What fasting insulin concentration ranges indicate normal, insulin resistance, hyperinsulinemia, or type 1 diabetes for diagnosing diabetes?

Fasting Insulin Levels for Diabetes Diagnosis

Fasting insulin is not recommended for diagnosing diabetes itself; instead, use fasting glucose, HbA1c, or oral glucose tolerance testing per ADA guidelines, but fasting insulin can help identify insulin resistance when levels are ≥20 μU/mL or assess for hyperinsulinemic hypoglycemia when ≥3 mIU/L during documented hypoglycemia. 1, 2, 3

Insulin Levels Are NOT Used to Diagnose Diabetes

• Diabetes diagnosis relies on glucose-based criteria only: fasting plasma glucose ≥126 mg/dL (7.0 mmol/L), HbA1c ≥6.5%, or 2-hour glucose ≥200 mg/dL during oral glucose tolerance test 1
• Insulin measurements play no role in the standard diagnostic criteria for type 1 or type 2 diabetes 1
• The ADA explicitly does not recommend routine fasting insulin testing for diabetes screening or diagnosis 2

When Fasting Insulin IS Clinically Useful

For Identifying Insulin Resistance (Not Diabetes)

• Normal fasting insulin: <15 μU/mL 2
• Borderline high: 15-20 μU/mL 2
• High (indicates insulin resistance): ≥20 μU/mL 2
• Only interpret fasting insulin when simultaneous fasting glucose is ≤220 mg/dL (12.5 mmol/L) 2
• Research suggests fasting insulin >9.0 μU/mL identifies prediabetes with 80% accuracy, though this is not a formal diagnostic criterion 4

Important caveat: The American Heart Association recommends assessing insulin resistance syndrome through clinical parameters (BMI, waist circumference, blood pressure, fasting glucose, lipid profile) rather than fasting insulin alone 2

For Diagnosing Hyperinsulinemic Hypoglycemia

• Inappropriate insulin secretion during hypoglycemia: insulin ≥3 mIU/L when glucose ≤45 mg/dL (2.5 mmol/L) has 100% specificity for organic hypoglycemia 3
• However, 11-35% of insulinoma patients fail to reach this threshold, so insulin <3 mIU/L does not rule out the diagnosis 3
• C-peptide is more reliable: ≥0.6 ng/mL during hypoglycemia (glucose ≤45 mg/dL) provides nearly 100% sensitivity and specificity 3

Reference Ranges from Research (Context-Dependent)

While not used diagnostically, population studies show:

• Brazilian population: 2.52-13.14 μU/mL (15.2-79.2 pmol/L) for both sexes 5
• Iranian population: 2.11-12.49 μU/mL overall, with slight age-related decline 6
• French population: median 5.6 μU/mL, 95th percentile 12.6 μU/mL for most adults 7

These ranges vary by assay method, population characteristics, and exclusion criteria, limiting their clinical applicability 5, 6, 7

Type 1 vs Type 2 Diabetes Distinction

• Type 1 diabetes: characterized by absolute insulin deficiency with low or undetectable C-peptide (<0.2 nmol/L or ~0.6 ng/mL after stimulation) 8
• Type 2 diabetes: characterized by relative insulin deficiency with preserved or elevated C-peptide (>0.6 nmol/L after stimulation) 8
• Fasting insulin alone cannot reliably distinguish type 1 from type 2 diabetes; stimulated C-peptide testing (mixed-meal tolerance test) is the preferred method 8

Critical Pitfalls to Avoid

• Do not use insulin levels to diagnose diabetes: This violates established diagnostic criteria and will lead to misclassification 1, 2
• Assay variability is substantial: Different laboratories use different calibration standards, making absolute values difficult to compare 9, 3
• Renal insufficiency invalidates standard thresholds: Use beta-hydroxybutyrate or glucagon stimulation instead 3, 8
• Hemolysis falsely lowers insulin: Strict specimen handling is essential 9
• Anti-insulin antibodies interfere: Measure free insulin if antibodies are present 9

Practical Algorithm for Clinical Use

1. For suspected diabetes: Measure fasting glucose, HbA1c, or perform OGTT—not insulin 1
2. For insulin resistance assessment in high-risk patients (obesity, PCOS, acanthosis nigricans): Consider fasting insulin ≥20 μU/mL as high, but prioritize clinical/metabolic parameters 2
3. For hypoglycemia evaluation: Measure simultaneous glucose, insulin, C-peptide, and proinsulin during documented hypoglycemia (glucose ≤45 mg/dL); insulin ≥3 mIU/L or C-peptide ≥0.6 ng/mL indicates inappropriate endogenous secretion 3, 8
4. For distinguishing diabetes type: Use stimulated C-peptide (90-minute post-meal): <0.2 nmol/L suggests type 1, >0.6 nmol/L suggests type 2 8