Type 2 diabetes in children was unheard of a generation ago. Even though very rare event today, the obesity epidemic is placing kids at risk.
The incidence of type 2 diabetes mellitus in children and adolescents has increased dramatically over the past 2 decades.1,2 SEARCH for Diabetes in Youth-a US population-based study-reported that the number of new cases of type 2 diabetes as a proportion of total new cases of diabetes in older children ranges from 15% for non-Hispanic white children to 86% for American Indians.3 On the basis of these results, SEARCH has estimated that approximately 3,700 new cases of type 2 diabetes are diagnosed annually in US children aged 10 to 19 years.5
Because type 2 diabetes can be asymptomatic during its early stages, and undiagnosed diabetes is common in adults, it had been assumed that the condition would also be underdiagnosed in children. Guidelines for screening children at risk of developing glucose intolerance and diabetes based on combinations of body mass index (BMI), ethnicity, family history of cardiovascular disease, and signs of insulin resistance have been developed by the American Diabetes Association (ADA) and the International Society for Pediatric and Adolescent Diabetes (ISPAD).5,6
The oral glucose tolerance test
In one nationally representative sample of US adolescents aged 12 to 19 years, no cases of type 2 diabetes were identified.6 Similarly, in a study of 1,740 high-risk, obese, minority eighth-grade students, only 0.1% had diabetes, identified by an oral glucose tolerance test (OGTT).8 Such findings raise the question of whether screening asymptomatic children for diabetes is appropriate or cost effective. Given the continuing low prevalence of type 2 diabetes, even among high-risk children, a large number of OGTTs would need to be performed to identify a case that was not clinically apparent.
In addition, because the OGTT is not reliable enough for a diagnosis in an asymptomatic patient to be made based on a single test, repeat testing is needed and rarely confirms the original abnormal result.9 Thus, because of the low concordance between repeated OGTTs and the low a priori prevalence of the disorder, the positive predictive value of an abnormal OGTT is low in this population.
In the past, the ADA had not recommended the use of hemoglobin A1c (A1c) as a diagnostic test, largely because of the lack of standardization of the assay.10 However, by 2010, the ADA recognized that laboratory-based (not point-of-care) A1c was being measured in most laboratories in the United States in adherence to the National Glycohemoglobin Standardization Program. Furthermore, an international expert panel convened to review this question noted that review of epidemiologic data supported a relationship between A1c and the risk of retinopathy similar to what had been shown for fasting plasma glucose (FPG) and 2-hour OGTT.11 Therefore, in 2010, the ADA added an A1c of 6.5% or higher as a diagnostic criterion for diabetes.
There are several potential practical benefits for the use of A1c as a component of a type 2 diabetes diagnosis. First, the patient does not need to be fasting, and testing does not require a return visit to the physician or laboratory.12 In addition, A1c has less variability and is more reproducible than FPG and OGTT. Finally, A1c is less subject to acute changes in glycemia because of stress-induced hyperglycemia or short-term exposure to medications.
Conversely, A1c does not provide a perfect estimation of mean blood glucose.13 Diseases such as iron-deficiency anemia, sickle-cell disease, thalassemia, and other hemoglobinopathies will alter serum A1c values because of hemoglobin variants and because of transfusions.14 Perhaps for this reason, A1c values are higher in blacks than in whites for the same mean blood glucose over a 3-month period.15
Several recent studies have pointed out additional potential flaws of using A1c for the diagnosis of diabetes, specifically in the pediatric population, including low sensitivity, poor negative predictive value, and poor operating characteristics for identifying diabetes and prediabetes diagnosed by FPG or OGTT.16-18 Taken together, these studies suggest that A1c may lead to misdiagnosis in children compared with OGTT or FPG.
However, studies comparing A1c to existing methods of diagnosing diabetes in pediatric populations are handicapped by the fact that FPG and OGTT are themselves not validated in this population.19 A truly validated definition of diabetes requires insight into the relationship of the proposed definitions to relevant aspects of medium-term and long-term health.
Such studies have not been available in pediatric patients because of the long time frames required to correlate disease definition thresholds to vascular complications. Thus, although the numbers of pediatric patients diagnosed by A1c, FPG, and OGTT differ, it is not yet clear which of these measures will be more closely related to long-term complications, the only true gold standard for defining glucose abnormalities.
Despite current guidelines for screening, the low prevalence of undiagnosed diabetes, even among otherwise high-risk adolescents, raises questions about the value of these recommendations in asymptomatic persons. The use of A1c as a screening test in obese adolescents offers substantial logistical benefits and, thus, reduces the cost and burden of FPG- and OGTT-based screening, but the results must still be interpreted with caution given the low prevalence of disease.