HOMA-IR Calculator
Estimate fasting insulin resistance from glucose and insulin with the HOMA-IR model.
What it does
HOMA-IR turns a fasting glucose and insulin pair into one number that estimates how hard the body works to hold blood sugar steady. Higher values suggest the tissues respond less to insulin.
The model came from Matthews and colleagues in 1985 as a simple surrogate for clamp studies. Their constant, 405, makes the calculation line up with the reference method when glucose is in mg/dL and insulin in uIU/mL.
The tool above returns the index, an interpretation band, and the exact formula. The defaults, 100 mg/dL and 10 uIU/mL, give 2.47, which sits in the typical adult range for many labs.
Insulin resistance lives on a spectrum, not on a switch. Common cutoffs cluster near 2.5 for adults, but age, ethnicity, pregnancy, and assay calibration all shift what counts as high, so trends in one person beat single readings.
HOMA-IR is not a diagnosis and not a glucose tolerance test. It answers one narrow question: given fasting levels, how much insulin is needed for the current glucose? This page presents the math, a worked example, and practice checks for education only.
How to use
- Confirm the glucose and insulin came from the same fasting draw.
- Pick mg/dL or mmol/L to match the lab report.
- Enter the fasting glucose.
- Enter the fasting insulin in uIU/mL.
- Read the index and the band.
- Track values over months rather than reacting to one result.
Formula
- glucose
- Fasting glucose in mg/dL (divide mmol/L by 0.0555 to convert).
- insulin
- Fasting insulin in uIU/mL (some labs report mIU/L, numerically the same).
- 405
- Matthews normalization constant; the metric twin is 22.5.
The two constants are the same relationship scaled by the glucose unit: 405 divided by 18 equals 22.5, and 100 mg/dL equals 5.55 mmol/L. Insulin units mIU/L and uIU/mL are interchangeable. The model assumes steady fasting conditions, so a non-fasting draw or recent insulin dose breaks the assumptions entirely.
Worked example
Example: glucose 100, insulin 10
Fasting glucose 100 mg/dL with fasting insulin 10 uIU/mL (defaults).
- Multiply: 100 x 10 = 1,000.
- Divide by 405: 1,000 / 405 = 2.469.
- Rounded to two decimals: 2.47.
- Metric mirror: 5.55 x 10 / 22.5 = 2.47, the same index.
- Tool shows 2.47 with the typical range band.
Reference table
| HOMA-IR | Reading | Context |
|---|---|---|
| Under 1.4 | Below typical | More insulin sensitive than average. |
| 1.4 to 2.4 | Typical range | Common in healthy adults. |
| 2.5 to 4.9 | Resistance band | 2.5 is the most cited cutoff. |
| 5 and above | High | Often pairs with metabolic syndrome markers. |
Limits
- A surrogate, not a clamp study; individual accuracy varies widely.
- Requires a true fasting draw, typically 8 hours or more.
- Insulin assays differ between labs, so compare within one lab.
- Not validated for people on insulin therapy or with type 1 diabetes.
Common mistakes
- Using a random glucose or insulin instead of fasting values.
- Comparing readings from two different labs with different assays.
- Treating 2.5 as a diagnosis line rather than a screening hint.
- Measuring soon after exercise or illness, which skews insulin.
- Watching a single number instead of the trend over months.
Try it yourself
Quick checks. Answers sit tucked under each question.
Glucose 90 mg/dL, insulin 5?
1.11, below typical.
Glucose 126 mg/dL, insulin 12?
3.73, in the resistance band.
Glucose 7 mmol/L, insulin 15?
4.67, in the resistance band.
Questions
What is a normal HOMA-IR?
Many adult labs call 1 to 2.5 typical. Cutoffs from 2.0 to 2.9 appear in the literature, so the band matters less than your trend with one lab.
Why 405 and 22.5?
Matthews picked 405 so the index matches clamp-derived insulin sensitivity with mg/dL glucose. Dividing glucose by 18 to get mmol/L converts 405 into 22.5.
Do I need to fast?
Yes. The model assumes fasting steady state. Eating changes both glucose and insulin and invalidates the estimate.
Is HOMA-IR the same as HOMA-B?
No. HOMA-B estimates beta cell function with a different equation. This page computes the insulin resistance index only.
Can children use it?
Studies use HOMA-IR in children but with different cutoffs, often near 2.5 to 3.2 depending on the population. Use age-specific references.
How often should it be checked?
Research protocols typically repeat every 6 to 12 months. Clinical decisions belong to your care team, not a web calculation.
Sources
- Matthews et al 1985: Homeostasis model assessment (PubMed 3899825)
- HOMA-IR methodology review (PMC3083884)
- Wallace et al 2004: use and abuse of HOMA modeling (PubMed 15111519)