Methods	Study Design: Randomized controlled trial
	Unit of Randomization: Individually randomised
	Type of study: Balanced Energy Protein
Participants	Location/Setting: Farming communities, West Kiang region, Gambia
	Population: 12,000
	Sample size: 2047
	Drop outs/withdrawal: Not specified
	Socio‐demographics
	Mean (SD) age/age range: Intervention: 24 (6.2) Control: 23.7 (6.4)/15–45 years
	Occupation: Farming
	Race: Not specified
	Education: Not specified
	Family income: Not specified
	Inclusion criteria: Women of reproductive age
	Exclusion criteria: Not specified
Interventions	Intervention (sample size):
	Frequency of intervention was the daily intake of two biscuits
	Composition: The biscuits were made of roasted groundnuts, rice flour, sugar, groundnut oil and they provided a maximum intake of 4250 kJ energy, 22 g protein, 56 g fat, 47 mg Calcium and 1.8 mg Iron. The biscuits were made at a single village. Duration of intervention was about 20 weeks (82 days). Supplementation was delivered by two birth attendants in each village (n = 1010)
	Control (sample size):
	Pregant females who received supplementation with high energy biscuits after delivery. Supplentation was delivered by two birth attendants in each village (n = 1037)
	Concomitant interventions:
	Both groups received:
	1. Routine antenatal care (interview; abdominal palpation; blood pressure and haemoglobin and urine protein concentrations; treatment or referral as indicated) from midwives in a mobile clinic that visited each village twice monthly 2. Iron and folate supplements (according to haemoglobin concentration) 3. Tetanus toxoid to not previously protected women 4. prophylactic dose of chloroquine (in hungry season)
	Training:
	The field workers were trained in anthropometric techniques and Parkin scoring in the delivery wards of the Gambia's main hospital
	Follow‐up:
	Maternal height (baseline), maternal weight (fortnightly), parity, birth weight, birth length, head circumference, and gestational age were assessed by eight field workers within 48 h of delivery
	Stillbirths were recorded.
	Infant deaths were recorded by a governmentappointed village recorder and cross checked by field workers when each infant became due for follow‐up anthropometry
Outcomes	Primary outcome:
	Stillbirth
	Perinatal mortality
	Neonatal mortality
	Infant mortality
	Secondary outcome:
	Low birthweight
	Birth weight
	Birth length
	Head circumference
	Timing of outcome assessment: Within 48 h of delivery
Notes	Study start date: July 1989
	Study end date: October 1994
	Time period: 5 years
	Study country: Gambia, West Africa
	Study limitations:
	Had inadequate sample power to test the prediction of a 37% reduction in neonatal mortality (equivalent to an odds ratio of 0.60)
	Funding source:
	Medical Research Council, Overseas Development Administration, and Nestlé Foundation
	Conflict of interest: None