Jump to comment:

• Re:There is no need for further placebo-controlled trials
  Ane B. Fisker
  Published on: 23 November 2015
• There is no need for further placebo-controlled trials
  Evan Mayo-Wilson
  Published on: 23 November 2015

• Published on: 23 November 2015

  Re:There is no need for further placebo-controlled trials

  • Ane B. Fisker, Post doc
  • Other Contributors:
    • Peter Aaby, Christine S. Benn

  We thank Mayo-Wilson and colleagues for their comments(1). They have used our paper on the effect of vitamin A campaigns on mortality(2) not to question the paper itself, but to repeat a previous discussion(3, 4) about whether new trials of vitamin A supplementation (VAS) are needed to optimise the vitamin A policy. This is clearly a very important discussion.

  We think that new trials are needed. We have found i...

  Show More

  We thank Mayo-Wilson and colleagues for their comments(1). They have used our paper on the effect of vitamin A campaigns on mortality(2) not to question the paper itself, but to repeat a previous discussion(3, 4) about whether new trials of vitamin A supplementation (VAS) are needed to optimise the vitamin A policy. This is clearly a very important discussion.

  We think that new trials are needed. We have found in several studies - including the present(2) - that VAS may interact negatively with routine vaccinations producing no overall benefit of VAS on child survival or a negative effect in some subgroups, e.g. children who have received diphtheria-tetanus-pertussis vaccine (DTP) as their most recent vaccination(5-7). We have emphasised that the randomised trials of VAS showing a beneficial effect - as summarised in a recent meta-analysis by Mayo-Wilson and colleagues(8) are from the 1980s and early 1990s when vaccination coverage was much lower(7) than today. More recent observational studies(2, 5-7) and a few randomised trials(9, 10) have not found a beneficial effect of VAS. This could be due to increasing DTP coverage.

  Mayo-Wilson and colleagues disagree(1). In their meta-analysis they include 16 trials with 194,413 children. They conclude that vitamin A is associated with a 24% reduction in overall mortality. In a sensitivity analysis they also include a recent yet unpublished Indian trial (DEVTA) of more than 1 mill children which showed no effect of vitamin A on mortality(9). Including this trial they still find a 12% significant reduction in overall mortality. They conclude that no further trials are needed(1, 4, 8).

  A 12% reduction in mortality is obviously important but the problem is that by lumping inhomogeneous data, the authors discard the possibility that the effect of VAS may have ceased to be beneficial and may be negative in some situations(3).

  Could the effect of VAS have ceased to be beneficial? Mayo-Wilson and colleagues(1) assert as untrue our claim that the trials with a positive effect are from the 1980s and early 1990s and that more recent data is not supporting current policy: "In our analysis of mortality, we included 8 studies published after 1993. Two studies (reported in 1994 and 1995) report no significant difference in mortality whilst six studies (reported from 1994 to 2002) report larger than average reductions in mortality. Contrary to what Fisker et al. imply, we also included the DEVTA trial in an analysis of mortality, and the effect remained statistically and clinically significant".

  The 8 studies(11-20) published after 1993 (i.e. the time of the first meta-analyses(21,22)) were all conducted before 1995: Dibley 1996 in 1989- 90(11), Barreto 1994 and Agarwal 1995 in 1990-91(12, 13), Pant 1996 in 1990-92(16), Venkatarao 1996 and Donnen 1998 in 1991-93(14, 15) and Benn 1997 in 1993-95(17). The last study (Chowdhury 2002)(18) had sub-studies submitted by 1994(19) and 1995(20). Thus all trials were conducted in the early 1990s. In contrast, DEVTA was conducted between 1999 and 2004(9). We did not imply the DEVTA trial was not included in the analysis of mortality but importantly that the trial was not included in the analysis of a possible secular trend(3).

  If DEVTA is included there is a significant secular trend; the 12% significant reduction in overall mortality consists of 24% significant reduction 15-25 years ago and an insignificant 4% reduction in the most recent and largest trial ever(9). Hence, the data is compatible with the notion that VAS was beneficial in the 1980s and early 1990s, but has ceased to be beneficial. We have supported this notion with reference to a VAS-placebo randomised trial conducted in Guinea-Bissau between 2007 and 2011 in which VAS had no effect on survival(10). The possibility of no effect should be of real concern. Why waste millions and millions of dollars on distributing 1.500.000.000 annual doses of VAS globally(23) if it no longer has any effect on overall child survival?

  Could the effect of vitamin A be negative in some situations? In 2003 we hypothesised that vitamin A amplifies the non-specific effect of vaccines and that VAS therefore would be beneficial when given with measles vaccine or BCG but potentially negative when given with DTP(24). It is this hypothesis which we have pursued and found supported in the present paper(2), in the Ghana VAST reanalysis(7), in observational studies of campaign VAS co-delivered with vaccines(5) and in trials of neonatal VAS(6, 25). We have repeatedly found that the combination of VAS and DTP has negative effects. Such consistent findings are unlikely to be "chance findings"(1). The findings could explain why the effect of VAS is no longer beneficial as DTP coverage has increased significantly since the first VAS trials were conducted. Importantly, they are not refuted by the Mayo-Wilson meta-analysis since no analysis by vaccination status has been carried out(8).

  In a world that increasingly demands evidence-based preventive policies it is embarrassing that our evidence for the VAS policy is more than 15 years old and contradicted by all recent studies. While it is true that VAS is inexpensive, we can no longer be sure that it is safe and beneficial. In such a situation randomised trials are needed.

  References

  1. Mayo-Wilson E, Imdad A, Herzer K, Bhutta ZA. There is no need for further placebo-controlled trials. BMJ open 2012;2:e000448. 2. Fisker AB, Aaby P, Bale C, Balde I, Biering-Sorensen S, Agergaard J, et al. Does the effect of vitamin A supplements depend on vaccination status? An observational study from Guinea-Bissau. BMJ open 2012;2(1):e000448. 3. Benn CS, Fisker AB, Aaby P. With heterogeneous and potentially harmful effects in existing trials it would be unethical not to conduct further studies of the effect of vitamin A supplementation. BMJ 2011;343:d5094. 4. Mayo-Wilson E, Imdad A, Herzer K, Yakoob MY, Bhutta ZA. Response to Benn et al. BMJ 2011;343:d5094. 5. Benn CS, Martins C, Rodrigues A, Ravn H, Fisker AB, Christoffersen D, et al. The effect of vitamin A supplementation administered with missing vaccines during national immunization days in Guinea-Bissau. International journal of epidemiology 2009;38(1):304-11. 6. Benn CS, Rodrigues A, Yazdanbakhsh M, Fisker AB, Ravn H, Whittle H, et al. The effect of high-dose vitamin A supplementation administered with BCG vaccine at birth may be modified by subsequent DTP vaccination. Vaccine 2009;27(21):2891-98. 7. Benn CS, Aaby P, Nielsen J, Binka FN, Ross DA. Does vitamin A supplementation interact with routine vaccinations? An analysis of the Ghana Vitamin A Supplementation Trial. Am J Clin Nutr 2009;90(3):629-39. 8. Mayo-Wilson E, Imdad A, Herzer K, Yakoob MY, Bhutta ZA. Vitamin A supplements for preventing mortality, illness, and blindness in children aged under 5: systematic review and meta-analysis. BMJ 2011;343:d5094. 9. Awasthi S, Peto R, Read S, Bundy D. Six-monthly Augmented Vitamin A Supplementation from 1 to 6 Years of Age: Block Randomized Trial in One Million Children in Northern India. Available at: www.ctsu.ox.ac.uk/projects/devta/istanbul-vit-alecture. Presented at Micronutrient Forum; 2007; Istanbul. 10. Fisker AB. Evaluation of the WHO vitamin A supplementation policy in Guinea-Bissau. [PhD Thesis]. University of Aarhus, 2011. 11. Dibley MJ, Sadjimin T, Kjolhede CL, Moulton LH. Vitamin A supplementation fails to reduce incidence of acute respiratory illness and diarrhea in preschool-age Indonesian children. J Nutr 1996;126(2):434-42. 12. Barreto ML, Santos LM, Assis AM, Araujo MP, Farenzena GG, Santos PA, et al. Effect of vitamin A supplementation on diarrhoea and acute lower- respiratory-tract infections in young children in Brazil. Lancet 1994;344(8917):228-31. 13. Agarwal DK, Pandey CM, Agarwal KN. Vitamin A administration and preschol child mortality. Nutr Res 1995;15(5):669-80. 14. Venkatarao T, Ramakrishnan R, Nair NG, Radhakrishnan S, Sundaramoorthy L, Koya PK, et al. Effect of vitamin A supplementation to mother and infant on morbidity in infancy. Indian Pediatr 1996;33(4):279-86. 15. Donnen P, Brasseur D, Dramaix M, Vertongen F, Zihindula M, Muhamiriza M, et al. Vitamin A supplementation but not deworming improves growth of malnourished preschool children in eastern Zaire. J Nutr 1998;128(8):1320- 27. 16. Pant CR, Pokharel GP, Curtale F, Pokhrel RP, Grosse RN, Lepkowski J, et al. Impact of nutrition education and mega-dose vitamin A supplementation on the health of children in Nepal. Bull World Health Organ 1996;74(5):533-45. 17. Benn CS, Aaby P, Bale C, Olsen J, Michaelsen KF, George E, et al. Randomised trial of effect of vitamin A supplementation on antibody response to measles vaccine in Guinea-Bissau, West Africa. Lancet 1997;350(9071):101-05. 18. Chowdhury S, Kumar R, Ganguly NK, Kumar L, Walia BN. Effect of vitamin A supplementation on childhood morbidity and mortality. Indian journal of medical sciences 2002;56(6):259-64. 19. Chowdhury S, Kumar R, Ganguly NK, Kumar L, Nain CK, Walia BN. Conjunctival impression cytology with transfer (CICT) to detect pre- clinical vitamin A deficiency among slum children in India. The British journal of nutrition 1996;75(5):785-90. 20. Chowdhury S, Kumar R, Ganguly NK, Kumar L, Verma M, Walia BN. Dynamics of conjunctival impression cytologic changes after vitamin A supplementation. The British journal of nutrition 1997;77(6):863-9. 21. Beaton GH, Martorell R, McCabe G, L'abb? KA, Edmonston B, Ross AC. Effectiveness of vitamin A supplementation in the control of young child morbidity and mortality in developing countries: United Nations Administrative Committee on Coordination/Sub-Committee on Nutrition, 1993. 22. Fawzi WW, Chalmers TC, Herrera MG, Mosteller F. Vitamin A supplementation and child mortality. A meta-analysis. JAMA 1993;269(7):898 -903. 23. Wieringa FT, Dijkhuizen MA, Berger J. Vitamin A supplementation in children and hearing loss. BMJ 2012;344:d7603. 24. Benn CS, Bale C, Sommerfelt H, Friis H, Aaby P. Hypothesis: Vitamin A supplementation and childhood mortality: amplification of the non-specific effects of vaccines? Int.J.Epidemiol. 2003;32(5):822-28. 25. Benn CS, Fisker AB, Napirna BM, Roth A, Diness BR, Lausch KR, et al. Vitamin A supplementation and BCG vaccination at birth in low birthweight neonates: two by two factorial randomised controlled trial. BMJ 2010;340:c1101.

  Conflict of Interest:

  None declared

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top
• Published on: 23 November 2015

  There is no need for further placebo-controlled trials

  • Evan Mayo-Wilson, Senior Research Associate (1)
  • Other Contributors:
    • Aamer Imdad (2), Kurt Herzer (3), Zulfiqar A Bhutta (2)

  To the Editor:

  We concur with Fisker et al. that factors moderating and mediating the effect of vitamin A supplementation on mortality are not fully understood.[1] Further observational studies could elucidate this relationship. New trials might examine the impact of delivering vitamin A at different times of year or with different doses and frequencies; however, we disagree with Fisker et al. that further tri...

  Show More

  To the Editor:

  We concur with Fisker et al. that factors moderating and mediating the effect of vitamin A supplementation on mortality are not fully understood.[1] Further observational studies could elucidate this relationship. New trials might examine the impact of delivering vitamin A at different times of year or with different doses and frequencies; however, we disagree with Fisker et al. that further trials in children should use placebo controls.

  In a systematic review,[2] we analysed 44 randomised controlled trials including the large Indian trial they reference.[3] We found that over 1.2 million children have already been randomised to receive vitamin A or placebo. Taken together, these trials demonstrate that vitamin A is important for normal functioning of several systems. In addition to reducing mortality, randomised trials show that vitamin A supplementation reduces diarrhoea, measles, and blindness. Another trial in a population at risk of vitamin A deficiency would add little to what we know, and it would deny some children an intervention that reduces illness, death, and disability. Under these circumstances, a placebo-controlled trial would be unethical.

  Fisker et al. make claims that are too strong based on evidence that demonstrates great uncertainty.[1] They report that vitamin A supplementation was not significantly associated with mortality in an observational study, but they observed only 64 events (deaths) across all groups. Despite showing no main effect, the authors then conducted a series of subgroup analyses that include only 58 events. Split by vitamin A group and most recent vaccination, only one subgroup included more than 10 events. The results have very wide confidence intervals and could be explained by chance. Fisker et al. bolster their argument by citing a previous paper about the possible interaction of sex and vaccination status,[4] an analysis they conducted sixteen years after the results of the original trial were published.[5] In that trial, vitamin A supplementation was associated with a 19% reduction in mortality, which was similar across sexes. Across eight groups (split by vitamin A group, sex, and health card), there were 173 events. The absolute difference in mortality between girls with and without a health card (a proxy for vaccination status) was 0.7%. Again, these differences could be explained by chance, particularly if the reported analyses are a subset of the post- hoc tests they conducted.

  Fisker et al. err in privileging the results of these studies over meta-analyses of much larger randomised controlled trials. We conducted an analysis including all randomised controlled trials reporting separate effects of vitamin A supplementation for boys and girls. Results from trials with 107,474 children suggest that the effect of vitamin A supplementation is not moderated by sex. Whilst some individual study results were not significant, all of the effects for girls favour vitamin A supplementation over placebo.[2]

  In their discussion, Fisker et al. reiterate a claim that Benn made in response to our systematic review, that our meta-analysis was "essentially based on trials from the late 1980s and early 1990s". They also assert "The only recent trials are the mega-trial from India showing no effect and a placebo-controlled trial of VAS administered with vaccines in Guinea-Bissau..." Neither claim is true. In our analysis of mortality, we included 8 studies published after 1993. Two studies (reported in 1994 and 1995) report no significant difference in mortality whilst six studies (reported from 1994 to 2002) report larger than average reductions in mortality. Both of the negative trials referenced by Fisker et al. remain unpublished and unavailable for full scrutiny. Contrary to what Fisker et al. imply, we also included the DEVTA trial in an analysis of mortality, and the effect remained statistically and clinically significant.

  We concur with colleagues who explain that further work is needed to refine supplementation programmes and to increase their coverage,[6] but further placebo controlled trials would cause more harm than good. The WHO reports that 190 million children remain vitamin A deficient, and widespread deficiency persists among millions of women.[7] Evidence from clinical trials shows that supplementation benefits boys and girls alike. We can no longer claim equipoise. Remaining questions about vitamin A supplementation must be answered using existing datasets or through studies that do not deny children an intervention that is safe, beneficial, and inexpensive.

  References 1. Fisker AB, Aaby P, Bale C, et al. Does the effect of vitamin A supplements depend on vaccination status? An observational study from Guinea-Bissau. BMJ Open. 2012;2:e000448.

  2. Mayo-Wilson E, Imdad A, Herzer K, Yakoob MY, Bhutta ZA. Vitamin A supplements for preventing mortality, illness, and blindness in children aged under 5: systematic review and meta-analysis. BMJ. 2011;343:d5094.

  3. Awasthi S, Peto R, Read S, Bundy D. Six-monthly vitamin A from 1 to 6 years of age. DEVTA: cluster randomised trial in 1 million children in North India. www.ctsu.ox.ac.uk/projects/devta/istanbul-vit-alecture. 2008.

  4. Benn CS, Aaby P, Nielsen J, Binka FN, Ross DA. Does vitamin A supplementation interact with routine vaccinations? An analysis of the Ghana Vitamin A Supplementation Trial. Am J Clin Nutr. 2009;90:629-639.

  5. Vitamin A supplementation in northern Ghana: effects on clinic attendances, hospital admissions, and child mortality. Ghana VAST Study Team. Lancet. 1993;342:7-12.

  6. Thorne-Lyman A, Fawzi WW. Improving child survival through vitamin A supplementation. BMJ. 2011;343:d5294.

  7. WHO. Guideline: Vitamin A supplementation in infants and children 6-59 months of age. Geneva: World Health Organization; 2011.

  Conflict of Interest:

  None declared

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top