Prediabetes — The Decade Before the Diagnosis

The definition of prediabetes is an HbA1c between 42 and 47 mmol/mol, or a fasting glucose between 6.1 and 6.9 mmol/L. Cross either threshold and you receive the diagnosis. Stay below them and you are told your results are normal. This binary — normal or prediabetic — is one of the most clinically misleading constructs in modern medicine, because the metabolic deterioration that produces prediabetes does not begin at the threshold. It begins roughly a decade earlier, in people whose fasting glucose is entirely normal and whose HbA1c is unremarkable, and it is detectable years before any standard screening test would find it.

The marker that reveals it is fasting insulin. Not fasting glucose. Insulin.

The Compensation Mechanism That Standard Testing Misses

Insulin resistance — the reduced sensitivity of cells to insulin’s glucose-uptake signal — does not immediately raise blood glucose. The pancreatic beta cells respond to rising insulin resistance by producing more insulin. More insulin overcomes the reduced sensitivity and keeps blood glucose normal. The glucose looks fine. The insulin is elevated. Standard screening tests glucose, not insulin, and declares the person well.

This compensatory hyperinsulinaemia can persist for years — even a decade — before beta cell capacity is exhausted and glucose finally begins to rise. During those years, the elevated insulin is not benign. High circulating insulin drives fat storage, particularly visceral abdominal fat. It drives the synthesis of triglycerides in the liver, creating the early fatty liver pattern visible on liver enzymes before any formal diagnosis is made. It promotes inflammation through NF-κB activation. It impairs thyroid T4-to-T3 conversion. It raises uric acid production. It suppresses sex hormone binding globulin, altering the balance of free testosterone and oestrogen. A cascade of systemic effects, years before fasting glucose rises enough to attract clinical attention.

A person with fasting glucose of 5.4 mmol/L and fasting insulin of 18 mIU/L is not in normal metabolic health. Their pancreas is producing approximately four times the insulin required to maintain that glucose level. The glucose looks normal because the system is compensating. The compensation is the disease. Standard screening sees only the glucose.

The Timeline — What Is Actually Happening Over Ten Years

Years 1–3 · Early insulin resistance

Fasting glucose normal. Fasting insulin elevated.

Cells become less responsive to insulin signalling. The pancreas compensates with higher insulin output. HOMA-IR rises above 1.5 — the functional threshold for insulin resistance — while fasting glucose remains entirely within normal range. No standard screening test catches this. The person is asymptomatic or has subtle signs: stronger afternoon carbohydrate cravings, slightly increased difficulty losing weight despite unchanged diet.

Years 3–6 · Established insulin resistance

HOMA-IR above 2.5. Triglycerides rising. Fatty liver beginning.

The liver, responding to chronically high insulin, increases de novo lipogenesis — converting excess carbohydrate to fat and depositing it in liver cells. Hepatic steatosis begins. GGT starts to rise within the normal range. The triglyceride:HDL ratio climbs, reflecting the atherogenic lipid pattern driven by insulin resistance. Visceral abdominal fat accumulates. The person may notice clothes fitting differently around the abdomen without significant weight change elsewhere. Morning joint stiffness begins — driven by low-grade systemic inflammation from the metabolic disruption.

Years 6–8 · Compensatory capacity stressing

HbA1c creeping up within normal range. Beta cell stress.

The sustained demand on beta cells to produce excess insulin begins to affect their function. Post-meal glucose excursions — not captured by fasting testing — become larger and more prolonged. HbA1c rises from 35 toward 40 mmol/mol — still entirely normal by standard criteria, entirely abnormal by functional criteria. Hair thinning may appear — driven by the combined effect of elevated androgens (from suppressed SHBG), thyroid conversion impairment, and chronic low-grade inflammation affecting the follicle cycle.

Years 8–10 · Threshold approach

Fasting glucose rising. Standard screening begins to detect.

Fasting glucose rises above 5.5 mmol/L. HbA1c approaches 42 mmol/mol. Beta cell function is measurably reduced. By the time standard screening catches the person in the prediabetic range, ten years of metabolic deterioration has already occurred. Fatty liver may already be established. Cardiovascular risk markers are significantly worsened. The window for the simplest, most effective intervention has been substantially narrowed.

Standard diagnosis point · ~Year 10

HbA1c 42–47 or fasting glucose 6.1–6.9. “Prediabetes” diagnosed.

The first clinical conversation about blood sugar occurs. Advice: watch your diet, lose weight, come back in a year. No fasting insulin tested. No HOMA-IR calculated. No explanation of what has been happening metabolically for a decade. No discussion of the specific dietary, exercise, and metabolic interventions with evidence for reversal. The opportunity that existed years earlier — when reversal required only modest lifestyle intervention on a system that still had full compensatory capacity — has become harder.

The Markers That See It Coming

Marker GP threshold Functional optimal What it signals

Fasting insulin

Rarely tested

3–8 mIU/L

Above 10 = compensatory hyperinsulinaemia. Above 15 = established resistance.

HOMA-IR

Rarely calculated

Below 1.5

1.5–2.5 = early resistance. Above 2.5 = established. Above 4 = significant.

Fasting glucose

Below 6.1 = normal

4.0–5.0 mmol/L

5.0–5.5 warrants fasting insulin. 5.5+ with normal insulin = different picture.

HbA1c

Below 42 = normal

Below 37 mmol/mol

37–42 is within normal range and functionally concerning. Trend matters.

Triglyceride:HDL ratio

Not routinely assessed

Below 1.0

Above 1.5 is atherogenic. The most accessible insulin resistance surrogate on standard bloods.

GGT (trend)

Below 50 = normal

Below 20 U/L

Rising GGT within normal range reflects hepatic fatty change and oxidative stress years before ALT moves.

Uric acid

Below 360 µmol/L (women), 420 (men)

Below 300 µmol/L

Fructose drives uric acid production directly. Elevated uric acid is an early metabolic syndrome marker.

Why Reversal Is Straightforward Earlier Than Later

The physiological distinction between early insulin resistance and established type 2 diabetes is beta cell function. In early insulin resistance, beta cells are stressed but intact — they are producing excess insulin, which is demanding, but their cellular machinery is undamaged. The insulin resistance is in the peripheral tissues and the liver. Reduce the demand on the beta cells — through dietary carbohydrate reduction, improved insulin sensitivity through exercise, and addressing the upstream drivers — and both the insulin resistance and the beta cell function recover.

In established type 2 diabetes, beta cell mass is reduced through glucotoxicity and lipotoxicity — the sustained damage from years of high glucose and fatty acid exposure. Reversal is still possible — the DiRECT trial demonstrated substantial remission through very low calorie dietary intervention — but it requires more intensive intervention and the recovery ceiling is lower than it would have been ten years earlier.

This is the clinical argument for early identification. Not to label people with a diagnosis. To identify the metabolic deterioration at the point where the intervention required is a moderate dietary change and a structured exercise programme, not a pharmaceutical protocol managing a condition that has already caused organ damage.

The Interventions That Work

Reduce liquid fructose first

Liquid fructose — fruit juice, sugar-sweetened drinks, high-fructose corn syrup — is the most hepatotoxic dietary carbohydrate. It bypasses the regulatory controls that glucose goes through and is metabolised almost exclusively by the liver, directly driving de novo lipogenesis and hepatic fat accumulation. Eliminating liquid fructose is the single dietary change with the most direct impact on fatty liver and insulin resistance and requires no other dietary restriction to achieve significant metabolic benefit.

Resistance training three times weekly

Skeletal muscle is the primary site of insulin-mediated glucose disposal — responsible for approximately 80–90% of postprandial glucose uptake. Increasing muscle mass and improving neuromuscular insulin sensitivity through progressive resistance training directly reduces the glucose-clearing demand on an already stressed beta cell population. The effect on HOMA-IR from twelve weeks of resistance training is equivalent to or greater than first-line pharmaceutical intervention in multiple trials.

Carbohydrate timing and protein anchoring

Total carbohydrate reduction is effective for insulin resistance. Carbohydrate timing — concentrating carbohydrate intake earlier in the day, avoiding significant carbohydrate in the evening when insulin sensitivity is naturally lower — adds additional metabolic benefit. Anchoring every meal with adequate protein (30g minimum) slows gastric emptying, reduces the glycaemic load of the meal, and supports the muscle mass that drives insulin-mediated glucose disposal. These are sequence changes, not deprivation.

Intermittent fasting — meal timing extension

A minimum twelve-hour overnight fast — not eating after 7pm and not eating before 7am — allows insulin to fall to its genuinely fasted baseline, promoting fat oxidation and giving beta cells recovery time. Extending to fourteen or sixteen hours produces additional benefit in people with established insulin resistance. This is not aggressive fasting. It is restoring the meal timing pattern that human biology was designed for — meals during daylight, fasting overnight.

Sleep — the underestimated metabolic lever

A single night of five hours of sleep reduces insulin sensitivity by 25% the following day. Chronic sleep restriction produces persistent insulin resistance independent of diet and exercise. GLP-1 response to food is significantly blunted after poor sleep. Any metabolic intervention that does not include sleep optimisation is working against itself. Sleep is not downstream from metabolic health. It is upstream of it.

Stress management — the cortisol-insulin axis

Cortisol raises blood glucose through hepatic glucose output, reduces insulin sensitivity in peripheral tissues, and promotes visceral fat deposition. Chronic HPA axis activation is a metabolic intervention in its own right — in the wrong direction. Addressing the stress physiology is not separate from addressing insulin resistance. In someone with significant HPA dysregulation, dietary and exercise changes alone will produce partial results until the cortisol pattern is addressed.

The test that changes the conversation

Ask your GP for a fasting insulin alongside the next fasting glucose. If they decline, it can be ordered privately. The HOMA-IR is calculated from both: fasting glucose (mmol/L) × fasting insulin (mIU/L) ÷ 22.5. A result above 1.5 warrants intervention. A result above 2.5 makes intervention urgent — not because you are diabetic, but because you are ten years closer to it than your glucose would suggest.

The other markers to ask about: triglyceride:HDL ratio (calculable from a standard lipid panel), GGT trend over the past few years, and HbA1c trend rather than single value. Together these give a metabolic picture that fasting glucose alone cannot provide.

Reversal is genuinely achievable at the early and middle stages of insulin resistance — not merely delay. The metabolic system at this stage has significant plasticity. The interventions required are available to almost everyone. The window closes, but it closes slowly, and it is almost certainly still open.

✦ ✦ ✦

The person who reaches a prediabetes diagnosis having never been told their fasting insulin was elevated, never had their HOMA-IR calculated, and never received any information about the decade of metabolic deterioration that preceded the threshold — that person was not failed by luck. They were failed by a screening system that watches one marker in isolation and calls everything below the threshold normal. Normal and optimal are not the same thing. The decade before the diagnosis is the decade that matters. It is the decade when the conversation about fasting insulin, HOMA-IR, and the specific, targeted, evidence-based interventions that restore insulin sensitivity should be happening. Not ten years later when the glucose finally crosses the line.