Bruce A Dye1, Robert J Genco. 1. Centers for Disease Control and Prevention, National Center for Health Statistics, 3311 Toledo Road, Room 4416, Hyattsville, MD 20782, USA. bfd1@cdc.gov
Abstract
SUBJECTS: A total of 506 adults participated in this study. The study population was recruited from a pool of new patients presenting to the Columbia University College of Dental Medicine over a 12-month period (April 2009 to March 2010). New patients were screened to determine potential eligibility for participation based on 2 criteria: age and knowledge of their diabetes status. Non-Hispanic white adults were required to be 40 years old or older and Hispanic or non-white adults were required to be 30 years or older. Additionally, all potential participants had to respond that a health care provider had never told them that they had diabetes or prediabetes. This screening yielded 601 individuals. From this group, 535 were selected based on having 1 of 4 self-reported risk factors (hypertension, high cholesterol, overweight, or a family history of diabetes). These 535 adults continued with a periodontal examination and an HbA1c test. Five hundred six participants returned for a follow-up visit to collect blood for a fasting plasma glucose (FPG) test. KEY RISK/STUDY FACTOR: HbA1c test, and dentate and periodontal status. MAIN OUTCOME MEASURE: FPG level. MAIN RESULTS: Among the 535 individuals participating in the study, 161 were determined potentially to be prediabetic (FPG was 100-125 mg/dL) and 21 potentially to be diabetic (FPG ≥ 126 mg/dL). Receiver operating characteristics (ROCs) via logistic regression was used to assess model performance and calculate key findings. The area under the curve of a multivariate regression model that included oral health status indicators and all 4 self-reported risk factors had a predicted value of 0.68 (confidence interval [CI]: 0.63, 0.73) for abnormal FPG (≥100 mg/dL). A model with just the percentage of periodontal pockets ≥5 mm and the number of missing teeth had a predicted value of 0.65 (CI: 0.60, 0.70). When the test results from HbA1C were added to the simpler model, the predictive value improved to 0.79 (CI: 0.75, 0.83). The authors also determined that the presence of 4 or more missing teeth or having pocket depth of 5 mm or more involving more than 25% of the sites identified 73% of the true cases (ie, abnormal FPG). CONCLUSIONS: Dental care professionals have the potential to identify patients at risk for diabetes and refer them to a physician for follow-up evaluation.
SUBJECTS: A total of 506 adults participated in this study. The study population was recruited from a pool of new patients presenting to the Columbia University College of Dental Medicine over a 12-month period (April 2009 to March 2010). New patients were screened to determine potential eligibility for participation based on 2 criteria: age and knowledge of their diabetes status. Non-Hispanic white adults were required to be 40 years old or older and Hispanic or non-white adults were required to be 30 years or older. Additionally, all potential participants had to respond that a health care provider had never told them that they had diabetes or prediabetes. This screening yielded 601 individuals. From this group, 535 were selected based on having 1 of 4 self-reported risk factors (hypertension, high cholesterol, overweight, or a family history of diabetes). These 535 adults continued with a periodontal examination and an HbA1c test. Five hundred six participants returned for a follow-up visit to collect blood for a fasting plasma glucose (FPG) test. KEY RISK/STUDY FACTOR: HbA1c test, and dentate and periodontal status. MAIN OUTCOME MEASURE: FPG level. MAIN RESULTS: Among the 535 individuals participating in the study, 161 were determined potentially to be prediabetic (FPG was 100-125 mg/dL) and 21 potentially to be diabetic (FPG ≥ 126 mg/dL). Receiver operating characteristics (ROCs) via logistic regression was used to assess model performance and calculate key findings. The area under the curve of a multivariate regression model that included oral health status indicators and all 4 self-reported risk factors had a predicted value of 0.68 (confidence interval [CI]: 0.63, 0.73) for abnormal FPG (≥100 mg/dL). A model with just the percentage of periodontal pockets ≥5 mm and the number of missing teeth had a predicted value of 0.65 (CI: 0.60, 0.70). When the test results from HbA1C were added to the simpler model, the predictive value improved to 0.79 (CI: 0.75, 0.83). The authors also determined that the presence of 4 or more missing teeth or having pocket depth of 5 mm or more involving more than 25% of the sites identified 73% of the true cases (ie, abnormal FPG). CONCLUSIONS: Dental care professionals have the potential to identify patients at risk for diabetes and refer them to a physician for follow-up evaluation.
Authors: Saba Kassim; Badr Othman; Sakher AlQahtani; Alemad Mustafa Kawthar; Sterling M McPherson; Barbara L Greenberg Journal: BMC Oral Health Date: 2019-08-06 Impact factor: 2.757