Fasting Insulin Index Calculator

1. Enter the fasting glucose from your lab report and pick its unit, mg/dL or mmol/L.
2. Enter the fasting insulin from the same blood draw and pick its unit, uIU/mL or pmol/L.
3. Read HOMA-IR and QUICKI, then check the fasting glucose range and the working line under the result.

A fasting insulin index is a number calculated from fasting glucose and fasting insulin that estimates how strongly the body resists or responds to insulin. The two best known are HOMA-IR and QUICKI.

Insulin resistance is a condition in which the body does not respond to insulin the way it should, as the NIDDK describes it. The pancreas then has to release more insulin to hold blood glucose steady. A fasting index tries to capture that pattern from one blood sample.

HOMA-IR comes from Matthews and colleagues at Oxford, published in Diabetologia in 1985. QUICKI comes from Katz and colleagues at the US National Institutes of Health, published in 2000.

Both were built for research. The NIDDK states that health care professionals may not test for insulin resistance, and that the test is used mainly in research studies. A fasting index is therefore a talking point for a clinic visit. It is not a test result that stands alone.

Both indices use the product of fasting glucose and fasting insulin. HOMA-IR divides that product by a constant. QUICKI takes the logarithm of the product and inverts it.

1. Convert glucose to mmol/L when it is in mg/dL: divide by 18.0156. Glucose has a molar mass of 180.156 g/mol.
2. Convert insulin to uIU/mL when it is in pmol/L: divide by 6.00.
3. Multiply glucose in mmol/L by insulin in uIU/mL and divide by 22.5 to get HOMA-IR.
4. Add the base 10 logs of insulin in uIU/mL and glucose in mg/dL, then take 1 divided by that sum to get QUICKI.

Many texts write HOMA-IR as glucose in mg/dL times insulin, divided by 405. That is the same formula with 18 used as the glucose factor, since 22.5 x 18 = 405. The moles calculator shows how a molar mass links grams to moles.

Glucose 90 mg/dL, insulin 10 uIU/mL. Glucose is 90 / 18.0156 = 5.00 mmol/L.

HOMA-IR is 5.00 x 10 / 22.5 = 2.22. QUICKI is 1 / (log10 10 + log10 90) = 1 / (1 + 1.954) = 0.338.

The glucose of 90 mg/dL sits in the normal fasting range of the NIDDK table. The index adds the insulin side of the picture. Whether 2.22 matters for you is a question for your clinician, because no agreed cut-off exists.

SI units. Glucose 5.4 mmol/L and insulin 72 pmol/L. Insulin is 72 / 6 = 12 uIU/mL, so HOMA-IR is 5.4 x 12 / 22.5 = 2.88 and QUICKI is 0.326.

The Insulin Assay

Laboratories measure insulin with different assays, and they do not all agree. Two labs can give different HOMA-IR values for the same person. Compare results from the same lab where you can.

The Insulin Unit Factor

Knopp and colleagues (2019) showed that the correct factor for human insulin is 1 uIU/mL = 6.00 pmol/L. Many online tools use 6.945, which the same paper lists as a common incorrect factor. It makes a pmol/L result come out about 13.6 percent too low.

How Long You Fasted

Both formulas assume a true fasting sample. The NIDDK describes a fasting glucose test as taken after at least 8 hours without food. Food, or a short fast, raises both numbers.

Insulin Medicine

Injected insulin changes the measured insulin level. MedlinePlus notes that insulin medicines may affect the accuracy of an insulin blood test. The indices are not meant for that case.

Day to Day Variation

Matthews and colleagues reported a coefficient of variation of 31 percent for the insulin resistance estimate. One reading is a rough estimate. The last result card shows that spread.

HOMA-IR and QUICKI use the same two inputs and rank people in the same order. They differ in scale and direction.

Katz and colleagues reported average QUICKI values of 0.382 in their nonobese group, 0.331 in their obese group and 0.304 in their group with diabetes. These are group averages from one study of 56 adults. They are not diagnostic cut-offs.

Average glucose over months is a different question, answered by the HbA1c calculator.

Checking a Figure on a Lab Report

Some reports print HOMA-IR. Enter the glucose and insulin to see how the lab reached its figure and which insulin factor it used.

Converting Units Before a Visit

A report in mmol/L and pmol/L can be hard to compare with material written in mg/dL and uIU/mL. The result cards show both.

Following a Research Paper

Wallace, Levy and Matthews (2004) describe cohort and epidemiological studies as the appropriate use of HOMA. The calculator lets you reproduce a paper's worked figures.

Preparing Questions for a Clinician

Bring the inputs and the result. Body size also shapes the conversation, and the BMI calculator gives that figure.

1. Mixing Glucose Units

Putting 90 mg/dL into a formula that expects mmol/L gives a HOMA-IR 18 times too high. Always check the unit selector.

2. Using 6.945 for Insulin

Dividing pmol/L by 6.945 turns 60 pmol/L into 8.64 uIU/mL. The correct value is 10 uIU/mL.

3. Using a Non-fasting Sample

The model describes the steady fasting state. A sample taken after food does not fit it.

4. Reading One Cut-off as Universal

Published cut-offs differ by study, country, age group and insulin assay. A number that flags risk in one study does not carry over to everyone.

5. Treating the Index as a Diagnosis

Prediabetes and diabetes are diagnosed with glucose-based tests such as fasting plasma glucose and A1C, as the NIDDK explains. A fasting index does not replace them.

6. Using Natural Logs for QUICKI

QUICKI uses base 10 logarithms. Natural logs give a much smaller, wrong figure.

The arithmetic is exact, but HOMA-IR and QUICKI are estimates of insulin resistance with wide error bars. Matthews and colleagues wrote that the low precision of the model limits its use.

This page publishes no HOMA-IR or QUICKI threshold, because no single agreed cut-off exists. The fasting glucose range shown is the one the NIDDK gives, sourced from the American Diabetes Association, for people who are not pregnant: 99 mg/dL or below is normal, 100 to 125 mg/dL is prediabetes and 126 mg/dL or above is diabetes. Pregnancy uses different cut-offs.

HOMA-IR equals fasting glucose in mmol/L times fasting insulin in uIU/mL, divided by 22.5. With glucose in mg/dL, the common form divides the product by 405.

No single normal value is agreed. Published cut-offs differ by study population and insulin assay, so ask your clinician how your laboratory reads the figure.

QUICKI equals 1 divided by the sum of the base 10 log of fasting insulin in uIU/mL and the base 10 log of fasting glucose in mg/dL.

Divide pmol/L by 6.00. So 72 pmol/L is 12 uIU/mL. The older factor of 6.945 is listed as incorrect for human insulin by Knopp and colleagues (2019).

Divide mg/dL by 18.0156. So 90 mg/dL is 5.00 mmol/L and 126 mg/dL is 7.0 mmol/L.

No. HOMA-IR is a research estimate. Diabetes and prediabetes are diagnosed by a clinician with glucose-based tests such as fasting plasma glucose or A1C.

A higher QUICKI points toward greater insulin sensitivity. Katz and colleagues reported an average of 0.382 in their nonobese group and 0.304 in their group with diabetes.

Labs may use the 405 shortcut, a different insulin unit factor or the HOMA2 computer model. Each choice shifts the figure a little.

Yes. Both indices assume fasting values. The NIDDK describes a fasting glucose test as taken after at least 8 hours without food.

It is a rough estimate. The original 1985 paper reported a coefficient of variation of 31 percent for the insulin resistance estimate.

No. The calculator runs in your browser and nothing goes to our servers. Saved results stay in this browser only.

• HOMA Calculator: what the homeostasis model assessment estimates, from the unit that developed it (University of Oxford, Radcliffe Department of Medicine, Diabetes Trials Unit).
• Assessment of insulin sensitivity/resistance, with the HOMA-IR and QUICKI equations and the reported QUICKI group values (Gutch et al., Indian Journal of Endocrinology and Metabolism, 2015).
• Insulin units and conversion factors: 1 uIU/mL = 6.00 pmol/L (Knopp, Holder-Pearson and Chase, Journal of Diabetes Science and Technology, 2019).
• Diabetes tests and diagnosis, with the fasting plasma glucose table (National Institute of Diabetes and Digestive and Kidney Diseases).
• Insulin resistance and prediabetes (National Institute of Diabetes and Digestive and Kidney Diseases).