A Normal HbA1c Can Hide a Decade of Insulin Resistance

Episode Summary

HbA1c turns up on almost every routine blood panel, and in this Monday mini episode Mikki looks at how much it can genuinely tell you. She starts with the physiology — haemoglobin picks up glucose non-enzymatically, red cells live 100 to 120 days, and the result is a weighted average of glucose exposure that leans heavily on the most recent month. That matters more than it first appears, because anything that shortens or lengthens red cell lifespan moves the number without blood glucose changing at all. Iron deficiency, heavy menstrual bleeding, foot strike haemolysis in runners, thalassaemia and sickle trait, recent blood donation and B12 deficiency can each push a result in one direction or the other.

From there Mikki works through the research-based critiques: insulin resistance and beta cell decline can precede an abnormal HbA1c by many years, the test is highly specific but only about 50 per cent sensitive against an oral glucose tolerance test, an average conceals the shape of the glucose curve, and both genetics and ordinary measurement noise move the number independently of glucose. She closes on the practical side — where HbA1c does earn its place, why she still thinks a result of 40 or 41 deserves a conversation with your doctor, and which markers are worth reading alongside it.

Key Topics

• What HbA1c actually measures. Glycation is non-enzymatic with no regulatory brake, so it proceeds in proportion to how much glucose is around and how long the cell is exposed. A red cell cannot repair it, making it a passive record of the preceding 8 to 12 weeks.
• The "three-month average" is weighted, not even. About 50 per cent of the movement toward a new value happens in the first month, which is why a result can shift over six weeks.
• Falsely low results. Haemolysis of any cause, haemoglobinopathies such as sickle trait and thalassaemia trait, ongoing blood loss, blood donation, recent transfusion, advanced liver disease, and starting iron treatment in someone iron deficient.
• Falsely high results. Untreated iron deficiency anaemia, B12 and folate deficiency, splenectomy or functional asplenia, and aplastic anaemia.
• Athletes get it from both directions. Foot strike haemolysis lowers A1c relative to true glucose exposure, while the iron deficiency common in female athletes raises it — and the two distortions can partly cancel out, so the number alone will not tell you which is happening.
• Correcting iron can drop A1c without any dietary change. Mikki's worked example is a woman going from 42 to 37 mmol/mol once her iron is sorted, and her concern is that the change gets credited to a nutrition intervention she never needed.
• The cut-offs, in New Zealand and US units. Roughly 30 to 39 mmol/mol normal, 41 mmol/mol pre-diabetes and 50 mmol/mol diabetes locally, against WHO and ADA thresholds of 42 and 48 mmol/mol. Type 2 diabetes throughout, not type 1.
• HbA1c is a lagging signal. Insulin markers can diverge from controls up to 20 years before diagnosis, and beta cell function is already down around 50 per cent by the time someone meets diagnostic criteria — so a normal HbA1c is compatible with a long stretch of compensated insulin resistance.
• Specific but not sensitive. Against an oral glucose tolerance test, sensitivity of about 50 per cent with specificity around 97 per cent, and screening work in which the HbA1c criteria missed a large majority of pre-diabetes.
• What to read alongside it. Fasting glucose, ideally fasting insulin, waist-to-height ratio, triglyceride-to-HDL ratio and a fasted lipid panel — plus your own historic trend rather than the population cut-off.

Chapters

00:00 How HbA1c works and what distorts it
03:41 Athletes, foot strike haemolysis and iron
06:23 New Zealand and US cut-offs
08:40 Why HbA1c lags insulin resistance
12:12 Averages, genetics and measurement noise
15:41 Your own trend, and ethnic differences
18:41 Where HbA1c earns its place
21:31 Sign-off and new recipes

Guest / Resources

Solo episode — no guest.

Referenced in the episode

• World Health Organization — pre-diabetes around 42 mmol/mol (6%), diabetes at 48 mmol/mol (6.5%)
• American Diabetes Association — diabetes at 48 mmol/mol (6.5%)

Mikki discusses findings from a number of published studies through the episode. They are described in the audio rather than cited by name, so no reference list is given here.

Tests and markers discussed

HbA1c · fasting glucose · fasting insulin · oral glucose tolerance test · continuous glucose monitor · HOMA2-IR · HOMA2-B · ferritin · fasted lipid panel · triglyceride-to-HDL ratio · waist-to-height ratio

Find Mikki

• Instagram, Threads and X — Mikki Williden
• Facebook — Mikki Williden Nutrition
• Website — mikkiwilliden.com — one-on-one calls and the recipe portal

Contact Mikki:
https://mikkiwilliden.com/
https://www.facebook.com/mikkiwillidennutrition
https://www.instagram.com/mikkiwilliden/
https://linktr.ee/mikkiwilliden
NZ listeners - save 10% off Calocurb by using the code Mikkipedia10 at www.calocurb.co.nz
Save 20% on all Nuzest Products WORLDWIDE with the code MIKKI at www.nuzest.co.nz, <a h