• Birth weight (<3 days)
  
• Perinatal mortality
  

• Maternal haemoglobin (8 weeks pp)
  
• Maternal anthropometry (8 weeks pp)
  
• Infant growth (8 weeks pp)
  

Prenatal maternal micronutrient supplementation has been suggested as a means to reduce the proportion of low birth weight babies in low-income countries. The effects of prenatal multi-micronutrient supplements on birth weight and perinatal mortality were studied in a randomised controlled trial among 2100 pregnant women in Guinea-Bissau. Women up to 37 weeks pregnant were individually randomised to daily supplements until delivery of identically looking tablets containing 1) Iron (60 mg) + folate (400 µg), 2) One recommended dietary allowance (RDA) of 5 minerals and 10 vitamins, including iron (30 mg) and folate (400 µg), or 3) Two RDA’s of 5 minerals and 10 vitamins, including iron (30 mg) and folate (800 µg). Supplements were provided in known excess at fortnightly home visits until delivery. Compliance was assessed by tablet count.

Women were interviewed about age, civil status, obstetric history and socio-economic status at enrolment. Maternal anthropometry was measured at enrolment, at delivery, and eight weeks after delivery. Maternal malaria parasitaemia and haemoglobin were measured at enrolment and eight weeks after delivery. Maternal status of a wide range of micronutrients was assessed at enrolment and eight weeks after delivery in a sub-cohort of 600 women. Further, survival of the infants will be followed until 3 years of age, within the routines of the Bandim Health Project surveillance system.

Women were provided impregnated bed nets and weekly malaria prophylaxis until delivery. Women with high malaria parasitaemia at enrolment were in addition given a full treatment at enrolment.

Sample size considerations: A sample size of 638 women in each treatment group will be required to detect a reduction in perinatal mortality from 9% to 5% or less in a treatment group using 80% power and a 5% significance level. With an expected loss to follow up of 10%, 2100 women are required. With a sample size of 2100 the study will be able to detect at least 75 g difference, i.e. a 2.5% change in birth weight.