Written by: Dr. Albana Greca Sejdini, MD, MMedSc
Medically reviewed by: Dr. Ruden Cakoni, MD, Endocrinologist
Last medically reviewed: July 2026
Insulin resistance means that muscle, liver, and fat cells do not respond to insulin as effectively as expected. The pancreas may compensate by producing more insulin for years, so fasting glucose and HbA1c can remain normal during early insulin resistance.
There are methods that directly measure insulin sensitivity, but no single inexpensive, standardized test is routinely used to diagnose insulin resistance in the same way that fasting glucose or HbA1c is used to diagnose diabetes.
The reason is not simply cost. Insulin resistance is a continuous biological characteristic rather than a yes-or-no condition, and different tissues can respond differently. For example, the liver may be insulin resistant while muscle insulin sensitivity differs. Results also vary with age, puberty, pregnancy, body composition, illness, medicines, ethnicity, laboratory insulin assays, and the method used.
In ordinary clinical care, the practical questions are often:
Enter fasting laboratory glucose and fasting insulin drawn at the same time. This calculator estimates HOMA-IR and QUICKI. It does not diagnose insulin resistance and should not be used with one random glucose reading.
Important limitations: HOMA-IR and QUICKI are surrogate estimates. There is no universally accepted HOMA-IR cutoff. Results depend on the insulin assay and population. They may be misleading during acute illness, pregnancy, severe hyperglycemia, after prolonged fasting, with major changes in insulin secretion, and in people using injected insulin. HOMA-IR is not appropriate for deciding medication or insulin doses.
The hyperinsulinemic-euglycemic clamp is widely regarded as the reference, or “gold-standard,” method for measuring insulin sensitivity.
During the procedure:
A person who is more insulin sensitive generally requires more infused glucose to maintain euglycemia during the standardized insulin infusion. A person who is more insulin resistant generally requires less.
The clamp is time-consuming, labor-intensive, costly, and technically demanding. It is mainly used in metabolic research and specialized clinical investigations rather than routine screening.
The old description that it “consists in measuring the blood insulin level” was incomplete. The procedure requires controlled insulin and glucose infusions with repeated glucose monitoring and protocol-specific calculations.
A fasting insulin test measures insulin concentration after an overnight fast. When glucose is still normal, a higher insulin level may suggest that the pancreas is compensating for reduced insulin sensitivity.
However, fasting insulin alone cannot diagnose insulin resistance because:
Be cautious with online claims that one narrow fasting-insulin range is “optimal” for everyone. A value must be interpreted with the laboratory method, simultaneous glucose, health history, and reason for testing.
The Homeostatic Model Assessment of Insulin Resistance estimates insulin resistance from fasting glucose and fasting insulin.
HOMA-IR = fasting insulin × fasting glucose (mg/dL) ÷ 405
or fasting insulin × fasting glucose (mmol/L) ÷ 22.5
A higher HOMA-IR generally suggests lower insulin sensitivity within the population and method being studied. It mainly reflects fasting, or basal, glucose-insulin physiology and is influenced strongly by hepatic insulin resistance.
There is no universal cutoff because proposed thresholds differ by:
HOMA2 is an updated computer model that accounts for some nonlinear aspects of glucose-insulin physiology. HOMA2 results are not interchangeable with the simple HOMA-IR formula.
The Quantitative Insulin Sensitivity Check Index also uses fasting glucose and insulin:
Lower QUICKI values generally indicate lower insulin sensitivity. Like HOMA-IR, QUICKI is a surrogate estimate and does not have one universal diagnostic threshold.
Fasting glucose and HbA1c do not directly measure insulin resistance. They identify whether the pancreas can still compensate sufficiently to keep glucose below prediabetes or diabetes thresholds.
| Test | Prediabetes | Diabetes |
|---|---|---|
| Fasting plasma glucose | 100–125 mg/dL | 126 mg/dL or higher |
| HbA1c | 5.7%–6.4% | 6.5% or higher |
| Two-hour 75-g OGTT glucose | 140–199 mg/dL | 200 mg/dL or higher |
A person may have insulin resistance with normal glucose tests because the pancreas is producing additional insulin. Conversely, abnormal glucose can occur for reasons other than ordinary type 2 insulin resistance, including type 1 diabetes, pancreatic disease, pregnancy, medication effects, or endocrine disease.
Without unequivocal hyperglycemia, a diabetes-range result generally requires confirmation. Read the blood tests for diabetes guide.
A standard 75-g oral glucose tolerance test measures plasma glucose before and two hours after a glucose drink. It is used to diagnose prediabetes and diabetes—not to directly diagnose insulin resistance.
Some specialists or researchers measure insulin at fasting and at several points during an OGTT. These values can help study insulin secretion and insulin sensitivity, but routine interpretation is difficult because:
An “insulin curve” should not be interpreted from one online chart. The complete pattern must be considered with glucose, symptoms, medicines, pregnancy status, and clinical context.
High triglycerides and low HDL cholesterol commonly occur with insulin resistance. They identify cardiometabolic risk but do not prove insulin resistance by themselves.
Increased abdominal or visceral fat is associated with insulin resistance. Body measurements are screening tools, not direct metabolic measurements.
High blood pressure often clusters with insulin resistance, abnormal lipids, and elevated glucose. Treating blood pressure is important regardless of whether a formal insulin-resistance estimate is calculated.
Fatty liver disease is strongly associated with insulin resistance. Normal ALT and AST do not exclude liver fat or fibrosis, and abnormal liver tests have many causes.
The TyG index is calculated from fasting triglycerides and glucose and is used increasingly in research as a surrogate marker. It is not a universally accepted diagnostic test for individual clinical care.
C-peptide estimates endogenous insulin secretion. It is useful in selected cases to help classify diabetes or assess beta-cell function, especially in people treated with insulin. It is not a direct test of insulin resistance.
Insulin resistance itself usually causes no specific symptoms. Many people feel well while the pancreas compensates with higher insulin production.
Possible clinical clues include:
Fatigue, hunger, cravings, brain fog, or sleepiness after meals are nonspecific and cannot diagnose insulin resistance.
Thirst, frequent urination, weight loss, or blurry vision may indicate that glucose has already risen into the diabetes range rather than proving early insulin resistance.
Learn more in What Is Insulin Resistance?
Follow the laboratory and clinician’s instructions. Common preparation includes:
A prolonged fast is not better and may change glucose and insulin physiology. Acute infection, sleep loss, pain, stress, and recent major diet changes can also affect results.
The clinician should interpret the complete picture rather than one score:
Management is generally based on risk and associated conditions rather than achieving one “perfect” fasting-insulin or HOMA-IR number.
Possible next steps include:
Insulin resistance may improve substantially, but no laboratory estimate can guarantee that type 2 diabetes or cardiovascular disease will never develop. Follow-up should focus on glucose, blood pressure, lipids, weight or waist trends, liver health, and the treatment goals relevant to you.
See insulin resistance treatment strategies and the prediabetes guide.
There is no universal normal range. Cutoffs vary by age, population, insulin assay, pregnancy status, and study method. Use the laboratory and clinician’s interpretation rather than one internet number.
Yes. The pancreas may compensate by making more insulin. Normal glucose does not prove normal insulin sensitivity.
Yes. Insulin secretion may be declining, the assay range may be broad, or resistance may not be captured accurately by a fasting measurement.
No. Diabetes is diagnosed with glucose or HbA1c criteria. High fasting insulin may suggest compensation, but it does not diagnose diabetes.
No. A meter measures glucose, not insulin sensitivity. Repeated patterns can identify abnormal glucose but cannot calculate a definitive insulin-resistance diagnosis.
A standard OGTT diagnoses impaired glucose tolerance or diabetes. It may indirectly reveal difficulty handling glucose, but it is not a direct measurement of insulin sensitivity.
Not necessarily. Testing is useful only when it will add information or change care beyond routine metabolic-risk evaluation.
Medical disclaimer: This calculator and article provide general education. They do not diagnose insulin resistance, prediabetes, diabetes, PCOS, fatty liver disease, or another condition. Do not start, stop, or change medicine based on fasting insulin, HOMA-IR, QUICKI, or an online calculation.