ESPE Abstracts (2014) 82 P-D-2-3-397

Combined Evaluation of Glucose Levels at Fasting State and 1-H After Glucose Load Can Safely Predict Prediabetes in Obese Youth

Eleni P Kotanidou, Ioannis Kyrgios & Assimina Galli-Tsinopoulou

4th Department of Pediatrics, Faculty of Medicine, Papageorgiou General Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece

Background: According to current literature 1-h glucose ≥155 mg/dl during an oral glucose tolerance test (OGTT) is considered an independent predictor of β-cell deterioration and future progression to prediabetes in obese individuals.

Objective and hypotheses: To investigate whether determination of glucose level obtained 1-h after glucose load can improve predictive ability of prediabetes.

Method: Ninety-eight overweight/obese youngsters aged 2.6–16.5 years were included for analysis. History taking, physical examination and an OGTT were performed. In order to predict prediabetes, defined as impaired fasting glucose and/or impaired glucose tolerance, four models were evaluated: i) a basic model including demographics (gender, pubertal stage, BMI percentile, ethnic origin, hypertension, hypercholesterolemia, acanthosis nigricans, history of gestational diabetes, parental family history of diabetes type 2 or obesity and birth weight below 5th percentile-for-gestational-age), ii) a second one combining the basic model plus measurement of fasting glucose, iii) a third one that compiled the basic model and the 1-h glucose level obtained during the OGTT, and iv) a fourth model that incorporated both fasting and 1-h glucose level into the basic model.

Results: Prediabetes was diagnosed in 67/98 participants (69.4%). The basic model significantly contributed to prediction of prediabetes (area under the curve (AUC): 0.74, 95% CI: 0.636–0.851). The additional determination of fasting glucose or 1-h glucose level obtained during the OGTT also increased the predictive ability (AUC: 0.84, 95 %CI: 0.75–0.92 and AUC: 0.80, 95 %CI: 0.70–0.89 respectively).The fourth model, comprised of all previous parameters (basic model+fasting+1-h glucose), predicted prediabetes to even a further degree (AUC: 0.87, 95 %CI: 0.8–0.94).

Conclusion: The 1-h glucose appears to be a reliable parameter to detect prediabetes, especially in combination with demographic and clinical characteristics as well as fasting glucose levels.

Article tools

My recent searches

No recent searches.