Is seven hours of sleep really the best amount?

The claim

Seven hours is the sleep duration with the lowest risk of dying.

What the evidence says

Pooled analyses of prospective studies find the lowest risk of dying at about seven hours of sleep and a higher risk at both ends. Short sleep: RR 1.12 (95% CI 1.06–1.18). Long sleep: RR 1.30 (1.22–1.38). That is 16 studies with 1.38 million people (Cappuccio 2010). Per hour below 7 hours: RR 1.06. Per hour above: RR 1.13 (Yin 2017).

The numbers

• Cappuccio 2010: 16 studies (27 cohort samples), 1,382,999 people, 112,566 deaths, follow-up 4–25 years; short sleep RR 1.12 (1.06–1.18), long sleep RR 1.30 (1.22–1.38).
• Yin 2017: U-shaped dose-response, lowest risk at about 7 hours; RR 1.06 (1.04–1.07) per hour below 7 h, RR 1.13 (1.11–1.15) per hour above.
• Sleep measured by questionnaire; observational.

Verdict: Mostly proven

The direction is consistent across cohorts; the size varies. As a rule for you, it isn’t proven.

Why only this far: Self-reported sleep, observational data, no trial. An average across people is not a target for one person.

Evidence level B: Large cohort studies. How verdicts work.

What we don’t know

Sleep was self-reported, and these are observations. Does sleep length cause the difference, or reflect health?

Sources

Checked 15 Oct 2026. Every link opens the original paper or its PubMed record.

1. Cappuccio et al., Sleep 2010Meta-analysis of prospective cohort studies (16 studies, 27 samples)DOI 10.1093/sleep/33.5.585 · PubMed 20469800
2. Yin et al., J Am Heart Assoc 2017Dose-response meta-analysis of prospective cohort studiesDOI 10.1161/JAHA.117.005947 · PubMed 28889101

Follow for the reel on Instagram