Abstract
BACKGROUND: Diabetes is an important risk factor for cardiovascular disease, certain types of cancer, and death, and self-reports are one of the most convenient methods for ascertaining diabetes status. We evaluated the validity of diabetes self-reports among Japanese who participated in a health checkup. METHODS: Self-reported diabetes was cross-sectionally compared with confirmed diabetes among 2535 participants aged 28 to 85 years in the Saku cohort study. Confirmed diabetes was defined as the presence of at least 1 of the following: fasting plasma glucose (FPG) level of 126 mg/dL or higher, 2-hour post-load glucose (2-hPG) level of 200 mg/dL or higher after a 75-gram oral glucose tolerance test, glycated hemoglobin (HbA1c) level of 6.5% or higher, or treatment with hypoglycemic medication(s). RESULTS: Of the 251 participants with self-reported diabetes, 121 were taking hypoglycemic medication(s) and an additional 69 were classified as having diabetes. Of the 2284 participants who did not self-report diabetes, 80 were classified as having diabetes. These data yielded a sensitivity of 70.4%, a specificity of 97.3%, a positive predictive value of 75.7%, and a negative predictive value of 96.5%. The frequency of participants with undiagnosed diabetes was 3.0%. Of these, 64.2% had FPG within the normal range and were diagnosed by 2-hPG and/or HbA1c. CONCLUSIONS: Our findings provide additional support for the use of self-reported diabetes as a measure of diabetes in epidemiologic studies performed in similar settings in Japan if biomarker-based diagnosis is difficult.