The Effects of a Nutrition‐Sensitive Agricultural Intervention on Social Support, Food Security and Maternal Self‐Efficacy in Complementary Feeding

Abstract

IntroductionNutrition‐sensitive agriculture has been the subject of much recent global interest, however, the empirical evidence of the individual pathways to improved nutrition is scant. Care behaviors are recognized as central to child nutrition, yet caregiver resources to perform feeding behaviors are understudied, generally, and specifically in the context of agricultural interventions. Self‐efficacy is an important caregiver resource that influences maternal behaviors around young child feeding practices. We evaluated the impacts of a multi‐faceted intervention (homegardening, gender sensitization) with and without nutrition education on maternal self‐efficacy in complementary feeding. We further explored the mediating roles of spousal support and household food security.MethodsRealigning Agriculture for Improved Nutrition (RAIN) was a four year homegardening intervention targeting rural women in the Central province of Zambia who were pregnant or had a child under two years of age. The intervention was implemented by Concern Zambia and evaluated using a cluster‐randomized design by the International Food Policy Research Institute. The unit of randomization was the census survey area, within which a number of women's groups consisting of 15–20 member were created based on the number of eligible women. Bi‐monthly group meetings, led by community agricultural educators and community health volunteers (CHV), served as the main platform for intervention delivery and were supplemented by home visits. We administered a mid‐term survey, two years after the intervention began, to a random sample of 759 women with children under two years of age in intervention and control groups. We conducted both intention‐to‐treat and per‐protocol analyses using SEM in STATA 13.0.ResultsThe agricultural intervention significantly improved maternal self‐efficacy (β = 0.82; p <0.001) and this was partially mediated by increasing gains in spousal support (β = 0.43; p <0.001) and food security (β = 0.78; p <0.001). We found similar results in women who additionally received nutrition education, but there were no additional gains in self‐efficacy when compared to women who received the agricultural intervention without nutrition education. We discovered substantial variability in the delivery of nutrition education due to low CHV participation. Only 73% of CHVs attended their last meeting, compared to 98% of agricultural educators and 25% of women could not recall their CHV's name. The average number of home visits performed by CHVs was 1.6 compared to 7.5 by agricultural educators. Our per protocol analysis showed additional gains to self‐efficacy for women with high levels of delivery and participation in the agricultural intervention (β = 0.97; p <0.001) with the greatest gains in women who also received nutrition education (β = 1.10; p <0.001). The difference in selfefficacy between women who did and did not receive nutrition education was not significant.ConclusionThis home‐gardening intervention improved maternal self‐efficacy in complementary feeding through improvements in spousal support and food security. The addition of nutrition education provided no additional gains to self‐efficacy in complementary feeding. Leveraging existing health infrastructure, notably community health volunteers, to add nutrition education to an agricultural project presented challenges to delivery in this setting.Support or Funding InformationNSF‐IGERT and USAID