Identifying resilience factors for adolescent mental health with cyberbullying victimisation as a risk factor
Aaron Kandola,Rosie Mansfield,Yvonne Kelly,Yasmin Rahman,Karmel Choi,Chris Hollis,Ellen Townsend,Digital Youth,Praveetha Patalay
DOI: https://doi.org/10.1101/2024.11.18.24317457
2024-11-19
Abstract:Background
Promoting resilience could reduce the pervasive and harmful mental health impact of cyberbullying in adolescents, but most studies only focus on a narrow set of factors. We aimed to identify novel mental health resilience factors by systematically scanning and examining many variables that could moderate the association between cyberbullying victimisation and psychological distress during adolescence.
Methods
We conducted a longitudinal cohort study using data from 9,969 adolescents in the Millennium Cohort Study (MCS) at ages 14-15 (baseline) and 17 (follow-up). We replicated our analyses in 4,240 adolescents from the Longitudinal Study of Australian Children (LSAC) at ages 14-15 (baseline) and 16-17 (follow-up). The outcome was psychological distress at follow-up (Kessler Psychological Distress Scales and short Moods and Feelings Questionnaire). We used an exposure-wide approach to systematically identify resilience factors (exposures) as any potentially modifiable variables measured at baseline in MCS associated with psychological distress (outcome) at follow-up in separate regression models (aim one). We then examined in separate models whether the aim one resilience factors (moderators) interacted with the cyberbullying (exposure) and distress (outcome) association in MCS (aim two). Cyberbullying was measured with a single-item response to a question about experiencing online bullying. We then checked whether aim two findings replicated in LSAC.
Results
Of 1,466 baseline variables, 478 were potential resilience factors associated with distress in MCS after confounder adjustment (aim one), which we examined in aim two as moderators. Of the 478 potential resilience factors, 31 moderated the cyberbullying-distress association in MCS and were mostly individual domain factors (54.84%, n = 17, e.g., happiness with friends). In LSAC, 15 models replicated, and most were still in the individual (n=8, e.g., happiness with life in general), then the family and friends (n = 3, e.g., sexual activities), structural (n = 2, e.g., sources of income), and learning environment and (n = 2, e.g., misbehaviour in class) domains.
Conclusion
We leveraged the breadth of data available in large cohort studies to identify several promising factors to increase mental health resilience against cyberbullying in adolescents, such as happiness with friends and life in general, parental supervision, and avoiding alcohol use or dieting to lose weight. These factors warrant further investigation as possible intervention targets for reducing the harm of cyberbullying in adolescents.