Background: We aimed to compare the prevalence of abnormal glucose tolerance identified by the 1985 WHO and the 1997 American Diabetes Association (ADA) diagnostic categories based on information collected in the Cardiovascular Health Study, an epidemiological study of elderly people.
Methods: We measured glucose concentrations during fasting and 2 h after a 75 g oral glucose-tolerance test in participants aged 65-100 years in the Cardiovascular Health Study. From a 1989 cohort, we analysed the glucose measurements of 4515 individuals without a previous diagnosis of diabetes and of 262 additional measurements from an African-American cohort recruited in 1992-93.
Findings: In the 1989 cohort, the prevalence of untreated diabetes with ADA diagnostic fasting criteria was 7.7% versus a prevalence of 14.8% by the WHO criteria. In the African-American cohort, the prevalence of untreated diabetes was 2.7% with ADA criteria and 11.8% with WHO criteria. 3509 (77.7%) of the 4515 participants in the 1989 cohort had normal glucose concentrations according to ADA fasting criteria, compared with 2401 (53.2%) according to WHO criteria. In the African-American cohort, the corresponding numbers were 239 (91.2%) versus 153 (58.4%). All differences in prevalence of abnormal glucose tolerance between ADA and WHO classifications were significant (p<0.0001).
Interpretation: Among elderly individuals, there was a significant difference in the prevalence of diabetes identified by the WHO diagnostic criteria based on oral glucose-tolerance test and the ADA fasting criteria. Consequently, many individuals currently classified as non-diabetic according to ADA criteria would previously have had a diagnosis of diabetes according to WHO criteria. Longitudinal studies are needed to assess the value of the criteria in the identification of individuals at increased risk of diabetes-associated chronic complications.