How to Check for C-Peptide
Measure C-peptide in the fasting state when the simultaneous fasting plasma glucose is ≤220 mg/dL (12.5 mmol/L), or use a random non-fasting sample with concurrent glucose measurement when the patient is seen in clinic. 1
Practical Testing Approaches
For Diabetes Classification (Type 1 vs Type 2)
Fasting C-peptide is the preferred method when you need to distinguish between diabetes types in ambiguous cases 1. The 2023 ADA/EASD guidelines provide clear direction:
- Only test C-peptide in insulin-treated patients 2, 3
- Ensure fasting plasma glucose is ≤220 mg/dL (12.5 mmol/L) at time of testing 1
- Do not test within 2 weeks of a hyperglycemic emergency (DKA/HHS) 2, 3
- Must measure C-peptide before insulin discontinuation to exclude severe insulin deficiency 2, 3
Interpretation thresholds:
- <200 pmol/L (<0.6 ng/mL): Severe insulin deficiency, consistent with type 1 diabetes 2, 3
- 200-600 pmol/L (0.6-1.8 ng/mL): Indeterminate, usually type 1 or MODY, but can occur in insulin-treated type 2 diabetes 2, 3
- >600 pmol/L (>1.8 ng/mL): Consistent with type 2 diabetes 2, 3
Alternative: Random Non-Fasting C-Peptide
A random non-fasting blood sample is highly practical and clinically valid 4. This approach allows testing when the patient presents to clinic without special preparation:
- Collect random C-peptide with concurrent glucose measurement 2, 3
- If concurrent glucose is <4 mmol/L (<70 mg/dL) or patient may have been fasting, repeat the test 2, 3
- If result is >600 pmol/L (>1.8 ng/mL), the circumstances of testing don't matter 2, 3
- Very low levels (<80 pmol/L [<0.24 ng/mL]) don't need repeating 2, 3
The random sample within 5 hours of eating can replace formal stimulation testing for classification purposes 2, 3. Research shows random C-peptide correlates strongly with stimulated values (r=0.91-0.96) and maintains excellent sensitivity (100%) and specificity (93%) for detecting severe insulin deficiency 4.
For Hypoglycemia Investigation
C-peptide measurement is essential when investigating non-diabetic hypoglycemia, particularly to rule out surreptitious insulin administration 1, 5.
During documented hypoglycemia (glucose ≤45 mg/dL [2.5 mmol/L]):
- Measure C-peptide simultaneously with glucose, insulin, and proinsulin 6
- C-peptide >0.6 ng/mL (>200 pmol/L) indicates inappropriate insulin secretion with nearly 100% sensitivity and specificity 6
- Also measure beta-hydroxybutyrate and screen for sulfonylureas 7
Critical caveat: These thresholds are not valid in renal insufficiency 6. In kidney disease, use plasma beta-hydroxybutyrate assay and glucagon stimulation test instead 6.
Glucagon Stimulation Test
When formal stimulation is needed (e.g., equivocal results, research protocols):
- Administer 1 mg glucagon intravenously 7
- Measure C-peptide at baseline and post-stimulation 1
- This test balances sensitivity with practicality 8
Urine C-Peptide Alternative
Urine C-peptide creatinine ratio (UCPCR) is a non-invasive alternative 9, 4:
- Collect single post-stimulated urine sample 9
- UCPCR <0.2 nmol/mmol has 83% sensitivity and 93% specificity for severe insulin deficiency 4
- Stable for 3 days at room temperature when collected in boric acid 9
- Well-correlated with serum C-peptide (r=0.82) 4
Important Clinical Pitfalls
Assay variability is a major limitation. Despite traceability to international standards, different C-peptide immunoassays show significant discrepancies—up to 36.6% mean bias between some platforms 10. This means:
- Results near decision thresholds require careful interpretation
- Longitudinal monitoring should use the same assay platform
- The lack of full standardization implementation affects clinical diagnosis 11, 10
When NOT to measure C-peptide:
- Routine screening in type 2 diabetes or cardiovascular risk assessment 1
- Polycystic ovary syndrome evaluation 1
- Within 2 weeks of ketoacidosis 2, 3
- When clinical presentation and treatment response clearly define diabetes type 1
Special populations requiring caution: