HOMA-IR Calculator

HOMA-IR in brief

HOMA-IR (the Homeostatic Model Assessment of Insulin Resistance) is a score calculated from a single fasting blood sample. It combines fasting glucose and fasting insulin into one number that estimates how resistant the body is to insulin. The formula is:

• HOMA-IR = fasting glucose (mg/dL) × fasting insulin (µIU/mL) ÷ 405, or
• HOMA-IR = fasting glucose (mmol/L) × fasting insulin (µIU/mL) ÷ 22.5

A fasting glucose of 95 mg/dL with an insulin of 8 µIU/mL gives 95 × 8 ÷ 405 = 1.88. The calculator above accepts glucose in mg/dL or mmol/L and insulin in µIU/mL or pmol/L, converts as needed, and places your result in a band with an explanation.

HOMA-IR is widely used in research and increasingly requested by people interested in metabolic health. It can be informative, but it also has important limitations: there is no universal cut-off, insulin assays vary, and it is not a diagnostic test for diabetes. This guide explains how the score was developed, how to calculate it correctly, how to interpret it, and when it is – and is not – useful.

What insulin resistance means

After a meal, glucose enters the blood and the pancreas releases insulin, which signals muscle, fat and liver cells to take up and store glucose and tells the liver to stop making new glucose. Overnight, insulin falls to a low baseline, and the liver releases just enough glucose to keep levels steady.

In insulin resistance, cells respond less strongly to insulin. To keep glucose normal, the pancreas compensates by producing more insulin. For years – sometimes decades – this compensation keeps fasting glucose in the normal range, but at the cost of higher insulin levels. Type 2 diabetes develops when the beta cells of the pancreas can no longer keep up.

Insulin resistance is closely associated with:

• excess abdominal (visceral) fat and fat in the liver
• physical inactivity and poor sleep
• metabolic syndrome, high triglycerides and low HDL cholesterol
• polycystic ovary syndrome (PCOS)
• metabolic dysfunction-associated steatotic liver disease (MASLD)
• family history of type 2 diabetes and certain ancestries
• some medicines, such as corticosteroids and some antipsychotics
• pregnancy (a normal, temporary increase)

Where HOMA-IR comes from

HOMA was developed by David Matthews, Robert Turner and colleagues at Oxford and published in Diabetologia in 1985. The model describes the feedback loop between the liver (which produces glucose) and the pancreas (which produces insulin) in the fasting state. From a fasting glucose and insulin, it estimates:

• HOMA-IR – insulin resistance, where 1.0 represents a reference value in young, lean, healthy adults
• HOMA-%B – beta-cell function (insulin secretion), where 100% is the reference
• HOMA-%S – insulin sensitivity, the inverse of HOMA-IR × 100

The simple formula used in this calculator is the original HOMA1 approximation. The constant 22.5 (or 405 in mg/dL) is the product of a normal fasting insulin of 5 µIU/mL and a normal fasting glucose of 4.5 mmol/L (81 mg/dL), so that a "normal" person scores about 1.0.

A computer-based version, HOMA2, was published in 1998. It accounts for more physiology, including the effect of high glucose on insulin secretion and kidney glucose loss, and can use C-peptide instead of insulin. HOMA2 values are generally lower than HOMA1 values from the same inputs. Published cut-offs differ between HOMA1 and HOMA2, so it matters which version a study used.

How to calculate HOMA-IR step by step

1. Collect a fasting blood sample (at least 8 hours without food; water is fine) and measure glucose and insulin from the same draw.
2. Make sure insulin is in µIU/mL (also written µU/mL or mU/L). If it is in pmol/L, divide by 6.
3. If glucose is in mg/dL, multiply glucose by insulin and divide by 405.
4. If glucose is in mmol/L, multiply glucose by insulin and divide by 22.5.
5. Round to one or two decimal places.

Worked examples

• Insulin-sensitive: glucose 85 mg/dL, insulin 4 µIU/mL → 85 × 4 ÷ 405 = 0.84.
• Typical: glucose 5.0 mmol/L, insulin 8 µIU/mL → 5.0 × 8 ÷ 22.5 = 1.78.
• Borderline: glucose 100 mg/dL, insulin 10 µIU/mL → 100 × 10 ÷ 405 = 2.47.
• Insulin resistant: glucose 110 mg/dL, insulin 20 µIU/mL → 110 × 20 ÷ 405 = 5.43.
• Insulin reported in pmol/L: glucose 5.5 mmol/L, insulin 90 pmol/L → insulin = 90 ÷ 6 = 15 µIU/mL → 5.5 × 15 ÷ 22.5 = 3.67.

The most common mistake

Entering insulin in pmol/L into the formula without converting it makes HOMA-IR six times too high. In the last example, using 90 directly would give 22.0 instead of 3.67. Another frequent error is mixing glucose units – using glucose in mg/dL with the divisor 22.5 makes the result 18 times too high. The calculator above prevents both errors by asking for units.

Interpreting HOMA-IR

There is no universally agreed normal range for HOMA-IR. Cut-offs depend on the population studied, the insulin assay used, age, sex, ethnicity and body weight. As a rough guide used in many studies of adults:

Examples of published cut-offs illustrate the variation:

• In the original Matthews paper, healthy non-obese adults had a mean HOMA-IR of about 1.0, as designed.
• Studies in different populations have defined insulin resistance using cut-offs from about 1.8 to 3.8, often based on the 75th or 90th percentile of a reference group.
• Studies of children and adolescents use different thresholds, partly because insulin resistance rises naturally during puberty.

So a value of 2.5 may be "high" in one study and "normal" in another. HOMA-IR is best used to track an individual over time with the same laboratory, or to compare groups in research, rather than as a pass-or-fail test.

When HOMA-IR is useful – and when it is not

Useful

• Research and population studies, where it is a validated, inexpensive estimate of insulin resistance that correlates reasonably well with the gold-standard euglycaemic clamp (correlation coefficients of about 0.6–0.8 in many studies).
• Tracking change in an individual over time – for example, after weight loss or a new exercise programme – using the same laboratory and conditions.
• Supporting a clinical picture of insulin resistance in specialist settings, such as some PCOS or fatty liver clinics.

Not useful or unreliable

• Diagnosing diabetes or prediabetes – these are diagnosed with glucose and HbA1c criteria, not HOMA-IR.
• People who inject insulin – injected insulin makes the insulin measurement meaningless for this purpose.
• People with significant beta-cell failure – in advanced type 2 diabetes or type 1 diabetes, insulin production is low, so HOMA-IR can be falsely reassuring.
• Non-fasting samples – any food intake invalidates the result.
• Comparing across laboratories – insulin assays differ by 20% or more, which directly changes HOMA-IR.
• Very high glucose – HOMA1 becomes less accurate at fasting glucose above about 140–180 mg/dL (7.8–10 mmol/L).

The American Diabetes Association and most clinical guidelines do not recommend routine measurement of insulin or HOMA-IR to screen for insulin resistance, because it does not change management for most people: the same lifestyle measures are recommended for anyone with risk factors, whatever their HOMA-IR.

HOMA-IR in specific groups

Polycystic ovary syndrome

Insulin resistance is common in PCOS, including in women of normal weight, and contributes to raised androgen levels. HOMA-IR is frequently used in PCOS research, but the 2023 international PCOS guideline does not recommend measuring insulin resistance with HOMA-IR or other tests in routine care, because the tests are poorly standardised and do not change treatment. Instead, it recommends assessing glucose status (with an oral glucose tolerance test in higher-risk women), lipids and blood pressure.

Pregnancy

Insulin resistance increases physiologically in the second and third trimesters, so HOMA-IR rises in normal pregnancy. Non-pregnant cut-offs do not apply, and gestational diabetes is diagnosed with glucose tests.

Children and adolescents

HOMA-IR rises during puberty and falls again afterwards, and levels differ by sex and pubertal stage. Studies in children have proposed cut-offs from about 2.5 to over 4, depending on age and population. Paediatric HOMA-IR values should be interpreted with age- and puberty-specific data, preferably by a specialist.

Ethnicity

At the same body mass index, people of South Asian, East Asian, Hispanic and Black African ancestry often have greater insulin resistance or higher diabetes risk than people of European ancestry. Some studies have proposed different cut-offs for different populations. This is another reason a single "normal" HOMA-IR cannot be applied to everyone.

Tracking HOMA-IR over time: an example

A 45-year-old man with a waist of 104 cm has a fasting glucose of 104 mg/dL and insulin of 18 µIU/mL, giving a HOMA-IR of 4.6. Over six months he increases his activity, changes his diet and loses 8% of his body weight. Repeat tests at the same laboratory show glucose 94 mg/dL and insulin 9 µIU/mL, a HOMA-IR of 2.1 – a reduction of more than half. His triglycerides, waist size and blood pressure also improved. The HOMA-IR change is consistent with the other markers, which together give a more complete picture than any single number.

HOMA-%B: the other half of the model

Insulin resistance alone does not cause diabetes; it is the failure of beta cells to compensate that matters. HOMA-%B estimates beta-cell function: HOMA-%B = 20 × insulin (µIU/mL) ÷ (glucose in mmol/L − 3.5). A high HOMA-IR with a high HOMA-%B suggests the pancreas is still compensating; a high HOMA-IR with a falling HOMA-%B suggests progression towards diabetes. HOMA-%B is less reliable than HOMA-IR in individuals and is mainly a research tool.

Other measures of insulin resistance

• Euglycaemic hyperinsulinaemic clamp – the gold standard, measuring how much glucose must be infused to keep glucose steady during a fixed insulin infusion. Accurate but laborious; used only in research.
• QUICKI – the Quantitative Insulin Sensitivity Check Index, 1 ÷ (log insulin + log glucose in mg/dL). It is mathematically related to HOMA-IR.
• Matsuda index – uses glucose and insulin values throughout an oral glucose tolerance test; reflects both liver and muscle insulin sensitivity.
• Triglyceride–glucose (TyG) index – ln(triglycerides × glucose ÷ 2), using mg/dL. It does not require an insulin measurement and correlates reasonably well with clamp studies, making it attractive where insulin assays are unavailable.
• Triglyceride/HDL ratio – a simple surrogate, though its performance varies by ethnicity.
• Clinical markers – waist circumference, blood pressure, triglycerides, HDL, fasting glucose and the presence of fatty liver often identify insulin resistance without special tests.

HOMA-IR is not a diagnostic test for diabetes. To see where your long-term glucose stands, check your result with our A1C calculator and chart, which shows normal, prediabetes and diabetes A1C ranges.

Improving insulin sensitivity

Whatever your HOMA-IR, these measures have the strongest evidence for improving insulin sensitivity:

• Physical activity. Both aerobic exercise and resistance training increase glucose uptake by muscle; effects begin within days and accumulate over weeks. Guidelines recommend at least 150 minutes a week of moderate activity plus muscle-strengthening on two or more days.
• Weight loss in people with excess weight, particularly reduction of abdominal and liver fat. Losing 5–10% of body weight produces substantial improvements.
• Diet quality. Diets lower in refined carbohydrates and added sugars and higher in fibre, vegetables, legumes, whole grains and unsaturated fats are associated with better insulin sensitivity. Several dietary patterns – Mediterranean, lower-carbohydrate, and others – can work if they lead to sustained weight loss.
• Sleep. Even a few nights of short sleep reduce insulin sensitivity; aiming for seven or more hours helps.
• Medicines, where indicated: metformin, GLP-1 receptor agonists, pioglitazone and SGLT2 inhibitors all improve glucose control, and some directly improve insulin sensitivity.

Repeating HOMA-IR after three to six months of sustained change, under the same fasting conditions and in the same laboratory, can show progress, but changes in waist size, triglycerides, fasting glucose and HbA1c are just as informative.

Preparing for a HOMA-IR test

• Fast for at least 8–12 hours; drink only water.
• Avoid strenuous exercise and alcohol the day before.
• Have the sample taken in the morning.
• Tell the laboratory or clinician about medicines that affect glucose or insulin, such as steroids, metformin or diabetes medicines.
• Make sure glucose and insulin come from the same blood draw.

When to see a doctor

See a doctor if your HOMA-IR is high together with a fasting glucose of 100 mg/dL (5.6 mmol/L) or more, an HbA1c of 5.7% (39 mmol/mol) or more, high triglycerides, high blood pressure or a large waist size, or if you have symptoms such as increased thirst, frequent urination or unexplained tiredness. If you have PCOS, fatty liver or a strong family history of type 2 diabetes, ask about appropriate screening with glucose or HbA1c rather than relying on HOMA-IR alone.