Overview
Quick answer
If a report came back with high blood sugar or the word prediabetes, the first job is to read the Australian stationery, not a US webpage. Fasting glucose is printed in mmol/L. HbA1c is printed in mmol/mol, often with a percentage beside it. Those two numbers are the same result in different units, not two diagnoses.
In Australian primary care, prediabetes is likely when fasting glucose is 6.1 to 6.9 mmol/L (impaired fasting glucose), when a 75 g oral glucose tolerance test (OGTT) two-hour value is 7.8 to under 11.1 mmol/L (impaired glucose tolerance), or when HbA1c is 42 to 47 mmol/mol (6.0% to 6.4%). Diabetes is a different line: fasting glucose 7.0 mmol/L or above, two-hour OGTT 11.1 mmol/L or above, or HbA1c 48 mmol/mol (6.5%) or above, confirmed on a second test in someone without symptoms.
The 2020 Australian prediabetes position statement, endorsed by the Australian Diabetes Society and four other colleges, says nearly one in six adults over 25 have prediabetes. That figure cites a 2012 Baker IDI report, not the 2022-24 ABS blood survey. A flagged line is common. Tracking the same markers over time is the useful next step, not a diet protocol from a search result.
Overview
Key takeaways
Australian reports use mmol/L for glucose and mmol/mol for HbA1c. A US page quoting 100 mg/dL or 5.7% is a different scale and a different cut-off.
Impaired fasting glucose in Australia is 6.1 to 6.9 mmol/L. The American Diabetes Association starts that band at 5.6 mmol/L. Do not mix the two.
The Australian HbA1c high-risk band is 42 to 47 mmol/mol (6.0% to 6.4%). The US prediabetes band starts at 39 mmol/mol (5.7%).
The Australian Diabetes Society's 2023 diagnosis guidance treats HbA1c of 48 mmol/mol (6.5%) or above as diabetes. It is more cautious about calling 42 to 47 mmol/mol "prediabetes" than the 2020 primary-care statement, which marks that band as prediabetes likely.
The three tests identify overlapping but different groups. Most people sit in one or two states, not all three. An OGTT is the only test that can name impaired glucose tolerance.
Fasting insulin and HOMA-IR are earlier metabolic signals. They are not Australian diagnostic criteria for prediabetes.
Hemexa does not run a 75 g OGTT. The product story is repeating HbA1c, fasting glucose, fasting insulin, and HOMA-IR on one record.
Blood cannot tell you whether you will progress to type 2 diabetes, and it cannot image the pancreas.
The frame
Why a high blood sugar result is not a diagnosis
Prediabetes is a lab description, not a disease name you can feel. The same word covers a fasting glucose of 6.2 mmol/L on a Heart Health Check, an HbA1c of 44 mmol/mol in someone who did not fast, impaired glucose tolerance that only an OGTT can see, and a person whose insulin has been high for a long time while glucose still looks ordinary.
That is why a good next step is to match the printed number to the Australian rule that produced it, then decide whether you have enough information. The page people join from is not "how to reverse prediabetes." It is what the blood can show, what it cannot, and which extra test is actually doing a job.
A flag is a review prompt, not a verdict
A result in the high-risk band means average or fasting glucose is higher than the usual laboratory cut-off. It does not mean you have type 2 diabetes, and it does not mean you will get it. The 2020 position statement still treats it as a reason to talk about risk, not as a lifelong label.
Colleges do not use the word the same way
The 2015 Medical Journal of Australia paper that implemented HbA1c diagnosis, still cited by the May 2023 ADS update, says labelling people just under 48 mmol/mol with prediabetes is not recommended, because the evidence for an HbA1c definition is thinner than for glucose. The 2020 primary-care statement, also ADS-endorsed, marks 42 to 47 mmol/mol as prediabetes likely. Both documents are current. The honest reading is "higher risk," not a fight about the noun.
RACGP treats it as a prevention window, not a fixed label
The RACGP handbook devotes a chapter to preventing progression to type 2 diabetes and recommends the same core interventions, movement, food, weight, smoking, blood pressure, regardless of which threshold someone has crossed. Its own rescreening rule is even more cautious with the HbA1c band than Bell: repeat annually below 5.7% (39 mmol/mol), or at six months for 5.7% to 6.4% (39 to 47 mmol/mol), the ADA band, not the 42 to 47 mmol/mol line the 2020 statement uses. The label is a sorting tool for testing and follow-up, not agreement on where risk starts.
If you came here hoping the report would tell you whether damage has started, this is the honest frame: these thresholds were set to mark future diabetes and cardiovascular risk, not to image an organ. A clear or borderline result still needs a clinician, not a webpage, to decide what happens next.
The tests
What the three tests measure
Australian guidance lets a GP use fasting venous glucose, HbA1c, or a 75 g OGTT to look for prediabetes. Each test watches a different part of glucose regulation, so they do not always agree. That disagreement is biology, not a lab error.
Fasting glucose
A venous sample after an overnight fast. It is a same-day snapshot of how much glucose is left in the blood when the liver is the main supplier. Impaired fasting glucose is closely tied to hepatic insulin resistance: the liver is releasing more glucose overnight than insulin can restrain.
HbA1c (glycated haemoglobin)
Glucose attached to haemoglobin inside red blood cells, accumulated over roughly two to three months. It does not need fasting. It cannot tell impaired fasting glucose from impaired glucose tolerance. Conditions that change red-cell lifespan can push it without a matching change in glucose.
Oral glucose tolerance test (OGTT)
A fasting venous sample, a 75 g glucose drink, then a two-hour venous sample. Impaired glucose tolerance is a two-hour value of 7.8 to under 11.1 mmol/L. It is the only test that names that state, and the 2020 statement says the clearest trial evidence for a structured lifestyle program sits here.
Finger-prick glucose in the rooms is a screen, not the diagnostic number. Point-of-care HbA1c devices in Australian general practice are funded for monitoring people already diagnosed with diabetes, not for diagnosing a new case. Diagnosis needs a laboratory result from an accredited pathology provider.
Reading your results
How to read an Australian glucose 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.
"Glucose" or "Fasting glucose"
In mmol/L. Impaired fasting glucose is 6.1 to 6.9. Diabetes is 7.0 or above. 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.
"HbA1c - NGSP" or a percentage
The older percentage printed for continuity. It describes the same glycation. 6.0% to 6.4% is 42 to 47 mmol/mol. 6.5% is 48 mmol/mol. Do not average the two or treat them as separate tests.
"Glucose 2 hr" or "OGTT 2 hour"
In mmol/L, only if a 75 g OGTT was done. Impaired glucose tolerance is 7.8 to under 11.1. Diabetes is 11.1 or above. A one-hour value, if printed, is not the Australian diagnostic line.
If the report only shows a percentage, or the two HbA1c units do not seem to match, ask the GP or the lab which figure is the diagnostic result. Do not convert it yourself from a US blog.
Australian practice
Australian diagnostic thresholds
These are the numbers the 2020 Australian primary-care statement uses. RACGP's own diabetes column matches exactly (7.0 mmol/L, 11.1 mmol/L, 48 mmol/mol); its printed high-risk HbA1c cut-off, used for setting a rescreening interval rather than the label, is narrower, at 5.7%. In someone without symptoms, a result in the diabetes column should be confirmed on a second test, on a separate day, before the diagnosis is written down.
| Test | High-risk / prediabetes band | Diabetes |
|---|---|---|
| Fasting glucose | 6.1 to 6.9 mmol/L (impaired fasting glucose) | 7.0 mmol/L or above |
| OGTT, 2-hour glucose | 7.8 to under 11.1 mmol/L (impaired glucose tolerance) | 11.1 mmol/L or above |
| HbA1c | 42 to 47 mmol/mol (6.0% to 6.4%) | 48 mmol/mol (6.5%) or above |
Australia uses a narrower HbA1c band than the United States
The American Diabetes Association starts prediabetes at 39 mmol/mol (5.7%). Bell et al. chose 42 to 47 mmol/mol (6.0% to 6.4%) because the evidence that intervening on an HbA1c-only definition helps is weaker, and because Canada and the UK use that narrower band. A result in the high 30s mmol/mol can look "prediabetic" on a US page and normal on an Australian report.
Australia uses the WHO fasting-glucose floor, not the ADA one
WHO and the International Diabetes Federation define impaired fasting glucose as 6.1 to 6.9 mmol/L. The ADA floor is 5.6 mmol/L. RACGP and Bell both use 6.1. A fasting glucose of 5.8 mmol/L is not impaired fasting glucose in Australian criteria, even if a US article says it is.
Diabetes still needs a confirmatory test when you have no symptoms
ADS May 2023 follows NHMRC: two results in the diabetes range, on separate days, in an asymptomatic person. Medicare will accept a single elevated diagnostic HbA1c (item 66841) as established diabetes and will not fund a second diagnostic HbA1c in the same 12 months. The Society itself calls that a conflict between practice guidelines and Medicare rules.
HbA1c cannot be used to diagnose diabetes in pregnancy. Pregnancy changes red-cell turnover. An OGTT is used instead. That exception belongs with a GP or obstetric team, not with a preventative membership page.
Mechanism
Why the three tests can disagree
The 2020 statement is explicit: each test identifies a slightly different group. Some people meet all three definitions. Most meet one or two. That is because the tests watch different tissues.
Impaired fasting glucose is mostly a liver story
Overnight, the liver releases glucose to keep the brain supplied. Insulin is supposed to turn that tap down. When the liver is insulin resistant, fasting glucose rises even if muscle still handles a meal reasonably well. A fasting sample can catch that. An HbA1c averaged across the whole day might not.
Impaired glucose tolerance is mostly a muscle story
After a glucose drink, muscle is the main sink. Peripheral insulin resistance plus beta cells that cannot raise insulin fast enough leaves the two-hour value high. Fasting glucose can still sit under 6.1. Only the OGTT names this state.
HbA1c is an average, so it can miss both
HbA1c accumulates on haemoglobin for as long as a red cell lives, weighted toward recent weeks. A person with high spikes and ordinary fasting values, or the reverse, can land in a different band from their glucose tests. Red-cell conditions move it again, in a direction that has nothing to do with insulin.
This is why the 2020 statement recommends an OGTT before referring someone into a structured, intensive lifestyle program: the trial evidence for that intensity is strongest in impaired glucose tolerance. People with impaired fasting glucose or a raised HbA1c alone still get general risk advice. They are not a lesser category. They are a different evidence base.
Earlier signal
Fasting insulin and HOMA-IR
Prediabetes sits on a continuum. Insulin resistance comes first. The pancreas compensates by releasing more insulin, which can keep fasting glucose and HbA1c inside the reference interval. When the beta cells can no longer compensate, glucose rises and the diagnostic labels appear. Fasting insulin and HOMA-IR watch that earlier phase. They do not replace the three diagnostic tests.
Fasting insulin is not an Australian diagnostic criterion
Bell, RACGP, and ADS diagnose on glucose and HbA1c. A high fasting insulin with a normal glucose is a metabolic signal, not a prediabetes diagnosis. Assays also vary between labs, which is one reason colleges have not adopted a single diagnostic insulin cut-off.
HOMA-IR combines the two fasting numbers
Australian labs calculate HOMA-IR as (fasting glucose in mmol/L × fasting insulin in mIU/L) ÷ 22.5. It is not a separate blood draw. It is not on a standard GP diabetes screen. It is useful when you already have both fasting results and want one index of how hard the pancreas is working.
A normal HbA1c does not close the metabolic file
HbA1c can sit below 42 mmol/mol while fasting insulin is already high. That is the sequence, not a claim about a fixed number of years. The marker guides for fasting insulin and HOMA-IR cover ranges, Medicare, and how to order. This page does not rebuild them.
If the question is "do I meet an Australian prediabetes definition," use glucose and HbA1c. If the question is "is the pancreas already compensating," add fasting insulin and HOMA-IR. Those are different questions.
For ranges, private costs, and how to order, see the fasting insulin guide and the HOMA-IR guide. For mmol/mol vs %, and when HbA1c misleads, see the HbA1c test guide.
Australia
How common is prediabetes?
Nearly one in six adults over 25, as of a 2012 estimate
Bell et al. 2020 write that nearly one in six Australian adults over 25 have prediabetes, more than two million people. The citation behind that sentence is Shaw and Tanamas, Diabetes: the silent pandemic and its impact on Australia (Baker IDI, 2012). It is the figure Australian colleges still quote. It is not a 2022-24 ABS blood-survey count.
About one in three progress within a decade, per the same statement
The 2020 position statement says that without intervention, about one in three people with prediabetes develop type 2 diabetes within ten years. That is the statement's figure, used to justify screening, not a single named trial result unpacked on this page.
Diagnosed diabetes is a different, newer number
The ABS National Health Measures Survey 2022-24, reported by AIHW, estimated 6.5% of Australian adults, about 1.3 million people, had diabetes. That survey is the current measured diabetes figure. It does not publish a matching prediabetes rate to replace Shaw and Tanamas.
Cardiovascular risk rises before the diabetes line
Bell cites a large meta-analysis in which prediabetes was associated with higher cardiovascular risk, and AusDiab, in which impaired fasting glucose independently predicted cardiovascular mortality. The point of catching the band is not only future diabetes. It is current risk that blood pressure, lipids, and smoking still drive.
Who to test
Who should consider testing?
Screening in Australia starts with risk, not with a shopping list of assays. AUSDRISK estimates five-year risk of type 2 diabetes from age, sex, ethnicity, family history, waist, blood pressure medicine, smoking, fruit and vegetable intake, and activity. A score of 12 or more is high risk.
AUSDRISK 12 or above, or a high-risk condition
Bell recommends pathology (fasting glucose, HbA1c, or OGTT) for intermediate (6 to 11) and high (≥12) AUSDRISK scores. ADS May 2023 uses HbA1c for diagnosis in high-risk asymptomatic people: AUSDRISK ≥12, a high-risk medical condition, or a high-risk ethnicity. High-risk conditions in that paper include cardiovascular disease, prior gestational diabetes, existing IFG or IGT, polycystic ovary syndrome, and antipsychotic medicine.
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, not a claim that every result will be abnormal.
A previous high result
Someone already in the high-risk band is usually retested annually. The 2020 statement supports annual HbA1c once prediabetes is identified, and Medicare funds a diagnostic HbA1c once every 12 months in eligible asymptomatic high-risk adults.
Symptoms that sound like new diabetes (thirst, passing a lot of urine, unexplained weight loss, blurred vision together) are not a preventative-testing question. ADS says those people need a glucose measurement, not an HbA1c used as a convenience test, because a rapidly rising glucose can still leave HbA1c behind.
Limits
What blood tests cannot show
A useful report answers a narrow question. Here is what these tests are not doing.
Whether you will progress
A result in the high-risk band raises the probability of type 2 diabetes. It does not date the event. Plenty of people stay in the band, and some return below it. Blood cannot tell you which group you are in.
Type 1 versus type 2, or damage already done
These assays measure glucose exposure. They do not type the diabetes, and they do not image eyes, kidneys, or nerves. Microvascular risk is why the diabetes threshold exists. Crossing a prediabetes line is not the same as having those complications.
Impaired fasting glucose versus impaired glucose tolerance, from HbA1c alone
HbA1c cannot split those states. Only an OGTT can. Hemexa does not run that test. If a structured program needs the distinction, that is a GP and laboratory OGTT, not a membership panel.
Diet quality, sleep, or activity
Nothing on the printout records what you ate, how you slept, or how much you moved. Those change the next result. They are not themselves blood tests.
Gestational diabetes
HbA1c is not used to diagnose diabetes in pregnancy. An OGTT is. A preventative HbA1c from before pregnancy does not replace antenatal screening.
Very high glucose with confusion, vomiting, or drowsiness is an emergency-care question, not a prediabetes-guide question.
How to order
How Australians get these tests
| Approach | Best for | Typical cost | What it covers |
|---|---|---|---|
| GP-ordered Medicare diagnostic HbA1c | Asymptomatic adults at high risk (for example AUSDRISK ≥12) being tested for a new diagnosis | Often bulk-billed or low gap when eligible. MBS item 66841, once per 12 months. | Laboratory HbA1c for diagnosis. Not a confirmatory second diagnostic HbA1c in the same year. |
| GP-ordered fasting glucose, with or without OGTT | A first flagged result, symptoms, or when IFG versus IGT would change the next step | Often bulk-billed or low gap when clinically indicated | Venous fasting glucose, and a 75 g OGTT if the GP requests it. Not fasting insulin. |
| GP-ordered Medicare monitoring HbA1c | People already diagnosed with diabetes | Often bulk-billed or low gap when eligible. MBS item 66551, up to four times per 12 months. | Laboratory HbA1c for established diabetes. Not a prediabetes screen. |
| Private fasting insulin, then HOMA-IR | A normal or high-normal glucose where the question is compensatory insulin, not the diagnostic label | Usually out of pocket. Medicare rarely funds insulin for routine screening in asymptomatic adults. | Fasting insulin. HOMA-IR is calculated if fasting glucose is on the same form. |
| Membership platforms (e.g. Hemexa) | Repeating HbA1c, fasting glucose, fasting insulin, and HOMA-IR 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 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
- What is prediabetes on an Australian blood test?
- It means glucose is higher than the usual cut-off and still below the diabetes line. In Australian primary care that is fasting glucose 6.1 to 6.9 mmol/L, an OGTT two-hour value of 7.8 to under 11.1 mmol/L, or HbA1c 42 to 47 mmol/mol (6.0% to 6.4%). It is a risk category, not a finding of organ damage.
- What HbA1c is prediabetes in Australia?
- The 2020 position statement treats 42 to 47 mmol/mol (6.0% to 6.4%) as prediabetes likely. The ADS 2023 diagnosis paper is more cautious about using the word prediabetes for that HbA1c band, and treats 48 mmol/mol (6.5%) or above as diabetes. Both are Australian college documents. A US page that starts prediabetes at 5.7% (39 mmol/mol) is using a different rule.
- What fasting glucose is prediabetes in Australia?
- Impaired fasting glucose is 6.1 to 6.9 mmol/L. Diabetes is 7.0 mmol/L or above. The American Diabetes Association starts the impaired band at 5.6 mmol/L. A result of 5.8 mmol/L is not impaired fasting glucose under Australian or WHO criteria.
- Do I need to fast for these tests?
- Fasting glucose and fasting insulin need an overnight fast, usually 8 to 12 hours, water only. HbA1c does not. An OGTT needs a fast beforehand and a two-hour appointment after the drink. If the same form includes lipids and glucose, you fast for the visit.
- Why does my report show two HbA1c numbers?
- Australian labs report HbA1c in IFCC mmol/mol, often with the older NGSP percentage beside it. They describe the same result. 48 mmol/mol and 6.5% are the diabetes threshold in both units. Do not average them.
- Can HbA1c and fasting glucose disagree?
- Yes. They measure different windows and different physiology. Impaired fasting glucose is largely hepatic. Impaired glucose tolerance is largely peripheral. HbA1c is a months-long average and can also move with anaemia, iron deficiency, kidney disease, or a recent transfusion. Disagreement is a reason to talk to your GP, not a reason to pick the more reassuring number.
- Does a high fasting insulin mean I have prediabetes?
- No. Australian diagnostic criteria use glucose and HbA1c. A high fasting insulin with a normal glucose is a signal that the pancreas may already be compensating. It is useful. It is not the label.
- How common is prediabetes in Australia?
- The 2020 college position statement says nearly one in six adults over 25, citing Shaw and Tanamas 2012. That is the figure Australian guidance still uses. The ABS 2022-24 survey published a diabetes rate (about 6.5% of adults), not a replacement prediabetes rate.
- How much do prediabetes blood tests cost in Australia?
- A GP-ordered diagnostic HbA1c is usually bulk-billed or low-cost when it meets Medicare item 66841 (once every 12 months in eligible asymptomatic high-risk adults). Monitoring HbA1c for established diabetes is item 66551, up to four times a year. Fasting glucose is often funded when clinically indicated. Fasting insulin is usually private. Memberships that include these markers on a panel of 76–80 signature markers start around AU$1,199 per year.
- Can I diagnose prediabetes with a finger-prick test?
- No. A rooms glucose is a screen. Point-of-care HbA1c in Australian general practice is funded for monitoring known diabetes, not for a new diagnosis. Diagnosis needs a laboratory result.
- Does Hemexa test for prediabetes?
- Hemexa includes HbA1c, fasting glucose, fasting insulin, and HOMA-IR 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 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 flagged glucose or HbA1c line is common. The useful product move is to put the diagnostic pair next to the earlier insulin pair on one record, then read the trend, rather than treating each PDF as a one-off scare.
HbA1c, glucose, insulin, and HOMA-IR twice a year
Hemexa includes HbA1c, fasting glucose, fasting insulin, and HOMA-IR 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
Live Hemexa range-policy rows, with no external source citation, currently treat fasting glucose of 4 to 5 mmol/L as the preventative target and 6.0 mmol/L as the watch ceiling; HbA1c at or below 39 mmol/mol as the preventative target and 41 mmol/mol as the watch ceiling; fasting insulin of 3 to 8 mIU/L as the target; and HOMA-IR at or below 1.5 as the target, with 2.8 as the watch ceiling. 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 result
An HbA1c of 48 mmol/mol or above, or a fasting glucose of 6.1 mmol/L 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. The app does not silently convert an imported HbA1c percentage into mmol/mol.
Collection through Healius Pathology
Healius Pathology is Hemexa's collection partner, with regional brands that differ by state. Hemexa does not run 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 ↗
Australian Diabetes Society. (2023). Guidance concerning the use of glycated haemoglobin for the diagnosis of diabetes mellitus (May 2023 update). View source ↗
d'Emden, M. C., Shaw, J. E., Jones, G. R., and Cheung, N. W. (2015). Guidance concerning the use of glycated haemoglobin (HbA1c) for the diagnosis of diabetes mellitus. Medical Journal of Australia, 203(2), 89-90. 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 ↗