longevitywatch

How much sleep do I really need?

For most adults, 7 hours of sleep is the safe minimum, with 7 to 8 hours representing the best-supported optimum. Both too little and too much sleep are associated with greater health risks.

Most adults do best with 7 to 8 hours of sleep per night. This is shown by an overview of 11 systematic reviews involving more than 4.4 million participants from 30 countries. The strength of the evidence varies by outcome, from low to high1.

Consistently sleeping fewer than 7 hours is associated with a greater risk of obesity, diabetes, high blood pressure, cardiovascular disease, stroke, mental health problems and premature death. More than a third of American adults did not reach this minimum in 20142. These are observational findings, so cause and effect have not been established.

Sleeping too much also carries risks. In a cohort study of more than 322,000 participants followed for an average of 14 years, the risk of death at 10 or more hours per night was 34% higher in men and 48% higher in women compared with 7 hours. The same U-shaped pattern, in which both too little and too much sleep are associated with higher mortality, also applied to cardiovascular disease and cancer3.

Similar thresholds apply to brain health in older adults. Among older Chinese adults, sleeping fewer than 6 hours was associated with a 22% higher likelihood of mild cognitive decline. Sleeping more than 8 hours was associated with a 78% higher likelihood, compared with 6 to 8 hours. The most favourable point was around 7 hours. A short afternoon nap of less than 30 minutes was associated with a 27% lower likelihood of decline compared with not napping at all. This is a cross-sectional snapshot, so cause and effect are uncertain here as well4.

Do you exercise in the evening? Then pay attention to timing. Vigorous exercise within 4 hours of bedtime is associated with a later time falling asleep, shorter sleep and a higher nighttime heart rate. If you stop more than 4 hours before bedtime, or opt for light activity, this disadvantage does not apply5. A note of caution: this study was partly conducted by researchers affiliated with the company that supplied the measurement equipment.

The evidence
5 studies · 1 meta-analyses · ≈ 4,727,410 participants

All claims are based on observational research (cohort and cross-sectional studies and systematic reviews thereof). No randomised controlled trials are available for sleep duration as a long-term lifestyle intervention. Causal conclusions are therefore not possible. The Asian cohort study is large and long-running, which makes the associations more reliable, but cultural and geographical bias is possible.

Last updated: June 2026
Related answers
How much does a consistent sleep schedule actually matter?
How much melatonin should you take to sleep better?
Moderate evidence
What can I do to sleep better?
Strong evidence
Will my memory improve if I sleep more?
Yes · Moderate evidence
What is the difference between REM sleep and deep sleep, and do I need both?
Yes · Moderate evidence
What is sleep debt and can you really pay it back?
Yes · Moderate evidence
Related research
28 Jul
Better brain drainage cuts tau buildup in mice
27 Jul
Sleep loss in Alzheimer’s is caused by immune cells, not plaques
12 Jun
Faking sleep in mice actually works
Don't see your question?
Ask it and we'll look it up for you.
Ask a question
Newsletter

Stay in the loop

Twice a week, the most important longevity research in your inbox.