What are the diagnostic criteria for diabetes?

Diagnostic Criteria for Diabetes Mellitus

Diabetes is diagnosed when any one of four criteria is met: A1C ≥6.5%, fasting plasma glucose ≥126 mg/dL, 2-hour plasma glucose ≥200 mg/dL during an oral glucose tolerance test, or random plasma glucose ≥200 mg/dL with classic hyperglycemic symptoms. 1, 2

Primary Diagnostic Thresholds

The American Diabetes Association establishes four distinct pathways to diagnosis, each with specific technical requirements: 1

1. Hemoglobin A1C ≥6.5% (≥48 mmol/mol)

• Must be performed in a laboratory using a method certified by the National Glycohemoglobin Standardization Program (NGSP) and standardized to the Diabetes Control and Complications Trial (DCCT) assay 1, 2
• Point-of-care A1C testing should be restricted to FDA-approved devices at CLIA-certified laboratories when used for diagnosis 3

2. Fasting Plasma Glucose (FPG) ≥126 mg/dL (≥7.0 mmol/L)

• Fasting is defined as no caloric intake for at least 8 hours 1, 2
• Samples must be spun and separated immediately after collection to avoid preanalytic variability 1

3. 2-Hour Plasma Glucose ≥200 mg/dL (≥11.1 mmol/L) During OGTT

• Test performed using a 75-g anhydrous glucose load dissolved in water, as described by the World Health Organization 1, 2
• Patient must fast for 8 hours before the test 3
• Individuals should consume a mixed diet with at least 150g of carbohydrates in the 3 days preceding the test to avoid falsely elevated glucose levels 2

4. Random Plasma Glucose ≥200 mg/dL (≥11.1 mmol/L)

• Only diagnostic when accompanied by classic symptoms of hyperglycemia: polyuria, polydipsia, unexplained weight loss 1, 2
• Or in the setting of hyperglycemic crisis 1, 2

Confirmation Requirements

In the absence of unequivocal hyperglycemia with acute metabolic decompensation, diagnosis requires two abnormal test results. 1, 3, 4

The confirmation strategy depends on the clinical scenario: 1

• Same test repeated: If A1C is 7.0% and repeat is 6.8%, diabetes is confirmed 1
• Two different tests both elevated: If both A1C and FPG are above diagnostic thresholds (even from different samples), diabetes is confirmed 1, 3
• Discordant results: If one test is above and another below the threshold, repeat the test that was elevated 1, 3
• No confirmation needed: When random plasma glucose ≥200 mg/dL occurs with classic hyperglycemic symptoms or hyperglycemic crisis 1, 4

Important Caveat on Test Variability

Preanalytic and analytic variability means a repeated test may fall below the diagnostic threshold even when initially elevated. 1 This is least likely with A1C, somewhat more likely with FPG, and most likely with 2-hour plasma glucose. 1 For patients with borderline results, consider close follow-up and repeat testing in 3-6 months. 1

Critical Limitations of A1C Testing

A1C should NOT be used for diagnosis in conditions affecting red blood cell turnover; use only plasma glucose criteria in these situations. 1, 2, 3

Conditions where A1C is unreliable include: 1, 2, 4

• Hemoglobinopathies (sickle cell disease, thalassemia)
• Pregnancy (second and third trimesters)
• Glucose-6-phosphate dehydrogenase deficiency
• Hemodialysis
• Recent blood loss or transfusion
• Erythropoietin therapy
• HIV treated with certain drugs
• Iron-deficiency anemia

Racial and Ethnic Considerations

African Americans may have A1C levels approximately 0.4% higher than non-Hispanic whites at similar glucose levels, though the association with complications appears similar across groups. 1 Genetic variants like G6PD G202A (carried by 11% of African Americans) can decrease A1C by 0.7-0.8% in homozygotes. 1

Prediabetes Categories (Increased Risk for Diabetes)

Three distinct categories identify individuals at high risk who require intervention: 1, 2, 3

• Impaired Fasting Glucose (IFG): FPG 100-125 mg/dL (5.6-6.9 mmol/L) 1, 2, 3
• Impaired Glucose Tolerance (IGT): 2-hour plasma glucose 140-199 mg/dL (7.8-11.0 mmol/L) during OGTT 1, 2, 3
• Elevated A1C: 5.7-6.4% (39-47 mmol/mol) 2, 3, 4

Risk is continuous across these ranges, becoming disproportionately greater at the higher ends. 1

Practical Algorithm for Diagnosis

Step 1: Select initial test based on clinical context

• If classic symptoms present: Random plasma glucose (immediate diagnosis if ≥200 mg/dL) 3
• If asymptomatic: A1C or FPG (most practical for routine screening) 1
• If mild FPG elevation or high clinical suspicion: Consider OGTT 5, 6

Step 2: Confirm diagnosis (unless hyperglycemic crisis)

• Repeat the same test on a different day, OR 1
• Perform a different test if both are elevated 1

Step 3: Assess for A1C interference

• If marked discrepancy between A1C and plasma glucose, suspect hemoglobin variants 2, 3
• Switch to plasma glucose criteria if interference suspected 1, 2

Common Pitfalls to Avoid

Pitfall 1: Using A1C alone in populations with hemoglobinopathies or conditions affecting red blood cell turnover—this can lead to misdiagnosis in either direction. 1, 2

Pitfall 2: Failing to confirm diagnosis with repeat testing when there are no acute symptoms—single elevated values require confirmation to avoid false-positive diagnoses. 1, 3

Pitfall 3: Relying solely on FPG for screening—this approach misses many diabetic patients, particularly in Asian and European populations where postprandial hyperglycemia predominates. 5, 6

Pitfall 4: Allowing glucose samples to remain at room temperature without prompt centrifugation—this causes falsely low glucose readings due to ongoing glycolysis. 1

Pitfall 5: Performing OGTT without adequate carbohydrate loading—patients need at least 150g carbohydrates daily for 3 days before testing to avoid falsely elevated results. 2

Gestational Diabetes Mellitus (GDM)

Screen pregnant women not previously known to have diabetes at 24-28 weeks gestation using a 75-g 2-hour OGTT. 1 Women with risk factors should be screened at the first prenatal visit using standard diagnostic criteria. 1 Screen women with GDM for persistent diabetes at 6-12 weeks postpartum using a test other than A1C. 1