Abstract
BACKGROUND: Studies assessing the impact of sleep restriction (SR) on blood pressure (BP) are limited by short study length, extreme SR (<4 hours a night), and lack of attention to psychological distress as a possible mediator. METHODS: A community-based cohort was assembled with 237 women (age 34.1 ± 13.5 years; body mass index 25.4 ± 5.4 kg/m(2)), and a randomized, crossover, intervention study was conducted in 41 women (24 completed: age 30.2 ± 6.5 years; body mass index 24.3 ± 2.8 kg/m(2)) to determine the causal effect of SR on BP. Sleep was maintained as usual (HS) or reduced by 1.5 hours a night (SR) for 6 weeks. In the cohort, associations between sleep and psychosocial factors were evaluated using multivariable models adjusted for demographic and clinical confounders. In the intervention study, in-office BP was measured weekly; ambulatory BP was measured at end point. Psychological factors were assessed at baseline and end point. Mixed-model analyses with total sleep time (TST, main predictor), week and fraction of time spent in physical activity (covariates), and subject (random effect) were performed. RESULTS: Among the community cohort, higher perceived stress, stressful events and distress, and lower resilience were associated with shorter sleep, worse sleep quality, and greater insomnia symptoms (P < .05). In the intervention, systolic BP increased as TST decreased (TST × week interaction, [coefficient ± standard error] -0.0097 ± 0.0046, P = .036). Wake ambulatory diastolic blood pressure (-0.059 ± 0.022, P = .021) and mean arterial pressure (-0.067 ± 0.023, P = .018) were higher after SR versus HS. Psychological distress variables were not affected by TST and did not mediate the effects of SR on BP. CONCLUSIONS: These results suggest that SR influences CVD risk in women via mechanisms independent of psychological stressors.