What this test measures
Most diabetes screening measures the wrong end of the problem.
Insulin is the hormone that moves glucose out of the bloodstream and into cells. When cells respond to it poorly, the pancreas compensates by making more. Glucose stays normal — sometimes for a decade or more — because the system is being held there by escalating effort. A fasting glucose and an HbA1c both read the outcome. Fasting insulin reads the effort.
Fasting insulin measures how much insulin is circulating after an overnight fast, when it ought to be near its lowest. A raised value means your pancreas is working hard just to maintain a resting state.
HOMA-IR — Homeostatic Model Assessment of Insulin Resistance — combines the two numbers into one estimate: fasting insulin multiplied by fasting glucose, divided by 405. The logic is that resistance is best described by the relationship between what your pancreas is putting in and what your blood glucose is doing in response. High insulin holding a normal glucose scores higher than the same glucose held by a low insulin.
This matters more in India than in most places. South Asians develop insulin resistance at lower body weights, carry more fat around and inside the organs at the same BMI, and reach type 2 diabetes a decade earlier than European populations. The stage where insulin is raised and glucose is still normal is therefore both longer and more common here — and it is the stage where standard screening finds nothing.
One caution worth carrying into the results: insulin assays are not standardised. Two laboratories can return meaningfully different values on the same sample, which limits how much any single number deserves.
What your result means
Read against the range on your own report first, then read it with two things in mind.
The first is that the printed range for fasting insulin is unusually generous. Reference ranges are built from the population that walks into laboratories, and in India that population already includes a very large proportion of people with insulin resistance. A value of 18 µIU/mL can sit comfortably inside a printed range while describing a pancreas under real strain. Many metabolic clinicians read under 10 µIU/mL as unremarkable, 10 to 15 as worth attention alongside other findings, and above 15 as a clear signal.
The second is that HOMA-IR is an estimate derived from a mathematical model, not a measurement of anything. Thresholds around 2.0 to 2.5 are commonly quoted for Indian populations, and studies have proposed figures on either side of that. It is a useful way of holding two numbers together — not a line separating health from disease.
A normal HOMA-IR with a normal glucose is genuinely reassuring, though a single fasting sample says little about how you handle glucose after meals.
A raised HOMA-IR with a normal glucose and a normal HbA1c is the finding this test exists to catch. It describes compensation that is currently working. It is not diabetes, not prediabetes, and not a diagnosis — it is information, arriving early enough to be useful.
A raised HOMA-IR alongside a raised HbA1c adds little. Once glucose has moved, the more standard tests are already telling the story.
Certain situations distort it. Illness, physical stress, steroids and a poorly fasted sample all raise insulin. Late pregnancy raises it physiologically. In long-standing diabetes, a falling insulin can lower HOMA-IR while things are getting worse rather than better.
What each value means
Reference ranges differ between laboratories. Always read your result against the range printed on your own report.
Fasting insulin
2 – 10 µIU/mL
How much insulin your pancreas is releasing after an overnight fast. Laboratories often print a range reaching 20–25 µIU/mL, because that range was built from a general population in which insulin resistance is already widespread. Most metabolic clinicians read values under about 10 µIU/mL as unremarkable and treat the printed upper limit with caution.
Fasting plasma glucose
70 – 99 mg/dL
The other half of the calculation. It stays normal for years while insulin climbs to keep it there, which is exactly why a normal fasting glucose offers less reassurance than it appears to.
HOMA-IR
Reference range
0.5 – 2 index
- What it means
- Fasting insulin multiplied by fasting glucose, divided by 405 when glucose is in mg/dL. It is an estimate of how hard your pancreas is working to hold glucose steady. Cut-offs differ between populations and studies — values around 2.0 to 2.5 are commonly used as the upper limit in Indian cohorts, and some laboratories print a higher one.
- Why it matters
- A HOMA-IR of 3.4 (from a fasting insulin of 14.5 µIU/mL and a glucose of 95 mg/dL) describes a pancreas working roughly twice as hard as it needs to, with a glucose that still reads entirely normal. This is the stage where the underlying process is present and the standard tests are silent — and it is also the stage at which change works best.
- Next step
- Confirm on a second properly fasted sample, and read it alongside an HbA1c, a lipid profile, waist measurement, blood pressure and a liver assessment. HOMA-IR is a research-derived estimate, not a diagnostic threshold, and it earns its place only as part of that picture.
Interpret in context. One result doesn't tell the whole story.
Triglyceride to HDL ratio
0.5 – 3 ratio
Triglycerides divided by HDL, both in mg/dL. A cheap stand-in for insulin resistance that uses a lipid profile you have probably already had. A ratio above about 3 tracks with insulin resistance in most populations, though it performs less consistently in South Asians than in white European cohorts.
HbA1c
4 – 5.6 %
Average glucose over roughly three months. It is a later signal than insulin — by the time HbA1c moves out of range, insulin has usually been raised for years. It is included here because HOMA-IR is interpreted against it, not instead of it.
This is general education and cannot account for your history, your examination or your reports. If you need advice specific to your health, we’re always happy to see you in consultation.
What to do next
Repeat before concluding, on a properly fasted sample from the same laboratory. Biological variation in fasting insulin between two mornings is considerable, and comparing across laboratories adds assay variation on top of it.
Put it next to the rest of the picture. Waist circumference, blood pressure, a fasting lipid profile, an HbA1c and a liver assessment together tell you far more than HOMA-IR does alone. Insulin resistance rarely travels by itself: raised triglycerides, a low HDL, fatty liver on ultrasound and irregular cycles are the company it usually keeps.
Ask what is being watched. If the aim is to see whether the picture is improving, HOMA-IR can be repeated after three to six months. Repeating it sooner mostly measures the test's own variation.
Direct effort at the things with real evidence behind them. Resistance training at least twice a week, because muscle is where most glucose is disposed of. Reduced refined carbohydrate. Sustained weight reduction where it applies, particularly around the waist. Adequate sleep, since short sleep worsens insulin sensitivity within days. These work substantially better at this stage than they do later.
Discuss any medication question with your own doctor, with the full picture in front of them.
This information is educational and not a diagnosis.
Alitheau Learning Sessions
Sustainable Weight Loss
- Why the body defends a weight it has reached, and what that means for how fast to go
- Why BMI misreads South Asian bodies, and which measurements are more useful
- How insulin resistance, sleep, stress and thyroid function affect the picture
One result rarely tells the whole story
Have this read alongside everything else.
Conditions this result comes up in
Questions people ask
There is no single agreed figure. Values under 2.0 are widely treated as unremarkable and values above about 2.5 as suggestive of insulin resistance, but the cut-offs come from research cohorts rather than from a diagnostic standard, and they differ by population and by insulin assay. Read yours against the range your laboratory prints and against the rest of your results.
Fasting insulin in µIU/mL multiplied by fasting glucose in mg/dL, divided by 405. If your glucose is reported in mmol/L, divide by 22.5 instead. Both samples have to come from the same properly fasted blood draw for the figure to mean anything.
It is not part of a standard health check and is usually added deliberately — most often when PCOS, fatty liver, a strong family history of diabetes or unexplained weight gain is being assessed. Insulin assays are not standardised between laboratories, so comparing values across labs is unreliable.
Yes, in several ways. It estimates fasting insulin resistance, which is largely a liver phenomenon, and can look reassuring in someone whose muscle handles glucose poorly after meals. It is unreliable once someone is on insulin, in advanced diabetes when the pancreas is failing, and in pregnancy. A single value also carries considerable biological variation.