HOMA-IR Calculator

HOMA stands for Homeostasis Model Assessment. The original model was published by Matthews and colleagues in 1985 to estimate insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations.

This page calculates the widely used linear approximation called HOMA1-IR. It does not directly measure insulin sensitivity, and it does not perform the later HOMA2 computer model.

Fasting insulin is entered in µIU/mL or the numerically equivalent mIU/L. The two equations return nearly identical results when the glucose conversion is handled accurately.

Oxford's HOMA group states that there is no absolute value for HOMA indices and no defined universal threshold separating normal from abnormal. Results depend on the assays used for glucose, insulin, and C-peptide.

Population studies have proposed different cutoffs based on age, sex, ethnicity, health status, study design, percentile choice, and the outcome researchers were trying to identify. A Spanish population study, for example, reported values ranging from about 1.85 to 3.46 depending on the subgroup and criterion. Those are study-specific findings, not universal personal targets.

Oxford explains that HOMA2 does not use a simple published equation. Its iterative model may return results that differ considerably from the HOMA1 linear approximation, especially at more extreme glucose and insulin values.

If a study, laboratory, or clinician reports HOMA2-IR, do not compare it directly with the HOMA1-IR number from this page unless a qualified reviewer has established that the comparison is appropriate.

These examples show that the output is driven by the product of both inputs. They do not prove that one person has or does not have insulin resistance, and they do not establish a treatment target.

Use values from the same draw

Glucose and insulin fluctuate. A fasting insulin value from Monday and a fasting glucose value from a different month did not describe the same steady state and should not be multiplied together.

Confirm the fasting instructions

Follow the fasting instructions given by the ordering clinician or laboratory. Do not invent a fasting duration based on a generic calculator page. If the sample was not fasting or you are unsure, the page can demonstrate the arithmetic but the result may not represent HOMA's intended use.

Check the insulin unit carefully

µIU/mL and mIU/L are numerically equivalent because of the way the prefixes and volume units cancel. A value in pmol/L is not the same number. Published sources and laboratory systems have used conflicting factors when moving between insulin international units and molar units, so this calculator deliberately rejects pmol/L inputs.

Use the laboratory value, not a home glucose reading

HOMA was designed around paired fasting blood measurements. Combining a laboratory insulin value with a separate finger-stick or continuous-glucose reading adds timing, specimen, and device differences that the equation does not correct.

Record specimen and assay details

Oxford notes differences between serum and plasma insulin values and recommends documenting serum use as a limitation. If you are tracking repeated results, keep the laboratory name, specimen type, assay, date, fasting duration, medication context, and both raw values.

HOMA1-IR may be difficult to interpret when the fasting steady-state assumption is not reasonable or when treatment and pancreatic function alter the relationship between glucose and insulin.

• Exogenous insulin can make a measured insulin concentration different from endogenous pancreatic insulin production.
• Insulin secretagogues and insulin-sensitizing medicines can alter one or both inputs.
• Type 1 diabetes, advanced beta-cell failure, pancreatic disease, or prior pancreatic surgery can weaken the assumptions behind the index.
• Pregnancy changes insulin sensitivity and requires pregnancy-specific clinical interpretation.
• Acute illness, infection, recent surgery, steroid exposure, severe sleep loss, and major stress can affect fasting glucose and insulin.
• Marked hyperglycemia or extreme insulin values can make the linear approximation especially unreliable.

Do not stop, start, or change medication based on this result. The calculator also cannot tell whether a medication caused a particular value.

NIDDK notes that testing for insulin resistance is used primarily in research studies and that healthcare professionals may not test for it routinely. Prediabetes and diabetes are evaluated with established glucose-based tests such as A1C, fasting plasma glucose, and in some settings an oral glucose tolerance test.

A HOMA1-IR result cannot by itself diagnose or exclude:

• Insulin resistance
• Prediabetes or diabetes
• Metabolic syndrome
• Polycystic ovary syndrome
• Fatty liver disease
• Cardiovascular disease
• Pancreatic beta-cell dysfunction

It also cannot predict when a condition will develop or show whether one intervention caused a change.

1. Were my glucose and insulin collected under the fasting conditions you intended?
2. Were the specimens plasma or serum, and which insulin assay did the laboratory use?
3. Does this HOMA1-IR calculation add useful information in my situation?
4. Are any of my medicines, supplements, diagnoses, or recent illnesses likely to affect the inputs?
5. Should the raw fasting insulin or glucose be repeated before comparing a trend?
6. Which established tests are more important for my clinical decisions?
7. If you use a threshold, what population, assay, and outcome is that threshold based on?

Bring the full laboratory report rather than only the calculated number. You can organize dated glucose and other markers in the Metabolic Progress Tracker.