Can Exercise Help with Insomnia? What the Research Says

9 min read

In a hurry? Scroll to the bottom for a summary.

What you’ll learn in this article

• Whether exercise can improve insomnia, and by how much
• The biological reasons exercise and sleep are connected
• What type, how much, and how often: based on a 2025 analysis of 86 clinical trials
• How long before you’ll notice a difference
• The point where more exercise starts making sleep worse, not better

If you’re searching this question, you probably haven’t slept well in a while and you’re looking for something that works. The short answer is yes: exercise improves sleep quality, and the evidence behind that claim is substantial. But the longer answer matters, because exercise is a tool with a specific range of effective use. Understanding that range is the difference between a habit that genuinely shifts your sleep and one that spins its wheels.

This post works through each of those questions in order.

The evidence: what exercise does to sleep quality

A 2025 network meta-analysis published in BMC Public Health synthesised 86 randomised controlled trials involving 7,276 participants and compared six types of exercise against control groups (Wang et al., 2025). Every single modality tested (aerobic exercise, resistance training, combined training, yoga, Pilates, and Traditional Chinese Sports) produced a statistically significant improvement in sleep quality. That’s a consistent finding across the evidence base.

To translate the effect size: imagine a dial that runs from sleeping very poorly to sleeping well. Exercise moves that dial. For aerobic exercise, the standardised mean difference was –1.21, which sits in the medium-to-large range by conventional benchmarks. It’s not a switch; you won’t wake up transformed after your first run. It’s a dial, and it moves meaningfully over weeks of consistent effort.

One honest caveat: the majority of studies in this meta-analysis covered people with poor sleep broadly, not people with a diagnosed insomnia disorder specifically. Clinical insomnia involves a layer of conditioned arousal and cognitive patterns that exercise doesn’t directly address (more on this at the end). The effect is real, but it’s likely more modest if your insomnia has been running for years.

Key finding A 2025 meta-analysis of 86 trials found that every form of exercise tested (aerobic, resistance, yoga, Pilates) produced a significant improvement in sleep quality. The effect is real and consistent across modalities.

Why exercise helps sleep: the biological pathways

Exercise doesn’t improve sleep through one mechanism. It works through several.

Sleep drive. Your brain keeps a running tab throughout the day. Every hour of wakefulness accumulates adenosine, a neurochemical that creates the growing pull toward sleep. Think of it as a debt your body is carrying: the heavier it gets, the harder sleep pulls at you by evening. Physical activity accelerates that accumulation. Sedentary days often produce restless nights not because something is wrong with your sleep, but because you haven’t run up enough of a tab.

Stress hormone clearance. People with insomnia tend to carry elevated baseline levels of cortisol and adrenaline, the physiological signature of a nervous system that won’t fully stand down. These hormones, when chronically elevated, interfere with the body clock’s ability to initiate and maintain sleep (Ehrnstrom & Brosse, 2016). Exercise acts as a pressure release: it metabolises these hormones in a way that sustained worrying or lying still in the dark does not.

Deep sleep. A 2021 crossover trial found that moderate daytime exercise produced a 33% increase in slow-wave sleep compared to a no-exercise control condition (Park et al., 2021). Slow-wave sleep is where the most physically restorative processes occur: cell repair, memory consolidation, immune function. If you wake feeling like you barely slept despite adequate hours, this is the stage worth supporting.

Body temperature. Exercise raises your core body temperature. The drop that follows, over the hours after you finish, mimics the natural temperature decline that signals the brain it’s time to sleep. This mechanism is one reason exercise has a mild preference for earlier in the day, though as we’ll cover, the timing question is less important than most people think.

What type of exercise is best for sleep

The Wang et al. 2025 meta-analysis is the largest and most detailed comparison of exercise types for sleep currently available, and its rankings are worth knowing.

By cumulative ranking probability (a statistical measure of relative effectiveness), Pilates came first at 91.7%, followed by aerobic exercise at 69.7%, combined aerobic and resistance training at 59.4%, resistance training at 58.6%, Traditional Chinese Sports at 40.5%, and yoga at 30.1%.

The Pilates ranking deserves a note of caution: it was supported by fewer studies than aerobic exercise, which means the confidence interval around that estimate is wider. Treat it as a promising signal rather than a firm verdict. Aerobic exercise has the deepest evidence base and the most consistent findings across studies.

What this means practically: if you run, cycle, walk briskly, or swim, you’re in the best-supported category. Any sustained aerobic activity that elevates your heart rate qualifies. Resistance training is also effective and meaningfully better than doing nothing. Yoga ranked lowest among the six, though still statistically significant. If yoga is what you’ll do consistently, do yoga.

Key finding Aerobic exercise has the strongest and most consistent evidence for sleep improvement. Resistance training is also effective. The best type of exercise for your sleep is the one you’ll do regularly.

How much exercise, and where more stops helping

Wang et al. 2025 found that the relationship between exercise dose and sleep quality follows a U-shaped curve. It’s not linear; more is not simply better. Think of it the way you’d think about training load in any sport: the right amount of stress produces adaptation; too much produces breakdown. Sleep responds to exercise volume the same way your body responds to overtraining.

The overall optimal dose was approximately 920 MET-minutes per week. MET (metabolic equivalent of task) is the standard unit exercise scientists use to quantify activity across different intensities and types. To translate this into something usable:

• A brisk walk at about 5 km/h burns roughly 4 METs per hour. Three 45-minute brisk walks per week gets you to about 270 MET-min/week, below the optimal range but still meaningfully better than sedentary.
• Moderate jogging (around 8 km/h, 8 METs) for 30 minutes, four times per week lands you at roughly 960 MET-min/week, near optimal.
• An intense hour-long cycling class (10 METs) seven days per week puts you at approximately 4,200 MET-min/week, well past the curve’s peak, into the zone where sleep benefits reverse.

The optimal dose also varies by exercise type. For Pilates, the sweet spot sits much lower, around 390 MET-min/week. For aerobic exercise, it’s approximately 1,100 MET-min/week. This means the “more is more” assumption that works in some fitness domains actively works against you when it comes to sleep.

Key finding The relationship between exercise volume and sleep quality is U-shaped. Benefits peak around 920 MET-minutes per week, roughly 150 minutes of moderate aerobic activity. Above that, gains flatten and can reverse.

The research also gives guidance on session parameters. A 2024 network meta-analysis of 58 RCTs found that sessions of 30 minutes or under produced stronger sleep effects than sessions of 40 to 55 minutes, and that 4 sessions per week was the optimal frequency (Wang, 2024). The likely reason for the session-length finding: prolonged exercise sustains cortisol elevation past the point that helps, tipping it from stress-clearance into stress-accumulation.

The practical prescription from the combined evidence:

• Frequency: 3 to 4 sessions per week
• Duration: 30 to 60 minutes per session
• Intensity: moderate. You should be working, not destroyed.
• Weekly volume: roughly 150 minutes of moderate activity, which aligns with standard public health guidelines

That last point isn’t a coincidence. The dose that’s optimal for sleep is the same dose that’s optimal for cardiovascular health, metabolic function, and mood. You don’t need a special sleep-exercise protocol. You need to meet the baseline.

How long before you notice a difference

This is the question most people don’t ask but should.

The Wang 2024 meta-analysis found that meaningful sleep improvements emerged at around 9 to 10 weeks of consistent exercise, with effects maintained and often strengthened at 24 weeks. Think of it the way you’d think about starting a new training block: you don’t expect cardiovascular gains in week one. The first few weeks are adaptation: your physiology is recalibrating, your adenosine system is resetting, your cortisol baseline is slowly coming down. Sleep benefits follow the same cumulative curve as fitness benefits.

This matters because the most common point to abandon an exercise habit is around weeks three to five, precisely when the early discomfort hasn’t yet converted into visible results. If you quit at week four, you leave before the intervention has run its course.

Two to three weeks in, you may notice modest changes in how heavy you feel at bedtime, or slightly less waking during the night. These are early signals, not the full effect. Stay with it.

When exercise starts working against your sleep

There is no clean threshold where exercise flips from helpful to harmful. It’s a direction you can drift in, often without noticing.

The research on overtraining distinguishes between functional overreaching (a short training spike your body recovers from in days) and non-functional overreaching, where accumulated load outpaces recovery over weeks. Sleep is one of the first systems to show the strain. A 2024 meta-analysis found that overreaching from endurance training produced a measurable drop in objective sleep efficiency, even when participants reported no change in how their sleep felt (Murphy et al., 2024). The body registers the problem before the brain does. Like the warning light coming on before the engine starts running rough.

What makes this tricky is that the early signs of too much exercise and too little sleep look nearly identical: daytime fatigue, low motivation, mood that doesn’t recover between sessions. People in this zone often respond by training harder. That tends to make things worse.

Signals worth paying attention to if your training load is high:

• You’re tired during the day but alert or anxious at night
• Your sleep feels unrestored despite adequate hours
• Your performance has plateaued or declined despite consistent effort
• Recovery between sessions takes longer than it used to

If several of these are present, reducing volume is a reasonable first experiment, before adjusting timing, changing your sleep environment, or concluding that you have an unfixable sleep problem.

Key finding Overreaching can reduce objective sleep efficiency while leaving subjective sleep quality apparently unchanged. People often don’t notice their sleep is getting worse until the deficit is significant.

A note on timing

Evening exercise has a reputation for disrupting sleep that the evidence doesn’t support. A real-world study tracking more than 12,600 people found that moderate-to-vigorous exercise within two hours of bedtime was associated with slightly more sleep and easier sleep onset, not worse sleep (Kahn et al., 2021). A 2022 network meta-analysis added a useful refinement: high-intensity exercise late in the evening can delay sleep onset in some people, but moderate-intensity exercise at night does not (Yang et al., 2022).

The practical upshot: exercise when it fits your life. If hard evening sessions reliably leave you wired, dial back the intensity rather than eliminating evening exercise altogether. Timing is a secondary variable. Dose and consistency are what matter.

What exercise won’t do

Exercise is a useful tool for sleep. It’s not a treatment for insomnia disorder.

Chronic insomnia involves conditioned wakefulness (the bed becomes a cue for alertness rather than sleep) and cognitive patterns (catastrophising about sleep, hypermonitoring, performance anxiety) that movement doesn’t touch. These patterns have a physiological signature; they’re not just in your head. They keep the nervous system in a state of readiness at the exact moment it needs to stand down.

If you’ve been sleeping poorly for more than three months, exercise consistently, and still aren’t sleeping well, the problem has almost certainly developed a cognitive and physiological dimension that movement alone won’t resolve. Cognitive Behavioural Therapy for Insomnia (CBT-I) is the first-line recommended treatment, with the strongest and most durable evidence for chronic insomnia. It addresses the parts of the problem exercise can’t reach.

What this looks like in practice

A client I’ll call Marcus came in exercising six days a week. Mid-forties, works in finance. He’d been sleeping poorly for two years and had doubled down on training, reasoning that more physical exhaustion would eventually tip him into sleep. His daytime energy was poor, his motivation had dropped, and he was waking most mornings between four and five. When we looked at his weekly training volume, it was sitting well above the range the evidence identifies as optimal. He reduced to four sessions, shortened their duration, and kept his evening sessions in place. His sleep didn’t change quickly. It rarely does. But eight weeks in, the early-morning waking had become less consistent, and his daytime energy had started to return. He’d been using exercise as a solution to a problem that exercise was partly sustaining.

TL;DR: What we covered

• Yes, exercise improves sleep. A 2025 meta-analysis of 86 trials found that every major form of exercise (aerobic, resistance, yoga, Pilates) produced significant improvements in sleep quality. The effect is real and consistent.
• The biology runs through several pathways: adenosine accumulation builds sleep pressure, exercise clears excess cortisol and adrenaline, deep sleep increases, and the post-exercise temperature drop supports sleep onset. These work independently; any one of them would be worth having.
• Aerobic exercise leads, but type matters less than consistency. Pilates ranked first in the 2025 analysis but has fewer studies behind it. Aerobic exercise is the best-supported option. The best type is the one you’ll do.
• Optimal dose: roughly 150 minutes of moderate activity per week, spread across 3 to 4 sessions of 30 to 60 minutes. More than this can reverse the gains. The dose curve is U-shaped, not linear.
• Give it 9 to 10 weeks. Meaningful improvements typically emerge around this point, with effects that strengthen over six months. Most people quit around week four, right before the gains arrive.

Work with a sleep specialist in BC

If what you’ve read here resonates and you’re still not sleeping well despite exercising consistently, exercise may not be the missing piece. If you’d like to talk through what’s keeping you awake, I offer free 15-minute consultations. No commitment. Just a conversation about your specific situation.

Book a free consultation →

Graeme Thompson, MA, RCC is a Registered Clinical Counsellor in British Columbia (RCC #21951, BCACC) with specialised training in Cognitive Behavioural Therapy for Insomnia (CBT-I). His practice, BC CBT-I, is a virtual insomnia therapy service serving clients across BC. His mission is to improve the waking lives of people struggling with insomnia and sleep.

References

Ehrnstrom, C., & Brosse, A. L. (2016). End the insomnia struggle: A step-by-step guide to help you get to sleep and stay asleep. New Harbinger Publications.

Kahn, M., Korhonen, T., Leinonen, L., Martinmaki, K., Kuula, L., Pesonen, A. K., & Gradisar, M. (2021). Is it time we stop discouraging evening physical activity? New real-world evidence from 150,000 nights. Frontiers in Public Health, 9, 772376. https://doi.org/10.3389/fpubh.2021.772376

Murphy, C., Svansdottir, S. A., Dupuy, O., & Louis, J. (2024). Does overreaching from endurance-based training impair sleep: A systematic review and meta-analysis. PLOS ONE, 19(5), e0303748. https://doi.org/10.1371/journal.pone.0303748

Oda, S., & Shirakawa, K. (2014). Sleep onset is disrupted following pre-sleep exercise that causes large physiological excitement at bedtime. European Journal of Applied Physiology, 114(9), 1789–1799.

Park, I., Díaz, J., Matsumoto, S., Iwayama, K., Nabekura, Y., Ogata, H., Kayaba, M., Aoyagi, R., Yajima, K., Satoh, M., Tokuyama, K., & Vogt, K. E. (2021). Exercise improves the quality of slow-wave sleep by increasing slow-wave stability. Scientific Reports, 11, 4410. https://doi.org/10.1038/s41598-021-83817-6

Wang, H., Xin, X., & Pan, Y. (2025). The best approaches and doses of exercise for improving sleep quality: A network meta-analysis and dose-response relationship study. BMC Public Health, 25, 1371. https://doi.org/10.1186/s12889-025-22570-1

Wang, P. (2024). Optimal exercise dose and type for improving sleep quality: A systematic review and network meta-analysis of RCTs. Frontiers in Psychology, 15, 1466277. https://doi.org/10.3389/fpsyg.2024.1466277

Yang, Y., Shin, J. C., Li, D., & An, R. (2022). Different intensities of evening exercise on sleep in healthy adults: A systematic review and network meta-analysis. Nature and Science of Sleep, 14, 2157–2177. https://doi.org/10.2147/NSS.S388863