Sleep Duration Irregularity and Risk for Incident Cardiovascular Disease in the UK Biobank

Emerging evidence supports a link between circadian disruption as measured by higher night-to-night variation in sleep duration and increased risk of cardiovascular disease (CVD). It remains unclear whether this association varies by CVD types or may be modified by average sleep duration and genetic risk for CVD.

Our prospective analysis included 86 219 UK Biobank participants who were free from CVD when completing 7 days of accelerometer measurement from 2013 to 2015. Sleep irregularity was evaluated by the SD of accelerometer-measured sleep duration over 7 days. Incident major CVD events, defined as fatal or nonfatal myocardial infarction and stroke, were identified through linkage to Hospital Episode Statistics data until May 31, 2022. Multivariable-adjusted Cox proportional hazard models were used to estimate hazard ratios (HRs) and 95% CIs for associations of sleep duration SD with risk for major CVD events overall and for myocardial infarction and stroke separately.

We documented 2310 incident cases of major CVD events (myocardial infarction: 1183, stroke: 1175) over 636 258 person-years of follow-up. After adjusting for sociodemographic factors and family history of CVD, the HR associated with a 1-hour increase in sleep duration SD was 1.19 (95% CI, 1.10-1.27) for CVD (P-trend<0.0001), 1.23 (95% CI, 1.11-1.35) for myocardial infarction (P-trend<0.0001), and 1.17 (95% CI, 1.05-1.29) for stroke (P-trend=0.003). Additional adjustment for lifestyle factors, comorbidities, and sleep-related factors modestly attenuated these associations. Higher sleep irregularity was associated with higher CVD risk irrespective of genetic risk (P-interaction=0.43), but this association was stronger among individuals with longer average sleep duration >8 hours (P-interaction=0.006).

Higher night-to-night variation in accelerometer-measured sleep duration was associated with consistently higher risks for major CVD events. The association did not seem to be modified by genetic risk for CVD and was more pronounced in long sleepers.