RESUMEN
Objective: To evaluate whether maternal and child nutrition activities provided through the Indian Integrated Child Development Services scheme in India were associated with improved nutritional knowledge and practices among beneficiary women. Methods: We used a multistage sampling design to randomly select 4400 pregnant women or mothers of children younger than 2 years for a cross-sectional telephone survey. The respondents were beneficiaries of the scheme from across 11 Indian states. We used multivariate regression models controlling for sociodemographic factors to estimate the association between: scheme activities and nutrition messages heard; and scheme activities and nutrition practices. We also estimated the proportion of the total association with nutrition practices which was mediated by nutrition messages. Results: Among 110 regression models testing unique pairs of seven activities and 18 nutrition messages, 103 showed a statistically significant positive relationship (median risk ratio, RR: 1.14). For activities and nine nutrition practices, 39 out of 54 tested pairs were significantly associated (median RR: 1.16). We observed statistically significant mediation through nutrition messages for 28 out of 42 tested pairs of activities and nutrition practices. Conclusion: Receipt of the scheme's activities was associated with improved nutrition knowledge and practices. Improvements in practices were statistically mediated by improvements in knowledge. These findings suggest that a large-scale nutrition scheme with a strong counselling component could successfully change beneficiary behaviours.