Modeling pathways from a behavior change communications intervention to complementary feeding outcomes: a proof of concept study

Abstract

BackgroundBehavior change interventions to improve complementary feeding have met with varying success. We propose that research to identify, understand and address determinants of recommended complementary feeding behaviors will lead to more efficacious outcomes. In this study we examine the role of maternal self‐efficacy for complementary feeding, a mother's belief in her capability to organize and execute courses of action for complementary feeding, as a factor that affects responses to 2 specific complementary feeding behaviors in rural Bangladesh.ObjectiveIn the context of the cluster‐randomized evaluation of Alive & Thrive (A&T), a behavior change intervention in Bangladesh, we examined the role of maternal self‐efficacy for 2 complementary feeding behaviors, using a theoretically derived model of determinants of complementary feeding behaviors.MethodsWe developed a set of questions that comprised an 8 point additive scale to measure maternal self‐efficacy for complementary feeding and inserted them into a program survey (n=457 mothers of children 6–24 mo of age). Based on qualitative research conducted as part of the A&T process evaluation, we selected 2 measured behaviors from the behaviors the project promoted—reported feeding of green leafy vegetables (GLV) in the last 24 h and the timely introduction of egg. These 2 behaviors differed in terms of mother's potential control over implementation requirements, ease of putting them into practice, and potential cultural constraints. With data from the survey on multiple maternal and household characteristics we examined direct and mediated paths of maternal self‐efficacy, in both program exposed and control group mothers using ordinary least squared regression, logistic regression, and structural equation modeling.ResultsThe A&T intervention significantly increased both behaviors (11 percentage‐point increase in the predicted probability for egg (p=0.036); 16 percentage‐point increase in the predicted probability for GLV (p<0.001)). For GLV maternal self‐efficacy for complementary feeding mediated (β =0.355, p=0.010) and potentiated (positively modified) (β=0.3888, p=0.039) the effect of the intervention. This means that in the intervention group, compared to the control, the mean value of self‐efficacy was higher and, simultaneously, at higher levels of self‐efficacy there was a greater likelihood of feeding GLV. In contrast, there was neither mediation nor potentiation of the effect of the intervention on timely introduction of egg.ConclusionMaternal self‐efficacy for complementary feeding was a significant mediator and potentiator for feeding GLV, but not for egg. As a proof of concept study, these findings support the value of examining and targeting maternal self‐efficacy in behavior change interventions that target complementary feeding. Our divergent findings for GLV vs the timely introduction of egg highlight the complex determinants of infant feeding behaviors and the importance of identifying features of cultural context that affect maternal capability to adopt intervention‐targeted behaviors.Support or Funding InformationSupport: Alive & Thrive, funded by the Bill & Melinda Gates Foundation and managed by FHI360; NIH (5T32HD007331).