Insulin Resistance: The Decade-Long Warning Before Diabetes

A fasting glucose of 126 mg/dL is the diagnostic threshold for type 2 diabetes. A fasting glucose of 125 mg/dL is classified as prediabetes. A fasting glucose of 99 mg/dL is called normal. The implication is that health and disease are separated by a single milligram per deciliter — a cliff edge that a single blood test can detect. The physiology does not work that way. The transition from metabolically healthy to diabetic is a gradient that unfolds over roughly a decade, and the warning signs appear long before fasting glucose drifts into the prediabetic range.

What the Whitehall II Study Actually Showed

In 2009, Adam Tabák and colleagues published one of the most clarifying papers on diabetes progression ever produced. The Whitehall II cohort followed over 6,500 British civil servants for more than a decade, measuring glucose, insulin, and insulin sensitivity at repeated intervals. The team then looked backward from the diagnosis of type 2 diabetes to trace what had changed, and when.

The trajectory was not gradual. Fasting glucose rose modestly for years before diabetes onset, then accelerated sharply in the final two to three years. Insulin sensitivity, measured by HOMA-IR, declined steadily over the entire observation period. Insulin secretion — the pancreatic beta-cell response — rose to compensate, maintained a plateau, then collapsed in the final years before diagnosis. The visible failure that a fasting glucose test catches is the last stage of a much longer process. The compensation had already run its course by the time the diagnostic threshold was crossed.

The clinical implication is significant. Patients who receive a prediabetes diagnosis are not at the beginning of metabolic disease. They are roughly eight to ten years into it.

The Reaven Framework

The conceptual foundation for understanding this process was laid by Gerald Reaven in his 1988 Banting Lecture. Reaven proposed that insulin resistance — the reduced ability of tissues to respond to insulin — was the central defect underlying a cluster of conditions that included hyperglycemia, hypertension, elevated triglycerides, reduced HDL cholesterol, and cardiovascular disease. He called the cluster "Syndrome X." The medical community now calls it metabolic syndrome, and the criteria used to define it are direct descendants of Reaven's original description.

Reaven's key insight was that the pancreas could compensate for insulin resistance for years by producing more insulin. Fasting glucose would remain normal during this period, because the elevated insulin was doing the work of suppressing hepatic glucose production and promoting peripheral uptake. A patient in this state would appear healthy on a routine fasting glucose test. They were not healthy. They were working harder, metabolically, to maintain the same output — and the beta cells were aging under the strain.

DeFronzo's 2009 Banting Lecture extended this picture with the "ominous octet" — eight tissues and processes that contribute to hyperglycemia in established type 2 diabetes. The list includes impaired insulin secretion, reduced incretin effect, increased glucagon secretion, elevated hepatic glucose output, reduced glucose uptake in muscle, increased lipolysis, increased renal glucose reabsorption, and neurotransmitter dysfunction in the brain. The dysfunction is systemic, and it is progressive.

What Fasting Glucose Misses

The problem with relying on fasting glucose as the primary screening tool is that it is the last metric to drift in the progression. By the time it has crossed the prediabetic threshold, insulin sensitivity has often declined by 30 to 50%, and beta-cell function has typically lost a substantial fraction of its reserve. The earlier indicators are more sensitive but less commonly measured in routine primary care:

• Fasting insulin rises before fasting glucose does. A fasting insulin above 10 µIU/mL in the context of normal glucose suggests the pancreas is working overtime. Optimal is generally below 7 µIU/mL.
• HOMA-IR, calculated from fasting glucose and fasting insulin, directly quantifies insulin resistance. A HOMA-IR above 2.5 is commonly used as a cutoff, though there is no universally agreed threshold.
• Hemoglobin A1c, the three-month average of blood glucose, begins to drift upward years before fasting glucose does. The ADA uses 5.7% to 6.4% as the prediabetic range, but within the "normal" range of 5.0% to 5.6%, higher values correlate with higher downstream risk.
• Triglyceride-to-HDL ratio is a rough surrogate for insulin resistance derived from a standard lipid panel. A ratio above 3.0 is associated with meaningful insulin resistance in most populations studied.
• Waist circumference reflects central adiposity, which is both a cause and a marker of insulin resistance independent of overall body weight.

None of these are perfect. In combination, they paint a much earlier picture than fasting glucose alone.

The Lifestyle Intervention That Actually Works

The Diabetes Prevention Program, published in 2002 and followed for over two decades now, remains the most important clinical trial on reversing the progression. The study enrolled adults with prediabetes and randomized them to lifestyle intervention, metformin, or placebo. The lifestyle arm — modest weight loss of 7% of body weight, combined with 150 minutes per week of moderate activity — reduced progression to diabetes by 58%. The metformin arm reduced it by 31%. Lifestyle outperformed pharmaceutical intervention by nearly a factor of two.

The specific mechanisms of the lifestyle intervention matter more than the number on the scale suggests. Weight loss reduces visceral fat, which is metabolically active and directly contributes to hepatic insulin resistance. Physical activity improves muscle insulin sensitivity through mechanisms partially independent of weight loss — contracting muscle takes up glucose via GLUT4 transporters in a largely insulin-independent manner, reducing the pancreatic workload at rest. Resistance training adds a further layer by increasing muscle mass, which expands the tissue sink for glucose disposal.

Dietary composition also matters, though the ideal pattern remains contested. Most of the intervention studies that have reduced insulin resistance specifically have involved lower-glycemic-load carbohydrates, higher fiber intake, adequate protein, and reduced ultra-processed food consumption. The common denominator is not a specific macronutrient distribution but the elimination of the postprandial glucose spikes that keep the pancreas working overtime.

The Screening Case

The argument for earlier and more comprehensive metabolic screening — specifically, adding fasting insulin and A1c to routine bloodwork for adults over 35, rather than waiting for fasting glucose to drift — is clinical, not speculative. The Whitehall II trajectory data makes the case clearly: waiting until fasting glucose crosses the prediabetic threshold means waiting until the damage is already substantial. The CDC's 2024 report estimates that 38% of U.S. adults have prediabetes, and more than 80% of them are unaware of it. A fasting insulin test costs roughly the same as a fasting glucose test. The clinical information is fundamentally different, and it arrives substantially earlier in the disease trajectory.

Type 2 diabetes is not inevitable. But it is silent, and it is slow, and the markers that would catch it early are not the ones most commonly looked at. The decade of warning is there for almost every patient who eventually develops the disease. Most of it goes unread.