HOMA-IR estimates insulin resistance from a single fasting blood draw, using just two numbers: fasting glucose and fasting insulin. It was designed as a low-cost proxy for far more invasive gold-standard tests, and it remains one of the most widely used research and screening tools for insulin resistance today.
How the HOMA-IR Calculator works
HOMA-IR multiplies fasting glucose by fasting insulin and divides by a constant (405 for glucose in mg/dL, or 22.5 for glucose in mmol/L). The formula comes from the Homeostatic Model Assessment introduced by Matthews and colleagues in a 1985 Diabetologia paper, which modeled the feedback loop between the liver, pancreas, and peripheral tissues to approximate insulin resistance from steady-state fasting values rather than a multi-hour clamp study.
Because HOMA-IR is a product of two values, a rise in either glucose or insulin raises the score. In practice, insulin tends to rise earlier and more sharply than glucose as insulin resistance develops — the pancreas compensates by secreting more insulin to hold glucose steady — so fasting insulin is often the input that moves the result most for a given person.
Inputs and what they mean
Fasting glucose must come from a blood draw after at least 8 hours without food, and is accepted in mg/dL or mmol/L via the unit toggle (1 mmol/L = 18 mg/dL). Fasting insulin must come from the same draw, in microunits per milliliter (uIU/mL) — numerically the same as mIU/L. If your lab reports insulin in pmol/L, use the Unit Conversion tab (divide by 6.945) before entering it.
Both values must be truly fasting — a non-fasting sample, recent illness, or recent strenuous exercise can transiently shift insulin and invalidate the result. Because HOMA-IR is a ratio-like product, small measurement errors in either input propagate directly into the score.
Limits and edge cases
HOMA-IR has no single validated cutoff. Studies in different populations — by ethnicity, age, sex, and body-mass index — have derived meaningfully different optimal thresholds, and the result also depends on which insulin assay a lab uses, since assays are not fully standardized against each other. This calculator's bands (optimal, normal, borderline, resistant) are a general screening reference, not a diagnostic threshold, and cutoffs commonly cited elsewhere (for example NHANES's frequently used 2.5 threshold) sit within this calculator's borderline band rather than at its edge.
HOMA-IR also assumes steady-state beta-cell function and does not directly apply to people on insulin therapy, with significant beta-cell failure (e.g. long-standing type 1 diabetes), or with acute illness affecting glucose or insulin. In those situations, or whenever a result is unexpected, the next step is a conversation with a clinician — not a self-directed diagnosis from this calculator.