|
|
| |||||||||||||||||||||||||||||||||||||
Background Prematurity and low birth weight are associated with high perinatal and infant mortality, especially in developing countries. Maternal micronutrient deficiencies may contribute to these adverse outcomes.
Methods In a double-blind trial in Dar es Salaam, Tanzania, we randomly assigned 8468 pregnant women (gestational age of fetus, 12 to 27 weeks) who were negative for human immunodeficiency virus infection to receive daily multivitamins (including multiples of the recommended dietary allowance) or placebo. All the women received prenatal supplemental iron and folic acid. The primary outcomes were low birth weight (<2500 g), prematurity, and fetal death.
Results The incidence of low birth weight was 7.8% among the infants in the multivitamin group and 9.4% among those in the placebo group (relative risk, 0.82; 95% confidence interval [CI], 0.70 to 0.95; P=0.01). The mean difference in birth weight between the groups was modest (67 g, P<0.001). The rates of prematurity were 16.9% in the multivitamin group and 16.7% in the placebo group (relative risk, 1.01; 95% CI, 0.91 to 1.11; P=0.87), and the rates of fetal death were 4.3% and 5.0%, respectively (relative risk, 0.87; 95% CI, 0.72 to 1.05; P=0.15). Supplementation reduced both the risk of a birth size that was small for gestational age (<10th percentile; 10.7% in the multivitamin group vs. 13.6% in the placebo group; relative risk, 0.77; 95% CI, 0.68 to 0.87; P<0.001) and the risk of maternal anemia (hemoglobin level, <11 g per deciliter; relative risk, 0.88; 95% CI, 0.80 to 0.97; P=0.01), although the difference in the mean hemoglobin levels between the groups was small (0.2 g per deciliter, P<0.001).
Conclusions Multivitamin supplementation reduced the incidence of low birth weight and small-for-gestational-age births but had no significant effects on prematurity or fetal death. Multivitamins should be considered for all pregnant women in developing countries. (ClinicalTrials.gov number, NCT00197548
[ClinicalTrials.gov]
.)
The micronutrient status of the mother is an important determinant of fetal growth and survival.5 Iron and folate supplements are routinely provided to pregnant women in many developing countries.6,7 The prenatal dietary intake of other micronutrients is frequently considered to be insufficient to meet increased requirements during pregnancy, particularly in developing countries.8 Prenatal multivitamin supplements are routinely provided in many developed countries; however, this practice is not based on findings from randomized trials. Two placebo-controlled trials involving pregnant women in Sarlahi, Nepal,9 and in Mexico,10 most of whom presumably were negative for human immunodeficiency virus (HIV) infection, showed no significant effects of micronutrient supplements on birth weight or other pregnancy outcomes. In a trial conducted in Janakpur, Nepal, supplements reduced the incidence of low birth weight but had no effect on preterm birth.11 Pooled analyses of data from the two trials in Nepal suggest that micronutrient supplementation may increase the risk of perinatal death.12
In a randomized, controlled trial involving HIV-positive pregnant women in Tanzania, multivitamin supplements significantly reduced the risks of fetal death, low birth weight, and preterm birth and increased maternal CD4 cell counts and hemoglobin levels.13 Because these benefits might not apply to HIV-negative women, we conducted a similarly designed trial involving HIV-negative women in Tanzania to assess the effects of these supplements on birth outcomes and maternal health indicators.
Methods
Patients
Pregnant women who attended antenatal clinics in Dar es Salaam, Tanzania, between August 2001 and July 2004 were invited to participate in the trial. Requirements for eligibility included a negative test for HIV infection, a plan to stay in the city until delivery and for 1 year thereafter, and an estimated gestational age between 12 and 27 weeks according to the date of the last menstrual period. HIV type 1 serologic status was ascertained from all women who consented to participate in the trial by means of two sequential enzyme-linked immunosorbent assay tests.14 All women provided written informed consent to participate. The study was approved by the institutional review boards at Muhimbili University College of Health Sciences in Dar es Salaam and at the Harvard School of Public Health in Boston.
Study Design
The women were randomly assigned to receive a daily oral dose of either a multivitamin supplement or placebo from the time of enrollment until 6 weeks after delivery. The supplements included 20 mg of vitamin B1, 20 mg of vitamin B2, 25 mg of vitamin B6, 100 mg of niacin, 50 µg of vitamin B12, 500 mg of vitamin C, 30 mg of vitamin E, and 0.8 mg of folic acid. On average, these amounts were twice the recommended dietary allowance (RDA) for vitamin E and 6 to 10 times the RDA for vitamin C and several B vitamins. Vitamin A and zinc were not included because earlier studies had shown harmful effects of vitamin A during pregnancy and at delivery in HIV-infected women in Tanzania15 and Zimbabwe,16 and several trials of zinc administered prenatally had shown no evidence of a benefit.17 The active tablets and placebo were similar in shape, size, and color and were packaged in identical coded bottles. A list was prepared according to a randomization sequence in blocks of 20; at enrollment, each eligible woman was assigned to the next numbered bottle. At every monthly visit, a new bottle was given to each woman, and the pills remaining in the used bottles were counted. Research assistants who assessed the study outcomes were unaware of the intervention groups. Tishcon commercially prepared the tablets but had no involvement in the study design, implementation, or reporting of the findings.
Our research team provided the participants with standard prenatal care. All women, irrespective of the assigned study regimen, were given daily doses of iron (60 mg of elemental iron) and folic acid (0.25 mg). They were also given malaria prophylaxis in the form of sulfadoxine–pyrimethamine tablets (Fansidar, Roche) at 20 weeks and 30 weeks of gestation. All women completed a baseline questionnaire that included their sociodemographic characteristics and obstetrical history. At randomization and at monthly visits thereafter, questionnaires were administered to evaluate interim medical problems. Laboratory investigations at baseline included tests for syphilis, gonorrhea, and trichomoniasis; routine urine and stool tests; and evaluation of blood films for malaria. The total and differential white-cell counts were calculated with the use of a CBC5 counter (Coulter). Counts of T-cell subgroups (CD4+, CD8+, and CD3+) in a random sample of women were calculated with the use of the FACScount or FACScan system (Becton Dickinson). Blood counts were obtained at baseline and 6 weeks post partum. The laboratory results were available to the women's physicians, who prescribed treatment, if indicated.
Full-time research midwives attended to the women at delivery. The weights of the babies and the placentas were measured to the nearest 10 g. The infants' body length and head circumference were measured to the nearest 0.1 cm. Women who did not come to the study clinic for their monthly appointments were visited at home when possible and were asked to come to the clinic if their condition allowed.
Study Outcomes
The primary outcomes were low birth weight (<2500 g), preterm delivery (before 37 weeks of gestation), and fetal death. Secondary outcomes included a birth weight below 2000 g, extremely preterm delivery (before 34 weeks), a size at birth that was small for gestational age (defined as a birth weight below the 10th percentile for gestational age, according to the standards of Brenner et al.18), and fetal death and death in the first 6 weeks of life. As in other studies of small-for-gestational-age infants,13 we used U.S. standards, which are considered to provide a better reflection of growth potential unimpeded by nutritional deprivation than are local norms and thus to provide a reasonable "reference" population. Other secondary outcomes included the following continuous variables: birth weight, length, head circumference, gestational age, and placental weight; risk of cesarean section; maternal mortality, including deaths up to 6 weeks post partum; hematologic status, assessed on the basis of both continuous hemoglobin levels and two categorical definitions of anemia (hemoglobin level <11 g per deciliter and <8.5 g per deciliter); and immune status, assessed on the basis of both continuous CD4+, CD8+, and CD3+ cell counts and categorical counts reflecting the baseline median values as cutoff points (CD4+ count, <775 per cubic millimeter; CD8+ count, <480 per cubic millimeter; and CD3+ count, <1350 per cubic millimeter).
Statistical Analysis
We calculated that we would need to enroll 6000 women in order to examine the effects of the vitamin supplements on the primary outcomes. This calculation was based on the following assumed risks in the placebo group (obtained from pilot data in the same setting): fetal death, 8.8% of pregnancies; low birth weight, 17.6%; and preterm births, 22.6%. Assuming a 10% loss to follow-up, this sample size provided a statistical power of more than 90% to detect protective effects of vitamin supplementation that would be equivalent to at least a 30% reduction in the rates of low birth weight and prematurity and a power of more than 80% to detect at least a 25% reduction in the risk of fetal death. The sample size was later increased to 8468 to increase the power because the observed rates of these outcomes were lower than anticipated. (This increase also permitted the enrollment of 2400 women who would undergo a second round of random assignments to the same two regimens 6 weeks post partum so that we could examine the effects of postnatal multivitamin supplementation on infant health — a separate analysis that is not reported here.)
Treatment effects were assessed by means of an intention-to-treat analysis. For outcomes for children, generalized estimating equations19 with a compound symmetry working correlation matrix were used to account for correlations due to twinning in the analysis of infant outcomes (155 pairs of twins and 1 set of triplets). Binary end points were assessed with the use of the log link and binomial variance function, and continuous end points were assessed with the use of the identity link and the gaussian variance function.20 To assess the statistical significance of treatment effects, P values were obtained from the robust score test. To assess the statistical significance of the treatment effects for end points for women, the chi-square test, or Fisher's exact test when warranted, was used for binary variables, and the Wilcoxon rank-sum test was used for continuous variables.21 A data safety and monitoring board met six times during the study and reviewed the results with regard to efficacy and the safety end points. The protocol called for the use of the Peto stopping boundary with a nominal P value of 0.001 for early discontinuation.22
Results
Of the 8468 women who were enrolled, 40 were not eligible for the study. Among the remaining 8428 women, data on birth outcomes were available for 8379 (99.4%); 6 women died before delivery, and the other 43 were lost to follow-up by the time of delivery (Figure 1). Of 8379 women with known birth outcomes (miscarriage, stillbirth, or live birth [with a baby born with any evidence of life, such as breathing or heartbeat, considered to be liveborn]), 8137 gave birth to live babies and were eligible for the analyses of birth weight and prematurity outcomes. In this group of women, a birth weight was not recorded for 271 babies (3.3%) because the birth occurred at home or at another medical facility.
|
|
|
|
Among all the women, 8.4% in the multivitamin group and 7.3% in the placebo group underwent cesarean section (relative risk, 1.15; 95% CI, 0.99 to 1.33; P=0.06). Multivitamins had no significant effect on the risk of maternal death (P=0.27). Maternal anemia was less likely in the multivitamin group than in the placebo group, and mean hemoglobin levels and CD4+ cell counts were significantly higher among women assigned to receive multivitamins (P<0.001 for both comparisons), but the absolute differences between the groups were small (Table 4).
|
HIV-negative Tanzanian women who received prenatal supplementation with vitamin B complex and vitamins C and E did not have significantly reduced risks of prematurity and fetal death, but they did have significantly reduced risks of low birth weight. Multivitamins also reduced the risk of a birth size that was small for gestational age and resulted in significant albeit modest improvements in hemoglobin levels and CD4+ cell counts among mothers 6 weeks after delivery.
We previously reported that the same multivitamin regimen improved outcomes among HIV-infected women enrolled in a similarly designed study in Tanzania.9 In the present study, which involved HIV-negative women, the magnitude of the benefit was smaller. The risks of low birth weight and a birth size that was small for gestational age were reduced by 18% and 23%, respectively, among HIV-negative women, as compared with corresponding reductions of 44% and 43% among HIV-infected women.
Multivitamins, including B vitamins and antioxidant vitamins C and E, may lead to better birth outcomes in several ways. By enhancing maternal nutritional status and immunity during pregnancy, multivitamins may reduce the risk of intrauterine infections. However, we were not able to examine this hypothesis directly. In observational studies, positive associations were reported between maternal vitamin status and birth outcomes.23,24,25,26 Better birth outcomes may also occur by means of improvements in maternal hematologic status. Low hemoglobin levels are associated with increased risks of adverse pregnancy outcomes, including low birth weight.27 In our previous trial, supplementation with vitamin B complex and vitamins C and E significantly increased the mean hemoglobin level in HIV-positive women at 6 weeks post partum (mean increase, 0.6 g per deciliter),13 as compared with an increase of 0.2 g per deciliter in the HIV-negative women in the current study. Although the small increase in the HIV-negative women may not be important on an individual level, a small shift of the population distribution toward higher hemoglobin levels may be beneficial from the public health perspective. Vitamin C and riboflavin improve the intestinal absorption of iron. Riboflavin is also necessary for synthesis of the globin component of hemoglobin.28
Our findings differ from those of two placebo-controlled trials among presumably HIV-negative pregnant women from Sarlahi, Nepal,9 and Mexico,10 which showed no significant effects of multiple micronutrient supplements on birth weight and other pregnancy outcomes. In both trials, the supplement included doses of micronutrients containing the RDA. In a placebo-controlled trial in Janakpur, Nepal, multiple micronutrients containing the RDA resulted in a significant mean difference in birth weight of 77 g and a reduction of 25% in the risk of low birth weight, as compared with placebo; this trial did not show a significant effect of vitamin supplementation on the risk of preterm birth.11 In a trial in Guinea-Bissau, which compared supplements containing one or two times the RDA of micronutrients with placebo,29 the incidence of low birth weight was significantly higher among infants of mothers who received placebo than among infants of mothers who received supplements containing twice the RDA (13.6% vs. 10.0%; mean difference in birth weight, 95 g) but was not significantly different from the incidence among infants of mothers who received supplements containing the RDA (low birth weight, 12.0%; mean difference in birth weight between this group and the placebo group, 53 g; P=0.009 for linear trend in birth weight).
The efficacy of supplements that include the RDA may vary in developing countries. In our previous study involving HIV-positive Tanzanian women13 and in the current study involving HIV-negative Tanzanian women, the supplement contained twice the RDA of vitamin E and 6 to 10 times the RDAs for several B vitamins and vitamin C. HIV-infected persons are likely to require an increased vitamin intake to maintain an adequate nutritional status.30 Given that the RDA is the level recommended for healthy women in North America, even in the absence of HIV infection, this level of supplementation may be inadequate to meet the requirements of pregnant women in many developing countries because of the higher burden of undernutrition and parasitic infections in these countries.
In an analysis of pooled data from the two Nepalese trials, perinatal mortality was paradoxically increased with multivitamin use.12 The authors suggested that higher birth weight, at least among short and chronically undernourished South Asian women, could lead to difficulty in labor, presumably because of cephalopelvic disproportion, with associated perinatal mortality. The trial in Mexico did not show an adverse effect of supplementation.10 Fetal mortality was significantly reduced (by 39%) with supplementation in our previous study of HIV-positive Tanzanian women.13 In the current study, supplementation had no significant effect on fetal mortality; the same was true in the study in Guinea-Bissau.29 In both the current study and the study in Zimbabwe,31 the risks of cesarean section did not differ significantly between the multivitamin and placebo groups. Thus, the availability of standard antenatal and obstetrical care that provides for early identification and management of conditions associated with obstructed labor may mitigate concerns about the potentially adverse effects of micronutrient supplements; such care was provided in the African trials that were carried out in large urban centers.
In our current study, the rate of low birth weight in the placebo group was substantially lower than previously reported rates in the general population. This low rate may reflect the high standard of prenatal care provided according to national and World Health Organization guidelines. Even so, multivitamin supplementation had beneficial effects on some pregnancy outcomes. It is not clear whether the effects of supplementation would be different in populations of women for whom prenatal care is less available. Future trials could assess the effects of varying doses and the addition of other nutrients not included in our regimen.
In conclusion, prenatal multivitamin supplementation significantly reduced the risks of low birth weight and a birth size that was small for gestational age, but it had no significant effects on the risks of prematurity or fetal death. In light of these benefits and the low cost of the supplements, multivitamins should be considered for all pregnant women. Many developing countries have a system for providing prenatal iron and folate supplements; the supplements are produced in bulk by the United Nations Children's Fund (UNICEF) at an estimated cost of less than $1 per person for the duration of pregnancy.32 The increase in cost of incorporating the RDA of additional nutrients is conservatively estimated to be about 20%,32 and scaling up prenatal multiple micronutrient supplementation could be a highly cost-effective approach to improving birth outcomes among pregnant women in developing countries.
Supported by a grant from the National Institute of Child Health and Human Development (NICHD R01 37701).
No potential conflict of interest relevant to this article was reported.
We thank the mothers and children, the field teams, including nurses, midwives, supervisors, and laboratory staff, and the administrative staff who made the study possible; Said Aboud, Illuminata Ballonzi, Julia Finkelstein, David Hunter, Sylvia Kaaya, Karim Manji, Heavengton Mshiu, Christina Nyhus, Ruilan Wei, and all other members of the Harvard–Tanzania collaboration; and the members of the data safety and monitoring board: Paul Jacques, the late Valerian Kimati, Andrew Kitua, Zul Premji, and Meir Stampfer. We also thank the permanent secretary of the Tanzanian Ministry of Health, the authorities at Muhimbili University College of Health Sciences, and the City of Dar es Salaam Regional Health Authority for their institutional support.
Source Information
From the Departments of Nutrition (W.W.F., P.P., W.C.W.), Epidemiology (W.W.F., E.H., W.C.W., D.S.), and Biostatistics (D.S.), Harvard School of Public Health, Boston; and the Departments of Community Health (G.I.M.) and Microbiology and Immunology (W.U.), Muhimbili University College of Health Sciences, Dar es Salaam, Tanzania.
Address reprint requests to Dr. Fawzi at the Department of Nutrition, Harvard School of Public Health, 665 Huntington Ave., Boston, MA 02115, or at mina{at}hsph.harvard.edu.
References
| |||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
