Pre-diabetes is not a phase. It is a five-year window.

Rafael is 44. A software engineer, sedentary for a decade, has gained 9 kg (20 lbs) since 35. Diabetic father, a cousin had a foot amputated at 51. He came in with the lab in his hand: HbA1c 5.9%, fasting glucose 108 mg/dL.

The interpretation he had received from a health-plan physician: your sugar is slightly altered, let's redo it in a year and see. The interpretation he wanted: somewhere between relief and despair.

It did not have to be either. It had to be a plan.

What pre-diabetes actually means

By the criteria of the American Diabetes Association, updated in the 2026 Standards of Care, pre-diabetes is the state in which at least one of these three markers is present:

• HbA1c between 5.7% and 6.4% (39 to 47 mmol/mol)
• Fasting glucose between 100 and 125 mg/dL (impaired fasting glucose — IFG)
• 2-hour glucose after a 75 g oral glucose tolerance test between 140 and 199 mg/dL (impaired glucose tolerance — IGT)

Above those values, it is established type-2 diabetes. Below, it is normal. In the middle — that apparently comfortable zone — is where the most underestimated diagnosis in the office lives.

The Tabák and colleagues review, published in Lancet in 2012, synthesized decades of prospective cohorts. Five to ten percent of people with pre-diabetes progress to type-2 diabetes each year. The cumulative conversion rate over 5 to 10 years reaches 35–50% without intervention. The same number of people, however, regress to normoglycemia in the same period — when there is intervention. The nature of that balance is what defines prognosis.

The authors also described something that changes the clinical reading: pancreatic beta-cell dysfunction begins to appear 10 to 15 years before the diagnosis of diabetes. By the time glucose is at 108, the beta cell has been working in overload for more than a decade. An HbA1c of 5.9% is not the beginning of a disease. It is the end of a silence.

Why the window is narrow

The question that matters is not “when will I become diabetic.” It is “how much time do I have to reverse this path.”

The answer, supported by the Diabetes Prevention Program (DPP) — published by Knowler and colleagues in the NEJM in 2002 — is clearer than imagined. The DPP randomized 3,234 adults with pre-diabetes (impaired fasting glucose, impaired post-load glucose, average BMI 34) into three groups:

• Placebo + generic guidance.
• Metformin 850 mg twice daily + generic guidance.
• Intensive lifestyle change program: target of losing 7% of body weight and doing at least 150 minutes per week of moderate physical activity, with structured follow-up.

Average follow-up of 2.8 years. Result:

• The intensive lifestyle group had a 58% reduction in diabetes incidence compared to placebo.
• The metformin group had a 31% reduction.
• The behavioral intervention beat the medication, and the effect was maintained in follow-up analyses at 10, 15, and 21 years.

To prevent one new case of diabetes in three years, it took intervening in 6.9 people in the lifestyle group. For metformin, in 13.9. Those are enormous numbers in preventive medicine — comparable to the best documented cardiovascular treatments.

And more: when the DPP was reanalyzed to see whether remission to normoglycemia predicted long-term risk reduction, the answer was yes — returning to HbA1c below 5.7% reduced the future probability of microvascular complications, even in subsequent years when glycemia oscillated again.

Every year of hyperglycemia counts. Every kilogram lost counts. Every walk counts.

And if it has already passed?

Here is where the DiRECT study, published by Lean and colleagues in Lancet in 2018, changed the conversation.

DiRECT recruited 306 adults aged 20 to 65 with type-2 diabetes diagnosed in the last 6 years, BMI between 27 and 45 kg/m², not on insulin. Half received standard primary-care care. Half entered an intensive program led by trained nurses and dietitians:

• Suspension of antidiabetic and antihypertensive medications.
• Total dietary replacement with a liquid formula of 825–853 kcal/day for 3 to 5 months.
• Stepped reintroduction of foods over 2 to 8 weeks.
• Structured support for weight maintenance.

At 12 months: 46% of the intervention group entered complete remission of diabetes (HbA1c below 6.5%, no medication), vs. 4% in the control. Among those who lost more than 15 kg (33 lbs), 86% achieved remission. At two years, more than a third remained in remission.

The reading is hard and liberating at the same time: type-2 diabetes is not, in most cases, an irreversible and progressive disease. It is a reversible disease as long as the beta cell still works. The window exists — it just isn't eternal.

What to do with pre-diabetes in real life

The DPP recipe is disappointingly simple on paper and rarely executed alone:

• Lose 5 to 7% of body weight, and keep it. In someone who weighs 90 kg (198 lbs), that means 4.5 to 6.3 kg less. It is not fad diet — it is sustained reduction of excess visceral adipose tissue.
• 150 minutes per week of moderate aerobic activity plus 2 to 3 strength training sessions. The DPP focused on aerobic; the recent literature is unanimous about the necessity of strength — muscle is the principal sink of post-prandial glucose.
• Regular sleep of 7 to 8 hours with consistent timing — sleep deprivation worsens insulin resistance within days, demonstrable in controlled studies.
• Reduce refined carbohydrates and sugary drinks — not forever, not with fanaticism, but as a reorganization of the plate.
• Consider metformin in selected cases (HbA1c near 6.4%, age below 60, BMI above 35, women with a history of gestational diabetes) — the additional risk reduction is modest but real, and the drug is safe and inexpensive.

Reassess HbA1c at 3 to 6 months. Body composition by DEXA or bioimpedance at 6 months. Fasting insulin + HOMA-IR alongside HbA1c — because someone with severe insulin resistance (high insulin with glucose still controlled) is closer to progression than the HbA1c number suggests.

What happened with Rafael

Rafael lost 8 kg (18 lbs) in 6 months, returned to training three times a week (exercise physiologist, a plan combining squat, deadlift, bench press, and a Zone 2 session on the weekend), began sleeping 7 hours with consistent timing. He cut out white bread at breakfast, did not cut wine.

HbA1c at 6 months: 5.4%. Fasting glucose: 89. Fasting insulin dropped from 19 to 7 µIU/mL. HOMA-IR dropped from 5.1 to 1.5.

He did not become diabetic. More importantly: he stepped off the path. In five years, without that turn, the probability was above 30% of being on metformin at 49, with his brother's pediatric nephrologist becoming the next family conversation. Instead, he is in a functional metabolic range that, maintained, can give him two more decades without established disease.

Why this requires follow-up, not a single visit

Pre-diabetes does not get resolved in one annual return visit. It gets resolved in panel — physician who reads the entire picture, nutritionist who rebuilds the plate without magic formulas, exercise physiologist who designs the strength plan, psychologist who steps in when fatigue, stress, and emotional eating make execution impossible. That is exactly the design of Continuum Plenya: integrated team, score that evolves over time, follow-up at 3 and 6 months to adjust.

Pre-diabetes is not a phase. It is a window. And windows close.