e-Library of Evidence for Nutrition Actions (eLENA)

Micronutrient supplementation in pregnant women with HIV infection

Systematic review summary

This document has been produced by the World Health Organization. It is a summary of findings and some data from the systematic review may therefore not be included. Please refer to the original publication for a complete review of findings.

Original publication
Siegfried N, Irlam JH, Visser ME, Rollins NN. Micronutrient supplementation in pregnant women with HIV infection. Cochrane Database of Systematic Reviews 2012, Issue 3. Art. No.: CD009755. DOI: 10.1002/14651858.CD009755.

Key findings

  • All four included trials were conducted in urban Tanzania in women who were not receiving antiretroviral therapy, and thus findings may not be generalizable to all populations
  • Selenium supplementation during and after pregnancy reduced the risk of low birth weight and maternal diarrhoea, but no beneficial effects were found for zinc supplementation in pregnancy
  • Supplementation with multivitamins in pregnancy reduced the risk of low birth weight, very premature birth, maternal depressive symptoms, child diarrhoea and several other secondary outcomes
  • Higher dose multivitamins conferred no extra benefit over multivitamins containing the recommended daily allowance (RDA)
  • Pregnant HIV-infected women should be given multivitamin supplements in single RDA formulations during the antenatal period and for at least for six weeks post-partum, especially if breastfeeding

1. Objectives

To evaluate the safety and efficacy of micronutrient supplementation for the reduction of morbidity and mortality in pregnant and lactating women with HIV infection and their infants

2. How studies were identified

The following databases were searched up to 19 July 2011:

  • CENTRAL (The Cochrane Library 2011, Issue 3)
  • PubMed

The AIDS Education Global Information System was searched for relevant conference abstracts, reference lists were hand-searched and researchers and policymakers were also contacted.

3. Criteria for including studies in the review

3.1 Study type

Systematic reviews and randomized controlled trials

3.2 Study participants

Pregnant and lactating women with HIV infection or AIDS, with or without malnutrition. Participants may have been receiving antiretroviral therapy (ART)

(Pregnant and lactating mothers <14 years old were excluded)

(AIDS was defined using either 2008 CDC classification Stage 3 or WHO stages of classification [Stages 3 and 4])

3.3 Interventions

Micronutrient supplementation (vitamins, trace elements, or combinations of these) compared with other supplements, placebo or no treatment for a minimum duration of four weeks

3.4 Primary outcomes

Pregnancy and maternal outcomes:

  • Low birth weight
  • Premature delivery
  • Quality of life and physical functioning of women

Child-related outcomes:

  • Mother-to-child transmission of HIV
  • Child morbidity
3.5 Secondary outcomes

Secondary outcomes included: maternal morbidity including diarrhoea; maintenance of breast-feeding; maternal haematological indicators and immune cell counts; maternal progression to HIV/AIDS; maternal and infant survival

4. Main results

4.1 Included studies

Four randomized controlled trials, enrolling 3529 women, were included in this review.

  • Pregnant women were enrolled between 12 and 27 weeks’ gestation. The majority of women had CD4 cell counts >400 cells/microL and were not receiving ART
  • One study was a four-arm factorial trial comparing vitamin A (30mg b-carotene plus 5000 IU preformed vitamin A, vitamin A), vitamin A and multivitamins, multivitamins, and placebo
  • Two studies compared micronutrient supplementation (zinc 25 mg, selenium 200 microg) to a placebo
  • One study compared supplements providing multiple strength recommended daily allowance (RDA) multivitamins (>100% RDA) to single strength RDA multivitamins (100% RDA)
  • Additional standard care antenatal co-interventions were daily supplementation with iron and folic acid and malaria prophylaxis. In all studies except a factorial trial, multivitamins were given as a co-intervention to treatment and control groups, and nevirapine was given during labour to reduce the risk of mother-to-child HIV transmission
  • Multivitamin supplements in all trials contained thiamine, riboflavin, niacin, vitamin B6, vitamin B12, vitamin E and folic acid
4.2 Study settings

All four studies were conducted by the same research group in an urban setting in Tanzania. Pregnant women were recruited in hospital-based antenatal clinics.

4.3 Effect of intervention on primary outcomes

4.3.1 How the data were analysed

Four main comparisons were made:

  • Multivitamin supplementation versus no multivitamins
  • Zinc supplementation versus placebo
  • Selenium supplementation versus a placebo
  • Multiple RDA multivitamins versus single RDA multivitamins

Outcome measures for dichotomous data were calculated as relative risks (RR) in intervention compared to control groups with 95% confidence intervals (CI). For continuous data, the mean difference (MD) between intervention and control groups and CI were calculated. For the four-arm trial, groups were combined for analysis on the basis of multivitamin supplementation. Data were intended to be modelled with random-effects meta-analysis, but due to a lack of clinical homogeneity, data were not synthesized for meta-analysis across the separate trials.

4.3.2 Summary of effects

Pregnancy and maternal outcomes
Low birth weight
In the four-arm trial, infants born to mothers receiving multivitamins were statistically significantly less likely to have a low birth weight <2500 g compared with infants whose mothers did not receive multivitamins (RR 0.55, 95% CI [0.38 to 0.81], p=0.003; 1085 infants). There was a borderline significant difference for very low birth weight <2000 g in favour of the multivitamin groups (RR 0.42, 95% CI [0.17 to 1.00], p=0.05; 1085 infants).

Zinc supplementation did not reduce the risk of low birth weight <2500 g (RR 1.06, 95% CI [0.58 to 1.93], p=0.86; 358 infants). However, supplementation with selenium resulted in a statistically significant reduced risk of low birth weight <2500 g (RR 0.68, 95% CI [0.47 to 0.98], p=0.037; 842 infants), but not very low birth weight <2000 g (RR 0.59, 95% CI [0.32 to 1.10], p=0.098; 842 infants). In comparison to single RDA multivitamins, multiple RDA multivitamins did not reduce the risk of low birth weight <2500 g (RR 1.13, 95% CI [0.79 to 1.61], p=0.51; 1000 infants) or very low birth weight <2000 g (RR 1.09, 95% CI [0.61 to 1.94], p=0.78; 1000 infants).

Premature delivery
While infants of mothers in the multivitamin groups of the factorial trial were not less likely to be born preterm <37 weeks (RR 0.86, 95% CI [0.68 to 1.10], p=0.23; 1085 infants), they were statistically significantly less likely to be born severely preterm <34 weeks (RR 0.61, 95% CI [0.38 to 0.95], p=0.03; 1085 infants).

Preterm birth <37 weeks did not differ between treatment and control groups in the zinc supplementation trial (RR 1.10, 95% CI [0.70 to 1.76], p=0.67; 366 infants), or the selenium trial (RR 1.01, 95% CI [0.77 to 1.37], p=0.96; 842 infants). In the multiple RDA versus single RDA multivitamin trial, no differences were evident between groups for preterm birth <37 weeks (RR 1.04, 95% CI [0.81 to 1.35], p=0.73; 1030 infants). The risk of severely preterm birth <34 weeks was not reduced with selenium (RR 1.07, 95% CI [0.60 to 1.90], p=0.82; 842 infants) or multiple RDA multivitamin supplementation (RR 1.12, 95% CI [0.70 to 1.78], p=0.63; 1030 infants).

Improved quality of life or physical functioning of women
In the four-arm trial, women underwent psychosocial assessments at two months after enrolment, then every two months until delivery, every six months until 2011, and every 12 months thereafter. The authors of a separate paper using this data reported that women receiving multivitamins had a statistically significant reduced incidence of elevated depressive symptoms compared with those not receiving multivitamins (RR 0.78, 95% CI [0.66 to 0.92], p=0.005; 1013 women) (Smith et al. HIV Med 2007; 8:203–12). Other quality of life indicators, including physical functioning, physical role, bodily pain, general health, vitality, social functioning, and mental health were also significantly improved in the groups receiving multivitamins.

Child-related outcomes
Mother-to-child transmission of HIV
The incidence of HIV infection in infants at birth was not reduced with multivitamin supplementation (RR 1.53, 95% CI [0.94 to 2.50], p=0.09; 739 infants), or at 6 weeks postpartum (RR 1.18, 95% CI [0.81 to 1.71], p=0.38; 453 infants) or at 24 months postpartum (RR 1.06, 95% CI [0.87 to 1.30], p=0.57; 898 infants).

Maternal zinc supplementation also showed no statistically significant effect on the risk of mother-to-child transmission of HIV at birth (RR 1.37, 95% CI [0.49 to 3.85], p=0.55; 286 infants), or at 6 weeks postpartum (RR 1.40, 95% CI [0.67 to 2.95], p=0.37; 286 infants).

Mother-to-child transmission of HIV infection was not reported in the selenium or multiple RDA versus single RDA multivitamin trials.

Child morbidity
Child morbidity surveillance was performed at monthly clinic visits between six and 24 months of age in the factorial multivitamin and vitamin A trial (reported in Fawzi et al., Clin Infect Dis 2003; 36:1053–62). Children whose mothers received multivitamin supplementation had statistically significantly fewer episodes of diarrhoea up to 24 months of age (RR 0.83, 95% CI [0.71 to 0.98], p=0.03; 788 children). Among children who were not HIV-infected at six weeks of age, CD4 cell counts at 24 months were improved in the multivitamin groups compared to those whose mothers did not receive multivitamins (MD 151 cells/microL, 95% CI [64 to 237], p=0.0006).

Adverse effects
In a sub-group of women in the factorial trial, samples of breast milk were obtained every three months for the first two years postpartum. Compared with women receiving the placebo, women receiving only multivitamins had an increased risk of sub-clinical mastitis (RR 1.33, 95% CI [1.09 to 1.61], p=0.005; 674 women). In a subsample of 255 women, zinc supplementation was statistically significantly associated with a three-fold increased risk of reaching a mid-upper arm circumference (MUAC) of <22 cm among women whose baseline MUAC was ≥ 22 cm.

Secondary outcomes
Overall, there were statistically significantly fewer foetal deaths in the multivitamin groups when miscarriages and stillbirths were considered as a composite variable (RR 0.61, 95% CI [0.39 to 0.94], p=0.03; 1021 infants). Infants of women who received multivitamins were less likely to be small-for-gestational age (RR 0.57, 95% CI [0.40 to 0.83], p=0.003; 768 infants). The risks of small-for-gestational age, miscarriage, stillbirth, foetal loss, perinatal death and neonatal death were not reduced with zinc, selenium, or multiple RDA multivitamin supplementation.

The risk of low weight gain in the third trimester was reduced by 30% among women receiving multivitamins in the factorial trial (95% CI [0.55 to 0.90], p=0.005; 837 women]. Women receiving multivitamins were also reported to have statistically significantly improved haemoglobin concentrations (MD 0.7 g/dL, 95% CI [0.44 to 0.96], p<0.00001; 1021 women) and CD4 cell counts (MD 55.00 cells/microL, 95% CI [12.51 to 97.49], p=0.01; 722 women) and from baseline to six weeks postpartum. Maternal mean change in CD4 cell count remained statistically significantly greater in the multivitamin groups at 30 weeks postpartum (MD 40.00 cells/microL, 95% CI [7.25 to 72.85], p=0.02; 507 women). In a separate paper, compared to placebo women receiving multivitamins alone were reported to have a significantly reduced viral load in the first four years of follow-up (-0.18 log, p≤0.002), and be significantly less likely to progress to WHO stage 4 or die of AIDS-related causes (RR 0.71, 95% CI [0.51 to 0.98], p=0.04) (Fawzi et al., NEMJ 2004;351:23–32). However, these findings were not statistically significant for the multivitamin plus vitamin A arm. No statistically significant differences were observed for maternal haemoglobin concentrations, CD4 cell counts or viral load between women randomised to treatment or control groups in the zinc and selenium trials, or in the multiple versus single RDA multivitamin trial. Selenium supplementation reduced the risk of maternal diarrhoeal-related morbidity (acute or persistent diarrhoea) by 40% (95% CI [0.42 to 0.84], p=0.003), acute diarrhoea by 41% (95% CI [0.42 to 0.83], p=0.003) and watery diarrhoea by 44% (95% CI [0.39 to 0.81], p=0.002). No effect of selenium on other measures of morbidity or mortality was detected.

5. Additional author observations*

All four studies were conducted in the same location in urban Tanzania, and participants attended antenatal clinics served by a large tertiary hospital. Therefore, the findings may not apply to women living in rural areas where nutritional deficiencies and other diseases may be more prevalent. In addition to being conducted in the same location by the same researchers, different micronutrient interventions were investigated in each trial. Thus, despite being relatively large and well-conducted trials, the lack of replication limits the applicability of the findings to other populations and settings. The factorial trial was conducted prior to the widespread use of ART, thus preventing the replication of these results in future trials. However, more convincing evidence would be gained from replicating the other trials across more diverse participant populations.

While there is no conclusive evidence that zinc or selenium supplementation alone is advantageous, the results from one trial suggest that the provision of single RDA multivitamins is as beneficial as providing multiple RDA multivitamins to pregnant women with HIV infection.

Further research is needed to investigate the effect of other micronutrients including iodine and vitamin D in pregnant HIV-infected women, as well as the effect of ART therapy during pregnancy on micronutrient concentrations and lipid profiles.

*The authors of the systematic review alone are responsible for the views expressed in this section.