Diagnostic Criteria for Pre-Diabetes: WHO and ADA Guidelines
Direct Answer
Both the WHO and ADA define pre-diabetes using fasting plasma glucose and 2-hour oral glucose tolerance test thresholds, but they differ on the lower fasting glucose cutoff and HbA1c criteria. 1
ADA Diagnostic Criteria for Pre-Diabetes
The American Diabetes Association recognizes three laboratory thresholds for diagnosing pre-diabetes, any one of which is sufficient: 1, 2
- HbA1c: 5.7–6.4% (39–47 mmol/mol) measured in an NGSP-certified laboratory 1, 2
- Impaired Fasting Glucose (IFG): 100–125 mg/dL (5.6–6.9 mmol/L) after ≥8-hour fast 1
- Impaired Glucose Tolerance (IGT): 140–199 mg/dL (7.8–11.0 mmol/L) at 2 hours during a 75-g oral glucose tolerance test 1
The ADA lowered the IFG threshold from 110 mg/dL to 100 mg/dL in 2003 to equalize the risk of progression to diabetes between IFG and IGT categories. 3
WHO Diagnostic Criteria for Pre-Diabetes
The World Health Organization uses more conservative thresholds and does not endorse HbA1c for pre-diabetes diagnosis: 1, 3
- Impaired Fasting Glucose (IFG): 110–125 mg/dL (6.1–6.9 mmol/L) with 2-hour glucose <140 mg/dL 1
- Impaired Glucose Tolerance (IGT): 140–199 mg/dL (7.8–11.0 mmol/L) at 2 hours during OGTT, with fasting glucose <126 mg/dL 1
- HbA1c is NOT recommended by WHO for diagnosing pre-diabetes 3
The WHO maintains the higher IFG cutoff of 110 mg/dL (6.1 mmol/L) because it identifies individuals at genuinely elevated risk for progression to diabetes, whereas the ADA's lower threshold captures many who will spontaneously revert to normal. 3
Key Differences Between WHO and ADA Criteria
Fasting Glucose Threshold
The most significant divergence is the IFG lower limit: ADA uses 100 mg/dL while WHO uses 110 mg/dL. 1, 3 This 10 mg/dL difference has major epidemiologic implications—the ADA criteria classify approximately 50% more individuals as having pre-diabetes. 3
HbA1c for Pre-Diabetes
The ADA endorses HbA1c 5.7–6.4% for pre-diabetes diagnosis; the WHO explicitly rejects HbA1c for this purpose. 3 The European guidelines and Canadian Diabetes Association have adopted an intermediate position, recommending HbA1c 6.0–6.4% (42–47 mmol/mol) for pre-diabetes. 3
2-Hour OGTT Values
Both organizations agree on the IGT threshold of 140–199 mg/dL at 2 hours, though the WHO places greater emphasis on performing OGTTs to detect IGT, which carries independent cardiovascular risk. 1, 4
Clinical Implications of Differing Criteria
Progression Risk
The WHO criteria identify individuals at higher absolute risk of progression to diabetes. 3 With ADA criteria, approximately one-half of those diagnosed with pre-diabetes test normal on repeat testing, one-third spontaneously revert to normal over time, and two-thirds never develop diabetes in their lifetimes. 3
Test Selection
Fasting plasma glucose alone misses the majority of pre-diabetes cases, particularly in high-risk populations. 5 In a cohort of young African Americans, FPG testing detected only 27.4% of pre-diabetic individuals, whereas a complete OGTT detected 87.1%. 5 Even lowering the IFG threshold to 100 mg/dL identified only 13 of 45 IGT-only cases. 5
Cardiovascular Risk
IGT carries independent cardiovascular disease risk beyond diabetes progression, whereas IFG primarily predicts future diabetes. 4, 6 This distinction supports the WHO's emphasis on OGTT performance when cardiovascular risk stratification is the goal. 6
Practical Testing Recommendations
Which Test to Use
All three tests (HbA1c, fasting glucose, OGTT) are equally appropriate for detecting pre-diabetes according to ADA guidelines. 1, 2 However:
- HbA1c offers the greatest convenience (no fasting required) and superior pre-analytical stability 1, 2
- Fasting plasma glucose is most practical for routine screening but has 12–15% day-to-day variability 2
- OGTT is most sensitive but cumbersome and poorly reproducible 1, 5
When HbA1c Cannot Be Used
Use plasma glucose criteria exclusively in these conditions: 1, 2
- Pregnancy (second/third trimesters)
- Hemoglobinopathies (sickle cell disease/trait)
- Hemodialysis
- Recent blood loss or transfusion
- Erythropoietin therapy
- G6PD deficiency (lowers HbA1c by ~0.7–0.8% in homozygotes) 2
Screening and Follow-Up
Who Should Be Screened
Universal screening should begin at age 45 years; earlier screening is indicated for adults with BMI ≥25 kg/m² (≥23 kg/m² in Asians) plus any additional risk factor: 1, 2
- First-degree relative with diabetes
- High-risk ethnicity (African American, Latino, Native American, Asian, Pacific Islander)
- History of cardiovascular disease
- Hypertension (≥140/90 mmHg or on therapy)
- HDL <35 mg/dL or triglycerides >250 mg/dL
- Polycystic ovary syndrome
- Physical inactivity
- Prior gestational diabetes
- Clinical insulin resistance (severe obesity, acanthosis nigricans)
Testing Intervals
- Normal results: repeat every 3 years minimum 1, 2
- Pre-diabetes diagnosed: annual testing 1, 2
- History of gestational diabetes: lifelong screening every 3 years 1
Common Pitfalls
Over-Reliance on Fasting Glucose
Do not screen high-risk populations with fasting glucose alone—this approach misses 70–85% of IGT cases, particularly in African Americans and other high-risk ethnic groups. 5
Ignoring Test Limitations
Confirm borderline results with repeat testing on a separate day because fasting glucose has 12–15% day-to-day variability that can mimic true metabolic change. 2 When results are near diagnostic thresholds, repeat testing in 3–6 months. 1
Using Point-of-Care HbA1c
Point-of-care HbA1c assays lack the analytical precision required for diagnostic purposes—only NGSP-certified laboratory assays should be used. 2
Misapplying HbA1c in Anemia
Marked discrepancies between HbA1c and plasma glucose should prompt suspicion of unreliable HbA1c results due to altered red blood cell turnover. 1, 2