Can Hemoglobin A1c (HbA1c) levels be used to diagnose diabetes?

Can HbA1c Be Used to Diagnose Diabetes?

Yes, HbA1c ≥6.5% (≥48 mmol/mol) can definitively diagnose diabetes, but the test must be performed using an NGSP-certified method in an accredited laboratory, and confirmation with a repeat test is recommended unless the patient presents with unequivocal hyperglycemia. 1

Diagnostic Threshold and Requirements

• An HbA1c value of ≥6.5% (≥48 mmol/mol) is diagnostic for diabetes 1
• The test must be performed using a method certified by the National Glycohemoglobin Standardization Program (NGSP) and standardized to the Diabetes Control and Complications Trial (DCCT) assay 1
• Confirmation requires either a repeat HbA1c test on a different sample or a glucose-based test (fasting glucose or OGTT), unless the patient has classic symptoms of hyperglycemia with random glucose ≥200 mg/dL 1, 2

Critical Advantages of HbA1c Testing

HbA1c offers several practical benefits over glucose-based testing:

• No fasting required, making it more convenient for patients 1
• Greater preanalytical stability compared to glucose samples 1
• Less day-to-day biological variability and less affected by acute stress, diet changes, or illness 1
• Reflects average glucose levels over the preceding 60-90 days 1

When HbA1c CANNOT Be Used for Diagnosis

You must use only plasma glucose criteria (not HbA1c) in the following conditions that falsely lower HbA1c: 1, 2

• Hemoglobinopathies including sickle cell disease 1, 2
• Pregnancy (second and third trimesters and postpartum period) 1, 2
• Conditions with increased red blood cell turnover:
  • Recent blood loss or transfusion (within 2-3 months) 1, 2
  • Hemolytic anemias 2
  • Hemodialysis 1, 2
  • Erythropoietin therapy 1, 2
• Glucose-6-phosphate dehydrogenase (G6PD) deficiency 1, 2

Point-of-Care Testing Limitations

• Point-of-care HbA1c devices should NOT be used for diagnosis unless they are specifically FDA-approved for diagnostic purposes and performed in CLIA-certified laboratories of moderate complexity or higher 1
• Most point-of-care devices are only validated for monitoring glycemic control, not diagnosis 1

Important Caveats and Pitfalls

Marked discordance between HbA1c and plasma glucose levels should raise suspicion for assay interference and prompt consideration of using an interference-free assay or relying on plasma glucose criteria instead 1, 2

Racial and Ethnic Considerations

• African American individuals may have HbA1c levels 0.3-0.8% higher than non-Hispanic White individuals with similar glucose levels, independent of hemoglobin variants 2
• This can impact both diagnosis and treatment decisions 2

Sensitivity Limitations

• HbA1c at the 6.5% threshold diagnoses only about 30% of diabetes cases that would be identified collectively using HbA1c, fasting glucose, or 2-hour glucose testing 1
• HbA1c has lower sensitivity compared to glucose-based criteria, meaning it may miss some individuals with diabetes 1, 3, 4

Prediabetes Diagnosis

• HbA1c levels of 5.7% to 6.4% (39-46 mmol/mol) indicate prediabetes or high risk for developing diabetes 1
• This range identifies individuals who would benefit from diabetes prevention interventions 1

Gestational Diabetes Exception

HbA1c is NOT recommended for screening or diagnosing gestational diabetes mellitus 1