Examining Family-Level Mechanisms and Processes that Affect Children’s Sleep and Weight Health within the Neighborhood

Sleep and weight health during childhood can be the essential building blocks that influence later physical and psychological health. Researchers note how neighborhood effects on health may be mediated or moderated by parenting or family-related factors. This dissertation expanded on the efforts through an examination of contextual predictors of children’s health using the Arizona Twin Project dataset (N = 933; 51.7% female; 56.5% White; 25.6% Latino). The family stress model (FSM), contextual relevance model (CRM), and the integrated model were used to test the relations between neighborhood and family factors on children sleep and weight health. Parent perceptions of neighborhood danger and childhood opportunities index were tested using multilevel- mediation, moderation, and moderated mediation models to address some of the limitations in previous reviews. The FSM was partially supported: parental stress and strain mediated the association between perceived neighborhood danger with sleep efficiency. However, the FSM was not fully supported as parental stress and disrupted parenting did not serially mediate the associations between neighborhood conditions and sleep efficiency. Moderation analyses revealed that negative parenting exacerbates associations between perceived neighborhood danger and sleep duration, demonstrating amplified disadvantages processes of the CRM. Negative parenting also moderated the associations between perceived neighborhood danger and sleep efficiency; the relationship was beyond what was noted in the CRM. Greater neighborhood opportunity was more strongly related to greater body fat percentage in families with higher positive parenting, as compared to families with lower positive parenting. Familism significantly moderated the association between neighborhood contexts and parental stress and strain within a subsample (Latino descent only). Last, the exploratory multilevel moderated mediation models suggested that most associations did not vary by the subgroups explored (i.e., positive parenting, familism cultural value). These results join a growing initiative to explore the science of ecological context, culture, and family interplay.