Early Subtypes of Depressive, Anxious, and Aggressive Symptoms: A Longitudinal Analysis of Pathways to Suicide Attempt

Introduction: Suicide is a leading cause of death for all age groups.  Unfortunately, after declining for several years, rates for individuals 10- to 24-years-old have begun to increase. Understanding the mechanisms underlying suicidal thoughts and behaviors is an important aspect of developing effective suicide prevention.  Identifying homogeneous subtypes of individuals who attempt suicide is an important step in this process. Along with internalizing symptoms such as depression and anxiety, aggressive characteristics have been indicated, although somewhat inconsistently, as important in the development of suicidal behaviors.

 Methods: Data was drawn from a randomized, longitudinal, epidemiologically-based, preventive-intervention trial conducted by the Johns Hopkins Center for Prevention and Early Intervention Research Center. Longitudinal latent class analysis was used to explore concurrent trajectories of depressive, anxious, and aggressive symptoms in 2^nd, 4^th, 5^th, 6^th, and 7^thgrades in order to begin to identify early subtypes of those at risk for suicide attempt. Two studies were conducted in order to understand these trajectories.  First, trajectories were established for individuals who report lifetime suicide attempts in young adulthood (n=191). Distal outcomes and covariates were added to the model to further understand what might predict the development of, as well as to differentiate, these early subtypes of individuals who attempt suicide. Second, trajectories of the same symptoms were explored for individuals who did not report a lifetime suicide attempt (n=1835).  Distal outcomes and covariates were again added to the model.  Finally, all subtypes were compared for level of symptoms and differences between covariates and distal outcomes. 

Results: Three subtypes of suicide attempt reporters were identified; those with 1) the lowest levels of aggressive, depressive, and anxious symptoms (LOW; 17.2% of the sample), 2) the highest levels of aggressive, depressive, and anxious symptoms (HIGH; 48.3%), and 3) the highest levels of depressive and anxious symptoms and lowest levels of aggressive symptoms (HI/LO; 34.4%).  Covariate and distal outcome differences were evident among these classes.  Five subtypes were identified for non-attempters, with differences among covariate and distal outcomes again identified. Interesting comparisons are made across all subtypes and trajectories.

Conclusion: These results suggest that early subtypes regarding depressive, anxious and aggressive symptoms offer important information for prevention in regards to suicidal behaviors.  Aggression appears to be a particularly important factor in these early processes.  Implications for public health and prevention will be discussed.