Is a fasting insulin of 12 normal?

Most laboratories would print it as normal. In the cohorts summarized here, 12 µIU/mL sits at or near the top quartile, where the risk of later diabetes and heart disease was highest. It is a reason to repeat the test, look at glucose, triglycerides and waist size alongside it, and talk to a physician, not a diagnosis.

What is an optimal fasting insulin level?

No guideline defines one. The best available evidence places the lowest-risk groups below about 8 µIU/mL, in the single digits, and shows risk rising from about 9 to 12 upward. That is a description of where risk was lowest, not a target proven by a trial.

Why does my fasting insulin change between labs?

Insulin assays are not standardized across manufacturers; the same sample can differ by 20 to 30 percent. Day-to-day biological variation adds another quarter. Compare results from the same laboratory and look at the trend.

Longevity Library · Lab reference

Written by Daniel Tagge, MD
Updated October 7, 2026
Reviewed October 10, 2026
13 sources · Measured in µIU/mL

System: Metabolism & Metabolomics

Fasting insulinSerum insulin, fasting · Basal insulin level

Name: Fasting insulin: normal range, what the evidence supports, and what moves it
Creator: Daniel Tagge, MD
Published: 2026-10-07

No guideline defines an optimal fasting insulin. In the cohorts that followed healthy adults for years, the lowest-risk groups sat in the single digits, and risk of later diabetes and heart disease rose from about 9 to 12 µIU/mL upward, well inside most laboratory reference ranges.

Fasting insulin is the hormone level your pancreas holds while you are not eating. It can climb for years while glucose stays normal. Laboratories call anything up to about 20 to 25 µIU/mL normal; the long cohorts put the lowest-risk groups in the single digits.

01 · The scale

Where the laboratory range and the evidence sit.

Risk rises continuously with fasting insulin. The lowest-risk groups in long cohorts sit below about 8 µIU/mL; the highest-risk fifth or quarter begins around 12 to 15. No trial has tested a target number.

051015202530

Common laboratory reference range
2 to 25 µIU/mL · Typical US laboratories
Lowest-risk quartiles and quintiles in cohorts
2 to 8 µIU/mL · Johnson 2010; Ausk 2010
Higher odds of prediabetes above 9
9 to 12 µIU/mL · Johnson 2010
Top quartile or quintile: higher diabetes, heart and mortality risk
12 and above µIU/mL · Dankner 2009; Ausk 2010; Després 1996

Laboratory range
Evidence supports
Guideline goal or unstudied
Higher risk in studies

Scale in µIU/mL. µIU/mL is the same as mIU/L. To convert to pmol/L, multiply by about 6.9. Bands show where studies found something, not a recommendation for any one person.

Watch

Fasting insulin, explained in person.

Your fasting insulin is “normal.” Here’s the range that matters. On YouTube

02 · What it is

What Fasting insulin measures.

Insulin moves glucose out of the blood and into cells. When muscle, liver and fat respond to it less well, the pancreas compensates by making more, so a rising fasting insulin is one of the earliest visible signs of insulin resistance, often years before glucose or hemoglobin A1c move.^1,2Grade B

Prolonged high insulin is not only a marker. Genetic studies that compare people born with higher or lower insulin production find more cardiovascular disease in the higher group, which argues that the hormone itself does harm over time.³Grade B

03 · The laboratory range

What a laboratory calls normal.

About 2 to 25 µIU/mL, depending on the laboratory.

A laboratory reference range describes the people the laboratory tested, most of whom were never followed to see what happened to them. It is a statistical interval, not an outcome threshold. For fasting insulin the interval also depends on the assay: when the American Diabetes Association compared ten commercial insulin assays, the same samples differed by a median of 24 percent from one assay to another.⁴Grade A

In a healthy adult population screened to exclude metabolic disease, the central 95 percent of fasting insulin values ran from about 2.6 to 14.6 µIU/mL, with a median near 7.7, noticeably tighter than the ranges most laboratories print.⁵Grade C

04 · The evidence

What the studies support.

In 515 healthy, normoglycemic Israeli adults followed for 24 years, the fifth with the highest baseline fasting insulin were the most likely to progress to prediabetes or diabetes, after adjusting for weight, blood pressure and triglycerides. Half the cohort developed some form of dysglycemia; the insulin measured decades earlier was the strongest predictor.¹Grade B

Among 965 endocrinology-clinic patients, mostly overweight, the odds of prediabetes doubled in the second quartile of fasting insulin (mean 8.0 µIU/mL), tripled in the third (mean 12.2) and were five times higher in the fourth (mean 25.9), compared with the lowest quartile (mean 4.9). A fasting insulin above 9.0 µIU/mL identified 80 percent of those with prediabetes. This is a cross-sectional study of a selected clinic population, so it describes association, not a threshold proven in the general public.⁶Grade C

In 5,511 US adults without diabetes from NHANES III, followed for a mean of 8.5 years, the top quartile of HOMA-IR (above 2.8, which at a fasting glucose of 90 mg/dL corresponds to a fasting insulin of about 12.6 µIU/mL) had a 64 percent higher all-cause mortality over follow-up than the bottom quartile (HOMA-IR 1.4 or less, about 6.3 µIU/mL at the same glucose). The association was confined to people with a body mass index under 25.⁷Grade B

In the Quebec Cardiovascular Study, men who went on to have ischemic heart disease had fasting insulin 18 percent higher at baseline than matched controls, and each standard deviation of insulin carried 1.7 times the odds of heart disease after adjusting for blood pressure, family history, triglycerides, ApoB, LDL and HDL cholesterol.⁸Grade B

In the Helsinki Policemen Study, 970 men free of heart disease were followed for 22 years. Those in the highest fifth of insulin response during a glucose tolerance test had 2.4 to 3.3 times the risk of a major coronary event in the first 5 to 10 years, independent of other risk factors; the excess faded by 22 years, as other risk factors caught up.⁹Grade B

05 · What moves it

The levers with evidence behind them.

What is listed here has been tested. The size of each effect, and the size of the study, are in the sentence, so a small effect reads as a small effect.

Weight loss of about 5 percent
lowers it
In a randomized study of adults with obesity, losing 5 percent of body weight improved insulin sensitivity in liver, muscle and fat tissue and lowered plasma insulin; further loss to 11 and 16 percent improved muscle insulin sensitivity further. The study was small (19 people reached the 5 percent loss, against 14 who held their weight) but used gold-standard clamp measurements.¹⁰Grade B
Exercise, including interval training
lowers insulin resistance
A meta-analysis of 50 studies found that high-intensity interval training reduced insulin resistance compared with no exercise (standardized mean difference -0.49) and with continuous moderate training (-0.35), with the largest fasting-glucose benefit in people at risk of or with type 2 diabetes.¹¹Grade A
Sleep
short sleep raises insulin resistance
When 11 healthy young men were restricted to 4 hours in bed for 6 nights, their glucose tolerance fell to a level resembling early diabetes, evening cortisol rose and sympathetic activity increased; all reversed with recovery sleep. A small, short laboratory study, but a direct demonstration that sleep debt alone shifts carbohydrate metabolism.¹²Grade C

06 · Reading it well

Caveats, and what belongs with a physician.

A single fasting insulin is noisy. Within one healthy person, fasting insulin varies by about 26 percent from day to day, so a value of 10 one week and 7 the next can be the same physiology. The trend across repeated fasting draws is more informative than any one result.¹³Grade B

Insulin assays are not harmonized. The same sample can read 20 to 30 percent differently on two laboratories' instruments, so a change between laboratories is not a change in you. Compare results from the same laboratory where possible.⁴Grade A

A normal fasting insulin does not exclude hyperinsulinemia after meals. In the Kraft database of 4,185 people with normal glucose tolerance, about half had a hyperinsulinemic pattern on a multi-hour insulin test, and the fasting value alone had limited ability to find them.²Grade C

See a physician

A fasting glucose of 100 mg/dL or higher, or a hemoglobin A1c of 5.7 percent or higher, on standard testing.
Increased thirst, frequent urination, blurred vision or unexplained weight loss.
Any insulin result during pregnancy, or while taking steroids or medications that affect blood sugar.
A very low fasting insulin with low blood sugar symptoms (shakiness, sweating, confusion when fasting).

This page is education, not individual medical advice, and reading it creates no physician-patient relationship.

07 · Open questions

What the literature does not settle.

No randomized trial has targeted a fasting insulin number and measured outcomes, so every threshold on this page is drawn from where risk rose in cohorts, not from a tested goal.
Whether lowering insulin itself, beyond the weight loss and glucose control that usually lower it, changes outcomes is unproven in humans.
The cohorts are mostly men, mostly of European or Israeli descent, and were measured decades ago on assays that no longer exist; the absolute numbers transfer imperfectly to today's laboratories.
Post-meal insulin testing may detect hyperinsulinemia that a fasting value misses, but there is no agreed protocol or threshold for routine use.

08 · In practice

How Dr. Tagge reads it in his own practice.

I like to see fasting insulin in the single digits, and I never read it alone: it sits next to glucose, triglycerides, waist size and the direction it has moved since the last draw. One high reading earns a repeat, not a label.

A practice judgment, labeled as one (Grade D): it is how one physician reads the number for the people he cares for, not a recommendation for you.

Questions

10 · Sources

13 sources, read in full.

Numbered in the order they appear. Each line says what this page relies on from the paper; the link opens the record at PubMed or the publisher.

1.
Dankner R, Chetrit A, Shanik MH, Raz I, Roth J. Basal-state hyperinsulinemia in healthy normoglycemic adults is predictive of type 2 diabetes over a 24-year follow-up: a preliminary report. Diabetes Care. 2009. PMID 19435961. DOI 10.2337/dc09-0153.
Prospective cohort · Highest fifth of fasting insulin in normoglycemic adults was the strongest independent predictor of dysglycemia 24 years later.
2.
Crofts C, Schofield G, Zinn C, Wheldon M, Kraft J. Identifying hyperinsulinaemia in the absence of impaired glucose tolerance: an examination of the Kraft database. Diabetes Research and Clinical Practice. 2016. PMID 27344544. DOI 10.1016/j.diabres.2016.06.007.
Cross-sectional study · About half of people with normal glucose tolerance showed hyperinsulinemia on dynamic testing; fasting insulin alone found few of them.
3.
Kolb H, Kempf K, Röhling M, Martin S. Insulin: too much of a good thing is bad. BMC Medicine. 2020. PMID 32819363. DOI 10.1186/s12916-020-01688-6.
Review · Reviews Mendelian randomization evidence that genetically higher insulin raises cardiovascular risk.
4.
Marcovina S, Bowsher RR, Miller WG, et al.. Standardization of insulin immunoassays: report of the American Diabetes Association Workgroup. Clinical Chemistry. 2007. PMID 17272483. DOI 10.1373/clinchem.2006.082214.
Laboratory method study · Among ten commercial insulin assays, between-assay variation had a median of 24 percent.
5.
Paudel P, et al.. Reference intervals for fasting insulin and insulin-related indices in healthy adults: a cross-sectional study. BMJ Open. 2026. PMID 41857869.
Cross-sectional study · In screened healthy adults the fasting-insulin reference interval was 2.63 to 14.56 µIU/mL, median 7.69.
6.
Johnson JL, Duick DS, Chui MA, Aldasouqi SA. Identifying prediabetes using fasting insulin levels. Endocrine Practice. 2010. PMID 19789156. DOI 10.4158/EP09031.OR.
Cross-sectional study · Odds of prediabetes rose across fasting-insulin quartiles (means 4.9, 8.0, 12.2, 25.9 µIU/mL); above 9.0 identified 80 percent of cases.
7.
Ausk KJ, Boyko EJ, Ioannou GN. Insulin resistance predicts mortality in nondiabetic individuals in the U.S.. Diabetes Care. 2010. PMID 20200308. DOI 10.2337/dc09-2110.
Prospective cohort · Top quartile of HOMA-IR (above 2.8) had 1.64 times the all-cause mortality of the bottom quartile (1.4 or less), in people with BMI under 25.
8.
Després JP, Lamarche B, Mauriège P, et al.. Hyperinsulinemia as an independent risk factor for ischemic heart disease. New England Journal of Medicine. 1996. PMID 8596596. DOI 10.1056/NEJM199604113341504.
Prospective cohort · Each standard deviation of fasting insulin carried 1.7 times the odds of ischemic heart disease in men, independent of lipids.
9.
Pyörälä M, Miettinen H, Laakso M, Pyörälä K. Hyperinsulinemia predicts coronary heart disease risk in healthy middle-aged men: the 22-year follow-up results of the Helsinki Policemen Study. Circulation. 1998. PMID 9714089. DOI 10.1161/01.cir.98.5.398.
Prospective cohort · Highest insulin quintile had 2.4 to 3.3 times the coronary risk over 5 to 10 years, fading by 22 years.
10.
Magkos F, Fraterrigo G, Yoshino J, et al.. Effects of moderate and subsequent progressive weight loss on metabolic function and adipose tissue biology in humans with obesity. Cell Metabolism. 2016. PMID 26916363. DOI 10.1016/j.cmet.2016.02.005.
Randomized trial · Five percent weight loss improved insulin sensitivity in liver, muscle and fat and lowered plasma insulin.
11.
Jelleyman C, Yates T, O'Donovan G, et al.. The effects of high-intensity interval training on glucose regulation and insulin resistance: a meta-analysis. Obesity Reviews. 2015. PMID 26481101. DOI 10.1111/obr.12317.
Systematic review and meta-analysis · Across 50 studies, interval training reduced insulin resistance versus control (SMD -0.49) and versus continuous training (-0.35).
12.
Spiegel K, Leproult R, Van Cauter E. Impact of sleep debt on metabolic and endocrine function. Lancet. 1999. PMID 10543671. DOI 10.1016/S0140-6736(99)01376-8.
Randomized trial · Six nights of 4 hours in bed lowered glucose tolerance and raised evening cortisol in healthy young men.
13.
Widjaja A, Morris RJ, Levy JC, Frayn KN, Manley SE, Turner RC. Within- and between-subject variation in commonly measured anthropometric and biochemical variables. Clinical Chemistry. 1999. PMID 10102917.
Laboratory method study · Within-person coefficient of variation for fasting insulin was 26 percent.

Keep reading.

Reading a number well is the start of the work, not the end of it. If you would like a physician to read yours over time, Explore working with Dr. Tagge.