HOMA-IR: How to Interpret Your Result

The direct answer

A HOMA-IR number by itself does not tell you whether you have a disease. It tells you, roughly, how hard your pancreas is working to keep your fasting glucose normal. A low score means your cells respond efficiently to a small amount of insulin. A rising score means insulin is climbing to compensate for cells that respond less well, and that compensation itself carries downstream cardiometabolic risk, independent of whether fasting glucose has crossed into the prediabetes or diabetes range yet. The number is most useful as a trend tracked on the same lab's assay over time, not as a one-time verdict.

What HOMA-IR is and how it is calculated

HOMA-IR estimates insulin resistance from two fasting values: serum insulin and plasma glucose. The formula, introduced by Matthews and colleagues in a widely cited 1985 paper and still in routine use, is:

(fasting insulin in µIU/mL × fasting glucose in mmol/L) ÷ 22.5

Insulin normally allows cells to take up glucose from the blood. When cells become resistant to that signal, the pancreas secretes more insulin to hold glucose steady. Because insulin tends to rise before fasting glucose climbs into an abnormal range, HOMA-IR can flag compensated insulin resistance before a standard fasting glucose or HbA1c test does. That said, HOMA-IR was validated primarily against the hyperinsulinemic-euglycemic clamp, the reference-standard direct measure of insulin sensitivity, and correlation between the two methods is moderate, not perfect. It is a screening estimate, useful at a population level, with real noise at the individual level.

A later refinement (HOMA2) models variable insulin secretion and hepatic insulin clearance more precisely than the original linear formula, and is available as a standalone calculator from academic groups that maintain it. If your result is borderline and decisions hinge on precision, ask your clinician whether a HOMA2 calculation or a dynamic test (see below) is more appropriate than the basic HOMA-IR formula.

Unit conversion

Most US labs report glucose in mg/dL, not mmol/L. Divide mg/dL by 18.0 to convert. Example: a fasting glucose of 90 mg/dL equals 5.0 mmol/L. With a fasting insulin of 8 µIU/mL, HOMA-IR = (8 × 5.0) ÷ 22.5 = 1.78.

What counts as a normal HOMA-IR result?

The ranges below are the ones most commonly cited in clinical and research use. They are approximate, not fixed diagnostic cutoffs, and the exact threshold a given guideline uses can differ from what is shown here.

• Below 1.0, optimal insulin sensitivity, common in lean, physically active adults.
• 1.0 to 1.9, within the range most labs and studies treat as unremarkable for non-diabetic Western adults.
• 2.0 to 2.9, a range where risk for prediabetes and metabolic syndrome rises meaningfully; many clinicians begin lifestyle-focused conversations here.
• Roughly 2.9 and above, treated by some professional guidance, including position statements from clinical endocrinology societies, as an actionable threshold in non-diabetic adults, though the precise cut-point varies by document and should be confirmed against the current version of that guidance rather than assumed.

Population and lab context change what "normal" means. Reference distributions differ between countries, ethnic groups, and even between insulin assay platforms at different labs, because fasting insulin immunoassays are not standardized the way fasting glucose assays are. A result that looks unremarkable against a US reference population can still reflect more metabolic dysfunction than the same number would in a lower-baseline-risk cohort. Practically, this means: interpret your number against your own lab's reference range, and prioritize your own trend over time above a single population percentile.

Age and sex shift the baseline. HOMA-IR tends to rise with age, and postmenopausal women tend to run higher than premenopausal women at a similar body weight, consistent with estrogen's role in insulin sensitivity. Men typically run slightly higher than premenopausal women at comparable weight. None of this changes the interpretation framework above, but it means a same-lab, same-assay comparison across two different people is less informative than tracking one person over time.

What a high HOMA-IR means

A HOMA-IR at or above roughly 2.0, and especially above 2.9, means insulin is elevated relative to glucose, which reflects cells that are not responding efficiently to insulin's signal. The compensatory hyperinsulinemia that produces this pattern is not metabolically neutral; sustained high insulin levels are themselves associated with several downstream conditions.

Elevated HOMA-IR has been associated in observational and cohort studies with:

• Type 2 diabetes risk. Cohort studies have repeatedly shown that higher HOMA-IR predicts incident type 2 diabetes over multi-year follow-up, independent of fasting glucose alone.
• Cardiovascular disease markers. Large multi-ethnic cohort studies have found HOMA-IR associated with subclinical atherosclerosis after adjusting for traditional risk factors.
• Non-alcoholic fatty liver disease (NAFLD). HOMA-IR performs reasonably well as a screening discriminator between NAFLD and healthy controls in pooled analyses, though it is not a substitute for liver imaging or biopsy when NAFLD is suspected.
• Polycystic ovary syndrome (PCOS). Endocrine Society guidance identifies insulin resistance as a central driver of PCOS, and elevated HOMA-IR is common in affected women across a range of BMIs.
• Metabolic syndrome. The components used to define metabolic syndrome (elevated fasting glucose, elevated triglycerides, reduced HDL, abdominal obesity, elevated blood pressure) track closely with rising HOMA-IR, though HOMA-IR itself is not one of the formal diagnostic criteria.

Specific effect sizes reported for these associations vary substantially by study population, follow-up duration, and adjustment model. Treat any single percentage or hazard ratio you see quoted elsewhere as illustrative of a general pattern rather than a number that applies to your individual risk; verifying the exact figure against the primary paper is worthwhile before repeating it as fact.

When a high result needs prompt follow-up rather than routine follow-up

A HOMA-IR clearly above the high end of typical ranges (roughly 5 and above) in a non-diabetic adult, especially combined with a fasting glucose already above 100 mg/dL, is worth discussing with a clinician sooner rather than at a routine annual visit. Very high compensatory insulin can also, in some cases, mark a transition point where beta-cell function is starting to fail rather than continuing to compensate, which changes the urgency and the workup. This is a reason for timely evaluation, not a same-day emergency; go to urgent or emergency care instead for symptoms of acute hyperglycemia (excessive thirst, frequent urination, unexplained weight loss, confusion, or very high home glucose readings) or for any symptoms of diabetic ketoacidosis.

What a low HOMA-IR means

A HOMA-IR below roughly 1.0 generally reflects efficient insulin action: the pancreas does not need to overproduce insulin to keep glucose normal. This pattern is associated with lower cardiometabolic risk and is common in lean, physically active people.

A very low result is not always reassuring. In a symptomatic patient, a HOMA-IR below roughly 0.5, especially paired with a low fasting insulin (below about 3 µIU/mL) and symptoms suggesting hyperglycemia, can reflect inadequate insulin secretion rather than exceptional insulin sensitivity. This pattern deserves consideration of autoimmune diabetes in adults (sometimes called LADA), which is diagnosed through antibody testing (for example, GAD-65) rather than through HOMA-IR. HOMA-IR was not designed to detect insulin deficiency and should not be relied on to rule it out.

HOMA-IR after intentional weight loss

Weight loss, whether through lifestyle change or GLP-1 receptor agonist therapy (semaglutide, marketed as Ozempic or Wegovy; tirzepatide, marketed as Mounjaro or Zepbound), commonly lowers HOMA-IR as insulin sensitivity improves. The magnitude of improvement tracks with the amount of weight lost and varies by individual; a specific before-and-after number from a trial population should not be read as a guarantee of an equivalent personal result. Semaglutide and tirzepatide are FDA-approved for chronic weight management in adults with obesity, or with overweight and at least one weight-related comorbidity (current as of this writing; confirm current label status before relying on it, since indications and dosing have been updated over time).

An emerging, unresolved link: insulin resistance and mood symptoms

A 2025 sub-study within the GRADE trial cohort (adults with type 2 diabetes) examined how somatic versus cognitive-affective symptoms of depression relate to inflammation and insulin resistance, using both cross-sectional and longitudinal data (Diabetes and Emotional Distress Sub-Study of GRADE, PubMed). The findings point to differential associations, meaning somatic symptoms and cognitive-affective symptoms did not relate to insulin resistance and inflammation in the same way. This is observational evidence in a population that already has type 2 diabetes, so it does not establish that treating insulin resistance improves mood, or that mood symptoms alone signal insulin resistance in people without diabetes. It is worth knowing about if you have both insulin resistance and persistent depressive symptoms, as a reason to mention both to your clinician rather than treating them as unrelated, but it should not be over-extended into a general claim about HOMA-IR and mental health.

Evidence-based interventions that lower HOMA-IR

Multiple interventions have reduced HOMA-IR in controlled studies. The rough pattern across the literature is consistent even though exact effect sizes differ by study population, baseline HOMA-IR, diet composition, and follow-up duration:

• Aerobic exercise, sustained over months, lowers HOMA-IR in randomized trials of adults with insulin resistance or type 2 diabetes.
• Reduced-carbohydrate or lower-glycemic-load diets lower HOMA-IR in both randomized and non-randomized intervention studies, generally more so when combined with weight loss.
• Metformin, the best-studied pharmacologic agent for this purpose in non-diabetic adults with insulin resistance, reduced progression to type 2 diabetes in the landmark Diabetes Prevention Program and is commonly used off-label at lower insulin-resistance stages under specialist guidance.
• GLP-1 receptor agonists (semaglutide, liraglutide, tirzepatide) lower HOMA-IR largely through weight loss, with some evidence of direct pancreatic effects independent of weight change.
• Time-restricted eating and resistance training have shown HOMA-IR reductions in smaller trials; the evidence base for these is less mature than for aerobic exercise, standard dietary change, or metformin.
• Sleep restriction moves HOMA-IR in the wrong direction: even a single night of inadequate sleep has been shown to raise insulin resistance measures acutely in controlled sleep studies, and chronic short sleep is associated with meaningfully higher HOMA-IR in observational cohorts.

Where this article's earlier draft quoted specific numeric effect sizes (for example, an exact unit reduction from a named trial), those figures require verification against the primary literature before being repeated as fact; the general direction of each intervention above is well supported, but precise magnitudes vary enough by population that a single borrowed number can mislead an individual reader about what to expect.

A note on thyroid, testosterone, and estrogen

Thyroid dysfunction (including subclinical hypothyroidism), low testosterone in men, and the estrogen decline of menopause are all associated with higher HOMA-IR in observational studies, and correcting the underlying hormonal abnormality (levothyroxine for confirmed hypothyroidism, testosterone replacement for confirmed hypogonadism, hormone therapy decisions made with a clinician for menopause) has been reported to modestly improve HOMA-IR in some trials. These are associations worth raising with a clinician if your HOMA-IR does not fit your overall risk picture, not stand-alone reasons to start hormone therapy; hormone treatment decisions carry their own risk-benefit calculus that is separate from a HOMA-IR number.

Limitations you should know before acting on one result

HOMA-IR is a screening tool built from an assumption of steady-state fasting physiology. Several situations distort it in ways that have nothing to do with true insulin resistance:

• Exogenous insulin use. If you inject insulin, HOMA-IR is not interpretable, because injected insulin does not reflect endogenous secretion.
• Liver disease. Altered hepatic insulin clearance can inflate measured insulin and therefore HOMA-IR independent of true cellular resistance.
• Kidney disease. Reduced renal clearance of insulin can also falsely raise HOMA-IR.
• Acute illness or major stress. Infection, surgery, and other acute stressors raise cortisol and glucagon, temporarily elevating both glucose and insulin.
• Assay differences between labs. Fasting insulin immunoassays are not standardized across manufacturers the way glucose assays are, and cross-lab comparisons of the same numeric HOMA-IR result can represent different underlying physiology. Track your trend using the same lab and, where possible, the same assay platform.

HOMA-IR versus the glucose clamp

The hyperinsulinemic-euglycemic clamp remains the reference-standard direct measure of insulin sensitivity, but it is invasive, time-consuming, and impractical outside research settings. HOMA-IR correlates with clamp-derived measures well enough for population screening but not precisely enough to drive individual treatment decisions on its own. In ambiguous cases, a clinician may add a 2-hour oral glucose tolerance test with insulin measured at multiple time points, which captures dynamic insulin response in a way a single fasting value cannot.

Evidence boundary: what is established, what is plausible, what is not

Established: HOMA-IR is a validated, widely used screening calculation correlated with directly measured insulin sensitivity; higher values track with higher risk of type 2 diabetes, cardiovascular disease markers, NAFLD, and PCOS in observational cohorts; lifestyle change, weight loss, and metformin lower HOMA-IR in controlled studies; the test is not interpretable in patients on exogenous insulin.

Plausible but not firmly established for an individual reader: that a specific numeric improvement seen in a trial population will transfer at the same magnitude to any one person; that correcting subclinical thyroid or hormonal abnormalities will meaningfully change an individual's HOMA-IR; that the depression-symptom association described above generalizes beyond adults with type 2 diabetes.

Not established: a single universal HOMA-IR cutoff that applies identically across labs, assay platforms, ethnicities, and ages; that HOMA-IR alone is sufficient to diagnose or rule out prediabetes, diabetes, or insulin deficiency without other testing; that HOMA-IR should guide individualized medication dosing.

A decision framework for acting on your HOMA-IR result

This framework is meant to help you frame a conversation with a clinician, not to replace one. It assumes a fasting draw done correctly (see preparation notes below) and no medications or acute illness distorting the result.

Exceptions that override this table entirely: current insulin therapy (HOMA-IR not interpretable), significant liver or kidney disease (falsely elevated result likely), any acute illness or recent surgery within the past few weeks (retest once recovered), and results drawn after intense exercise the day before or after a non-fasted or afternoon draw (repeat under correct conditions before trusting the number).

Seek urgent or emergency care instead of waiting for a repeat lab if you have symptoms of very high blood sugar: confusion, rapid unintentional weight loss, persistent vomiting, fruity-smelling breath, or very high home glucose readings. HOMA-IR is not a tool for that situation.

How to prepare for an accurate test

• Fast 8 to 12 hours. Water is fine; coffee can raise cortisol and may shift fasting insulin.
• Have blood drawn in the morning where possible, since cortisol follows a diurnal pattern that can influence insulin.
• Avoid intense exercise the day before, since a recent hard workout can transiently lower fasting insulin.
• Tell your clinician about all current medications; corticosteroids, some antipsychotics, and certain other drugs raise fasting glucose or insulin independent of true insulin resistance.
• If tracking trends, try to use the same lab and, if you can find out, the same insulin assay method each time.