Context: The optimal screening test for diabetes and prediabetes in obese youth is controversial.
Objective: We examined whether glycosylated hemoglobin (HbA1c) or the oral glucose tolerance test (OGTT) is a better predictor of free-living glycemia as measured by continuous glucose monitoring (CGM).
Design: This was a cross-sectional study of youth 10-18 years old, body mass index (BMI) 85th percentile or greater, with diabetes risk factors.
Setting and participants: Participants (n = 118) with BMI 85th percentile or greater, not on medications for glucose management, were recruited from primary care and pediatric endocrinology clinics around Denver, Colorado.
Intervention: HbA1c, fasting plasma glucose, and 2-hour glucose were collected and all participants wore a blinded CGM for 72 hours.
Main outcome measures: CGM outcomes were determined and descriptive statistics calculated. Performance characteristics at current American Diabetes Association cutpoints were compared with CGM outcomes.
Results: CGM data were successfully collected on 98 obese youth. Those with prediabetes had significantly higher average glucose, area under the curve (AUC), peak glucose, and time greater than 120 and greater than 140 mg/dL (P < .01) on CGM than youth with normal HbA1c or OGTT. HbA1c had a greater magnitude of correlation to CGM average glucose, AUC, and minimum glucose; 2-hour glucose had a greater magnitude of correlation to CGM SD, peak glucose, and time greater than 140 and greater than 200 mg/dL. However, there were no overall differences in the strength comparisons between 2-hour glucose and HbA1c correlations to CGM outcomes.
Conclusions: In obese youth, HbA1c and 2-hour glucose performed equally well at predicting free-living glycemia on CGM, suggesting that both are valid tests for dysglycemia screening.