Examining bystander intervention for peer depression and sociodemographic correlates among university students in Singapore.

Introduction: The Advancing Research To Eliminate Mental Illness Stigma (ARTEMIS) study evaluated the impact of an intervention developed and implemented in Singapore on attitudes towards depression in university students. We aimed to assess the likelihood of university students intervening when their peers suffer from depression, before and after the ARTEMIS intervention. Methods: 390 students were recruited from a university in Singapore. The ARTEMIS intervention comprised a lecture by a trained mental health professional, a sharing session by a person with lived experience of depression, and a question-and-answer segment with a panel. The Bystander Intervention Scale for Depression (BISD) was administered at baseline, post-intervention, and 3-month follow-up. BISD assessed four factors: acceptance of responsibility to intervene, knowledge on how to intervene, awareness of depression among peers, and vigilance towards possible symptoms of depression. Linear mixed models were conducted to investigate associations. Sociodemographic correlates were also examined. Results: A favourable shift in all factors was observed at post-intervention, which weakened at 3-month follow-up. Having past experience in the mental health field (b=1.50) and older age (b=0.18) were significantly associated with knowledge on how to intervene. Having social contact with mental illness (SCMI) and past experience in the mental health field (PEMHF) were significantly associated with awareness of depression among peers (SCMIb=0.89, PEMHF b=0.43) and vigilance towards possible symptoms of depression (SCMI b=0.39, PEMHF b=0.61). Discussion: The short-term results of the intervention appeared promising across all BISD factors; however, these results were not sustained after 3 months. Future research should include the impact of 'booster' interventions over time. Sociodemographic factors that were identified to be significant correlates should also be considered when planning for future interventions. [ABSTRACT FROM AUTHOR]