndash091) with no evidence of interstudy het-erogeneity (I2frac14 0 Pheterogeneity 081) (Figure 2 and
Figure S1 in the Supporting Information online)Regarding specific types of nuts consumption of tree
nuts ([RR 085 95CI 079ndash091] I2 0
Figure 1 Flow diagram of the literature search and selection process
Nutrition ReviewsVR Vol 77(10)691ndash709 695
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Char
acte
rist
ics
ofth
ein
clud
edst
udie
sev
alua
ting
the
asso
ciat
ion
betw
een
nuts
and
risk
ofca
rdio
vasc
ular
dise
ase
outc
omes
Stud
yCo
untr
ySt
udy
nam
ePo
pula
tion
Nut
cons
umpt
ion
asse
ssm
ent
met
hod
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Fras
eret
al(1
992)
29U
SASe
vent
h-da
yAd
vent
ists
2647
3M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
51
serv
ings
wk
vslt
1se
rvin
gw
k
25
6N
onfa
talM
I13
4Ag
ency
7CH
Dm
orta
lity
463
Albe
rtet
al(2
002)
30U
SAPh
ysic
ians
rsquoH
ealth
Stud
y21
454
MSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gm
o
40ndash8
417
CHD
mor
talit
y56
6Ag
ency
7N
onfa
talM
I10
37
Blom
hoff
etal
(200
6)44
USA
Iow
aW
omen
rsquosH
ealth
Stud
y31
778
FSe
lf-ad
min
is-
tere
dFF
QN
uts
plus
pean
utbu
tter
7se
rvin
gsw
kvs 0
5se
rvin
gsw
k
55ndash6
915
CVD
mor
talit
y16
75Ag
ency
6CH
Dm
orta
lity
948
Djo
uss e
etal
(200
8)56
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2097
6M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gw
k
40ndash8
419
6H
eart
failu
re10
93N
R6
Net
tleto
net
al(2
008)
57U
SAAt
hero
scle
rosi
sRi
skin
Com
mun
ities
Stud
y
1415
3M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
erCo
ntin
uous
1se
rvin
gd
45ndash6
413
Hea
rtfa
ilure
1140
Agen
cy8
Djo
uss e
etal
(201
0)52
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2107
8M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
7se
rvin
gsw
kvs lt
1se
rvin
gw
k
40ndash8
421
1St
roke
1424
Agen
cy6
Isch
emic
stro
ke11
89H
emor
rhag
icst
roke
219
Bern
stei
net
al(2
012)
54U
SAN
urse
srsquoH
ealth
Stud
yI
8001
0F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
034
serv
ings
dvs
0se
rvin
gsd
30ndash5
526
Hem
orrh
agic
stro
ke13
83Ag
ency
6
Hea
lthPr
ofes
sion
als
Follo
w-U
pSt
udy
4315
0M
060
serv
ings
dvs
0se
rvin
gsd
40ndash7
522
Hem
orrh
agic
stro
ke82
9
Khaw
aja
etal
(201
2)55
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2105
4M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
6
serv
ings
wk
vs0
serv
ings
wk
40ndash8
424
Atria
lfib
rilla
tion
3317
Agen
cy7
Yaem
siri
etal
(201
2)58
USA
Wom
enrsquos
Hea
lthIn
itiat
ive
Obs
erva
tiona
lSt
udy
8702
5F
Self-
adm
inis
-te
red
FFQ
Not
spec
ified
Cont
inuo
us1
serv
ing
d50
ndash79
76
Isch
emic
stro
ke10
49Ag
ency
8
von
Rues
ten
etal
(201
3)59
Ger
man
yEP
IC-P
otsd
amst
udy
2353
1M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Pean
uts
wal
nuts
br
azil
nuts
Cont
inuo
us1
serv
ing
d35
ndash65
8CV
D36
3Ag
ency
8
(con
tinue
d)
696 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Gua
sch-
Ferr
eet
al(2
013)
46Sp
ain
PRED
IMED
stud
y72
16M
FIn
terv
iew
-adm
in-
iste
red
vali-
date
dse
mi-
quan
titat
ive
FFQ
Alm
onds
pea
-nu
tsh
azel
-nu
tsp
ista
-ch
ios
pine
nuts
gt3
serv
ings
wk
vs neve
r
55ndash8
04
8CV
Dm
orta
lity
81Ag
ency
-In
dust
ry9
Bona
ccio
etal
(201
5)48
Italy
Mol
i-san
istu
dy10
509
F88
77M
Valid
ated
sel
f-ad
min
iste
red
FFQ
Wal
nuts
haz
el-
nuts
alm
onds
pe
anut
s
Inta
kevs no
inta
ke
gt35
43
CVD
mor
talit
y10
4Ag
ency
6CH
Dm
orta
lity
39St
roke
mor
talit
y19
diG
iuse
ppe
etal
(201
5)32
Ger
man
yEP
IC-P
otsd
amst
udy
2599
7M
FVa
lidat
eds
elf-
adm
inis
tere
dFF
Q
Pean
uts
wal
nuts
br
azil
nuts
142
gd
vs 0g
d
F49
2M
52
58
3St
roke
288
Agen
cy8
Isch
emic
stro
ke23
5St
roke
mor
talit
y36
Gop
inat
het
al(2
015)
51Au
stra
liaBl
ueM
ount
ains
Eye
Stud
y13
12F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
490
to10
0g
dvs 0
to0
50g
d
49
15CV
Dm
orta
lity
258
Agen
cy7
IHD
mor
talit
y18
8St
roke
mor
talit
y10
1
1581
MCV
Dm
orta
lity
288
IHD
mor
talit
y24
2St
roke
mor
talit
y75
Har
ing
etal
(201
4)50
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1206
6M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Not
spec
ified
06
serv
ings
dvs
01
serv
ings
d45
ndash64
22CH
D11
47Ag
ency
9
Har
ing
etal
(201
5)53
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1160
1M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
er1
serv
ings
dvs 0
serv
ings
d
45ndash6
422
7St
roke
699
Agen
cy8
Hem
orrh
agic
stro
ke11
4
Isch
emic
stro
ke59
8H
shie
het
al(2
015)
47U
SAPh
ysic
ians
rsquoH
ealth
Stud
y20
742
MU
nval
idat
eds
elf-
adm
inis
tere
dFF
Q
Not
spec
ified
5
serv
ings
wk
vs lt1
serv
ing
wk
40ndash8
49
6CV
Dm
orta
lity
760
Agen
cy6
Stro
kem
orta
lity
143
Luu
etal
(201
5)31
USA
Sout
hern
Com
mun
ityCo
hort
Stud
y
7176
4M
FSe
mi-q
uant
ita-
tive
FFQ
Tota
lnut
san
dpe
anut
butt
er
184
5g
dvs lt
095
gd
40ndash7
95
4CV
Dm
orta
lity
1857
Agen
cy8
IHD
mor
talit
y79
3Is
chem
icst
roke
mor
talit
y12
1
Hem
orrh
agic
stro
kem
orta
lity
96
(con
tinue
d)
Nutrition ReviewsVR Vol 77(10)691ndash709 697
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Chin
aSh
angh
aiW
omen
rsquosH
ealth
Stud
yan
dSh
angh
aiM
enrsquos
Hea
lthSt
udy
134
265
MF
Pean
ut
254
gd
vs lt0
14gr
amd
ay
40ndash7
0an
d40
ndash74
122
and
65
CVD
mor
talit
y25
87IH
Dm
orta
lity
631
Isch
emic
stro
kem
orta
lity
588
Hem
orrh
agic
stro
kem
orta
lity
597
van
den
Bran
dtan
dSc
hout
en(2
015)
26
Net
herla
nds
Net
herla
nds
Coho
rtSt
udy
3202
MF
(sub
coho
rt)
Self-
adm
inis
-te
red
valid
ated
FFQ
Tree
nuts
pe
anut
s19
6g
dvs 0
gd
55ndash6
99
CVD
mor
talit
y29
85N
R7
IHD
mor
talit
y14
88St
roke
mor
talit
y56
5
Wan
get
al(2
016)
60Ch
ina
Linx
ian
NIT
coho
rt24
45M
FFF
QPe
anut
sch
est-
nuts
wal
nuts
Cont
inuo
us3
serv
ings
mo
40ndash6
926
Hea
rtdi
seas
em
orta
lity
355
Agen
cy6
Ger
man
ySt
roke
mor
talit
y45
2
Esla
mpa
rast
etal
(201
7)49
Iran
Gol
esta
nCo
hort
Stud
y28
257
FVa
lidat
eds
elf-
adm
inis
tere
dse
miq
uant
ita-
tive
FFQ
Pean
uts
tree
nuts
3
serv
ings
wk
vs neve
r
40ndash8
77
CVD
mor
talit
y91
1Ag
ency
720
855
M11
05
Gua
sch-
Ferr
eet
al(2
017)
21U
SAN
urse
srsquoH
ealth
Stud
yI
7636
4F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Pean
uts
othe
rnu
tsa
ndw
al-
nuts
(ifav
aila
ble)
5
serv
ings
wk
vs neve
rora
lmos
tne
ver
30ndash5
528
7CV
D67
27Ag
ency
7CV
Dm
orta
lity
1770
CHD
3552
CHD
mor
talit
y99
6St
roke
3322
Stro
kem
orta
lity
773
Isch
emic
stro
ke16
35N
urse
srsquoH
ealth
Stud
yII
9294
6F
25ndash4
221
5CV
D19
15CV
Dm
orta
lity
82CH
D67
0CH
Dm
orta
lity
46St
roke
1262
Stro
kem
orta
lity
36
Isch
emic
stro
ke22
0H
ealth
Prof
essi
onal
sFo
llow
-Up
Stud
y
4152
6M
40ndash7
522
5CV
D54
94CV
Dm
orta
lity
2599
CHD
4168
CHD
mor
talit
y19
21St
roke
1326
Stro
kem
orta
lity
367
Isch
emic
stro
ke74
2
(con
tinue
d)
698 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Pheterogeneity 070) peanuts ([RR 087 95CI
081ndash093] I2 0 Pheterogeneity 067) and walnuts ([RR081 95CI 071ndash092] I2 73 Pheterogeneity 003) was
associated with a lower risk of CVD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S2ndashS4 in the SupportingInformation online) No association was reported be-tween peanut butter consumption and the risk of CVD
incidence ([RR 098 95CI 093ndash103] I2 89Pheterogeneitylt001) (Table 3 and Figure S5 in the
Supporting Information online)
Nuts and cardiovascular disease mortality Fifteen co-hort comparisons analyzed the association between
nut consumption and CVD mortality including413 727 participants and 14 475 cases The summary
RR (95CI) for high vs low categories of nut con-sumption was 077 (072ndash082) with no evidence of
interstudy heterogeneity (I2 3 Pheterogeneity 042)(Figure 2 and Figure S6 in the Supporting Information
online) Regarding specific types of nuts only peanutshave been studied in relation to CVD mortality show-
ing a summary RR of 077 (95CI 070ndash085) for highvs low categories of consumption with no evidence of
interstudy heterogeneity (I2 0 Pheterogeneity 081)(Table 3 and Figure S7 in the Supporting Information
online)
Nuts and coronary heart disease incidence Seven cohortcomparisons analyzed the association between nut con-
sumption and CHD incidence including 275 812 par-ticipants and 12 654 cases The summary RR (95CI)
for high vs low categories of nut consumption was 082(069ndash096) with evidence of substantial interstudy het-
erogeneity (I2 74 Pheterogeneitylt001) (Figure 2 andFigure S8 in the Supporting Information online)
Regarding specific types of nuts consumption of treenuts ([RR 077 95CI 070ndash084] I2 61
Pheterogeneity 008) peanuts ([RR 085 95CI 079ndash092] I2 0 Pheterogeneity 055) and walnuts ([RR0791 95CI 066ndash094] I2 69 Pheterogeneity 004) was
associated with a lower risk of CHD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S9ndashS11 in the SupportingInformation online) No association was reported be-
tween peanut butter consumption and the risk of CHDincidence ([RR 100 95CI 094ndash107] I2 43
Pheterogeneity 017) (Table 3 and Figure S12 in theSupporting Information online)
Nuts and coronary heart disease mortality Thirteen co-
hort comparisons analyzed the association between nutconsumption and CHD mortality including 396 041
participants and 7877 cases The summary RR (95CI)Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Lars
son
etal
(201
8)22
Swed
enCo
hort
ofSw
edis
hM
en32
911
MFF
QN
uts
(not
incl
ud-
ing
coco
nuto
rch
estn
uts)
3
serv
ings
wk
vs neve
r
45ndash8
317
MI
4983
Agen
cy8
MIm
orta
lity
917
Hea
rtfa
ilure
3160
Atria
lfib
rilla
tion
7550
Isch
emic
stro
ke37
82In
trac
ereb
ral
hem
orrh
age
543
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CVD
car
diov
ascu
lard
isea
seE
PIC
Euro
pean
Pros
pect
ive
Inve
stig
atio
nin
toCa
ncer
and
Nut
ritio
nF
fem
ales
FFQ
foo
dfr
eque
ncy
ques
tionn
aire
IH
Di
sche
mic
hear
tdis
ease
Mm
ales
MI
myo
card
iali
nfar
ctio
nN
ITN
utrit
ion
Inte
rven
tion
Tria
lsN
OS
New
cast
le-O
ttaw
asc
ale
NR
Not
repo
rted
PRE
DIM
EDP
reve
nci o
nco
nD
ieta
MED
iterr
anea
Nutrition ReviewsVR Vol 77(10)691ndash709 699
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
for high vs low categories of nut consumption was 076(067ndash086) with evidence of moderate interstudy het-
erogeneity (I2 46 Pheterogeneity 004) (Figure 2 andFigure S13 in the Supporting Information online)
Regarding specific types of nuts peanut consumption
was inversely associated with the risk of CHD mortalityafter comparing high vs low categories of consumption
([RR 075 95CI 064ndash088] I2 0 Pheterogeneity 046)(Table 3 and Figure S14 in the Supporting Information
online)
Table 3 Summary RR of cardiovascular disease outcomes by different types of nut consumption (comparing highest vslowest categories)
No of cohorts No of participants No of cases RR (95CI) P for heterogeneity I2 ()
Tree nutsCVD 3 210 836 14 136 085 (079 091) 070 0CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 100 (089 111) 093 0Stroke mortality 3 118 962 1851 093 (077 113) 044 0CHD 3 210 836 8390 077 (070 084) 008 61CHD mortality ndash ndash ndash ndash ndash ndash
PeanutsCVD 3 210 836 14 136 087 (081 093) 067 0CVD mortality 2 134 265 5572 077 (070 085) 081 0Stroke 3 210 836 5910 090 (081 099) 032 13Stroke mortality 4 253 227 3036 083 (073 095) 007 57CHD 3 210 836 8390 085 (079 092) 055 0CHD mortality 2 134 265 2119 075 (064 088) 046 0
WalnutsCVD 3 144 021 5255 081 (071 092) 003 73CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 144 021 5910 085 (071 102) 019 39Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 144 021 2685 079 (066 094) 004 69CHD mortality ndash ndash ndash ndash ndash ndash
Peanut butterCVD 3 210 836 14 136 098 (093 103) lt001 89CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 094 (087 102) lt001 86Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 210 836 8390 100 (094 107) 017 43CHD mortality ndash ndash ndash ndash ndash ndash
Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovascular disease RR risk ratio
Figure 2 Summary plots of effect estimates from prospective cohort studies evaluating the association between nut consumptionand the risk of different cardiovascular outcomes Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovasculardisease RR risk ratio
700 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Nuts and stroke incidence Seven cohort comparisonsanalyzed the association between nut consumption and
stroke incidence including 302 888 participants and12 646 cases The summary RR (95CI) for high vs low
categories of nut consumption was 100 (092ndash109)with no evidence of interstudy heterogeneity (I2 0
Pheterogeneity 097) (Figure 2 and Figure S15 in theSupporting Information online) Regarding specific
types of nuts peanut consumption was associated witha lower risk of stroke incidence after comparing highest
vs lowest categories of consumption ([RR 090 95CI081ndash099] I2 13 Pheterogeneity 032) (Table 3 andFigure S16 in the Supporting Information online) No
association was observed between tree nut walnut andpeanut butter consumption and the risk of stroke inci-
dence (Table 3 and Figures S17ndashS19 in the SupportingInformation online)
Nuts and stroke mortality Twelve cohort comparisons
analyzed the association between nut consumption andstroke mortality including 351 618 participants and
2332 cases The summary RR (95CI) for high vs lowcategories of nut consumption was 083 (075ndash093)
with no evidence of interstudy heterogeneity (I2 0Pheterogeneity 045) (Figure 2 and Figure S20 in the
Supporting Information online) Regarding specifictypes of nuts peanut consumption was associated with
a lower risk of stroke mortality after comparing highestvs lowest categories of consumption ([RR 085 95CI
079ndash092] I2 0 Pheterogeneity 055) (Table 3 andFigure S21 in the Supporting Information online) No
association was reported between tree nut consumptionand the risk of stroke death (Table 3 and Figure S22 in
the Supporting Information online)
Nuts and hemorrhagic stroke Five cohort comparisonsanalyzed the association between nut consumption and
hemorrhagic stroke incidence involving 188 750 partic-ipants and 3088 cases The summary RR (95CI) for
high vs low categories of nut consumption was 102(077ndash134) with no evidence of interstudy heterogene-ity (I2 15 Pheterogeneity 032) (Figure 2 and Figure S23
in the Supporting Information online) No studyanalyzed the association between different types of nuts
and the risk of hemorrhagic stroke
Nuts and ischemic stroke Seven cohort comparisonsanalyzed the association between nut consumption and
ischemic stroke incidence involving 302 423 partici-pants and 8401 cases The summary RR (95CI) for
high vs low categories of nut consumption was 099(089ndash110) with no evidence of interstudy heterogene-
ity (I2 0 Pheterogeneity 062) (Figure 2 and Figure S24 in
the Supporting Information online) No study analyzed
the association between different types of nuts and therisk of hemorrhagic stroke
Nuts and atrial fibrillation Two cohort comparisonsanalyzed the association between nut consumption and
AF involving 53 965 participants and 10 867 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 085 (073ndash099) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 052)
(Figure 2 and Figure S25 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of AF
Nuts and heart failure Two cohort comparisonsanalyzed the association between nut consumption and
HF involving 53 887 participants and 4253 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 100 (086ndash116) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 085)
(Figure 2 and Figure S26 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of HF
Dose-response analyses
Figures S27ndashS34 in the Supporting Information onlineshow the linear and non-linear dose-response analyses
between total nut consumption and CVD outcomesThe summary RR (95CI) for a 28-gd increment was
087 (081ndash093) for CVD incidence 071 (061ndash084)for CVD mortality 075 (064ndash088) for CHD incidence
067 (052ndash087) for CHD mortality 106 (097ndash115) forstroke incidence 101 (088ndash118) for stroke mortality
105 (077ndash143) for hemorrhagic stroke and 106(086ndash131) for ischemic stroke
Total nut consumption and the risk of CVD inci-dence (Figure S27 in the Supporting Information on-
line) CVD mortality (Figure S28 in the SupportingInformation online) stroke mortality (Figure S32 in theSupporting Information online) and hemorrhagic
stroke (Figure S33 in the Supporting Information on-line) showed a non-linear association (Pnon-line-
aritylt001) The reduction in the risk of CVD incidencewas observed up to a consumption of 10 gd with no
further reduction with higher consumptions (Table S6in the Supporting Information online) For CVD mor-
tality and CHD mortality there was a steeper reductionin the risk at approximately 15ndash20 gd with no further
reduction with a higher consumption (Table S6 in theSupporting Information online) The reduction in the
risk of stroke mortality was observed up to a consump-tion of 5 gd with no significant reductions above this
amount (Table S6 in the Supporting Information
Nutrition ReviewsVR Vol 77(10)691ndash709 701
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
7 study comparisons (5 reports) for ischemic
stroke2122325253 2 study comparisons for AF2255 and2 study comparisons for HF2256
Four studies57ndash60 that reported the risk estimate oncontinuous scale instead of categories of nut consump-
tion were identified and therefore were not included inthe high vs low categories of consumption analyses and
the dose-response analyses
Study characteristics
The characteristics of the included studies are presentedin Table 221222629ndash324446ndash60 Publication date ranged
from 1992 to 2018 Six of the studies originated fromEurope 13 from America 1 from Asia 1 from
Australia 1 from both China and the USA and 1 fromboth China and Germany The duration of follow-up
ranged from 43 to 287 years All the studies assessed
nut intake via a food frequency questionnaire The vast
majority of studies (709) were of high quality accord-ing to the NOS scale Tables S1ndashS5 in the Supporting
Information online describe the characteristics of theincluded studies by type of nut consumption
High vs low categories of consumption analyses
Nuts and cardiovascular disease incidence Three cohort
comparisons involving 210 836 participants and 14 136cases analyzed the association between nut consump-
tion and the risk of CVD incidence The summary RR(95CI) for high vs low categories of nut consumption
was 085 (080ndash091) with no evidence of interstudy het-erogeneity (I2frac14 0 Pheterogeneity 081) (Figure 2 and
Figure S1 in the Supporting Information online)Regarding specific types of nuts consumption of tree
nuts ([RR 085 95CI 079ndash091] I2 0
Figure 1 Flow diagram of the literature search and selection process
Nutrition ReviewsVR Vol 77(10)691ndash709 695
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Char
acte
rist
ics
ofth
ein
clud
edst
udie
sev
alua
ting
the
asso
ciat
ion
betw
een
nuts
and
risk
ofca
rdio
vasc
ular
dise
ase
outc
omes
Stud
yCo
untr
ySt
udy
nam
ePo
pula
tion
Nut
cons
umpt
ion
asse
ssm
ent
met
hod
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Fras
eret
al(1
992)
29U
SASe
vent
h-da
yAd
vent
ists
2647
3M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
51
serv
ings
wk
vslt
1se
rvin
gw
k
25
6N
onfa
talM
I13
4Ag
ency
7CH
Dm
orta
lity
463
Albe
rtet
al(2
002)
30U
SAPh
ysic
ians
rsquoH
ealth
Stud
y21
454
MSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gm
o
40ndash8
417
CHD
mor
talit
y56
6Ag
ency
7N
onfa
talM
I10
37
Blom
hoff
etal
(200
6)44
USA
Iow
aW
omen
rsquosH
ealth
Stud
y31
778
FSe
lf-ad
min
is-
tere
dFF
QN
uts
plus
pean
utbu
tter
7se
rvin
gsw
kvs 0
5se
rvin
gsw
k
55ndash6
915
CVD
mor
talit
y16
75Ag
ency
6CH
Dm
orta
lity
948
Djo
uss e
etal
(200
8)56
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2097
6M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gw
k
40ndash8
419
6H
eart
failu
re10
93N
R6
Net
tleto
net
al(2
008)
57U
SAAt
hero
scle
rosi
sRi
skin
Com
mun
ities
Stud
y
1415
3M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
erCo
ntin
uous
1se
rvin
gd
45ndash6
413
Hea
rtfa
ilure
1140
Agen
cy8
Djo
uss e
etal
(201
0)52
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2107
8M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
7se
rvin
gsw
kvs lt
1se
rvin
gw
k
40ndash8
421
1St
roke
1424
Agen
cy6
Isch
emic
stro
ke11
89H
emor
rhag
icst
roke
219
Bern
stei
net
al(2
012)
54U
SAN
urse
srsquoH
ealth
Stud
yI
8001
0F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
034
serv
ings
dvs
0se
rvin
gsd
30ndash5
526
Hem
orrh
agic
stro
ke13
83Ag
ency
6
Hea
lthPr
ofes
sion
als
Follo
w-U
pSt
udy
4315
0M
060
serv
ings
dvs
0se
rvin
gsd
40ndash7
522
Hem
orrh
agic
stro
ke82
9
Khaw
aja
etal
(201
2)55
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2105
4M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
6
serv
ings
wk
vs0
serv
ings
wk
40ndash8
424
Atria
lfib
rilla
tion
3317
Agen
cy7
Yaem
siri
etal
(201
2)58
USA
Wom
enrsquos
Hea
lthIn
itiat
ive
Obs
erva
tiona
lSt
udy
8702
5F
Self-
adm
inis
-te
red
FFQ
Not
spec
ified
Cont
inuo
us1
serv
ing
d50
ndash79
76
Isch
emic
stro
ke10
49Ag
ency
8
von
Rues
ten
etal
(201
3)59
Ger
man
yEP
IC-P
otsd
amst
udy
2353
1M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Pean
uts
wal
nuts
br
azil
nuts
Cont
inuo
us1
serv
ing
d35
ndash65
8CV
D36
3Ag
ency
8
(con
tinue
d)
696 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Gua
sch-
Ferr
eet
al(2
013)
46Sp
ain
PRED
IMED
stud
y72
16M
FIn
terv
iew
-adm
in-
iste
red
vali-
date
dse
mi-
quan
titat
ive
FFQ
Alm
onds
pea
-nu
tsh
azel
-nu
tsp
ista
-ch
ios
pine
nuts
gt3
serv
ings
wk
vs neve
r
55ndash8
04
8CV
Dm
orta
lity
81Ag
ency
-In
dust
ry9
Bona
ccio
etal
(201
5)48
Italy
Mol
i-san
istu
dy10
509
F88
77M
Valid
ated
sel
f-ad
min
iste
red
FFQ
Wal
nuts
haz
el-
nuts
alm
onds
pe
anut
s
Inta
kevs no
inta
ke
gt35
43
CVD
mor
talit
y10
4Ag
ency
6CH
Dm
orta
lity
39St
roke
mor
talit
y19
diG
iuse
ppe
etal
(201
5)32
Ger
man
yEP
IC-P
otsd
amst
udy
2599
7M
FVa
lidat
eds
elf-
adm
inis
tere
dFF
Q
Pean
uts
wal
nuts
br
azil
nuts
142
gd
vs 0g
d
F49
2M
52
58
3St
roke
288
Agen
cy8
Isch
emic
stro
ke23
5St
roke
mor
talit
y36
Gop
inat
het
al(2
015)
51Au
stra
liaBl
ueM
ount
ains
Eye
Stud
y13
12F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
490
to10
0g
dvs 0
to0
50g
d
49
15CV
Dm
orta
lity
258
Agen
cy7
IHD
mor
talit
y18
8St
roke
mor
talit
y10
1
1581
MCV
Dm
orta
lity
288
IHD
mor
talit
y24
2St
roke
mor
talit
y75
Har
ing
etal
(201
4)50
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1206
6M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Not
spec
ified
06
serv
ings
dvs
01
serv
ings
d45
ndash64
22CH
D11
47Ag
ency
9
Har
ing
etal
(201
5)53
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1160
1M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
er1
serv
ings
dvs 0
serv
ings
d
45ndash6
422
7St
roke
699
Agen
cy8
Hem
orrh
agic
stro
ke11
4
Isch
emic
stro
ke59
8H
shie
het
al(2
015)
47U
SAPh
ysic
ians
rsquoH
ealth
Stud
y20
742
MU
nval
idat
eds
elf-
adm
inis
tere
dFF
Q
Not
spec
ified
5
serv
ings
wk
vs lt1
serv
ing
wk
40ndash8
49
6CV
Dm
orta
lity
760
Agen
cy6
Stro
kem
orta
lity
143
Luu
etal
(201
5)31
USA
Sout
hern
Com
mun
ityCo
hort
Stud
y
7176
4M
FSe
mi-q
uant
ita-
tive
FFQ
Tota
lnut
san
dpe
anut
butt
er
184
5g
dvs lt
095
gd
40ndash7
95
4CV
Dm
orta
lity
1857
Agen
cy8
IHD
mor
talit
y79
3Is
chem
icst
roke
mor
talit
y12
1
Hem
orrh
agic
stro
kem
orta
lity
96
(con
tinue
d)
Nutrition ReviewsVR Vol 77(10)691ndash709 697
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Chin
aSh
angh
aiW
omen
rsquosH
ealth
Stud
yan
dSh
angh
aiM
enrsquos
Hea
lthSt
udy
134
265
MF
Pean
ut
254
gd
vs lt0
14gr
amd
ay
40ndash7
0an
d40
ndash74
122
and
65
CVD
mor
talit
y25
87IH
Dm
orta
lity
631
Isch
emic
stro
kem
orta
lity
588
Hem
orrh
agic
stro
kem
orta
lity
597
van
den
Bran
dtan
dSc
hout
en(2
015)
26
Net
herla
nds
Net
herla
nds
Coho
rtSt
udy
3202
MF
(sub
coho
rt)
Self-
adm
inis
-te
red
valid
ated
FFQ
Tree
nuts
pe
anut
s19
6g
dvs 0
gd
55ndash6
99
CVD
mor
talit
y29
85N
R7
IHD
mor
talit
y14
88St
roke
mor
talit
y56
5
Wan
get
al(2
016)
60Ch
ina
Linx
ian
NIT
coho
rt24
45M
FFF
QPe
anut
sch
est-
nuts
wal
nuts
Cont
inuo
us3
serv
ings
mo
40ndash6
926
Hea
rtdi
seas
em
orta
lity
355
Agen
cy6
Ger
man
ySt
roke
mor
talit
y45
2
Esla
mpa
rast
etal
(201
7)49
Iran
Gol
esta
nCo
hort
Stud
y28
257
FVa
lidat
eds
elf-
adm
inis
tere
dse
miq
uant
ita-
tive
FFQ
Pean
uts
tree
nuts
3
serv
ings
wk
vs neve
r
40ndash8
77
CVD
mor
talit
y91
1Ag
ency
720
855
M11
05
Gua
sch-
Ferr
eet
al(2
017)
21U
SAN
urse
srsquoH
ealth
Stud
yI
7636
4F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Pean
uts
othe
rnu
tsa
ndw
al-
nuts
(ifav
aila
ble)
5
serv
ings
wk
vs neve
rora
lmos
tne
ver
30ndash5
528
7CV
D67
27Ag
ency
7CV
Dm
orta
lity
1770
CHD
3552
CHD
mor
talit
y99
6St
roke
3322
Stro
kem
orta
lity
773
Isch
emic
stro
ke16
35N
urse
srsquoH
ealth
Stud
yII
9294
6F
25ndash4
221
5CV
D19
15CV
Dm
orta
lity
82CH
D67
0CH
Dm
orta
lity
46St
roke
1262
Stro
kem
orta
lity
36
Isch
emic
stro
ke22
0H
ealth
Prof
essi
onal
sFo
llow
-Up
Stud
y
4152
6M
40ndash7
522
5CV
D54
94CV
Dm
orta
lity
2599
CHD
4168
CHD
mor
talit
y19
21St
roke
1326
Stro
kem
orta
lity
367
Isch
emic
stro
ke74
2
(con
tinue
d)
698 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Pheterogeneity 070) peanuts ([RR 087 95CI
081ndash093] I2 0 Pheterogeneity 067) and walnuts ([RR081 95CI 071ndash092] I2 73 Pheterogeneity 003) was
associated with a lower risk of CVD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S2ndashS4 in the SupportingInformation online) No association was reported be-tween peanut butter consumption and the risk of CVD
incidence ([RR 098 95CI 093ndash103] I2 89Pheterogeneitylt001) (Table 3 and Figure S5 in the
Supporting Information online)
Nuts and cardiovascular disease mortality Fifteen co-hort comparisons analyzed the association between
nut consumption and CVD mortality including413 727 participants and 14 475 cases The summary
RR (95CI) for high vs low categories of nut con-sumption was 077 (072ndash082) with no evidence of
interstudy heterogeneity (I2 3 Pheterogeneity 042)(Figure 2 and Figure S6 in the Supporting Information
online) Regarding specific types of nuts only peanutshave been studied in relation to CVD mortality show-
ing a summary RR of 077 (95CI 070ndash085) for highvs low categories of consumption with no evidence of
interstudy heterogeneity (I2 0 Pheterogeneity 081)(Table 3 and Figure S7 in the Supporting Information
online)
Nuts and coronary heart disease incidence Seven cohortcomparisons analyzed the association between nut con-
sumption and CHD incidence including 275 812 par-ticipants and 12 654 cases The summary RR (95CI)
for high vs low categories of nut consumption was 082(069ndash096) with evidence of substantial interstudy het-
erogeneity (I2 74 Pheterogeneitylt001) (Figure 2 andFigure S8 in the Supporting Information online)
Regarding specific types of nuts consumption of treenuts ([RR 077 95CI 070ndash084] I2 61
Pheterogeneity 008) peanuts ([RR 085 95CI 079ndash092] I2 0 Pheterogeneity 055) and walnuts ([RR0791 95CI 066ndash094] I2 69 Pheterogeneity 004) was
associated with a lower risk of CHD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S9ndashS11 in the SupportingInformation online) No association was reported be-
tween peanut butter consumption and the risk of CHDincidence ([RR 100 95CI 094ndash107] I2 43
Pheterogeneity 017) (Table 3 and Figure S12 in theSupporting Information online)
Nuts and coronary heart disease mortality Thirteen co-
hort comparisons analyzed the association between nutconsumption and CHD mortality including 396 041
participants and 7877 cases The summary RR (95CI)Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Lars
son
etal
(201
8)22
Swed
enCo
hort
ofSw
edis
hM
en32
911
MFF
QN
uts
(not
incl
ud-
ing
coco
nuto
rch
estn
uts)
3
serv
ings
wk
vs neve
r
45ndash8
317
MI
4983
Agen
cy8
MIm
orta
lity
917
Hea
rtfa
ilure
3160
Atria
lfib
rilla
tion
7550
Isch
emic
stro
ke37
82In
trac
ereb
ral
hem
orrh
age
543
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CVD
car
diov
ascu
lard
isea
seE
PIC
Euro
pean
Pros
pect
ive
Inve
stig
atio
nin
toCa
ncer
and
Nut
ritio
nF
fem
ales
FFQ
foo
dfr
eque
ncy
ques
tionn
aire
IH
Di
sche
mic
hear
tdis
ease
Mm
ales
MI
myo
card
iali
nfar
ctio
nN
ITN
utrit
ion
Inte
rven
tion
Tria
lsN
OS
New
cast
le-O
ttaw
asc
ale
NR
Not
repo
rted
PRE
DIM
EDP
reve
nci o
nco
nD
ieta
MED
iterr
anea
Nutrition ReviewsVR Vol 77(10)691ndash709 699
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
for high vs low categories of nut consumption was 076(067ndash086) with evidence of moderate interstudy het-
erogeneity (I2 46 Pheterogeneity 004) (Figure 2 andFigure S13 in the Supporting Information online)
Regarding specific types of nuts peanut consumption
was inversely associated with the risk of CHD mortalityafter comparing high vs low categories of consumption
([RR 075 95CI 064ndash088] I2 0 Pheterogeneity 046)(Table 3 and Figure S14 in the Supporting Information
online)
Table 3 Summary RR of cardiovascular disease outcomes by different types of nut consumption (comparing highest vslowest categories)
No of cohorts No of participants No of cases RR (95CI) P for heterogeneity I2 ()
Tree nutsCVD 3 210 836 14 136 085 (079 091) 070 0CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 100 (089 111) 093 0Stroke mortality 3 118 962 1851 093 (077 113) 044 0CHD 3 210 836 8390 077 (070 084) 008 61CHD mortality ndash ndash ndash ndash ndash ndash
PeanutsCVD 3 210 836 14 136 087 (081 093) 067 0CVD mortality 2 134 265 5572 077 (070 085) 081 0Stroke 3 210 836 5910 090 (081 099) 032 13Stroke mortality 4 253 227 3036 083 (073 095) 007 57CHD 3 210 836 8390 085 (079 092) 055 0CHD mortality 2 134 265 2119 075 (064 088) 046 0
WalnutsCVD 3 144 021 5255 081 (071 092) 003 73CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 144 021 5910 085 (071 102) 019 39Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 144 021 2685 079 (066 094) 004 69CHD mortality ndash ndash ndash ndash ndash ndash
Peanut butterCVD 3 210 836 14 136 098 (093 103) lt001 89CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 094 (087 102) lt001 86Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 210 836 8390 100 (094 107) 017 43CHD mortality ndash ndash ndash ndash ndash ndash
Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovascular disease RR risk ratio
Figure 2 Summary plots of effect estimates from prospective cohort studies evaluating the association between nut consumptionand the risk of different cardiovascular outcomes Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovasculardisease RR risk ratio
700 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Nuts and stroke incidence Seven cohort comparisonsanalyzed the association between nut consumption and
stroke incidence including 302 888 participants and12 646 cases The summary RR (95CI) for high vs low
categories of nut consumption was 100 (092ndash109)with no evidence of interstudy heterogeneity (I2 0
Pheterogeneity 097) (Figure 2 and Figure S15 in theSupporting Information online) Regarding specific
types of nuts peanut consumption was associated witha lower risk of stroke incidence after comparing highest
vs lowest categories of consumption ([RR 090 95CI081ndash099] I2 13 Pheterogeneity 032) (Table 3 andFigure S16 in the Supporting Information online) No
association was observed between tree nut walnut andpeanut butter consumption and the risk of stroke inci-
dence (Table 3 and Figures S17ndashS19 in the SupportingInformation online)
Nuts and stroke mortality Twelve cohort comparisons
analyzed the association between nut consumption andstroke mortality including 351 618 participants and
2332 cases The summary RR (95CI) for high vs lowcategories of nut consumption was 083 (075ndash093)
with no evidence of interstudy heterogeneity (I2 0Pheterogeneity 045) (Figure 2 and Figure S20 in the
Supporting Information online) Regarding specifictypes of nuts peanut consumption was associated with
a lower risk of stroke mortality after comparing highestvs lowest categories of consumption ([RR 085 95CI
079ndash092] I2 0 Pheterogeneity 055) (Table 3 andFigure S21 in the Supporting Information online) No
association was reported between tree nut consumptionand the risk of stroke death (Table 3 and Figure S22 in
the Supporting Information online)
Nuts and hemorrhagic stroke Five cohort comparisonsanalyzed the association between nut consumption and
hemorrhagic stroke incidence involving 188 750 partic-ipants and 3088 cases The summary RR (95CI) for
high vs low categories of nut consumption was 102(077ndash134) with no evidence of interstudy heterogene-ity (I2 15 Pheterogeneity 032) (Figure 2 and Figure S23
in the Supporting Information online) No studyanalyzed the association between different types of nuts
and the risk of hemorrhagic stroke
Nuts and ischemic stroke Seven cohort comparisonsanalyzed the association between nut consumption and
ischemic stroke incidence involving 302 423 partici-pants and 8401 cases The summary RR (95CI) for
high vs low categories of nut consumption was 099(089ndash110) with no evidence of interstudy heterogene-
ity (I2 0 Pheterogeneity 062) (Figure 2 and Figure S24 in
the Supporting Information online) No study analyzed
the association between different types of nuts and therisk of hemorrhagic stroke
Nuts and atrial fibrillation Two cohort comparisonsanalyzed the association between nut consumption and
AF involving 53 965 participants and 10 867 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 085 (073ndash099) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 052)
(Figure 2 and Figure S25 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of AF
Nuts and heart failure Two cohort comparisonsanalyzed the association between nut consumption and
HF involving 53 887 participants and 4253 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 100 (086ndash116) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 085)
(Figure 2 and Figure S26 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of HF
Dose-response analyses
Figures S27ndashS34 in the Supporting Information onlineshow the linear and non-linear dose-response analyses
between total nut consumption and CVD outcomesThe summary RR (95CI) for a 28-gd increment was
087 (081ndash093) for CVD incidence 071 (061ndash084)for CVD mortality 075 (064ndash088) for CHD incidence
067 (052ndash087) for CHD mortality 106 (097ndash115) forstroke incidence 101 (088ndash118) for stroke mortality
105 (077ndash143) for hemorrhagic stroke and 106(086ndash131) for ischemic stroke
Total nut consumption and the risk of CVD inci-dence (Figure S27 in the Supporting Information on-
line) CVD mortality (Figure S28 in the SupportingInformation online) stroke mortality (Figure S32 in theSupporting Information online) and hemorrhagic
stroke (Figure S33 in the Supporting Information on-line) showed a non-linear association (Pnon-line-
aritylt001) The reduction in the risk of CVD incidencewas observed up to a consumption of 10 gd with no
further reduction with higher consumptions (Table S6in the Supporting Information online) For CVD mor-
tality and CHD mortality there was a steeper reductionin the risk at approximately 15ndash20 gd with no further
reduction with a higher consumption (Table S6 in theSupporting Information online) The reduction in the
risk of stroke mortality was observed up to a consump-tion of 5 gd with no significant reductions above this
amount (Table S6 in the Supporting Information
Nutrition ReviewsVR Vol 77(10)691ndash709 701
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Char
acte
rist
ics
ofth
ein
clud
edst
udie
sev
alua
ting
the
asso
ciat
ion
betw
een
nuts
and
risk
ofca
rdio
vasc
ular
dise
ase
outc
omes
Stud
yCo
untr
ySt
udy
nam
ePo
pula
tion
Nut
cons
umpt
ion
asse
ssm
ent
met
hod
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Fras
eret
al(1
992)
29U
SASe
vent
h-da
yAd
vent
ists
2647
3M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
51
serv
ings
wk
vslt
1se
rvin
gw
k
25
6N
onfa
talM
I13
4Ag
ency
7CH
Dm
orta
lity
463
Albe
rtet
al(2
002)
30U
SAPh
ysic
ians
rsquoH
ealth
Stud
y21
454
MSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gm
o
40ndash8
417
CHD
mor
talit
y56
6Ag
ency
7N
onfa
talM
I10
37
Blom
hoff
etal
(200
6)44
USA
Iow
aW
omen
rsquosH
ealth
Stud
y31
778
FSe
lf-ad
min
is-
tere
dFF
QN
uts
plus
pean
utbu
tter
7se
rvin
gsw
kvs 0
5se
rvin
gsw
k
55ndash6
915
CVD
mor
talit
y16
75Ag
ency
6CH
Dm
orta
lity
948
Djo
uss e
etal
(200
8)56
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2097
6M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
2
serv
ings
wk
vslt
1se
rvin
gw
k
40ndash8
419
6H
eart
failu
re10
93N
R6
Net
tleto
net
al(2
008)
57U
SAAt
hero
scle
rosi
sRi
skin
Com
mun
ities
Stud
y
1415
3M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
erCo
ntin
uous
1se
rvin
gd
45ndash6
413
Hea
rtfa
ilure
1140
Agen
cy8
Djo
uss e
etal
(201
0)52
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2107
8M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
7se
rvin
gsw
kvs lt
1se
rvin
gw
k
40ndash8
421
1St
roke
1424
Agen
cy6
Isch
emic
stro
ke11
89H
emor
rhag
icst
roke
219
Bern
stei
net
al(2
012)
54U
SAN
urse
srsquoH
ealth
Stud
yI
8001
0F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
034
serv
ings
dvs
0se
rvin
gsd
30ndash5
526
Hem
orrh
agic
stro
ke13
83Ag
ency
6
Hea
lthPr
ofes
sion
als
Follo
w-U
pSt
udy
4315
0M
060
serv
ings
dvs
0se
rvin
gsd
40ndash7
522
Hem
orrh
agic
stro
ke82
9
Khaw
aja
etal
(201
2)55
USA
Phys
icia
nsrsquo
Hea
lthSt
udy
2105
4M
Unv
alid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
6
serv
ings
wk
vs0
serv
ings
wk
40ndash8
424
Atria
lfib
rilla
tion
3317
Agen
cy7
Yaem
siri
etal
(201
2)58
USA
Wom
enrsquos
Hea
lthIn
itiat
ive
Obs
erva
tiona
lSt
udy
8702
5F
Self-
adm
inis
-te
red
FFQ
Not
spec
ified
Cont
inuo
us1
serv
ing
d50
ndash79
76
Isch
emic
stro
ke10
49Ag
ency
8
von
Rues
ten
etal
(201
3)59
Ger
man
yEP
IC-P
otsd
amst
udy
2353
1M
FSe
lf-ad
min
is-
tere
dse
mi-
quan
titat
ive
FFQ
Pean
uts
wal
nuts
br
azil
nuts
Cont
inuo
us1
serv
ing
d35
ndash65
8CV
D36
3Ag
ency
8
(con
tinue
d)
696 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Gua
sch-
Ferr
eet
al(2
013)
46Sp
ain
PRED
IMED
stud
y72
16M
FIn
terv
iew
-adm
in-
iste
red
vali-
date
dse
mi-
quan
titat
ive
FFQ
Alm
onds
pea
-nu
tsh
azel
-nu
tsp
ista
-ch
ios
pine
nuts
gt3
serv
ings
wk
vs neve
r
55ndash8
04
8CV
Dm
orta
lity
81Ag
ency
-In
dust
ry9
Bona
ccio
etal
(201
5)48
Italy
Mol
i-san
istu
dy10
509
F88
77M
Valid
ated
sel
f-ad
min
iste
red
FFQ
Wal
nuts
haz
el-
nuts
alm
onds
pe
anut
s
Inta
kevs no
inta
ke
gt35
43
CVD
mor
talit
y10
4Ag
ency
6CH
Dm
orta
lity
39St
roke
mor
talit
y19
diG
iuse
ppe
etal
(201
5)32
Ger
man
yEP
IC-P
otsd
amst
udy
2599
7M
FVa
lidat
eds
elf-
adm
inis
tere
dFF
Q
Pean
uts
wal
nuts
br
azil
nuts
142
gd
vs 0g
d
F49
2M
52
58
3St
roke
288
Agen
cy8
Isch
emic
stro
ke23
5St
roke
mor
talit
y36
Gop
inat
het
al(2
015)
51Au
stra
liaBl
ueM
ount
ains
Eye
Stud
y13
12F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
490
to10
0g
dvs 0
to0
50g
d
49
15CV
Dm
orta
lity
258
Agen
cy7
IHD
mor
talit
y18
8St
roke
mor
talit
y10
1
1581
MCV
Dm
orta
lity
288
IHD
mor
talit
y24
2St
roke
mor
talit
y75
Har
ing
etal
(201
4)50
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1206
6M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Not
spec
ified
06
serv
ings
dvs
01
serv
ings
d45
ndash64
22CH
D11
47Ag
ency
9
Har
ing
etal
(201
5)53
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1160
1M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
er1
serv
ings
dvs 0
serv
ings
d
45ndash6
422
7St
roke
699
Agen
cy8
Hem
orrh
agic
stro
ke11
4
Isch
emic
stro
ke59
8H
shie
het
al(2
015)
47U
SAPh
ysic
ians
rsquoH
ealth
Stud
y20
742
MU
nval
idat
eds
elf-
adm
inis
tere
dFF
Q
Not
spec
ified
5
serv
ings
wk
vs lt1
serv
ing
wk
40ndash8
49
6CV
Dm
orta
lity
760
Agen
cy6
Stro
kem
orta
lity
143
Luu
etal
(201
5)31
USA
Sout
hern
Com
mun
ityCo
hort
Stud
y
7176
4M
FSe
mi-q
uant
ita-
tive
FFQ
Tota
lnut
san
dpe
anut
butt
er
184
5g
dvs lt
095
gd
40ndash7
95
4CV
Dm
orta
lity
1857
Agen
cy8
IHD
mor
talit
y79
3Is
chem
icst
roke
mor
talit
y12
1
Hem
orrh
agic
stro
kem
orta
lity
96
(con
tinue
d)
Nutrition ReviewsVR Vol 77(10)691ndash709 697
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Chin
aSh
angh
aiW
omen
rsquosH
ealth
Stud
yan
dSh
angh
aiM
enrsquos
Hea
lthSt
udy
134
265
MF
Pean
ut
254
gd
vs lt0
14gr
amd
ay
40ndash7
0an
d40
ndash74
122
and
65
CVD
mor
talit
y25
87IH
Dm
orta
lity
631
Isch
emic
stro
kem
orta
lity
588
Hem
orrh
agic
stro
kem
orta
lity
597
van
den
Bran
dtan
dSc
hout
en(2
015)
26
Net
herla
nds
Net
herla
nds
Coho
rtSt
udy
3202
MF
(sub
coho
rt)
Self-
adm
inis
-te
red
valid
ated
FFQ
Tree
nuts
pe
anut
s19
6g
dvs 0
gd
55ndash6
99
CVD
mor
talit
y29
85N
R7
IHD
mor
talit
y14
88St
roke
mor
talit
y56
5
Wan
get
al(2
016)
60Ch
ina
Linx
ian
NIT
coho
rt24
45M
FFF
QPe
anut
sch
est-
nuts
wal
nuts
Cont
inuo
us3
serv
ings
mo
40ndash6
926
Hea
rtdi
seas
em
orta
lity
355
Agen
cy6
Ger
man
ySt
roke
mor
talit
y45
2
Esla
mpa
rast
etal
(201
7)49
Iran
Gol
esta
nCo
hort
Stud
y28
257
FVa
lidat
eds
elf-
adm
inis
tere
dse
miq
uant
ita-
tive
FFQ
Pean
uts
tree
nuts
3
serv
ings
wk
vs neve
r
40ndash8
77
CVD
mor
talit
y91
1Ag
ency
720
855
M11
05
Gua
sch-
Ferr
eet
al(2
017)
21U
SAN
urse
srsquoH
ealth
Stud
yI
7636
4F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Pean
uts
othe
rnu
tsa
ndw
al-
nuts
(ifav
aila
ble)
5
serv
ings
wk
vs neve
rora
lmos
tne
ver
30ndash5
528
7CV
D67
27Ag
ency
7CV
Dm
orta
lity
1770
CHD
3552
CHD
mor
talit
y99
6St
roke
3322
Stro
kem
orta
lity
773
Isch
emic
stro
ke16
35N
urse
srsquoH
ealth
Stud
yII
9294
6F
25ndash4
221
5CV
D19
15CV
Dm
orta
lity
82CH
D67
0CH
Dm
orta
lity
46St
roke
1262
Stro
kem
orta
lity
36
Isch
emic
stro
ke22
0H
ealth
Prof
essi
onal
sFo
llow
-Up
Stud
y
4152
6M
40ndash7
522
5CV
D54
94CV
Dm
orta
lity
2599
CHD
4168
CHD
mor
talit
y19
21St
roke
1326
Stro
kem
orta
lity
367
Isch
emic
stro
ke74
2
(con
tinue
d)
698 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Pheterogeneity 070) peanuts ([RR 087 95CI
081ndash093] I2 0 Pheterogeneity 067) and walnuts ([RR081 95CI 071ndash092] I2 73 Pheterogeneity 003) was
associated with a lower risk of CVD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S2ndashS4 in the SupportingInformation online) No association was reported be-tween peanut butter consumption and the risk of CVD
incidence ([RR 098 95CI 093ndash103] I2 89Pheterogeneitylt001) (Table 3 and Figure S5 in the
Supporting Information online)
Nuts and cardiovascular disease mortality Fifteen co-hort comparisons analyzed the association between
nut consumption and CVD mortality including413 727 participants and 14 475 cases The summary
RR (95CI) for high vs low categories of nut con-sumption was 077 (072ndash082) with no evidence of
interstudy heterogeneity (I2 3 Pheterogeneity 042)(Figure 2 and Figure S6 in the Supporting Information
online) Regarding specific types of nuts only peanutshave been studied in relation to CVD mortality show-
ing a summary RR of 077 (95CI 070ndash085) for highvs low categories of consumption with no evidence of
interstudy heterogeneity (I2 0 Pheterogeneity 081)(Table 3 and Figure S7 in the Supporting Information
online)
Nuts and coronary heart disease incidence Seven cohortcomparisons analyzed the association between nut con-
sumption and CHD incidence including 275 812 par-ticipants and 12 654 cases The summary RR (95CI)
for high vs low categories of nut consumption was 082(069ndash096) with evidence of substantial interstudy het-
erogeneity (I2 74 Pheterogeneitylt001) (Figure 2 andFigure S8 in the Supporting Information online)
Regarding specific types of nuts consumption of treenuts ([RR 077 95CI 070ndash084] I2 61
Pheterogeneity 008) peanuts ([RR 085 95CI 079ndash092] I2 0 Pheterogeneity 055) and walnuts ([RR0791 95CI 066ndash094] I2 69 Pheterogeneity 004) was
associated with a lower risk of CHD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S9ndashS11 in the SupportingInformation online) No association was reported be-
tween peanut butter consumption and the risk of CHDincidence ([RR 100 95CI 094ndash107] I2 43
Pheterogeneity 017) (Table 3 and Figure S12 in theSupporting Information online)
Nuts and coronary heart disease mortality Thirteen co-
hort comparisons analyzed the association between nutconsumption and CHD mortality including 396 041
participants and 7877 cases The summary RR (95CI)Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Lars
son
etal
(201
8)22
Swed
enCo
hort
ofSw
edis
hM
en32
911
MFF
QN
uts
(not
incl
ud-
ing
coco
nuto
rch
estn
uts)
3
serv
ings
wk
vs neve
r
45ndash8
317
MI
4983
Agen
cy8
MIm
orta
lity
917
Hea
rtfa
ilure
3160
Atria
lfib
rilla
tion
7550
Isch
emic
stro
ke37
82In
trac
ereb
ral
hem
orrh
age
543
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CVD
car
diov
ascu
lard
isea
seE
PIC
Euro
pean
Pros
pect
ive
Inve
stig
atio
nin
toCa
ncer
and
Nut
ritio
nF
fem
ales
FFQ
foo
dfr
eque
ncy
ques
tionn
aire
IH
Di
sche
mic
hear
tdis
ease
Mm
ales
MI
myo
card
iali
nfar
ctio
nN
ITN
utrit
ion
Inte
rven
tion
Tria
lsN
OS
New
cast
le-O
ttaw
asc
ale
NR
Not
repo
rted
PRE
DIM
EDP
reve
nci o
nco
nD
ieta
MED
iterr
anea
Nutrition ReviewsVR Vol 77(10)691ndash709 699
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
for high vs low categories of nut consumption was 076(067ndash086) with evidence of moderate interstudy het-
erogeneity (I2 46 Pheterogeneity 004) (Figure 2 andFigure S13 in the Supporting Information online)
Regarding specific types of nuts peanut consumption
was inversely associated with the risk of CHD mortalityafter comparing high vs low categories of consumption
([RR 075 95CI 064ndash088] I2 0 Pheterogeneity 046)(Table 3 and Figure S14 in the Supporting Information
online)
Table 3 Summary RR of cardiovascular disease outcomes by different types of nut consumption (comparing highest vslowest categories)
No of cohorts No of participants No of cases RR (95CI) P for heterogeneity I2 ()
Tree nutsCVD 3 210 836 14 136 085 (079 091) 070 0CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 100 (089 111) 093 0Stroke mortality 3 118 962 1851 093 (077 113) 044 0CHD 3 210 836 8390 077 (070 084) 008 61CHD mortality ndash ndash ndash ndash ndash ndash
PeanutsCVD 3 210 836 14 136 087 (081 093) 067 0CVD mortality 2 134 265 5572 077 (070 085) 081 0Stroke 3 210 836 5910 090 (081 099) 032 13Stroke mortality 4 253 227 3036 083 (073 095) 007 57CHD 3 210 836 8390 085 (079 092) 055 0CHD mortality 2 134 265 2119 075 (064 088) 046 0
WalnutsCVD 3 144 021 5255 081 (071 092) 003 73CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 144 021 5910 085 (071 102) 019 39Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 144 021 2685 079 (066 094) 004 69CHD mortality ndash ndash ndash ndash ndash ndash
Peanut butterCVD 3 210 836 14 136 098 (093 103) lt001 89CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 094 (087 102) lt001 86Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 210 836 8390 100 (094 107) 017 43CHD mortality ndash ndash ndash ndash ndash ndash
Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovascular disease RR risk ratio
Figure 2 Summary plots of effect estimates from prospective cohort studies evaluating the association between nut consumptionand the risk of different cardiovascular outcomes Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovasculardisease RR risk ratio
700 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Nuts and stroke incidence Seven cohort comparisonsanalyzed the association between nut consumption and
stroke incidence including 302 888 participants and12 646 cases The summary RR (95CI) for high vs low
categories of nut consumption was 100 (092ndash109)with no evidence of interstudy heterogeneity (I2 0
Pheterogeneity 097) (Figure 2 and Figure S15 in theSupporting Information online) Regarding specific
types of nuts peanut consumption was associated witha lower risk of stroke incidence after comparing highest
vs lowest categories of consumption ([RR 090 95CI081ndash099] I2 13 Pheterogeneity 032) (Table 3 andFigure S16 in the Supporting Information online) No
association was observed between tree nut walnut andpeanut butter consumption and the risk of stroke inci-
dence (Table 3 and Figures S17ndashS19 in the SupportingInformation online)
Nuts and stroke mortality Twelve cohort comparisons
analyzed the association between nut consumption andstroke mortality including 351 618 participants and
2332 cases The summary RR (95CI) for high vs lowcategories of nut consumption was 083 (075ndash093)
with no evidence of interstudy heterogeneity (I2 0Pheterogeneity 045) (Figure 2 and Figure S20 in the
Supporting Information online) Regarding specifictypes of nuts peanut consumption was associated with
a lower risk of stroke mortality after comparing highestvs lowest categories of consumption ([RR 085 95CI
079ndash092] I2 0 Pheterogeneity 055) (Table 3 andFigure S21 in the Supporting Information online) No
association was reported between tree nut consumptionand the risk of stroke death (Table 3 and Figure S22 in
the Supporting Information online)
Nuts and hemorrhagic stroke Five cohort comparisonsanalyzed the association between nut consumption and
hemorrhagic stroke incidence involving 188 750 partic-ipants and 3088 cases The summary RR (95CI) for
high vs low categories of nut consumption was 102(077ndash134) with no evidence of interstudy heterogene-ity (I2 15 Pheterogeneity 032) (Figure 2 and Figure S23
in the Supporting Information online) No studyanalyzed the association between different types of nuts
and the risk of hemorrhagic stroke
Nuts and ischemic stroke Seven cohort comparisonsanalyzed the association between nut consumption and
ischemic stroke incidence involving 302 423 partici-pants and 8401 cases The summary RR (95CI) for
high vs low categories of nut consumption was 099(089ndash110) with no evidence of interstudy heterogene-
ity (I2 0 Pheterogeneity 062) (Figure 2 and Figure S24 in
the Supporting Information online) No study analyzed
the association between different types of nuts and therisk of hemorrhagic stroke
Nuts and atrial fibrillation Two cohort comparisonsanalyzed the association between nut consumption and
AF involving 53 965 participants and 10 867 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 085 (073ndash099) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 052)
(Figure 2 and Figure S25 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of AF
Nuts and heart failure Two cohort comparisonsanalyzed the association between nut consumption and
HF involving 53 887 participants and 4253 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 100 (086ndash116) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 085)
(Figure 2 and Figure S26 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of HF
Dose-response analyses
Figures S27ndashS34 in the Supporting Information onlineshow the linear and non-linear dose-response analyses
between total nut consumption and CVD outcomesThe summary RR (95CI) for a 28-gd increment was
087 (081ndash093) for CVD incidence 071 (061ndash084)for CVD mortality 075 (064ndash088) for CHD incidence
067 (052ndash087) for CHD mortality 106 (097ndash115) forstroke incidence 101 (088ndash118) for stroke mortality
105 (077ndash143) for hemorrhagic stroke and 106(086ndash131) for ischemic stroke
Total nut consumption and the risk of CVD inci-dence (Figure S27 in the Supporting Information on-
line) CVD mortality (Figure S28 in the SupportingInformation online) stroke mortality (Figure S32 in theSupporting Information online) and hemorrhagic
stroke (Figure S33 in the Supporting Information on-line) showed a non-linear association (Pnon-line-
aritylt001) The reduction in the risk of CVD incidencewas observed up to a consumption of 10 gd with no
further reduction with higher consumptions (Table S6in the Supporting Information online) For CVD mor-
tality and CHD mortality there was a steeper reductionin the risk at approximately 15ndash20 gd with no further
reduction with a higher consumption (Table S6 in theSupporting Information online) The reduction in the
risk of stroke mortality was observed up to a consump-tion of 5 gd with no significant reductions above this
amount (Table S6 in the Supporting Information
Nutrition ReviewsVR Vol 77(10)691ndash709 701
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Gua
sch-
Ferr
eet
al(2
013)
46Sp
ain
PRED
IMED
stud
y72
16M
FIn
terv
iew
-adm
in-
iste
red
vali-
date
dse
mi-
quan
titat
ive
FFQ
Alm
onds
pea
-nu
tsh
azel
-nu
tsp
ista
-ch
ios
pine
nuts
gt3
serv
ings
wk
vs neve
r
55ndash8
04
8CV
Dm
orta
lity
81Ag
ency
-In
dust
ry9
Bona
ccio
etal
(201
5)48
Italy
Mol
i-san
istu
dy10
509
F88
77M
Valid
ated
sel
f-ad
min
iste
red
FFQ
Wal
nuts
haz
el-
nuts
alm
onds
pe
anut
s
Inta
kevs no
inta
ke
gt35
43
CVD
mor
talit
y10
4Ag
ency
6CH
Dm
orta
lity
39St
roke
mor
talit
y19
diG
iuse
ppe
etal
(201
5)32
Ger
man
yEP
IC-P
otsd
amst
udy
2599
7M
FVa
lidat
eds
elf-
adm
inis
tere
dFF
Q
Pean
uts
wal
nuts
br
azil
nuts
142
gd
vs 0g
d
F49
2M
52
58
3St
roke
288
Agen
cy8
Isch
emic
stro
ke23
5St
roke
mor
talit
y36
Gop
inat
het
al(2
015)
51Au
stra
liaBl
ueM
ount
ains
Eye
Stud
y13
12F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Not
spec
ified
490
to10
0g
dvs 0
to0
50g
d
49
15CV
Dm
orta
lity
258
Agen
cy7
IHD
mor
talit
y18
8St
roke
mor
talit
y10
1
1581
MCV
Dm
orta
lity
288
IHD
mor
talit
y24
2St
roke
mor
talit
y75
Har
ing
etal
(201
4)50
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1206
6M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Not
spec
ified
06
serv
ings
dvs
01
serv
ings
d45
ndash64
22CH
D11
47Ag
ency
9
Har
ing
etal
(201
5)53
USA
Athe
rosc
lero
sis
Risk
inCo
mm
uniti
esSt
udy
1160
1M
FIn
terv
iew
-adm
in-
iste
red
FFQ
Nut
spl
uspe
anut
butt
er1
serv
ings
dvs 0
serv
ings
d
45ndash6
422
7St
roke
699
Agen
cy8
Hem
orrh
agic
stro
ke11
4
Isch
emic
stro
ke59
8H
shie
het
al(2
015)
47U
SAPh
ysic
ians
rsquoH
ealth
Stud
y20
742
MU
nval
idat
eds
elf-
adm
inis
tere
dFF
Q
Not
spec
ified
5
serv
ings
wk
vs lt1
serv
ing
wk
40ndash8
49
6CV
Dm
orta
lity
760
Agen
cy6
Stro
kem
orta
lity
143
Luu
etal
(201
5)31
USA
Sout
hern
Com
mun
ityCo
hort
Stud
y
7176
4M
FSe
mi-q
uant
ita-
tive
FFQ
Tota
lnut
san
dpe
anut
butt
er
184
5g
dvs lt
095
gd
40ndash7
95
4CV
Dm
orta
lity
1857
Agen
cy8
IHD
mor
talit
y79
3Is
chem
icst
roke
mor
talit
y12
1
Hem
orrh
agic
stro
kem
orta
lity
96
(con
tinue
d)
Nutrition ReviewsVR Vol 77(10)691ndash709 697
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Chin
aSh
angh
aiW
omen
rsquosH
ealth
Stud
yan
dSh
angh
aiM
enrsquos
Hea
lthSt
udy
134
265
MF
Pean
ut
254
gd
vs lt0
14gr
amd
ay
40ndash7
0an
d40
ndash74
122
and
65
CVD
mor
talit
y25
87IH
Dm
orta
lity
631
Isch
emic
stro
kem
orta
lity
588
Hem
orrh
agic
stro
kem
orta
lity
597
van
den
Bran
dtan
dSc
hout
en(2
015)
26
Net
herla
nds
Net
herla
nds
Coho
rtSt
udy
3202
MF
(sub
coho
rt)
Self-
adm
inis
-te
red
valid
ated
FFQ
Tree
nuts
pe
anut
s19
6g
dvs 0
gd
55ndash6
99
CVD
mor
talit
y29
85N
R7
IHD
mor
talit
y14
88St
roke
mor
talit
y56
5
Wan
get
al(2
016)
60Ch
ina
Linx
ian
NIT
coho
rt24
45M
FFF
QPe
anut
sch
est-
nuts
wal
nuts
Cont
inuo
us3
serv
ings
mo
40ndash6
926
Hea
rtdi
seas
em
orta
lity
355
Agen
cy6
Ger
man
ySt
roke
mor
talit
y45
2
Esla
mpa
rast
etal
(201
7)49
Iran
Gol
esta
nCo
hort
Stud
y28
257
FVa
lidat
eds
elf-
adm
inis
tere
dse
miq
uant
ita-
tive
FFQ
Pean
uts
tree
nuts
3
serv
ings
wk
vs neve
r
40ndash8
77
CVD
mor
talit
y91
1Ag
ency
720
855
M11
05
Gua
sch-
Ferr
eet
al(2
017)
21U
SAN
urse
srsquoH
ealth
Stud
yI
7636
4F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Pean
uts
othe
rnu
tsa
ndw
al-
nuts
(ifav
aila
ble)
5
serv
ings
wk
vs neve
rora
lmos
tne
ver
30ndash5
528
7CV
D67
27Ag
ency
7CV
Dm
orta
lity
1770
CHD
3552
CHD
mor
talit
y99
6St
roke
3322
Stro
kem
orta
lity
773
Isch
emic
stro
ke16
35N
urse
srsquoH
ealth
Stud
yII
9294
6F
25ndash4
221
5CV
D19
15CV
Dm
orta
lity
82CH
D67
0CH
Dm
orta
lity
46St
roke
1262
Stro
kem
orta
lity
36
Isch
emic
stro
ke22
0H
ealth
Prof
essi
onal
sFo
llow
-Up
Stud
y
4152
6M
40ndash7
522
5CV
D54
94CV
Dm
orta
lity
2599
CHD
4168
CHD
mor
talit
y19
21St
roke
1326
Stro
kem
orta
lity
367
Isch
emic
stro
ke74
2
(con
tinue
d)
698 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Pheterogeneity 070) peanuts ([RR 087 95CI
081ndash093] I2 0 Pheterogeneity 067) and walnuts ([RR081 95CI 071ndash092] I2 73 Pheterogeneity 003) was
associated with a lower risk of CVD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S2ndashS4 in the SupportingInformation online) No association was reported be-tween peanut butter consumption and the risk of CVD
incidence ([RR 098 95CI 093ndash103] I2 89Pheterogeneitylt001) (Table 3 and Figure S5 in the
Supporting Information online)
Nuts and cardiovascular disease mortality Fifteen co-hort comparisons analyzed the association between
nut consumption and CVD mortality including413 727 participants and 14 475 cases The summary
RR (95CI) for high vs low categories of nut con-sumption was 077 (072ndash082) with no evidence of
interstudy heterogeneity (I2 3 Pheterogeneity 042)(Figure 2 and Figure S6 in the Supporting Information
online) Regarding specific types of nuts only peanutshave been studied in relation to CVD mortality show-
ing a summary RR of 077 (95CI 070ndash085) for highvs low categories of consumption with no evidence of
interstudy heterogeneity (I2 0 Pheterogeneity 081)(Table 3 and Figure S7 in the Supporting Information
online)
Nuts and coronary heart disease incidence Seven cohortcomparisons analyzed the association between nut con-
sumption and CHD incidence including 275 812 par-ticipants and 12 654 cases The summary RR (95CI)
for high vs low categories of nut consumption was 082(069ndash096) with evidence of substantial interstudy het-
erogeneity (I2 74 Pheterogeneitylt001) (Figure 2 andFigure S8 in the Supporting Information online)
Regarding specific types of nuts consumption of treenuts ([RR 077 95CI 070ndash084] I2 61
Pheterogeneity 008) peanuts ([RR 085 95CI 079ndash092] I2 0 Pheterogeneity 055) and walnuts ([RR0791 95CI 066ndash094] I2 69 Pheterogeneity 004) was
associated with a lower risk of CHD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S9ndashS11 in the SupportingInformation online) No association was reported be-
tween peanut butter consumption and the risk of CHDincidence ([RR 100 95CI 094ndash107] I2 43
Pheterogeneity 017) (Table 3 and Figure S12 in theSupporting Information online)
Nuts and coronary heart disease mortality Thirteen co-
hort comparisons analyzed the association between nutconsumption and CHD mortality including 396 041
participants and 7877 cases The summary RR (95CI)Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Lars
son
etal
(201
8)22
Swed
enCo
hort
ofSw
edis
hM
en32
911
MFF
QN
uts
(not
incl
ud-
ing
coco
nuto
rch
estn
uts)
3
serv
ings
wk
vs neve
r
45ndash8
317
MI
4983
Agen
cy8
MIm
orta
lity
917
Hea
rtfa
ilure
3160
Atria
lfib
rilla
tion
7550
Isch
emic
stro
ke37
82In
trac
ereb
ral
hem
orrh
age
543
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CVD
car
diov
ascu
lard
isea
seE
PIC
Euro
pean
Pros
pect
ive
Inve
stig
atio
nin
toCa
ncer
and
Nut
ritio
nF
fem
ales
FFQ
foo
dfr
eque
ncy
ques
tionn
aire
IH
Di
sche
mic
hear
tdis
ease
Mm
ales
MI
myo
card
iali
nfar
ctio
nN
ITN
utrit
ion
Inte
rven
tion
Tria
lsN
OS
New
cast
le-O
ttaw
asc
ale
NR
Not
repo
rted
PRE
DIM
EDP
reve
nci o
nco
nD
ieta
MED
iterr
anea
Nutrition ReviewsVR Vol 77(10)691ndash709 699
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
for high vs low categories of nut consumption was 076(067ndash086) with evidence of moderate interstudy het-
erogeneity (I2 46 Pheterogeneity 004) (Figure 2 andFigure S13 in the Supporting Information online)
Regarding specific types of nuts peanut consumption
was inversely associated with the risk of CHD mortalityafter comparing high vs low categories of consumption
([RR 075 95CI 064ndash088] I2 0 Pheterogeneity 046)(Table 3 and Figure S14 in the Supporting Information
online)
Table 3 Summary RR of cardiovascular disease outcomes by different types of nut consumption (comparing highest vslowest categories)
No of cohorts No of participants No of cases RR (95CI) P for heterogeneity I2 ()
Tree nutsCVD 3 210 836 14 136 085 (079 091) 070 0CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 100 (089 111) 093 0Stroke mortality 3 118 962 1851 093 (077 113) 044 0CHD 3 210 836 8390 077 (070 084) 008 61CHD mortality ndash ndash ndash ndash ndash ndash
PeanutsCVD 3 210 836 14 136 087 (081 093) 067 0CVD mortality 2 134 265 5572 077 (070 085) 081 0Stroke 3 210 836 5910 090 (081 099) 032 13Stroke mortality 4 253 227 3036 083 (073 095) 007 57CHD 3 210 836 8390 085 (079 092) 055 0CHD mortality 2 134 265 2119 075 (064 088) 046 0
WalnutsCVD 3 144 021 5255 081 (071 092) 003 73CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 144 021 5910 085 (071 102) 019 39Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 144 021 2685 079 (066 094) 004 69CHD mortality ndash ndash ndash ndash ndash ndash
Peanut butterCVD 3 210 836 14 136 098 (093 103) lt001 89CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 094 (087 102) lt001 86Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 210 836 8390 100 (094 107) 017 43CHD mortality ndash ndash ndash ndash ndash ndash
Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovascular disease RR risk ratio
Figure 2 Summary plots of effect estimates from prospective cohort studies evaluating the association between nut consumptionand the risk of different cardiovascular outcomes Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovasculardisease RR risk ratio
700 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Nuts and stroke incidence Seven cohort comparisonsanalyzed the association between nut consumption and
stroke incidence including 302 888 participants and12 646 cases The summary RR (95CI) for high vs low
categories of nut consumption was 100 (092ndash109)with no evidence of interstudy heterogeneity (I2 0
Pheterogeneity 097) (Figure 2 and Figure S15 in theSupporting Information online) Regarding specific
types of nuts peanut consumption was associated witha lower risk of stroke incidence after comparing highest
vs lowest categories of consumption ([RR 090 95CI081ndash099] I2 13 Pheterogeneity 032) (Table 3 andFigure S16 in the Supporting Information online) No
association was observed between tree nut walnut andpeanut butter consumption and the risk of stroke inci-
dence (Table 3 and Figures S17ndashS19 in the SupportingInformation online)
Nuts and stroke mortality Twelve cohort comparisons
analyzed the association between nut consumption andstroke mortality including 351 618 participants and
2332 cases The summary RR (95CI) for high vs lowcategories of nut consumption was 083 (075ndash093)
with no evidence of interstudy heterogeneity (I2 0Pheterogeneity 045) (Figure 2 and Figure S20 in the
Supporting Information online) Regarding specifictypes of nuts peanut consumption was associated with
a lower risk of stroke mortality after comparing highestvs lowest categories of consumption ([RR 085 95CI
079ndash092] I2 0 Pheterogeneity 055) (Table 3 andFigure S21 in the Supporting Information online) No
association was reported between tree nut consumptionand the risk of stroke death (Table 3 and Figure S22 in
the Supporting Information online)
Nuts and hemorrhagic stroke Five cohort comparisonsanalyzed the association between nut consumption and
hemorrhagic stroke incidence involving 188 750 partic-ipants and 3088 cases The summary RR (95CI) for
high vs low categories of nut consumption was 102(077ndash134) with no evidence of interstudy heterogene-ity (I2 15 Pheterogeneity 032) (Figure 2 and Figure S23
in the Supporting Information online) No studyanalyzed the association between different types of nuts
and the risk of hemorrhagic stroke
Nuts and ischemic stroke Seven cohort comparisonsanalyzed the association between nut consumption and
ischemic stroke incidence involving 302 423 partici-pants and 8401 cases The summary RR (95CI) for
high vs low categories of nut consumption was 099(089ndash110) with no evidence of interstudy heterogene-
ity (I2 0 Pheterogeneity 062) (Figure 2 and Figure S24 in
the Supporting Information online) No study analyzed
the association between different types of nuts and therisk of hemorrhagic stroke
Nuts and atrial fibrillation Two cohort comparisonsanalyzed the association between nut consumption and
AF involving 53 965 participants and 10 867 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 085 (073ndash099) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 052)
(Figure 2 and Figure S25 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of AF
Nuts and heart failure Two cohort comparisonsanalyzed the association between nut consumption and
HF involving 53 887 participants and 4253 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 100 (086ndash116) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 085)
(Figure 2 and Figure S26 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of HF
Dose-response analyses
Figures S27ndashS34 in the Supporting Information onlineshow the linear and non-linear dose-response analyses
between total nut consumption and CVD outcomesThe summary RR (95CI) for a 28-gd increment was
087 (081ndash093) for CVD incidence 071 (061ndash084)for CVD mortality 075 (064ndash088) for CHD incidence
067 (052ndash087) for CHD mortality 106 (097ndash115) forstroke incidence 101 (088ndash118) for stroke mortality
105 (077ndash143) for hemorrhagic stroke and 106(086ndash131) for ischemic stroke
Total nut consumption and the risk of CVD inci-dence (Figure S27 in the Supporting Information on-
line) CVD mortality (Figure S28 in the SupportingInformation online) stroke mortality (Figure S32 in theSupporting Information online) and hemorrhagic
stroke (Figure S33 in the Supporting Information on-line) showed a non-linear association (Pnon-line-
aritylt001) The reduction in the risk of CVD incidencewas observed up to a consumption of 10 gd with no
further reduction with higher consumptions (Table S6in the Supporting Information online) For CVD mor-
tality and CHD mortality there was a steeper reductionin the risk at approximately 15ndash20 gd with no further
reduction with a higher consumption (Table S6 in theSupporting Information online) The reduction in the
risk of stroke mortality was observed up to a consump-tion of 5 gd with no significant reductions above this
amount (Table S6 in the Supporting Information
Nutrition ReviewsVR Vol 77(10)691ndash709 701
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Chin
aSh
angh
aiW
omen
rsquosH
ealth
Stud
yan
dSh
angh
aiM
enrsquos
Hea
lthSt
udy
134
265
MF
Pean
ut
254
gd
vs lt0
14gr
amd
ay
40ndash7
0an
d40
ndash74
122
and
65
CVD
mor
talit
y25
87IH
Dm
orta
lity
631
Isch
emic
stro
kem
orta
lity
588
Hem
orrh
agic
stro
kem
orta
lity
597
van
den
Bran
dtan
dSc
hout
en(2
015)
26
Net
herla
nds
Net
herla
nds
Coho
rtSt
udy
3202
MF
(sub
coho
rt)
Self-
adm
inis
-te
red
valid
ated
FFQ
Tree
nuts
pe
anut
s19
6g
dvs 0
gd
55ndash6
99
CVD
mor
talit
y29
85N
R7
IHD
mor
talit
y14
88St
roke
mor
talit
y56
5
Wan
get
al(2
016)
60Ch
ina
Linx
ian
NIT
coho
rt24
45M
FFF
QPe
anut
sch
est-
nuts
wal
nuts
Cont
inuo
us3
serv
ings
mo
40ndash6
926
Hea
rtdi
seas
em
orta
lity
355
Agen
cy6
Ger
man
ySt
roke
mor
talit
y45
2
Esla
mpa
rast
etal
(201
7)49
Iran
Gol
esta
nCo
hort
Stud
y28
257
FVa
lidat
eds
elf-
adm
inis
tere
dse
miq
uant
ita-
tive
FFQ
Pean
uts
tree
nuts
3
serv
ings
wk
vs neve
r
40ndash8
77
CVD
mor
talit
y91
1Ag
ency
720
855
M11
05
Gua
sch-
Ferr
eet
al(2
017)
21U
SAN
urse
srsquoH
ealth
Stud
yI
7636
4F
Valid
ated
sel
f-ad
min
iste
red
FFQ
Pean
uts
othe
rnu
tsa
ndw
al-
nuts
(ifav
aila
ble)
5
serv
ings
wk
vs neve
rora
lmos
tne
ver
30ndash5
528
7CV
D67
27Ag
ency
7CV
Dm
orta
lity
1770
CHD
3552
CHD
mor
talit
y99
6St
roke
3322
Stro
kem
orta
lity
773
Isch
emic
stro
ke16
35N
urse
srsquoH
ealth
Stud
yII
9294
6F
25ndash4
221
5CV
D19
15CV
Dm
orta
lity
82CH
D67
0CH
Dm
orta
lity
46St
roke
1262
Stro
kem
orta
lity
36
Isch
emic
stro
ke22
0H
ealth
Prof
essi
onal
sFo
llow
-Up
Stud
y
4152
6M
40ndash7
522
5CV
D54
94CV
Dm
orta
lity
2599
CHD
4168
CHD
mor
talit
y19
21St
roke
1326
Stro
kem
orta
lity
367
Isch
emic
stro
ke74
2
(con
tinue
d)
698 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Pheterogeneity 070) peanuts ([RR 087 95CI
081ndash093] I2 0 Pheterogeneity 067) and walnuts ([RR081 95CI 071ndash092] I2 73 Pheterogeneity 003) was
associated with a lower risk of CVD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S2ndashS4 in the SupportingInformation online) No association was reported be-tween peanut butter consumption and the risk of CVD
incidence ([RR 098 95CI 093ndash103] I2 89Pheterogeneitylt001) (Table 3 and Figure S5 in the
Supporting Information online)
Nuts and cardiovascular disease mortality Fifteen co-hort comparisons analyzed the association between
nut consumption and CVD mortality including413 727 participants and 14 475 cases The summary
RR (95CI) for high vs low categories of nut con-sumption was 077 (072ndash082) with no evidence of
interstudy heterogeneity (I2 3 Pheterogeneity 042)(Figure 2 and Figure S6 in the Supporting Information
online) Regarding specific types of nuts only peanutshave been studied in relation to CVD mortality show-
ing a summary RR of 077 (95CI 070ndash085) for highvs low categories of consumption with no evidence of
interstudy heterogeneity (I2 0 Pheterogeneity 081)(Table 3 and Figure S7 in the Supporting Information
online)
Nuts and coronary heart disease incidence Seven cohortcomparisons analyzed the association between nut con-
sumption and CHD incidence including 275 812 par-ticipants and 12 654 cases The summary RR (95CI)
for high vs low categories of nut consumption was 082(069ndash096) with evidence of substantial interstudy het-
erogeneity (I2 74 Pheterogeneitylt001) (Figure 2 andFigure S8 in the Supporting Information online)
Regarding specific types of nuts consumption of treenuts ([RR 077 95CI 070ndash084] I2 61
Pheterogeneity 008) peanuts ([RR 085 95CI 079ndash092] I2 0 Pheterogeneity 055) and walnuts ([RR0791 95CI 066ndash094] I2 69 Pheterogeneity 004) was
associated with a lower risk of CHD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S9ndashS11 in the SupportingInformation online) No association was reported be-
tween peanut butter consumption and the risk of CHDincidence ([RR 100 95CI 094ndash107] I2 43
Pheterogeneity 017) (Table 3 and Figure S12 in theSupporting Information online)
Nuts and coronary heart disease mortality Thirteen co-
hort comparisons analyzed the association between nutconsumption and CHD mortality including 396 041
participants and 7877 cases The summary RR (95CI)Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Lars
son
etal
(201
8)22
Swed
enCo
hort
ofSw
edis
hM
en32
911
MFF
QN
uts
(not
incl
ud-
ing
coco
nuto
rch
estn
uts)
3
serv
ings
wk
vs neve
r
45ndash8
317
MI
4983
Agen
cy8
MIm
orta
lity
917
Hea
rtfa
ilure
3160
Atria
lfib
rilla
tion
7550
Isch
emic
stro
ke37
82In
trac
ereb
ral
hem
orrh
age
543
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CVD
car
diov
ascu
lard
isea
seE
PIC
Euro
pean
Pros
pect
ive
Inve
stig
atio
nin
toCa
ncer
and
Nut
ritio
nF
fem
ales
FFQ
foo
dfr
eque
ncy
ques
tionn
aire
IH
Di
sche
mic
hear
tdis
ease
Mm
ales
MI
myo
card
iali
nfar
ctio
nN
ITN
utrit
ion
Inte
rven
tion
Tria
lsN
OS
New
cast
le-O
ttaw
asc
ale
NR
Not
repo
rted
PRE
DIM
EDP
reve
nci o
nco
nD
ieta
MED
iterr
anea
Nutrition ReviewsVR Vol 77(10)691ndash709 699
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
for high vs low categories of nut consumption was 076(067ndash086) with evidence of moderate interstudy het-
erogeneity (I2 46 Pheterogeneity 004) (Figure 2 andFigure S13 in the Supporting Information online)
Regarding specific types of nuts peanut consumption
was inversely associated with the risk of CHD mortalityafter comparing high vs low categories of consumption
([RR 075 95CI 064ndash088] I2 0 Pheterogeneity 046)(Table 3 and Figure S14 in the Supporting Information
online)
Table 3 Summary RR of cardiovascular disease outcomes by different types of nut consumption (comparing highest vslowest categories)
No of cohorts No of participants No of cases RR (95CI) P for heterogeneity I2 ()
Tree nutsCVD 3 210 836 14 136 085 (079 091) 070 0CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 100 (089 111) 093 0Stroke mortality 3 118 962 1851 093 (077 113) 044 0CHD 3 210 836 8390 077 (070 084) 008 61CHD mortality ndash ndash ndash ndash ndash ndash
PeanutsCVD 3 210 836 14 136 087 (081 093) 067 0CVD mortality 2 134 265 5572 077 (070 085) 081 0Stroke 3 210 836 5910 090 (081 099) 032 13Stroke mortality 4 253 227 3036 083 (073 095) 007 57CHD 3 210 836 8390 085 (079 092) 055 0CHD mortality 2 134 265 2119 075 (064 088) 046 0
WalnutsCVD 3 144 021 5255 081 (071 092) 003 73CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 144 021 5910 085 (071 102) 019 39Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 144 021 2685 079 (066 094) 004 69CHD mortality ndash ndash ndash ndash ndash ndash
Peanut butterCVD 3 210 836 14 136 098 (093 103) lt001 89CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 094 (087 102) lt001 86Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 210 836 8390 100 (094 107) 017 43CHD mortality ndash ndash ndash ndash ndash ndash
Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovascular disease RR risk ratio
Figure 2 Summary plots of effect estimates from prospective cohort studies evaluating the association between nut consumptionand the risk of different cardiovascular outcomes Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovasculardisease RR risk ratio
700 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Nuts and stroke incidence Seven cohort comparisonsanalyzed the association between nut consumption and
stroke incidence including 302 888 participants and12 646 cases The summary RR (95CI) for high vs low
categories of nut consumption was 100 (092ndash109)with no evidence of interstudy heterogeneity (I2 0
Pheterogeneity 097) (Figure 2 and Figure S15 in theSupporting Information online) Regarding specific
types of nuts peanut consumption was associated witha lower risk of stroke incidence after comparing highest
vs lowest categories of consumption ([RR 090 95CI081ndash099] I2 13 Pheterogeneity 032) (Table 3 andFigure S16 in the Supporting Information online) No
association was observed between tree nut walnut andpeanut butter consumption and the risk of stroke inci-
dence (Table 3 and Figures S17ndashS19 in the SupportingInformation online)
Nuts and stroke mortality Twelve cohort comparisons
analyzed the association between nut consumption andstroke mortality including 351 618 participants and
2332 cases The summary RR (95CI) for high vs lowcategories of nut consumption was 083 (075ndash093)
with no evidence of interstudy heterogeneity (I2 0Pheterogeneity 045) (Figure 2 and Figure S20 in the
Supporting Information online) Regarding specifictypes of nuts peanut consumption was associated with
a lower risk of stroke mortality after comparing highestvs lowest categories of consumption ([RR 085 95CI
079ndash092] I2 0 Pheterogeneity 055) (Table 3 andFigure S21 in the Supporting Information online) No
association was reported between tree nut consumptionand the risk of stroke death (Table 3 and Figure S22 in
the Supporting Information online)
Nuts and hemorrhagic stroke Five cohort comparisonsanalyzed the association between nut consumption and
hemorrhagic stroke incidence involving 188 750 partic-ipants and 3088 cases The summary RR (95CI) for
high vs low categories of nut consumption was 102(077ndash134) with no evidence of interstudy heterogene-ity (I2 15 Pheterogeneity 032) (Figure 2 and Figure S23
in the Supporting Information online) No studyanalyzed the association between different types of nuts
and the risk of hemorrhagic stroke
Nuts and ischemic stroke Seven cohort comparisonsanalyzed the association between nut consumption and
ischemic stroke incidence involving 302 423 partici-pants and 8401 cases The summary RR (95CI) for
high vs low categories of nut consumption was 099(089ndash110) with no evidence of interstudy heterogene-
ity (I2 0 Pheterogeneity 062) (Figure 2 and Figure S24 in
the Supporting Information online) No study analyzed
the association between different types of nuts and therisk of hemorrhagic stroke
Nuts and atrial fibrillation Two cohort comparisonsanalyzed the association between nut consumption and
AF involving 53 965 participants and 10 867 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 085 (073ndash099) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 052)
(Figure 2 and Figure S25 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of AF
Nuts and heart failure Two cohort comparisonsanalyzed the association between nut consumption and
HF involving 53 887 participants and 4253 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 100 (086ndash116) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 085)
(Figure 2 and Figure S26 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of HF
Dose-response analyses
Figures S27ndashS34 in the Supporting Information onlineshow the linear and non-linear dose-response analyses
between total nut consumption and CVD outcomesThe summary RR (95CI) for a 28-gd increment was
087 (081ndash093) for CVD incidence 071 (061ndash084)for CVD mortality 075 (064ndash088) for CHD incidence
067 (052ndash087) for CHD mortality 106 (097ndash115) forstroke incidence 101 (088ndash118) for stroke mortality
105 (077ndash143) for hemorrhagic stroke and 106(086ndash131) for ischemic stroke
Total nut consumption and the risk of CVD inci-dence (Figure S27 in the Supporting Information on-
line) CVD mortality (Figure S28 in the SupportingInformation online) stroke mortality (Figure S32 in theSupporting Information online) and hemorrhagic
stroke (Figure S33 in the Supporting Information on-line) showed a non-linear association (Pnon-line-
aritylt001) The reduction in the risk of CVD incidencewas observed up to a consumption of 10 gd with no
further reduction with higher consumptions (Table S6in the Supporting Information online) For CVD mor-
tality and CHD mortality there was a steeper reductionin the risk at approximately 15ndash20 gd with no further
reduction with a higher consumption (Table S6 in theSupporting Information online) The reduction in the
risk of stroke mortality was observed up to a consump-tion of 5 gd with no significant reductions above this
amount (Table S6 in the Supporting Information
Nutrition ReviewsVR Vol 77(10)691ndash709 701
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Pheterogeneity 070) peanuts ([RR 087 95CI
081ndash093] I2 0 Pheterogeneity 067) and walnuts ([RR081 95CI 071ndash092] I2 73 Pheterogeneity 003) was
associated with a lower risk of CVD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S2ndashS4 in the SupportingInformation online) No association was reported be-tween peanut butter consumption and the risk of CVD
incidence ([RR 098 95CI 093ndash103] I2 89Pheterogeneitylt001) (Table 3 and Figure S5 in the
Supporting Information online)
Nuts and cardiovascular disease mortality Fifteen co-hort comparisons analyzed the association between
nut consumption and CVD mortality including413 727 participants and 14 475 cases The summary
RR (95CI) for high vs low categories of nut con-sumption was 077 (072ndash082) with no evidence of
interstudy heterogeneity (I2 3 Pheterogeneity 042)(Figure 2 and Figure S6 in the Supporting Information
online) Regarding specific types of nuts only peanutshave been studied in relation to CVD mortality show-
ing a summary RR of 077 (95CI 070ndash085) for highvs low categories of consumption with no evidence of
interstudy heterogeneity (I2 0 Pheterogeneity 081)(Table 3 and Figure S7 in the Supporting Information
online)
Nuts and coronary heart disease incidence Seven cohortcomparisons analyzed the association between nut con-
sumption and CHD incidence including 275 812 par-ticipants and 12 654 cases The summary RR (95CI)
for high vs low categories of nut consumption was 082(069ndash096) with evidence of substantial interstudy het-
erogeneity (I2 74 Pheterogeneitylt001) (Figure 2 andFigure S8 in the Supporting Information online)
Regarding specific types of nuts consumption of treenuts ([RR 077 95CI 070ndash084] I2 61
Pheterogeneity 008) peanuts ([RR 085 95CI 079ndash092] I2 0 Pheterogeneity 055) and walnuts ([RR0791 95CI 066ndash094] I2 69 Pheterogeneity 004) was
associated with a lower risk of CHD incidence aftercomparing highest vs lowest categories of consumption
(Table 3 and Figures S9ndashS11 in the SupportingInformation online) No association was reported be-
tween peanut butter consumption and the risk of CHDincidence ([RR 100 95CI 094ndash107] I2 43
Pheterogeneity 017) (Table 3 and Figure S12 in theSupporting Information online)
Nuts and coronary heart disease mortality Thirteen co-
hort comparisons analyzed the association between nutconsumption and CHD mortality including 396 041
participants and 7877 cases The summary RR (95CI)Tabl
e2
Cont
inue
dSt
udy
Coun
try
Stud
yna
me
Popu
latio
nN
utco
nsum
ptio
nas
sess
men
tm
etho
d
Type
ofnu
tsN
utin
take
Age
yFo
llow
-up
(mea
nm
edia
nor
rang
e)y
Out
com
eIn
cide
ntca
ses
Fund
ing
sour
ceN
OS
scor
e
Lars
son
etal
(201
8)22
Swed
enCo
hort
ofSw
edis
hM
en32
911
MFF
QN
uts
(not
incl
ud-
ing
coco
nuto
rch
estn
uts)
3
serv
ings
wk
vs neve
r
45ndash8
317
MI
4983
Agen
cy8
MIm
orta
lity
917
Hea
rtfa
ilure
3160
Atria
lfib
rilla
tion
7550
Isch
emic
stro
ke37
82In
trac
ereb
ral
hem
orrh
age
543
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CVD
car
diov
ascu
lard
isea
seE
PIC
Euro
pean
Pros
pect
ive
Inve
stig
atio
nin
toCa
ncer
and
Nut
ritio
nF
fem
ales
FFQ
foo
dfr
eque
ncy
ques
tionn
aire
IH
Di
sche
mic
hear
tdis
ease
Mm
ales
MI
myo
card
iali
nfar
ctio
nN
ITN
utrit
ion
Inte
rven
tion
Tria
lsN
OS
New
cast
le-O
ttaw
asc
ale
NR
Not
repo
rted
PRE
DIM
EDP
reve
nci o
nco
nD
ieta
MED
iterr
anea
Nutrition ReviewsVR Vol 77(10)691ndash709 699
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
for high vs low categories of nut consumption was 076(067ndash086) with evidence of moderate interstudy het-
erogeneity (I2 46 Pheterogeneity 004) (Figure 2 andFigure S13 in the Supporting Information online)
Regarding specific types of nuts peanut consumption
was inversely associated with the risk of CHD mortalityafter comparing high vs low categories of consumption
([RR 075 95CI 064ndash088] I2 0 Pheterogeneity 046)(Table 3 and Figure S14 in the Supporting Information
online)
Table 3 Summary RR of cardiovascular disease outcomes by different types of nut consumption (comparing highest vslowest categories)
No of cohorts No of participants No of cases RR (95CI) P for heterogeneity I2 ()
Tree nutsCVD 3 210 836 14 136 085 (079 091) 070 0CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 100 (089 111) 093 0Stroke mortality 3 118 962 1851 093 (077 113) 044 0CHD 3 210 836 8390 077 (070 084) 008 61CHD mortality ndash ndash ndash ndash ndash ndash
PeanutsCVD 3 210 836 14 136 087 (081 093) 067 0CVD mortality 2 134 265 5572 077 (070 085) 081 0Stroke 3 210 836 5910 090 (081 099) 032 13Stroke mortality 4 253 227 3036 083 (073 095) 007 57CHD 3 210 836 8390 085 (079 092) 055 0CHD mortality 2 134 265 2119 075 (064 088) 046 0
WalnutsCVD 3 144 021 5255 081 (071 092) 003 73CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 144 021 5910 085 (071 102) 019 39Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 144 021 2685 079 (066 094) 004 69CHD mortality ndash ndash ndash ndash ndash ndash
Peanut butterCVD 3 210 836 14 136 098 (093 103) lt001 89CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 094 (087 102) lt001 86Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 210 836 8390 100 (094 107) 017 43CHD mortality ndash ndash ndash ndash ndash ndash
Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovascular disease RR risk ratio
Figure 2 Summary plots of effect estimates from prospective cohort studies evaluating the association between nut consumptionand the risk of different cardiovascular outcomes Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovasculardisease RR risk ratio
700 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Nuts and stroke incidence Seven cohort comparisonsanalyzed the association between nut consumption and
stroke incidence including 302 888 participants and12 646 cases The summary RR (95CI) for high vs low
categories of nut consumption was 100 (092ndash109)with no evidence of interstudy heterogeneity (I2 0
Pheterogeneity 097) (Figure 2 and Figure S15 in theSupporting Information online) Regarding specific
types of nuts peanut consumption was associated witha lower risk of stroke incidence after comparing highest
vs lowest categories of consumption ([RR 090 95CI081ndash099] I2 13 Pheterogeneity 032) (Table 3 andFigure S16 in the Supporting Information online) No
association was observed between tree nut walnut andpeanut butter consumption and the risk of stroke inci-
dence (Table 3 and Figures S17ndashS19 in the SupportingInformation online)
Nuts and stroke mortality Twelve cohort comparisons
analyzed the association between nut consumption andstroke mortality including 351 618 participants and
2332 cases The summary RR (95CI) for high vs lowcategories of nut consumption was 083 (075ndash093)
with no evidence of interstudy heterogeneity (I2 0Pheterogeneity 045) (Figure 2 and Figure S20 in the
Supporting Information online) Regarding specifictypes of nuts peanut consumption was associated with
a lower risk of stroke mortality after comparing highestvs lowest categories of consumption ([RR 085 95CI
079ndash092] I2 0 Pheterogeneity 055) (Table 3 andFigure S21 in the Supporting Information online) No
association was reported between tree nut consumptionand the risk of stroke death (Table 3 and Figure S22 in
the Supporting Information online)
Nuts and hemorrhagic stroke Five cohort comparisonsanalyzed the association between nut consumption and
hemorrhagic stroke incidence involving 188 750 partic-ipants and 3088 cases The summary RR (95CI) for
high vs low categories of nut consumption was 102(077ndash134) with no evidence of interstudy heterogene-ity (I2 15 Pheterogeneity 032) (Figure 2 and Figure S23
in the Supporting Information online) No studyanalyzed the association between different types of nuts
and the risk of hemorrhagic stroke
Nuts and ischemic stroke Seven cohort comparisonsanalyzed the association between nut consumption and
ischemic stroke incidence involving 302 423 partici-pants and 8401 cases The summary RR (95CI) for
high vs low categories of nut consumption was 099(089ndash110) with no evidence of interstudy heterogene-
ity (I2 0 Pheterogeneity 062) (Figure 2 and Figure S24 in
the Supporting Information online) No study analyzed
the association between different types of nuts and therisk of hemorrhagic stroke
Nuts and atrial fibrillation Two cohort comparisonsanalyzed the association between nut consumption and
AF involving 53 965 participants and 10 867 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 085 (073ndash099) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 052)
(Figure 2 and Figure S25 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of AF
Nuts and heart failure Two cohort comparisonsanalyzed the association between nut consumption and
HF involving 53 887 participants and 4253 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 100 (086ndash116) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 085)
(Figure 2 and Figure S26 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of HF
Dose-response analyses
Figures S27ndashS34 in the Supporting Information onlineshow the linear and non-linear dose-response analyses
between total nut consumption and CVD outcomesThe summary RR (95CI) for a 28-gd increment was
087 (081ndash093) for CVD incidence 071 (061ndash084)for CVD mortality 075 (064ndash088) for CHD incidence
067 (052ndash087) for CHD mortality 106 (097ndash115) forstroke incidence 101 (088ndash118) for stroke mortality
105 (077ndash143) for hemorrhagic stroke and 106(086ndash131) for ischemic stroke
Total nut consumption and the risk of CVD inci-dence (Figure S27 in the Supporting Information on-
line) CVD mortality (Figure S28 in the SupportingInformation online) stroke mortality (Figure S32 in theSupporting Information online) and hemorrhagic
stroke (Figure S33 in the Supporting Information on-line) showed a non-linear association (Pnon-line-
aritylt001) The reduction in the risk of CVD incidencewas observed up to a consumption of 10 gd with no
further reduction with higher consumptions (Table S6in the Supporting Information online) For CVD mor-
tality and CHD mortality there was a steeper reductionin the risk at approximately 15ndash20 gd with no further
reduction with a higher consumption (Table S6 in theSupporting Information online) The reduction in the
risk of stroke mortality was observed up to a consump-tion of 5 gd with no significant reductions above this
amount (Table S6 in the Supporting Information
Nutrition ReviewsVR Vol 77(10)691ndash709 701
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
for high vs low categories of nut consumption was 076(067ndash086) with evidence of moderate interstudy het-
erogeneity (I2 46 Pheterogeneity 004) (Figure 2 andFigure S13 in the Supporting Information online)
Regarding specific types of nuts peanut consumption
was inversely associated with the risk of CHD mortalityafter comparing high vs low categories of consumption
([RR 075 95CI 064ndash088] I2 0 Pheterogeneity 046)(Table 3 and Figure S14 in the Supporting Information
online)
Table 3 Summary RR of cardiovascular disease outcomes by different types of nut consumption (comparing highest vslowest categories)
No of cohorts No of participants No of cases RR (95CI) P for heterogeneity I2 ()
Tree nutsCVD 3 210 836 14 136 085 (079 091) 070 0CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 100 (089 111) 093 0Stroke mortality 3 118 962 1851 093 (077 113) 044 0CHD 3 210 836 8390 077 (070 084) 008 61CHD mortality ndash ndash ndash ndash ndash ndash
PeanutsCVD 3 210 836 14 136 087 (081 093) 067 0CVD mortality 2 134 265 5572 077 (070 085) 081 0Stroke 3 210 836 5910 090 (081 099) 032 13Stroke mortality 4 253 227 3036 083 (073 095) 007 57CHD 3 210 836 8390 085 (079 092) 055 0CHD mortality 2 134 265 2119 075 (064 088) 046 0
WalnutsCVD 3 144 021 5255 081 (071 092) 003 73CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 144 021 5910 085 (071 102) 019 39Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 144 021 2685 079 (066 094) 004 69CHD mortality ndash ndash ndash ndash ndash ndash
Peanut butterCVD 3 210 836 14 136 098 (093 103) lt001 89CVD mortality ndash ndash ndash ndash ndash ndashStroke 3 210 836 5910 094 (087 102) lt001 86Stroke mortality ndash ndash ndash ndash ndash ndashCHD 3 210 836 8390 100 (094 107) 017 43CHD mortality ndash ndash ndash ndash ndash ndash
Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovascular disease RR risk ratio
Figure 2 Summary plots of effect estimates from prospective cohort studies evaluating the association between nut consumptionand the risk of different cardiovascular outcomes Abbreviations CHD coronary heart disease CI confidence interval CVD cardiovasculardisease RR risk ratio
700 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Nuts and stroke incidence Seven cohort comparisonsanalyzed the association between nut consumption and
stroke incidence including 302 888 participants and12 646 cases The summary RR (95CI) for high vs low
categories of nut consumption was 100 (092ndash109)with no evidence of interstudy heterogeneity (I2 0
Pheterogeneity 097) (Figure 2 and Figure S15 in theSupporting Information online) Regarding specific
types of nuts peanut consumption was associated witha lower risk of stroke incidence after comparing highest
vs lowest categories of consumption ([RR 090 95CI081ndash099] I2 13 Pheterogeneity 032) (Table 3 andFigure S16 in the Supporting Information online) No
association was observed between tree nut walnut andpeanut butter consumption and the risk of stroke inci-
dence (Table 3 and Figures S17ndashS19 in the SupportingInformation online)
Nuts and stroke mortality Twelve cohort comparisons
analyzed the association between nut consumption andstroke mortality including 351 618 participants and
2332 cases The summary RR (95CI) for high vs lowcategories of nut consumption was 083 (075ndash093)
with no evidence of interstudy heterogeneity (I2 0Pheterogeneity 045) (Figure 2 and Figure S20 in the
Supporting Information online) Regarding specifictypes of nuts peanut consumption was associated with
a lower risk of stroke mortality after comparing highestvs lowest categories of consumption ([RR 085 95CI
079ndash092] I2 0 Pheterogeneity 055) (Table 3 andFigure S21 in the Supporting Information online) No
association was reported between tree nut consumptionand the risk of stroke death (Table 3 and Figure S22 in
the Supporting Information online)
Nuts and hemorrhagic stroke Five cohort comparisonsanalyzed the association between nut consumption and
hemorrhagic stroke incidence involving 188 750 partic-ipants and 3088 cases The summary RR (95CI) for
high vs low categories of nut consumption was 102(077ndash134) with no evidence of interstudy heterogene-ity (I2 15 Pheterogeneity 032) (Figure 2 and Figure S23
in the Supporting Information online) No studyanalyzed the association between different types of nuts
and the risk of hemorrhagic stroke
Nuts and ischemic stroke Seven cohort comparisonsanalyzed the association between nut consumption and
ischemic stroke incidence involving 302 423 partici-pants and 8401 cases The summary RR (95CI) for
high vs low categories of nut consumption was 099(089ndash110) with no evidence of interstudy heterogene-
ity (I2 0 Pheterogeneity 062) (Figure 2 and Figure S24 in
the Supporting Information online) No study analyzed
the association between different types of nuts and therisk of hemorrhagic stroke
Nuts and atrial fibrillation Two cohort comparisonsanalyzed the association between nut consumption and
AF involving 53 965 participants and 10 867 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 085 (073ndash099) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 052)
(Figure 2 and Figure S25 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of AF
Nuts and heart failure Two cohort comparisonsanalyzed the association between nut consumption and
HF involving 53 887 participants and 4253 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 100 (086ndash116) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 085)
(Figure 2 and Figure S26 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of HF
Dose-response analyses
Figures S27ndashS34 in the Supporting Information onlineshow the linear and non-linear dose-response analyses
between total nut consumption and CVD outcomesThe summary RR (95CI) for a 28-gd increment was
087 (081ndash093) for CVD incidence 071 (061ndash084)for CVD mortality 075 (064ndash088) for CHD incidence
067 (052ndash087) for CHD mortality 106 (097ndash115) forstroke incidence 101 (088ndash118) for stroke mortality
105 (077ndash143) for hemorrhagic stroke and 106(086ndash131) for ischemic stroke
Total nut consumption and the risk of CVD inci-dence (Figure S27 in the Supporting Information on-
line) CVD mortality (Figure S28 in the SupportingInformation online) stroke mortality (Figure S32 in theSupporting Information online) and hemorrhagic
stroke (Figure S33 in the Supporting Information on-line) showed a non-linear association (Pnon-line-
aritylt001) The reduction in the risk of CVD incidencewas observed up to a consumption of 10 gd with no
further reduction with higher consumptions (Table S6in the Supporting Information online) For CVD mor-
tality and CHD mortality there was a steeper reductionin the risk at approximately 15ndash20 gd with no further
reduction with a higher consumption (Table S6 in theSupporting Information online) The reduction in the
risk of stroke mortality was observed up to a consump-tion of 5 gd with no significant reductions above this
amount (Table S6 in the Supporting Information
Nutrition ReviewsVR Vol 77(10)691ndash709 701
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Nuts and stroke incidence Seven cohort comparisonsanalyzed the association between nut consumption and
stroke incidence including 302 888 participants and12 646 cases The summary RR (95CI) for high vs low
categories of nut consumption was 100 (092ndash109)with no evidence of interstudy heterogeneity (I2 0
Pheterogeneity 097) (Figure 2 and Figure S15 in theSupporting Information online) Regarding specific
types of nuts peanut consumption was associated witha lower risk of stroke incidence after comparing highest
vs lowest categories of consumption ([RR 090 95CI081ndash099] I2 13 Pheterogeneity 032) (Table 3 andFigure S16 in the Supporting Information online) No
association was observed between tree nut walnut andpeanut butter consumption and the risk of stroke inci-
dence (Table 3 and Figures S17ndashS19 in the SupportingInformation online)
Nuts and stroke mortality Twelve cohort comparisons
analyzed the association between nut consumption andstroke mortality including 351 618 participants and
2332 cases The summary RR (95CI) for high vs lowcategories of nut consumption was 083 (075ndash093)
with no evidence of interstudy heterogeneity (I2 0Pheterogeneity 045) (Figure 2 and Figure S20 in the
Supporting Information online) Regarding specifictypes of nuts peanut consumption was associated with
a lower risk of stroke mortality after comparing highestvs lowest categories of consumption ([RR 085 95CI
079ndash092] I2 0 Pheterogeneity 055) (Table 3 andFigure S21 in the Supporting Information online) No
association was reported between tree nut consumptionand the risk of stroke death (Table 3 and Figure S22 in
the Supporting Information online)
Nuts and hemorrhagic stroke Five cohort comparisonsanalyzed the association between nut consumption and
hemorrhagic stroke incidence involving 188 750 partic-ipants and 3088 cases The summary RR (95CI) for
high vs low categories of nut consumption was 102(077ndash134) with no evidence of interstudy heterogene-ity (I2 15 Pheterogeneity 032) (Figure 2 and Figure S23
in the Supporting Information online) No studyanalyzed the association between different types of nuts
and the risk of hemorrhagic stroke
Nuts and ischemic stroke Seven cohort comparisonsanalyzed the association between nut consumption and
ischemic stroke incidence involving 302 423 partici-pants and 8401 cases The summary RR (95CI) for
high vs low categories of nut consumption was 099(089ndash110) with no evidence of interstudy heterogene-
ity (I2 0 Pheterogeneity 062) (Figure 2 and Figure S24 in
the Supporting Information online) No study analyzed
the association between different types of nuts and therisk of hemorrhagic stroke
Nuts and atrial fibrillation Two cohort comparisonsanalyzed the association between nut consumption and
AF involving 53 965 participants and 10 867 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 085 (073ndash099) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 052)
(Figure 2 and Figure S25 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of AF
Nuts and heart failure Two cohort comparisonsanalyzed the association between nut consumption and
HF involving 53 887 participants and 4253 cases Thesummary RR (95CI) for high vs low categories of nut
consumption was 100 (086ndash116) with no evidence ofinterstudy heterogeneity (I2 0 Pheterogeneity 085)
(Figure 2 and Figure S26 in the Supporting Informationonline) No study analyzed the association between dif-
ferent types of nuts and the risk of HF
Dose-response analyses
Figures S27ndashS34 in the Supporting Information onlineshow the linear and non-linear dose-response analyses
between total nut consumption and CVD outcomesThe summary RR (95CI) for a 28-gd increment was
087 (081ndash093) for CVD incidence 071 (061ndash084)for CVD mortality 075 (064ndash088) for CHD incidence
067 (052ndash087) for CHD mortality 106 (097ndash115) forstroke incidence 101 (088ndash118) for stroke mortality
105 (077ndash143) for hemorrhagic stroke and 106(086ndash131) for ischemic stroke
Total nut consumption and the risk of CVD inci-dence (Figure S27 in the Supporting Information on-
line) CVD mortality (Figure S28 in the SupportingInformation online) stroke mortality (Figure S32 in theSupporting Information online) and hemorrhagic
stroke (Figure S33 in the Supporting Information on-line) showed a non-linear association (Pnon-line-
aritylt001) The reduction in the risk of CVD incidencewas observed up to a consumption of 10 gd with no
further reduction with higher consumptions (Table S6in the Supporting Information online) For CVD mor-
tality and CHD mortality there was a steeper reductionin the risk at approximately 15ndash20 gd with no further
reduction with a higher consumption (Table S6 in theSupporting Information online) The reduction in the
risk of stroke mortality was observed up to a consump-tion of 5 gd with no significant reductions above this
amount (Table S6 in the Supporting Information
Nutrition ReviewsVR Vol 77(10)691ndash709 701
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
online) The association between total nuts and
hemorrhagic stroke appeared to be J-shaped with a riskreduction up to 5 gd but there was a slight non-
significant positive association at intakes of 25 gd(Table S6)
There was no evidence of non-linear associationfor the other outcomes
Sensitivity analyses
Table S7 in the Supporting Information online showsthe sensitivity analysis by the removal of one study at a
time Regarding total nut consumption no trial modi-fied the magnitude direction or significance of the
pooled estimates or the evidence for heterogeneity fortotal CVD incidence and mortality stroke incidence
and mortality HF or AF Removal of the Guasch-Ferreet al study (NHSI)21 changed the pooled estimates of to-
tal CHD incidence from significant to nonsignificantRemoval of the Larsson et al22 study explained the het-
erogeneity for CHD death (I2 16 Pheterogeneity 028)With regard to different types of nuts removal of the
Guasch-Ferre et al study (HPFS) and Guasch-Ferrestudy (NHSII)21 modified the significance of the pooled
RR for peanut consumption and total stroke from sig-nificant to nonsignificant and removal of the Bao
et al45 study explained the heterogeneity (I2 0Pheterogeneity 077) In the case of peanut butter removal
of the Guasch-Ferre et al study (NHSII)21 explained theheterogeneity for total CVD (I2 0 Pheterogeneity 060)
and for total stroke (I2 0 Pheterogeneity 053)Furthermore removal of the Guasch-Ferre et al study
(NHSI)21 changed the significance of the pooled esti-mates for total stroke from nonsignificant to significant
Regarding walnuts removal of the Guasch-Ferre et alstudy (HPFS)21 changed the magnitude of the RR for
total CVD and CHD and the pooled estimates becamesignificant for total stroke Removal of the Guasch-
Ferre et al study (NHSI)21 changed the significance ofthe pooled estimates for total CVD and CHD from sig-nificant to nonsignificant Finally removal of the
Guasch-Ferre et al study (NHSII)21 explained the het-erogeneity (I2 0 Pheterogeneity 053) for CHD and
changed the significance of the pooled estimates fromsignificant to nonsignificant
Subgroup analyses
Subgroup analyses could only be conducted for CVD
CHD and stroke death Figures S35ndashS40 in theSupporting Information online show the a priori sub-
group analyses for the aforementioned outcomes Themeta-regression analysis revealed no evidence of effect
modification by sex duration of follow-up or NOS
quality score and its individual domains However the
risk of CVD death was modified by geographical area(Pfrac14 003) In studies conducted in America nut con-
sumption was inversely associated with the risk of CVDmortality (RR 068 95CI 061ndash077) whereas no as-
sociation was observed in those studies conducted inEurope (RR 091 95CI 072ndash114) or Oceania (RR095 95CI 071ndash128) Geographical area explained
389 of the total heterogeneity (I2 47Pheterogeneity 004) No effect modification by geographi-
cal area was observed for CHD mortality and strokemortality
Publication bias
Figures S41ndashS43 in the Supporting Information onlineshow the funnel plots used to assess publication bias for
death from CVD CHD and stroke (the only outcomeswith more than 10 study comparisons in the analyses)
There was no statistical evidence of small study effectsbased on visual inspection of the funnel plots with ei-
ther Eggerrsquos test or Beggrsquos test (all Pgt 005)
Grading of the evidence
Table 4 shows the GRADE assessment for the certainty
of the evidence for the association between total nutconsumption and the risk of CVD outcomes The evi-dence was rated as moderate for CVD mortality and
CHD mortality low for CVD incidence and strokemortality and very low for CHD incidence stroke inci-
dence hemorrhagic stroke ischemic stroke AF andHF Tables S8ndashS11 in the Supporting Information on-
line show the GRADE assessment for the associationbetween subtypes of nut consumption and the risk of
CVD outcomes The overall certainty of the evidencewas graded as very low for all subtypes of nut consump-
tion and CVD outcomes
DISCUSSION
The results of the present systematic review and meta-analysis of prospective cohort studies showed a signifi-
cant inverse association between total nut consumptionand the risk of CVD incidence and mortality CHD in-
cidence and mortality and AF There was no associa-tion between total nut consumption and stroke
incidence or mortality hemorrhagic stroke ischemicstroke and HF Regarding specific types of nuts tree
nut consumption was associated with a lower risk ofCVD and CHD incidence while peanut consumption
was associated with a lower incidence of and mortalityfrom CVD stroke and CHD and walnut consumption
with a lower incidence of CVD stroke and CHD No
702 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Tabl
e4
GRA
DE
asse
ssm
ent
ofth
esy
stem
atic
revi
ewan
dm
eta-
anal
ysis
ofpr
ospe
ctiv
eco
hort
stud
ies
asse
ssin
gth
eas
soci
atio
nbe
twee
nto
taln
utco
nsum
ptio
nan
dca
rdio
vasc
u-la
rdi
seas
eou
tcom
esO
utco
me
No
ofst
udie
sSt
udy
desi
gnRi
skof
bias
Inco
nsis
tenc
yIn
dire
ctne
ssIm
prec
isio
nO
ther
cons
ider
atio
nsa
Rela
tive
(95
CI)
Qua
lity
CVD
inci
denc
e3
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usb
Not
serio
usD
ose-
resp
onse
grad
ient
cRR
085
(08
0to
091
)
LOW
CVD
mor
talit
y14
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
ntd
RR0
77(0
72
to0
82)
MO
DER
ATE
CHD
inci
denc
e7
Obs
erva
tiona
lstu
dies
Not
serio
usSe
rious
eSe
rious
fSe
rious
gD
ose-
resp
onse
grad
ient
hRR
082
(06
9to
096
)
VERY
LOW
CHD
mor
talit
y12
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Not
serio
usN
otse
rious
Dos
e-re
spon
segr
adie
nti
RR0
76(0
67
to0
86)
MO
DER
ATE
Stro
kein
cide
nce
7O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
jSe
rious
kN
one
RR1
00(0
92
to1
09)
VERY
LOW
Stro
kem
orta
lity
11O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usN
otse
rious
Serio
usl
Dos
e-re
spon
segr
adie
ntm
RR0
87(0
76
to1
00)
LOW
Hem
orrh
agic
stro
ke5
Obs
erva
tiona
lstu
dies
Serio
usn
Not
serio
usSe
rious
oSe
rious
pD
ose-
resp
onse
grad
ient
qRR
102
(07
7to
134
)
VERY
LOW
Isch
emic
stro
ke7
Obs
erva
tiona
lstu
dies
Not
serio
usN
otse
rious
Serio
usr
Serio
uss
Non
eRR
099
(08
9to
110
)
VERY
LOW
Atria
lfib
rilla
tion
2O
bser
vatio
nals
tudi
esN
otse
rious
Not
serio
usSe
rious
tSe
rious
uN
one
RR0
85(0
73
to0
99)
VERY
LOW
Hea
rtfa
ilure
2O
bser
vatio
nals
tudi
esSe
rious
vN
otse
rious
Serio
usw
Serio
usx
Non
eRR
100
(08
6to
116
)
VERY
LOW
Abbr
evia
tions
CH
Dc
oron
ary
hear
tdis
ease
CI
conf
iden
cein
terv
alC
VDc
ardi
ovas
cula
rdis
ease
GLS
Tge
nera
lized
leas
tsqu
ares
tren
dG
RAD
EG
radi
ngof
Reco
mm
enda
tions
Ass
essm
ent
Dev
elop
men
tan
dEv
alua
tion
NO
SN
ewca
stle
-Ott
awa
scal
eRR
ris
kra
tio
a Publ
icat
ion
bias
coul
dno
tbe
asse
ssed
inm
eta-
anal
yses
that
incl
udedlt
10tr
ialc
ompa
rison
sTh
eref
ore
fort
hese
outc
omes
no
dow
ngra
des
wer
em
ade
forp
ublic
atio
nbi
as
bSe
rious
indi
rect
ness
forC
VDin
cide
nce
asth
ein
clud
edst
udie
sw
ere
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(69
30
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
mal
es
c Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
inci
denc
e(Plt
001
)se
eFi
gure
S27
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ed
Upg
rade
fora
dose
-res
pons
egr
adie
nta
sth
eM
KSPL
INE
dose
-res
pons
ean
alys
essh
owed
asi
gnifi
cant
nonl
inea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CVD
mor
talit
y(Plt
001
)se
eFi
gure
S28
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ee Se
rious
inco
nsis
tenc
yfo
rCH
Din
cide
nce
due
tohi
ghde
gree
ofun
expl
aine
dhe
tero
gene
ity(I2frac14
74
Pfrac14
000
1)
f Serio
usin
dire
ctne
ssfo
rCH
Din
cide
nce
asgt
50
ofth
ew
eigh
t(55
4
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
gSe
rious
impr
ecis
ion
forC
HD
inci
denc
eas
the
95
CI(0
69ndash
096
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
hU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
GLS
Tdo
se-r
espo
nse
anal
yses
reve
aled
asi
gnifi
cant
linea
rinv
erse
rela
tions
hip
betw
een
tota
lnut
cons
umpt
ion
and
CHD
inci
denc
e(Plt
001
)se
eFi
gure
S29
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ei U
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dCH
Dm
orta
lity
(Plt
001
)se
eFi
gure
S30
inth
eSu
ppor
ting
Info
rmat
ion
onlin
ej Se
rious
indi
rect
ness
fors
trok
ein
cide
nce
asgt
50
ofth
ew
eigh
t(72
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
k Serio
usim
prec
isio
nfo
rstr
oke
inci
denc
eas
the
95
CI(0
92ndash
109
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
l Serio
usim
prec
isio
nfo
rstr
oke
mor
talit
yas
the
95
CI(0
76ndash
100
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
mU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dst
roke
mor
talit
y(Pfrac14
002
9)s
eeFi
gure
S32
inth
eSu
ppor
ting
Info
rmat
ion
onlin
en
Serio
usris
kof
bias
forh
emor
rhag
icst
roke
asgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
o Serio
usin
dire
ctne
ssfo
rhem
orrh
agic
stro
kea
sgt
50
ofth
ew
eigh
t(68
7
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
andgt
50
ofth
ew
eigh
t(55
7
)was
cont
rib-
uted
byst
udie
sco
nduc
ted
amon
gm
ales
p
Serio
usim
prec
isio
nfo
rhem
orrh
agic
stro
keas
the
95
CI(0
77ndash
134
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
qU
pgra
defo
rado
se-r
espo
nse
grad
ient
as
the
MKS
PLIN
Edo
se-r
espo
nse
anal
yses
show
eda
sign
ifica
ntno
nlin
eari
nver
sere
latio
nshi
pbe
twee
nto
taln
utco
nsum
ptio
nan
dhe
mor
rhag
icst
roke
(Pfrac14
001
)se
eFi
gure
S33
inth
eSu
ppor
ting
Info
rmat
ion
onlin
er Se
rious
indi
rect
ness
fori
sche
mic
stro
kea
sgt
50
ofth
ew
eigh
t(66
1
)was
cont
ribut
edby
stud
ies
cond
ucte
dam
ong
heal
thpr
ofes
sion
als
s Serio
usim
prec
isio
nfo
risc
hem
icst
roke
as
the
95
CI(0
89ndash
110
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
t Serio
usin
dire
ctne
ssfo
ratr
ialf
ibril
latio
nas
only
2av
aila
ble
stud
ies
wer
eco
nduc
ted
amon
gm
ales
u
Serio
usim
prec
isio
nfo
ratr
ialf
ibril
latio
nas
the
95
CI(0
73ndash
099
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)
v Serio
usris
kof
bias
forh
eart
failu
rea
sgt
50
ofth
ew
eigh
t(65
40
)was
cont
ribut
edby
ast
udy
cons
ider
edto
beat
high
risk
ofbi
as(N
OSlt
7)
wSe
rious
indi
rect
ness
forh
eart
failu
rea
son
ly2
avai
labl
est
udie
sw
ere
cond
ucte
dam
ong
mal
es
x Serio
usim
prec
isio
nfo
rhea
rtfa
ilure
as
the
95
CI(0
86ndash
116
)ove
rlapp
edw
ithth
em
inim
ally
impo
rtan
tdiff
eren
cefo
rclin
ical
bene
fit(R
R0
95)a
ndha
rm(R
R1
05)
Nutrition ReviewsVR Vol 77(10)691ndash709 703
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
association was observed between peanut butter con-
sumption and CVD outcomesSeveral previous meta-analyses have focused on
summarizing data regarding nut consumption and dif-ferent CVD outcomes8ndash20 It is important to highlight
that some limitations in terms of methodology werepresent such as the inclusion of studies with nuts plusseeds or fruits as exposure9101617 or the arbitrary com-
bination of different end points across the analyses (egthe inclusion of studies with a cause-specific CVD out-
come in CVD)91418ndash20 The present meta-analysisattempted to deal with these methodological issues by
including exclusively those studies that reported onlynut consumption as exposure Additionally in the anal-
yses of CVD incidence only those studies evaluating acomposite of non-fatal or a combination of nonfatal
and fatal CVD events were included Similarly forCVD mortality only studies that evaluated a composite
of fatal CVD events were considered For secondaryoutcomes the same definition criteria as in the primary
outcomes were appliedThe results of the present study regarding total nut
consumption are highly consistent with one of themost recent meta-analyses in this field15 which showed
an inverse association between nut consumption anddifferent CVD outcomes and which also took into ac-
count the aforementioned methodological issues at thetime of performing their analyses Recently one study
conducted only among individuals with diabetes ob-served similar results Those individuals consuming 5
servings of total nuts per week presented a lower risk ofCVD incidence (HR 083 95CI 071ndash098) CHD in-
cidence (HR 080 95CI 067ndash096) and CVD mortal-ity (HR 066 95CI 052ndash084) than those consuming
less than 1 serving per month25
Results regarding specific types of nuts are in line
with those of previous meta-analyses which alsoshowed an inverse association between tree nut con-
sumption and peanut consumption and the risk ofCVD incidence CVD mortality CHD incidence andCHD mortality91115 and no association with stroke in-
cidence915 In a recent analysis conducted only amongindividuals with diabetes the results were similar for
tree nut consumption which was also associated with alower risk of CVD incidence and mortality and CHD
incidence However the findings revealed no associa-tion between peanut consumption and CVD out-
comes25 Although the present meta-analysis alsoevaluated the association between walnut consumption
and peanut butter consumption and the risk of CVDCHD and stroke incidence the data was sourced from
only one report which included information from threedifferent cohorts the NHSI NHSII and HPFS21 The
lack of association between peanut butter consumption
and the risk of CVD outcomes may be due to the addi-
tion of salt and hydrogenated fats which could counter-act the beneficial effect of other nutrients present in raw
peanuts At present owing to the limited number ofstudies included in previous meta-analyses and in the
present analyses and considering the high degree ofinterstudy heterogeneity it is unclear whether differenttypes of nuts are associated with CVD outcomes
Different potential mechanisms have been pro-posed to explain the beneficial association observed be-
tween nut consumption and different CVD outcomesNuts are rich in unsaturated fatty acids proteins differ-
ent minerals (including potassium and magnesium)vitamins (including vitamin C and E) and phenolic
compounds This unique nutritional profile means thatnuts possess different properties that beneficially mod-
ify CVD risk factors and therefore reduce the risk ofCVD In fact the ability to lower total cholesterol and
low-density lipoprotein (LDL)-cholesterol levels isprobably one of the best-known properties of nuts as
was demonstrated by a pooled analysis of 25 interven-tion trials61 and more recently in one meta-analysis of
61 randomized controlled trials62 Other possible mech-anisms include a reduction in circulating levels of in-
flammatory cytokines (especially C-reactive protein)the modulation of nitric oxide production an improve-
ment in endothelial function and a reduction in oxida-tive stress6364
The present analysis has some strengths that shouldbe elucidated First a comprehensive systematic search
strategy was used to identify all available prospective co-hort studies Second studies reporting only nut con-
sumption as exposure were included Third the certaintyof the evidence was assessed using the GRADE
approachHowever the present systematic review and meta-
analysis also has some limitations Subgroup analyses formost of the outcomes could not be performed because
less than 10 study comparisons were availableMeasurement error in the evaluation of nut consumptioncould not be ruled out because all included studies used
food frequency questionnaires for this purpose Becauseof this limitation along with the possibility of residual
confounding because of the observational nature of theincluded studies GRADE-assessed prospective cohort
studies tend to be of lower quality than other types ofprospective studies Another important limitation is that
the certainty of the evidence in the effect estimates re-garding total nut consumption was moderate only for
two outcomes (CVD mortality and CHD mortality) andit was considered as low and very low for the others
mainly owing to downgrading for indirectness and im-precision Therefore future research is very likely to
change the confidence in the effect estimates
704 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
CONCLUSION
The present systematic review and meta-analysis pro-vides the most updated and comprehensive summary
estimates of the association between total nut consump-tion different subtypes and CVD outcomes The results
suggest a beneficial role of total nut consumption in re-ducing the incidence of and mortality from different
CVD outcomes Future research should focus on spe-cific types of nuts in order to better clarify their effect
on CVD outcomes
Acknowledgments
Author contributions NB-T and JS-S designed thestudy NB-T JS-S and IP-G contributed to data
acquisition NB-T JS-S and IP-G analyzed andinterpreted data NB-T JS-S IP-G CWCK
HK DR and JLS drafted and critically revised themanuscript for important intellectual content NB-Tand JS-S are the guarantors of this work have full ac-
cess to the data of the study and take responsibility forthe accuracy of the data analysis and the integrity of the
data
Funding The Diabetes and Nutrition Study Group(DNSG) of the European Association for the Study of
Diabetes (EASD) commissioned this systematic reviewand meta-analysis and provided funding and logistical
support for meetings as part of their development ofclinical practice guidelines for nutrition therapy JS-S
was funded by a PSI Graham Farquharson KnowledgeTranslation Fellowship Canadian Diabetes Association
Clinician Scientist award CIHR INMDCNS NewInvestigator Partnership Prize and Banting amp Best
Diabetes Centre Sun Life Financial New InvestigatorAward With the exception of the Clinical Practice
Guidelines Committee of the DNSG of the EASD noneof the sponsors had a role in any aspect of the present
study including design and conduct of the study col-lection management analysis and interpretation of the
data and preparation review approval of the manu-script or decision to publish IP-G has received grants
from the Spanish Ministry of Education Culture andSports (FPU 1701925)
Declaration of interest JS-S reports serving on the
board of the International Nut and Dried Fruit Counciland the Eroski Foundation and receiving grant support
from these entities through his institution He alsoreports serving on the Executive Committee of the
Instituto Danone Spain He has received research fund-ing from the Instituto de Salud Carlos III Spain
Ministerio de Educacion y Ciencia Spain Departamentde Salut Publica de la Generalitat de Catalunya
Catalonia Spain and the European Commission Hehas also received research funding from the California
Walnut Commission Sacramento CA USAPatrimonio Comunal Olivarero Spain La Morella
Nuts Spain and Borges SA Spain He reports receiv-ing consulting fees or travel expenses from Danone the
California Walnut Commission the Eroski Foundationthe Instituto Danone - Spain Nuts for Life the
Australian Nut Industry Council Nestle AbbottLaboratories and Font Vella Lanjaron He is on the
Clinical Practice Guidelines Expert Committee of theEuropean Association for the study of Diabetes
(EASD) and has served on the Scientific Committee ofthe Spanish Food and Safety Agency and the Spanish
Federation of Food Nutrition and Dietetic SocietiesHe is a member of the International Carbohydrate
Quality Consortium (ICQC) and Executive BoardMember of the Diabetes and Nutrition Study Group
(DNSG) of the EASD JLS has received research sup-port from the Canadian Foundation for Innovation
Ontario Research Fund Province of Ontario Ministryof Research and Innovation and Science Canadian
Institutes of health Research (CIHR) Diabetes CanadaPSI Foundation Banting and Best Diabetes Centre(BBDC) American Society for Nutrition (ASN) INC
International Nut and Dried Fruit Council FoundationNational Dried Fruit Trade Association The Tate and
Lyle Nutritional Research Fund at the University ofToronto The Glycemic Control and Cardiovascular
Disease in Type 2 Diabetes Fund at the University ofToronto (a fund established by the Alberta Pulse
Growers) and the Nutrition Trialists Fund at theUniversity of Toronto (a fund established by an inaugu-
ral donation from the Calorie Control Council) He hasreceived in-kind food donations to support a random-
ized controlled trial from the Almond Board ofCalifornia California Walnut Commission American
Peanut Council Barilla Unilever UnicoPrimo LoblawCompanies Quaker Kellogg Canada and WhiteWave
Foods He has received travel support speaker feesandor honoraria from Diabetes Canada Mottrsquos LLP
Dairy Farmers of Canada FoodMinds LLCInternational Sweeteners Association Nestlu Pulse
Canada Canadian Society for Endocrinology andMetabolism (CSEM) GI Foundation Abbott Biofortis
ASN Northern Ontario School of Medicine INCNutrition Research amp Education Foundation European
Food Safety Authority (EFSA) and PhysiciansCommittee for Responsible Medicine He has or has
had ad hoc consulting arrangements with Perkins CoieLLP Tate amp Lyle and Wirtschaftliche Vereinigung
Nutrition ReviewsVR Vol 77(10)691ndash709 705
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Zucker eV He is a member of the European Fruit JuiceAssociation Scientific Expert Panel He is on the
Clinical Practice Guidelines Expert Committees ofDiabetes Canada European Association for the study of
Diabetes (EASD) Canadian Cardiovascular Society(CCS) and Obesity Canada He serves or has served as
an unpaid scientific advisor for the Food Nutritionand Safety Program (FNSP) and the Technical
Committee on Carbohydrates of the International LifeScience Institute (ILSI) North America He is a member
of the International Carbohydrate Quality Consortium(ICQC) Executive Board Member of the Diabetes and
Nutrition Study Group (DNSG) of the EASD andDirector of the Toronto 3D Knowledge Synthesis and
Clinical Trials foundation His wife is an employee ofSobeys Inc CWCK has received grants or research
support from the Advanced Food Materials NetworkAgriculture and Agri-Foods Canada (AAFC) Almond
Board of California American Pistachio GrowersBarilla Calorie Control Council Canadian Institutes of
Health Research (CIHR) Canola Council of CanadaInternational Nut and Dried Fruit Council
International Tree Nut Council Research andEducation Foundation Loblaw Brands Ltd Pulse
Canada Saskatchewan Pulse Growers and Unilever Hehas received in-kind research support from the AlmondBoard of California American Peanut Council Barilla
California Walnut Commission Kellogg CanadaLoblaw Companies Quaker (PepsiCo) Primo Unico
Unilever and WhiteWave Foods He has received travelsupport andor honoraria from the American Peanut
Council American Pistachio Growers BarillaCalifornia Walnut Commission Canola Council of
Canada General Mills International Nut and DriedFruit Council International Pasta Organization Loblaw
Brands Ltd Nutrition Foundation of Italy OldwaysPreservation Trust Paramount Farms Peanut Institute
Pulse Canada Sabra Dipping Co Saskatchewan PulseGrowers Sun-Maid Tate amp Lyle Unilever and
WhiteWave Foods He has served on the scientific advi-sory board for the International Tree Nut Council
International Pasta Organization McCormick ScienceInstitute Oldways Preservation Trust Paramount
Farms and Pulse Canada He is a member of theInternational Carbohydrate Quality Consortium
(ICQC) Executive Board Member of the Diabetes andNutrition Study Group (DNSG) of the European
Association for the Study of Diabetes (EASD) is on theClinical Practice Guidelines Expert Committee for
Nutrition Therapy of the EASD and is a director of theToronto 3 D Knowledge Synthesis and Clinical Trials
foundation DR is the president of the Croatian
Society for Diabetes and Metabolic Disorders of theCroatian Medical Association He serves as an
Executive Committee member of the CroatianEndocrine Society Croatian Society for Obesity and
Croatian Society for Endocrine Oncology He was aboard member and secretary of IDF Europe and cur-
rently he is the chair of the IDF YLD Program He hasserved as an Executive Committee member of the
Diabetes and Nutrition Study Group of the EASD andcurrently he serves as a Executive Committee member
of Diabetes and Cardiovascular Disease Study Group ofthe EASD He has served as principal investigator or
coinvestigator in clinical trials of AstraZeneca Eli LillyMSD Novo Nordisk Sanofi Aventis Solvay and
Trophos He has received honoraria for speaking or ad-visory board engagements and consulting fees from
Abbott Amgen AstraZeneca Bayer BoehringerIngelheim Eli Lilly Lifescan ndash Johnson amp Johnson
Novartis Novo Nordisk MSD Merck Sharp amp DohmePfizer Pliva Roche Salvus Sanofi Aventis and Takeda
No competing interests were declared by NB-T IP-G and HK
Supporting Information
The following Supporting Information is available
through the online version of this article at the publish-errsquos website
Table S1 Search strategy
Table S2 Characteristics of the included studies
evaluating the association between tree nut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S3 Characteristics of the included studies
evaluating the association between peanut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S4 Characteristics of the included studies
evaluating the association between walnut consump-
tion and the risk of cardiovascular disease outcomes
comparing highest vs lowest categories
Table S5 Characteristics of the included studies
evaluating the association between peanut butter con-
sumption and the risk of cardiovascular disease out-
comes comparing highest vs lowest categories
Table S6 Risk ratio and 95CI from non-linear
dose-response analysis of total nuts and cardiovascu-
lar disease coronary heart disease and stroke
Table S7 Sensitivity analysis by systematic exclu-
sion of one study at a timea
Table S8 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
706 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
assessing the association between tree nut consump-
tion and cardiovascular disease outcomes
Table S9 GRADE assessment of the systematic re-
view and meta-analysis of prospective cohort studies
assessing the association between peanut consump-
tion and cardiovascular disease outcomes
Table S10 GRADE assessment of the systematic
review and meta-analysis of prospective cohort stud-
ies assessing the association between walnut con-
sumption and cardiovascular disease outcomes
Table S11 GRADE assessment of the
systematic review and meta-analysis of prospective
cohort studies assessing the association between pea-
nut butter consumption and cardiovascular disease
outcomes
Figure S1 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease
Figure S2 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S3 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S4 Association between walnut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease
Figure S5 Association between peanut butter con-
sumption (highest vs lowest categories) and the risk
of cardiovascular disease
Figure S6 Association between nut consumption
(highest vs lowest categories) and the risk of cardio-
vascular disease mortality
Figure S7 Association between peanut consump-
tion (highest vs lowest categories) and the risk of car-
diovascular disease mortality
Figure S8 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease
Figure S9 Association between tree nut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S10 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S11 Association between walnut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease
Figure S12 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of coronary heart disease
Figure S13 Association between nut consumption
(highest vs lowest categories) and the risk of coronary
heart disease mortality
Figure S14 Association between peanut consump-
tion (highest vs lowest categories) and the risk of cor-
onary heart disease mortality
Figure S15 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
Figure S16 Association between peanut consump-
tion (highest vs lowest categories) and the risk of stroke
Figure S17 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke
Figure S18 Association between walnut consump-
tion (highest vs lowest categories) and the risk of
stroke
Figure S19 Association between peanut butter
consumption (highest vs lowest categories) and the
risk of stroke
Figure S20 Association between nut consumption
(highest vs lowest categories) and the risk of stroke
mortality
Figure S21 Association between peanut consump-
tion (highest vs lowest categories) and the risk of
stroke death
Figure S22 Association between tree nut con-
sumption (highest vs lowest categories) and the risk
of stroke mortality
Figure S23 Association between nut consumption
(highest vs lowest categories) and the risk of hemor-
rhagic stroke
Figure S24 Association between nut consumption
(highest vs lowest categories) and the risk of ischemic
stroke
Figure S25 Association between nut consumption
(highest vs lowest categories) and the risk of atrial
fibrillation
Figure S26 Association between nut consumption
(highest vs lowest categories) and the risk of heart
failure
Figure S27 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
incidence
Figure S28 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of cardiovascular disease
mortality
Figure S29 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease incidence
Figure S30 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of coronary disease mortality
Figure S31 Linear and non-linear dose-response
relation between increasing total nut intak by 28-g
increments and the risk of stroke incidence
Nutrition ReviewsVR Vol 77(10)691ndash709 707
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
Figure S32 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of stroke mortality
Figure S33 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of hemorrhagic stroke
Figure S34 Linear and non-linear dose-response
relation between increasing total nut intake by 28-g
increments and the risk of ischemic stroke
Figure S35 Subgroup analyses of total nut con-
sumption and the risk of cardiovascular disease
mortality
Figure S36 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of cardiovascular disease mortality
Figure S37 Subgroup analyses of total nut con-
sumption and the risk of coronary heart disease
mortality
Figure S38 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of coronary heart disease mortality
Figure S39 Subgroup analyses of total nut con-
sumption and the risk of stroke mortality
Figure S40 Risk of bias (Newcastle-Ottawa Scale
[NOS]) subgroup analysis of total nut consumption
and the risk of stroke mortality
Figure S41 Funnel plot of the natural logarithm
risk ratio for cardiovascular disease mortality
Figure S42 Funnel plot of the natural logarithm
risk ratio for coronary heart disease mortality
Figure S43 Funnel plot of the natural logarithm
risk ratio for stroke mortality
REFERENCES
1 World Health Organization WHO j Cardiovascular diseases (CVDs) 2017 Availableat httpwwwwhointmediacentrefactsheetsfs317en Accessed March 272018
2 Yu E Malik VS Hu FB Cardiovascular disease prevention by diet modification JAm Coll Cardiol 201872914ndash926
3 Willett WC Sacks F Trichopoulou A et al Mediterranean diet pyramid a culturalmodel for healthy eating Am J Clin Nutr 1995611402Sndash1406S
4 Sacks FM Obarzanek E Windhauser MM et al Rationale and design of theDietary Approaches to Stop Hypertension trial (DASH) A multicenter controlled-feeding study of dietary patterns to lower blood pressure Ann Epidemiol19955108ndash118
5 Craig WJ Mangels AR American Dietetic Association Position of the AmericanDietetic Association vegetarian diets J Am Diet Assoc 20091091266ndash1282
6 Ros E Mataix J Fatty acid composition of nutsndashimplications for cardiovascularhealth Br J Nutr 200696S29ndashS35
7 Segura R Javierre C Lizarraga MA et al Other relevant components of nuts phy-tosterols folate and minerals Br J Nutr 200696S36ndashS44
8 Afshin A Micha R Khatibzadeh S et al Consumption of nuts and legumes andrisk of incident ischemic heart disease stroke and diabetes a systematic reviewand meta-analysis Am J Clin Nutr 2014100278ndash288
9 Aune D Keum N Giovannucci E et al Nut consumption and risk of cardiovas-cular disease total cancer all-cause and cause-specific mortality a system-atic review and dose-response meta-analysis of prospective studies BMCMed 201614207
10 Bechthold A Boeing H Schwedhelm C et al Food groups and risk of coro-nary heart disease stroke and heart failure a systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2017591071ndash1090
11 Chen G-C Zhang R Martınez-Gonzalez MA et al Nut consumption in relation toall-cause and cause-specific mortality a meta-analysis 18 prospective studiesFood Funct 201783893ndash3905
12 Grosso G Yang J Marventano S et al Nut consumption on all-cause cardiovascu-lar and cancer mortality risk a systematic review and meta-analysis of epidemio-logic studies Am J Clin Nutr 2015101783ndash793
13 Luo C Zhang Y Ding Y et al Nut consumption and risk of type 2 diabetes cardio-vascular disease and all-cause mortality a systematic review and meta-analysisAm J Clin Nutr 2014100256ndash269
14 Ma L Wang F Guo W et al Nut consumption and the risk of coronary artery dis-ease a dose-response meta-analysis of 13 prospective studies Thromb Res2014134790ndash794
15 Mayhew AJ de Souza RJ Meyre D et al A systematic review and meta-analysis ofnut consumption and incident risk of CVD and all-cause mortality Br J Nutr2016115212ndash225
16 Shao C Tang H Zhao W et al Nut intake and stroke risk a dose-response meta-analysis of prospective cohort studies Sci Rep 2016630394
17 Shi ZQ Tang JJ Wu H et al Consumption of nuts and legumes and risk of strokea meta-analysis of prospective cohort studies Nutr Metab Cardiovasc Dis2014241262ndash1271
18 Weng Y-Q Yao J Guo M-L et al Association between nut consumption and coro-nary heart disease Coron Artery Dis 201627227ndash232
19 Zhang Z Xu G Wei Y et al Nut consumption and risk of stroke Eur J Epidemiol201530189ndash196
20 Zhou D Yu H He F et al Nut consumption in relation to cardiovascular diseaserisk and type 2 diabetes a systematic review and meta-analysis of prospectivestudies Am J Clin Nutr 2014100270ndash277
21 Guasch-Ferre M Liu X Malik VS et al Nut consumption and risk of cardiovasculardisease J Am Coll Cardiol 2017702519ndash2532
22 Larsson SC Drca N Bjorck M et al Nut consumption and incidence of seven car-diovascular diseases Heart 20181041615ndash1620
23 Higgins JP Green S eds Cochrane Handbook for Systematic Reviews ofInterventions Version 510 United Kingdom The Cochrane Collaboration 2011
24 Stroup DF Berlin JA Morton SC et al Meta-analysis of observational studies in ep-idemiology a proposal for reporting Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group JAMA 20002832008ndash2012
25 Liu G Guasch-Ferre M Hu Y et al Nut consumption in relation to cardiovasculardisease incidence and mortality among patients with diabetes mellitus Circ Res2019124920ndash929
26 van den Brandt PA Schouten LJ Relationship of tree nut peanut and peanut but-ter intake with total and cause-specific mortality a cohort study and meta-analy-sis Int J Epidemiol 2015441038ndash1049
27 van den Brandt PA Red meat processed meat and other dietary protein sourcesand risk of overall and cause-specific mortality in The Netherlands Cohort StudyEur J Epidemiol 201934351ndash369 doi101007s10654-019-00483-9
28 Wells G Shea B OrsquoConnell D et al The NewcastlendashOttawa Scale (NOS) for assess-ing the quality of nonrandomised studies in meta-analyses Ottawa OttawaHospital Research Institute 2014 Available at wwwohricaprogramsclinical_epi-demiologyoxfordasp Accessed April 11 2018
29 Fraser GE Sabate J Beeson WL et al A possible protective effect of nut consump-tion on risk of coronary heart disease The Adventist Health Study Arch InternMed 19921521416ndash1424
30 Albert CM Gaziano JM Willett WC et al Nut consumption and decreased risk ofsudden cardiac death in the Physiciansrsquo Health Study Arch Intern Med20021621382ndash1387
31 Luu HN Blot WJ Xiang Y-B et al Prospective evaluation of the association of nutpeanut consumption with total and cause-specific mortality JAMA Intern Med2015175755ndash766
32 di Giuseppe R Fjeld MK Dierkes J et al The association between nut consump-tion and the risk of total and ischemic stroke in a German cohort study Eur J ClinNutr 201569431ndash435
33 Hamling J Lee P Weitkunat R et al Facilitating meta-analyses by deriving relativeeffect and precision estimates for alternative comparisons from a set of estimatespresented by exposure level or disease category Stat Med 200827954ndash970
34 Greenland S Longnecker MP Methods for trend estimation from summarizeddose-response data with applications to meta-analysis Am J Epidemiol19921351301ndash1309
35 Orsini N Bellocco R Greenland S Generalized least squares for trend estimationof summarized dosendashresponse data Stata J 2006640ndash57
36 Desquilbet L Mariotti F Dose-response analyses using restricted cubic spline func-tions in public health research Stat Med 2010291037ndash1057
37 Guyatt G Oxman AD Akl EA et al GRADE guidelines 1 Introduction-GRADE evi-dence profiles and summary of findings tables J Clin Epidemiol 201164383ndash394
38 Guyatt GH Oxman AD Vist G et al GRADE guidelines 4 Rating the quality ofevidencendashstudy limitations (risk of bias) J Clin Epidemiol 201164407ndash415
39 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 7 Rating the quality of ev-idencemdashinconsistency J Clin Epidemiol 2011641294ndash1302
40 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 8 Rating the quality of ev-idencemdashindirectness J Clin Epidemiol 2011641303ndash1310
708 Nutrition ReviewsVR Vol 77(10)691ndash709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019
41 Guyatt GH Oxman AD Kunz R et al GRADE guidelines 6 Rating the quality of ev-idencemdashimprecision J Clin Epidemiol 2011641283ndash1293
42 Guyatt GH Oxman AD Montori V et al GRADE guidelines 5 Rating the quality ofevidencemdashpublication bias J Clin Epidemiol 2011641277ndash1282
43 Guyatt GH Oxman AD Sultan S et al GRADE guidelines 9 Rating up the qualityof evidence J Clin Epidemiol 2011641311ndash1316
44 Blomhoff R Carlsen MH Andersen LF et al Health benefits of nuts potential roleof antioxidants Br J Nutr 200696S52ndashS60
45 Bao Y Han J Hu FB et al Association of nut consumption with total and cause-specific mortality N Engl J Med 20133692001ndash2011
46 Guasch-Ferre M Bullo M Martınez-Gonzalez MA et al Frequency of nut con-sumption and mortality risk in the PREDIMED nutrition intervention trial BMCMed 201311164 doi1011861741-7015-11-164
47 Hshieh TT Petrone AB Gaziano JM et al Nut consumption and risk of mortality inthe Physiciansrsquo Health Study Am J Clin Nutr 2015101407ndash412
48 Bonaccio M Di Castelnuovo A De Curtis A et al Nut consumption is in-versely associated with both cancer and total mortality in a Mediterraneanpopulation prospective results from the Moli-sani study Br J Nutr2015114804ndash811
49 Eslamparast T Sharafkhah M Poustchi H et al Nut consumption and total andcause-specific mortality results from the Golestan Cohort Study Int J Epidemiol20174675ndash85
50 Haring B Gronroos N Nettleton JA et al Dietary protein intake and coronaryheart disease in a large community based cohort results from the AtherosclerosisRisk in Communities (ARIC) study PLoS One 20149e109552
51 Gopinath B Flood VM Burlutksy G et al Consumption of nuts and risk of totaland cause-specific mortality over 15 years Nutr Metab Cardiovasc Dis2015251125ndash1131
52 Djousse L Gaziano JM Kase CS et al Nut consumption and risk of stroke in USmale physicians Clin Nutr 201029605ndash609
53 Haring B Misialek JR Rebholz CM et al Association of dietary protein consump-tion with incident silent cerebral infarcts and stroke the Atherosclerosis Risk inCommunities (ARIC) Study Stroke 2015463443ndash3450
54 Bernstein AM Pan A Rexrode KM et al Dietary protein sources and the risk ofstroke in men and women Stroke 201243637ndash644
55 Khawaja O Gaziano JM Djousse L Nut consumption and risk of atrial fibrillationin the Physiciansrsquo Health Study Nutr J 20121117
56 Djousse L Rudich T Gaziano JM Nut consumption and risk of heart failure in thePhysiciansrsquo Health Study I Am J Clin Nutr 200888930ndash933
57 Nettleton JA Steffen LM Loehr LR et al Incident heart failure is associated withlower whole-grain intake and greater high-fat dairy and egg intake in theAtherosclerosis Risk in Communities (ARIC) study J Am Diet Assoc20081081881ndash1887
58 Yaemsiri S Sen S Tinker L et al Trans fat aspirin and ischemic stroke in postmen-opausal women Ann Neurol 201272704ndash715
59 von Ruesten A Feller S Bergmann MM et al Diet and risk of chronic diseasesresults from the first 8 years of follow-up in the EPIC-Potsdam study Eur J ClinNutr 201367412ndash419
60 Wang J-B Fan J-H Dawsey SM et al Dietary components and risk of total cancerand cardiovascular disease mortality in the Linxian Nutrition Intervention Trials co-hort in China Sci Rep 2016622619
61 Sabate J Oda K Ros E Nut consumption and blood lipid levels Arch Intern Med2010170821ndash827
62 Del Gobbo LC Falk MC Feldman R et al Effects of tree nuts on blood lipids apoli-poproteins and blood pressure systematic review meta-analysis and dose-response of 61 controlled intervention trials Am J Clin Nutr 20151021347ndash1356
63 Coates A Hill A Tan S Nuts and cardiovascular disease prevention CurrAtheroscler Rep 20182048
64 Bitok E Sabate J Nuts and cardiovascular disease Prog Cardiovasc Dis20186133ndash37
Nutrition ReviewsVR Vol 77(10)691ndash709 709
Dow
nloaded from httpsacadem
icoupcomnutritionreview
sarticle-abstract77106915540978 by guest on 25 September 2019