Key takeaways
- PCOS in most women is a loop: insulin drives ovarian androgen production and suppresses SHBG, androgens stall ovulation, and central fat gain worsens the insulin.
- Birth control suppresses ovulation and metformin helps modestly, but neither improves the insulin resistance underneath.
- Measure fasting insulin first — a lean, insulin-sensitive phenotype exists and needs different emphasis.
- Labs move before cycles and cycles before appearance; returning ovulation is a fertility event to plan for.
Most women with PCOS are handed the same two things: a packet of birth control to make the bleeding regular, and metformin for the insulin. Both do something. Neither touches the loop that produces the condition, which is why so many women take them for a decade and remain as symptomatic as they were at diagnosis. The modern approach starts from a different premise — that in most cases this is a metabolic disorder expressing itself in the ovaries, and the ovaries are downstream of the thing worth treating.
The loop, stated mechanically
The name is misleading. Most affected women do not have cysts in any meaningful sense; what an ultrasound shows is a collection of small follicles that started to develop and stalled. That is a consequence, not a cause.
The sequence runs like this. Insulin resistance in muscle and liver forces the pancreas to secrete more insulin to do the same job. That excess insulin acts on the theca cells of the ovary, where it directly amplifies androgen production (Nestler et al., J Clin Endocrinol Metab), and simultaneously suppresses the liver's output of sex hormone binding globulin — so more androgen is produced and a greater fraction circulates unbound and active. Elevated androgens disrupt the orderly selection of a dominant follicle, so ovulation becomes irregular or stops. Meanwhile insulin is a storage hormone, weight accumulates around the middle, and visceral fat worsens insulin resistance further.
Every arrow feeds the next. Which is why interventions aimed at a single arrow — suppressing androgens, or forcing a bleed — rarely change the trajectory.
Where that model is honest, and where it is not
Insulin resistance is the dominant driver in most cases and by far the most treatable part of the loop. But it is not the whole story. Diagnosis rests on two of three features — irregular or absent ovulation, clinical or biochemical androgen excess, and polycystic ovarian morphology — with other causes excluded, and that definition captures more than one phenotype. A subset of women are lean, insulin-sensitive on testing, and still meet criteria; in those, adrenal androgen contribution or a primarily neuroendocrine pattern may explain more. There is also a genuine genetic component, which is why it clusters in families and why nobody caused it by eating badly.
The practical consequence: measure before assuming. Fasting insulin and HOMA-IR tell you which version you are treating.
Why the standard treatment falls short
Combined oral contraceptives do exactly what they are designed to do: suppress ovulation entirely and impose a withdrawal bleed. Cycles become predictable because there is no cycle. Androgen-driven skin and hair symptoms often improve, partly because oral estrogen raises SHBG and mops up free androgen. That is genuine relief, and for a woman not currently pursuing fertility it remains a reasonable option.
What it does not do is improve insulin sensitivity, and the underlying trajectory continues underneath it. Metformin improves insulin sensitivity modestly. The result for many women is a decade of managed symptoms with an unchanged metabolic position, and the diabetes risk that comes with it.
Where a GLP-1 medication acts on the loop
Incretin-based therapy enters at the origin rather than the output. It raises satiety signalling and slows gastric emptying, reducing energy intake; it prompts glucose-dependent insulin release; and the weight loss that follows reduces visceral fat, the tissue driving insulin resistance in the first place. Lower insulin means less theca-cell stimulation, less androgen, recovering SHBG.
Research on GLP-1 therapy in PCOS has reported restored ovulation, lower androgens and improved metabolic markers over months (Jensterle et al., Endocrine Connections 2020). The important nuance is that returning ovulation is a fertility consideration, not only a symptom improvement — women who assumed they could not conceive need to know that before it becomes a surprise. The PCOS metabolic phenotype and GLP-1 therapy in PCOS go further. Whether any medication is appropriate is a clinical decision made after evaluation.
Inositol, and what it is realistically worth
Inositols are second messengers in the insulin signalling pathway, and there is evidence that women with PCOS handle them abnormally — specifically a defect in converting myo-inositol to D-chiro-inositol in ovarian tissue. Supplementing both, in the roughly 40:1 ratio used in the trials, is intended to restore that balance. A meta-analysis of nine randomised trials reported ovulation restored in 47-62% of women within six months (Unfer et al., Endocr Connect 2017). Those trials are small and heterogeneous, so the honest description is a well-tolerated, inexpensive adjunct with reasonable supporting evidence — not a primary therapy, and worth a fair trial period before judging it.
Training and protein: the part that works regardless of phenotype
Resistance training improves insulin sensitivity by a mechanism that does not require weight loss at all. Contracting muscle takes up glucose through a pathway that runs independently of insulin (Richter & Hargreaves, Physiol Rev 2013), and training increases both the muscle available to do that and the machinery inside it. Reported improvements sit in the region of 30-40% independent of weight change — which is why this is the one intervention that helps the lean and the metabolic phenotypes equally.
Protein does two jobs: it protects lean tissue during weight loss, and it is the most satiating macronutrient, which makes intake easier to hold. Around 1.0 g per pound of goal weight is a workable rule of thumb, and a structured programme — the three-day women's foundations plan is a reasonable entry point — removes the decision fatigue that ends most attempts.
Selective hormonal support
Some women benefit from cyclical or continuous bioidentical progesterone, particularly where anovulation leaves the endometrium exposed to unopposed estrogen and where sleep is disturbed (progesterone for sleep and mood). Spironolactone is added in some cases for significant acne or hirsutism, on the understanding that it treats the symptom while something else treats the driver. Combined contraceptives remain useful for women not currently pursuing fertility. None of these are default first-line any more, and all are physician decisions after evaluation.
What to expect, and in what order
| Time | What typically changes |
|---|---|
| Month 1 | Appetite becomes manageable; early weight change; nothing yet visible in cycles |
| Month 3 | Fasting insulin falling; energy steadier; cycles beginning to regularise in some |
| Month 6 | Ovulation returning in many; androgens falling; skin and hair lagging behind the labs |
| Month 12 | Cycles regular in most; metabolic markers substantially improved; hirsutism improving slowly |
Two points about that table. Cycles respond before appearance does — hair growth cycles are slow, so hirsutism improves months after the androgens driving it have fallen, and women judging progress by the mirror conclude it is not working while their labs say otherwise. And the sequence is metabolic first, ovarian second: fasting insulin moves before androgens, and androgens before ovulation.
The shift: modern PCOS care treats this as a metabolic disease that manifests in the ovaries, not an ovarian disease that happens to affect metabolism. That single reframe changes what you measure, what you treat first, and what counts as progress.
Bottom line
PCOS in most women is a self-reinforcing loop: insulin resistance drives hyperinsulinaemia, which drives ovarian androgen production and suppresses SHBG, which disrupts ovulation and promotes central weight gain, which worsens insulin resistance. Treatment that enters at the origin — insulin sensitivity through weight loss, resistance training, adequate protein and, where clinically appropriate, GLP-1 therapy, with inositol as a reasonable adjunct — changes the trajectory in a way suppressing ovulation does not. Measure first, because the lean and insulin-resistant phenotypes need different emphasis. Expect labs to move before cycles and cycles before appearance. And be aware that returning ovulation is a fertility event as well as a sign of progress. The 60-second assessment routes the lab work and the treatment decision to a physician.
Educational content, not medical advice. Laboratory interpretation and any treatment decision are made by a licensed physician after individual evaluation. Individual results vary.
