Evaluation of the AIN-C program in Honduras

AIN-C was initially launched in the mid 1990s and intensely developed and supported by USAID over the period 1998-2005. An impact evaluation covering this period was conducted in 2002. The original quasi-experimental, "intention-to-treat" evaluation design called for preand post-comparisons between intervention and control communities. A multi-stage cluster sampling design was employed; data were collected on 1467 households at baseline and 1343 households in the final evaluation. Due to issues including non-equivalent groups, extensive contamination of control communities, and reduced intensity of implementation due to unanticipated diversion of funding during the evaluation period, the pre-post, interventioncontrol community comparison was replaced with an analysis based on individual, community-level participation in AIN-C. Data were analyzed from approximately 1300 households (600 AIN-C; 700 "No-GMP") and 1200 children (580 AIN-C; 611 "No-GMP"). The results demonstrate that AIN-C had a significant, positive impact on young child feeding and care practices and on nutritional status, especially among poorer households and children who participated more regularly in the program. Exclusive breastfeeding rates among infants younger than six months of age, the proportion of caregivers who fed children 6–23 months appropriately, the percentage of children who were fully immunized by 13 months, and the percentage of children through 23 months who received iron and vitamin A supplementation were all significantly higher in the AIN-C group than the group that did not participate in GMP. Controlling for socio-economic status and age, AIN-C participants had mean weight-for-height Z-scores .122 higher than non-participants. The impact of AIN-C was two to three times greater among poorer households. Intensity of participation also increased impact; each additional month of participation yielded a .042 increase in weight-for-age Zscore. (Author abstract, modified)