Christine Stabell Benn, MD, PhD

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)


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)


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)


Low-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


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


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)


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


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


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


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)


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: [email protected]: +45 3268 8354

Thank you for your attention!

mailto:[email protected]

Documents

Christine Stabell Benn, MD, PhD