Training around your cycle: what the evidence actually says

There is a popular framework in which you lift heavy in your follicular phase, do steady-state work in your luteal phase, and rest during your period. It is neat, it sells programmes, and the evidence behind it is considerably thinner than its confidence implies.

This is worth saying carefully, because the alternative claim — that the cycle has no effect at all — is also wrong. The honest position is more useful than either.

What the research actually shows

Systematic reviews and meta-analyses of menstrual cycle phase and exercise performance have generally reached the same conclusion: effects on performance are small and highly variable between individuals, and the underlying studies are mostly of low quality.

The problems with the literature are structural. Most studies have small samples. Many do not confirm ovulation, so participants assigned to a "luteal phase" group may not have been in one. Phase definitions differ between studies. Hormone levels are often not measured. Publication bias favours positive findings.

A frequently cited meta-analysis concluded that exercise performance might be trivially reduced during the early follicular phase compared with other phases, while explicitly noting that the evidence was low quality and that the effect was too small and too variable to justify prescribing training by phase.

That is a long way from the confident four-phase infographic.

What is better established

Some physiological differences across the cycle are real, even where their performance consequences are not clear.

Core temperature is higher in the luteal phase by a few tenths of a degree. This genuinely matters for thermoregulation in the heat — heat tolerance is reduced, and this is one of the more consistent findings. If you are training or competing in hot conditions, it is worth accounting for.

Perceived exertion is often higher in the luteal phase at the same objective workload. This is a real and consistently reported subjective effect, and it is not the same as reduced capacity.

Sleep is worse in the luteal phase for many people, and sleep affects training more than phase does. Sleep in the luteal phase.

Symptoms matter. Heavy bleeding, cramps and iron deficiency have obvious effects on training capacity, and these are usually much larger than any direct hormonal effect on muscle. How heavy is too heavy and iron deficiency in women.

Injury risk around ovulation has been examined, particularly for ACL injuries, with some evidence of increased laxity around the oestrogen peak. The findings are mixed and the practical implications remain unclear.

What is not established

That there is an optimal phase for strength gains that applies to everyone. That you should reduce training during your period as a rule. That cycle-based periodisation outperforms consistent, well-designed training. That the same schedule works for two different people.

The last point is the most important. The between-person variation in cycle experience is enormous — some people are barely affected, some are substantially affected, and the direction is not even consistent. Prescribing a population-level template to individuals whose responses vary that much is close to guaranteed to be wrong for most of them.

What actually matters more

If you rank the variables by effect size on training quality, cycle phase is well down the list:

1. Consistency. Training that happens beats training that is optimal.
2. Progressive overload. The load has to increase over time.
3. Sleep.
4. Adequate energy and protein. Underfuelling blunts adaptation and, taken far enough, stops your cycle altogether.
5. Iron status, particularly with heavy periods.
6. Then, possibly, cycle phase.

The risk of the cycle-training framework is that it puts a small variable at the top and produces a permission structure for skipping sessions — "I'm in my luteal phase" — that costs more than any phase effect ever gains.

The version that is defensible

Rather than periodising by phase, respond to symptoms. This sounds similar and is meaningfully different.

If you are exhausted, in pain or sleeping badly, adjust that session. If you feel good, train hard, whatever day it is. That approach responds to your actual state rather than to a predicted one, and it does not require the cycle to be behaving as expected — which it frequently is not.

For most people, the useful discovery is not "train hard in the follicular phase". It is something more specific: that days 25 to 27 are consistently poor sessions, or that heavy-bleeding days are worse than premenstrual days, or that nothing tracks with the cycle at all and it is all sleep.

Which means you have to look

None of that is knowable from a template, and it is not reliably knowable from memory either — people substantially over-attribute bad sessions to their cycle after the fact.

Logging session quality against cycle day, sleep, energy and symptoms for a few months answers it properly. For some people a clear pattern emerges and is worth planning around. For plenty of others, the pattern turns out to be sleep and nothing else, which is a genuinely useful finding.

Naked is built for that comparison — training, sleep, cycle and mood in one record, with the relationships surfaced rather than assumed. The point is not to follow the average woman's cycle. It is to find out whether you are one of the people this actually applies to.