Aim: To explore the risk factors associated with suicidal behavior (SB) using logistic regression analysis and the propensity score matching (PSM) method among Chinese patients suffering from late-life depression (LLD).
Method: Patient information sheets were retrieved with the International Classification of Diseases, Tenth Revision (ICD-10) code from an electronic database that comprised patient medical information. Herein, we set SB as a dependent variable, and gender, marital status, monthly income, quality of interpersonal relationships, hobbies, impulsivity, severity of depression, psychiatric symptoms or not, and having histories of smoking, drinking, major mental trauma as independent variables according to clinical experience and previous findings. For uncertain independent risk factors associated with SB generated by logistic regression analysis, PSM was performed for further verification.
Results: The differences between the SB group and non-SB group for marital status, severity of depression, a history of drinking, and a history of major mental trauma were found to be statistically significant in univariate comparisons (P < 0.05); binary logistic regression analysis and PSM analysis showed that the severity of depression, a history of drinking, and a history of major mental trauma were independent risk factors associated with SB of patients with LLD with an odds ratio greater than one.
Conclusion: The severity of depression, a history of drinking, and a history of major mental trauma were independently associated with the occurrence of SB of patients with LLD in China. Further longitudinal and prospective studies are warranted to examine the dynamic changes of confounding risk factors. Geriatr Gerontol Int 2021; 21: 849-854.
Keywords: late-life depression; logistic regression; propensity score matching; risk factor; suicidal behavior.
© 2021 Japan Geriatrics Society.