A High Fasting Glucose with a Normal HbA1c

The morning number is up while the three-month average is not. Two explanations cover nearly all of it: the fasting rise is genuine but the rest of the day is fine, so it does not move the average much; or the average is being pulled down by something to do with your red cells rather than your glucose.

The pattern on your report

  • Fasting glucose High · mild Key
  • HbA1c Normal Key
  • Hemoglobin Normal Key

Printed as: Fasting glucose in mmol/Lor mg/dL— An eighteenfold difference: 6.1 mmol/L is 110 mg/dL and 7.0 is 126.Hemoglobin 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%.

Why the numbers look like this

Overnight, blood glucose is set by the liver, which manufactures glucose steadily and is restrained from making too much by background insulin. When the liver becomes resistant to that restraint, the morning number climbs. Muscle uptake after meals is a separate mechanism, and while it still works, daytime excursions stay ordinary and the three-month average barely moves.

The dawn phenomenon sharpens this: growth hormone and cortisol pulse in the hours before waking and push the liver's output up further, so the fasting sample lands at the day's least flattering moment. One number ruled by hepatic physiology, one average ruled by everything else, and they can honestly disagree.

Not being flagged is not the same as normal

The two threshold schemes differ where this pattern lives. US practice draws impaired fasting glucose at 5.6 mmol/L, which is 100 mg/dL; UK guidance treats 5.5 to 6.9 mmol/L as the high-risk band. A fasting glucose of 7.0 mmol/L, 126 mg/dL, confirmed on a second occasion, is diabetes in both schemes. A result between those lines carries a different label depending on which side of the Atlantic interpreted it, which is a reason to ask which scheme your report used.

What else on the report can hide this

The blood count with a reticulocyte count decides whether the average can be trusted, because anything that shortens red cell life, hemolysis, recent bleeding, a transfusion, pulls the HbA1c down and makes it falsely reassuring. Triglycerides, a liver panel and a waist measurement fill in the hepatic insulin resistance picture that a genuine fasting rise usually belongs to. When the two tests still disagree after a properly fasted repeat, a glucose tolerance test asks the question neither of them can: what the hours after eating actually look like.

What usually causes it

Listed from most to least common — not from most to least serious.

  1. Very common

    Impaired fasting glucose with preserved post-meal control

    Hepatic insulin resistance appearing before anything else. The rest of the day is normal, so the average is barely affected.

  2. Very common

    An HbA1c reading falsely low

    Hemolysis, recent bleeding, a transfusion, pregnancy, or advanced kidney or liver disease. Here the glucose is right and the average is not.

  3. Common

    The sample was not truly fasting

    Eight hours with water only. Coffee with milk, a mint, or an early breakfast all invalidate it, and a repeat resolves it.

  4. Common

    The dawn phenomenon

    Overnight hormone release raises glucose towards morning. Exaggerated in insulin resistance and it affects the fasting number specifically.

  5. Common

    Stress, illness or steroids

    All raise fasting glucose, and steroids do so disproportionately. Recent courses count, not only current ones.

  6. Uncommon

    A hemoglobin variant interfering with the assay — in people of African, Mediterranean, Middle Eastern or Southeast Asian ancestry

    HbS, HbC, HbD or HbE trait. Whether it affects the result depends entirely on the laboratory's method.

  7. Uncommon

    Early type 1 diabetes or LADA

    A lean adult whose numbers are moving faster than insulin resistance would explain. GAD antibodies and C-peptide are the tests, and the HbA1c lags behind a rapid rise.

What is usually checked next

  • Full blood count with a reticulocyte count Identifies the red cell problems that make an HbA1c read falsely low, which is half the differential.
  • Repeat the fasting glucose properly Eight hours, water only. A surprising share of these resolve on a correctly taken sample.
  • Oral glucose tolerance test Shows what the rest of the day is doing, which neither the fasting number nor the average captures on its own.
  • Waist measurement, triglycerides and liver enzymes Establishes whether hepatic insulin resistance is the driver, which is what the treatment would address.
  • GAD antibodies and C-peptide, in a lean adult Separates autoimmune diabetes from insulin resistance, which changes the expected course entirely.

When to seek care sooner

  • Emergency Deep rapid breathing, vomiting, abdominal pain, or a sweet smell on the breath
  • Emergency Vomiting and abdominal pain on an SGLT2 inhibitor, even with a glucose that is not high
  • Same day Heavy thirst, passing large volumes of urine, blurred vision, or unexplained weight loss
  • Same day Rapid weight loss in a lean adult with rising glucose
  • Soon A fasting glucose of 7.0 mmol/L or above confirmed on a repeat

Questions worth bringing to your appointment

  1. Was I properly fasting for this sample?
  2. Which set of thresholds was this interpreted against?
  3. Could anything about my red cells be making the HbA1c read low?
  4. Would a glucose tolerance test tell us more than repeating either one?
  5. Given my build, is autoimmune diabetes worth excluding?

More from this panel

How to read blood sugar results →

Related patterns