Pre-diabetes: what the window really means

A category, not a disease

Pre‑diabetes is not a thing that has happened to you. It is a band on a scale — a blood sugar high enough to be above the reference population but below the line where diabetes is diagnosed. Somebody drew that line, and where they drew it is itself contested.

That does not make it meaningless. It makes it a probability statement rather than a diagnosis.

The two definitions, which disagree

The American Diabetes Association counts a fasting glucose of 100–125 mg/dL (5.6–6.9 mmol/L) or an HbA1c of 5.7–6.4%. The World Health Organization sets the fasting band higher, at 110–125 mg/dL (6.1–6.9), and has not endorsed an HbA1c range for it at all.

The practical consequence is not academic: a large group of people are pre‑diabetic under one definition and normal under the other. If you have been told you have it, it is worth knowing which criteria your laboratory used.

What the window actually means

The honest version has three parts, and clinics usually tell you only the first.

• Risk is genuinely raised. People in this band develop diabetes at a higher rate than those below it. That is consistent across populations.
• Most people in it do not progress. Over a given handful of years, the majority stay where they are or return to normal without treatment. Pre‑diabetes is not a waiting room with one exit.
• Regression is common. A meaningful proportion revert to normal glucose on repeat testing, partly through real change and partly through regression to the mean, which is why the repeat matters.

Anyone presenting this as an inevitable slide into diabetes is overstating it. Anyone presenting it as nothing is understating it.

The strongest evidence in metabolic medicine

This is the part worth taking seriously. The Diabetes Prevention Program — a large randomised trial in people with impaired glucose tolerance — compared an intensive lifestyle programme, metformin, and placebo.

The lifestyle arm outperformed the drug. Its targets were modest and specific: around 7% weight loss and about 150 minutes of moderate activity a week. Follow‑up studies have shown the benefit persisting for years after the programme ended.

It is unusual to have evidence this clear, and it points away from a prescription rather than towards one. A clinic that reaches for a medicine first in this band is not following the best evidence it has.

Why the number is not the whole picture

Two people with the same HbA1c of 6.0% can be in very different positions. One is 35, slim, with normal triglycerides and no family history. The other is 58, carrying weight around the middle, with raised triglycerides, low HDL and a raised ALT.

The second person's glucose is the least interesting thing about them. The surrounding markers — the ones on the same panel — say far more about what is coming than the glucose band does.

That is why pre‑diabetes is best read as one line in a panel rather than as a verdict.

What a sensible response looks like

1. Repeat it. One result in this band is not a finding. Same laboratory, same conditions, after enough time to mean something.
2. Look at the rest of the panel — triglycerides, HDL, ALT, waist, blood pressure. The company it keeps determines how much it matters.
3. Address what the evidence actually supports: modest weight loss, regular activity, sleep, alcohol. Dull, and better supported than anything else on offer.
4. Retest at a sensible interval, which for HbA1c is months rather than weeks.

The boundary

Whether anything should be prescribed in this band depends on your age, your other results, your history and how the repeat looks — and that is a judgement for a physician. The one thing worth resisting is treating a band on a scale as if it were an illness that must be medicated.