Risk and Protective Factors Associated With Suicidal Ideation and Attempts in a Longitudinal Study of Adolescents

• In Event: 1-039 - Addressing Mental Health Needs in Vulnerable Youth

Integrative Statement

Introduction.
Suicide is the second leading cause of death among 15-24-year-olds in the US, and rates have increased since 1999 (CDC, 2016). Prevention of suicide is a key public health priority, but existing population-level prevention strategies have not reduced suicide rates among adolescents and young adults. Identification of modifiable risk and protective factors is necessary to improve prevention programs.

Method.
We analyzed risk and protective factors for adolescent suicidal behavior in a prospective epidemiological study of adolescents interviewed at baseline (n=3981; ages 11-17), 1-year (n=2985), and 5-year (n=2391) follow-ups. Thirty-seven percent of respondents were White, 37% were Black, and 26% were Latino. We first analyzed the concurrent association of socio-demographic characteristics and past-year DSM-IV disorders with past-year suicidal ideation and suicide attempts. We then conducted longitudinal analyses of the associations of past DSM-IV disorders and self-reported risk/protective factors with first incidence of suicide attempt at a later wave, adjusting for sociodemographic factors. Risk/protective factors included exposure to potential stressors (neighborhood stress, school stress, economic stress, bullying, and experiences of trauma) and personal resources (self-esteem, positive coping, mastery of goals, social support, and positive parent-youth relationship). Analyses were conducted using multilevel logit models with random intercepts at the individual level and robust standard errors.

Results.
Unweighted and unadjusted prevalence rates of past year suicidal ideation across waves ranged from 2.7% to 3.5%; past year suicide attempts ranged from .8% to 1.2%. The lifetime suicide attempt rate ranged from 4% at Wave 1 to 10.3% at Wave 3. In analyses of concurrent associations controlling for sociodemographics, all disorder categories were associated with increased risk of attempts, with mood disorders having the strongest association (b=9.90, p<0.01). In analysis of modifiable risk and protective factors for first incidence of suicide attempt between Waves 1 and 2 (1 year later), accounting for sociodemographic and clinical characteristics, only experiencing school-related stress was associated with increased risk. Several sociodemographic characteristics were associated with decreased risk (Black or Latino race/ethnicity; having married parents; being in the youngest age group). In the same analysis for first incidence of suicide attempt between Waves 2 and 3 (five years later), better self-esteem and better parent-child relationships were associated with reduced risk of first incidence, while report of experiencing abuse (factor including sexual abuse/assault and family physical abuse) was associated with increased risk. The only sociodemographic characteristic still associated with reduced risk was Black race/ethnicity.

Conclusion.
Potential risk and protective factors may affect suicide risk differently at different points in development. School stress appears to impact short-term risk of suicide attempt in adolescence, perhaps because of daily exposure to this stressor. Attempts in late adolescence/young adulthood, on the other hand, appear to be impacted by self-esteem, positive parent/child relationships, and experiences of abuse reported five years earlier. These risk and protective factors may have a longer-term impact on individual well-being. The role of race/ethnicity, though not a modifiable risk factor, merits further investigation, particularly in terms of identifying what protective factors may be particularly salient for Black adolescents and young adults

Authors

• Kiara Alvarez, Massachusetts General Hospital
  Presenting Author

• Mirko Fillbrunn, Harvard Medical School
  Non-Presenting Author

• Mario A Cruz-Gonzalez, Massachusetts General Hospital
  Non-Presenting Author

• Margarita Alegria, Massachusetts General Hospital
  Non-Presenting Author

• Robert E Roberts, University of Texas
  Non-Presenting Author