Background: We report the prevalence of diabetes in a rural, multiethnic community in Hawaii, of predominantly Asian and Native Hawaiian ancestry, by using 1997 World Health Organization diagnostic criteria applied to a two-hour oral glucose tolerance test.
Methods: This cross-sectional survey included 1452 men and nonpregnant women who were >18 years of age. Blood was drawn in the fasting and postchallenge states. Individuals under pharmacologic treatment for diabetes were excluded. Information obtained included demographics, medical history, dietary intake, physical activity, and anthropometric measurements.
Results: Prevalence of diabetes was approximately three-fold higher among Asian and Native Hawaiian ancestry groups than among Caucasians, even after adjusting for other risk factors. Furthermore, diabetes prevalence was similar among all non-Caucasian ethnic groups despite significant differences in body mass indices.
Conclusions: These findings indicate that earlier reports of high prevalence of diagnosed diabetes among Asians and Hawaiian ethnic groups were not due to detection bias, since our study revealed similar prevalence of previously unrecognized diabetes. Furthermore, similar prevalence among these groups was observed despite significant differences in body mass indices, diet, and physical activity. This apparent paradox may reflect limitations in the measurement of these risk factors; differences in the impact of these risk factors on diabetes risk in different ethnic groups; or ethnic differences in lifestyle, biochemical, or genetic factors that were not examined in this study.