Morning versus evening exercise and sleep
Systematic review (25 studies, 4233 participants) with a meta-analysis of 16 RCTs (n=611): no clear overall advantage of morning over evening exercise for most sleep outcomes; evening/night exercise went with more wake after sleep onset (SMD −0.51, very low certainty, small-study effects), mostly at moderate intensity
A systematic review of comparative studies of morning versus evening/night exercise and sleep, with a meta-analysis restricted to RCTs. Content from the abstract plus the full text (marked).
Participants
25 studies: five observational, two non-randomized intervention studies, 18 RCTs; 16 RCTs in the meta-analysis. 4233 participants in all; 611 in the meta-analysis (full text).
Protocol
Morning sessions 06:00-12:00 (mean ~08:14), evening/night 16:00-23:00 (mean ~18:22) (full text); mostly aerobic; acute (8 studies) and repeated (≥5 days, 17 studies) exposures (full text).
Key Results
At the group level the evidence does not indicate a clear overall advantage of morning over evening/night exercise for most sleep outcomes.
Evening/night exercise went with higher wake after sleep onset (WASO): SMD = −0.51 (95% CI −0.89 to −0.12); potential small-study effects; GRADE very low, I² 63% (full text).
Sleep onset latency: no timing effect (k = 13, SMD 0.04, 95% CI −0.16 to 0.24; GRADE low) (full text).
Intensity moderated the WASO effect: at moderate intensity SMD −0.62 (−0.88 to −0.36) favouring morning; at low intensity a small opposite trend; the high-intensity estimate (−1.41) came from a single study (full text).
Authors' conclusion: support flexible exercise timing that prioritises adherence and individual constraints; evening moderate-to-vigorous exercise may impair sleep continuity.
Practical Application
"Will evening training wreck my sleep?" A 2026 meta-analysis of 16 RCTs found no clear overall advantage of morning exercise for most sleep outcomes; evening exercise went with somewhat more waking in the night (overall SMD −0.51, very low certainty), clearest at moderate intensity (SMD −0.62, full text), while falling asleep was not affected.
Train when the session will actually happen; if the evening is the only slot, keep the hard work as early in the evening as the day allows and watch for night waking. Not from this paper; the base's own practice (SYNTH-07).
Consistent with REF-974 (a single evening session did not harm sleep; vigorous work ending ≤1 h before bed might).
Limitations
Mostly non-athletes; evening sessions mostly early evening; few late-night trials; the high-intensity WASO estimate rests on a single study (6 of 16 RCTs were high-intensity overall).
WASO result: very low certainty, heterogeneous, possible small-study effects.
Timing defined by clock time in most studies, not by chronotype or wake time.
How this entry was checked and changed
Why this level
SR/MA of controlled trials, but small trials, heterogeneity (WASO I² 63%) and small-study effects; GRADE low to very low