IADR Abstract Archives
Predicting Undiagnosed Diabetes and Prediabetes in Asian Patients With Periodontitis
Objectives: Presently in Southeast Asia, 90 million adults live with diabetes and almost 50% of them remain undiagnosed causing risk for serious morbidity. Data available in the dental setting can serve as a tool for early diabetes (DM) or prediabetes (PreDM) identification. Here, we sought to identify predictors in patient and periodontal profile to screen for undiagnosed DM and PreDM.
Methods: The study included 1074 adult patients with or without periodontitis who reported no DM or PreDM. A periodontal examination and a screening fingerstick HbA1c were performed. The POC HbA1c was considered the outcome, classified into no DM (NDM), at risk of PreDM, or at risk of DM. Patients’ sociodemographic characteristics, general health status, and periodontal status were recorded as potential predictors. The best model was chosen using univariable analysis of DM and PreDM by logistic regression followed by multi-variable logistic regression analysis. The performance and clinical values of the model were determined.
Results: 66(6.1%) and 386(35.9%) individuals were identified with HbA1c levels corresponding to the American Diabetes Association guideline levels for DM and PreDM, respectively. Patients’ BMI, family history of diabetes, non-smoking status and either having Stage III or IV Periodontitis were significantly associated with the diabetic status of the patients. The model showed moderate to good discrimination with area under the curve (AUC) values of 0.720 (95% CI: 0.656, 0.783) for predicting DM and 0.694 (95% CI: 0.662, 0.737) for predicting PreDM, yielding a sensitivity of 0.803 and specificity of 0.546 for DM prediction and a sensitivity of 0.731 and specificity of 0.563 for PreDM prediction.
Conclusions: Predictors in patient and periodontal profiles for screening of DM and PreDM in patients with periodontitis were identified. The clinical values of the model were acceptable. The model may well assist dentists to screen Asian periodontal patients for undiagnosed DM and PreDM reliably in dental settings.
Methods: The study included 1074 adult patients with or without periodontitis who reported no DM or PreDM. A periodontal examination and a screening fingerstick HbA1c were performed. The POC HbA1c was considered the outcome, classified into no DM (NDM), at risk of PreDM, or at risk of DM. Patients’ sociodemographic characteristics, general health status, and periodontal status were recorded as potential predictors. The best model was chosen using univariable analysis of DM and PreDM by logistic regression followed by multi-variable logistic regression analysis. The performance and clinical values of the model were determined.
Results: 66(6.1%) and 386(35.9%) individuals were identified with HbA1c levels corresponding to the American Diabetes Association guideline levels for DM and PreDM, respectively. Patients’ BMI, family history of diabetes, non-smoking status and either having Stage III or IV Periodontitis were significantly associated with the diabetic status of the patients. The model showed moderate to good discrimination with area under the curve (AUC) values of 0.720 (95% CI: 0.656, 0.783) for predicting DM and 0.694 (95% CI: 0.662, 0.737) for predicting PreDM, yielding a sensitivity of 0.803 and specificity of 0.546 for DM prediction and a sensitivity of 0.731 and specificity of 0.563 for PreDM prediction.
Conclusions: Predictors in patient and periodontal profiles for screening of DM and PreDM in patients with periodontitis were identified. The clinical values of the model were acceptable. The model may well assist dentists to screen Asian periodontal patients for undiagnosed DM and PreDM reliably in dental settings.