Diagnosis of Diabetes Using A1C: Two Abnormal Test Results Required
Two abnormal A1C test results (≥6.5%) are required to diagnose diabetes, either from the same sample or from two separate test samples, unless the patient presents with classic symptoms of hyperglycemia or hyperglycemic crisis with a random plasma glucose ≥200 mg/dL. 1, 2
Diagnostic Criteria for Diabetes Using A1C
A1C testing has become a standard method for diagnosing diabetes due to its convenience and reliability. When using A1C for diagnosis:
- A1C must be ≥6.5% (48 mmol/mol) to be considered diagnostic 1
- The test must be performed in a laboratory using a method certified by the NGSP and standardized to the Diabetes Control and Complications Trial (DCCT) assay 1
- Confirmation requires either:
Diagnostic Algorithm
- Initial A1C test ≥6.5%
- Confirmation options:
- If discordant results (e.g., A1C ≥6.5% but FPG <126 mg/dL):
Special Considerations and Limitations
When A1C Should Not Be Used for Diagnosis
A1C is less reliable in conditions with altered red blood cell turnover, including:
- Sickle cell disease or trait (use A1C assay without interference) 1
- Pregnancy (second and third trimesters) 1
- Hemodialysis 1
- Recent blood loss or transfusion 1
- Erythropoietin therapy 1
- HIV treated with certain drugs 1
- Iron-deficient anemia 1
In these conditions, only plasma blood glucose criteria should be used for diagnosis 1.
Test Performance Considerations
- A1C has high specificity (97-98%) but lower sensitivity (47-67%) compared to fasting glucose criteria 3
- When A1C and plasma glucose results are discordant, consider possible A1C assay interference 1
- Point-of-care A1C testing is not standardized for diagnostic purposes 2
- Plasma glucose samples must be spun and separated immediately after collection to prevent falsely low results 1, 2
Exception to the Two-Test Rule
The only scenario where a single test is sufficient for diagnosis is when a patient presents with:
- Classic symptoms of hyperglycemia (polyuria, polydipsia, unexplained weight loss) AND
- Random plasma glucose ≥200 mg/dL (11.1 mmol/L) 1
Clinical Implications
Using A1C for diagnosis offers several advantages including greater convenience (no fasting required) and better preanalytic stability 4. However, the requirement for two abnormal test results is critical to avoid misdiagnosis, as A1C has analytical variability and can be affected by various clinical conditions 1, 2.
For patients with test results near the diagnostic threshold, follow-up testing in 3-6 months is recommended 1, 2.