A cluster randomized controlled trial of a community-based initiative to reduce stunting in rural Indonesia

We evaluate the impacts of a $120 million project in Indonesia conducted between 2014 and 2018 that sought to reduce stunting through a combination of (1) community-driven development grants targeted at health and education outcomes, (2) training for health providers on infant and young child feeding and growth monitoring and (3) training for sanitarians on a local variation of community-led total sanitation. This cluster randomized controlled trial involved 95 treatment and 95 control subdistricts across South Sumatra, West Kalimantan, and Central Kalimantan provinces. Overall, we find no significant impacts on stunting, the study's primary outcome measure (0.5 pp; 95% confidence interval [CI]: -3.0 to 4.1 percentage points [pp]), or other longer-term undernutrition outcomes about 1 year after the end of the project. The project had a modest impact on some secondary, more proximal outcomes related to maternal and child nutrition, including the percentage of mothers consuming the recommended number of iron-folic acid pills during pregnancy (8.7 pp; 95% CI: 4.1-13.3 pp), 0-5-month-olds being exclusively breastfed (8.7 pp; 95% CI: 1.8-15.6 pp) and 6-23-month-olds receiving the number of recommended meals per day (8.5 pp; 95% CI: 3.8-13.2 pp). However, there were no significant impacts on other proximal outcomes like the number of pre-natal and post-natal checkups, child dietary diversity, child vitamin A receipt or the incidence of child diarrhoea. Our findings highlight that successfully implementing an integrated package of interventions to reduce child stunting may be challenging in practice. Project design needs to consider implementation reality along with best practice-for example, by piloting the synchronous implementation of multifaceted interventions or phasing them in more gradually over a longer timeframe.