Relationship between fasting plasma glucose and glycosylated hemoglobin: potential for false-positive diagnoses of type 2 diabetes using new diagnostic criteria

JAMA. 1999 Apr 7;281(13):1203-10. doi: 10.1001/jama.281.13.1203.


Context: New criteria for the diagnosis of type 2 diabetes mellitus have recently been introduced that lowered the diagnostic fasting plasma glucose (FPG) concentration from 7.8 to 7.0 mmol/L (140 to 126 mg/dL).

Objective: To determine if individuals with diabetes diagnosed by the new FPG concentration criterion would have excessive glycosylation (elevated hemoglobin [HbA1c] levels).

Definitions: We determined the distribution of HbA1c levels in individuals using 4 classifications: (1) normal by the new criterion (FPG concentration <6.1 mmol/L [110 mg/dL]); (2) impaired fasting glucose by the new criterion (FPG concentration of 6.1-6.9 mmol/L [110-125 mg/dL]); (3) diabetes diagnosed solely by the new FPG concentration criterion of 7.0 through 7.7 mmol/L (126-139 mg/dL); and (4) diabetes diagnosed by the previous FPG concentration criterion of 7.8 mmol/L (140 mg/dL) or higher.

Design: Cross-sectional analysis of 2 large data sets (NHANES III and Meta-Analysis Research Group [MRG] on the Diagnosis of Diabetes Using Glycated Hemoglobin) that contained individuals in whom FPG concentrations, 2-hour glucose concentrations using an oral glucose tolerance test, and an HbA1c level were simultaneously measured. We cross-tabulated FPG concentrations (<6.1 mmol/L [110 mg/dL], 6.1-6.9 mmol/L [110-125 mg/dL], 7.0-7.7 mmol/L [126-139 mg/dL], and > or =7.8 mmol/L [140 mg/dL]) and HbA1c levels separated into 3 intervals: normal, less than the upper limit of normal (ULN); slightly elevated, ULN to ULN plus 1%; and high, higher than ULN plus 1%.

Results: Among subjects with normal FPG concentrations, HbA1c levels in the NHANES III (and the MRG) data sets were normal in 97.3% (96.2%), slightly elevated in 2.7% (3.6%), and high in 0.1% (0.2%). Among individuals with impaired fasting glucose, HbA1c concentrations were normal in 86.7% (81.4%), slightly elevated in 13.1% (16.4%), and high in 0.2% (2.2%). Among diabetic patients diagnosed by the new FPG criterion only, HbA1c levels were normal in 60.9% (59.6%), slightly elevated in 35.8% (32.8%), and high in 3.4% (7.6%). In diabetic patients diagnosed by the former FPG criterion, HbA1c levels were normal in 18.6% (16.7%), slightly elevated in 32.5% (21.0%), and high in 48.9% (62.3%).

Conclusions: About 60% of the new cohort of diabetic patients in both data sets have normal HbA1c levels. We believe that diabetes should not be diagnosed in those with FPG concentrations less than 7.8 mmol/L (140 mg/dL) unless excessive glycosylation is evident. Individuals without excessive glycosylation but with moderate elevations of FPG concentrations (6.1-7.7 mmol/L [110-139 mg/dL]) should be diagnosed as having impaired fasting glucose and treated with an appropriate diet and exercise. This diagnostic labeling achieves the goal of early intervention without subjecting these persons to the potentially negative insurance, employment, social, and psychological consequences of a diagnosis of diabetes mellitus.

MeSH terms

  • Adult
  • Blood Glucose / analysis*
  • Cross-Sectional Studies
  • Diabetes Mellitus, Type 2 / blood
  • Diabetes Mellitus, Type 2 / diagnosis*
  • False Positive Reactions
  • Fasting / blood
  • Glucose Tolerance Test
  • Glycated Hemoglobin A / analysis*
  • Humans


  • Blood Glucose
  • Glycated Hemoglobin A