Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Exercise and sleep

Older adults [1]
Depressed adults [2]
Postmenopausal women [3]

Imagine going to sleep at 10 and waking at seven the next morning. For some that is an unachievable nirvana, because of nights of broken sleep, or no sleep. A good night's sleep, even occasionally, is something many would love to have.

Being tired would seem to be a useful factor in promoting good sleep, but it isn't always like that. Or perhaps it is the right sort of tired - the tired that comes from physical, rather than mental, exercise. Bandolier performed a quick search for studies that might shed some light on this. It found three [1-3], but they hold out only limited hope.

Older adults [1]

The eligibility criteria for this trial were extensive (12 items), but included an age range of 50-76 years, being sedentary, with a moderate sleep complaint (getting to sleep, waking during the night, or waking and getting up in the morning, and not taking sleep medicines. The intervention involved four exercise sessions a week for 16 weeks, two in a class and two at home, with each session lasting 60 minutes, half of which was endurance training.

Participants were randomised to exercise or no intervention control. Outcomes were standard sleep scales measured before, during, and after the intervention period, using sleep diaries and sleep quality indices.

Forty-eight people (average age 62 years) were randomised, with 20 completing the exercise programme and 23 completing the no intervention control period. Adherence to exercise was high. Initial sleep onset was about 27 minutes in exercise and control groups at baseline, and fell to 15 minutes after 16 weeks of exercise (Figure 1). Initial sleep duration was about six hours in exercise and control groups, and rose to 6.8 hours after 16 weeks of exercise (Figure 2). Better exercise duration was associated with better night-time sleep, and less daytime sleep.

Figure 1: Average time to fall asleep at the start and end of therapy for exercise group and no intervention control

Figure 2: Average sleep duration at the start and end of therapy for exercise group and no intervention control

Depressed adults [2]

Eligibility for this study was based on a proper diagnosis of unipolar major or minor depression of dysthymia, and age over 60 years, as well as aerobic exercise more than twice a week. The intervention was high-intensity progressive resistance training three days a week for 10 weeks. The control was enrolment in a health education programme without exercise training. Outcomes were standard sleep scales measured before and after the intervention period, using sleep diaries and sleep quality indices.

Twenty-eight participants had an average age of 71 years, and adherence to the exercise programme was high. Exercise reduced the proportion of poor sleepers (Figure 3), and more exercise participants reported improved sleep, compared with controls (Table 1). Six of 15 (40%) depressed adults in the exercise group reported improved sleep, compared with none in the control group. Though numbers were small, this implies a number needed to treat of about 2.5 for one depressed adult to have improved sleep. Actually, it may be better than that, because two patients in the control group actually reported worse sleep.

Figure 3: Poor sleepers among depressed adults at start and end of intervention for exercise and control groups

Table 1: Change in sleep quality at the end of the intervention for exercise and control groups

Exercise (n=15)
Control (n=13)

Postmenopausal women [3]

Eligible women were post-menopausal, taking no hormone replacement therapy. Smokers, and those with medical conditions where exercise was contraindicated were excluded. They were randomised to an exercise programme, or a stretching programme. Exercise involved moderate aerobic exercise five days a week for a year, some at a centre, and some at home. Stretching involved a 60-minute low intensity stretching and relaxation session each week for a year.

Over a year exercise or stretching made little difference to sleep quality indicators, though stretchers used less sleep medication. Because the study was designed to assess morning versus evening exercise, an analysis showed that morning exercisers who exercised for more than 225 minutes a week had significantly less trouble falling asleep, while evening exercisers who exercised more than 180 minutes a week had more trouble falling asleep.


This is not be the whole literature on exercise and sleep. There are other studies that may have included some exercise with other interventions, and there are certainly observational studies. Some of these also associate improved sleep with more exercise.

The down side is that there are not many trials of exercise for sleep. They are in people with different sources of problem (depressed, postmenopausal), and while they may be randomised none could be blind. They are small studies, with just over 200 participants.

Yet all show a consistent pattern of improved sleep measures with exercise. It may be that some threshold of exercise is needed to obtain better sleep, or perhaps morning exercise is better than evening exercise, but some measures of improved sleep were consistently found. Exercise is good for the heart, for weight, for lungs, and for the bones. Whatever the effect on sleep, it is going to do some good. So for those who have poor sleep and want to do something about it, a useful new year's resolution might be to get in at least four to six hours of moderate exercise a week.


  1. AC King et al. Moderate-intensity exercise and sleep-rated quality of sleep in older adults. A randomized controlled trial. JAMA 1997 277: 32-37.
  2. NA Singh et al. A randomized controlled trial of the effect of exercise on sleep. Sleep 1997 20: 95-101.
  3. SS Tworoger et al. Effects of a yearlong moderate-intensity exercise and a stretching intervention on sleep quality in postmenopausal women. Sleep 2003 26: 830-836.

previous or next story