How is C‑peptide measurement used to diagnose diabetes?

C-Peptide Is Not Used to Diagnose Diabetes—It Classifies Diabetes Type

C-peptide measurement does not diagnose diabetes itself; rather, it helps distinguish between type 1 and type 2 diabetes in ambiguous clinical presentations, particularly when classification based on clinical features alone is unclear. 1 Diabetes diagnosis relies on demonstrating hyperglycemia through plasma glucose or HbA1c measurements, not C-peptide levels.

Primary Clinical Role: Diabetes Classification

C-peptide serves as a biomarker of endogenous insulin secretion, produced in equimolar amounts with insulin. Its key utility emerges when:

• Adults present with features of both diabetes types (e.g., type 2 phenotype but presenting in ketoacidosis) 1
• Insulin-treated patients require classification to guide appropriate management strategies 1
• 3-5 years post-diagnosis, when persistent C-peptide helps differentiate type 2 or monogenic diabetes from type 1 2

Diagnostic Thresholds for Classification

Based on the most recent 2023 ADA/EASD guidelines 1:

For Distinguishing Type 1 from Type 2:

• C-peptide <0.20 nmol/L (<0.6 ng/mL): Indicates type 1 diabetes with severe insulin deficiency
• C-peptide ≥0.20 nmol/L (≥0.6 ng/mL): Suggests preserved beta-cell function (type 2 or monogenic diabetes)
• C-peptide 0.20-0.60 nmol/L (0.6-1.8 ng/mL): Intermediate zone, usually consistent with type 1 or MODY but may occur in insulin-treated type 2 diabetes 3

The 2025 ADA Standards further refined this: C-peptide >600 pmol/L (>1.8 ng/mL) strongly indicates type 2 diabetes, while values <80 pmol/L (<0.24 ng/mL) confirm severe insulin deficiency and don't require repeat testing 3.

Testing Methodology

When to Measure:

• Fasting C-peptide when simultaneous fasting plasma glucose is ≤220 mg/dL (12.5 mmol/L) 1
• Random (non-fasting) C-peptide within 5 hours of eating is acceptable and performs comparably to formal stimulation tests 3, 4
• Do NOT test within 2 weeks of ketoacidosis or hyperglycemic emergency 3

Sample Collection:

• Serum or plasma samples are acceptable
• Results expressed as pmol/L or ng/mL (conversion: pmol/L × 0.003 = ng/mL) 5
• Random spot urine C-peptide:creatinine ratio is emerging as a convenient alternative 2

When C-Peptide Should NOT Be Used

Routine C-peptide testing is NOT recommended for: 1

• Initial diabetes diagnosis (use glucose/HbA1c instead)
• Routine monitoring of type 2 diabetes
• Polycystic ovary syndrome evaluation
• Most people with diabetes or cardiovascular disease risk

Special Diagnostic Applications Beyond Classification

Hypoglycemia Investigation:

C-peptide is essential for evaluating non-diabetic hypoglycemia 1, 6:

• C-peptide ≥0.6 ng/mL with glucose <45 mg/dL (2.5 mmol/L): Indicates endogenous hyperinsulinism (insulinoma, nesidioblastosis) with nearly 100% sensitivity and specificity 5
• C-peptide <0.6 ng/mL with high insulin: Suggests surreptitious exogenous insulin administration 5

Insurance Requirements:

Some payers require fasting C-peptide measurement to approve insulin pump therapy coverage 1

Critical Caveats and Pitfalls

Assay Standardization Issues:

Major limitation: C-peptide assays are NOT standardized across manufacturers, causing significant inter-assay variability 7. The same sample may yield different results depending on the laboratory method used. Apply published thresholds cautiously and validate them with your specific laboratory's assay 5.

Clinical Factors Affecting Interpretation:

• Renal insufficiency: Elevates C-peptide levels, making thresholds unreliable 5
• Hepatic insufficiency: Thresholds not validated 5
• Post-pancreatectomy: Thresholds not applicable 5
• Insulin resistance in type 2 diabetes: May elevate C-peptide despite beta-cell dysfunction 8
• Pubertal stage, age, and BMI: All independently influence C-peptide levels 9

Timing Considerations:

C-peptide measurement has greatest utility 3-5 years after diagnosis when detectable levels help distinguish MODY from type 2 diabetes, or when absent C-peptide confirms absolute insulin requirement regardless of apparent etiology 2.

Algorithmic Approach to C-Peptide Use

Step 1: Confirm diabetes diagnosis using glucose or HbA1c (NOT C-peptide)

Step 2: Assess if classification is clinically ambiguous:

• Clear type 1 presentation (young, lean, ketoacidosis, rapid onset) → No C-peptide needed
• Clear type 2 presentation (older, overweight, gradual onset) → No C-peptide needed
• Ambiguous presentation → Proceed to Step 3

Step 3: Measure C-peptide appropriately:

• Random sample within 5 hours of eating (most convenient) 3
• OR fasting sample when glucose ≤220 mg/dL
• Avoid testing within 2 weeks of ketoacidosis

Step 4: Interpret results:

• <0.2 nmol/L → Type 1 diabetes, absolute insulin requirement
• 0.2-0.6 nmol/L → Likely type 1 or MODY; consider genetic testing if MODY features present

• 0.6 nmol/L → Type 2 or monogenic diabetes; insulin may not be immediately necessary

The evidence consistently demonstrates that C-peptide's role is classification, not diagnosis—a distinction critical for appropriate clinical application.