Exercise linked to lower postpartum depression scores in 24 trials

Low mood after childbirth is common, and for some women it deepens into postpartum depression, a problem that affects both mother and baby. Medication and talking therapies are not always accessible, acceptable or practical in the newborn period, so interest in exercise as a supportive option has grown. Plenty of individual trials exist, but they differ in the type of movement offered, how much of it, and whether they aimed to prevent symptoms or treat them. This review set out to pool that randomised evidence and ask not just whether exercise helps, but which form and weekly amount look most promising.

The researchers searched eight databases from inception through 23 July 2026, plus three trial registries in the final update, for randomised controlled trials of postpartum exercise reporting scores on the Edinburgh Postnatal Depression Scale. Twenty-four trials with 1,792 unique participants qualified, 912 assigned to exercise and 880 to control. Standard pairwise meta-analysis compared exercise against usual care or no exercise, while network meta-analyses attempted to rank exercise types against each other and to compare pragmatic categories of weekly exercise volume.

Exercise was associated with Edinburgh scale scores 2.73 points lower than usual care, an average across trials whose results varied enormously. How the sessions were delivered made no detectable difference in an overall comparison, and the composition of the intervention showed no subgroup difference either. Programmes shorter than 12 weeks showed a larger pooled effect than longer ones, though the authors flag this as exploratory. In the rankings, yoga and mind-body exercise came out top and was estimated to beat aerobic exercise, but that comparison was entirely indirect and rated very low certainty. A weekly volume of 45 to 90 minutes ranked highest, yet no volume comparison reached statistical significance.

The headline direction fits the broader literature suggesting movement supports mental health, and this review extends it by attempting to rank types and doses rather than just asking whether exercise works at all. Those rankings, however, rest on star-shaped networks in which active treatments were rarely compared head to head, so the yoga signal is a hypothesis rather than a finding. The included trials also mixed women recruited to prevent symptoms with women already experiencing them, and many ran alongside other support. Which exercise type genuinely works best, and at what dose, remains open.

For a new mother weighing up how to protect her mood, this is evidence that structured movement is associated with meaningfully lower depressive symptom scores, whatever form it takes. It is not evidence that yoga beats a brisk pram walk, since that comparison was too indirect to trust. The authors are clear that exercise sits alongside, not instead of, proper clinical assessment and treatment where symptoms are significant. Knowing the benefit appeared across many formats may be the most useful part: the best exercise is plausibly the one that fits newborn life.