The prevalence of type 2 diabetes is increasing in developed countries. The failure to recognize and treat diabetes in the general population is a cause of increased morbidity and mortality. In 1997, the American Diabetes Association (ADA) adopted new criteria for the diagnosis of diabetes by requiring a fasting blood glucose level of at least 126 mg per dL (7.0 mmol per L) instead of 140 mg per dL (7.8 mmol per L). The ADA also suggested that oral glucose tolerance tests are not necessary in routine practice for the diagnosis of diabetes. These criteria were suggested to decrease the cases of undiagnosed diabetes by using more aggressive and simpler screening criteria. The ADA recommendations identify borderline patients who have diabetes with fasting blood glucose levels between 126 and 139 mg per dL (7.0 and 7.7 mmol per L). Tenenbaum and colleagues evaluated the prevalence and prognostic significance of unrecognized and borderline diabetes in patients with healed myocardial infarction or stable angina pectoris.

The final study included 14,629 patients screened for participation in the Bezafibrate Infarction Prevention (BIP) study. The majority of patients were men. All patients underwent detailed evaluation and were followed for a mean of 7.7 years. There were 335 patients without a previous history of diabetes who were found to have borderline diabetes (fasting blood glucose levels of 126 to 139 mg per dL), and 487 patients presenting with previously undiagnosed diabetes (fasting blood glucose levels of at least 140 mg per dL).

All-cause ischemic heart disease and cerebrovascular accident mortality rates were lower in patients without diabetes than in patients with diabetes. Both undiagnosed and established (but not borderline) diabetes were associated with a significant stroke-related mortality excess. Among patients previously undiagnosed with diabetes, a surprising 4.2 percent were found to be hyperglycemic on routine fasting blood glucose analysis.

The authors conclude that it is unclear whether the identification of borderline diabetes better identifies patients with established coronary artery disease who are at risk for increased long-term mortality. There is, however, a clear independent association between diagnosed and undiagnosed diabetes and increased mortality. The greatest impact of undiagnosed and established diabetes is in stroke mortality excess.