Surveys indicate that up to 12% of Americans suffer from chronic insomnia. While effective treatment in the form of cognitive behavioral therapy for insomnia (CBT-I) is a solution for some, CBT-I can be time- and cost-intensive and doesn’t work for everyone as a standalone treatment. Others opt for prescription-based or over-the-counter sleep aids, but as explained on The Drive by sleep expert Dr. Matthew Walker, these options generally fail in reproducing key aspects of natural sleep patterns, such as deep sleep and REM sleep, and many come with risks to other aspects of health. Recently, however, researchers conducted a meta-analysis of randomized controlled trials (RCTs) to evaluate another intervention for its ability to promote quality sleep – an intervention that offers countless benefits to other areas of health and yet is accessible to virtually everyone and comes with a low price tag: exercise.1
What they did
The study sought to determine the effects of exercise interventions on both objective and subjective sleep metrics among participants with insomnia. The investigators pooled data from 19 studies that used exercise alone (i.e., not in conjunction with other sleep interventions) in a total of 1,233 participants who did not have other sleep-related disorders. Included studies varied in duration between four weeks and 12 months.
The effect of exercise was compared to inactive controls for four objective sleep metrics measured via polysomnography and actigraphy: total sleep time (TST), sleep onset latency (SOL, the amount of time it takes to fall asleep), wake after sleep onset (WASO, the total time awake during the night after falling asleep), and sleep efficiency (SEI, the ratio of TST to total time in bed). Additionally, the researchers used data from sleep questionnaires to assess three subjective sleep metrics: subjective TST (i.e., the participant’s perception of their total sleep time), subjective SOL (the participant’s perception of their sleep onset latency), and sleep quality (SQ, how rested participants felt upon waking and their overall satisfaction with sleep).
What they found
Despite the significant heterogeneity across study interventions, exercise interventions had a small but significant positive effect on objective sleep parameters, as indicated by a standardized mean difference (SMD) of 0.37 (95% CI: 0.17 to 0.57). (For a detailed explanation of SMD, see this newsletter, but in brief, this is a summary statistic to assess the magnitude of a given effect in pooled data, with values between 0.2–0.5 indicating a small effect, values between 0.5–0.8 indicating a moderate effect, and values over 0.8 indicating a large effect.) The impact of exercise on objective sleep parameters collectively was driven by improvements in WASO (SMD=0.50; 95% CI: 0.14 to 0.86) and SEI (SMD=0.66; 95% CI: 0.27 to 1.06). However, exercise did not improve TST or SOL, meaning that subjects did not fall asleep faster and did not get more total sleep, though they were awake less throughout the night and spent more of their total time in bed asleep.
Exercise interventions also had a significant, large effect on subjective sleep parameters collectively (SMD=0.90; 95% CI: 0.61 to 1.19), though these metrics are more susceptible to bias than objective measures. The strongest effect was observed with subjective sleep quality (SMD=1.16; 95% CI: 0.70 to 1.61), though subjective TST (SMD=0.69; 95% CI: 0.34 to 1.05), and subjective SOL (SMD=0.48; 95% CI: 0.22 to 0.75) also exhibited significant improvements. This indicates that participants felt like they fell asleep faster, slept longer, and were more rested and satisfied with their sleep, even if these sentiments did not align in all cases with objective measurements.
Of note, the magnitude of improvement in sleep parameters from exercise interventions was greater with higher exercise intensity (though intensities for all included studies were considered light to low-moderate) and in studies with higher average participant age or a higher proportion of females. Insomnia can be due to age-related decreases in physical fitness, at least partially explaining why exercise intervention may be greater for older individuals. The greater effect in women could likewise be explained by lower average physical fitness in females due to historically lower rates of exercising, and potentially to worse sleep at baseline from hormone-related (i.e., menopausal) deficits in sleep quality.
Study limitations
It’s not a stretch to conclude that exercise is good for sleep, but efforts to derive more nuanced insights from this analysis are limited by the studies included within it. Criteria for chronic insomnia varied across studies, indicating different levels of severity for insomnia diagnoses, which could affect the magnitude of the effect from an exercise intervention. Similarly, nine of the included studies did not include information about the use of sleep aids, which also could impact the effects of an exercise intervention (either positively or negatively).
However, the biggest limitation of this study is the intensity range of the exercise interventions. The controls (no exercise intervention) were assigned an intensity of 0 metabolic equivalents (METs), and the exercise interventions had intensities of 1-4 METs (most studies used 3 METs). This is not how METs are typically represented. Usually, one MET is the amount of oxygen consumed while at rest, so this analysis included exercise interventions of 2-5 METs, oversampling the light end of the intensity spectrum.
It’s difficult to call an intensity of 2 METs exercising, as this is the energy required to put away laundry or eat while standing. An intensity of 3 METs would be achieved by walking on a treadmill at 2.0 to 2.4 mph at a 0% grade, while 4 METs corresponds to leisurely cycling at a speed of <10 mph, and 5 METs is the intensity from shooting baskets or kayaking at a moderate pace. Since the majority of included studies were at an intensity of 4 METs, there is limited ability to determine the dose-response effect of exercise intensity on sleep parameters, since sleep parameters very likely improve when the exercise intervention ramps up to an intensity higher than what was tested in the included studies. The limited intensity range also means that we don’t know whether an individual who is very active at baseline might see improvements in sleep by adding more exercise to their routine, though it certainly seems likely that ceasing all exercise would make insomnia worse.
The significant heterogeneity between interventions (type of exercise, frequency, duration) makes it impossible to give a specific exercise “prescription” for insomnia. Although exercise may be the last thing on your mind if it’s been a while since your last good night of sleep, the significant improvements in sleep quality at the low range of exercise intensity tell us that if you are not exercising, adding any form of physical activity is likely to improve your sleep. And even if you start at very low intensity, increasing exercise intensity over time may lead to further sleep improvements.
The bottom line
After a night (or week, or month) of poor sleep, exercise is often the last thing we want to do. Yet although daytime grogginess and fatigue from inadequate or low-quality sleep might seem like reasonable reasons to deprioritize your fitness routine, remaining sedentary can feed the vicious cycle of fatigue and poor sleep. Exercise, by contrast, has been shown to significantly improve sleep quality in people with insomnia. Although more research is needed to determine how exercise and other interventions (particularly CBT-I) might complement each other, exercise on its own is a cost-effective, accessible intervention to improve both objective and subjective measures of sleep.
For a list of all previous weekly emails, click here.
References
- Riedel A, Benz F, Deibert P, et al. The effect of physical exercise interventions on insomnia: A systematic review and meta-analysis. Sleep Med Rev. 2024;76(101948):101948. doi:10.1016/j.smrv.2024.101948