Sleep, cognition, and behavioral problems in school-age children: A century of research meta-analyzed

86 studies, 35,936 healthy children aged 5-12

Healthy school-age children, 5-12 years, pooled internationally

Comprehensive meta-analysis of the correlational literature on sleep, cognition, and behaviour in children. The overwhelming majority of pooled primary studies are cross-sectional, so the pooled correlations carry the full weight of family and genetic confounding. Notable for reporting cognitive subdomains separately, which is the outcome-class discipline this archive requires.

Sleep duration correlates with cognitive performance at r = .08 (CI .06-.10). By subdomain: executive functioning r = .07 (CI .02-.13), multi-domain tasks r = .10 (CI .05-.16), school performance r = .09 (CI .06-.12), and — the load-bearing null — NOT with intelligence. Unlike in adults, sleep duration was not associated with sustained attention or memory in children. Shorter sleep correlates with more behavioural problems, r = .09 (CI .07-.11), holding for both internalizing (r = .09) and externalizing (r = .08) problems.

High and unaddressed. These are pooled cross-sectional correlations in a domain where the predictor (sleep duration) is substantially heritable and the outcomes (cognition, school performance, behaviour) are more so. The absence of an association with intelligence is nonetheless notable — it is what one would expect if sleep affects state-dependent performance rather than the underlying trait.