The Effects of Oral Contraceptives on Exercise Performance in Women: A Systematic Review and Meta-Analysis
Participants
42 studies, 590 participants in total (OCP users and naturally menstruating comparators). Most studies (83%) were graded moderate, low or very low quality; 17% high quality.
Analysis
Between-group: OCP users vs naturally menstruating women, phase-matched two ways — (1) OCP withdrawal vs early follicular phase; (2) OCP consumption vs all phases except early follicular
Within-group: performance during OCP consumption vs OCP withdrawal
Bayesian framework; effects expressed as the probability of at least a small effect (d ≥ 0.2)
Key Results
Probability of a small effect favouring OCP users: effectively zero (p < 0.001) under both comparison methods
Probability of a small effect favouring naturally menstruating women: moderate under method (1) (p = 0.40) and small under method (2) (p = 0.19) — i.e. if anything, OCP users perform slightly worse, but the group-level effect is most likely trivial
Relatively large between-study variance for both between-group comparisons (τ 0.16 [95% CrI 0.01-0.44] and 0.22 [95% CrI 0.06-0.45])
Within the OCP cycle: almost zero probability of a small effect in either direction (p ≤ 0.001) — performance was consistent across pill-taking and withdrawal days
Authors' conclusion: the evidence does not warrant general guidance on OCP use vs non-use; an individualised approach is more appropriate when performance is a priority
Mechanism proposed in the abstract's background: OCPs downregulate endogenous oestradiol and progesterone while supplying exogenous oestrogen and progestin daily — a hormonal milieu that differs from eumenorrheic women
Practical Application
Across 42 studies and 590 women, oral-contraceptive users performed at most trivially worse than naturally menstruating women (probability of even a small deficit 0.19-0.40 depending on phase matching), and their performance did not change between pill-taking and withdrawal days. Do not expect a meaningful performance cost from the pill, and do not expect a "phase" effect within a pill cycle.
An OCP cycle gave a flat performance profile in this analysis; for an athlete already on OCP that is a stable baseline for tracking training response.
Inference, not tested: formulation (mono- vs multiphasic, oestrogen dose) may matter — the abstract does not analyse it, and between-study variance was large, so any individual athlete may deviate from the pooled picture.
The authors: individualise; this is not a performance-first decision and belongs with medical input.
Limitations
83% of the included studies were of moderate, low or very low quality
Large between-study variance; OCP formulation is not analysed as a subgroup in the abstract
Training status of the pooled cohorts is not broken down in the abstract
How this entry was checked and changed
Content correction 2026-09-27
"SMD ≈ −0.06 to −0.17" -> not in the abstract, which reports Bayesian probabilities of a small effect (0.40 and 0.19 favouring naturally menstruating women; ~0 favouring OCP users) and between-study τ; replaced.
"n ~590 OCP users vs comparators" -> 590 participants in total across both groups.
"Most studies ≤3 months on OCP" -> not in the abstract; removed. Added the 83%/17% quality split and the within-OCP-cycle result.
Tier "4-5 Elite" -> Tier 1-3 (the entry's own limitations said elite were under-represented); Sport "cycling" -> general.