International dental journal · 2025
Associations Between Thyroid Function and Periodontitis: A Machine Learning Approach Using NHANES.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Wang Han, Tian Suyan, Bai Sen, Yang Chunqi, Jiang Zhengang, Li Nannan
Аннотация
INTRODUCTION: Thyroid dysfunction alters bone metabolism and is associated with osteoporosis. Since periodontitis involves alveolar bone loss, and thyroid disorders may impair immune regulation as well as bone remodeling, we hypothesized that thyroid hormone (TH) levels could predict the risk and severity of periodontitis. However, this potential association remains understudied to date.
METHODS: This study utilized cross-sectional data from the 2009-2010 and 2011-2012 National Health and Nutrition Examination Survey (NHANES) cycles, comprising 2352 eligible participants: 704 without periodontitis and 1648 with periodontitis. We analyzed the data using multiple machine learning (ML) approaches, including eXtreme Gradient Boosting (XGBoost), random forest (RF), support vector machine (SVM), logistic regression (LR), and multilayer perceptron (MLP). Predictive models were developed using thyroid function parameters and clinical characteristics.
RESULTS: XGBoost demonstrated the highest performance for both binary and multi-class classification tasks, while MLP performed worse compared to other models. For binary classification, the XGBoost model achieved an average area under the ROC curve (AUC) of 0.854, a recall of 0.760, a precision of 0.760, and an F1-score of 0.754 across 5-fold cross-validations (CV). Thyroid-stimulating hormone (TSH) emerged as a top predictor in feature importance analysis, alongside age and body mass index (BMI). Similarly, for multi-classclassification, TSH ranked highly in both feature importance and SHAP value analyses. Models excluding thyroid function variables exhibited significantly inferior performance compared to those incorporating them.
CONCLUSIONS: Our findings suggest that thyroid parameters, particularly TSH, might predict risk for periodontitis and serve as its biomarkers, thereby enabling early intervention and personalized patient care. Further validation through larger, prospective studies is warranted to confirm these observations.
METHODS: This study utilized cross-sectional data from the 2009-2010 and 2011-2012 National Health and Nutrition Examination Survey (NHANES) cycles, comprising 2352 eligible participants: 704 without periodontitis and 1648 with periodontitis. We analyzed the data using multiple machine learning (ML) approaches, including eXtreme Gradient Boosting (XGBoost), random forest (RF), support vector machine (SVM), logistic regression (LR), and multilayer perceptron (MLP). Predictive models were developed using thyroid function parameters and clinical characteristics.
RESULTS: XGBoost demonstrated the highest performance for both binary and multi-class classification tasks, while MLP performed worse compared to other models. For binary classification, the XGBoost model achieved an average area under the ROC curve (AUC) of 0.854, a recall of 0.760, a precision of 0.760, and an F1-score of 0.754 across 5-fold cross-validations (CV). Thyroid-stimulating hormone (TSH) emerged as a top predictor in feature importance analysis, alongside age and body mass index (BMI). Similarly, for multi-classclassification, TSH ranked highly in both feature importance and SHAP value analyses. Models excluding thyroid function variables exhibited significantly inferior performance compared to those incorporating them.
CONCLUSIONS: Our findings suggest that thyroid parameters, particularly TSH, might predict risk for periodontitis and serve as its biomarkers, thereby enabling early intervention and personalized patient care. Further validation through larger, prospective studies is warranted to confirm these observations.