Physical exercise linked to improved sleep in patients with psychiatric disorders

Subgroup analyses revealed which exercise patterns showed the strongest associations with better sleep. Interventions with a frequency of three or more sessions per week (g = −0.60), aerobic modalities (g = −0.69), and low-to-moderate intensity (g = −0.70) showed larger effect sizes than other approaches. Single sessions lasting 30 minutes or longer (g = −0.58) and total intervention durations under 12 weeks (g = −0.58) were also associated with greater sleep improvements. However, univariate meta-regression analyses found no significant linear moderating effects of intervention duration, frequency, volume, or total weekly exercise time on sleep quality outcomes (all P > 0.05), suggesting that the relationship between exercise dose and sleep benefit may not follow a simple dose-response pattern. The authors conclude that exercise represents a potent complementary treatment for enhancing sleep in individuals with mental health conditions, though substantial heterogeneity among studies and variability in outcome measurement tools represent important limitations to consider when interpreting these findings.

The meta-analysis showed that aerobic exercise at low-to-moderate intensity produced the largest sleep improvements (g = −0.69), larger than high-intensity or resistance-only approaches. Notably, shorter intervention periods (under 12 weeks) showed effects comparable to longer programs (g = −0.58), which contrasts with some fitness domains where longer duration drives stronger effects. The consistent benefits across three-or-more-times-weekly schedules suggests a threshold rather than a linear dose effect—increasing sessions beyond three per week did not proportionally amplify sleep gains. The authors noted that outcome measurement heterogeneity (different sleep scales and subjective vs. objective measures across studies) and modest sample sizes in many individual trials limit confidence in the effect estimates. Subgroup results for psychiatric diagnosis (depression, anxiety, bipolar, psychosis) were not separately reported in the abstract, making it unclear whether benefits hold equally across all mental health conditions.