Normal A1c but High Fasting Insulin: What That Combination Means
A1c and fasting glucose report the outcome of glucose handling. Fasting insulin reports the effort it took.
Photo: Franco Debartolo / Unsplash
Yes, you can have meaningful insulin resistance while your A1c comes back in range. An A1c and a fasting glucose both report the outcome of glucose handling, meaning whether your blood sugar ended up where it should be. A fasting insulin reports the effort behind that result, which is how much insulin your pancreas had to release to get it there. The body can compensate for years before it shows on an A1c.
An in-range result is good news, and it isn't the whole answer. The diabetes hub gathers the rest of this series.
What an A1c and a fasting glucose are measuring
An A1c, or glycated hemoglobin, is the percentage of your hemoglobin that has glucose stuck to it. Hemoglobin is the oxygen-carrying protein inside red blood cells, and glucose attaches to it in proportion to how much is circulating. Red blood cells live a few months, so the test estimates your average glucose across roughly the previous two to three months. A fasting glucose is a different snapshot: your basal level after an overnight fast. What neither one is built to report is what it cost your body to keep that number there.
Can insulin resistance hide behind a normal A1c?
Yes, and the Whitehall II study measured how long it can go on. Tabák and colleagues followed 6,538 British civil servants who were free of diabetes at the start, publishing the analysis in The Lancet in 2009. Of those, 505 developed type 2 diabetes, and the researchers traced each person's measurements backwards from the date of diagnosis.
In the group who went on to develop diabetes, fasting glucose rose slowly and in a straight line from 13 years before diagnosis until 3 years before it. The move was from 5.47 to 5.79 mmol/L, or roughly 99 to 104 mg/dL. Both of those are ordinary results. Only in the last 3 years did the curve turn steep, reaching 7.40 mmol/L, about 133 mg/dL, at diagnosis.
The insulin side of those same people looked different. Across the long stretch from 13 to 5 years before diagnosis, their insulin sensitivity was already 34.2 percent lower than in people who never developed diabetes, and it stayed lower throughout. It was declining at the same rate as everyone else's. It had simply started from a worse place, and a steeper decline began only in the final 5 years. Calculated beta-cell function, which estimates how hard the insulin-producing cells are working, climbed between years 4 and 3 before diagnosis, from 85.0 to 92.6 percent, then fell to 62.4 percent. Those are percentages of a normal reference population: a short surge of extra effort, then a drop to roughly two-thirds of normal.
In their discussion, the authors described this shape as a long compensatory period, when insulin secretion increases to compensate insulin resistance without any major changes in glucose values. That compensation is the part I want you to hold onto. It's why a glucose panel can read normally for years while the pancreas works harder each year to hold it there.
A few limits temper this too. Whitehall II measured glucose instead of A1c, so it describes what happens underneath a normal glucose. The authors add in their limitations that HOMA uses fasting values, so it mostly describes insulin resistance in the liver and may date the beta-cell decline late. These are also population trajectories from a cohort that was mostly male and mostly white. No individual curve is forecast by them.
What is HOMA-IR, and why can't you read your own number against a cutoff?
HOMA-IR stands for homeostatic model assessment of insulin resistance. It isn't a separate blood test but a calculation, combining a fasting glucose and a fasting insulin from the same draw into a single index of insulin resistance. Researchers use it heavily, yet applying it to one person's result is harder than it looks.
Isokuortti and colleagues set out to define reference values for HOMA-IR, and in a 2017 paper in Diabetologia they did propose an upper limit of normal from two Finnish population cohorts. Then they tested how well such a number travels, sending samples from ten people to seven European laboratories for analysis. A value of 2.0 in the Helsinki laboratory came back as 1.3, 1.6, 1.8, 1.8, 2.0 and 2.1 in the other six. The inter-laboratory coefficient of variation was 25 percent, so the same blood produced meaningfully different answers depending on who ran it. The insulin measurement drove that, at 25 percent against the glucose at 5 percent. Isokuortti and colleagues wrote that standardization of insulin assays would be the first step toward defining normal values for HOMA-IR.
So a reference value exists, and it doesn't yet travel between laboratories. That's why you won't find a target fasting insulin or a HOMA-IR cutoff in this post: a number that moves that much isn't a line you can hold your own result against.
What else can make an A1c misleading?
An A1c and a fasting glucose disagree more often than you would expect. The NCD Risk Factor Collaboration pooled 117 population-based studies from 45 countries, covering 364,825 adults with both tests measured, and published it in Nature Medicine in 2023. Among those whose diabetes was found by screening, 29 percent had only a raised fasting glucose, 37 percent had only a raised A1c, and 34 percent had both. Roughly two in three were caught by one test and not the other.
That analysis asks how diabetes gets diagnosed at its thresholds. Detecting insulin resistance early is a different question. The narrower point still carries: two tests aimed at the same question don't classify the same people.
The authors also list non-glycemic factors that move A1c and fasting glucose differently. Anemia from iron deficiency or malaria, certain hemoglobin variants and other hemoglobinopathies, liver and kidney disease, and some medications all appear, and they note that guidelines recommend a glucose test for diagnosis in those people. Anemia is the overlap that reaches the most readers here, since an A1c is an average carried on red blood cells. If iron is part of your picture, what a ferritin result can and cannot tell you covers that side of it.
Where this leaves your next blood draw
This isn't an argument for ordering your own panel and reading it yourself. A fasting insulin isn't part of standard screening, its interpretation isn't standardized, and reference ranges differ between laboratories. If a normal A1c hasn't settled the question you brought, the clinician who orders your labs is the person to ask what a fasting insulin would add in your case.
The continuous glucose monitor post argues something similar about glucose data. Where insulin resistance is already established, where the evidence is strongest covers what to do about it.
Symptoms of frank high blood sugar need medical assessment now, not a blog post and a lab plan: unusual thirst, urinating far more than normal, unexplained weight loss, or a change in your vision.
If you want a more structured approach to metabolic health, the Wayfinder's Well membership includes the full educational library, and Beyond Blood Sugar, an upcoming course on diabetes, will arrive in three parts.
FAQ
Can your blood sugar be normal while your insulin is high?
Yes, and that pairing has a name: the compensated stage of insulin resistance. The pancreas releases more insulin to hold glucose in its usual range, so the glucose result stays ordinary while the effort behind it rises. In the Whitehall II cohort, people who later developed type 2 diabetes already had insulin sensitivity 34.2 percent lower than people who did not, across the years from 13 to 5 before diagnosis, while their fasting glucose was still reading between 5.47 and 5.79 mmol/L.
What does a fasting insulin measure that an A1c doesn't?
An A1c estimates your average glucose over roughly the previous two to three months, so it reports the result of glucose handling. A fasting insulin measures the hormone doing that handling, which is closer to the effort involved. Where the two separate, the insulin is generally the one that shifts first.
What is HOMA-IR, and is there a normal range?
HOMA-IR is an index of insulin resistance worked out from two fasting values, a glucose and an insulin, taken from the same blood draw. A reference value has been proposed in the literature, and it doesn't yet travel between laboratories. A 2017 Diabetologia study found that a HOMA-IR of 2.0 in one laboratory read anywhere from 1.3 to 2.1 in six others, with the variation coming from the insulin assay instead of the glucose, and the authors wrote that standardizing insulin assays would be the first step toward defining normal values.
Why would my A1c and my fasting glucose disagree?
They measure different things, so disagreement is common instead of a lab error. Across 117 population-based studies pooled by the NCD Risk Factor Collaboration, roughly two in three people whose diabetes was found by screening had only one of the two tests above the diagnostic threshold. An A1c can also be pushed around by anything that changes red blood cell turnover, including anemia and certain hemoglobin variants, which is why guidelines recommend a glucose test for diagnosis in people with those conditions.
Should I ask for a fasting insulin?
That's a question for the clinician who orders your labs, not one to settle from a blog post. It isn't part of standard diabetes screening, interpretation of it isn't standardized, and reference ranges vary between laboratories. What it can add is a view of the compensation that a normal A1c leaves out, which is most useful when a clinician is reading it beside your symptoms, your history, and the rest of your results.
References
- 1.
- 2.
- 3.
