The menstrual cycle barely moves exercise performance, pooled data suggest

Sex hormone concentrations fluctuate across the menstrual cycle, and a growing industry of cycle-based training advice assumes those fluctuations meaningfully change what women can do in the gym or on the track. Yet the underlying studies have long conflicted, with no consensus on whether performance actually varies by phase. This systematic review and meta-analysis set out to settle whether exercise performance differs across menstrual cycle phases in eumenorrheic women, meaning women with regular natural cycles, and whether any evidence-based recommendations can be made.

The reviewers searched four databases for experimental studies measuring at least one performance outcome in two or more defined menstrual cycle phases, and found 78 eligible studies. Data from 51 studies fed an initial analysis comparing the early follicular phase against all other phases combined, covering both endurance and strength outcomes. Seventy-three studies had enough data for a network meta-analysis comparing all phases against each other, with phases ranked by the likelihood that performance was poorest. Study quality was formally assessed, and the models were built on Bayesian multilevel methods.

Performance was trivially reduced in the early follicular phase compared with all other phases, an effect so small its uncertainty range included zero. The largest single comparison, early follicular versus late follicular, was still small. The ranking analysis gave the early follicular phase a 30% likelihood of being among the poorest phases for performance, with every other phase clustered between 53 and 55%, suggesting little separates them. Between-study variation was large, many included studies were of poor quality, and the overall evidence was graded as low, at 42%.

This is the most comprehensive pooling of menstrual cycle and performance data to date, and it complicates the popular narrative rather than confirming it: whatever effect cycle phase has on performance appears trivial at the group level. That does not rule out meaningful effects in some individuals, which is precisely why the authors decline to issue general guidelines and instead recommend a personalised approach. The field's biggest open problem is study quality, with better-controlled research needed before firmer conclusions are possible.

If you have encountered confident advice to restructure training around your cycle, this is a reason for scepticism: pooled across all available studies, phase effects on performance were trivial, and the evidence behind them was rated low quality. It is not a reason to dismiss your own experience, since responses vary widely between individuals and the reviewers themselves suggest a personalised approach based on how you actually perform across your cycle. The group average and your own pattern are two different questions.