Four modeling approaches to study restrictions on everyday life and social activities due to chronic diseases with consequences of suicidal behavior
Myoungjee Jung,Kwang Ok Lee,Hae-Rim Kim,Sang-Baek Koh,Jeong-An Gim
DOI: https://doi.org/10.1016/j.jpsychires.2024.03.037
Abstract:The purpose of this study was to discover the association between disability in everyday life and social activities due to chronic diseases and suicidal ideation (SI), suicidal plan (SP), and suicidal attempt (SA) from the Korea National Health and Nutrition Examination Survey (KNHANES), considering the cross-sectional design of this study, 2016-2018 dataset. Variables for finding the associated factors of SI, SP, and SA were confirmed through random forest (RF), decision tree, generalized linear model (GLM), and support vector machine (SVM), and the performance of each model is listed. A total of 17,323 (males: 7,530, females: 9793) responders from the KNHANES from 2016 to 2018 were employed for the study. The relationship between restrictions on daily life, social activities, and three stages of suicidal behaviors due to diseases were analyzed using the R function (R version 4.2.0), randomForest, ctree, glm, and ksvm. The F1-score is a measure used to evaluate the accuracy of the performance of a model, in the binary classification. The score of 1 indicates good performance, whereas a score of 0 signifies poor performance. Due to chronic diseases, disability in everyday life and social activities lead to suicide behaviors. In our study, we examined the impact of limitations in daily living and social activities on suicidal behaviors among participants. Our findings revealed that for those experiencing such limitations, the odds ratios (ORs) for SIs were 6.10 (95% CI: 3.99-9.34) for males and 2.61 (1.79-3.81) for females. SPs were 3.69 (2.36-5.78) for males and 3.94 (2.70-5.75) for females. Similarly, the odds ratios for SAs were 5.04 (2.51-10.13) for males and 2.71 (1.48-4.98) for females, indicating a significant association between these limitations and increased suicidal behaviors, with variances observed between genders. These results underscore the necessity of addressing daily living and social activity restrictions when considering mental health interventions and suicide prevention strategies. In RF, GLM, and SVM, F1-score were 0.8192, 0.6887, and 0.9687 in SA, respectively. Among the patients with chronic disease, those with sequelae, low incomes, and low levels of education had limitations in daily activities and social activities, which increased the likelihood of suicidal thoughts, planning, and attempts.