Abstract
AIM: To validate the Diabetes Risk Assessment in Dentistry Score (DDS) in a US population-based sample and compare its performance with the American Diabetes Association (ADA) risk calculator and the Leicester Risk Assessment (LRA). METHODS: Data were obtained from the National Health and Nutrition Examination Survey (NHANES) covering the 2009-2014 cycles. The study focused on participants aged 40 years and older who included complete data for DDS and ADA risk score assessment. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC), while decision curve analysis (DCA) was used to evaluate the clinical utility of the model. RESULTS: Of the 6259 participants included, the average age was 57.3 years (±12.0 years, range: 40-80 years) and the sample was evenly distributed by sex (50.8% female). DDS showed limited discriminative ability with an AUC of 0.64 (95% confidence interval [CI]: 0.62-0.65), and DCA analysis showed higher net benefit than the 'Treat None' strategy across most probability thresholds, indicating added clinical value in decision making. The ADA risk calculator and LRA showed an AUC of 0.61 (95% CI: 0.59-0.63) and 0.631 (95% CI: 0.615-0.647), which were below the DDS performance. CONCLUSION: The DDS demonstrated acceptable performance for American adults aged 40 years or older, and showed marginally superior performance compared with the ADA diabetes risk calculator and LRA, highlighting its potential utility in dental practice settings as a complementary screening tool.