The association between serum chloride levels and asthma prevalence: A cross-sectional study with variable screening...

Prior evidence suggests that ion channel dysfunction plays a significant role in the pathogenesis of asthma, which provides a theoretical basis for the potential link between chloride homeostasis and asthma. However, epidemiological evidence regarding the association between circulating chloride levels and asthma risk remains limited. To our knowledge, no large-scale cross-sectional study has yet explored this relationship. This study aims to investigate the association between chloride levels and asthma, thereby providing new insights into the pathogenesis of asthma. We analyzed data from the National Health and Nutrition Examination Survey 2007 to 2012, including 8241 participants (1165 asthma cases). Seven machine learning algorithms (eXtreme Gradient Boosting, Random Forest, Light Gradient Boosting Machine, Categorical Boosting, Adaptive Boosting, Gradient Boosting, and logistic regression) were employed to rank feature importance using SHapley Additive exPlanations values, supplemented by least absolute shrinkage and selection operator regression for feature selection. Model performance was evaluated via receiver operating characteristic curves. Key variables were further examined using subgroup analyses and restricted cubic spline regression to assess linear and nonlinear associations while adjusting for confounders (age, sex, body mass index, smoking, socioeconomic status, and occupational exposures). Of the biochemical indicators analyzed, chloride was a stronger predictor of asthma than other conventional biomarkers in both least absolute shrinkage and selection operator and SHapley Additive exPlanations-based logistic regression analyses. Serum chloride was inversely associated with asthma prevalence (adjusted odds ratios = 0.96 per mmol/L, 95% confidence intervals: 0.94-0.98, P < .001). While this relationship was consistent across models, the overall predictive performance was moderate (area under the curve approximately 0.65). These findings suggest a potential statistical association between systemic chloride levels and asthma, though causality cannot be inferred from this cross-sectional design. Among numerous conventional biochemical indicators, serum chloride levels show the most consistent negative correlation with asthma prevalence. However, the effect size is small, the discriminative performance of models incorporating this marker is modest, and the observed association may be confounded by unmeasured factors including dietary patterns, comorbidities, and medication use. The clinical utility and biological mechanism of this association require further research incorporating detailed asthma phenotyping and longitudinal follow-up.