A Randomized Controlled Trial Offering Higher- Compared with Lower-Dairy Second Meals Daily in Preschools in Guinea-Bissau Demonstrates an Attendance-Dependent Increase in Weight Gain for Both Meal Types and an Increase in Mid-Upper Arm Circumference for the Higher-Dairy Meal

Background: Controversy remains over the most effective approaches to prevent childhood malnutrition. Objectives: We tested the feasibility and effectiveness of delivering ready-to-use supplementary foods (RUSFs) as a second daily meal in preschool children aged 3–5 y in Guinea-Bissau, and compared RUSFs with different levels of dairy protein. Methods: This study was a 3 mo cluster-randomized controlled pilot trial of 2 RUSFs differing in dairy protein in 533 boys and girls from 9 preschools. Children receiving RUSFs were compared with wait-listed controls, and all students received a daily school lunch. The RUSFs were delivered 5 d/wk for 3 mo and contained 478 kcal and 11.5 g protein per 92-g daily serving. Deliveries included a ready-to-use supplementary food with 15% of protein from dairy sources (RUSF-15%) or one with 33% of protein from dairy sources (RUSF-33%). Intention-to-treat (ITT) and per-protocol analyses (>50 d of RUSF consumption) were conducted. Changes in the weight-for-age z score (WAZ) and height-for-age z score were primary outcomes. Additional outcomes included changes in mid-upper arm circumference (MUAC), hemoglobin, and retinol binding protein. Results: Baseline anthropometry was not different between groups (WAZ, 20.48 6 1.04) and increased significantly over time (P 50 d had a significantly greater increase in WAZ relative to the increase in controls (+0.40 and +0.32 for RUSF-15% and RUSF-33%, respectively, compared with +0.24 in controls, P <0 .01 andP < 0.05, respectively). RUSF-33%, but not RUSF-15%, also eliminated a decrease in MUAC observed in controls (20.01 cm in RUSF-33% compared with 20.34 cm in controls, P < 0.05). The only difference between RUSF-15% and RUSF-33% was a mean decrease in hemoglobin in children receiving RUSF-15% (20.5 compared with 20.002 g/dL, P =0 .05). Conclusions: Implementation of 2-meal preschool feeding programs is feasible in low-income countries, and there are measurable benefits relative to 1-meal programs in children attending preschool regularly. In addition, MUAC and hemoglobin measurements indicate that meals with 33% compared with 15% of protein from dairy may help prevent wasting and anemia. J Nutr doi: 10.3945/jn.115.218917.