A1C vs Fasting Glucose vs HOMA-IR

The 60-second summary

The staging sequence — what becomes abnormal first

Type 2 diabetes doesn\'t appear overnight. The metabolic dysfunction progresses through stages, each detectable by a different marker:

1. Stage 1 — early insulin resistance (5–15 years pre-diabetes): Fasting insulin starts climbing as the pancreas compensates for tissue insulin resistance. HOMA-IR rises; fasting glucose and HbA1c remain normal because the extra insulin keeps glucose in range.
2. Stage 2 — impaired post-meal glucose: Post-meal glucose excursions get larger and longer, but fasting glucose stays normal. HbA1c starts edging up. CGM detects this; standard fasting labs may not.
3. Stage 3 — impaired fasting glucose (prediabetes): Fasting glucose enters 100–125 mg/dL range. HbA1c reaches 5.7–6.4%. HOMA-IR has typically been elevated for 5–10 years by this point.
4. Stage 4 — diabetes: Fasting glucose ≥126 mg/dL or HbA1c ≥6.5%. Pancreatic beta-cell function is declining; insulin resistance is severe.

Most people only get tested at stage 3 or 4 because fasting glucose and HbA1c are on standard panels and HOMA-IR isn\'t. By that point the metabolic dysfunction is well advanced. The Diabetes Prevention Program (NEJM 2002) showed lifestyle intervention reduced T2D progression by 58% over 3 years — but the intervention is far more effective at stage 1 (insulin resistance only) than at stage 3 (full prediabetes).

Head-to-head comparison

Fasting glucose — the snapshot

Fasting plasma glucose is what your blood glucose was at the moment of the morning lab draw, after 8–12 hours of fasting. Strengths: cheap, universal, fast turnaround, easy to interpret. The ADA diagnostic thresholds are widely known — normal <100 mg/dL, prediabetes 100–125, diabetes ≥126.

Weaknesses: it\'s a snapshot, and a normal fasting glucose doesn\'t rule out post-prandial hyperglycemia, insulin resistance, or developing diabetes. Day-to-day variation is ±5–10 mg/dL even in stable subjects. A single elevated reading isn\'t diagnostic; the ADA requires confirmation on a second day before diagnosis. For comprehensive assessment, pair fasting glucose with HbA1c and ideally HOMA-IR.

HbA1c — the 90-day average

Hemoglobin A1c measures the percentage of hemoglobin molecules that have undergone irreversible glycation. Because red blood cells live ~120 days, A1C reflects average glucose exposure across that window, weighted slightly toward the most recent 30 days. Higher average glucose = more glycation = higher A1C.

HbA1c is the gold standard for diabetes diagnosis and long-term treatment monitoring. Strengths: integrates the full glycemic picture (fasting + post-meal + overnight) into one number, doesn\'t require fasting, less day-to-day variable than fasting glucose. Weaknesses: anemia, recent blood loss, hemoglobinopathies, pregnancy, and CKD all shift the relationship between average glucose and A1C — sometimes substantially. The Nathan 2008 ADAG formula gives the population-average translation (eAG mg/dL = 28.7 × A1C − 46.7).

HOMA-IR — the early warning

HOMA-IR (Homeostatic Model Assessment of Insulin Resistance) is computed from fasting glucose and fasting insulin: (insulin × glucose) ÷ 405 (mg/dL) or ÷ 22.5 (mmol/L). It estimates the basal insulin needed to maintain fasting glucose — higher numbers mean higher insulin resistance.

The key clinical insight: fasting insulin rises years to decades before fasting glucose goes abnormal. The pancreas compensates for tissue insulin resistance by producing more insulin, which keeps glucose in range. By the time fasting glucose enters prediabetes range, insulin resistance has typically been present for 5–15 years. HOMA-IR catches it earlier — exactly when lifestyle interventions are most effective.

Why it\'s not on standard panels: historical inertia. Diabetes was traditionally diagnosed by glucose alone, so fasting insulin wasn\'t routinely tested. The 2017 endocrinology consensus increasingly recommends adding fasting insulin to standard CMPs for adults with family history of T2D, central obesity, or metabolic syndrome features.

When to use each

• Screening (general adult population): fasting glucose + HbA1c, annually after age 35 (ADA 2024 recommendation).
• Early detection (family history T2D, central obesity, metabolic syndrome): add fasting insulin → HOMA-IR, every 1–2 years starting from young adulthood.
• Diabetes monitoring (T1D or T2D): HbA1c every 3 months. CGM for daily management. Fasting insulin / HOMA-IR less relevant once beta-cell function is the main issue.
• Prediabetes monitoring: HbA1c every 6 months, fasting glucose at the same draw. HOMA-IR every 6 months tracks intervention response faster than HbA1c does.
• CGM users (T1D, T2D, or non-diabetic): CGM for daily TIR + average glucose; lab HbA1c quarterly; lab fasting insulin annually for HOMA-IR.

Try the calculators

• HOMA-IR Calculator
• A1C ↔ eAG Converter
• A1C from Average Glucose Calculator
• Glucose Unit Converter
• All diabetes tools
• All CGM tools

FAQ

Which marker catches diabetes risk earliest?

HOMA-IR. Joseph Kraft's 1975 work and subsequent insulin response studies established that fasting insulin rises 5–15 years before fasting glucose goes abnormal in most subjects who eventually develop type 2 diabetes. By the time fasting glucose hits 100 mg/dL (prediabetes), insulin resistance has typically been present for over a decade. HOMA-IR catches the underlying metabolic dysfunction in that long pre-diabetic window when interventions are most effective. Fasting glucose and A1C only become elevated after pancreatic beta-cells start failing to compensate for the resistance — by then the disease process is well-established. The catch: fasting insulin isn't on standard CMPs and has to be specifically requested.

Why does my HbA1c not match my fasting glucose?

Because they measure different things. Fasting glucose is a single timepoint snapshot — what your glucose was the morning of the test. HbA1c is the percentage of hemoglobin glycated over the previous ~120 days — it integrates all the glucose excursions (fasting, post-meal, overnight) across that window. Someone with elevated post-meal spikes but normal fasting glucose can have an elevated HbA1c that "doesn't match." Conversely, someone with stable but slightly elevated fasting glucose without big spikes can have a lower HbA1c than fasting glucose alone would predict. For complete picture: get both, plus a 14+ day CGM if available.

Can I just use a CGM instead of these blood tests?

For some purposes, yes — but not all. A CGM gives you Time in Range, average glucose, and GMI (estimated A1C from CGM data). GMI is the same Nathan ADAG math as the A1C-from-average-glucose calculator. CGM misses fasting insulin (and therefore HOMA-IR), and the 2018 Bergenstal consensus paper specifically noted that GMI and lab A1C can disagree by 0.3–0.7% in many subjects — both numbers can be right, they're measuring slightly different things. For comprehensive metabolic assessment: CGM + quarterly lab A1C + annual fasting insulin (for HOMA-IR) covers the full picture.