Practical Takeaways

Put the insomnia patient through 60 minutes of gentle Hatha/Restorative yoga, three times a week, for 12 straight weeks. Average nightly effect: + 110 min total sleep time, + 15 % sleep efficiency, – 56 min wake‑after‑sleep‑onset, – 29 min time‑to‑fall‑asleep, PSQI drops ≈ 6 points.

Run a 60 ‑min Yang‑style (24‑form) Tai Chi session, three mornings a week, for 12 weeks.
Average nightly effect: + 52 min total sleep time, + 7 % efficiency, – 36 min nighttime wake, – 25 min sleep‑latency. PSQI ↓ 4–5 pts and the gains hold up to a year.

Prescribe 30 min at 60–75 % HR max (can talk but puffing) on five days each week, maintained for 12 weeks. Outcome: ISI score falls by about 10 points — a clinically large improvement in perceived insomnia severity.

Effects of various exercise interventions in insomnia patients: a systematic review and network meta-analysis

Researchers began by searching six large medical databases — PubMed, Cochrane CENTRAL, Embase, Web of Science, SPORTDiscus, and PsycINFO. They took everything these databases contained from their first entries right through 1 April 2025 and kept only studies that were randomized controlled trials.

Next, they filtered those studies to adults with insomnia. Patients either had a formal diagnosis (DSM‑IV, DSM‑5, or ICSD‑3) or scored 8 or higher on the Insomnia Severity Index. After removing anything that did not fit, 22 trials with a combined total of 1,348 participants remained.

Each trial tested at least one of seven named exercise programs (yoga, Tai Chi, walking or jogging, aerobic plus strength training, strength‑only training, aerobic exercise combined with therapy, or mixed aerobic routines) and compared them with one of six control approaches (cognitive‑behavioural therapy, sleep‑hygiene advice, ayurveda, acupuncture or massage, usual‑care–style active controls, or a wait‑list).

The exercise programs lasted between 4 and 26 weeks, depending on the type:

  • Yoga: 4–16 weeks
  • Tai Chi: 12–26 weeks
  • Walking or jogging: 12–26 weeks
  • Aerobic + strength: 12–16 weeks
  • Strength‑only: 16 weeks
  • Aerobic + therapy: 4–12 weeks
  • Mixed aerobics: 6–16 weeks

Sleep outcomes were collected right after each program ended. A smaller set of studies also checked participants again 6–7 months later, and an even smaller set followed them for 1–2 years.

Different Effects

Most of the exercise and non‑exercise approaches outside the top three produced at least a small benefit, but none reached the pre‑set “clinically important” marks with reliable evidence. Here is what the review actually found, written straight and simply:

Aerobic plus strength sessions (12–16 weeks)

People slept a little better—insomnia scores fell by about two to three points, diary‑based sleep efficiency rose three to five percent, and objective monitors showed about 22 minutes less wake time after falling asleep. Those numbers were all smaller than the cut‑offs the researchers set for a change that matters to patients, so it was considered helpful but not decisively so. The supporting studies were judged moderate quality for the objective measures and low for everything else.

Strength‑training alone (16 weeks)

The two small trials (around sixty participants in total) never showed a consistent shift in any major sleep score. Because the samples were tiny and the results varied, certainty was rated low.

Aerobic exercise with a therapy add‑on (4–12 weeks)

One study combined treadmill work with sleep‑hygiene counselling. Participants’ sleep quality score (PSQI) dropped by about 5 points and their insomnia severity (ISI) by roughly 9 points—big enough changes to notice. But all of that evidence came from a single trial with some risk‑of‑bias problems, so confidence stayed low.

Mixed aerobic classes (6–16 weeks of varied cardio)

People logged 45–50 minutes more total sleep and six percent higher sleep efficiency in their diaries, yet those gains disappeared in sensitivity checks. Results differed a lot between studies, and no objective or long‑term figures backed them up, so certainty was graded very low.

Cognitive‑behavioural therapy (CBT)

Although not an exercise, CBT served as a benchmark. Across several trials it cut PSQI by 6 points, lengthened sleep by almost an hour, raised sleep efficiency 15 percent, and shortened wake periods by nearly an hour. Evidence ranged from moderate to low certainty, and because CBT is already the standard treatment, the paper used it as a reference rather than featuring it among the exercise options.

Sleep hygiene advice alone

One small trial achieved about a 3.9‑point drop in PSQI — just at the threshold of clinical importance —but the study was too small to lift certainty above very low.

Acupuncture or massage

Results were mixed and mostly not significant – risk of bias was high and protocols varied widely, leaving certainty very low.

Ayurveda

A single, under‑powered study showed no meaningful change, so no conclusions could be drawn.

In short, only yoga, Tai Chi and brisk walking or jogging produced changes large enough — and consistent enough — to meet the study’s predefined standards for a worthwhile, trustworthy improvement in insomnia. The other activities may help a little, but current data are too weak or inconsistent to say so with confidence.