Abstract
OBJECTIVE: To examine associations between prolonged early household food insecurity (FI) during pregnancy, infancy, and toddlerhood, and child feeding practices, and the mediating role of dysfunctional parent-child interactions. METHODS: We conducted secondary longitudinal analyses of data from the Starting Early Program (StEP) randomized controlled trial, which studied a primary care-based child obesity prevention program for low-income Hispanic families. Our independent variable was FI, using the USDA Food Security Module, during the third trimester of pregnancy and at child ages 10- and 19-months. Frequency of reported FI was defined by the number of periods with FI (0, 1, 2, or 3). Our dependent variables were feeding practices at child age 28-months using the Comprehensive Feeding Practices Questionnaire. Our mediating variable was dysfunctional parent-child interactions using the Parenting Stress Index subscale at age 19-months. We used linear regression to determine associations between frequency of reported FI and feeding practices adjusting for covariates, and mediation analyses to determine if dysfunctional parent-child interactions mediate these associations. RESULTS: Three hundred and forty four mothers completed assessments at child age 28-months. Of the 12 feeding practices examined, higher frequency of reported FI was positively associated with using food as a reward, restriction of food for weight control, and using food for emotional regulation, and was negatively associated with monitoring of less healthy foods. There was a significant indirect effect of frequency of reported FI on these practices through dysfunctional parent-child interactions. CONCLUSION: Higher frequency of reported FI was associated with four feeding practices, through dysfunctional parent-child interactions. Understanding these pathways can inform preventive interventions.