J Affect Disord. 2025 Oct 26:120520. doi: 10.1016/j.jad.2025.120520. Online ahead of print.
ABSTRACT
BACKGROUND: Depression is prevalent among asthma patients, negatively impacting their quality of life, treatment adherence, and prognosis. This study aimed to develop and validate a predictive model estimating the risk of depressive episodes in asthma patients.
METHODS: In this study, data from NHANES were used to collect clinically accessible variables. We applied a combined selection strategy to identify core variables associated with the risk of depression in asthma patients. Multiple machine learning methods were subsequently employed to develop predictive models, and the optimal model was selected and visualized through the integration of various model parameters. Finally, external validation was performed.
RESULTS: A total of 3517 asthma patients were included in this study for analysis, and the cohort was divided into a training set and a validation set based on the survey period. 1951 patients and 1566 patients formed the training cohort for model development and the validation cohort for model performance evaluation, respectively. Based on the ranking of the importance of feature variables, last year’s asthma acute attack, smoking, PIR, total bilirubin, COPD, and serum glucose were ultimately selected as the final variables for model construction. Multiple models were comprehensively evaluated, and the Logistic model with the best predictive performance was selected. This model still demonstrated excellent predictive performance and clinical utility in the validation set.
CONCLUSION: We developed a risk prediction model for the occurrence of depressive symptoms in asthma patients and visualized it in the form of a nomogram for clinical application.
PMID:41151724 | DOI:10.1016/j.jad.2025.120520
Recent Comments