PCOS and the Insulin Connection: Why Metabolic Treatment Often Outperforms Hormonal Treatment

Roughly 8 to 13 percent of women of reproductive age meet the diagnostic criteria for polycystic ovary syndrome, making it one of the most common endocrine conditions in this population. The 2023 international evidence-based guideline, led by Teede and an international expert panel, represents the most comprehensive consensus on diagnosis and management currently available. The guideline reflects a substantial shift from earlier framings of the condition. PCOS is no longer treated primarily as an ovarian or hormonal disorder. It is treated as a metabolic disorder that produces hormonal manifestations.

This reordering has direct implications for treatment, and it explains why interventions targeting insulin resistance tend to outperform interventions targeting androgens or menstrual cyclicity directly.

What the Diagnostic Criteria Actually Capture

The widely used Rotterdam criteria require two of three features for PCOS diagnosis: oligo-ovulation or anovulation, clinical or biochemical hyperandrogenism, and polycystic ovarian morphology on ultrasound. The criteria deliberately allow heterogeneous presentations because the underlying biology is heterogeneous. A woman might present with irregular cycles and acne but normal ovaries on imaging. Another might have multiple cysts and high testosterone but cycle regularly. Both would meet diagnostic criteria. Their underlying physiology may differ substantially.

The unifying feature across most clinical PCOS phenotypes is insulin resistance. Diamanti-Kandarakis and Dunaif, in their definitive 2012 review of the mechanism, summarized the evidence that insulin resistance is present in 50 to 70 percent of women with PCOS, including women who are not overweight. The insulin resistance is not a complication that develops downstream from PCOS. It is, in most cases, mechanistically upstream of the visible hormonal abnormalities.

The pathway works approximately like this. Insulin acts on the ovaries to stimulate androgen production by theca cells, particularly testosterone and androstenedione. In the presence of normal insulin signaling, this effect is modest and well-regulated. In the presence of insulin resistance, the compensatory hyperinsulinemia drives chronically elevated theca cell androgen output. The hyperinsulinemia also suppresses hepatic production of sex hormone-binding globulin, increasing the free fraction of circulating testosterone available to act on target tissues. The result is the clinical and biochemical hyperandrogenism that is one of the diagnostic features. The same hyperinsulinemia disrupts the LH:FSH ratio in ways that interfere with normal follicular development, contributing to the anovulation and cyst formation visible on ultrasound.

In this framework, the polycystic ovaries, the irregular menses, and the elevated androgens are downstream symptoms. The insulin signaling defect is the upstream driver in the majority of cases.

Why Metabolic Interventions Tend to Outperform

This causal model predicts that interventions targeting insulin sensitivity should produce broad improvements across the cluster of PCOS features, while interventions targeting only one feature should produce narrower improvements. This is, in fact, what the trial data show.

Moran and colleagues, in a Cochrane review of lifestyle interventions in PCOS, pooled data from multiple trials and demonstrated that even modest weight loss in overweight women with PCOS produced improvements in androgen levels, menstrual regularity, ovulation rates, and metabolic markers. The improvements tracked the metabolic intervention rather than any direct hormonal targeting. The mechanism is consistent with the insulin-driven model: improving insulin sensitivity reduces the upstream driver, and the downstream features improve in parallel.

Metformin, originally developed as a diabetes medication, has been studied extensively in PCOS and produces reliable improvements in ovulatory function, menstrual regularity, and metabolic parameters. The 2023 Teede guideline recommends metformin as first-line pharmacotherapy for several PCOS phenotypes, particularly when metabolic features are prominent or when fertility is the primary concern. Hormonal contraception, by contrast, masks the menstrual irregularity but does not address the underlying insulin resistance and is associated with no metabolic benefit during use.

Inositol has emerged as a particularly interesting intervention. Myo-inositol and D-chiro-inositol are signaling molecules involved in insulin transduction, and inositol deficiency in PCOS appears to contribute to the cellular insulin resistance characteristic of the condition. Unfer and colleagues, in a 2017 meta-analysis of randomized trials, demonstrated that myo-inositol supplementation at doses around 4 grams daily produced improvements in ovulation rates, cycle regularity, and androgen levels comparable to or greater than metformin in some subgroups, with a substantially better tolerability profile. The supplement is widely available and the safety data is strong, though access and clinical familiarity vary by region.

The Lean PCOS Subtype

A common clinical pitfall is the assumption that lean women cannot have insulin resistance and therefore cannot have insulin-driven PCOS. This assumption is incorrect. Approximately 30 to 40 percent of lean women with PCOS show measurable insulin resistance using gold-standard methods like the euglycemic-hyperinsulinemic clamp. The resistance in these women is often muscle-specific or hepatic, and standard fasting glucose tests will miss it almost entirely. HOMA-IR may capture it inconsistently. The diagnostic challenge is real, and many lean women with PCOS receive years of hormonal treatment before the underlying metabolic feature is identified and addressed.

This subtype illustrates why broad metabolic screening, including consideration of fasting insulin and not just fasting glucose, is appropriate at PCOS diagnosis even in normal-weight women. The treatment implications are substantial. A lean woman with PCOS and demonstrable insulin resistance has the same upstream driver as her overweight counterpart and benefits from the same upstream interventions, even though weight loss is not part of the prescription.

The Practical Synthesis

For most women diagnosed with PCOS, the treatment hierarchy that has emerged from the recent literature looks like this. Lifestyle interventions targeting insulin sensitivity — exercise, dietary patterns that reduce postprandial glucose excursions, modest weight loss when appropriate — form the foundation. These produce improvements across the diagnostic features and address the upstream driver. Pharmacological metabolic intervention, primarily metformin or inositol, is added when lifestyle alone is insufficient or when faster results are needed. Hormonal contraception remains useful for managing specific symptoms like menstrual irregularity or hyperandrogenic skin findings, but should be understood as treating downstream features rather than the underlying condition.

The reframing of PCOS as primarily metabolic does not change the diagnostic criteria, but it does change which interventions tend to be tried first and which are reserved for symptom management. The earlier framing positioned hormonal contraception as primary therapy with metabolic intervention as adjunctive. The current framing, supported by the 2023 international guideline and twenty years of accumulated trial data, has effectively reversed that order. The clinical implication is that women with new PCOS diagnoses are increasingly seen primarily by reproductive endocrinologists who think metabolically rather than by gynecologists who think hormonally. The change is recent, the access varies, but the direction of clinical practice is clear.