Trajectories of suicide attempts from early adolescence to emerging adulthood: Prospective 11-year follow-up of a Canadian cohort

Background: Suicide is a leading cause of mortality in youth, yet the course of suicide attempts is poorly documented. We explored the vulnerable transition from adolescence to emerging adulthood to identify group trajectories and risk factors.

Methods: The National Longitudinal Survey of Children and Youth is a prospective representative cohort of Canadian children. We followed participants aged 7-11 years in 1994-95 to age 23 (2008-09). We modelled self-reported past-year suicide attempts (ages 12 to 23 years) using growth mixture models. We analysed risk factors from self- and parent-report questionnaires at pre-adolescence (ages 10-11) and early adolescence (ages 12-13) using multinomial logistic regressions. Analyses were adjusted for sample non-response and attrition.

Results: In 2233 participants answering questions on teen and adult suicide attempts, we identified three trajectories: never attempted (96.0%), adolescence-limited (2.0%) and persisting into adulthood (2.0%). Adolescent girls aged 12-13 with depression/anxiety symptoms, and with mothers experiencing depression had higher risks of adolescence-limited than never-attempted [relative risk RR 9.27 (95% confidence interval: 1.73-49.82); 2.03 (1.02-3.32), for each standard deviation increase; 1.07 (1.00-1.15); respectively]. Preteen ADHD symptoms increased the risk of attempts persisting into adulthood as compared to never-attempted [RR 2.05 (1.29-3.28) for each standard deviation increase]. Suicide death of schoolmate/acquaintance increased risks of an adulthood trajectory as compared to never-attempted and adolescence-limited [RR 8.41 (3.04-23.27) and 6.63 (1.29-34.06), respectively].

Conclusion: In half the participants attempting suicide, attempts continued into adulthood. We stress the need for preventive strategies in early adolescence and differential clinical/educational interventions as identified for each trajectory.