Sleep and Longevity: All-Cause Mortality and the U-Shaped Curve
U-shaped mortality curve: 7–8h sleep is optimal; <6h sleep is associated with 12% higher all-cause mortality and >9h with 30% higher mortality in meta-analyses of 1.3 million participants across 16 studies.
| Measure | Value | Unit | Notes |
|---|---|---|---|
| Optimal sleep duration for longevity | 7–8 | hours/night | Lowest all-cause mortality across multiple large cohort studies |
| All-cause mortality at <6h | +12 | % higher risk | Cappuccio et al. 2010 meta-analysis; 16 studies, 1.3 million participants |
| All-cause mortality at >9h | +30 | % higher risk | Long sleep likely reflects underlying illness rather than being causally harmful |
| CVD mortality at <6h | +48 | % higher risk | Meta-analysis; strongest associations for cardiovascular causes |
| Follow-up duration of studies | 4–25 | years | 22-year Finnish cohort (Hublin) is among longest; consistent U-shape |
The U-Shaped Mortality Curve
The relationship between sleep duration and mortality follows a robust U-shape across all major longitudinal studies: people sleeping 7–8 hours per night have the lowest all-cause mortality, with risk increasing on both ends.
Cappuccio et al. (2010) pooled data from 16 prospective cohort studies representing approximately 1.3 million participants and found:
| Sleep Duration | Relative Mortality Risk |
|---|---|
| <6 hours/night | +12% all-cause mortality |
| 6–7 hours | Slightly elevated |
| 7–8 hours | Reference (lowest risk) |
| 8–9 hours | Slightly elevated |
| >9 hours | +30% all-cause mortality |
The risk associations were consistent across demographic groups, geographic regions, and follow-up periods ranging from 4 to 25 years.
Short Sleep: Causal Risk?
The causal mechanisms linking short sleep to mortality are well-established:
- Cardiovascular disease: elevated blood pressure, impaired nocturnal dipping, sympathetic hyperactivation
- Metabolic disease: insulin resistance, obesity, type 2 diabetes
- Immune impairment: reduced infection defense, potentially increased cancer immune surveillance failure
- Inflammation: elevated CRP, IL-6 as independent cardiovascular risk factors
- Accident risk: impaired psychomotor vigilance, microsleeps, industrial accidents
The Finnish twin cohort (Hublin, 2007) — with 22-year follow-up — found that short sleep predicted mortality even after controlling for other health behaviors, suggesting an independent causal pathway rather than merely a marker of other risk factors.
Long Sleep: Reverse Causation
The elevated mortality associated with long sleep (>9h) is largely interpreted as reverse causation: people who are already ill (depression, cardiovascular disease, cancer, hypothyroidism) sleep more as a symptom of their condition, not as a cause of their mortality.
Experimental extension of sleep in healthy individuals does not increase mortality risk — healthy subjects allowed to sleep 9–10 hours in controlled settings show improved performance and health markers. This contrasts with short sleep restriction, which causally worsens health outcomes in controlled experiments.
Sleep as a Longevity Lever
Walker (2017) estimated that if chronic short sleep (<7h) were eliminated across the US population, all-cause mortality rates could decline meaningfully. While not a magic bullet, optimizing sleep duration and quality is one of the highest-leverage, lowest-cost health interventions with evidence across virtually every disease category examined.
Related Pages
Sources
- Cappuccio FP et al. — Sleep duration and all-cause mortality: a systematic review and meta-analysis. Sleep (2010)
- Youngstedt SD & Kripke DF — Long sleep and mortality: rationale for sleep restriction. Sleep Med Rev (2004)
- Hublin C et al. — Sleep and mortality: a population-based 22-year follow-up study. Sleep (2007)
- Grandner MA et al. — Sleep symptoms associated with intake of specific dietary nutrients. J Sleep Res (2014)
Frequently Asked Questions
How many hours of sleep do you need to live longest?
Meta-analyses consistently identify 7–8 hours as the optimal sleep duration for lowest all-cause mortality. The relationship is U-shaped: both sleeping too little (<6h) and too much (>9h) are associated with higher mortality. The increased risk at long sleep durations is thought to primarily reflect reverse causation — people who are ill sleep more, rather than long sleep itself being harmful.
Is it possible to sleep too much?
Consistently sleeping >9–10 hours per night in an otherwise healthy adult may warrant medical evaluation, as prolonged sleep often reflects underlying health conditions (depression, thyroid dysfunction, sleep disorders). In controlled settings where healthy individuals are allowed to sleep as long as they wish, most converge at 8–9 hours — suggesting extended sleep in the absence of health issues resolves within days of catching up on a deficit.