It is believed that many diabetes cases—approximately 25%—in the United States are undiagnosed (1, 2). This conclusion is based on epidemiologic surveys of respondents who had 1 test of glycemia to measure their level of fasting plasma glucose (FPG), hemoglobin A_1c (HbA_1c), or 2-hour glucose as part of a glucose tolerance test. However, for persons with no diabetes symptoms, the American Diabetes Association (ADA) currently recommends repeating a positive test to confirm the diagnosis. Confirmation may be done by using the same test (for example, a repeated measurement of FPG level if the previous one was elevated) or a second evaluation, such as an HbA_1c test (which often is recommended because it is helpful in making treatment decisions) (3). This approach is recommended, in part, because all tests have variability that may produce values above and below a diagnostic cut point when repeated (3). In terms of clinical measures of glycemia, the 2-hour oral glucose tolerance test value is the least reproducible and thus is not recommended for nonpregnant persons (3). Measurements of FPG and HbA_1c levels are more reliable, although several factors, such as race, anemia, and hemoglobinopathy, may affect glycation (4).

据信，在美国，许多糖尿病病例（约25％）未得到诊断（1、2）。该结论基于对接受1次血糖测试以测量其空腹血浆葡萄糖（FPG），血红蛋白A _1c（HbA _1c）或2小时葡萄糖水平作为糖耐量测试一部分的受访者的流行病学调查。但是，对于没有糖尿病症状的人，美国糖尿病协会（ADA）当前建议重复进行阳性试验以确认诊断。可以通过使用相同的测试（例如，如果先前的FPG水平升高则重复测量FPG水平）或第二次评估（例如HbA _{1c）来进行确认}测试（通常建议这样做，因为它有助于做出治疗决定）（3）。之所以建议使用这种方法，部分原因是，所有测试均具有可变性，重复进行时可能会产生高于和低于诊断切点的值（3）。就血糖的临床测量而言，2小时口服葡萄糖耐量测试值的再现性最低，因此不建议非孕妇服用（3）。尽管种族，贫血和血红蛋白病等多种因素可能会影响糖基化，但测量FPG和HbA _1c的水平更为可靠（4）。