Socio-Economic Differences in Mortality by Cause of Deathandrewc/ARCresources/Cairns... · 11...
Transcript of Socio-Economic Differences in Mortality by Cause of Deathandrewc/ARCresources/Cairns... · 11...
,
Socio-Economic Differences in Mortality by Causeof Death
Andrew J.G. Cairns
Heriot-Watt University, Edinburgh
Director, Actuarial Research Centre, IFoA
Joint work with D. Blake, K. Dowd, M. Kallestrup-Lamb, C. Rosenskjold, C. Redondo, A.S.Macdonald
Longevity 13, Taipei, September 2017
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 1 / 38
,
The Actuarial Research Centre (ARC)A gateway to global actuarial research
The Actuarial Research Centre (ARC) is the Institute and Faculty ofActuaries’ (IFoA) network of actuarial researchers around the world.The ARC seeks to deliver cutting-edge research programmes that addresssome of the significant, global challenges in actuarial science, through apartnership of the actuarial profession, the academic community andpractitioners.The ’Modelling, Measurement and Management of Longevity andMorbidity Risk’ research programme is being funded by the ARC, theSoA and the CIA.
www.actuaries.org.uk/arc
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 2 / 38
,
Outline
Danish data:
affluenceeducationcause of death
Statistical significance
US cause of death data by education group
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 3 / 38
,
Purpose of looking at cause of death data
What are the key drivers of all-cause mortality?
How are the key drivers changing over time?Which causes of death have high levels of inequality:
by education;by affluence?
Beware of
changes in ICD classification of deathsdrift in how deaths are classifiedchanging education levels (grade inflation)
Insight into mortality underpinning life insurance andpensions
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 4 / 38
,
Danish Data: Cairns et al. Longevity 11, Lyon
Statistics Denmark National Register Database
Key data (amongst others) for each individual:
Date of birth (⇒ age)Date of deathWealthIncomeAffluence=Wealth+15×IncomeEducationCause of Death
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 5 / 38
,
Education and Affluence Levels
EducationLow education Primary and lower secondary educationMedium education Upper secondary educationHigh education Tertiary education
AffluenceLevel 1 Low affluence decile
......
Level 10 High affluence decile
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 6 / 38
,
Model-Inferred Underlying Death Rates 2012
60 70 80 90
0.00
20.
010
0.05
00.
200
Group 1Group 2Group 3Group 4Group 5Group 6Group 7Group 8Group 9Group 10
Males Crude m(t,x); 2012
Age
m(t
,x)
(log
scal
e)
60 70 80 90
0.00
20.
010
0.05
00.
200
Group 1Group 2Group 3Group 4Group 5Group 6Group 7Group 8Group 9Group 10
Males CBD−X Fitted m(t,x); 2012Point Estimates
Age
m(t
,x)
(log
scal
e)
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 7 / 38
,
Education as an Alternative Covariate
Level of Educational Attainment also known to be a goodpredictor
Various US studies
Mackenbach et al. (2003) including Denmark: Std. Mortality RatesBrønnum-Hansen and Baadsgaard (2012) Denmark: LE(x = 30)
As close as possible on a like for like basis:
Crude death rates; age 30+; matching years.
Affluence ⇒Wider spread of SMR’s than M. et al. (2003)Wider spread of LE (30) than BHB (2012)
Issue: “grade inflation” distorts results
More to be done.
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 8 / 38
,
Education as an Alternative Covariate
1985 1990 1995 2000 2005 2010
05
1015
20
Year
Age
Sta
ndar
dise
d M
orta
lity
Rat
e (
per
1000
per
son
year
s)
Affluence Group 1Low EducationHigh EducationAffluence Group 10
Age Standardised Mortality Rates per 1000Ages 45−54; European Standard Population (1976)
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 9 / 38
,
Cause of Death Data – Health Inequalities
Deaths subdivided into 29 CoD groups
Age groups 31-35, 36-40, ..., 91-95
Year groups 1985-89, 1990-94, 1995-99,2000-2004, 2005-2009
Compare affluence groups
Compare education groups
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 10 / 38
,
Cause of Death Data – Health Inequalities
1 Infectious diseases incl. tuberculosis 2 Cancer: mouth, gullet, stomach3 Cancer: gut, rectum 4 Cancer: lung, larynx, ..5 Cancer: breast 6 Cancer: uterus, cervix7 Cancer: prostate, testicular 8 Cancer: bones, skin9 Cancer: lymphatic, blood-forming tissue 10 Benign tumours11 Diseases: blood 12 Diabetes13 Mental illness 14 Meningitis + nervous system (Alzh.)15 Blood pressure + rheumatic fever 16 Ischaemic heart diseases17 Other heart diseases 18 Diseases: cerebrovascular19 Diseases: circulatory 20 Diseases: lungs, breathing21 Diseases: digestive 22 Diseases: urine, kidney,...23 Diseases: skin, bone, tissue 24 Senility without mental illness25 Road/other accidents 26 Other causes27 Alcohol → liver disease 28 Suicide29 Accidental Poisonings
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 11 / 38
,
Denmark: Cause of Death Data 2007
Compare education with affluence as covariates:
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by EducationIschaemic heart diseases
Age Group
Dea
th R
ate
(log
scal
e)
Low EduMediumHigh
40 50 60 70 80 901e
−05
1e−
031e
−01
Danish Males by AffluenceIschaemic heart diseases
Age Group
Dea
th R
ate
(log
scal
e)
12345678910
Affluence ⇒ wider spread
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 12 / 38
,
Denmark: Cause of Death Data 2007
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by EducationDiabetes
Age Group
Dea
th R
ate
(log
scal
e)
Low EduMediumHigh
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by AffluenceDiabetes
Age GroupD
eath
Rat
e (lo
g sc
ale)
12345678910
Affluence ⇒ much wider
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 13 / 38
,
Denmark: Cause of Death Data 2007
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by EducationCancer: lung, larynx, ..
Age Group
Dea
th R
ate
(log
scal
e)
Low EduMediumHigh
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by AffluenceCancer: lung, larynx, ..
Age GroupD
eath
Rat
e (lo
g sc
ale)
12345678910
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 14 / 38
,
Denmark: Cause of Death Data 2007
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by EducationAlcohol −> liver
Age Group
Dea
th R
ate
(log
scal
e)
Low EduMediumHigh
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by AffluenceAlcohol −> liver
Age GroupD
eath
Rat
e (lo
g sc
ale)
12345678910
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 15 / 38
,
Denmark: Cause of Death Data 2007
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by EducationAccidental Poisonings
Age Group
Dea
th R
ate
(log
scal
e)
Low EduMediumHigh
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by AffluenceAccidental Poisonings
Age GroupD
eath
Rat
e (lo
g sc
ale)
12345678910
Low affluence ⇒ over 20× at young ages
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 16 / 38
,
Denmark: Cause of Death Data 2007
Many causes of death have known risk factors ordriverse.g. smoking, diet, healthy lifestyle etc.⇒ clear socio-economic differences
Biggest differences at ages < 60
Affluence ⇒ stronger predictor than education(sometimes very much stronger)
Other diseases do not have strong differences:
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 17 / 38
,
Denmark: Cause of Death Data 2007
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by EducationCancer: prostate, testicular
Age Group
Dea
th R
ate
(log
scal
e)
Low EduMediumHigh
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by AffluenceCancer: prostate, testicular
Age GroupD
eath
Rat
e (lo
g sc
ale)
12345678910
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 18 / 38
,
Denmark: Cause of Death Data 2007
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by EducationCancer: lymphatic, blood−forming tissue
Age Group
Dea
th R
ate
(log
scal
e)
Low EduMediumHigh
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by AffluenceCancer: lymphatic, blood−forming tissue
Age GroupD
eath
Rat
e (lo
g sc
ale)
12345678910
Education ⇒ no effectAffluence ⇒ small effect
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 19 / 38
,
Denmark: Cause of Death Data 2007
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by EducationCancer: gut, rectum
Age Group
Dea
th R
ate
(log
scal
e)
Low EduMediumHigh
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Danish Males by AffluenceCancer: gut, rectum
Age GroupD
eath
Rat
e (lo
g sc
ale)
12345678910
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 20 / 38
,
Denmark: Cause of Death Data – Health Inequalities
Some causes of death have no obvious link tolifestyle/affluence/educatione.g. Prostate CancerCancerUK:Prostate cancer is not clearly linked to anypreventable risk factors.
But Affluence ⇒ inequalities
Possible explanations (a very non-expert view)
onset is not dependent on lifestyle/affluence/educationBUT less affluent/educated ⇒
??? later diagnosis??? engage less well with treatment process??? lower quality housing
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 21 / 38
,
CoD Death Rates: Different Shapes & Patterns
40 50 60 70 80 90
1e−
051e
−04
1e−
031e
−02
1e−
01Infectious diseases incl. tuberculosis
Dea
th R
ate
(log
scal
e)
12345678910
40 50 60 70 80 90
1e−
051e
−04
1e−
031e
−02
1e−
01
Meningitis + nervous system (Alzh.)
Dea
th R
ate
(log
scal
e)
12345678910
40 50 60 70 80 90
1e−
051e
−04
1e−
031e
−02
1e−
01
Ischaemic heart diseases
Dea
th R
ate
(log
scal
e)
12345678910
40 50 60 70 80 90
1e−
051e
−04
1e−
031e
−02
1e−
01
Diseases: circulatory
Dea
th R
ate
(log
scal
e)
12345678910
40 50 60 70 80 90
1e−
051e
−04
1e−
031e
−02
1e−
01Diseases: lungs, breathing
Dea
th R
ate
(log
scal
e)
12345678910
40 50 60 70 80 90
1e−
051e
−04
1e−
031e
−02
1e−
01
Diseases: urine, kidney,...
Dea
th R
ate
(log
scal
e)
12345678910
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 22 / 38
,
CoD Death Rates: Different Shapes & Patterns
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Cancer: gut, rectum
Dea
th R
ate
(log
scal
e)
12345678910
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Cancer: lung, larynx, ..
Dea
th R
ate
(log
scal
e)
12345678910
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Cancer: prostate, testicular
Dea
th R
ate
(log
scal
e)
12345678910
40 50 60 70 80 90
1e−
051e
−03
1e−
01
Cancer: bones, skin
Dea
th R
ate
(log
scal
e)
12345678910
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 23 / 38
,
Shapes: Conclusions
Typically:Non-cancerous diseases ⇒ approximately exponentialgrowthNeoplasms (cancers) ⇒ subexponential ???polynomial
What does this reveal about different diseasemechanisms?
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 24 / 38
,
Denmark Males: Statistical Significance
Which CoD’s are significantly affected bysocio-economic status?
H0: Affluence groups all have the same CoDdeath rate mi(c , t, x) = mj(c , t, x) ∀i 6= jversus
H1: Affluence groups do not all have the sameCoD death rates
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 25 / 38
,
Denmark Males: Statistical Significance
For each cause of death (29), and age group (13)
Rank the death rates for the 10 groups i = 1, . . . , 10
For each year group, tR(i , t) = rank of m(i , t) out of m(1, t), . . . ,m(10, t)Rank 1: highest death rateRank 10: lowest death rate
Data (i ,R(i , t))
Test statistic, S = cor(i ,R(i , t)
)Under H0 the ranks are a random permutation of 1, . . . , 10
Under H0, S is approximately N(0, σ2) where σ = 0.149.
One-sided test: Reject H0 if S > σΦ−1(α)
Large S ⇒ low affluence ∼ high CoD mortality
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 26 / 38
,
Cause of Death Inequalities: p-values
p−values for Danish Male Affluence Group Rankings
CoD Group
Age
Ban
d
2
4
6
8
10
12
5 10 15 20 25
0
0.0010.0050.010.05
1
Very low or zero mortality: CoD 5, 6, 24 & low agesHigh age convergence
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 27 / 38
,
US Education Data
Males and Females (2)
Single ages 55-75 (21)
Single years 1989-2015 (27)
Causes of death (29)
Low, medium & high education level (3)
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 28 / 38
,
US Education Data: Growing Inequality
US Males All Cause MortalityRatio of Low to High Education Mortality
Year 1989−2015
Age
555657585960616263646566676869707172737475
1990 1994 1998 2002 2006 2010 2014
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 29 / 38
,
US Education Data
55 60 65 70 75
1e−
051e
−04
1e−
031e
−02
1e−
01 Low EduMediumHigh
Year 2000 Cancer: prostate, testicular
Age
CoD
Dea
th R
ate
55 60 65 70 75
1e−
051e
−04
1e−
031e
−02
1e−
01 Low EduMediumHigh
Year 2015 Cancer: prostate, testicular
Age
CoD
Dea
th R
ate
Recall: Denmark ⇒ very narrow gap
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 30 / 38
,
US Education Data
55 60 65 70 75
1e−
051e
−04
1e−
031e
−02
1e−
01 Low EduMediumHigh
Year 2000 Ischaemic heart diseases
Age
CoD
Dea
th R
ate
55 60 65 70 75
1e−
051e
−04
1e−
031e
−02
1e−
01 Low EduMediumHigh
Year 2015 Ischaemic heart diseases
Age
CoD
Dea
th R
ate
Widening gap
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 31 / 38
,
US Education Data
55 60 65 70 75
1e−
051e
−04
1e−
031e
−02
1e−
01 Low EduMediumHigh
Year 2000 Meningitis + nervous system (Alzh.)
Age
CoD
Dea
th R
ate
55 60 65 70 75
1e−
051e
−04
1e−
031e
−02
1e−
01 Low EduMediumHigh
Year 2015 Meningitis + nervous system (Alzh.)
Age
CoD
Dea
th R
ate
Widening gap
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 32 / 38
,
US Education Data
55 60 65 70 75
1e−
051e
−04
1e−
031e
−02
1e−
01 Low EduMediumHigh
Year 2000 Accidental Poisonings
Age
CoD
Dea
th R
ate
55 60 65 70 75
1e−
051e
−04
1e−
031e
−02
1e−
01 Low EduMediumHigh
Year 2015 Accidental Poisonings
Age
CoD
Dea
th R
ate
Case & Deaton (2015) ⇒ Accidental poisoning ↗
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 33 / 38
,
US Education Data
55 60 65 70 75
1e−
051e
−04
1e−
031e
−02
1e−
01 Low EduMediumHigh
Year 2000 Alcohol −> liver
Age
CoD
Dea
th R
ate
55 60 65 70 75
1e−
051e
−04
1e−
031e
−02
1e−
01 Low EduMediumHigh
Year 2015 Alcohol −> liver
Age
CoD
Dea
th R
ate
Widening gap
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 34 / 38
,
US Males: Low versus High Education
Do Low and High education groups have the sameCoD rate?
Four × 5-year age groups
29 causes of death
Signs Test (count low edu. > high edu. mort.)
29× 4 = 116 individual tests
115/116 ⇒ reject H0 equality
Accept H0 (p = 0.08) for only one pairing(Meningitis + nervous system (Alzh.), 70-74)
Most p-values < 10−6
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 35 / 38
,
4. Summary
Affluence better than education for all CoD ifyou have the data
Impact of affluence/education varies with CoD
Different growth patterns cancers versus otherdiseases
Work in progress!
E: [email protected] W: www.macs.hw.ac.uk/∼andrewc
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 36 / 38
,
Thank You!
Questions?
E: [email protected] W: www.macs.hw.ac.uk/∼andrewc
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 37 / 38
,
Education as an Alternative Covariate
Dig a bit deeper:Affluence + Education: average ASMR’s over 5 years
Mortality Improvement Rates (%)Period 1987−2009; Age Band 45−54By Affluence and Education Group
Affluence Group
Edu
catio
n Le
vel
LOW
MED
HIGH
1 2 3 4 5 6 7 8 9 10
−1012345
Andrew J.G. Cairns Socio-Economic Differences in Mortality by Cause of Death 38 / 38