"Exercise more" is the advice almost every woman with PCOS/PMOS hears from her GP, and it is — technically — not wrong. But the type of exercise matters enormously, and most of the guidance skips that part entirely. If you've been grinding through HIIT classes and wondering why HIIT might be working against you, the mechanism is cortisol: sustained high-intensity cardio spikes cortisol levels and, in women with PCOS/PMOS, can actively worsen the insulin resistance that underlies so much of the symptom burden. Roughly 70–80% of women with PCOS/PMOS have some degree of insulin resistance — even lean women, even women whose GPs have told them their bloods are normal.
Resistance training works through completely different mechanisms. It doesn't rely on sustained cardiovascular effort, it doesn't produce the prolonged cortisol response that aerobic exercise does, and it targets insulin resistance more directly than any other form of exercise. Understanding why — and knowing what to track — is the difference between exercising generally and using movement as a targeted metabolic tool.
Why Resistance Training Works for PCOS/PMOS: Four Mechanisms
1. GLUT4 Upregulation
This is the primary mechanism, and it's worth understanding in detail. Glucose enters your cells via GLUT4 transporter proteins. In most circumstances, those transporters need insulin to function — insulin acts as the key that unlocks the cell door. In women with PCOS/PMOS and insulin resistance, that key doesn't work as efficiently as it should: the cells resist insulin's signal, glucose stays in the bloodstream, and the pancreas produces yet more insulin in response. The consequence is the insulin-androgen loop — high insulin drives LH-stimulated androgen overproduction in the ovaries, which drives the symptoms most people associate with PCOS/PMOS.
Here's what's remarkable about resistance training: muscle contraction directly stimulates GLUT4 transporter expression independently of insulin. When you perform a squat or a deadlift, your muscle fibres literally pull GLUT4 transporters to the cell surface through a separate signalling pathway — no insulin required. This means glucose can enter muscle cells during and after exercise without an insulin spike. Over time, with consistent training, GLUT4 expression increases persistently even at rest. The result is improved insulin sensitivity that's measurable, cumulative, and doesn't require weight loss to occur.
2. Androgen Metabolism
Increased muscle mass improves androgen metabolism through two routes. First, muscle tissue increases the expression of sex hormone-binding globulin (SHBG) — the protein that binds free testosterone and renders it metabolically inactive. Women with PCOS/PMOS typically have low SHBG, which is why free testosterone (rather than total testosterone) is a more meaningful marker. More muscle means more SHBG, which means less free testosterone circulating. Less free testosterone means reduced hirsutism, reduced androgen-driven acne, and — over time — reduced androgenic hair loss on the scalp.
Second, adipose tissue, particularly visceral fat, is itself a site of androgen production via the aromatase enzyme pathway. Reducing visceral fat (which resistance training does disproportionately — see mechanism 4) reduces this peripheral androgen source.
3. Cortisol Buffering
Cortisol and stress are central to the PCOS/PMOS picture. The cortisol-insulin interaction is a bidirectional loop: high cortisol worsens insulin resistance, and insulin dysregulation makes the HPA axis more reactive. Sustained aerobic exercise — particularly HIIT — produces a prolonged cortisol elevation that can last two to four hours post-session. For women with an already hyperreactive HPA axis, this is metabolically counterproductive.
Resistance training produces a different pattern: a sharp, short cortisol spike during the session itself, which resolves within 60–90 minutes and is followed by a prolonged cortisol reduction phase lasting several hours. This is not a small distinction. It means that properly programmed resistance training actively lowers the resting cortisol set point over time — the opposite of what chronic cardio does. The key word is "properly programmed": sessions that are too long or too frequent can begin to mimic the sustained cardio cortisol pattern. Three sessions per week, 45–60 minutes each, is the sweet spot.
4. Adipokine Profile
Visceral fat — the fat stored around the organs in the abdominal cavity — is metabolically active. It secretes inflammatory signalling molecules called adipokines, including TNF-α and IL-6, which directly drive insulin resistance and systemic inflammation. Subcutaneous fat (the fat under the skin) secretes fewer of these molecules and is metabolically less aggressive.
Resistance training reduces visceral fat disproportionately compared to subcutaneous fat. The mechanisms include GLUT4-mediated glucose clearance (less glucose is available to be converted to fat), improved insulin signalling (less fat storage stimulus), and direct post-exercise lipolysis in visceral depots. The result is an improved adipokine profile — lower TNF-α and IL-6 — which reduces the inflammatory drive on insulin resistance and androgen production. For more on how to track inflammation signals day-to-day, the PCOS/PMOS inflammation tracking guide covers the check-in markers most sensitive to this pathway.
What the Research Says
The evidence base for resistance training in PCOS/PMOS is substantially stronger than the generic "exercise more" advice implies. A 2019 meta-analysis published in the Journal of Clinical Endocrinology & Metabolism found that resistance training improved HOMA-IR — the standard clinical marker for insulin resistance — by 25–30% over 12 weeks in women with PCOS. That's a clinically meaningful improvement, comparable to what some studies show with low-dose metformin.
Multiple randomised controlled trials have shown improvements in testosterone levels and LH/FSH ratios, particularly with compound movement patterns — squat, deadlift, and hip hinge exercises — rather than isolated machine work. The hypothesis is that compound movements recruit more total muscle mass, producing a greater systemic GLUT4 and SHBG response.
Three sessions per week appears to be the sweet spot for hormonal outcomes specifically. Studies comparing two, three, and four sessions per week have found that four sessions produces no additional hormonal benefit over three, and may marginally worsen the cortisol picture in women with a hyperreactive HPA axis.
Critically: the improvements are independent of weight loss. In trials where researchers controlled for body weight, women who did not lose weight still showed significant improvements in insulin resistance markers, testosterone levels, and cycle regularity. This is the key differentiator from calorie-restriction-focused approaches, and it matters enormously for women who have been told they need to lose weight first before their hormones can improve. The evidence says the hormonal improvement comes first.
The Five Atlas Signals to Watch
Each mechanism maps to a trackable Atlas check-in signal. Here's where to look and when to expect movement:
1. Morning energy (energy check-in)
GLUT4 upregulation improves overnight glucose stability — the mechanism behind the 3am cortisol spike that wakes some women with PCOS/PMOS. As GLUT4 expression improves with consistent resistance training, glucose is cleared into muscle cells more efficiently, blood sugar doesn't drop as far overnight, and the early-morning energy score begins to improve. This is typically the first signal that changes — most women notice it between weeks 3 and 5. Watch for it specifically on the morning after a training day.
2. Cravings score
As insulin sensitivity improves, reactive hypoglycaemia episodes reduce in frequency and intensity. The mid-afternoon cravings crash — that specific 3–4pm need for something sweet — is driven by blood sugar volatility that improved GLUT4 function addresses directly. Cravings tend to drop before energy, mood, or skin changes, making it a leading indicator that the mechanism is working. Log it every day and look for the trend across weeks 4–6.
3. Mood score
Resistance training produces a measurable same-day mood response via serotonin and endorphin pathways. In Atlas, this shows up as a clear pattern differential between training days and rest days: mood scores on training days should run higher than rest-day scores. Look for this pattern beginning in week 2, and watch whether the gap widens over time as the cumulative hormonal effects build. If training-day mood isn't running higher than rest-day mood, the session may be too long or too intense — pushing the cortisol pattern in the wrong direction.
4. Skin flag (daily check-in)
Androgen-related changes are the slowest-moving signal — 8–12 weeks — but they're measurable if you track consistently. The androgen pathway (SHBG upregulation, reduced free testosterone) takes time to express itself in skin changes. Log the skin flag daily even when it feels premature. At week 12, comparing your skin flag frequency in weeks 1–2 (baseline) against weeks 10–12 gives you a meaningful data point. Don't wait until you feel like you should see a difference — by then, you'll have lost the baseline.
5. Sleep quality
Resistance training improves slow-wave sleep (stage 3 NREM) more effectively than aerobic exercise. Slow-wave sleep is when cortisol is at its lowest and growth hormone is at its highest — both of which matter directly for PCOS/PMOS recovery. The mechanism is well-established: the adenosine drive generated by muscle activity during resistance training produces deeper sleep onset. The sleep and cortisol connection in PCOS/PMOS explains how this plays out in practice and which morning energy patterns are most tightly linked to the slow-wave mechanism. Watch the sleep quality signal against the morning energy score — as slow-wave sleep improves, the next-morning energy score should begin to correlate more tightly with training nights.
Three 12-Week Tracking Experiments
Experiment 1: Baseline-first (Weeks 1–2)
Before changing anything, log 14 days of Atlas check-ins. Don't add resistance training yet. The goal is to establish your cravings, energy, and mood baseline — your personal control period. This sounds obvious, but almost nobody does it, and without it, any improvement you see after starting training is anecdotal rather than measurable. What is your average morning energy score? What does your cravings pattern look like across the week? How does it vary around your cycle phase? Know those numbers before you start.
Experiment 2: 3x/Week Compound Protocol (Weeks 3–14)
Three sessions per week, compound movements only: a squat pattern, a hip hinge, an upper-body push, and an upper-body pull. That's it — no HIIT, no sustained cardio sessions over 30 minutes, no additional gym classes during the 12-week experiment period. Three to four sets of 8–12 repetitions per movement, with 60–90 seconds' rest between sets. Sessions should take 45–50 minutes.
In Atlas, track energy, cravings, and mood every day. The critical comparison is training-day morning signals (the day after a session) versus rest-day morning signals. Look for the first training-day morning energy signal between weeks 3 and 5. If you're not seeing a differential by week 6, the protocol may need adjusting.
Experiment 3: Recovery Window Experiment (Ongoing)
From week 4 onwards, add a deliberate note to your sleep-quality log on training nights versus rest nights. The goal is to see slow-wave sleep improvement show up in the morning energy score: training night → better sleep quality → higher morning energy the following day. If the pattern isn't emerging by week 8, the session volume may be too high. Overtraining — sessions that are too long, too frequent, or too intense — suppresses the slow-wave benefit by sustaining elevated cortisol into the evening. If you're not seeing sleep improvement on training nights, cut the session length by 15 minutes and reassess at week 10.
GP Framing and a Practical Starting Point
Most GPs won't prescribe resistance training specifically. They'll say "exercise more" and leave the type, frequency, and intensity entirely unspecified. What to bring to your appointment: your Atlas training-day versus rest-day energy and craving differentials, printed or screenshotted. A chart showing that your morning energy scores are consistently 1.5 points higher the morning after resistance training sessions than on rest days — across 8 weeks of data — is the kind of n=1 evidence that turns a lifestyle conversation into something evidence-based. GPs are trained to respond to data. Anecdote is easily dismissed. Your own pattern data is not.
Practically: you do not need a gym. Bodyweight squats, hip hinges (a Romanian deadlift performed with a water bottle or backpack as load), push-ups, and rows using a sturdy table edge or resistance band hit all four mechanisms. The load matters far less than the movement pattern and the consistency. You're stimulating GLUT4 translocation and SHBG production with movement quality and neuromuscular effort — not with the weight on a barbell. Begin with bodyweight, add load progressively as the movements feel stable, and prioritise showing up three times per week over maximising any single session.
Track What's Actually Changing
The research is clear: resistance training is one of the most powerful interventions available for PCOS/PMOS, and the benefits are independent of weight loss. But "resistance training helps PCOS/PMOS" is a population-level finding. The only way to know if it's working for your specific body, on your specific hormonal profile, in your specific circumstances, is to track the signals.
Atlas captures exactly the check-in data that maps to these mechanisms — energy, cravings, mood, skin, sleep quality — so the pattern engine can surface the training-day differentials, the week-3-to-5 energy shift, and the slow-wave sleep signal that tells you whether the protocol is working. If you haven't yet read about why HIIT might be working against you, that's the natural companion piece — the mechanisms are the mirror image of everything covered here.