Skip to content
Sleep & circadian healthTrial meta-analysis8 min read

Exercise improved sleep-quality scores in adults 45 and older with sleep disorders

MoveVitals Editorial Collective · Study published · Reading updated

Original paper:Comparative effects of exercise modalities on sleep quality in middle-aged and older adults with sleep disorders: A systematic review with pairwise, network, and dose-responsemeta-analyses · Journal of Exercise Science & Fitness · 2026 · Original paper ↗

Key takeaway

Pooling 23 randomized trials with 1,478 adults aged 45 to 77 who had sleep disorders, exercise improved Pittsburgh Sleep Quality Index scores (SMD 0.62, 95% CI 0.40–0.83), and aerobic exercise showed the largest effect in the network analysis (SMD 1.34, 95% CrI 0.86–1.82). Certainty was low to moderate, the outcomes were questionnaires, and the best weekly dose of about 539 MET-min/week is a model estimate, not a prescription.

Study design

Trial meta-analysis

Reading basis

Full text

Reading focus

Design & applicability

The research question

Sleep problems grow more common with age, and the introduction puts their prevalence near 50% in adults over 65; chronic sleep disorders are also linked to high blood pressure, diabetes, cardiovascular disease and other conditions. Exercise is safe, cheap and easy to scale, and earlier meta-analyses found it improves subjective sleep quality, but most pooled all exercise together or looked at general populations or older adults, so they could not rank one type against another, and few asked how much exercise is enough. The authors set out to compare exercise types within one Bayesian network meta-analysis in middle-aged and older adults with sleep disorders, and to model how total weekly exercise dose relates to improvement in sleep quality.

Who and how

This was a systematic review with random-effects pairwise, Bayesian network and Bayesian dose-response meta-analyses, registered in PROSPERO and reported to PRISMA standards. Two reviewers searched PubMed, Web of Science, Embase and the Cochrane Library from inception to May 2026 and screened reference lists. Of 13,058 records, 2375 were duplicates and 60 reached full-text assessment, leaving 23 randomized controlled trials with 1,478 participants aged 45 to 77. Eligible adults were aged 45 or older with a clinically diagnosed sleep disorder or clinically significant sleep disturbance on validated instruments, so the trials mixed insomnia, obstructive sleep apnoea and screened poor sleep quality. Any exercise counted (for example Tai chi, walking or high-intensity interval training), compared with sedentary or non-exercise controls such as health education or standard care; exercise combined with medication was excluded. The 24 exercise arms comprised mind–body exercise (9), aerobic exercise (7), combined exercise (5), walking (2) and high-intensity interval training (1).

The primary outcome was subjective sleep quality on the Pittsburgh Sleep Quality Index (PSQI); secondary outcomes were insomnia severity (Insomnia Severity Index, ISI) and daytime sleepiness (Epworth Sleepiness Scale, ESS), taken immediately after the intervention and analysed pairwise only. Effects are standardized mean differences (SMD) arranged so a positive value means benefit. Heterogeneity was measured with I² and τ², and publication bias with Egger’s and Begg’s tests. The network analysis (PSQI only) reported SMDs with 95% credible intervals (CrI) and ranked treatments by SUCRA. For the dose-response analysis, weekly dose in MET-min/week was calculated as MET value times session minutes times weekly frequency, using the 2011 Compendium of Physical Activities, and sorted into pragmatic nodes of 250, 500, 750 and 1000; several curve shapes were compared and the quadratic model gave the peak. Risk of bias used Cochrane RoB 2 and certainty of evidence used GRADE.

What it found

  1. 1
    Exercise improved sleep quality and insomnia severity but not clearly daytime sleepiness. Pooled across trials, PSQI improved (SMD 0.62, 95% CI 0.40–0.83, I² 65.8%) and ISI improved (SMD 0.64, 0.34–0.94, I² 49.2%), while ESS did not reach significance (SMD 0.70, 95% CI −0.05 to 1.46, I² 90.1%). With heterogeneity that high and an interval that crosses zero, the sleepiness result is uncertain rather than a clear null.
  2. 2
    In the network analysis of PSQI, all three exercise types beat control. Aerobic exercise gave an SMD of 1.34 (95% CrI 0.86–1.82), mind–body exercise 0.64 (0.30–0.97) and combined exercise 0.60 (0.13–1.07).
  3. 3
    Aerobic exercise ranked first but with caveats. Its SUCRA was 99.1%, against 52.1% for mind–body, 48.6% for combined and 0.2% for control, and it beat mind–body (SMD 0.70, 95% CrI 0.12–1.29) and combined exercise (0.74, 0.07–1.41), while mind–body and combined did not differ (0.04, −0.54 to 0.61). The pairwise aerobic estimate (SMD 1.52, 95% CI 0.57–2.48, I² 83.0%) shrank under sensitivity analyses with heterogeneity persisting, and the authors say rankings remain uncertain because the evidence is sparse and heterogeneous.
  4. 4
    The dose-response curve was an inverted U. The predicted benefit rose with weekly dose and peaked at about 539 MET-min/week (predicted SMD 1.09, 95% CrI 0.63–1.62; credible interval for the optimal dose 449 to 900), then faded. A significant effect first appeared at about 225 MET-min/week (SMD 0.71, 0.35–1.12) and held up to about 900 (SMD 0.65, 95% CrI 0.004–1.392); alternative models gave broadly similar ranges.
  5. 5
    The benefit did not clearly differ by type of sleep disorder. PSQI improved among people with insomnia or poor sleep quality (SMD 0.62, 95% CI 0.21–1.03), and the test for subgroup differences was not significant (χ² = 0.15, df = 2, P = 0.927). The authors report no significant PSQI improvement in the sleep apnoea subgroup, which they partly attribute to few studies.
  6. 6
    Secondary outcomes varied by type. Mind–body exercise improved ISI (SMD 0.78, 95% CI 0.30–1.25) and was the only ESS subgroup reported as significant (0.48, 0.01–0.96, a lower bound barely above zero), while combined exercise improved ISI (0.57, 0.06–1.08). Egger’s and Begg’s tests found no significant publication bias (P > 0.05), though with few trials for ESS and ISI they have little power.

Limits & uncertainty

  1. 1
    The authors note that incomplete reporting prevented adjusting for effect modifiers such as type and severity of sleep disorder, comorbidities, psychological status, physical function, medication and lifestyle, and that some trials were small with considerable heterogeneity in exercise protocols, control conditions and outcome measures. These factors limit robustness and generalisability.
  2. 2
    The authors note that each exercise type had few studies and dose gradients were insufficient, so no dose-response curve could be drawn for individual types; ISI, ESS and sleep-disorder subgroups were analysed pairwise only; exercise type was partly tied to population and geography, so residual confounding cannot be excluded; and language restrictions and the lack of long-term follow-up may have affected completeness.
  3. 3
    The authors stress that the 539 MET-min/week peak is an exploratory, model-dependent estimate rather than a validated target, given its wide credible interval, the absence of studies with a low overall risk of bias, heterogeneity, imprecision and possible small-study effects.
  4. 4
    The authors note that sleep quality was mostly rated on subjective scales and that participant blinding is difficult in exercise trials, so self-report and expectation effects may inflate results. Of 23 trials, 5 were rated high risk of bias and 18 had some concerns, so none was rated low risk; certainty was low to moderate.
  5. 5
    Our observation: the dose-response analysis could classify only 14 of the 24 exercise arms, with session durations borrowed from similar studies when a trial did not report them, and arms were snapped to four dose nodes. The population is also a mix of insomnia, sleep apnoea and screened poor sleep quality, so a single pooled effect hides quite different patients.

What to take from it

Regular aerobic exercise, mind–body exercise such as Tai Chi or yoga, or a mix of both were each linked to better self-rated sleep quality in adults 45 and older with sleep problems. The range with a significant effect was roughly 225 to 900 MET-min/week. For scale, the paper’s own table codes 35-minute aerobic sessions four times a week at 500 MET-min/week and 60-minute sessions three times a week at 750. Exercise looks like a worthwhile addition to care for sleep problems, not a replacement for assessing them; anyone with persistent insomnia, suspected sleep apnoea or heavy daytime sleepiness should speak to a clinician.

It does not show that aerobic exercise is best for you, that 539 MET-min/week is your personal optimum, or that more exercise makes sleep worse; the inverted U is a modelled average from a small set of trials. It also does not show that exercise fixes daytime sleepiness. Certainty is low to moderate for the improvement in sleep scores and low for the ranking and the dose peak.

Reading your own data

The outcomes were questionnaire scores, so judge progress with how you feel and, if you like, a repeat PSQI-style questionnaire every few weeks, alongside your wearable’s sleep trend; do not translate the effect into extra deep-sleep minutes. The included trials lasted from 4 to 52 weeks, so compare at least a month or two of weekly active minutes before and after a change, using the same device. Wearable ‘active minutes’ or zone scores are not the intensity-weighted MET-min/week the paper used, so treat any dose conversion as rough. Illness, shift changes and a new medication can move sleep data independent of exercise.

How to read this evidence

Pooling trials can improve precision, but compare populations, controls, bias and uncertainty. Network rankings include indirect evidence and do not identify a universal best treatment.

Sources & attribution

Comparative effects of exercise modalities on sleep quality in middle-aged and older adults with sleep disorders: A systematic review with pairwise, network, and dose-responsemeta-analyses

Journal of Exercise Science & Fitness · 2026-10-01

DOI: 10.1016/j.jesf.2026.200526 ↗PubMed · 42898167 ↗

MoveVitals editorial interpretation · educational content · no individual clinical review recorded

Sources checked 2026-11-01

Practical guides

Read in context

·Cohort study

Sleep regularity tracked mortality more closely than sleep duration in 60,977 UK adults

Among 60,977 UK Biobank adults who wore a wrist accelerometer for a week, the most regular fifth of sleepers had a fully adjusted all-cause mortality HR of 0.70 (95% CI 0.59–0.83) against the least regular fifth, over about 6.3 years and 1859 deaths. Regularity fitted the data better than sleep duration, but the study is observational and rests on one week of sleep data.

Sleep

In-depth explainersRead study →
·Randomised trial

Sleep extension cut daily energy intake by about 270 kcal in adults with overweight

In 80 adults with overweight who usually slept under 6.5 hours, two weeks of sleep-extension counselling added about 1.2 hours of sleep and lowered measured energy intake by 270.4 kcal/d versus control (95% CI −393.4 to −147.4). The trial was short, single-centre and in a narrow group, so it does not show lasting weight loss or any benefit for people who already sleep enough.

JAMA Internal Medicine

In-depth explainersRead study →
·Cohort study

Six hours or less of sleep at 50 and 60 linked to higher dementia risk

In 7959 British civil servants followed for about 25 years, sleeping six hours or less at age 50 (HR 1.22, 95% CI 1.01–1.48) or age 60 (HR 1.37, 1.10–1.72) was linked to a higher dementia rate than sleeping seven hours, and persistent short sleep gave an HR of 1.30 (1.00–1.69). Sleep was mostly self-reported and the study is observational, so early brain disease or other factors may explain part of the link.

Nature Communications

In-depth explainersRead study →
·Cohort study

Students with irregular sleep had later melatonin timing and lower grades

Among 61 Harvard undergraduates tracked for about 30 days, the most irregular sleepers had melatonin onset at 00:08 versus 21:32 in the most regular sleepers (p < 0.003), and sleep regularity correlated with grade point average (r = 0.37, p < 0.004). The sample was small and the design observational, so the data cannot show that irregular sleep lowers grades or shifts the body clock.

Scientific Reports

In-depth explainersRead study →
Sleep & circadian health →