| Title: | Dose-Response Associations Between Sleep Duration and Health Outcomes in Adults: Comparison Between Self-Reported and Device-Based Measures |
| Journal: | Sleep |
| Published: | 15 Jul 2026 |
| Pubmed: | https://pubmed.ncbi.nlm.nih.gov/42454954/ |
| DOI: | https://doi.org/10.1093/sleep/zsag193 |
Publication 20362
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
Abstract
STUDY OBJECTIVES: This study aimed to: (i) assess whether associations between sleep duration and health outcomes differ by measurement method (self-report vs. actigraphy); and (ii) compare dose-response curves to evaluate whether sleep duration recommendations should vary based on the method of assessment.</p>
METHODS: We conducted a prospective cohort study using data from UK Biobank participants aged 40-69 years who wore wrist-worn accelerometers for 7 days and self-reported their daily sleep duration. Outcome measures included all-cause mortality, incident cardiovascular disease (CVD), type 2 diabetes, dementia, and depression. Dose-response associations between sleep duration-measured using device-based metrics (total sleep time and sleep period time) as well as self-report-and these outcomes were examined, adjusting for multiple covariates.</p>
RESULTS: We analyzed data from 69,153 adults for cardiovascular disease (CVD) and up to 76,811 adults for mortality, with an average follow-up of 8.0 (SD: 0.9) years. Across all sleep measurement methods, adjusted dose-response curves showed a similar inverse J-shaped pattern (particularly for device-based measures), and a U-shaped association for incident depression. Associations were stronger for sleep period time than for the other sleep metrics. For example, a sleep period time of 7 hours was associated with a 30% lower risk of mortality relative to the 5-hour reference value (HR: 0.70; 95% CI: 0.61-0.79). In comparison, a total sleep time and a self-reported sleep duration of 7 hours were associated with 17% (HR: 0.83; 95% CI: 0.78-0.89) and 14% (HR: 0.86; 95% CI: 0.74-1.00) lower mortality risk relative to the 5-hour reference value, respectively. The lowest estimated risk across outcomes occurred at approximately 7.2 hours for self-reported sleep, 7.7 hours for sleep period time, and between 6.8 and 9.3 hours for total sleep time.</p>
CONCLUSIONS: Both self-reported and device-measured sleep durations demonstrated broadly similar dose-response relationships with health outcomes, although associations were substantially stronger for device-measured sleep period time. These findings support existing evidence that 7-9 hours of sleep per night is optimal for adults, regardless of how sleep is measured, and suggest that sleep period time may be more relevant than total sleep time for future device-based recommendations.</p>
6 Authors
- Jean-Philippe Chaput
- Raaj Kishore Biswas
- Matthew Ahmadi
- Peter A Cistulli
- Wenxin Bian
- Emmanuel Stamatakis
1 Application
| Application ID | Title |
|---|---|
| 25813 | Integrated paradigms to investigate the interplay between lifestyle behaviours, the social environment, and genetic risk in the development of cardio- metabolic disease and mental health. |
Enabling scientific discoveries that improve human health