Development and validation of the Suicide-related Behaviors Questionnaire-4 items across multiple high-risk populations.

Abstract

Suicidal ideation, suicide planning, non-suicidal self-injury (NSSI), and perceived controllability of suicidal thoughts represent distinct but related aspects of suicide-related experiences. However, brief instruments that assess these multiple aspects within a common timeframe are limited. Therefore, this study aimed to develop and validate the four-item Suicide-related Behaviors Questionnaire (SBQ-4) across three populations. A total of 405 university students (Mage = 19.20, SD = 1.21), 1318 daytime workers (Mage = 41.15, SD = 12.65), and 1260 pregnant women (Mage = 33.87, SD = 6.28) were included in the analyses. Confirmatory factor analyses supported a unidimensional structure of the SBQ-4 in all three populations. Multi-group confirmatory factor analysis further supported configural, metric, and scalar invariance across populations. Item response theory analyses using a graded response model indicated adequate item discrimination, although differential item functioning analyses suggested population differences in the NSSI and suicidal thought controllability items. Specifically, daytime workers required higher levels of the underlying latent trait (θ) to endorse higher response categories on these two items. Correlation analyses showed that the SBQ-4 was positively associated with suicidal ideation, depression, insomnia, and loneliness, supporting its construct validity. Overall, the SBQ-4 demonstrated a unidimensional structure and generally acceptable psychometric properties across the three populations. The SBQ-4 may therefore provide a brief research instrument for assessing recent suicide-related experiences in epidemiological and population-based studies, although caution is warranted when interpreting the NSSI and suicidal thought controllability items across populations.