Insulin Resistance Calculator

How it works

Download the template and fill in one row per subject.
Upload the completed workbook (.xlsx).
Review results for every subject, then download them as an Excel file.

Upload your data

Results

Estimates for research and educational use only — not for diagnosis, treatment, or any medical decision. Validate independently before relying on any result.

Subject	Fasting G mg/dL	Fasting I uU/mL	HOMA-IR	WBISI Matsuda	Minimal Model SI x10⁻⁴ dL/kg/min per uU/mL	Fit RMSE mg/dL	Fit Confidence

About the metrics

HOMA-IR: (fasting glucose [mg/dL] × fasting insulin [uU/mL]) / 405 — a simple fasting-state index of insulin resistance.
WBISI (Matsuda Index): 10000 / √[(fasting G × fasting I) × (mean G × mean I)] over the standard 0/30/60/90/120-min OGTT window — a whole-body insulin sensitivity index.
Cobelli Oral Minimal Model SI: Fits a two-compartment glucose/remote-insulin ODE model to the full OGTT glucose curve by nonlinear least squares, using measured insulin as a forcing input. SG and the distribution volume V are fixed to published population constants (Dalla Man, Caumo & Cobelli, 2002) rather than freely estimated, since SG and SI cannot both be reliably identified from OGTT data alone. SI is computed by a more flexible 8-parameter piecewise-linear gut-absorption model (the primary estimator shown here), with the simpler single-parameter gamma-shaped model retained and computed alongside it as a fast cross-check. Reported alongside a glucose-fit RMSE so low-confidence fits (poor convergence, a fitted parameter at its physiological bound, high fit error, or an SI value above the typical physiological range) are flagged — a low or near-zero SI is treated as a valid result (severe insulin resistance), not an error. When the two models disagree substantially on the same subject, that disagreement is shown as a note rather than hidden, since it signals genuine uncertainty in the fit for that individual's data.

This tool is provided for research and educational purposes only. It is not a medical device and must not be used to guide clinical decisions.

References & Methodology

This implementation is based on the following published methods:

Dalla Man C, Caumo A, Cobelli C. The oral glucose minimal model: estimation of insulin sensitivity from a meal test. IEEE Trans Biomed Eng. 2002;49(5):419-429.
Cobelli C, Dalla Man C, Toffolo G, Basu R, Vella A, Rizza R. The oral minimal model method. Diabetes. 2014;63(4):1203-1213.
Cobelli C, Dalla Man C. Minimal and maximal models to quantitate glucose metabolism. J Diabetes Sci Technol. 2022;16(5):1270-1298.
Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing. Diabetes Care. 1999;22(9):1462-1470. (WBISI/Matsuda index)
Matthews DR, et al. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7):412-419. (HOMA-IR)

Simplifications relative to the published method

This tool's Cobelli oral minimal model fit is a simplified implementation, not a certified equivalent of the original method. Specifically:

Gut glucose absorption Ra(t) is modeled with an 8-knot piecewise-linear curve (the primary estimator) rather than the full published absorption model; a simpler single-parameter gamma-shaped curve is computed alongside it as a fast cross-check.
SG and the distribution volume V are fixed to published population constants rather than individually estimated per subject, to avoid a known structural identifiability problem between SG and SI.

Results should be interpreted with these simplifications in mind and validated against established clinical methods before any downstream use.