An HbA1c Sitting Right on the Diagnostic Line
A single result at the threshold is not a diagnosis. Guidelines ask for confirmation on a second sample in anyone without symptoms, precisely because the test carries enough variation that a value at the line can land either side of it on a different day. Knowing that turns a frightening result into a defined next step.
The pattern on your report
- HbA1c High · mild Key
- Fasting glucose High-normal Key
- Hemoglobin Normal Key
- Ferritin Normal Key
Printed as: Fasting glucose in mmol/Lor mg/dL— An eighteenfold difference: 7.0 mmol/L is 126 mg/dL.Ferritin in ug/Lor ng/mLHemoglobin in g/Lor g/dLHbA1c in mmol/molor %— Not a simple multiplication. The conversion is (mmol/mol divided by 10.929) plus 2.15, so 48 mmol/mol is 6.5%. UK reports print mmol/mol, US reports the percentage.
Why the numbers look like this
HbA1c is a chemical measurement of how much glucose has attached to hemoglobin, and like every measurement it has a spread. Two components matter. Analytical variation comes from the machine, and modern assays are tight but not perfect. Biological variation comes from you: the same person on the same diet gives slightly different results week to week.
Add those together and a result sitting at the diagnostic cut-off has a real probability of falling below it when repeated, and a real probability of rising above.
Which is why the rule exists. In someone with classic symptoms a single abnormal result is enough, because the prior probability is already high. Without symptoms it is not, and confirming it is the difference between a diagnosis and a coin toss.
Not being flagged is not the same as normal
The line is 6.5%, or 48 mmol/mol, in both US and UK practice, and it is the same figure for a fasting glucose of 7.0 mmol/L. The threshold was chosen from the point at which the risk of diabetic eye disease begins to rise, so it marks a slope rather than a cliff: 6.4% and 6.6% are not two different conditions. They differ in which side of an administrative line they fall on, and that line still determines a great deal about monitoring and treatment.
What else on the report can hide this
Before accepting either result, check the red cells. Anything shortening their survival lowers HbA1c and anything lengthening it raises the figure, so a result near the line in someone with anemia, a recent transfusion or chronic kidney disease is better settled by a glucose-based test.
Iron deficiency is the one that pushes in the unhelpful direction here, because it raises HbA1c and can lift a borderline result across the line. Treating the iron and repeating is the correct order when ferritin is low.
If the two tests disagree, the guidance is to repeat the abnormal one, not to average them. A fasting glucose and an HbA1c measure different things, and using one to overrule the other misunderstands what each is for.
What usually causes it
Listed from most to least common — not from most to least serious.
- Very common
Genuine diabetes at an early stage
Confirmed by a second sample. Most people crossing the line for the first time have no symptoms at all, which is the reason screening exists.
- Very common
Ordinary measurement variation
A result within a few tenths of the line has a real chance of falling the other side on a repeat. This is why confirmation is required rather than optional.
- Common
Iron deficiency
Raises HbA1c independently of glucose and can lift a borderline value across the line. Check the ferritin, treat it, and repeat.
- Common
A recent course of steroids
Raises glucose for weeks and the HbA1c reflects it afterwards. Repeating once the effect has passed gives a truer baseline.
- Common
Prediabetes that has been drifting
The trend across previous years says more than the current value. A result that has climbed steadily is a different situation from one that has hovered.
- Uncommon
Anything lengthening red cell survival
Splenectomy or aplastic anemia. Cells live longer, accumulate more glucose, and the average overstates the glucose.
- Uncommon
A hemoglobin variant affecting the assay — in people of African, Mediterranean, Middle Eastern or Southeast Asian ancestry
Can push the result in either direction depending on which method the laboratory uses. The laboratory can say whether its assay is affected.
What is usually checked next
- A second HbA1c on a separate sample What guidelines ask for in someone without symptoms, and the step that turns a borderline number into an answer.
- Ferritin with a full blood count Iron deficiency raises HbA1c, and correcting it before repeating avoids a diagnosis made on a distorted number.
- Fasting glucose, or a glucose tolerance test Measures glucose directly, which is the right approach when anything makes the HbA1c unreliable.
- Your HbA1c results from previous years Distinguishes a value that has climbed steadily from one that has always sat close to the line.
- Waist measurement, blood pressure, lipids and liver enzymes Places the result in the metabolic picture, which drives what to do about it far more than the exact figure does.
When to seek care sooner
- Emergency Deep rapid breathing, vomiting, or a sweet smell on the breath
- Same day Heavy thirst, passing large volumes of urine, blurred vision, or unexplained weight loss
- Soon A confirmed HbA1c at or above 6.5%, which is 48 mmol/mol
- Soon New numbness or tingling in the feet, or a foot wound that is not healing
- Soon Pregnancy with a borderline result
Questions worth bringing to your appointment
- Does this need a second sample before it counts as a diagnosis?
- Has my ferritin been checked, since iron deficiency can lift the result?
- How does this compare with my HbA1c in previous years?
- Would a fasting glucose or tolerance test be more reliable for me?
- What would you want to see change before the next test?