Causal Associations Between Sedentary Time and Sleep Traits

Few observational studies have examined relationships between sedentary time and sleep traits, and their causal relationships remain unexplored in the general population.

We triangulated evidence using observational analyses and one-sample Mendelian randomization (MR) in the UK Biobank. Sedentary time was assessed by self-report and wrist-worn accelerometry (accelerometer substudy). Six sleep traits were examined: evening chronotype, difficulty waking up, insomnia, short sleep (< 6 h), long sleep (> 9 h), and snoring. Observational associations were estimated using logistic regression with stepwise adjustment, and exploratory observational mediation analyses evaluated BMI as a potential mediator. MR used a genetic risk score for sedentary time and included primary linear MR and secondary nonlinear MR to test departures from linearity. Statistical significance for the six prespecified outcomes in primary linear MR was evaluated using a Bonferroni correction (p < 0.05/6 = 8.3 × 10^{- 3}).

In linear MR, higher genetically predicted sedentary time was associated with increased odds of insomnia, short sleep, long sleep, and snoring, and these associations remained significant under the Bonferroni-corrected threshold. Nonlinear MR provided no robust evidence of departure from linearity after accounting for multiple testing. In observational analyses, self-reported sedentary time was associated with multiple adverse sleep traits. In accelerometer-based analyses, device-measured sedentary time was positively associated with short sleep and evening chronotype, inversely associated with insomnia and long sleep, and its association with snoring was attenuated toward null after full adjustment. Exploratory observational mediation models suggested that BMI may account for part of several associations, particularly that with snoring; these estimates do not establish a causal pathway.

Higher sedentary time may be detrimental to sleep health, with the strongest evidence for short sleep across complementary methods. For other sleep traits, methodological heterogeneity warrants cautious interpretation and further validation.