Skip to content
pmos·pcosGet started
PCOS glossary · updated 16 May 2026

HOMA-IR

Homeostasis Model Assessment of Insulin Resistance

HOMA-IR is an estimation tool, not a diagnosis. Its thresholds vary across studies and laboratories. Do not self-diagnose from this score: only your doctor can interpret it in your clinical context.

What it is

HOMA-IR is a mathematical index that estimates insulin resistance from two values measured while fasting: blood glucose and insulin. Its formula is:

HOMA-IR = (fasting insulin × fasting glucose) ÷ constant

Constant: 405 if glucose in mg/dL · 22.5 if glucose in mmol/L

The index was first described by Matthews et al. in 1985. It is based on a mathematical model of the insulin-glucose relationship at baseline, assuming a homeostatic equilibrium.

Why it matters in PCOS

Insulin resistance is present in 35–80% of people with PCOS, and plays a role in amplifying hyperandrogenism. HOMA-IR is sometimes used to assess the metabolic component of PCOS and guide management (physical activity, nutrition, possible metformin).

It is not included in the diagnostic criteria for PCOS, but may be part of a supplementary workup depending on the individual's profile.

Important limitations

HOMA-IR has several limitations that are essential to understand:

  • Thresholds are not universal — reference values vary across studies (often cited between 2 and 3.5), populations, ethnic groups, and insulin assay methods
  • Insulin measurement is not standardised — laboratories do not all use the same techniques, making comparisons between results difficult
  • It is not a diagnostic test — a high score points toward probable insulin resistance, but does not diagnose metabolic syndrome or type 2 diabetes
  • A score within the normal range does not rule out insulin resistance, particularly in mild forms or in lean people with PCOS

Key takeaways

  • Calculated from fasting glucose and fasting insulin
  • Estimates (without diagnosing) insulin resistance
  • Thresholds vary by study and laboratory — no universal value
  • Insulin assay not standardised: compare trends, not absolute numbers
  • Always interpret with your doctor, in clinical context