A multisectoral approach to eradication of malnutrition in vulnerable groups: a cluster-randomised trial

Abstract

Abstract Background Nutrition is essential for sustainable development. Yet in Uganda, malnutrition and poverty are still prevalent, affecting mostly women and children younger than 5 years. This pilot study tests Uganda's National Development Plan, which uses a multisectoral approach to eradicate malnutrition. We examined the effect of financial products and agricultural and health extension services on food security and nutrition of smallholder farmers. Methods The study, conducted in southwestern Uganda, used a cluster-randomised controlled design at village level. 180 villages were randomly assigned to six treatment groups (30 villages per treatment group) and 30 villages were assigned to a control group. Treatment groups were: agricultural training for the first year (treatment [T]1); agriculture (ie, T1) and health interventions (T6 and T2); agricultural credit and T2 (T3); a voucher scheme for subsidised inputs and T2 (T4); price insurance and T2 (T5); and health interventions—growth monitoring and promotion in children aged >2 years, quarterly health and nutrition community forums, and conditional food transfers to pregnant women (T6). Primary outcomes were food security, measured using the Food and Agricultural Organization's household food insecurity access scale (HFIAS), and smallholder family nutrition (as measured by prevalence of wasting, underweight, and stunting in children younger than 5 years). The HFIAS ranges from 0 to 27, with a lower scale implying higher food security. Analysis was by intention to treat. The trial was not registered. Findings We enrolled 7694 households from 210 villages; 196 households (2·55%) were lost to follow up, leaving data from 7498 households in analysis. Data were collected from April to June, 2014, and June to August, 2016. At the end of the intervention period, there was an increase in food security in group T3, compared with control. At baseline, HFIAS for the control changed from 10·5 to 6·74 while that for group T3 improved from 12·4 to 4·57 (p=0·027). Wasting was significantly reduced, compared with the control group, in only two treatment groups: T3 with a 14·22% reduction (p Interpretation The combination of agricultural and health interventions together with the voucher scheme for subsidised agricultural inputs was generally more effective in increasing food security and reducing wasting and underweight than other treatment interventions. The pilot did not achieve its objective with respect to stunting, especially in the agriculture-only treatment arm (T1). While we believe that contamination in the control group may have reduced the effect of the project, our findings do highlight the importance of health interventions in addition to simply having enough food. Funding Japan Social Development Fund, World Bank Group Africa Gender Innovation Lab.