Overview
Quick answer
If you searched insulin resistance, or "weight stuck," after a normal blood sugar result, the first job is to separate a mechanism from a diagnosis. Insulin resistance is how muscle, liver, and fat respond less well to insulin. It is not an Australian diagnosis. Australian colleges diagnose prediabetes and diabetes on fasting glucose, a 75 g oral glucose tolerance test (OGTT), or HbA1c. They do not diagnose insulin resistance from a fasting insulin or a HOMA-IR number.
What blood can show is the compensation. The 2020 Australian prediabetes position statement, endorsed by the Australian Diabetes Society and four other colleges, describes the sequence: insulin resistance is first counteracted by more insulin, so glucose can stay ordinary; when the beta cells can no longer compensate, glucose rises and the diagnostic labels appear. Fasting insulin and HOMA-IR watch that earlier phase. HbA1c can still sit inside the reference interval while they are already high.
The Whitehall II analysis (Tabak et al., Lancet 2009) is the primary human trajectory, not a year-count from a blog. People who later developed type 2 diabetes already had higher glucose and lower HOMA insulin sensitivity 13 years before diagnosis. The steep drop in HOMA insulin sensitivity sat in the last 5 years, and the steep glucose rise in the last 3 to 6. That is a sequence with measured windows, not a fixed multi-decade clock. Tracking the same fasting pair over time is the useful next step, not a diet protocol from a search result.
Overview
Key takeaways
Insulin resistance is a mechanism, not an Australian diagnostic label. RACGP and ADS diagnose on glucose and HbA1c.
Fasting insulin and HOMA-IR do not diagnose insulin resistance, prediabetes, or diabetes. They estimate how hard the pancreas is working while fasting.
HOMA-IR is calculated, not drawn: (fasting glucose in mmol/L × fasting insulin in mIU/L) ÷ 22.5. It needs both numbers from the same fasting sample.
HbA1c can look normal while fasting insulin is already high. That is the compensation described by Bell et al. 2020, not a lab error.
Whitehall II found metabolic differences 13 years before a diabetes diagnosis, with a steep HOMA-sensitivity drop in the last 5 years. It did not find a fixed multi-decade clock.
The research standard for insulin sensitivity is a hyperinsulinaemic-euglycaemic clamp. HOMA-IR is a fasting surrogate. Matthews 1985 reported a 31% coefficient of variation for the resistance estimate.
RACGP names acanthosis nigricans (velvety darkening on the neck or axillae) and dyslipidaemia as clinical evidence of insulin resistance when deciding who to screen. "Weight stuck" is a search query, not a criterion.
The 2023 PCOS guideline says clinically available insulin assays are not recommended in routine care. The glycaemic test it wants is an OGTT.
The frame
Why insulin resistance is not a diagnosis
Insulin resistance is a description of signalling, not a disease name you can feel. The same phrase covers a person with acanthosis nigricans and a high waist, a person whose fasting insulin is 14 mIU/L with a glucose of 5.2 mmol/L, a woman with PCOS whose guideline says not to order that insulin, and a search result that promises a reversal protocol.
That is why a good next step is to match the printed number to the job it can do. The page people join from is not "how to reverse insulin resistance." It is what the blood can show, what it cannot, and which extra test is actually doing a job.
Colleges diagnose glucose states, not insulin resistance
RACGP and the Australian Diabetes Society use fasting glucose, HbA1c, or a 75 g OGTT to name impaired fasting glucose, impaired glucose tolerance, prediabetes-likely HbA1c, or diabetes. Fasting insulin is not in those tables. AJGP 2021 is explicit: insulin is not recommended for routine screening for prediabetes or diabetes.
The position statement still needs the mechanism
Bell et al. 2020 open on insulin resistance and beta-cell failure as the continuum that produces prediabetes. They then diagnose on glucose and HbA1c. Both sentences are true. A high fasting insulin with a normal glucose is a metabolic signal. It is not the label.
A flag is a review prompt, not a verdict
A raised fasting insulin or HOMA-IR means the fasting pair is higher than a chosen cut-off. It does not mean you have type 2 diabetes, and it does not mean you will get it. Assays also vary between labs, which is one reason colleges have not adopted a single diagnostic insulin number.
If you came here hoping a HOMA-IR would name the feeling, this is the honest frame: these numbers were built to estimate fasting insulin action, not to image an organ or explain weight. A clear or borderline result still needs a clinician, not a webpage, to decide what happens next.
Symptoms
What people feel, and what colleges name
Most insulin resistance has no specific symptom. People search "weight stuck," afternoon crashes, and sugar cravings. Australian college documents name a shorter list when they decide who is at high risk and should have glucose or HbA1c measured.
Acanthosis nigricans is the sign RACGP actually writes down
The RACGP handbook lists acanthosis nigricans as clinical evidence of insulin resistance: hyperpigmentation, usually with a velvety change, commonly on the neck and axillae. In overweight adults aged 18 to 40, that finding (or dyslipidaemia) is one of the reasons to screen with fasting glucose or HbA1c. The skin change is a prompt to test glucose. It is not itself a blood result.
Skin tags and central adiposity sit beside it
The same RACGP section names skin tags and central obesity (waist, waist-to-hip, waist-to-thigh) among the features that cluster with insulin resistance. They raise the probability that a glucose test is worth doing. They do not tell you the fasting insulin.
"Weight stuck" is a search, not a criterion
Difficulty losing weight is why many people land here. Blood can show whether the pancreas is already compensating. It cannot tell you why weight is hard to move, and it cannot prescribe a deficit. Treating a HOMA-IR as a weight-loss diagnosis is the version of this page that is most likely to be wrong.
PCOS is the other named cluster. The 2023 international guideline, NHMRC-approved and summarised for Australia in the MJA, treats insulin resistance as a pathophysiological factor and still tells clinicians not to use routine insulin assays. The test it wants for glycaemia is the OGTT. See the PCOS hub for that workup.
For the named-condition workup, see the PCOS blood tests guide.
The tests
What the blood tests actually measure
Four numbers get mixed together because they sit on the same pathway. They do different jobs. Only two of them can name an Australian glucose diagnosis.
Fasting insulin
A venous immunoassay after an overnight fast. It estimates how much insulin the pancreas is releasing to hold fasting glucose. The RCPA Manual (reviewed 2 January 2024) gives 4 to 10 mU/L after an 8 hour fast when plasma glucose is normal, and says the application includes insulin-resistant states. mU/L and mIU/L are the same unit. It does not diagnose prediabetes.
HOMA-IR
Homeostatic Model Assessment of Insulin Resistance. Australian labs calculate it as (fasting glucose in mmol/L × fasting insulin in mIU/L) ÷ 22.5, from Matthews et al. 1985. It is not a separate blood draw. It is a fasting estimate of insulin action, not an Australian diagnostic criterion.
Fasting glucose
A same-day snapshot after an overnight fast, in mmol/L. Impaired fasting glucose in Australia is 6.1 to 6.9 mmol/L. Diabetes is 7.0 mmol/L or above. This is one of the diagnostic tests. It is closely tied to hepatic insulin resistance: the liver releasing more glucose overnight than insulin can restrain.
HbA1c (glycated haemoglobin)
Glucose attached to haemoglobin over roughly two to three months, printed in mmol/mol. The Australian high-risk band is 42 to 47 mmol/mol. Diabetes is 48 mmol/mol (6.5%) or above. It does not need fasting. It cannot see compensatory hyperinsulinaemia while glucose is still held.
An OGTT is the only test that can name impaired glucose tolerance. Hemexa does not run one. If that distinction would change the next step, that is a GP and laboratory OGTT, not a membership panel.
Reading your results
How to read an Australian insulin printout
Australian pathology reports print each marker with the lab's own name, unit, and reference interval. Field names vary. Match the row and the unit, not a remembered US cut-off or a blog HOMA band.
"Insulin" or "Fasting insulin"
In mIU/L or mU/L. Some labs print pmol/L. The RCPA Manual cites 4 to 10 mU/L after an 8 hour fast with a normal plasma glucose. Many commercial reports use a wider upper limit, often around 20 to 25 mIU/L. Being inside that wider interval is not the same as the RCPA fasting band. Do not convert a US µU/mL blog cut-off and treat it as Australian consensus.
"HOMA-IR" or "HOMA INDEX"
Unitless. Calculated from the two fasting numbers: (glucose mmol/L × insulin mIU/L) ÷ 22.5. Some labs print it. Others return glucose and insulin only. There is no college diagnostic cut-off. A US page quoting /405 is using glucose in mg/dL.
"Glucose" or "Fasting glucose"
In mmol/L. Impaired fasting glucose is 6.1 to 6.9. A US page quoting 100 to 125 mg/dL is the American Diabetes Association band (about 5.6 to 6.9 mmol/L), which starts lower than the Australian and WHO rule.
"HbA1c" or "HbA1c - IFCC"
The primary Australian result, in mmol/mol. 42 to 47 is the high-risk band used by the 2020 position statement. 48 or above is the diabetes threshold in the May 2023 ADS guidance. A percentage beside it is the same result, not a second test.
RCPA also says the insulin sample must be separated and frozen promptly, and that a simultaneous glucose is required for interpretation. If the report has insulin without a fasting glucose, HOMA-IR cannot be calculated honestly. Ask the lab or the GP which figure is fasting.
Mechanism
Why HbA1c can still look normal
The useful claim is the order of events, not a countdown. Insulin action fails first. The pancreas compensates. Glucose and HbA1c move later, and not in everyone.
Why muscle can hide the problem from HbA1c
Insulin binds its receptor and, through IRS, PI3K, and AKT, moves GLUT4 transporters to the muscle-cell surface so glucose can leave the blood after a meal (Saltiel and Kahn, 2001). When that signalling is impaired, muscle takes up less glucose. The pancreas raises insulin to compensate. Fasting glucose and the months-long HbA1c average can still look ordinary.
Why the liver writes the fasting glucose
Overnight, the liver releases glucose to keep the brain supplied. Insulin is supposed to turn that tap down. Bell et al. tie impaired fasting glucose to hepatic insulin resistance. A fasting sample can catch that. An HbA1c averaged across the whole day might not.
What Whitehall II actually measured
Tabak et al. followed 6,538 British civil servants. People who later developed diabetes already differed 13 years before diagnosis: higher fasting and two-hour glucose, lower HOMA insulin sensitivity. HOMA sensitivity then fell steeply in the last 5 years. Fasting and two-hour glucose rose steeply in the last 3 to 6. The paper's own interpretation is "as much as 3 to 6 years" for those steep changes, against a 13-year backdrop of earlier difference. It is not a multi-decade rule, and it is not an HbA1c study.
A normal HbA1c does not close the metabolic file. It answers a different question: has average glucose been high enough, for long enough, to glycate haemoglobin into the Australian high-risk or diabetes band? Fasting insulin answers whether the pancreas is already working harder than the glucose number shows.
For mmol/mol versus %, and when HbA1c misleads, see the HbA1c test guide. For ranges and how to order the earlier pair, see the fasting insulin guide and the HOMA-IR guide. For the diagnostic glucose labels, see the prediabetes blood tests guide.
Limits of the index
Clamp versus HOMA-IR
If a rigorous reader asks "how do you actually measure insulin resistance," the honest answer is: not with a standard GP form.
The clamp is the research standard
The hyperinsulinaemic-euglycaemic clamp infuses insulin at a fixed rate and measures how much glucose must be given to hold blood glucose steady. More glucose needed means more sensitive tissue. It is a research and tertiary-care method. It is not an Australian primary-care test, and Hemexa does not run it.
HOMA-IR is a fasting model of that loop
Matthews et al. 1985 built HOMA from the idea that fasting glucose and insulin sit in a feedback loop. Their resistance estimate correlated with the euglycaemic clamp (Rs = 0.88) and with fasting insulin itself (Rs = 0.81). They also reported a 31% coefficient of variation for the resistance estimate. Correlation with a clamp is not the same as being a clamp.
Wallace 2004 is still the caution
Wallace, Levy, and Matthews later wrote that HOMA is useful in large studies and in people not treated with insulin, and that it should not be used in isolation for clinical decisions. Insulin immunoassays are not standardised across platforms. A HOMA-IR of 2.1 at one lab is not automatically the same physiology as 2.1 at another.
That assay problem is why the 2023 PCOS guideline says clinically available insulin assays have limited clinical relevance and are not recommended in routine care. It is also why this page will not print a single "you are insulin resistant above X" number as if colleges agreed on it.
Who to test
Who should consider testing?
Australian screening for type 2 diabetes starts with risk, then glucose or HbA1c. Fasting insulin is an extra question, usually private, for when the glucose pair looks ordinary and the metabolic history does not.
RACGP high-risk features, then glucose first
The handbook screens high-risk adults with fasting glucose or HbA1c. High risk includes overweight with age 40 or above, younger overweight with hypertension or clinical insulin resistance (acanthosis nigricans, dyslipidaemia), a first-degree relative with type 2 diabetes, prior cardiovascular disease, high-risk ethnicity, gestational diabetes, PCOS, antipsychotic medicine, or AUSDRISK of 12 or above.
A normal glucose where the question is compensation
Family history, central adiposity, fatty liver, PCOS, or a "weight stuck" search with a normal HbA1c is the setting where fasting insulin and HOMA-IR add a signal. They still do not replace the diagnostic pair.
Aboriginal and Torres Strait Islander adults
RACGP recommends opportunistic or annual glucose testing from age 18, rather than relying on AUSDRISK alone. That is a screening interval for glucose, not a claim that every result will be abnormal, and not a reason to skip insulin if a clinician wants the earlier pair.
Symptoms that sound like new diabetes (thirst, passing a lot of urine, unexplained weight loss, blurred vision together) are not a preventative-testing question. Those people need a glucose measurement, not a HOMA-IR used as a convenience test.
Fatty liver is one of the metabolic settings where the earlier insulin pair can add a signal. See the fatty liver blood tests guide.
Limits
What blood tests cannot show
A useful report answers a narrow question. Here is what these tests are not doing.
Whether you have an Australian diagnosis
Fasting insulin and HOMA-IR do not diagnose insulin resistance, prediabetes, or diabetes. The diagnostic tests are glucose and HbA1c, and an OGTT when impaired glucose tolerance is the question.
Why weight is stuck
Nothing on the printout records intake, sleep, medicines, or how much you moved. A high fasting insulin can sit beside weight that is hard to move. It does not explain the weight, and it does not license a protocol from a search result.
Tissue insulin sensitivity the way a clamp would
HOMA-IR estimates the fasting loop. It does not measure how much glucose muscle takes up under a fixed insulin infusion. It does not image liver fat. It does not replace a clamp, and Hemexa does not run one.
Whether you will progress to type 2 diabetes
A compensated high-insulin pattern raises the probability of later glucose rise. Whitehall II described trajectories in a cohort, not your date. Plenty of people stay compensated. Blood cannot tell you which group you are in.
PCOS, or which diet works
Insulin assays are not a PCOS diagnostic criterion and are not recommended in routine PCOS care. No number on this page selects a diet composition.
Very high glucose with confusion, vomiting, or drowsiness is an emergency-care question, not an insulin-resistance-guide question.
How to order
How Australians get these tests
| Approach | Best for | Typical cost | What it covers |
|---|---|---|---|
| GP-ordered Medicare glucose or HbA1c | High-risk screening, or a first flagged sugar result | Often bulk-billed or low gap when eligible. Diagnostic HbA1c is item 66841, once per 12 months. | Laboratory fasting glucose and/or HbA1c. Not fasting insulin. |
| GP-ordered insulin under item 66695 | A clinician who wants insulin for a named indication (for example hypoglycaemia workup, or a selected metabolic question) | Rebate available when the request meets the item. Fee listed around $30.50; 85% benefit about $25.95. Not a dedicated "insulin resistance" item. | Quantitation of insulin as one hormone on the 66695 list. AJGP 2021 notes this is not recommended for routine prediabetes screening. |
| Private fasting insulin, then HOMA-IR | A normal or high-normal glucose where the question is compensatory insulin | Usually out of pocket if the request does not meet a rebatable indication. Standalone private insulin is often quoted around $25 to $60. | Fasting insulin. HOMA-IR is calculated if fasting glucose is on the same form. |
| GP-ordered 75 g OGTT | When impaired glucose tolerance would change the next step, or in PCOS glycaemic assessment | Often bulk-billed or low gap when clinically indicated | Fasting and two-hour venous glucose after a 75 g drink. Not fasting insulin. Hemexa does not run this. |
| Membership platforms (e.g. Hemexa) | Repeating fasting insulin, HOMA-IR, fasting glucose, and HbA1c on one record | ~$1,199/year (full membership) | Those four markers on the annual baseline and the included six-month retest, among 76–80 signature markers. Not a clamp. Not an OGTT. |
Pathology in Australia requires an authorised request from a registered medical practitioner. Collection for Hemexa members is through Healius Pathology, with regional brands that differ by state. A request form is not a diagnosis, and a membership is not a substitute for a GP visit about thirst, weight loss, or a result already in the diabetes range.
FAQ
Frequently asked questions
- Is insulin resistance a diagnosis in Australia?
- No. It is a mechanism. RACGP and the Australian Diabetes Society diagnose prediabetes and diabetes on fasting glucose, a 75 g OGTT, or HbA1c. Fasting insulin and HOMA-IR are not in those diagnostic tables.
- What are the symptoms of insulin resistance?
- Most people have no specific symptom. RACGP names acanthosis nigricans (velvety darkening on the neck or axillae), skin tags, central obesity, and, in the right context, PCOS features. "Weight stuck," cravings, and post-meal fatigue are common search terms. They are not diagnostic criteria.
- Can I have insulin resistance if my HbA1c is normal?
- Yes. Bell et al. 2020 describe insulin resistance being counteracted by more insulin while glucose stays ordinary. HbA1c averages glycation over months. It cannot see that compensation. Fasting insulin and HOMA-IR are the earlier pair. They still do not diagnose you.
- How is HOMA-IR calculated in Australia?
- HOMA-IR = (fasting glucose in mmol/L × fasting insulin in mIU/L) ÷ 22.5, from Matthews et al. 1985. It is not a separate blood test. A US calculator that divides by 405 is using glucose in mg/dL.
- What is a normal fasting insulin in Australia?
- The RCPA Manual cites 4 to 10 mU/L after an 8 hour fast when plasma glucose is normal. Many lab reports use a wider upper limit, often around 20 to 25 mIU/L. Always read the interval on your report. There is no college diagnostic insulin cut-off for insulin resistance.
- Do I need to fast for these tests?
- Fasting insulin, fasting glucose, and HOMA-IR need an overnight fast, usually 8 to 12 hours, water only. HbA1c does not. If the same form includes lipids, you fast for the visit. Morning collection is standard.
- Does Medicare cover a fasting insulin test?
- Insulin sits on MBS item 66695, a hormone quantitation item, not a dedicated insulin-resistance screen. AJGP 2021 notes insulin is not recommended for routine prediabetes or diabetes screening, so many preventative requests are private. Diagnostic HbA1c, when eligible, is item 66841.
- How much do insulin resistance blood tests cost in Australia?
- A GP-ordered diagnostic HbA1c is usually bulk-billed or low-cost when it meets item 66841. Fasting glucose is often funded when clinically indicated. Insulin under item 66695 has a listed fee around $30.50 when the request meets the item; otherwise private insulin is often $25 to $60. Memberships that include these markers on a panel of 76–80 signature markers start around AU$1,199 per year.
- Should I get a fasting insulin test if I have PCOS?
- The 2023 international PCOS guideline says insulin resistance is a pathophysiological factor, and that clinically available insulin assays are of limited clinical relevance and are not recommended in routine care. The glycaemic test it wants is a 75 g OGTT. Talk to your GP about that, not about using HOMA-IR to diagnose PCOS.
- Does a high HOMA-IR mean I will get type 2 diabetes?
- No. It raises probability. Whitehall II showed people who later developed diabetes already differed 13 years earlier, with a steep change 3 to 6 years before diagnosis. Plenty of people do not progress. Blood cannot date the event.
- Does Hemexa test for insulin resistance?
- Hemexa includes fasting insulin, HOMA-IR, fasting glucose, and HbA1c on the annual baseline and the included six-month retest, among 76–80 signature markers. Collection is through Healius Pathology. Hemexa does not run a hyperinsulinaemic-euglycaemic clamp or a 75 g OGTT. A membership does not replace a GP visit for a result already in the diabetes range.
How Hemexa fits
How Hemexa can help
A normal HbA1c with a metabolic history that still looks off is common. The useful product move is to put the earlier insulin pair next to the diagnostic glucose pair on one record, then read the trend, rather than treating each PDF as a one-off scare.
Insulin, HOMA-IR, glucose, and HbA1c twice a year
Hemexa includes fasting insulin, HOMA-IR, fasting glucose, and HbA1c on the annual baseline and again on the included six-month retest, part of 76–80 signature markers, in the Pancreatic Function & Glucose category. Those four are biannual because they move.
Dashboard bands are Hemexa policy, and HOMA-IR rows conflict
Live Hemexa range-policy rows, with no external source citation, currently treat fasting insulin of 3 to 8 mIU/L as the preventative target. HOMA-IR policy rows conflict (one set uses a target at or below 1.5 with a watch ceiling of 2.8; others use 1.0 / 1.9 or 1.5 / 2.5), so this page does not quote a single Hemexa optimal HOMA-IR as if it were settled. Fasting glucose of 4 to 5 mmol/L and HbA1c at or below 39 mmol/mol are the preventative targets on the rows that match the prediabetes hub. Those bands sit below the Australian diagnostic floors. They are company clinical policy, not ADS or RACGP treatment rules.
Clinical safety review on a diagnostic-range glucose
A fasting glucose of 6.1 mmol/L or above, or an HbA1c of 48 mmol/mol or above, is flagged in Hemexa's clinical safety review for follow-up with a doctor. Fasting glucose above 15 mmol/L is treated as urgent. There is no separate critical-value tier for fasting insulin or HOMA-IR.
Collection through Healius Pathology
Healius Pathology is Hemexa's collection partner, with regional brands that differ by state. Hemexa does not run a hyperinsulinaemic-euglycaemic clamp or a 75 g OGTT. Results from other Australian labs can be imported after your first baseline.
Sources
References
Bell, K., Shaw, J. E., Maple-Brown, L., et al. (2020). A position statement on screening and management of prediabetes in adults in primary care in Australia. Diabetes Research and Clinical Practice, 164, 108188. View source ↗
Australian Diabetes Society. (2020). A position statement on screening and management of prediabetes in adults in primary care in Australia (PDF, 30 April 2020). View source ↗
Royal Australian College of General Practitioners. Management of type 2 diabetes: a handbook for general practice. Defining and diagnosing type 2 diabetes (updated 14 November 2024). View source ↗
Tabak, A. G., Jokela, M., Akbaraly, T. N., Brunner, E. J., Kivimaki, M., and Witte, D. R. (2009). Trajectories of glycaemia, insulin sensitivity, and insulin secretion before diagnosis of type 2 diabetes: an analysis from the Whitehall II study. Lancet, 373(9682), 2215-2221. View source ↗
Matthews, D. R., Hosker, J. P., Rudenski, A. S., Naylor, B. A., Treacher, D. F., and Turner, R. C. (1985). Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia, 28(7), 412-419. View source ↗