Post on 30-Dec-2015
1
Neonatal vitamin A supplementation interacts with routine immunizations in
infancy - with consequences for mortality
Christine Stabell Benn, MD, PhDPeter Aaby, Ane Fisker, Carlitos Balé, Amabelia Rodrigeus,
and many moreBandim Health Project
INDEPTH NetworkStatens Serum Institut
Guinea-Bissau & Denmark
Disclosure: No competing financial interests
2
Background
• Vitamin A deficiency associated with increased mortality
• Randomised trials in late 80’s-early 90’s: Vitamin A supplementation associated with 23-30% reduction in overall mortality in children > 6 mo of age
• WHO policy: High-dose vitamin A supplements every 4-6 months to all children 6 mo- 5 yr in low-income countries - preferably linked to the immunization program (EPI) for logistic reasons
• Overall effect of WVO policy never tested in randomised trial
• Our hypothesis: Vitamin A supplementation beneficial when given with the live BCG and measles vaccine, but harmful with inactivated diphtheria-tetanus-pertussis (DTP) vaccine
3
Neonatal vitamin A supplementation
• Seven trials have been conducted:– Asia: Nepal (West 1995), Indonesia (Humphrey 1996),
India (Rahmathullah 2003), and Bangladesh (Klemm 2008)
– Africa: Zimbabwe (Malaba, Humphrey 2005/6), Guinea-Bissau (Benn 2008 and Benn, 2010)
Gogia BMJ 2009Heterogeneity
No effect
4
Neonatal vitamin A supplementation trials Guinea-Bissau
5
Vitamin A supplementation versus placebowith BCG to neonates in Guinea-Bissau
Hypothesis: BCG=
Benn et al. BMJ 2010
0.00
0.01
0.02
0.03
0.04
0.05
0.06
Cu
mul
ativ
e m
orta
lity
2066 2023 1758 1445Placebo1998 1926 1661 1368Vitamin A
Number at risk
0 1 2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
Normal-birth-weight: 1.07 (0.79-1.44)
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Cu
mul
ativ
e m
orta
lity
821 796 780 731 721 650Placebo (N=863)808 788 772 732 721 645Vitamin A (N=854)
Number at risk
0 1 2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
Meta-estimate of the two trials in Guinea-Bissau:
MRR=1.08 (0.87-1.33)
Low-birth-weight: 1.08 (0.79-1.47)
Benn et al. BMJ 2008
Vitamin A supplementation at birth and mortality by sex0.0
00.0
10.0
20.0
30.0
40.0
50.0
6C
um
ula
tive m
ort
alit
y
0 1 2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
Boys
0.0
00.0
10.0
20.0
30.0
40.0
50.0
6C
um
ula
tive m
ort
alit
y
0 1 2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
Girls
MRR=0.8 (0.6-1.3)
MRR=1.4 (0.9-2.1)
P=0.10 for interaction
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Cu
mul
ativ
e m
orta
lity
0 1 2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Cu
mul
ativ
e m
orta
lity
0 1 2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
MRR=0.7 (0.5-1.2) MRR=1.4 (0.9-2.2)
P=0.04 for interaction
Low-birth-weight:
Normal-birth-weight:
Vitamin A supplementation at birth and mortality by sex0.0
00.0
10.0
20.0
30.0
40.0
50.0
6C
um
ula
tive m
ort
alit
y
0 1 2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
Boys
0.0
00.0
10.0
20.0
30.0
40.0
50.0
6C
um
ula
tive m
ort
alit
y
0 1 2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
Girls
MRR=0.8 (0.6-1.3)
MRR=1.4 (0.9-2.1)
P=0.10 for interaction
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Cu
mul
ativ
e m
orta
lity
0 1 2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Cu
mul
ativ
e m
orta
lity
0 1 2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
MRR=0.7 (0.5-1.2) MRR=1.4 (0.9-2.2)
P=0.04 for interaction
Low-birth-weight:
Normal-birth-weight:
Meta-estimates of the two trials in Guinea-BissauBoys: 0.80 (0.58-1.09) Girls: 1.41 (1.04-1.90) P for interaction=0.01
8
Vitamin A at birth associated with higher mortality than placebo in girls when they receive DTP vaccine
Benn et al. IJE 2009
0.00
0.01
0.02
0.03
0.04
Cu
mul
ativ
e m
orta
lity
745 863 890 782 728 695 658 463Placebo695 783 821 721 662 639 614 414Vitamin A
Number at risk
2 3 4 5 6 7 8 9Age in months
Vitamin A boys Placebo boys
Boys
MRR=0.9 (0.4-1.8)
0.00
0.01
0.02
0.03
0.04
Cu
mul
ativ
e m
orta
lity
705 821 849 760 707 686 652 443Placebo655 774 821 688 643 607 574 354Vitamin A
Number at risk
2 3 4 5 6 7 8 9Age in months
Vitamin A girls Placebo girls
Girls
MRR=2.2 (1.1-4.4)P=0.08 for interaction
Normal-birth-weight:
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Cu
mul
ativ
e m
orta
lity
2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
Girls: Mortality by treatment while DTP is last vaccine
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Cu
mul
ativ
e m
orta
lity
2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
Boys: Mortality by treatment while DTP is last vaccine
MRR=0.7 (0.3-1.6) MRR=1.4 (0.8-2.8)
P=0.16 for interaction
Low-birth-weight:
9
Vitamin A at birth associated with higher mortality than placebo in girls when they receive DTP vaccine
Benn et al. IJE 2009
0.00
0.01
0.02
0.03
0.04
Cu
mul
ativ
e m
orta
lity
745 863 890 782 728 695 658 463Placebo695 783 821 721 662 639 614 414Vitamin A
Number at risk
2 3 4 5 6 7 8 9Age in months
Vitamin A boys Placebo boys
Boys
MRR=0.9 (0.4-1.8)
0.00
0.01
0.02
0.03
0.04
Cu
mul
ativ
e m
orta
lity
705 821 849 760 707 686 652 443Placebo655 774 821 688 643 607 574 354Vitamin A
Number at risk
2 3 4 5 6 7 8 9Age in months
Vitamin A girls Placebo girls
Girls
MRR=2.2 (1.1-4.4)P=0.08 for interaction
Normal-birth-weight:
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Cu
mul
ativ
e m
orta
lity
2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
Girls: Mortality by treatment while DTP is last vaccine
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Cu
mul
ativ
e m
orta
lity
2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
Boys: Mortality by treatment while DTP is last vaccine
MRR=0.7 (0.3-1.6) MRR=1.4 (0.8-2.8)
P=0.16 for interaction
Low-birth-weight:
Vitamin A supplementation at birth negative effect on survival after DTP vaccination in girls
Meta-estimate all girls after DTP: MRR=1.8 (1.1-2.8) p=0.02
100.
880.
900.
920.
940.
960.
981.
00S
urvi
val p
roba
bilit
y0 1 2 3 4 5 6 7 8 9 10 11 12
Age in months
Vitamin A Placebo
0.9
40.9
50.9
60.9
70.9
80.9
91.0
0
0 1 2 3 4 5 6 7 8 9 10 11 12Age in months
Vitamin A Placebo
Guinea-Bissau 2002-8 Zimbabwe 1997-2001
Indonesia 1992-4
Vitamin A
Placebo
India 1998-2001
Vitamin A
Placebo
Bangladesh 2004-7
11
Our interpretation of current evidence
• Neonatal VAS beneficial during the first months of life
• The effect may shift when the children receive DTP– areas with high mortality throughout infancy– areas with high DTP coverage – areas which follow the WHO recommended vaccination
schedule of first BCG and then DTP
• Of utmost importance that the three new WHO/Gates sponsored trials in Kintambo, Ifakara and India follow children to 12 months of age and analyse data by sex and vaccination status
12
Neonatal vitamin A supplementation and early measles vaccine (MV) trial
Guinea-Bissau
13
VAS versus Placebo at birth Children who had DTP3 at 4 mo and MV at 9 mo
From 4-8 months: DTP
RR VAS/placebo=1.1 (0.5-2.2)Girls: VAS/Placebo=1.7 (0.7-
4.3)
From 9-36 months: MV
RR VAS/placebo=0.8 (0.5-1.2)Girls: VAS/Placebo=0.5 (0.3-
1.0)
0
2
4
6
8
10
12
14
All Boys Girls
Vitamin A
Placebo
0
10
20
30
40
All Boys Girls
Vitamin A
Placebo
Significant inversion of VAS effect in girls from DTP to MV, p=0.04
14
VAS versus Placebo at birth Children who had MV at 4 mo and MV at 9 mo
From 4-8 months: 1st MV
RR VAS/placebo negative effect,
P=0.004
From 9-36 months: 2nd MV
RR VAS/placebo=1.6 (0.8-3.5)0
10
20
30
40
All Boys Girls
Vitamin A
Placebo
0
2
4
6
8
10
12
14
All Boys Girls
Vitamin A
Placebo
No deaths in placebo group
Overall VAS effect from 4-36 months: MRR=2.5 (1.2-5.3)
15
Summary: Neonatal vitamin A and vaccines
• Negative interaction between neonatal vitamin A and DTP, even given months apart, in girls (RR=1.8 (1.1-2.8)
• Negative interaction between vitamin A and early MV at 4 mo of age (4 weeks after DTP) (RR=2.5 (1.2-5.3))
• Potentially positive interaction between VAS and MV at 9 months of age (5 months after DTP) in girls (RR=0.5 (0.3-1.0))
16
Conclusions
• Vitamin A protects against vitamin A deficiency and thereby against mortality
• Vitamin A is also an immuno-modulator and the effect on mortality depends on what is going on in the immune system
• Vitamin A may be harmful in certain situations: 1-5 month-old-children, respiratory infections, +DTP vaccine in girls, +early measles vaccine
• Vitamin A may have sex-differential effects
• We can optimise the use of vitamin A supplementation if we take the immuno-modulatory sex-differential effects into account
17
Further information or questions?
• Please contact Christine Stabell Benn at
E-mail: cb@ssi.dkTel: +45 3268 8354
Thank you for your attention!