Insulin Resistance in PMOS (PCOS): How to Test It, Read It, and Treat It

How to test for it — and what not to use

The recommended screen: 75 g OGTT

The 75 g oral glucose tolerance test is the recommended screening test for glycaemic status in PMOS. Fasting glucose alone and HbA1c alone both under-detect impaired glucose tolerance in this population — a patient can have a normal fasting glucose and a normal HbA1c while having clearly abnormal post-load glucose handling.

Interpretation (WHO/ADA):

Frequency: at diagnosis, then every 1–3 years depending on risk factors (BMI, family history of type 2 diabetes, prior gestational diabetes, age).

HOMA-IR: useful for understanding, not for diagnosis

HOMA-IR = (fasting insulin µU/mL × fasting glucose mg/dL) ÷ 405

This is widely ordered in Indian practice and widely over-interpreted. Its genuine limitations:

• No internationally agreed cut-off. Values from 2.0 to 3.8 are variously quoted. Population-specific thresholds for Indian women are not well standardised.
• Insulin assays are not standardised between laboratories. The same sample can yield materially different results on different platforms, making values non-comparable across labs and across time.
• It is not a guideline-endorsed diagnostic criterion for PMOS or for insulin resistance in clinical care. It is primarily a research tool.
• It does not change management. Treatment decisions rest on glycaemic status, BMI, symptoms and reproductive goals — not on a HOMA-IR number.

Used sensibly, HOMA-IR can help explain the condition to a patient and can support a decision to start metformin in a borderline case. It should not be the basis of the diagnosis, and a normal HOMA-IR should never be used to reassure a patient that insulin resistance is absent.

Fasting insulin alone has the same assay problems and should not be used in isolation.

What else to measure

• Fasting lipid profile — at diagnosis, then per cardiovascular risk
• Blood pressure — every visit
• Waist circumference — ≥ 80 cm is the threshold for South Asian women; a better marker of visceral adiposity than BMI
• BMI using Asian cut-offs — overweight ≥ 23, obese ≥ 25 kg/m²
• Liver function / hepatic ultrasound where indicated — metabolic dysfunction-associated steatotic liver disease (MASLD, formerly NAFLD) is substantially more prevalent in PMOS and is frequently silent

Clinical signs of insulin resistance

• Acanthosis nigricans — velvety hyperpigmentation of the neck, axillae, groin or knuckles. This is a highly specific bedside marker and is frequently missed because it is mistaken for poor hygiene or dirt, including by patients themselves.
• Skin tags (acrochordons), often in the same distribution
• Central adiposity with a raised waist circumference despite an apparently acceptable BMI

---

Treatment: what the evidence actually supports

1. Lifestyle intervention — first-line, always

This remains the foundation of management and is recommended in every guideline. Key points that improve adherence:

• Modest weight loss works. A 5–10% reduction in body weight in those with excess weight produces meaningful improvement in ovulatory function, androgen levels and metabolic parameters. Patients routinely believe they must lose 20 kg to see benefit; they do not.
• Exercise benefits are partly weight-independent. Resistance training and moderate aerobic activity improve insulin sensitivity even without weight change — important messaging for lean patients and for those whose weight plateaus.
• No single diet is superior. The evidence does not support any specific macronutrient composition over another. Sustainability and total energy balance dominate. Low-glycaemic-index approaches are reasonable and often practical in an Indian dietary context, where refined carbohydrate load is typically high.
• Frame carefully. Weight-centric messaging carries real risk in a population with elevated rates of disordered eating and depression. Anchor goals to cycle regularity, energy, skin and long-term risk.

2. Metformin

The established pharmacological option for metabolic and anthropometric outcomes, recommended in the international guideline.

• Typical dosing: start 500 mg once daily with food, titrate slowly over weeks to 1500–2000 mg daily in divided doses. Slow titration is the single most effective way to improve tolerance.
• Extended-release formulations substantially reduce gastrointestinal side effects and improve adherence — worth using routinely rather than as a rescue.
• Benefits: improved insulin sensitivity, modest weight reduction, improved menstrual regularity, reduced progression to type 2 diabetes.
• Limitations: response is inconsistent between individuals, and gastrointestinal side effects (nausea, diarrhoea, abdominal pain) frequently cause discontinuation.
• Monitor vitamin B12 on long-term use — metformin-associated B12 deficiency is well documented and relevant given the high prevalence of baseline B12 deficiency in Indian vegetarian populations. This is under-monitored in routine practice.

3. Inositol

• Inositol may be considered based on individual patient preference, given limited harm and potential metabolic benefit. The guideline position is deliberately permissive rather than strongly recommending.
• Evidence for efficacy remains indeterminate — the honest position is that it may help, is unlikely to harm, and is not a substitute for lifestyle intervention.
• Commonly used as myo-inositol with D-chiro-inositol in a 40:1 ratio.
• Practical advantages: markedly better tolerated than metformin, available over the counter, acceptable to patients reluctant to take a "diabetes medicine". These adherence advantages are not trivial.
• Some evidence suggests benefit for pregnancy rates and insulin resistance markers in assisted reproduction contexts, though study quality is variable.

4. GLP-1 receptor agonists

The most rapidly evolving area, and one where careful framing is needed.

What the evidence shows: GLP-1 receptor agonists reduce body weight, BMI and insulin resistance in PMOS. Meta-analytic comparison suggests superiority over metformin for improving insulin sensitivity and reducing BMI and abdominal circumference.

What the evidence does not yet show: a 2026 systematic review found that while short-term weight loss is established, evidence for metabolic, reproductive and psychological outcomes remains uncertain because available studies are limited in number, size and duration. There is no good long-term data on live birth rates or on durability after discontinuation.

Practical cautions:

• Nausea, vomiting and dizziness are common
• These agents must be stopped before conception. Washout periods vary by agent — confirm for the specific drug. Effective contraception is required during use in patients who may conceive.
• Weight regain after discontinuation is well documented
• Cost is a substantial barrier in Indian practice
• Regulatory approval for PMOS specifically differs from approval for obesity or diabetes — prescribing may be off-label depending on the indication and agent

Reasonable position: consider in patients with obesity and PMOS where lifestyle intervention and metformin have been insufficient, with explicit discussion of the uncertainty, the contraception requirement, and the likelihood of weight regain on stopping.

5. Combined oral contraceptives — a caveat

COCs are first-line for cycle regulation and hyperandrogenic symptoms, but they do not treat insulin resistance and some formulations may modestly worsen glucose parameters and lipids. A patient on a COC with well-regulated cycles still requires ongoing metabolic screening. This is a common blind spot: cycles look controlled, so metabolic risk is assumed to be controlled. It is not.

---

For gynaecologists: practice points

Screen every patient with an OGTT, regardless of BMI. The most common error in real-world PMOS care is deferring metabolic screening in lean patients. Insulin resistance in PMOS is BMI-independent, and South Asian patients manifest it at lower BMI thresholds.

Look at the neck. Acanthosis nigricans takes five seconds to check and is frequently the finding that converts a reluctant patient into an engaged one.

Do not use HOMA-IR to rule out insulin resistance. A normal value in a symptomatic patient does not exclude it.

Titrate metformin slowly and use extended-release. Most metformin discontinuation is avoidable and results from starting at 1000 mg.

Check B12 annually on long-term metformin.

Re-screen at intervals. Metabolic status is not static. An OGTT at diagnosis and never again is a missed opportunity — particularly before planned pregnancy, where undiagnosed impaired glucose tolerance materially changes antenatal risk.