Sleep Duration Associated with the Lowest Risk of Depression/Anxiety in Adolescents

Abstract

Study objectives: To investigate sleep duration associated with the least depression/anxiety in adolescence.

Methods: Grades 7-12 Japanese students (n = 18,250, aged 12-18 y) from public junior high/high schools were studied in a cross-sectional design. Due to missing/implausible data, 15,637 out of the 18,250 students were statistically analyzed. Relationship between sleep duration on school nights and depression/anxiety, measured using self-report questionnaires, including the General Health Questionnaire-12 (GHQ-12), were studied by sex and grade, controlling for bedtime regularity.

Results: When sleep duration was classified by 1-h intervals, rate of adolescents with a GHQ-12 score ≥ 4 was the lowest in males and females who slept 8.5-9.5 h and 7.5-8.5 h, respectively, (designated "references") in both grades 7-9 and 10-12. The rate was significantly higher than the references in both males and females who slept < 7.5 h, regardless of grade (P < 0.05, logistic regression). GHQ-12 tended to be worse in adolescents (2.0%-13.5%) who slept longer than the references. Sleep duration for the minimum GHQ-12 score was estimated to be 8.8 and 8.5 h in males, and 8.0 and 7.5 h in females, in grades 7-9 and 10-12, respectively, using the General Additive Model.

Conclusions: Sleep duration of ≥ 8.5 h on school nights may be associated with the lowest risk of depression/anxiety on average in male adolescents. Although the duration was estimated to be shorter in females (≥ 7.5 h) than males, this should be interpreted carefully. Most adolescents may currently be sleeping less than the optimal duration.

Commentary: A commentary on this article appears in this issue on page 1491.

Keywords: General Health Questionnaire-12; adolescents; age; mental health; sex; sleep duration.

Figure 1

Relationship between sleep duration on school nights and odds ratio for depression/anxiety (General Health Questionnaire-12 score ≥ 4) by sex and school grade (i.e., grades 7–9 and 10–12). Odds of having depression/anxiety compared with the reference group (those who sleep 8.5–9.5 h in males and 7.5–8.5 h in females), adjusting for bedtime regularity and grades. Odds ratios and their 95% confidence intervals are indicated by circles and bars, respectively. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 2

Estimated dose-response relations between sleep duration on school nights and General Health Questionnaire-12 (GHQ-12) score using the General Additive Model by sex and grade (i.e., grades 7–9 and 10–12). Estimated sleep durations on school nights for the minimum GHQ-12 score are also shown. Dotted lines represent the upper and lower 95% confidence intervals. Natural cubic splines were applied to detect the best model shape.