Evaluating a Latent Measurement Model for Infant Sleep: From Intrinsic and Extrinsic Predictors to Cognitive Outcomes

The development of infant sleep is thought to be jointly guided by the dual processes of sleep consolidation and regulation. However, until now, there have been few empirical studies testing whether there is evidence for these latent processes. The current study uses structural equation modeling to test whether sleep consolidation and regulation can be modeled by two distinct latent variables. Using observed indicators from multiple sleep assessment methods, we found that a two factor model representing the processes of consolidation and regulation fit better than a one factor, undifferentiated model. These two latent factors were predicted by different intrinsic (i.e., infant) and extrinsic (i.e., parenting) factors, as well as interactions between the two classes of predictors. Specifically, we replicated the interaction of infant temperament and maternal emotional availability in predicting both consolidation and regulation. We also found that infant sleep regulation longitudinally predicted infant attention regulation, although this relationship was only true for children whose mothers held a college degree or higher. These findings contribute to the literature by providing a novel measurement model that appropriately accounts for measurement error. Further, our findings suggest that considering the interaction between child characteristics and parental input in promoting high quality sleep is a key avenue for future research. Finally, by partially replicating sleep-cognition linkages previously observed in adolescents and adults, we find support for the notion of hierarchically organized self-regulatory abilities, which motivates areas for future investigation and possible intervention.Doctor of Philosoph