Analyzing 506,582 person-days of accelerometer recordings from 73,647 UK Biobank participants followed for over 549,000 person-years, sleep regularity index (SRI) scores below 60 were strongly associated with elevated all-cause, cardiovascular, and cancer mortality — and this association was steeper and more dose-dependent in males than females. Males were disproportionately represented in the very low SRI (<60) group even after adjusting for demographic, health, and behavioral covariates. The interaction contrast comparing SRI <60 versus ≥90 differed meaningfully by sex, suggesting males face compounded mortality risk when sleep-wake timing is highly erratic.
This preprint, not yet peer-reviewed, extends a growing body of evidence positioning sleep regularity — rather than duration alone — as a primary circadian health metric. Prior work has shown SRI outperforms nightly sleep hours in predicting mortality, and this analysis adds a critical sex-stratification layer. The biological mechanisms remain speculative: males may have less robust circadian entrainment or fewer compensatory hormonal buffers. Sociocultural exposures like shift work, which skew male, could also confound apparent biological vulnerability. Crucially, this is observational epidemiology — causality cannot be established, and unmeasured confounders remain a concern. The omnibus SRI × sex interaction did not reach conventional significance, so the sex-specific finding, while suggestive, requires independent replication. Practically, men with chronically variable sleep schedules may warrant targeted clinical screening. Results should be treated as hypothesis-generating until peer review is complete.