Estimates insulin resistance from a fasting glucose and fasting insulin, using the homeostasis model assessment.
Result
Enter the fasting glucose and fasting insulin.
Sources
4 studies
4
studies
31,832
citations
22,169
PubMed papers
Evidence
Known limitations
Derived the homeostasis model and the approximation formula from fasting glucose and insulin. HOMA insulin resistance correlated with the euglycemic clamp (Rs = 0.88), but its coefficient of variation was 31%, which the authors said limits its use.
Diabetologia: Journal Metrics Score 7.8, Q1 · cited 31,832 times
Reviewed more than 500 publications using HOMA. Recommended it for cohort and epidemiological studies, advised against measuring beta-cell function in isolation, and stressed robust input data and careful interpretation.
Diabetes Care: Journal Metrics Score 12.6, Q1 · cited 5,203 times
In 2,459 Spanish adults, the cut-offs best linked to the metabolic syndrome in people without diabetes were 1.85 to 2.07, well below the 90th-percentile threshold of 3.46 from an earlier Spanish study. In women without diabetes the best cut-off rose with age.
BMC Endocr Disord: Journal Metrics Score 3.2, Q1 · cited 574 times
PubMed · Free full text · DOI
From a large Brazilian laboratory database, proposed a HOMA-IR reference interval of 0.39 to 2.86 in adults, similar in women and men, with fasting insulin of 2.52 to 13.14 µU/mL.
Arch Endocrinol Metab: Journal Metrics Score 1.8, Q2 · cited 16 times
PubMed · Free full text · DOI
The original paper has been cited 31,832 times, and 22,169 PubMed articles mention HOMA-IR in their title or abstract. Citation counts from OpenAlex and article counts from PubMed, retrieved November 11, 2026. Journal Metrics Scores and quartiles are from this site’s current release; follow the journal name for details.
After the result
High HOMA-IR · Borderline · Following change over time
Look at the whole metabolic picture. Interpret with your laboratory and population. Use the same laboratory each time.
Next steps
Avoid common errors
5 common errors to avoid
and 3 more
When HOMA-IR misleads
1Not a true fast
Insulin rises quickly after food, so a sample that is not fasting gives a falsely high HOMA-IR. Take blood after an overnight fast of at least 8 hours.
2Insulin treatment
In anyone injecting insulin, the measured insulin includes the injected hormone, so HOMA-IR is not valid. Insulin secretagogues such as sulfonylureas also raise insulin.
3Assay differences
Insulin immunoassays from different manufacturers can give quite different values for the same sample. Cut-offs from one laboratory or study do not transfer reliably to another.
4Unit mix-ups
Using insulin in pmol/L as if it were µU/mL makes HOMA-IR six times too high. Check the unit on the report.
5Children and pregnancy
Insulin resistance rises normally in puberty and pregnancy. Adult cut-offs do not apply.
Formula
Fasting glucose (mmol/L) × fasting insulin (µU/mL) / 22.5, or glucose (mg/dL) × insulin / 405.
glucose × insulin / 22.5
Fasting glucose (mmol/L) × fasting insulin (µU/mL) / 22.5, or glucose (mg/dL) × insulin / 405.
Formula and ranges
| Units | HOMA-IR |
|---|---|
| Glucose in mmol/L | glucose × insulin (µU/mL) / 22.5 |
| Glucose in mg/dL | glucose × insulin (µU/mL) / 405 |
| Insulin in pmol/L | divide by 6 to get µU/mL first |
| Result | HOMA-IR | Meaning |
|---|---|---|
| Below common cut-offs | below 2.0 | Below 2.0: under the insulin resistance cut-offs used in most adult population studies. |
| Borderline | 2.0 to 2.9 | 2.0 to 2.9: above the cut-off some studies use for insulin resistance (1.85 to 2.07 in Spanish adults without diabetes), yet within the reference interval of others (upper limit 2.86 in a Brazilian laboratory database). |
| High: suggests insulin resistance | 3.0 or higher | 3.0 or higher: above most published cut-offs and the upper reference limit in some populations (2.86 in a Brazilian laboratory database), suggesting insulin resistance. HOMA-IR is an estimate: it supports, but does not make, a diagnosis. |
Matthews 1985
Cut-offs
Units
Test cases
3 published cases, each also a test case
3 published cases; the calculator reproduces every result.
Worked examples
A 14-year-old girl with premature ovarian insufficiency had a fasting insulin of 37.9 µIU/mL and a fasting glucose of 5.27 mmol/L. The authors calculated HOMA-IR with the 22.5 formula and give about 8.88, well above their 2.5 cut-off.
“the HOMA-IR index of the patient was approximately 8.88”
The calculator gives 8.877, shown as 8.9. Adult cut-offs may not apply to adolescents, whose HOMA-IR rises in puberty.
Entered as: Fasting glucose: 5.27 mmol/L; Fasting insulin: 37.9 µU/mL (mU/L).
Source: Chen S et al., J Med Case Rep 2026 (PubMed · Free full text · DOI)
A 65-year-old woman with a partial deletion of the X chromosome long arm and obesity had, in blood tests 12 months before referral, prediabetes by HbA1c (5.9%), a fasting glucose of 99 mg/dL and a fasting insulin of 12.14 µIU/mL. The report gives a HOMA-IR of 3.0, read as insulin resistance.
“a HOMA-IR value of 3.0 [fasting plasma glucose: 99 mg/dL; fasting insulinemia: 12.14 μIU/mL]”
A normal-looking fasting glucose with a moderately raised insulin is the pattern HOMA-IR is meant to pick up.
Entered as: Fasting glucose: 99 mg/dL; Fasting insulin: 12.14 µU/mL (mU/L).
Source: Marzolla V et al., Reports (MDPI) 2026 (PubMed · Free full text · DOI)
A 48-year-old man with newly diagnosed type 2 diabetes and an HbA1c above 10% had a fasting glucose of 17.6 mmol/L and a fasting insulin of 78.3 pmol/L. The report’s table gives a HOMA-IR of 10.2, which fell to 2.54 and then 2.36 with lifestyle change alone.
The calculator converts 78.3 pmol/L to 13.05 µU/mL with the factor of 6 and gives 10.21, matching the report. At this degree of hyperglycemia much of the high value comes from the glucose, not the insulin; the later values are reproduced to within rounding (6.3 mmol/L with 54.3 pmol/L gives 2.53 against the reported 2.54).
Entered as: Fasting glucose: 17.6 mmol/L; Fasting insulin: 78.3 pmol/L.
Source: Ferreira A et al., Endocrinol Diabetes Metab Case Rep 2026 (PubMed · Free full text · DOI)
FAQ
6 questions and answers
and 4 more answers
Questions
HOMA-IR = fasting glucose (mmol/L) × fasting insulin (µU/mL) / 22.5. With glucose in mg/dL, divide by 405 instead. A fasting glucose of 99 mg/dL and insulin of 12.14 µU/mL give 99 × 12.14 / 405 = 3.0. The formula is the 1985 approximation of the homeostasis model by Matthews and colleagues.
There is no single normal value. Reference limits depend on the population and the insulin assay: the upper limit was 2.86 in a large Brazilian laboratory database, while a Spanish population study put the cut-offs linked to the metabolic syndrome at 1.85 to 2.07 in adults without diabetes. Below 2 is under most published adult cut-offs and 3 or more is above most of them; compare with your laboratory’s interval.
Divide by 6: 78 pmol/L is 13 µU/mL. Some laboratories use 6.945 instead, which gives a slightly lower HOMA-IR. µU/mL, µIU/mL and mU/L are the same unit.
Not if the patient injects insulin: the measured insulin is then not the body’s own. In type 2 diabetes HOMA-IR can be calculated, but failing insulin secretion pulls it down as the disease progresses, so it is best read alongside the glucose and HbA1c.
It is a supporting test, not a diagnosis. No guideline sets a HOMA-IR cut-off for diagnosing insulin resistance, prediabetes or polycystic ovary syndrome; those diagnoses rest on glucose, HbA1c and clinical criteria. HOMA-IR is most reliable for comparing groups or following one person over time on the same assay.
HOMA-IR here is the simple 1985 formula. HOMA2 is an updated computer model from Oxford that accounts for glucose resistance in the liver and periphery and gives insulin sensitivity (%S) and beta-cell function (%B). The two give different numbers, so do not mix their cut-offs.
Further reading
7 references with PubMed and DOI links
7
references
5
free full text
From Matthews DR et al., Diabetologia 1985 to the newest case reports.
References
Contributed by Dr. Meng Zhao (MD · Former Neurosurgeon · Medical AI Researcher), from the original publications listed here. Last updated November 11, 2026. LinkedIn
The calculator code is tested against every worked example on this page before each release.
This calculator supports, and does not replace, clinical judgment. It is meant for healthcare professionals and is not medical advice for patients. Report an error through the contact page.
clinical grade of aneurysmal subarachnoid hemorrhage, I to V.
total calcium adjusted for albumin, in mg/dL or mmol/L.
stroke severity from 15 bedside items, 0 to 42.
MAP from systolic and diastolic pressure, against the 65 mm Hg target.
eye, verbal and motor scores for preverbal children, 3 to 15.
oxygenation ratio with the Berlin ARDS severity ranges.
serum sodium corrected for high glucose, Katz or Hillier factor.
sodium − (chloride + bicarbonate), with the albumin-corrected gap.
urea nitrogen to creatinine ratio, in US or SI units.
estimated creatinine clearance with actual, ideal or adjusted weight.
cardiac output and index from oxygen consumption and saturations.
ANC from the white cell count, neutrophils and bands.
FENa from plasma and urine sodium and creatinine: prerenal, indeterminate or tubular pattern.
three-item stroke severity screen for large vessel occlusion, 0 to 4.
water deficit in hypernatremia from sodium, weight and body water.
expected PaCO2 in metabolic acidosis, 1.5 × HCO3 + 8 ± 2.
For sodium in hyperglycemia.
Which published the Wallace guidance on HOMA, with its current metrics.
Where the original HOMA paper appeared.