High insulin makes the ovaries produce more testosterone, and more testosterone makes insulin work less well. In the guides we publish here, that two-way loop is the part readers most often have not been told, because a PCOS appointment usually covers cycles and skin and stops there.

Insulin resistance is not one of the diagnostic criteria for polycystic ovary syndrome. It is present in a large share of people who have it, including people whose weight gives no hint of it, and it is the mechanism behind several symptoms that look purely hormonal.

The 2023 International Evidence-Based Guideline for the Assessment and Management of Polycystic Ovary Syndrome is the reference document for this, and it is freely readable.

The Verdict

If you have a PCOS diagnosis, get your glycaemic status assessed regardless of your weight, and repeat it every one to three years. The international guideline recommends exactly that, and it is the step most often dropped after the initial workup. Adding fasting insulin to the same draw is inexpensive and shows the compensation that glucose alone hides.

Why Insulin Resistance Shows Up in PCOS

Insulin resistance appears in PCOS at rates well above the background population, and the excess holds after accounting for body weight. That last point deserves attention. If the association were purely a consequence of higher average body mass, it would disappear once you compared like with like. It does not.

Part of the explanation lies in how insulin signalling fails in PCOS. Muscle and fat tissue respond weakly to insulin's effect on glucose uptake, while the ovaries and adrenal glands stay fully responsive to insulin's effect on steroid production. The pancreas compensates for the first problem by secreting more insulin, and that extra insulin lands on tissues that never stopped listening.

The result is selective. Glucose handling worsens while androgen production rises, driven by the same hormone.

The Loop Between Insulin and Androgens

Three linked effects run this loop, and each one is measurable on a blood panel.

  1. Insulin acts directly on ovarian theca cells. Those cells produce androgens, and insulin increases their output alongside luteinising hormone. More circulating insulin means more testosterone leaving the ovary.
  2. Insulin suppresses SHBG production in the liver. Sex hormone binding globulin is the protein that holds testosterone inactive in circulation. When it falls, the fraction of testosterone that is free rises even if total testosterone has not changed at all.
  3. Higher androgens push fat toward the abdomen. Visceral fat is metabolically active tissue that raises insulin demand further, which closes the loop.

Step two is the one that explains a confusing lab report. A woman with worsening acne and hair changes can be told her testosterone is normal, because the laboratory measured total testosterone and the number sat inside the range. Her SHBG had fallen, and the free fraction driving her symptoms had climbed. Asking for calculated free testosterone alongside SHBG makes that visible.

Lean PCOS Is Where the Testing Fails

Metabolic screening in most clinics is triggered by body weight, and that trigger misses the lean PCOS group entirely. Studies comparing lean women with PCOS against body-mass-index-matched controls consistently find reduced insulin sensitivity in the PCOS group.

The practical failure follows from the prompt rather than the biology. A 24-year-old with a BMI of 21, irregular cycles and mild acne is unlikely to be offered a glucose tolerance test, because nothing about her presentation matches the mental picture attached to metabolic risk. Her diagnosis is made on cycles and ultrasound, her treatment is a contraceptive pill, and her glucose is never looked at.

The guideline recommendation does not carry a weight qualifier. Assessment of glycaemic status is recommended for everyone with a PCOS diagnosis, and the repeat interval runs on risk factors rather than on the scale.

Which Markers to Run, and When

Six markers cover this, and they split into the glucose side and the androgen side. Ordering them from one fasted draw is the efficient version.

Blood markers used to assess insulin resistance and androgen status in PCOS.
MarkerWhat it measuresWhy it helps in PCOSIts limitation
Fasting insulin How hard the pancreas is working before food Rises earlier than glucose and is the marker that explains a normal HbA1c in someone with clear PCOS symptoms Left off standard panels. Ask for it alongside glucose in the same fasted draw
HOMA-IR Fasting insulin and fasting glucose combined into one index Easier to track across draws than two separate numbers Assay variation between laboratories makes comparisons across providers unreliable
Oral glucose tolerance test Glucose across two hours after a fixed 75 g drink The international PCOS guideline names it the most sensitive test for glycaemic status in PCOS Two hours in a clinic, and it needs a physician order
SHBG The protein that binds testosterone and keeps it inactive High insulin suppresses it, so a low SHBG is often the first laboratory hint of hyperinsulinaemia Also lowered by thyroid dysfunction and some medications
Free and total testosterone Circulating androgen, bound and unbound Free testosterone drives acne and hair changes, and it climbs when SHBG falls even if total testosterone holds still Assays for free testosterone are unreliable at low concentrations; calculated free testosterone is usually preferred
Triglyceride to HDL ratio Two numbers already on a lipid panel, divided Free if you have had a lipid panel in the past year, and it tracks insulin resistance better than either value alone A proxy rather than a measurement, and thresholds differ by ancestry

One point on timing. The androgen markers are best drawn in the early follicular phase where cycles are regular enough to identify one, and any hormonal contraceptive suppresses them to the point where the result describes the medication rather than the ovary. The glucose markers have no such constraint. Our page on the perimenopause biomarker panel covers the cycle-day timing rules in more detail, and most of them transfer directly.

What Lowers Insulin in PCOS

Four interventions have real evidence behind them, and they differ in how much they ask of you.

  • Resistance training. Skeletal muscle disposes of most post-meal glucose, and training increases that capacity independently of weight change. This is the intervention that works for the lean subgroup, where weight loss advice does not apply.
  • Reducing visceral fat. Trials in PCOS show improvements in cycle regularity and androgen levels at losses of roughly 5 to 10 percent of body weight. Waist circumference tracks the relevant fat better than the scale does.
  • Metformin. It improves insulin sensitivity and helps modestly with cycle regularity. The guideline positions it after lifestyle change, and for metabolic features specifically rather than for every presentation.
  • Myo-inositol. Small trials support an effect on insulin sensitivity and ovulation. The evidence base is limited rather than settled, and it is a reasonable low-risk addition rather than a replacement for the two items above.

Sleep belongs in this list more than most PCOS material admits. Short and fragmented sleep reduces insulin sensitivity measurably within a few nights, and it is common in PCOS. Our guide to how much sleep you need covers what the trials used as their intervention.

Who This Advice Does Not Fit

Two groups should take a different route.

  • Anyone currently trying to conceive. Ovulation induction has its own protocols and its own medication rules, and decisions about metformin, letrozole and timing belong with a fertility specialist rather than with a metabolic panel.
  • Anyone pregnant or recently pregnant. Insulin sensitivity falls during pregnancy as normal physiology, and gestational screening runs on its own thresholds through maternity care.

Our reading here would change if insulin assays were standardised across laboratories. The current guideline advice against using insulin diagnostically rests on that variation rather than on the marker being uninformative. A standardised assay would make fasting insulin a defensible diagnostic test in PCOS, and it would move the screening conversation years earlier than a glucose tolerance test can.

If your PCOS diagnosis is more than a year old and nobody has looked at your glucose since, ask for glycaemic status testing with fasting insulin added to the same draw.

Frequently Asked Questions

Does PCOS cause insulin resistance, or does insulin resistance cause PCOS?

Both directions operate, which is why the relationship is described as a loop rather than a cause. High insulin acts on ovarian theca cells and increases androgen production, and it suppresses SHBG in the liver so more testosterone circulates unbound. Higher androgens in turn worsen insulin sensitivity, partly through changes in fat distribution. The loop can be entered from either side, and treatment works by weakening one link rather than by identifying which came first.

Can you have PCOS without insulin resistance?

Yes. Insulin resistance is common in PCOS but it is not part of the diagnostic criteria, and a meaningful minority of women with PCOS have entirely normal insulin. The Rotterdam criteria are built on irregular ovulation, clinical or biochemical androgen excess, and polycystic ovarian morphology on ultrasound. Insulin is not one of the three. A normal fasting insulin does not rule out PCOS, and an elevated one does not diagnose it.

Do lean women with PCOS get insulin resistance?

Yes, and it is regularly missed because the screening prompt in most clinics is body weight. Studies of lean PCOS consistently find reduced insulin sensitivity compared with body-mass-index-matched controls without PCOS. The practical failure is that a woman with a BMI of 22 and irregular cycles is less likely to be offered a glucose test at all. If you have a PCOS diagnosis, the guideline recommendation to assess glycaemic status applies regardless of your weight.

What is the best test for insulin resistance in PCOS?

The international PCOS guideline recommends assessing glycaemic status with an oral glucose tolerance test, fasting glucose or HbA1c, and it recommends against measuring insulin for diagnostic purposes because no assay is standardised well enough. In practice many clinicians add fasting insulin anyway, because it moves earlier than glucose and gives a sense of how much compensation is happening. Read it as a trend across repeat draws at the same laboratory rather than as a single verdict.

Does metformin help PCOS?

Metformin improves insulin sensitivity and modestly helps with menstrual regularity and metabolic markers in PCOS. It is not a first-line treatment for every presentation. The international guideline positions lifestyle change first, with metformin added for metabolic features or for cycle regulation where the combined oral contraceptive pill is unsuitable or declined. Gastrointestinal side effects are the usual reason people stop, and extended-release formulations are better tolerated.

Does inositol work for insulin resistance in PCOS?

Myo-inositol has trial evidence for improving insulin sensitivity and ovulation in PCOS, but the trials are mostly small and the guideline treats the evidence as limited rather than settled. It is generally well tolerated at the doses studied. Treating it as a low-risk addition to lifestyle change is defensible. Treating it as a replacement for the interventions with larger effect sizes, particularly resistance training and reducing visceral fat, is not.

Will losing weight fix insulin resistance in PCOS?

Weight loss improves insulin sensitivity in PCOS, and modest amounts help. Trials in PCOS show improvements in cycle regularity and androgen levels with losses around 5 to 10 percent of body weight. It does not fix it for everyone, and it is not available as advice to the lean subgroup at all. Resistance training improves glucose disposal independently of weight change, which is why it belongs in the plan even when the scale does not move.

Does insulin resistance in PCOS raise diabetes risk?

Yes. PCOS is associated with a higher lifetime risk of type 2 diabetes and gestational diabetes, which is the reason the international guideline recommends assessing glycaemic status at diagnosis and repeating it every one to three years depending on other risk factors. That repeat schedule is the part most often dropped after the initial workup. If your PCOS diagnosis is several years old and nobody has re-tested glucose, that is worth raising at your next appointment.

Related