Negative life events, suicidal ideation and self-harm among nurses with depressive symptoms: A latent class analysis

Suicidal ideation and self-harm among nurses with depressive symptoms are public health concerns, given their elevated prevalence and potentially severe outcomes. Prior studies predominantly examined negative life events either as isolated variables within broader exploratory analyses or as a single continuous measure, without assessing distinct patterns. This study aimed to identify distinct patterns of negative life events among Chinese nurses with depressive symptoms and examine their associations with suicidal ideation and self-harm.

This cross-sectional study analyzed data from 73,371 nurses with depressive symptoms from the Nurses' Mental Health Study in China. Latent class analysis identified patterns of ten negative life events experienced in the past year. Two binomial logistic regression models were used to examine associations between negative life event patterns and suicidal ideation (n = 64,569) and self-harm (n = 68,038), adjusting for demographic, work-related, health-related, and childhood experience covariates. Sensitivity analyses included stratified logistic regressions by levels of depressive symptom severity and multinomial logistic regressions that treated "prefer not to answer" responses as a separate category, to assess the robustness of associations and potential underreporting bias.

Four distinct negative life event patterns emerged: Low-Stress Life Events (59.7 %), Health and Family Crisis (12.4 %), Economic and Relationship Struggles (23.1 %), and Widespread Life Crises (4.7 %). Compared to the Low-Stress group, nurses in the Economic and Relationship Struggles group had statistically significantly higher odds of suicidal ideation (AOR = 1.25, 95 % CI: 1.14 to 1.37) and self-harm (AOR = 1.20, 95 % CI: 1.04 to 1.39). The Widespread Life Crises group showed even stronger associations with suicidal ideation (AOR = 1.68, 95 % CI: 1.46 to 1.92) and self-harm (AOR = 1.86, 95 % CI: 1.54 to 2.24). The Health and Family Crisis group showed no statistically significant associations with either outcome. Sensitivity analyses supported the main findings, with consistent associations observed across depression severity subgroups. The multinomial logistic regression also showed elevated risk ratios for suicidal ideation and self-harm among respondents in the Widespread Life Crises group, including those who selected "prefer not to answer."

This study identified distinct patterns of negative life events among nurses with depressive symptoms and demonstrated that economic and relationship challenges, especially when occurring across multiple life domains, are significantly associated with the risk of suicidal ideation and self-harm. Targeted interventions addressing specific life stressor patterns may help reduce suicidality in this high-risk population.