Exposure to and experience of self-harm and self-harm related content: An exploratory network analysis. (January 2021)

• Record Type: Journal Article
• Authors: Kirtley, Olivia J.; Hussey, Ian; Marzano, Lisa
• Is Part Of: Psychiatry research. Volume 295(2021)
• Publication Date: 2021-01
• Subjects: Self-harm -- exposure -- network analysis; Psychiatry -- Periodicals; Psychiatry -- periodicals; Psychiatrie -- Périodiques; 616.89
• DOI: 10.1016/j.psychres.2020.113572
• Languages: English
• ISSNs: 0165-1781

Abstract

Highlights: Study examined wide variety of types of exposure to self-harm and related content First ever network analysis of diverse range of exposure mediums In-passing and online exposure directly associated with self-harm thoughts and behaviours Future research should investigate temporal dynamics of different exposure mediums Abstract: Exposure to the self-harm behaviour of others plays a role in individuals' own self-harm thoughts and behaviours, but there has been little consideration of the broader range of mediums through which exposure to self-harm related content may occur. N = 477 participants completed an online study, including questions regarding lifetime history of self-harm thoughts and behaviours and the frequency with which they had been exposed to self-harm via various mediums. Gaussian Markov random field network models were estimated using graphical LASSO and extended Bayesian information criterion. Bootstrapping revealed that exposure mediums with a direct connection to self-harm thoughts and behaviours were the internet ( rrp = .34, 95% CI [.26, .42]) and in-passing 'miscellaneous' exposure ( rrp = .14, 95% CI [.00, .23]). However, stability of the network centrality was low (expected influence stability = 0.52). The node with the greatest increase in expected influence within the network was miscellaneous "in-passing" exposure. In-passing exposure is an understudied exposure medium. Our results may suggest new types of exposure mediums for future research. Data were cross-sectional, so temporal relationships between exposure and behaviour could not be determined. Low stability of the networks suggests that future similar studies would benefit from larger sample sizes.