The risk of dying at your sleep length

Compared with 7 hours of sleep a day, which had the lowest risk of dying, the risk of dying was 15% higher for men and 7% higher for women at 5 hours or less, 6% higher for women at 6 hours (for men the difference could be chance), 6% higher for men and 7% higher for women at 8 hours, 13% higher for men and 17% higher for women at 9 hours, and 34% higher for men and 48% higher for women at 10 hours or more. (Svensson, 2021) This was in 322,721 adults in Japan, China, Singapore and Korea, and the study can't show sleep length caused the deaths. The table below puts each risk as deaths in groups of the same size. Every number links to its source.

The short version

• 7 hours of sleep a day came with the lowest risk of dying, for men and for women (Svensson, 2021).
• 8 hours, the most common amount for men and the second most common for women, came with a 6% higher risk of dying for men and 7% higher for women than 7 hours (Svensson, 2021).
• 10 hours or more came with the highest risk, 34% higher for men and 48% higher for women (Svensson, 2021). In groups of the same size, for every 6 deaths among men who slept 7 hours there were about 8 among men who slept 10 hours or more, and for every 6 among women who slept 7 hours, about 9 among women who slept 10 hours or more.
• This study watched what people already did, so it can't show sleep length caused the deaths. Sleep was asked about once, so later changes weren't counted (Svensson, 2021).

Your sleep length against 7 hours

Find your usual hours of sleep a day. Each row compares people who slept that long with people who slept 7 hours, after allowing for age, marriage, body mass index, smoking, alcohol, exercise, diabetes, high blood pressure and, for women, menopause (Svensson, 2021). The likely range is in brackets. The last two columns put the same risk as deaths in groups of the same size: for every 20 deaths among people who slept 7 hours, about how many among people who slept your hours.

The percentages compare risks between groups. They aren't percentage points, and they aren't what any one person would gain or lose by changing their sleep. Some of the 9 studies asked about night sleep only, and others counted naps in the total (Svensson, 2021).

For men, age mattered. Men under 65 had a higher risk of dying at every sleep length than at 7 hours, while men 65 and over had a clearly higher risk only at 9 hours and at 10 hours or more. For women, age made no clear difference (Svensson, 2021).

The study of 322,721 people

Svensson and colleagues pooled 9 follow-up studies from Japan, China, Singapore and Korea. People were 55 on average when this analysis began, and each said once how many hours they usually slept: when they joined in eight of the studies, and at a later survey in the Shanghai Women's Health Study. The analysis left out anyone who reported a history of cancer or heart and blood vessel disease, and anyone who died in the first 5 years. People were followed for 13 to 14 years on average, and 33,187 died, 19,419 men and 13,768 women (Svensson, 2021).

• The link between sleep length and the risk of dying was J-shaped for men and for women, with its lowest point at 7 hours (Svensson, 2021).
• For deaths from heart and blood vessel disease, 7 hours also had the lowest risk for men and women (Svensson, 2021).

How far to trust these numbers

• This study watched what people already did. The authors say it can't show sleep length causes death (Svensson, 2021).
• Sleep was asked about once for this analysis. Their sleep may have changed over the years, and people who sleep less may overstate how long they sleep (Svensson, 2021).
• People with illnesses other than cancer and heart and blood vessel disease stayed in the study. Illness can change sleep length, and the authors say this may make the link look stronger than it is (Svensson, 2021).
• The study had no record of how well people slept, of depression or of how well off they were (Svensson, 2021).
• Everyone lived in East Asia, where the link between long sleep and dying has looked stronger than in North America or Europe, so the results may not apply elsewhere (Svensson, 2021).

Sources

• Svensson T et al. Association of sleep duration with all- and major-cause mortality among adults in Japan, China, Singapore, and Korea. JAMA Netw Open, 2021.