Oral Infection Signs May Predict Heart Disease Risk
A machine learning model was developed to predict heart disease using indicators of oral infections and was presented at the 52nd Annual Meeting & Exhibition of The American Association for Dental, Oral, and Craniofacial Research (AADOCR) in Portland on March 15-18, 2023.
Developing a Machine Learning Model to Predict Heart Disease using Oral Infections
This finding was reported in the new study that analyses the relationship between self-reported cardiovascular disease (heart surgery, heart valve, heart murmur, irregular heartbeat, and congenital heart disease) and markers for oral infections in 5,188 subjects from the University of Pittsburgh School of Dental Medicine's Dental Registry and DNA Repository project.
Periodontal screening and recording data (PSR) available from 740 subjects and the decayed, missing, or filled teeth and surfaces (DMFT and DMFS) from 5010 subjects were used in the analyses.
‘Using machine learning to identify heart disease occurrence from common oral signs can help diagnosticians reduce misdiagnosis.’
The results pointed to a significant association between both DMFT and DMFS and cardiovascular disease that are independent of sex and tobacco use. The results of the analysis of covariance between DMFS and cardiovascular disease also remained significant (p-value = 0.0027) after controlling for participants' age.
The machine learning model predicted whether a subject had cardiovascular disease based on their DMFS score with an accuracy of 84.3% in the registry. The study confirmed the association between dental caries and cardiovascular disease and highlighted the potential for machine learning methods to improve cardiovascular disease prediction using indicators of oral infections.
Future directions include assessing if artificial intelligence methods can help predict improvement in cardiovascular disease markers with dental caries management.
Source: Eurekalert