Pregnant, HIV-negative Tanzanian women who are given multivitamins have reduced risks of some adverse perinatal and postnatal outcomes but not others, according to a double-blind, randomized trial.1 Compared with their counterparts given a placebo, women given multivitamins were less likely to deliver infants with a low birth weight (relative risk, 0.8) or a small size for gestational age (0.8) and were themselves less likely to have anemia postpartum (0.9). However, the risks of fetal death, preterm birth and maternal death were unaffected.

Pregnant women visiting prenatal clinics in Dar es Salaam, Tanzania, during 2001–2004 were eligible for the trial if they had a fetal gestational age of 12–27 weeks and tested negative for HIV. Participants received either a daily multivitamin or a daily placebo from enrollment until six weeks after delivery. The multi-vitamins contained niacin, folic acid and vitamins B1, B2, B6, B12, C and E in amounts providing, on average, 6–10 times the recommended dietary allowance for several of the B vitamins and vitamin C, and two times that for vitamin E. In addition, both groups received iron and folic acid supplements, as well as malaria prophylaxis. Social, demographic and obstetric information was collected and laboratory tests were performed at enrollment; thereafter, the women were examined monthly and received standard prenatal care. Midwives attended deliveries and weighed and measured infants. Outcomes were compared between groups according to the intended treatment.

Analyses were based on 8,428 women and their infants. On average, the women were 25 years old and the fetal gestational age was 21 weeks at enrollment. The majority of women had 5–7 years of education (67%) and had previously had at least one live birth (55%). About two-thirds were anemic (having a hemoglobin level of less than 11 g/dl). On average, the women took 88% of the pills they were given up to delivery and 80% during the entire trial, with no difference between the multivitamin and placebo groups.

Women in the multivitamin group had a lower incidence of low-birth-weight births (those with a birth weight less than 2,500 g) than did their counterparts in the placebo group (8% vs. 9%); the difference corresponded to a 20% reduction in risk (relative risk, 0.8). The result was similar when the analysis was restricted to singleton births. In addition, the average birth weight was higher in the multivitamin group, although only modestly so (3,148 g vs. 3,083 g).

Compared with women in the placebo group, those given multivitamins had a lower incidence of small-for-gestational-age births, defined as deliveries of infants with a birth weight below the 10th percentile for gestational age (11% vs. 14%); the decrease in risk was 20% (relative risk, 0.8). Average gestational age was older in the multivitamin group as well, but the difference was again small (39.5 vs. 39.4 weeks).

The two groups had statistically indistinguishable incidences of preterm birth, defined as birth before 37 weeks (17% in each group), a finding that did not change when the analysis was restricted to singleton births. The groups also had nearly identical rates of extremely preterm birth, defined as birth before 34 weeks (5–6%); fetal death, defined as death at any time before delivery (4–5%); perinatal death, defined as death between 28 weeks' gestation and one week after delivery (6–7%); postnatal death, defined as death during the first six weeks after delivery (3% each); and maternal death, defined as death at any time up to six weeks postpartum.

Six weeks after delivery, relative to their counterparts given placebos, women given multivitamins had a lower incidence of anemia (19% vs. 22%), corresponding to a 10% lower likelihood of this outcome (relative risk, 0.9). There was a difference, albeit a small one, in average hemoglobin level as well (12.1 vs. 11.9 g/dl). The women in the multivitamin group were also less likely to have a CD4+ immune cell count below the median baseline value (775 cells/mm3) at that time (31% vs. 38%), translating to a risk reduction of 20% (relative risk, 0.8). Average CD4+ values were modestly higher in this group (924 vs. 888 cells/mm3).

Discussing the findings, the researchers point out that many developing countries already routinely provide iron and folic acid supplements to pregnant women and could add micronutrients at the recommended dietary allowance level for an increase in cost of roughly 20%. They note that the multivitamin-associated risk reductions observed in HIV-negative women were smaller than those previously observed in HIV-positive women in the same setting, and concede that it is unclear if multivitamins will have the same effects in populations that have less access to prenatal care than the trial population. Nonetheless, given the observed benefits and the low cost of this intervention, "multivitamins should be considered for all pregnant women in developing countries," they recommend.—S. London


1. Fawzi WW et al., Vitamins and perinatal outcomes among HIV-negative women in Tanzania, New England Journal of Medicine, 2007, 356(14):1423–1431.