Which is worse, poor sleep or no exercise?

This study can't pick one. Each came with a similar rise in the risk of dying. Poor sleep came with a 23% higher risk of dying than good sleep (Huang, 2022). Doing no moderate or hard exercise came with a 25% higher risk of dying than exercising a lot (Huang, 2022). The study didn't test the two habits against each other. Poor sleep with no exercise came with the highest risk of dying, 57% higher than in good sleepers who exercised a lot (Huang, 2022). Poor sleepers who exercised did better than those who didn't. This was in 380,055 UK adults followed for about 11 years on average (Huang, 2022). The study can't show that either habit caused the deaths. Every percentage links to its source.

The short version

• Neither was clearly worse. Poor sleep came with a 23% higher risk of dying and no exercise with a 25% higher risk (Huang, 2022). Picture two groups of the same size. Either figure is about 5 deaths for every 4.
• Poor sleepers who did no exercise had the highest risk of dying, 57% higher than good sleepers who exercised a lot (Huang, 2022). In two groups of the same size, that is about 11 deaths for every 7.
• Poor sleepers who exercised did better. At the weekly activity minimum or above, most of poor sleep's extra risk of dying was gone (Huang, 2022). The World Health Organization sets that minimum at 150 minutes of moderate activity a week (WHO, 2026).
• At every sleep level, people who did no moderate or hard exercise had a higher risk of dying than good sleepers who exercised a lot (Huang, 2022).
• This study followed what people already did, so it can't show the habits caused the deaths (Huang, 2022).

Poor sleep and no exercise, side by side

Each row compares one group with another. The other habit and things like age, smoking and weight are taken into account. The likely range is in brackets. The last column pictures the same risk as deaths in two groups of the same size. Those counts are a way to picture the percentage. They are not deaths the paper counted.

The first two rows use different comparison groups, and the paper's text reports no direct test of one against the other. So the study can't say which is worse. Only about 3 in 100 people were poor sleepers (Huang, 2022). The percentages compare risks between groups. They aren't what any one person would gain or lose by changing a habit.

Poor sleepers who exercised did better

• The extra risk of dying that came with poor sleep was bigger in people who did less exercise (Huang, 2022).
• In people at or above the weekly activity minimum, most of that extra risk was gone. The authors write that this much activity "eliminated most of the deleterious associations of poor sleep with mortality" (Huang, 2022).
• The study's minimum is the lower limit of the World Health Organization's guideline, counted in the study as 600 MET-minutes a week (Huang, 2022). The WHO recommends at least 150 minutes of moderate activity a week (WHO, 2026).
• Poor sleep still came with some extra risk in the group that exercised most. The authors write that its effects "remain even within participants with a high PA level" (Huang, 2022). PA is physical activity.
• The paper's text gives no exact figure for poor sleepers who exercised (Huang, 2022).
• At every sleep level, people who did no moderate or hard exercise had a higher risk of dying than good sleepers who exercised a lot. The risk was highest when they also slept poorly (Huang, 2022).

The 5 signs of healthy sleep the study scored

People answered questions about their sleep at the start of the study. Each of these counted as one sign of healthy sleep (Huang, 2022):

• Being a morning person
• Sleeping 7 to 8 hours a day
• Not usually having insomnia
• Not snoring
• Not often feeling sleepy in the daytime

Good sleep meant 4 or 5 of the signs, and poor sleep meant at most 1. People with 2 or 3 were in between. About 56 in 100 people had good sleep. About 42 in 100 were in between. About 3 in 100 had poor sleep. The paper's shares are rounded, so they add to 101 (Huang, 2022). The score is this study's way of sorting people into groups. It is not a medical test.

What counted as exercise

People reported how much physical activity they did in their free time. Activity at work, for travel or around the house wasn't counted (Huang, 2022). "No exercise" here means they reported no moderate or hard activity. "Lots of exercise" means at least twice the study's weekly minimum, with some moderate or hard activity. That is the upper limit of the WHO guideline (Huang, 2022).

The study of 380,055 people

Huang and colleagues used the UK Biobank. People were 55.9 years old on average, and 55% were women. Nobody had a history of heart disease or cancer at the start. People with a history of obstructive sleep apnoea or with very severe obesity were left out too. Follow-up was about 11 years on average, and 15,503 people died, about 4 in 100 (Huang, 2022).

• For dying of heart and blood vessel disease, poor sleepers who did no moderate or hard exercise had a 67% higher risk than good sleepers who exercised a lot. For dying of cancer the figure was 45% higher (Huang, 2022).
• Poor sleep, compared with good sleep, came with a 39% higher risk of dying of heart and blood vessel disease (Huang, 2022).
• The results are adjusted for age, sex, body mass index, smoking, alcohol, fruit and vegetables, sitting time, mental health, work and how well off people were (Huang, 2022).

How far to trust these numbers

• This study followed what people already did. The authors say that rules out a firm conclusion about cause (Huang, 2022).
• The study used the sleep and exercise people reported at the start. A smaller group was asked again within 7 years. Of them, 59% had kept their activity level. 68% had kept their sleep group (Huang, 2022).
• People with healthier sleep were also younger and slimmer, and more of them had never smoked. The study took these into account, but other differences may remain (Huang, 2022).
• Some of the combined groups were small. The authors say that may have made real differences harder to detect (Huang, 2022).
• Only about 5.5% of the people invited joined the UK Biobank, so they aren't typical of the UK (Huang, 2022).
• The authors say the results may not apply to extremely active people, such as athletes (Huang, 2022).

Sources

• Huang BH et al. Sleep and physical activity in relation to all-cause, cardiovascular disease and cancer mortality risk. Br J Sports Med, 2022.
• The same paper, full text at the journal.
• World Health Organization. Physical activity, fact sheet. 2026.