HOMA-IR: the insulin resistance test, its formula and how to read it

HOMA-IR is a number calculated from fasting glucose and fasting insulin that estimates insulin resistance, the state in which the body needs more insulin than usual to keep blood sugar steady. It is not a separate blood test: it comes from two results on the same fasting sample. A higher number suggests more insulin resistance.

What HOMA-IR measures

HOMA stands for homeostasis model assessment. It was described in 1985 from a model of how fasting glucose and fasting insulin hold each other in balance through a feedback loop. If cells resist insulin, the pancreas has to run at a higher insulin level to keep glucose at the same point, and HOMA-IR rises.

In the original study, HOMA-IR tracked the euglycaemic clamp, the reference method for insulin resistance, with a correlation of 0.88. The clamp is invasive and laborious, so it is used in research, and HOMA-IR is one of several simpler stand-ins that need a single fasting sample. Its authors describe it as useful in cohort and population studies, provided the glucose and insulin results going into it are sound.

The formula

HOMA-IR = fasting glucose (mmol/L) × fasting insulin (µIU/mL) ÷ 22.5

For example, a fasting glucose of 5.0 mmol/L and a fasting insulin of 9 µIU/mL give 5.0 × 9 ÷ 22.5 = 2.0. If insulin is reported in pmol/L, it has to be converted to µIU/mL before it goes into the formula, and the factor matters: Mayo Clinic Laboratories uses 1 µIU/mL = 6.945 pmol/L.

A later computer version, HOMA2, gives different values. In the Qatar Biobank study its insulin-based cut-off was 1.128, against 1.878 for HOMA-IR, so the two are not interchangeable.

Is there a normal range?

There is no single one. The American Diabetes Association does not use HOMA-IR in its criteria for diagnosing diabetes, and a 2013 Spanish study described the published cut-offs as greatly variable. They come from different populations, different insulin methods and different definitions of insulin resistance: some studies take the top 10% or 20% of values in a healthy population, others look for the value that most accurately separates people with metabolic syndrome or diabetes.

The cut-offs also shift with the laboratory's insulin method, and the Spanish study found that age and sex changed them too. If your report gives a range, read the result against it, and with a doctor.

What a high result usually points to

A high HOMA-IR usually points to insulin resistance: the body is holding glucose steady by making more insulin. MedlinePlus lists prediabetes, type 2 diabetes and polycystic ovary syndrome among the situations where insulin resistance is more likely, and higher values travel with the features of metabolic syndrome. It is an estimate, not a diagnosis.

A doctor will usually read it with HbA1c, fasting glucose, triglycerides, HDL cholesterol, weight and blood pressure, several of which are on the ADA's list of reasons to screen for diabetes, and may repeat the fasting tests before reading much into one value.

What a low result usually points to

A low HOMA-IR generally means the body needs little insulin to hold fasting glucose steady. There is one exception worth knowing. If glucose is high and insulin is low, HOMA-IR can come out low because the pancreas is not making enough insulin, not because cells are sensitive to it. HOMA-IR also cannot be read in the usual way in people who use insulin, because injected insulin and insulin antibodies affect the insulin result.

What moves HOMA-IR

Anything that moves either input. On the glucose side: food within the fasting window, stress, illness, recent exercise and some medicines. On the insulin side: the length of the fast, haemolysis or delay in handling the sample, biotin with some methods, insulin medicines, and the insulin method itself.

Together these make it a noisy number. In the European Biological Variation Study, HOMA-IR varied by about 27% within the same person across 10 weekly fasting samples, driven largely by insulin, and the authors concluded that this limits the value of single measurements.

Preparing for the test

Preparation is the same as for its two inputs: fast for 8 to 12 hours, or as the laboratory instructs, and give the glucose and insulin samples at the same draw. Tell the person taking the sample about insulin or other diabetes medicines and any biotin supplement.

How Clare's panel handles it

In Clare's panel, HOMA-IR is calculated by Clare, not measured by the lab. It uses the fasting glucose and fasting insulin from the same panel: glucose (mmol/L) × insulin (µIU/mL) ÷ 22.5. Where insulin is reported in pmol/L, Clare converts it at 6.945 pmol/L per µIU/mL before the calculation. When Clare opens, it will appear as a derived biomarker beside the two results it comes from.

Why it matters in the UAE

We found no HOMA-IR cut-off developed for the UAE population. The closest data come from neighbours: 1.62 in Arab adults in Nizwa, Oman, and 1.878 in the Qatar Biobank, both lower than many cut-offs from Europe, Latin America and East Asia. The Qatar authors suggest their values could serve as a reference for Middle Eastern populations, and found that the TyG Index, which uses triglycerides instead of insulin, had the highest accuracy of the seven indices they compared.

Related tests

HOMA-IR comes from Fasting Glucose and Fasting Insulin, so those two are worth reading beside it, and the TyG Index offers a second estimate of insulin resistance from triglycerides and glucose. The article HbA1c, glucose and insulin: reading them together puts them side by side.

HOMA-IR over time

Because a single HOMA-IR moves so much, its value is in the trend: several results, taken after the same length of fast and ideally at the same laboratory, read across months alongside HbA1c, which changes more slowly. A sustained shift in one direction says more than any one number, and it is worth talking through with a doctor with the results in front of you.

Clare is opening with a waitlist at clare.health.