Similar 9-year mortality risks and reproducibility for the World Health Organization and American Diabetes Association glucose tolerance categories -The Hoorn study

OBJECTIVE- To compare the risks of all-cause and cardiovascular disease (CV D) mortality in the American Diabetes Association (ADA) and World Health Or ganization (WHO) glucose tolerance categories after 9 years of follow-up in the Hoorn Study and to study the test-retest reproducibility of those cate gories. RESEARCH DESIGN AND METHODS- In this population-based cohort study of 2,468 elderly men and women, subjects were classified according to both the WHO and the ADA criteria. Causes of death were extracted from the medical recor ds. Age- and sex-adjusted relative risks were estimated by Cox's proportion al hazards model. Reproducibility of the diagnostic criteria was assessed i n a sample of 1,109 subjects with duplicate oral glucose tolerance rests. RESULTS- Subjects with known diabetes had a four to five times higher risk of all-cause and CVD mortality compared with normal subjects (P < 0.05). Th e relative risks of all-cause mortality were 1.67 (95% CT 1.09-2.57) and 1. 56 (1.00-2.43) for newly diagnosed diabetic subjects according to the WHO a nd ADA criteria, respectively. The WHO and ADA criteria had similar levels of reproducibility. The overall kappa was 0.59 (0.54-0.64) for WHO criteria and 0.61 (0.56-0.66) for ADA criteria. For the category of newly diagnosed diabetes according to WHO or ADA, the percentages of agreement for the sec ond test compared with the first test were 77% (85/110) and 74% (74/100), r espectively. CONCLUSIONS- Both sets of diagnostic criteria identify criteria-specific di abetic subjects with an increased mortality risk compared with normal subje cts, and the reproducibility both criteria is similar.