Night-to-night sleep variability and all-cause and cardiovascular mortality: a prospective cohort study with a complementary in vitro circadian-disruption experiment

Abstract

Irregular sleep–wake patterns may disrupt circadian rhythms, but the dose–response relationship between night-to-night sleep variability and mortality and the association of sleep onset variability with death remain unclear. Using 4–7 valid days of wrist accelerometer data, including ≥1 weekend day, from 7,604 adults aged ≥20 years in the National Health and Nutrition Examination Survey 2011–2014, linked to mortality through December 2019, we examined the within-person SDs of sleep midpoint, sleep onset, and total sleep time across nights and the sleep regularity index (SRI) in relation to all-cause and cardiovascular disease (CVD) mortality. Survey-weighted Cox models were fitted across 20 imputed datasets and combined using Rubin's rules. Restricted cubic splines examined dose–response relationships. Complementary categorical, subgroup, and sensitivity analyses were conducted. Human umbilical vein endothelial cell (HUVEC) experiments repeated in three independent runs compared cellular responses under a regular dexamethasone schedule and four prespecified irregular schedule variants. Over a median follow-up of 81 months, 714 all-cause and 229 CVD deaths occurred. Each 1-hour increase in the across-night SD of sleep midpoint, sleep onset, and total sleep time was associated with all-cause mortality HRs of 1.22 (95% CI, 1.12–1.33), 1.19 (1.11–1.29), and 1.17 (1.05–1.31), respectively. The corresponding CVD mortality HRs were 1.21 (1.09–1.35), 1.22 (1.05–1.42), and 1.16 (0.94–1.42), respectively. Each 10-point lower SRI was associated with all-cause mortality (HR, 1.29; 95% CI, 1.21–1.38) and CVD mortality (HR, 1.31; 95% CI, 1.18–1.45). Across three independent HUVEC runs, the mean response across the four prespecified irregular schedule variants showed lower core clock-gene expression, lower Griess assay absorbance, and higher inflammatory and adhesion-marker expression than the regular schedule. Greater night-to-night sleep irregularity was associated with higher hazards of all-cause and CVD mortality, although these associations were modest compared with those of older age, hypertension, and diabetes in the same models. The complementary cellular experiment showed differences between the regular dexamethasone schedule and the prespecified irregular schedule set.