Assessing the effect of perfluoroalkyl and polyfluoroalkyl substances on cardiovascular-kidney-metabolic syndrome: Insights from an interpretable machine learning model.

Journal: The Science of the total environment
Published Date: Jul 2, 2025

Abstract

Cardiovascular-kidney-metabolic syndrome (CKM) and its association with exposure to emerging pollutants, particularly perfluoroalkyl and polyfluoroalkyl substances (PFAS), present significant challenges for environmental public health and risk prediction. This study utilized cross-sectional data from the National Health and Nutrition Examination Survey (NHANES), conducted from 2015 to 2020, involving a total of 1953 participants. The diagnosis and staging of CKM were performed according to the definitions established by the American Heart Association. PFAS concentrations were measured using online solid-phase extraction combined with high-performance liquid chromatography-isotope dilution-tandem mass spectrometry. Using multistage regression models, we identified a significant positive association between serum perfluorooctane sulfonic acid (PFOS) concentration and CKM (odds ratio [OR] = 1.10, 95 % confidence interval [CI]: 1.04-1.16). Restricted cubic spline models further supported this linear relationship. Additionally, we selected eight distinct machine learning algorithms based on factors such as problem type, data characteristics, and interpretability to construct a CKM risk prediction model associated with PFOS exposure. We evaluated the performance of the models using metrics including area under the curve (AUC), accuracy, sensitivity, specificity, positive predictive value, negative predictive value, and F1 score. The results indicated that the neural network model exhibited the best predictive performance, with an AUC of 0.90 (95 % CI: 0.88-0.92) and an accuracy of 0.83 (95 % CI: 0.80-0.85). Furthermore, sensitivity, specificity, positive predictive value, negative predictive value, and F1 score all exceeded 0.82. The SHAP model interpretation revealed that total PFOS concentration was the most significant influencing factor for the occurrence of CKM. These findings provide evidence that total PFOS exposure is an independent risk factor for CKM and offer an effective tool for predicting CKM associated with PFAS exposure, which has substantial clinical application value. Future research will utilize longitudinal cohort data to further validate these findings.

Authors

  • Zidan Zheng
    State Key Laboratory of Trauma and Chemical Poisoning, National Institute for Occupational and Poison Control, Chinese Center for Disease Control and Prevention, Beijing 100050, China.
  • Yuyan Xu
    National Institute for Environmental Health, Chinese Center for Disease Control and Prevention, Beijing 100021, China.. Electronic address: xuyuyan@nieh.chinacdc.cn.
  • Ning Kang
    CHESS Center, Institute of Portal Hypertension, First Hospital of Lanzhou University, Lanzhou, China.
  • Yiwen Dong
  • Xin Wang
    Key Laboratory of Bio-based Material Science & Technology (Northeast Forestry University), Ministry of Education, Harbin 150040, China.
  • Lin Zhang
    Laboratory of Molecular Translational Medicine, Centre for Translational Medicine, Key Laboratory of Birth Defects and Related Diseases of Women and Children, Ministry of Education, Clinical Research Center for Birth Defects of Sichuan Province, West China Second Hospital, Sichuan University, Chengdu, Sichuan, 610041, China. Electronic address: zhanglin@scu.edu.cn.
  • Hongying Bian
    State Key Laboratory of Trauma and Chemical Poisoning, National Institute for Occupational and Poison Control, Chinese Center for Disease Control and Prevention, Beijing 100050, China.
  • Qibing Zeng
    State Key Laboratory of Trauma and Chemical Poisoning, National Institute for Occupational and Poison Control, Chinese Center for Disease Control and Prevention, Beijing 100050, China. Electronic address: zengqb@niohp.chinacdc.cn.

Keywords

No keywords available for this article.

External Resources

Popular Topics
Recent Journals