Gentle relaxation beats harder workouts for perinatal sleep

The perinatal period runs from the start of pregnancy to one year after birth, and it is a stretch when sleep tends to suffer just as it matters most. Exercise is already known to improve sleep in general, but for pregnant and postpartum women the practical questions have gone unanswered: which kind of movement, how often, how long per session and over how many weeks. Previous trials tested single approaches in isolation, leaving no way to rank them against each other. This review set out to compare exercise types and doses within one framework and identify the most promising combination.

The authors searched nine databases, including PubMed, Web of Science, the Cochrane Library, Embase and Scopus, for randomised controlled trials of exercise and sleep quality in perinatal women, up to August 5, 2025. They found 23 eligible trials with 1,862 participants. Five exercise types were compared: aerobic exercise, aerobic combined with resistance training, yoga, Pilates and relaxation exercises. A network meta-analysis, which allows indirect comparisons across trials, then ranked programme lengths from 4 weeks or less up to 13-16 weeks, weekly frequencies of 1-2, 3 or 5 sessions, and session lengths of under 30 minutes, 30-60 minutes or over 60 minutes.

Relaxation exercises ranked highest of the five exercise types for improving sleep quality, with a clear pooled advantage over comparison conditions. The dose rankings favoured surprisingly light schedules. Programmes lasting 4 weeks or less ranked above longer ones, a frequency of 1-2 sessions per week ranked above 3 or 5 sessions, and sessions of 30-60 minutes ranked above both shorter and longer ones. Taken together, the analysis points to a gentle, low-frequency, moderate-length format as the relatively most effective combination in this population, though these are relative rankings rather than proof that other formats do nothing.

Earlier work had already suggested exercise helps sleep during pregnancy and after birth, but this is among the first attempts to rank exercise types and doses against each other in a single analysis rather than trial by trial. The result complicates the common assumption that more exercise, done more often and for longer, is always better for sleep. Because network meta-analysis leans on indirect comparisons, and because the included trials varied in design and populations, the rankings need confirming in head-to-head studies. The authors themselves call for more work on types, dosages and combinations before the recipe is treated as settled.

For anyone pregnant or in the first year after birth who is sleeping badly, this is a reason to think gentle, unhurried movement can be a legitimate sleep strategy, not just vigorous workouts. It also suggests the barrier to entry is lower than many expect, since the best-ranked schedule involved only one or two sessions a week and benefit appeared within about a month. None of this replaces individual medical guidance during pregnancy, and the rankings are relative rather than absolute, but the direction of the evidence favours starting small and calm.