Background: Capture-recapture methods have been widely employed in the study of wildlife populations and have recently been applied to count various human diseases and conditions. We have estimated the prevalence of type 2 diabetes mellitus by adjusting for the degree of undercount using a two-sample model of capture-recapture among men and women aged 50-69 in Oguni town, Japan.
Methods: Oguni town diabetes registry data were utilized as the first source. In the registry, only those who had experienced fasting plasma glucose of > or = 7.8 mmol/l (140 mg/dl) or 2 h plasma glucose after a 75 g oral glucose tolerance test (OGTT) of > or = 11.1 mmol/l (200 mg/dl) were counted as having diabetes. A second source was a sample study selecting 200 men and 200 women aged 50-69 randomly, which was conducted in August 1991. A 75 g OGTT was done in the morning. The 1985 World Health Organization criteria were used to classify the diabetes status of the participants. A two-sample model of capture-recapture methods was employed to estimate the total number of cases of diabetes and determine the ascertainment rates of the registry.
Results: The prevalence estimated by the diabetes registry was 7.1%. The prevalence from the sample study was 8.8% with a participation rate of 74%. Estimated prevalence employing the capture-recapture method was 13.1%. The ascertainment rate of the registry was 53.8%.
Conclusions: Little is known about the prevalence of type 2 diabetes in local areas in Japan, the US and the world. Capture-recapture methods are likely to provide a means to accurately assess the prevalence of diabetes.