Meta-analysis of short sleep duration and obesity in children and adults.
Level 3 - non-randomized controlled study
Systematic review and meta-analysis of observational studies.
PubMed 18517032 · doi:10.1093/sleep/31.5.619
What was done
A systematic review and meta-analysis of publications indexed across five databases (MEDLINE, EMBASE, AMED, CINAHL, PsycINFO up to late 2007) and manual searches without language restrictions. Studies reporting sleep duration as an exposure and BMI or obesity prevalence as an outcome were included. Results were pooled using random-effects models, with assessments for heterogeneity, sensitivity, and publication bias.
What was found
From 45 eligible studies, 30 studies (12 pediatric, 18 adult; comprising 36 population samples and 634,511 participants: 30,002 children and 604,509 adults, ages 2 to 102) were meta-analyzed. In children, the pooled odds ratio (OR) for short sleep duration and obesity was 1.89 (95% CI 1.46 to 2.43; P < 0.0001). In adults, the pooled OR was 1.55 (95% CI 1.43 to 1.68; P < 0.0001), with a pooled regression coefficient of -0.35 unit change in BMI per hour of sleep change (95% CI -0.57 to -0.12). No evidence of publication bias was found.
Why it matters
This review quantitatively synthesizes global evidence showing a consistent, dose-related association between short sleep duration and higher risk of obesity across all age groups.
Limits
The vast majority of included evidence consists of cross-sectional studies, which prevents causal inference. Prospective studies showed inconsistent evidence of temporal sequence, and studies frequently lacked adequate control for critical confounding variables.
Cited by
- supports Sleeping less than six hours per night increases the risk for obesity by 55%.