The 11th edition of the IDF Diabetes Atlas, published in The Lancet Diabetes & Endocrinology in December 2025, estimates that 589 million adults worldwide now have diabetes — roughly 1 in 9 people aged 20–79. [1] By 2050, that number is projected to reach 852.5 million. Globally, 42.8% of people with diabetes, over 250 million adults, don't know they have it. [2] In Germany, approximately one in five people with diabetes remain undiagnosed. [3]
But here's what those numbers obscure: diabetes is the end of a metabolic story, not the beginning. Insulin resistance, the underlying driver of type 2 diabetes, is thought to precede diagnosis by 10 to 15 years. [4] Some research puts the window at 20 years or more. [5] And standard blood tests, fasting glucose or HbA1c, are designed to detect the disease, not the decade of metabolic dysfunction that leads to it.
This article is about what happens in that decade, and why the blood markers most people rely on don't capture it.
When your GP checks your "blood sugar," the typical test is one or both of the following:
Fasting plasma glucose (FPG): A snapshot of your blood glucose after an overnight fast. Normal is below 100 mg/dL (5.6 mmol/L). Prediabetes: 100–125 mg/dL (5.6–6.9 mmol/L). Diabetes: 126 mg/dL (7.0 mmol/L) or higher.
HbA1c (glycated haemoglobin): A measure of average blood glucose over the past 2–3 months. It works by measuring the percentage of haemoglobin molecules that have had glucose attach to them — a process called glycation. Normal is below 5.7%. Prediabetes: 5.7–6.4%. Diabetes: 6.5% or higher.
Both of these tests share an important limitation: they measure glucose, not insulin. And in the early stages of metabolic dysfunction, glucose is the last thing to change. Precisely because your pancreas compensates by producing more insulin.
Think of it this way: insulin resistance means your cells, primarily in your muscles, liver, and fat tissue, don't respond to insulin as effectively as they should. When you eat, your blood glucose rises and your pancreas releases insulin to shuttle that glucose into cells. If those cells are resistant, the pancreas has to produce more insulin to get the same glucose-lowering effect. [4]
For years, often many years, this compensation works. Your blood glucose stays normal because your pancreas is working overtime. Fasting glucose: normal. HbA1c: normal. But insulin levels are elevated, often dramatically so. The metabolic dysfunction is already there. The standard tests just can't see it, because they're looking at the wrong thing.
It's only when the pancreas can no longer keep up, when beta-cell function starts to fail, that glucose finally rises above normal thresholds. By that point, significant metabolic damage has often already occurred. [4]
In 1975, clinical pathologist Dr. Joseph Kraft published what remains one of the most striking demonstrations of this blind spot. He performed glucose-insulin tolerance tests on 3,650 patients, measuring both glucose and insulin at multiple time points after a glucose load. [5]
The results were remarkable. Among patients with normal glucose tolerance by standard criteria, only 33% also had a normal insulin response. The remaining patients, roughly two-thirds, showed abnormal insulin patterns: exaggerated peaks, delayed return to baseline, or persistently elevated levels. Kraft classified these as "diabetes in-situ": metabolic dysfunction already present but invisible to glucose-only testing. [5]
When all patients with normal or equivocal glucose tolerance were combined (n=2,940), 75% had borderline or abnormal insulin response patterns. [5] A subsequent collaboration confirmed that at least 85% of people with impaired glycaemic responses also had elevated insulin concentrations, and that 2-hour postprandial insulin was more diagnostic than fasting insulin alone. [6]
Kraft's work introduced five insulin response patterns (Patterns I–V), ranging from normal to overt diabetes, based on the shape of the insulin curve after glucose ingestion. The critical insight: you could progress from Pattern I (normal) through Patterns II and III (compensated insulin resistance) while maintaining completely normal glucose values. Pattern IV showed both elevated insulin and glucose, overt diabetes. Pattern V showed insufficient insulin production entirely consistent with type 1 diabetes or exhausted beta cells. [5]
This research, now nearly 50 years old, makes a simple but powerful point: if you only measure glucose, you miss the earliest and most reversible stages of metabolic dysfunction. Glucose-only screening catches the fire after it's burning. Insulin measurement catches the smoke.
Standard blood sugar tests don't measure insulin. Aniva does. Fasting insulin, fasting glucose, HbA1c, HOMA-IR, plus the cofactors that affect how your body processes glucose. 140+ biomarkers. One blood draw.
If fasting glucose alone is insufficient, what should you be measuring instead? The most widely validated clinical surrogate for insulin resistance is HOMA-IR — the Homeostatic Model Assessment of Insulin Resistance.
Developed by Matthews and Turner in 1985, HOMA-IR is calculated from two simple fasting measurements: [7]
HOMA-IR = (fasting insulin [µU/mL] × fasting glucose [mg/dL]) / 405
Or in SI units: (fasting insulin [µU/mL] × fasting glucose [mmol/L]) / 22.5
The result gives you a single number that estimates how hard your body is working to maintain normal glucose levels. Here's how to interpret it: [7] [8]
The general clinical cutoff for diagnosing insulin resistance is a HOMA-IR above 2.5, though optimal values are below 1.0. [7] In large European and Brazilian cohorts, HOMA-IR predicted both prediabetes and overt diabetes, with cutoffs ranging between 1.8 and 3.6 depending on population and BMI. [9]
The power of HOMA-IR is in what it reveals that glucose alone does not. A 2023 study published in eClinicalMedicine confirmed that high HOMA-IR values are independently associated with high risks of developing prediabetes, and that this association is statistically independent of impaired glucose tolerance, obesity, and body fat distribution. [10] In other words, you can have a normal BMI, normal glucose tolerance, and still be insulin resistant, and HOMA-IR can detect it.
Among type 2 diabetics, higher HOMA-IR values independently predict future cardiovascular events, even after adjusting for traditional risk factors. [9] Insulin resistance is also associated with all-cause mortality: one study found an adjusted hazard ratio of 1.16 for all-cause mortality in non-diabetic patients with elevated HOMA-IR. [10]
Here's the frustrating part: calculating HOMA-IR requires only fasting glucose and fasting insulin. Both simple, inexpensive blood tests. But fasting insulin is not routinely measured in most standard GP panels across Europe or the US. [10] The standard approach screens with fasting glucose or HbA1c alone, which means the earliest and most actionable stage of metabolic dysfunction, compensated hyperinsulinaemia, goes undetected.
An estimated 25% of non-diabetic adults have significant insulin resistance. [11] Most of them have no idea, because the standard tests they receive don't measure it.
HbA1c deserves its own section because it's increasingly used as the sole screening and monitoring tool for blood sugar, and it has limitations that most people (and many clinicians) don't fully appreciate.
As mentioned, HbA1c reflects average blood glucose over the lifespan of your red blood cells, approximately 90–120 days. [12] It's a good measure of long-term glycaemic exposure. But its accuracy depends on a critical assumption: that your red blood cells behave normally.
They often don't.
Iron deficiency anaemia — one of the most common nutrient deficiencies worldwide, and particularly prevalent in women of reproductive age, is associated with falsely elevated HbA1c values. [13] The mechanism involves two processes: iron deficiency slows red blood cell production and turnover, giving more time for glycation to occur; and it increases malondialdehyde levels, which directly enhances haemoglobin glycation. [14]
A systematic review found that treating iron deficiency anaemia with iron supplementation reduced HbA1c by up to 1.2 percentage points, a clinically significant amount that could mean the difference between a "normal" and "prediabetic" classification. [15]
This matters enormously for screening: a woman with iron deficiency could receive a falsely elevated HbA1c result, be told she has prediabetes, and begin interventions for a metabolic problem she may not actually have. Conversely, treating the iron deficiency could "resolve" her prediabetes without any change in actual glucose metabolism.
The list of factors that can render HbA1c unreliable is longer than most people realize: [12] [15]
Falsely high HbA1c: Iron deficiency anaemia, B12/folate deficiency, chronic alcohol use, asplenia (absent spleen).
Falsely low HbA1c: Acute or chronic blood loss, haemolytic anaemia, pregnancy, splenomegaly, treatment with erythropoietin, recent blood transfusion.
Variable effects: Chronic kidney disease (multiple competing mechanisms), haemoglobin variants (Hb S, C, E. All common in certain populations), high-dose vitamin C or E supplementation.
The NGSP (National Glycohemoglobin Standardization Program) explicitly recommends alternative testing, such as glycated albumin or fructosamine, for patients with conditions affecting red blood cell lifespan. [13] But in practice, these alternatives are rarely ordered.
This is exactly why a single-marker approach to metabolic health is so limited. HbA1c has value,but its value increases dramatically when cross-referenced with fasting glucose, fasting insulin, iron/ferritin, vitamin B12, and a complete blood count. If your HbA1c says 5.9% (prediabetic range) but your ferritin is 8 ng/mL (severely iron-depleted), the HbA1c may be artificially elevated. Without the iron data, you can't know.
Metabolic health requires more than one number.Aniva measures fasting glucose, HbA1c, fasting insulin, HOMA-IR, iron, ferritin, vitamin B12, folate, and triglycerides. Because metabolic markers don't exist in isolation.
Fasting glucose is the simplest blood sugar test: how high is your glucose right now, after an overnight fast? It's quick, cheap, and tells you whether your body can maintain normal glucose at rest.
What it doesn't tell you: how your body handles glucose after a meal — which is where insulin resistance first manifests.
The progression of metabolic dysfunction follows a predictable sequence. First, postprandial (after-meal) glucose responses start to deteriorate. Your body takes longer to bring glucose back down after eating. But fasting glucose — which reflects overnight hepatic glucose production — may remain normal for years. [4] By the time fasting glucose crosses the 100 mg/dL threshold, the postprandial dysfunction has often been present for a long time.
This is why the WHO definition of impaired glucose tolerance uses a 2-hour oral glucose tolerance test (OGTT) rather than fasting glucose: it captures postprandial dysfunction that fasting measurements miss. The IDF Diabetes Atlas 11th edition estimates that 634.8 million adults globally (12.0%) have impaired glucose tolerance — and that the prevalence of impaired fasting glucose is a separate, often non-overlapping population of 487.7 million (9.2%). [2] These are not identical groups. Many people with IGT have normal fasting glucose, and vice versa. Testing only one will miss the other.
For practical purposes, most people aren't going to do a 2-hour OGTT as routine screening. But there's a far simpler intermediate step: add fasting insulin to the fasting glucose test. This turns a simple glucose snapshot into a HOMA-IR calculation — transforming a single data point into a measure of metabolic function.
If you already have a standard lipid panel, you may have an additional window into insulin resistance without any extra testing.
The triglyceride-to-HDL cholesterol ratio has been shown to correlate with insulin resistance and is increasingly used as a surrogate marker when fasting insulin isn't available. The relationship is logical: insulin resistance drives increased hepatic triglyceride production and reduces HDL cholesterol. The same metabolic pattern seen in metabolic syndrome.
A triglyceride/HDL ratio above 3.0 (using mg/dL) or above 1.4 (using mmol/L) is generally considered suggestive of insulin resistance. While less precise than HOMA-IR, it can flag individuals who might benefit from more detailed metabolic testing.
This is another example of why comprehensive blood panels, where multiple markers are measured simultaneously, provide dramatically more insight than isolated tests. The relationship between your triglycerides, HDL, fasting glucose, fasting insulin, and HbA1c tells a metabolic story that no single marker can.
The consequences of insulin resistance extend far beyond blood sugar. The StatPearls review on insulin resistance lists the metabolic consequences: hyperglycaemia, hypertension, dyslipidaemia, hyperuricaemia, elevated inflammatory markers, endothelial dysfunction, and a prothrombotic state. [4]
Specifically, insulin resistance is associated with or directly drives:
Type 2 diabetes. The primary downstream consequence. 30–50% of individuals with prediabetes progress to T2D within 5–10 years. [6]
Cardiovascular disease. Hyperinsulinaemia promotes endothelial dysfunction, arterial stiffness, and atherogenic dyslipidaemia (high triglycerides, low HDL, small dense LDL). HOMA-IR independently predicts cardiovascular events. [9]
Non-alcoholic fatty liver disease (NAFLD). Insulin resistance is the primary metabolic driver of NAFLD, now the most common chronic liver disease in Europe. HOMA-IR values above 2.0–2.5 strongly distinguish NAFLD patients from controls. [9]
Polycystic ovary syndrome (PCOS). Insulin resistance is present in approximately 70% of women with PCOS and drives many of its symptoms — hormonal imbalance, weight gain, infertility. [9]
Hyperuricaemia and gout. Elevated insulin reduces renal uric acid clearance, directly linking insulin resistance to elevated uric acid levels.
Chronic inflammation. Insulin resistance promotes pro-inflammatory cytokine production, linking metabolic dysfunction to systemic inflammation measurable via hs-CRP.
Certain cancers. Hyperinsulinaemia is associated with increased cancer risk through its effects on the insulin-IGF-1 axis, promoting cell proliferation. [6]
The breadth of this list is the point: insulin resistance isn't just a "blood sugar problem." It's a systemic metabolic disturbance that touches cardiovascular, hepatic, reproductive, inflammatory, and oncological health. Measuring only glucose captures only the most visible symptom while missing the underlying driver.
Insulin resistance affects nearly every system in your body. You should know if you have it.Aniva's 140+ biomarker panel measures metabolic, cardiovascular, inflammatory, hormonal, and nutritional markers together — because that's how your body actually works. Not in isolated numbers, but as an interconnected system.
Full membership: €199/year.
If you want to understand your metabolic health, not just whether you've already crossed a disease threshold, here's what you likely should be alsomeasuring and why:
1. Fasting glucose. The baseline. Tells you whether your body can maintain glucose homeostasis at rest. Necessary but insufficient alone.
2. Fasting insulin. The missing piece in standard screening. Reveals whether your pancreas is working overtime to maintain that "normal" glucose. Optimal: below 5–7 µU/mL. Above 15 µU/mL suggests significant hyperinsulinaemia. [5]
3. HOMA-IR. Calculated from the two above. The single most informative fasting measure of insulin resistance. Target: below 1.0 for optimal metabolic health.
4. HbA1c. Average glucose over 2–3 months. Useful in context, but must be interpreted alongside iron status, haemoglobin, and red blood cell health. [12]
5. Triglycerides and HDL cholesterol. The TG/HDL ratio provides an additional insulin resistance signal from your lipid panel. Also independently relevant for cardiovascular risk.
6. Iron and ferritin. Essential for interpreting HbA1c accurately. Iron deficiency can falsely elevate HbA1c by up to 1.2 percentage points. [15] Women of reproductive age are particularly vulnerable to this confound.
7. Vitamin B12 and folate. Deficiency in either can affect red blood cell turnover and distort HbA1c results. [12]
8. hs-CRP. Chronic inflammation both results from and contributes to insulin resistance. Provides a systemic context that glucose markers alone cannot.
9. Uric acid. Elevated by hyperinsulinaemia and independently associated with metabolic syndrome. An early signal that often appears before glucose abnormalities.
10. Liver enzymes (ALT, AST, GGT). NAFLD is the hepatic manifestation of insulin resistance and is now the most common chronic liver disease in Europe. Elevated liver enzymes may be the first clinical sign.
This is 10 markers altogether. A standard GP metabolic panel typically measures just 1–2 of them (fasting glucose and/or HbA1c). The information gap is enormous, not because the additional tests are expensive or exotic, but because the screening paradigm was designed to detect disease, not to detect the decade of dysfunction that precedes it.
The numbers are sobering. Updated 2024 estimates from the IDF Diabetes Atlas 11th edition report: [16]
The IDF notes that there is still no consensus definition of "prediabetes", at least five different definitions are endorsed by different clinical organisations. [2] This definitional ambiguity likely leads to substantial underestimation of the actual burden, particularly because most prevalence estimates are based on glucose criteria alone, not insulin measurement.
If Kraft's data are even directionally correct — that the majority of people with "normal" glucose tolerance already show abnormal insulin dynamics, then the real burden of early metabolic dysfunction is far greater than any glucose-based prevalence estimate suggests.
Here's why all of this matters practically: early insulin resistance is reversible. [4] The StatPearls review on insulin resistance is explicit: lifestyle modifications, calorie reduction, avoidance of foods that stimulate excessive insulin demand, and increased physical activity, should be the primary focus when treating insulin resistance. Resistance training in particular increases glucose disposal capacity by building muscle, the body's largest glucose-storage depot.
But you can only reverse something you've detected. And you can only detect insulin resistance if you measure insulin.
The tragedy of the current screening paradigm is timing. By the time fasting glucose hits 126 mg/dL or HbA1c reaches 6.5%, the patient has Type 2 diabetes. The pancreatic beta cells have been under compensatory strain for years. Some degree of beta-cell dysfunction may already be irreversible. The window for lifestyle intervention alone to resolve the problem has narrowed considerably.
Measuring fasting insulin and calculating HOMA-IR pushes the detection window back by a decade or more into the period when metabolic dysfunction is most responsive to intervention and least likely to require medication.
Standard metabolic screening measures glucose. Glucose is the last thing to change.
Insulin resistance drives type 2 diabetes, cardiovascular disease, fatty liver disease, PCOS, chronic inflammation, and increased cancer risk. It precedes elevated glucose by 10–20 years. It can be detected with a simple, inexpensive fasting insulin test. And in its early stages, it's reversible through lifestyle changes alone.
The Kraft data showed that 75% of glucose-normal individuals had abnormal insulin patterns. HbA1c, the most common monitoring tool, can be thrown off by iron deficiency, anaemia, kidney disease, and a dozen other conditions. The IDF estimates over 630 million people worldwide have impaired glucose tolerance, with prevalence continuing to rise.
Adding fasting insulin, HOMA-IR, and the cofactors that affect metabolic marker accuracy to a standard blood panel turns a disease-detection system into a disease-prevention system. That's the difference between waiting for diabetes and preventing it.
Medical disclaimer: This content is for informational purposes only and is not medical advice. If you have concerns about insulin resistance, prediabetes, or diabetes, consult a qualified healthcare professional.
