Self-awareness of diabetes is the first step to prevent this non-curable disease. Therefore, it is important to develop diabetes risk scores which can be easily calculated by general people. We developed self-assessable diabetes risk score using the data of Korean Genome and Epidemiology Study (KoGES). A total of 8,740 subjects without diabetes at baseline were followed-up biannually over 10 years. We assessed the significantly different clinical and non-laboratory variables at baseline between subjects developed diabetes mellitus and those free from diabetes at 10-year follow-up. Each of variables generated a maximum score of 100 in each gender group. Next, the 10-year probability of incident diabetes in each score was calculated. During a median follow-up of 9.7 years, 22.7% of subjects progressed to diabetes. The self-assessable diabetes risk scores included age, living location, smoking history, hypertension, family history of diabetes and waist circumference. The developed risk score yielded acceptable discrimination for incident diabetes in both male (AUC 0.661) and female (AUC 0.691). If the subject had a score more than 50, the probability of 10-year incident diabetes would be 43.4% and 36.1% in male and female, respectively.
In conclusion, our newly developed self-assessable diabetes risk score is easily applicable to people without diabetes to predict their future risk of diabetes by themselves without laboratory tests. This model would be helpful to make a diabetes prevention program and to identify high risk subjects who should undergo laboratory tests more frequently. Further studies will be needed to validate the model and to test its feasibility in real clinical settings.
T. Oh: None. J. Moon: None. S. Choi: None. S. Lim: None. Y. Cho: Consultant; Self; Hanmi Pharm. Co., Ltd. Research Support; Self; AstraZeneca, Sanofi-Aventis. K. Park: None. N.H. Cho: None. H. Jang: None.
Korea Centers for Disease Control & Prevention