Oral Glucose Tolerance Testing in Asymptomatic Obese Children: More Questions than Answers

Abstract

The prevalence of type 2 diabetes is increasing among youth and adults throughout the world, with striking rates of increase in rapidly developing countries (1–3). In adults, type 2 diabetes may remain undiagnosed for several years because hyperglycemia develops gradually and may not cause symptoms (2, 4). Therefore, it has been estimated that nearly 50% of prevalent cases of diabetes among adults remain undiagnosed (2, 5) Yet, it is also known that the risk for cardiovascular morbidity and mortality is increased well before the diagnosis of diabetes is made, including in subjects with impaired fasting glucose (IFG) or impaired glucose tolerance (IGT) (6). Therefore, the challenge of developing cost-effective approaches to the identification of undiagnosed diabetes among adults and the value of identifying IFG and IGT remain areas of substantial controversy. The challenge is even greater among youth. Although the obesity prevalence among children in the United States ranges from 10–30% (7), studies suggest that the prevalence of undiagnosed diabetes is surprisingly low (8), even among youth with multiple risk factors. Thus, there is little evidence to support wide-scale screening for diabetes among asymptomatic youth, because this would be a labor-intensive and expensive task with uncertain benefit, even when multiple risk factors are present. On the other hand, 40% of a cohort of eighth-graders had IFG, and rates of IFG and IGT are even higher among youth referred for evaluation in obesity clinics (9). Yet, the prognostic and therapeutic value of identifying IFG and IGT in youth is controversial, and the potential harm and long-term effects of labeling a population as abnormal are unknown. Furthermore, although there is evidence that the combination of IFG and IGT in youth is a strong predictor of progression to overt diabetes in the short term (10), no trial has yet established either that systematic screening for early glucose abnormalities in obese children or that early treatment improves health outcome. Nevertheless, there are clearly circumstances in which testing for undiagnosed diabetes, along with the identification of IFG and IGT may be desirable. For example, the identification of dysglycemia is important for establishing cohorts for research into the pathophysiology of these states in adolescents as well as for the exploration of potential approaches to intervention. Similarly, limited availability may require that clinical resources for the treatment of obese adolescents be focused on those most at risk for disease progression and comorbidities. However, a number of important questions remain. Who should be screened? Which test should be used? How often should the screening be repeated? Determination of hemoglobin A1c is not sufficiently sensitive to detect early or subtle abnormalities in glucose excursion and has no clear utility in identification of IFG and IGT. Furthermore, the utility of this simple measure of average glucose has been limited in the past by poor reproducibility and lack of standardization (11, 12). Measurement of fasting glucose allows the identification of both IFG and diabetes. However, ensuring adherence to fasting guidelines can be challenging in an obese adolescent population. Furthermore, the 2-h postload plasma glucose has been shown to have increased sensitivity for the diagnosis of both diabetes and IGT in adults (5) and children (9). In addition, it has been proposed that the oral glucose tolerance test (OGTT) is preferable to measurement of fasting glucose because IGT, which can only be identified through glucose challenge, has also been associated with an increased risk for development of cardiovascular disease (13, 14). However, the reproducibility of the test has been questioned. In this issue of the Journal, Ingrid Libman and her colleagues at the University of Pittsburgh (15) report a prospective study of the reproducibility of the OGTT in overweight children and confirm what has previously been reported in adults, namely that the reproducibility of the test is poor, particularly among those adolescents with the greatest degree of apparent abnormality. Clinically, these results mean that an isolated abnormal postload plasma glucose is not sufficient for the diagnosis of diabetes or impaired glucose metabolism in the pediatric population, further complicating the argument between the use of fasting glucose