Association between short sleep duration and risk behavior factors in middle school students

Abstract

Study Objectives: To examine the association between self-reported sleep duration (SD) and peer/individual factors predictive of risky behaviors (risk behavior factors) in a large socioeconomically diverse school-based sample of early adolescents. Design, Setting, and Participants: Survey data collected from 10 718 and 11 240 eighth-grade students in 2010 and 2012, respectively, were analyzed. Intervention: N/A Measurements and Results: Self-reported school night SD was grouped as ≤4 hours, 5 hours, 6 hours, 7 hours, 8 hours, 9 hours, and ≥10 hours. Scores on 10 peer/individual risk behavior factor scales were dichotomized according to national eigth-grade cut points. The percentage of students reporting an "optimal" SD of 9 hours was 14.8% and 15.6% in 2010 and 2012, respectively; 45.6% and 46.1% reported <7 hours. Adjusted for covariates of gender, race, and SES, multilevel logistic regression results showed that odds ratios (ORs) for 9 of 10 risk factor scales increased with SD <7 hours, with a dose-response effect for each hour less sleep compared to an SD of 9 hours. For example, ORs for students sleeping <7 hours ranged from 1.3 (early initiation of antisocial behavior) to 1.8 (early initiation of drug use). The risk factor scale ORs for <5 hours SD ranged from 3.0 (sensation seeking) to 6.4 (gang involvement). Conclusions: Middle school students are at high risk of insufficient sleep; in particular, an SD <7 hours is associated with increased risk behavior factors.