Screening for Type 2 Diabetes Mellitus in Asymptomatic Adults
In the United States, 5.4 million people have undiagnosed diabetes. This number represents approximately one-third of all diabetes. At diagnosis of type 2 diabetes, 239% of patients have diabetic retinopathy, 818% have nephropathy, 513% have neuropathy, and approximately 8% have cardiovascular disease. Based on the large number of people with undiagnosed diabetes and the high prevalence of complications at diagnosis, it seems intuitive that earlier diagnosis and treatment should reduce the burden of diabetes and its complications. Unfortunately, at the present time, there is no evidence of benefit from screening asymptomatic adults for undiagnosed diabetes.
Clearly, if a person exhibits symptoms of diabetes such as polydipsea, polyuria, nocturia, or unintentional weight loss, fasting glucose testing is indicated. Such testing, when based on symptoms or other clinical cues, must be distinguished from screening. Screening is defined as the examination of asymptomatic individuals to differentiate an individual at high risk from one at low risk of having a disease. Screening the asymptomatic population for undiagnosed type 2 diabetes requires substantial resources, and the yield (that is, the proportion of people who are screened who subsequently undergo definitive diagnostic testing, are found to have diabetes, and receive follow-up care) has generally been low.
The greatest opportunity for case-finding is in the clinical setting. Clinicians should have a high index of suspicion for undiagnosed diabetes in individuals with risk factors such as advanced age, obesity, physical inactivity, history of abnormal glucose tolerance (including gestational diabetes), hypertension, dyslipidemia, and family history of diabetes. In such individuals, periodic glucose testing should be performed.
Equally important is correct interpretation of the results of the testing performed. Fasting glucose levels >110 mg/dl are clearly suspicious for undiagnosed type 2 diabetes. Fasting plasma glucose levels persistently >126 mg/dl are diagnostic. Likewise, nonfasting, random, or postprandial plasma glucose levels >140 mg/dl are suspicious and require follow-up, usually with a fasting plasma glucose.
In retrospect, it is not unusual to discover multiple earlier plasma glucose levels between 110 and 200 mg/dl in the medical records of newly diagnosed patients. Often, these tests were obtained as part of "routine" biochemical panels but were simply not recognized as being suspicious and, thus, not acted upon. In many instances, the failure to make a timely diagnosis of type 2 diabetes is not so much a lack of testing as the failure to appropriately interpret the results and follow-up on abnormal test results.
The American Diabetes Association position statement that appears on the following pages provides a useful guide for clinical case-finding. In reviewing the position statement, it is important to recognize that no definitive studies on the benefits and risks of screening are available and that these recommendations rely on expert opinion and consensus. Questions remain about the best screening test, the cut points that provide optimal sensitivity and specificity, the optimal frequency of testing, and the impact of early treatment on long-term outcomes. Certainly, however, the large burden of undiagnosed diabetes, the high prevalence of complications at clinical diagnosis of type 2 diabetes, and the proven effectiveness of improved glycemic management in preventing the development and progression of the microvascular and neuropathic complications in type 2 diabetes support such testing. If benefits are to accrue from early detection, appropriate interpretation of test results, follow-up, and treatment must occur.
William H. Herman, MD, MPH, is an associate professor of internal medicine at the University of Michigan in Ann Arbor. He is an associate editor of Clinical Diabetes. Michael M. Engelgau, MD, MS, is chief of the Epidemiology and Statistics Branch of the Division of Diabetes Translation at the Centers for Disease Control and Prevention in Atlanta, Ga
Copyright © 2000American Diabetes
For Technical Issues contact firstname.lastname@example.org