Insulin resistance is common even without obesity, and its popular test isn’t standard

Key takeaways

• In national survey data from 2015 to 2018, about 40% of U.S. adults ages 18 to 44 without diabetes had insulin resistance by a common HOMA-IR cutoff, and nearly half of those affected weren’t obese³.
• Insulin resistance usually causes no symptoms, and NIDDK says doctors may not test for it directly; the standard checks are blood sugar tests such as the A1C¹.
• HOMA-IR scores depend on the lab: German researchers found three insulin tests disagreed enough that each needs its own cutoff⁷.
• In a 2026 NIH study of Indigenous Americans, the popular triglyceride-to-HDL ratio matched a gold-standard insulin resistance test only weakly, far less closely than HOMA-IR or a glucose-drink test⁴.
• Dark, velvety skin on the neck, armpits or groin is a physical sign of insulin resistance worth a checkup, while darkening that appears suddenly, especially in a middle-aged or older adult, needs prompt evaluation²¹⁰.

Pick a cutoff, and you pick how many people have a problem.

When researchers analyzed national data on 3,429 U.S. teens ages 12 to 15, they deliberately chose a low HOMA-IR cutoff, 2.3, to cast a wide net⁸. By that line, about 6 in 10 counted as insulin resistant⁸. In Germany, a study of especially healthy adults put the top of the normal range at 2.35⁷.

HOMA-IR is a score built from a fasting blood sugar and a fasting insulin level⁴. It sounds exact. As those two studies show, the cutoff that turns a score into a label is a choice.

None of this means insulin resistance is a myth. It is real and common, and it can set the stage for type 2 diabetes, fatty liver disease and heart risk¹.

This article explains what insulin resistance is, who has it, how it’s measured in research and in a doctor’s office, and which signs are worth a checkup. You’ll also find a clue sheet that points to the standard tests, not a self-diagnosis quiz.

Insulin resistance means your pancreas is working overtime

Insulin is the hormone that helps move sugar from your blood into your cells. In insulin resistance, cells in your muscles, fat and liver respond poorly to it¹.

Over time, the pancreas may not make enough insulin to keep up. Type 2 diabetes develops when the body’s insulin falls short of what it takes to overcome that weakened response¹⁵.

Before that, blood sugar drifts into the prediabetes range. NIDDK, the National Institute of Diabetes and Digestive and Kidney Diseases, says people with prediabetes have a high chance of developing type 2 diabetes within 5 to 10 years¹.

The early strain can show up before blood sugar does. A loss of the fast first burst of insulin after a meal may be the earliest detectable sign that insulin-making cells are struggling, and it predicts type 2 diabetes, a Garvan Institute review notes¹¹.

Insulin resistance usually causes no symptoms¹. That’s why it so often goes unnoticed, and why the clues later in this article matter.

About 4 in 10 young adults have it, and half aren’t obese

The best-known estimate comes from researchers at the University of Alabama at Birmingham¹². They analyzed national survey data on 6,247 adults ages 18 to 44 without diabetes or heart disease, from 2007 to 2018³. They defined insulin resistance as a HOMA-IR of 2.5 or more³.

By that measure, 44.8% had insulin resistance in 2007–2010 and 40.3% in 2015–2018, a dip too small to rule out chance³. Rates differed by group in the latest years³:

• Mexican American adults: 53.8%, the highest
• Non-Hispanic Black adults: 37.9%
• Non-Hispanic white adults: 35.7%
• Men: 41.0%, down from 50.2%
• Women: about 40%, little changed

Insulin resistance traveled with other risks. High blood pressure showed up in 31.3% of young adults with it, compared with 14.7% of those without; high cholesterol, in 16.0% compared with 7.0%³.

The surprise was weight. Nearly half of the young adults with insulin resistance weren’t obese, and measures of body fat explained only a small share of the differences in HOMA-IR³. The authors suggested screening young adults regardless of body mass index, or BMI³.

The study has limits its authors named. It captured one moment in time, so it can’t show cause. HOMA-IR mostly reflects insulin resistance in the liver, and it may underestimate insulin resistance in Black adults³. The insulin tests used also changed over the survey years³.

A later article in this series looks more closely at body weight versus metabolic health.

How it’s measured, from a research clamp to a ratio on your lab report

The gold standard is a research procedure. In a clamp study, staff infuse insulin and glucose into a vein continuously, adjusting the glucose to hold blood sugar steady⁴. The more glucose your body soaks up, the more sensitive you are to insulin. A 2026 study by an NIH research branch in Phoenix called the method invasive, expensive and laborious⁴. NIDDK says the formal test is used mainly in research¹.

HOMA-IR is a stand-in. It needs only a fasting blood sample, measured for glucose and insulin⁴. But the insulin half of the formula is a problem. When an American Diabetes Association workgroup compared 12 commercial insulin tests in 2007, the spread in results between tests ranged from 12% to 66%, with a typical value of 24%⁶. Our companion article on insulin covers a 2025 follow-up.

Cutoffs move with the lab. The German Gutenberg Health Study found a median HOMA-IR of 1.09 in its healthiest group, with an upper limit of 2.35⁷. When the team compared three insulin tests, they agreed poorly, so the authors said each test needs its own reference range⁷. The studies in this article alone use 2.3 and 2.5³⁸.

Ratios from a cholesterol panel are rough. The 2026 NIH study, in Indigenous Americans in the Southwest, compared 18 shortcut scores with clamp results in 286 people⁴. On a scale where 1 means a perfect match and 0 means none, a glucose-drink-based score reached 0.69 and HOMA-IR 0.64⁴. The triglyceride-to-HDL ratio managed 0.36, and the triglyceride-glucose index 0.33⁴.

In the same study’s larger group of 2,260 adults, 509 developed type 2 diabetes⁴. Scores from a glucose-drink test predicted it best, and HOMA-IR edged out the best score that skips insulin⁴.

Why your doctor uses other numbers. When liver specialists rewrote the definition of fatty liver disease in 2023, they left fasting insulin and HOMA-IR out on purpose, citing complexity, cost and lab-to-lab variation⁹. They used five everyday measures instead: waist or BMI, blood sugar, blood pressure, triglycerides and HDL cholesterol⁹.

What raises your risk, from habits to genes to medicines

NIDDK lists these risk factors for insulin resistance and prediabetes¹:

• Body size: overweight, obesity or a large waist
• Age: 35 or older, though children and teens can develop it, and risk rises with age
• Family and ancestry: a family history of diabetes; African American, American Indian, Asian American, Hispanic or Latino, or Pacific Islander heritage
• Habits: physical inactivity; smoking, including e-cigarettes and secondhand smoke
• Pregnancy history: gestational diabetes or a baby weighing 9 pounds or more
• Other conditions: polycystic ovary syndrome (PCOS), Cushing’s syndrome, acromegaly, sleep apnea, and viral illnesses such as COVID-19
• Medicines: long-term glucocorticoids (steroids), some antipsychotics and some HIV medicines

Ancestry changes where lines are drawn. The 2023 fatty liver criteria use a lower BMI cutoff for people in Asia, 23 instead of 25⁹. Because NIDDK lists those medicines as risk factors¹, ask your prescriber whether your blood sugar should be checked. Don’t stop a medicine on your own.

Sleep matters, too. Our article on sleep and metabolism covers how short nights blunt insulin’s effect.

Fatty liver got a new name, and insulin resistance is at its core

In 2023, liver groups led by the American Association for the Study of Liver Diseases and the European Association for the Study of the Liver renamed nonalcoholic fatty liver disease⁹. It’s now metabolic dysfunction-associated steatotic liver disease, or MASLD⁹.

In the consensus survey, 61% of respondents called the word “nonalcoholic” stigmatizing, and 66% said the same of “fatty”⁹. The new criteria still captured 98% of people previously diagnosed in a European registry⁹.

The statement called insulin resistance the key metabolic problem underlying MASLD⁹. NIDDK also links insulin resistance to high blood pressure, metabolic syndrome, unhealthy cholesterol and triglyceride levels, and blood vessel changes¹. Our triglycerides article covers that number in depth, and a later article in this series covers metabolic syndrome.

The insulin resistance clue sheet

This sheet is our own, built from the sources in this article. It points to standard tests to discuss with a clinician. It is not a way to diagnose yourself.

Clue	Why it matters	Standard test to ask about
Dark, thick, velvety skin on the neck, armpits or groin	NIDDK calls it a physical sign of insulin resistance2; high insulin may spur skin-cell growth10	A1C or fasting blood sugar1
Skin tags in those same folds	Can appear alongside it and hint at a metabolic problem10	A1C or fasting blood sugar1
Waist over 94 cm (about 37 inches) for men or 80 cm (about 31.5 inches) for women	One of five cardiometabolic criteria9; large waist is a risk factor1	Blood sugar test and lipid panel1
Triglycerides 150 or higher, or HDL at or below 40 (men) or 50 (women)	Criteria chosen because they’re thought to reflect insulin resistance9	A1C or fasting blood sugar1
Blood pressure 130/85 or higher	Another of the five criteria9	Blood sugar test1
Fat seen in the liver on a scan	Insulin resistance is the core problem in MASLD9	Blood sugar test and lipid panel19
Past gestational diabetes, a 9-pound baby, or PCOS	Each raises risk1	Regular blood sugar testing1
Long-term steroids, some antipsychotics or HIV medicines	Each raises risk1	Ask your prescriber about blood sugar checks
Age 35 or older, with family history or higher-risk ancestry	Each raises risk1	Testing; if normal, repeat at least every 3 years1

Waist conversions are by our math. These waist lines come from the 2023 liver criteria; other checklists, including the one for metabolic syndrome, use different ones, so ask which your clinician uses. If a blood sugar test comes back in the prediabetes range, A1C 5.7% to 6.4% or fasting glucose 100 to 125 mg/dL, NIDDK advises testing for type 2 diabetes every year¹.

What helps, and how strong the evidence is

The levers are the ones natural-health readers already value: food, movement, sleep and a healthy weight. NIDDK lists healthy eating, physical activity, weight management and enough sleep¹.

In the Diabetes Prevention Program, people at high risk who lost 5% to 7% of their starting weight cut their chance of type 2 diabetes¹. For someone who weighs 200 pounds, by our math, that’s 10 to 14 pounds. The metformin pill also helped, though less than lifestyle changes, and follow-up found benefits lasting at least 15 years¹. Dark skin patches can improve with weight loss, diet and exercise, too¹⁰.

A sibling article in this series ranks the levers that improve insulin sensitivity, and a later article examines supplements sold for blood sugar. Skepticism is healthy in both directions: a test marketed to diagnose insulin resistance deserves the same scrutiny as a pill marketed to fix it.

Claim	Evidence level	What the research shows
Insulin resistance is common in young U.S. adults, including many without obesity	Strong evidence	National survey data; the exact share depends on the cutoff38
Insulin resistance raises the risk of type 2 diabetes	Strong evidence	Federal summaries and long cohort studies145
One HOMA-IR result gives a precise personal diagnosis	Evidence says no	Insulin tests disagree, and cutoffs vary67
A triglyceride-to-HDL ratio can stand in for a clamp test	Not enough evidence	Only a weak match in the one clamp comparison we reviewed, a 2026 NIH study of Indigenous Americans4
Dark, velvety skin in body folds signals insulin resistance	Strong evidence	Recognized by NIDDK and clinical references210
Modest weight loss and activity lower diabetes risk in high-risk people	Strong evidence	Diabetes Prevention Program and its follow-up1

Questions readers ask

The bottom line

Insulin resistance is real, common and often silent, and it can show up in people who aren’t overweight, years before blood sugar crosses into diabetes. But the scores used to label it shift with the lab and the cutoff, so a single HOMA-IR number online is a weak basis for a diagnosis or a purchase.

Use the clues, such as dark skin folds, a growing waist, high triglycerides, a pregnancy history or a family history, as reasons to ask for the standard tests. And put your effort into the habits with the strongest evidence: moving more, eating well, sleeping enough and, if you carry extra weight, losing a modest amount.

Related reading

• Insulin’s patent sold for $1, and then came the fights over price and carbs
• Short sleep blunts insulin and stokes appetite, and adding sleep cut calories
• Habits can halve your triglycerides, and very high levels can inflame the pancreas
• Why one flagged lab number is not a diagnosis
• When the company reading your labs also sells the fix

Sources

1. National Institute of Diabetes and Digestive and Kidney Diseases. “Insulin Resistance & Prediabetes.” NIH, reviewed March 2025. niddk.nih.gov
2. National Institute of Diabetes and Digestive and Kidney Diseases. “Acanthosis nigricans” (dictionary definition). NIH. niddk.nih.gov
3. Parcha V, Heindl B, Kalra R, Li P, Gower B, Arora G, Arora P. “Insulin Resistance and Cardiometabolic Risk Profile Among Nondiabetic American Young Adults: Insights From NHANES.” Journal of Clinical Endocrinology & Metabolism, 2022. pubmed.ncbi.nlm.nih.gov
4. Vazquez L, Vazquez Arreola E, Nagul M, Krakoff J, Hanson RL. “Comparing surrogate indexes for insulin resistance as predictors of type 2 diabetes.” Journal of Clinical Endocrinology & Metabolism, 2026. pubmed.ncbi.nlm.nih.gov
5. National Institute of Diabetes and Digestive and Kidney Diseases. “Story of discovery: how different medications for diabetes and obesity emerged from basic research on one pancreatic hormone.” NIH, January 27, 2021. niddk.nih.gov
6. Marcovina S, Bowsher RR, Miller WG, et al. “Standardization of insulin immunoassays: report of the American Diabetes Association Workgroup.” Clinical Chemistry, 2007 (University of Minnesota record). experts.umn.edu
7. Gutenberg Health Study researchers. “Distribution of HOMA-IR in a population-based cohort and proposal for reference intervals.” Clinical Chemistry and Laboratory Medicine, 2021. degruyterbrill.com
8. Zhuo Y, He J, Wang L, Chow CB, Ho FK. “Association between birth weight and insulin resistance in US adolescents: A retrospective cohort study exploring the role of concurrent body mass index.” Nutrition, Metabolism & Cardiovascular Diseases, 2024 (University of Glasgow repository). eprints.gla.ac.uk
9. Rinella ME, et al. (NAFLD Nomenclature consensus group). “A multisociety Delphi consensus statement on new fatty liver disease nomenclature.” Journal of Hepatology, 2023 (University of Birmingham repository). research.birmingham.ac.uk
10. Hughes EK, Brady MF, Rawla P. “Acanthosis Nigricans.” StatPearls, NCBI Bookshelf, updated August 11, 2023. ncbi.nlm.nih.gov
11. Cheng K, Andrikopoulos S, Gunton JE. “First phase insulin secretion and type 2 diabetes.” Garvan Institute of Medical Research open-access record. publications.gimr.garvan.org.au
12. University of Alabama at Birmingham. “UAB researchers find that 40 percent of young American adults have insulin resistance and cardiovascular risk factors.” UAB News, September 20, 2021. uab.edu