Multiple micronutrient supplementation in Tanzanian infants born to HIV-infected mothers: A randomized, double-blind, placebo-controlled clinical trial

Background: Multiple micronutrients (vitamin B complex and vitamins C and E) were effective at reducing infectious disease morbidity, HIV disease progression, and poor pregnancy outcomes in HIV-infected women. Objective: The objective was to evaluate whether direct supplementation of these micronutrients to HIV-exposed infants reduces mortality and morbidity. Design: Infants born to HIV-infected women from Dar es Salaam, Tanzania, were randomly assigned to receive daily oral supplementation of multiple multivitamins (vitamin B complex and vitamins C and E) or placebo from age 6 wk for 24 mo. All-cause mortality, hospitalizations, and unscheduled clinic visits were recorded. Morbidities were recorded during monthly follow-up visits. All mothers received multiple micronutrients throughout the study. Results: A total of 1193 infants were randomly assigned to receive micronutrients and 1194 to receive placebo. There were 138 child deaths in the multivitamin group and 124 deaths in the placebo group (HR: 1.13; 95% CI: 0.88, 1.44; P = 0.33). Hospitalizations (RR: 0.83; 95% CI: 0.62, 1.13; P = 0.23), unscheduled clinic visits (RR: 0.97; 95% CI: 0.85, 1.10; P = 0.59), and maternal reports of diarrhea (RR: 0.97; 0.87, 1.10; P = 0.64) were not significantly different between the 2 groups. Fever (P = 0.02) and vomiting (P = 0.007) were significantly lower in the multivitamin group. Among 429 children whose mothers received antiretroviral (ARV) therapy, multivitamin use had no effect on mortality but was associated with a significant reduction in hospitalizations (P = 0.035), episodes of fever (P = 0.005), and episodes of fever and cough (P = 0.019). Conclusions: In the setting of maternal micronutrient supplementation, supplementation of HIV-exposed infants with vitamin B and vitamins C and E does not reduce mortality. Studies of nutrition supplementation in ARV-exposed infants may be warranted. This trial was registered at clinicaltrials.gov as NCT00197730. © 2012 American Society for Nutrition. The study was a randomized, double-blind, placebo-controlled trial. Women aged ≥18 y presenting for prenatal care at the 32nd week of gestation or earlier in 1 of 8 antenatal clinics in Dar es Salaam were offered HIV screening with pre- and posttest counseling. Women who tested HIV positive were further screened for eligibility, including the intention to reside in Dar es Salaam for the duration of follow-up. Maternal HIV-1 serostatus was determined by 2 sequential enzyme-linked immunosorbent assays with the use of Murex HIV antigen/antibody (Abbott Murex) followed by the Enzygnost anti-HIV-1/2 Plus (Dade Behring) (15); discordant results were resolved by a Western blot test (Bio-Rad Laboratories). Written informed consent was obtained from women for participation in the trial while still pregnant. Eligibility for infant participation in the trial included singleton birth and age between 5 and 7 wk at randomization. Excluded were infants born of multiple gestation or those with serious congenital anomalies or other conditions that would interfere with study procedures, including the ability to consume a daily micronutrient supplement. Hemoglobin concentrations were measured by using the AcT5 Diff AL hematology analyzer (Beckman Coulter). Weight was measured to the nearest 10 g with a digital infant balance (TANITA), and length was measured to the nearest 1 mm with a rigid length board with a movable foot piece. Anthropometric equipment was calibrated on a regular basis. Institutional approval was granted by the Harvard School of Public Health Human Subjects Committee, the Muhimbili University of Health and Allied Sciences Committee of Research and Publications, the Tanzanian National Institute of Medical Research, and the Tanzanian Food and Drugs Authority. A Data Safety and Monitoring Board (DSMB)5 met twice annually during the course of the study. Infants were randomly assigned to receive a daily oral dose of multivitamins or placebo from enrollment at age 6 wk for 24 mo. A randomization list from 1 to 2400 was prepared by the study biostatistician in Boston with the use of permuted blocks of size 20 and provided to the pharmacy Department in Dar es Salaam, with each number corresponding to a code denoting 1 of the 2 treatment arms. On-site study pharmacists stored the coded randomization list in a locked file cabinet and concealed allocation by covering the numeric regimen code on each blister pack with a sticker. Infants enrolled at the study clinic were provided the next consecutive number in series. Study physicians, research nurses, and participants were unaware of the treatment groups. From age 6 wk to 6 mo, infants in the multivitamin (active) arm received one capsule containing 60 mg vitamin C, 8 mg vitamin E, 0.5 mg thiamine, 0.6 mg riboflavin, 4 mg niacin, 0.6 mg vitamin B-6, 130 μg folate, and 1 mg vitamin B-12. From age 7 mo to the end of follow-up, 2 capsules were given daily. These dosages represent between 150% and 600% of the US Adequate Intake for children aged 0–6 mo and 200% and 400% for children aged 7–12 mo (16–20). The supplement used was a powder encapsulated in an opaque gelatinous capsule manufactured by Nutriset. Mothers were instructed how to push the capsule through the back of the blister pack, open it, and decant the powder into a small plastic cup. Sterile water (5 mL) supplied with the supplement was added to the powder, and the dose was given to the child orally. A pilot phase of open-label vitamin use in 12 infants and mothers confirmed that this supplement preparation and use was well accepted by the mothers and infants. Both the placebo and active capsules contained an orange powder of identical taste and appearance. All study personnel and participants were blinded to treatment assignment for the duration of the study. Only the study statisticians and the DSMB saw unblinded data, but none had any contact with the study participants. Mothers and children were asked to return monthly to a central clinic site on the campus of Muhimbili University of Health and Allied Sciences for research visits and standard clinical care. As part of standard medical care, all children received growth monitoring, immunizations, and routine medical care for illnesses. All children enrolled in the trial received periodic high-dose vitamin A supplementation, as per Tanzanian Ministry of Health guidelines (100,000 IU at 9 mo and 200,000 IU at 15 and 21 mo). Fortified complementary foods were not universally available, and universal iron supplementation was not provided to the children. Children found to be anemic were treated with iron supplementation. Mothers were counseled on the risks and benefits of exclusive breastfeeding. The provision of the study regimen was in keeping with WHO recommendation concerning exclusive breastfeeding, in that drops, syrups, or oral rehydration solutions are allowed. Women were instructed not to give any additional foods or fluids while they were exclusively breastfeeding, in keeping with WHO and Tanzanian Ministry of Health guidelines supporting exclusive breastfeeding. All mothers were provided with oral multivitamins, in keeping with our prior findings (10), from enrollment until study end postpartum. Maternal multivitamin doses were generally several times the Recommended Dietary Allowance (RDA) for B-complex vitamins, vitamin C, and vitamin E, but women who were started on antiretroviral therapy (ARV) were changed to single RDA multivitamin dosages. When the study was proposed, routine medical care for pregnant women with HIV infection included malaria prophylaxis, iron and folate supplementation, diagnosis and treatment of sexually transmitted infections and prophylaxis, and diagnosis and treatment of opportunistic infections. ARV medication was limited to nevirapine prophylaxis for maternal to child transmission (one dose given to the mother at the onset of labor and one dose given to the infant within 72 h of birth) (21). As the study progressed, the availability of ARVs increased substantially through programs such as the President’s Plan for AIDS Relief and other governmental and nongovernmental programs. Beginning in July 2005, women and children in the study were screened for ARV eligibility and treated according to Tanzanian Ministry of Health guidelines. For adults, eligibility was based on WHO stage IV HIV disease, or CD4 cell count <200 cells/μL, or WHO stage III and CD4 cell count <350 cells/μL. For children aged <18 mo, eligibility was based on CD4% <20 or Pediatric WHO Stage III; for children aged ≥18 mo, eligibility was based on Pediatric WHO Stage III or CD4% <15%. The standard first-line regimen was stavudine, lamivudine, and nevirapine for adults and zidovudine, lamivudine, and nevirapine for children; alternative drugs were available for specific circumstances. All children were tested for HIV infection at 6 wk of age by using the Amplicor HIV-1 DNA assay version 1.5 (Roche Molecular Systems Inc). Tests at 18 mo of age were performed by using HIV ELISAs followed by Enzygnost anti-HIV-1/2 Plus (Dade Behring); discordant results were resolved by using a Western blot test. Samples from children who tested positive at 18 mo were then back tested via polymerase chain reaction to estimate time of transmission. Time of transmission was estimated as the date halfway between the last negative and first positive HIV test result. During the monthly follow-up visits, mothers underwent standardized assessment of child morbidity with the prompting of symptom diaries that had been distributed at the previous visit. By using pictorials of illness symptoms (eg, diarrhea, vomiting, and fever), mothers were asked to check off days when their children suffered these symptoms. These diaries were reviewed by trained research nurses during the monthly visits to the clinic to document the occurrence of infectious morbidities, including the need for unscheduled clinic visits or hospitalizations. Diarrhea was defined as ≥3 loose or watery stools within a 24-h period. Rapid respiratory rate was defined as >60 breaths/min in infants younger than 2 mo, >50 breaths/min in 2–11-mo-olds, and >40 breaths/min in those aged ≥12 mo. Every 3 mo, and/or when acute complaints of illness were noted, children underwent a physical examination and medical treatment by study physicians. Standardized diagnostic criteria were used to make physician diagnoses. Compliance with the daily regimen was measured by research nurses who counted the number of unused pills. Mothers who were traveling out of Dar es Salaam temporarily were provided with extra regimen and water to suffice until the next visit to the research clinic. Mothers were reimbursed for their travel expenses to the study clinic, but received no other payment. Children who missed their monthly follow-up appointment were visited at home, and their vital status was confirmed through contact with immediate family members. A verbal autopsy was performed in cases of child death to determine the cause of death. Coding of the cause of death from the verbal autopsies was performed independently by 2 pediatricians (KPM and CD), and differences were adjudicated by a third pediatrician. The primary outcome was all-cause mortality. We initially estimated that the mortality rate in the placebo arm would be 25% and powered the trial to detect a 30% reduction in mortality. During the course of the trial, the mortality rate was substantially lower than expected. On the basis of a 2-sided test for a comparison of proportions with 80% power, we increased our sample size from 1600 to 2360 to detect a 30% reduction in child mortality, assuming a placebo arm mortality rate of 12.5%. Because of both this increased sample size and budgetary limits, follow-up time was truncated to before 24 mo for some of the infants. Data were double entered by using Microsoft Access software and converted to SAS software (version 9.1; SAS Institute) for analysis. Final data sets were uploaded onto a UNIX-based server in Boston, Massachusetts. Intention-to-treat analyses were carried out according to a preestablished analysis plan. To evaluate the effect of supplements on mortality, we created Kaplan-Meier curves and evaluated the effect of multivitamins with the use of the Cox proportional hazard test. For morbidity outcomes, the probability of a reported event at a visit was compared between arms by using generalized estimating equations, adjusted for the child’s age (<6, 6–18, or >18 mo). For physician diagnoses, which were made both during routine and additional sick visits, the rate per year of follow-up was compared between arms by using Poisson regression. Stratified analyses by prespecified (child sex, HIV status, low birth weight, maternal age, and parity) and post hoc (maternal ARV use) factors were also performed. All analyses were performed by using SAS software (version 9.2; SAS Institute).