The Causal Relationship Between Physical Activity and Sleep and Eating Disorders: Evidence From a Mendelian Randomization Analysis.

Abstract

Background: Sleep and eating disorders are significant global public health concerns. While physical activity (PA) has been suggested to influence these disorders, the nature of this relationship remains unclear. This study aims to investigate whether a causal link exists between PA and sleep and eating disorders.

Methods: This study uses Genome-Wide Association Study data; PA ("ebi-a-GCST90061432") is the exposure, with sleep disorders ("ukb-d-SLEEP") and eating disorders ("finn-b-F5_EATING") as outcomes variable. The study sample consists of a European population, and genetic instrumental variables are carefully selected. Two-sample Mendelian Randomization (MR) analysis is performed using inverse variance weighting, weighted median, and MR-Egger regression.

Results: MR analysis revealed a significant negative association between PA and sleep disorders (inverse variance weighting estimate β = -0.004, SE = 0.002, P = .008). Both the Weighted Median and MR-Egger methods showed negative trends, but not statistically significant (β = -0.003/-0.005, SE = 0.002/0.005, P = .191/0.346). For PA and eating disorders, the inverse variance weighting estimate was β = -0.375 (P = .153), with no significance in the other methods, suggesting a limited effect of PA on eating disorders. Sensitivity (MR-Egger, leave-one-out) and heterogeneity (Q = 29.87, P = .472; I2 = 0.004) analyses confirmed the robustness of these findings (F = 24.49).

Conclusion: This study suggests PA may reduce the risk of sleep disorders, while its impact on eating disorders appears limited. Future research should explore how different types, frequencies, and intensities of PA affect various health conditions to optimize its use in public health.