By Marc Dessein ·

PCOS and exercise, what training alone changes in your hormones

PCOS, now renamed PMOS, revolves largely around insulin. In many women with the condition, cells respond poorly to this hormone. The pancreas makes more of it to get the same effect, and insulin levels stay high. That excess pushes the ovaries to produce more androgens, the so-called male hormones. It also lowers SHBG, the liver protein that binds testosterone in the blood, so more of it circulates in its free form, the one that acts on skin, hair and the cycle.

Muscle sits right inside this loop. It is the body's biggest user of sugar, and when it works, it pulls sugar out of the blood with less need for insulin. The open question is what exercise changes on its own to PCOS hormones, with no diet or medication alongside.

An effect that is hard to isolate

In most PCOS studies, physical activity comes at the same time as a change in diet, weight loss or treatment. When blood results improve, there is no way to tell how much is down to exercise. The authors of this meta-analysis therefore kept only trials in which exercise alone was tested.

22 trials and 1,044 women

A team from Chinese university hospitals in Harbin and Hangzhou pooled the randomised trials published up to July 2025 in women with PCOS (Zou et al., 2026). In total, 22 trials and 1,044 women. The results appeared in September 2026 in the Journal of Obstetrics and Gynaecology Research.

Lower testosterone and lower insulin

Compared with control groups, exercise lowered total testosterone and free testosterone. It also lowered fasting insulin and HOMA-IR, an index calculated from fasting glucose and insulin that estimates insulin resistance. HDL, often called "good cholesterol", went up.

These results point the same way. Less insulin in circulation means a weaker signal telling the ovaries to make androgens. The study cannot say which drop drives the other, and the two move together.

The abstract states that these drops are statistically significant. It does not say by how much testosterone or insulin fell.

What duration and intensity change

In the subgroup analyses, the fall in total testosterone was larger when activity in the trials reached at least 150 minutes a week. Only moderate to low intensity exercise raised SHBG, the protein that binds testosterone.

These figures describe trials and do not make up a training plan. The abstract does not say which type of exercise works best, or how strength training compares with cardio. What it mainly shows is that moderate effort already counts, which includes simple activities such as walking.

Results that vary from woman to woman

The authors also split the results by region and by weight. In trials run in Asia, exercise lowered body mass index, total testosterone, LDL, fasting insulin and HOMA-IR. In South America the clearest effect was on total testosterone, in Europe on fasting insulin.

Women without obesity saw a larger fall in total testosterone than women with obesity. This also concerns slim women with PCOS, who are often diagnosed late because their weight looks normal. For them too, moving has an effect on androgens.

The authors conclude that body composition improves. The abstract, however, only details a drop in body mass index for the Asian trials. Weight alone therefore says little about what exercise does in PCOS.

The limits of this meta-analysis

With 1,044 women spread across 22 trials, each trial had fewer than 50 participants on average. The programmes varied widely in type of exercise, intensity, duration and the populations studied. The analyses by region, intensity or weight rest on smaller groups and remain exploratory.

The abstract gives no effect sizes. It reports no results on ovulation, cycle regularity, acne or hair growth. So there is no way to know whether the fall in testosterone is enough to ease those symptoms.

Exercise, one lever among others

This meta-analysis isolates exercise to measure its own effect. In everyday life, it works alongside food, sleep and stress, which all weigh on insulin. Food keeps its full place. Meals that limit blood sugar spikes, with protein at each one and at least 30 g at breakfast, work on the same loop as muscle.

Eating too little to make up for training pulls the other way. Long hours of cardio on too little food raise cortisol, which in turn raises blood sugar, and muscle gains nothing from it. Muscle builds slowly, and it responds to training at any age, as another meta-analysis on strength training at menopause shows.

Testosterone, SHBG and fasting insulin can all be measured with a blood test. Repeating them after a few months, with your doctor, shows what has shifted.

This article is for information only and does not replace medical advice.

Read next