Self-regulation from birth to age five: Associations among sleep, reactivity and persistence, and outcomes at age seven

Abstract

A substantive body of research documents the ability to self-regulate as a primary developmental goal during the first five years, with strong self-regulation skills considered to be a protective factor across the lifespan (Calkins & Williford, 2009). Whilst there appears to be a growing movement towards the conceptualisation of self-regulation as a broad construct encompassing a range of hierarchical domains (Blair, Calkins & Kopp, 2010), there is limited empirical work that constructs and measures self-regulation in this way. Specifically, although very early sleeping and eating difficulties, and excessive crying are considered to be signs of early regulatory problems (Wolke, Schmid, Schreier & Meyer, 2009), little has been done to empirically link these behaviours with later across-domain self-regulation. In this poster, we explore self-regulation in a nationally representative sample of 2880 children participating in the Longitudinal Study of Australian Children (LSAC) – B Cohort. First, we examine the feasibility of self-regulation measurement models that include sleep regulation, temperamental soothability (reactivity) and cognitive persistence. Second, we explore longitudinal profiles of self-regulation and the relationships between profile membership and social, emotional and behavioural outcomes at age 6 to 7.

At Wave 1 (birth to 1 year), Wave 2 (2 to 3 years) and Wave 3 (4 to 5 years), confirmatory factor analysis confirmed that parent-reported indicators of sleeping problems, feeding problems, soothability and persistence (first order factors) could be used as indicators in well-fitting self-regulation (second order factor) measurement models. Comparative fit indices (CFIs) averaged 0.985 and Root Mean Square Error of Approximations (RMSEAs), 0.031. Very few model modifications were required.

Having established the measurement model, these variables were then used in a longitudinal latent profile analysis (LPA) to describe patterns of self-regulation from birth to age five. LPA was selected over other options such as latent growth curve analysis due to the inability to document measurement invariance. Three profiles emerged. The normative profile had a likely membership of 69% (n=1988) of the sample and had consistently high scores on all measures at each wave. Sleep regulation steadily improved for this group with all members achieving ‘perfect’ sleep scores by 4-5 years. The moderately poor profile had a likely membership of 27% (n=777) of the sample and consistently lower scores on all measures, with sleep regulation becoming more problematic over time. The poor regulation profile had a likely membership of 4% (n=112) of the sample and significantly lower scores than the normative group on most measures, and lower than the moderately poor group on measures of sleep regulation across waves and reactivity at Wave 1. This profile showed a steep decline in sleep regulation from birth to 4-5 years of age. Profile membership predicted scores on the Total Problems Scale of the Strengths and Difficulties Questionnaire at age 5-7 in the anticipated direction, such that those in the normative profile had the lowest problem scores, and those in the poor profile had the highest problem scores. These findings will support the consideration of early regulatory indicators for use in screening children.