Developing machine learning models for predicting cardiovascular disease survival based on heavy metal serum and urine levels DOI Creative Commons
Hui Jin, Ling Zhang, Yan Sun

et al.

Frontiers in Public Health, Journal Year: 2025, Volume and Issue: 13

Published: May 21, 2025

Background Environmental exposure to heavy metals, such as arsenic, cadmium, and lead, is a known risk factor for cardiovascular diseases. Objective We aim examine the associations between metal mortality of patients with Methods analyzed data from NHANES 2003–2018, including urine blood concentrations 4,924 participants. Five machine learning models—CoxPHSurvival, FastKernelSurvivalSVM, GradientBoostingSurvival, RandomSurvivalForest, ExtraSurvivalTrees—were used predict mortality. Model performance was assessed concordance index (C-index), integrated Brier score, time-dependent AUC, calibration curves. SHAP analysis conducted using reduced background dataset created via K-means clustering. Results GradientBoostingSurvival (GBS) showed best hypertension (C-index: 0.780, mean AUC: 0.798). RandomSurvivalForest (RSF) top model coronary heart disease 0.592, 0.626) myocardial infarction 0.705, 0.743), while CoxPHSurvival excelled failure 0.642, 0.672) stroke 0.658, 0.691). ExtraSurvivalTrees performed in angina 0.652, 0.669). Calibration curves confirmed models’ accuracy. identified age most influential factor, metals like thallium significantly contributing risk. A user-friendly web calculator developed individualized survival predictions. Conclusion Machine models, CoxPHSurvival, ExtraSurvivalTrees, demonstrated strong predicting various Key were significant factors assessment.

Language: Английский

Joint Effects of Lifestyle Habits and Heavy Metals Exposure on Chronic Stress Among U.S. Adults: Insights from NHANES 2017–2018 DOI Creative Commons
Esther Ogundipe, Emmanuel Obeng-Gyasi

Journal of Xenobiotics, Journal Year: 2025, Volume and Issue: 15(1), P. 7 - 7

Published: Jan. 7, 2025

Chronic stress, characterized by sustained activation of physiological stress response systems, is a key risk factor for numerous health conditions. Allostatic load (AL), biomarker cumulative offers quantitative measure this burden. Lifestyle habits such as alcohol consumption and smoking, alongside environmental exposures to toxic metals like lead, cadmium, mercury, were individually implicated in increasing AL. However, the combined impact these lifestyle factors remains underexplored, particularly populations facing co-occurring exposures. This study aims investigate joint effects on AL, using data from NHANES 2017-2018 cycle. By employing linear regression Bayesian Kernel Machine Regression (BKMR), we identify predictors explore interaction effects, providing new insights into how contribute chronic stress. Results BKMR analysis underscore importance addressing exposures, synergistic cadmium consumption, managing Descriptive statistics calculated summarize dataset, multivariate was performed assess associations between employed estimate exposure-response functions posterior inclusion probabilities (PIPs), focusing identifying indicated that mean levels mercury 1.23 µg/dL, 0.49 1.37 µg/L, respectively. The allostatic 3.57. Linear significantly associated with increased AL (β = 0.0933; 95% CI [0.0369, 0.1497]; p 0.001). Other including lead -0.1056; [-0.2518 0.0408]; 0.157), -0.0001, [-0.2037 0.2036], 0.999), -0.0149; [-0.1175 0.0877]; 0.773), smoking 0.0129; [-0.0086 0.0345]; 0.508), not significant. confirmed alcohol's strong PIP 0.9996, highlighted non-linear effect (PIP 0.7526). showed stronger at higher exposure levels. In contrast, demonstrated minimal Alcohol identified contributors load, while other no significant associations. These findings emphasize

Language: Английский

Citations

1
A sustainable way to prevent oral diseases caused by heavy metals with phytoremediation DOI Creative Commons
Samira Salehi, Mahdi Pouresmaieli, Ali Nouri Qarahasanlou

et al.

Case Studies in Chemical and Environmental Engineering, Journal Year: 2025, Volume and Issue: unknown, P. 101106 - 101106

Published: Jan. 1, 2025

Language: Английский

Citations

1
Association between heavy metal exposure and asthma in adults: Data from the Korean National Health and Nutrition Examination Survey 2008–2013 DOI Creative Commons
Mijung Jang, Doh‐Hee Kim, Seunghee Lee

et al.

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(3), P. e0319557 - e0319557

Published: March 10, 2025

Risk factors for asthma include genetic, host, and environmental such as allergens, smoking, exposure to chemicals. Heavy metals from air pollution or contaminated water food can also trigger asthma. This study aimed identify the biological levels of blood lead, mercury, cadmium, determine association with single multiple exposures these heavy using data Korean National Health Nutrition Examination Survey (KNHANES) conducted between 2008 2013. A weighted analysis 40,328 adults aged ≥ 20 years was conducted. Variables included metal levels, health behaviors, demographic characteristics, status. Logistic regression used odds ratio (OR) in adults. The overall prevalence 3.0%. geometric mean values cadmium were 2.14 μg/dL, 3.72 μg/L, 0.96 respectively. An high lead observed, highest level group showing a statistically significant association. Blood mercury significantly associated quartile levels. After adjusting behavior variables, associations persisted quartiles all metals. Multiple showed demonstrated adults, emphasizing need reduce preventive measure against

Language: Английский

Citations

0
Developing machine learning models for predicting cardiovascular disease survival based on heavy metal serum and urine levels DOI Creative Commons
Hui Jin, Ling Zhang, Yan Sun

et al.

Frontiers in Public Health, Journal Year: 2025, Volume and Issue: 13

Published: May 21, 2025

Background Environmental exposure to heavy metals, such as arsenic, cadmium, and lead, is a known risk factor for cardiovascular diseases. Objective We aim examine the associations between metal mortality of patients with Methods analyzed data from NHANES 2003–2018, including urine blood concentrations 4,924 participants. Five machine learning models—CoxPHSurvival, FastKernelSurvivalSVM, GradientBoostingSurvival, RandomSurvivalForest, ExtraSurvivalTrees—were used predict mortality. Model performance was assessed concordance index (C-index), integrated Brier score, time-dependent AUC, calibration curves. SHAP analysis conducted using reduced background dataset created via K-means clustering. Results GradientBoostingSurvival (GBS) showed best hypertension (C-index: 0.780, mean AUC: 0.798). RandomSurvivalForest (RSF) top model coronary heart disease 0.592, 0.626) myocardial infarction 0.705, 0.743), while CoxPHSurvival excelled failure 0.642, 0.672) stroke 0.658, 0.691). ExtraSurvivalTrees performed in angina 0.652, 0.669). Calibration curves confirmed models’ accuracy. identified age most influential factor, metals like thallium significantly contributing risk. A user-friendly web calculator developed individualized survival predictions. Conclusion Machine models, CoxPHSurvival, ExtraSurvivalTrees, demonstrated strong predicting various Key were significant factors assessment.

Language: Английский

Citations

0