Subtypes of Substance Users and Suicide Behaviors: An Adolescent Risk Profile

Background & Purpose: Suicide and substance use are pervasive problems for adolescents in the U.S. Among U.S. youth, suicide is the second leading cause of death for those age 15-19. Drugs and alcohol also pose a significant public health concern as 8% of all U.S. youth are diagnosed with a substance use disorder, and 78.2% have consumed alcohol by late adolescence. Substance use, in general, is a known risk factor related to suicide among adolescents. Past postmortem analyses, for example, have found that up to 70% of adolescents who completed suicide were drug and alcohol users and that adolescent completers were more likely to have been under the influence of alcohol. The adolescent substance use literature has proposed varying typologies of users, with the four most reported being experimentation, regular use, abuse, and dependence/addiction. We apply a similar typology to examine the risk and profile of adolescents, including their demographic characteristics, who engage in suicidal behaviors.

Methods: Data were obtained from the 2015 National Youth Risk Behavior Survey (YRBS). The analytic sample includes 15,624 adolescents who reported a range of substance use behaviors. Marijuana, cigarette, and alcohol use were measured by both lifetime and current use (ever use, age at initiation, current frequency of use, and current intensity of use). We utilize the term “polysubstance use” to describe reported usage of all three substances (marijuana, cigarettes, and alcohol) by adolescents. Latent class analysis was used to identify homogeneous subgroups of adolescent substance use patterns. After determining the appropriate number of classes, a final model was run in which the associations between substance use class membership and suicidal behaviors were assessed while controlling for age, gender, and race/ethnicity effects on each through multinomial logistic regression. All analyses were conducted using Mplus 7.4 and incorporated sample weights, stratum, and cluster variables to account for the complex sample design of the YRBS.

Results: A four-class model provided the best fit to the data, which was conceptualized as experimental polysubstance users, frequent polysubstance users, moderate polysubstance users, and nonusers. Using the nonusers’ class as the reference group, all the other three groups had a higher likelihood of engagement in suicidal behaviors, with adolescents in the frequent polysubstance users group, which included older males, exhibited the highest odds of suicidal ideation (OR=2.28), suicide plan (OR=2.25), and suicide attempt (OR=3.152). The moderate polysubstance users group contained more Black and Hispanic/Latino youth than the nonusers group.

Conclusion & Implications: Adolescents who are frequent polysubstance users are at heightened risk for suicidal behaviors compared to nonusers, moderate users, and adolescents who use substances somewhat experimentally. Moderate and experimental polysubstance users are at even greater risk for suicidal thoughts, plans, and attempts when compared to nonusers. The findings provide important information regarding the profile of adolescents with high risk for engagement in suicidal behaviors. Practitioners and health professionals should consider screening for “polysubstance use” (and its frequency) in adolescents given the association with suicide risk.