Growth effects of exclusive breastfeeding promotion by peer counsellors in sub-Saharan Africa: the cluster-randomised PROMISE EBF trial
Ingunn Marie Stadskleiv Engebretsen, Debra Jackson, Lars Thore Fadnes, Victoria Nankabirwa, Abdoulaye Hama Diallo, Tanya Doherty, Carl Lombard, Sonja Swanvelder, Jolly Nankunda, Vundli Ramokolo, David Sanders, Henry Wamani, Nicolas Meda, James K Tumwine, Eva-Charlotte Ekström, Philippe Van de Perre, Chipepo Kankasa, Halvor Sommerfelt, Thorkild Tylleskär
BackgroundIn this multi-country cluster-randomized behavioural intervention trial promoting exclusive breastfeeding (EBF) in Africa, we compared growth of infants up to 6 months of age living in communities where peer counsellors promoted EBF with growth in those infants living in control communities.MethodsA total of 82 clusters in Burkina Faso, Uganda and South Africa were randomised to either the intervention or the control arm. Feeding data and anthropometric measurements were collected at visits scheduled 3, 6, 12 and 24 weeks post-partum. We calculated weight-for-length (WLZ), length-for-age (LAZ) and weight-for-age (WAZ) z-scores. Country specific adjusted Least Squares Means with 95% confidence intervals (CI) based on a longitudinal analysis are reported. Prevalence ratios (PR) for the association between peer counselling for EBF and wasting (WLZ < −2), stunting (LAZ < −2) and underweight (WAZ < −2) were calculated at each data collection point.ResultsThe study included a total of 2,579 children. Adjusting for socio-economic status, the mean WLZ at 24 weeks were in Burkina Faso −0.20 (95% CI −0.39 to −0.01) and in Uganda −0.23 (95% CI −0.43 to −0.03) lower in the intervention than in the control arm. In South Africa the mean WLZ at 24 weeks was 0.23 (95% CI 0.03 to 0.43) greater in the intervention than in the control arm. Differences in LAZ between the study arms were small and not statistically significant. In Uganda, infants in the intervention arm were more likely to be wasted compared to those in the control arm at 24 weeks (PR 2.36; 95% CI 1.11 to 5.00). Differences in wasting in South Africa and Burkina Faso and stunting and underweight in all three countries were small and not significantly different.ConclusionsThere were small differences in mean anthropometric indicators between the intervention and control arms in the study, but in Uganda and Burkina Faso, a tendency to slightly lower ponderal growth (weight-for-length z-scores) was found in the i