Specific traumatic events elevate the risk of a suicide attempt in a 10-year longitudinal community study on adolescents and young adults.

Traumatic events (TEs) have been associated with suicide attempts (SAs). However, the empirical status of some TEs is inconclusive. This also concerns community adolescents and young adults, known to be a high-risk group for SAs. We examined associations between (a) a range of prior TEs (physical attack, rape/sexual abuse, serious accident, and witnessing somebody else experiencing a TE) and a subsequent SA, and (b) the number of prior TEs and an SA, and (c) we estimated attributable proportions of SAs, in relation to each TE. Over a 10-year period, the Early Developmental Stages of Psychopathology (EDSP) study prospectively assessed community members, aged 14-24 years at baseline. Starting with 3021 subjects, each individual was assessed up to four times. Assessment was based on the Munich-Composite International Diagnostic Interview. Temporal associations were estimated using the Cox model with time-dependent covariates. Attributable proportions were based on the results of the Cox models. All four TEs elevated the risk for a subsequent SA, adjusting for confounders. Highest risk was found for the combined TE rape/sexual abuse. Results showed that 56-90% of SAs could be attributed to TEs in the exposed group; on the population level, attributable proportions ranged between 6.9% and 23.5%. Different TEs have been shown to elevate the risk of an SA in a young community sample. Our results suggest that both health professionals and health policy decision-makers consider specific TEs and the number of prior TEs as risk factors for SAs.

Mental disorders and the risk for the subsequent first suicide attempt: results of a community study on adolescents and young adults.

Adolescents and young adults represent the high-risk group for first onset of both DSM-IV mental disorders and lifetime suicide attempt (SA). Yet few studies have evaluated the temporal association of prior mental disorders and subsequent first SA in a young community sample. We examined (a) such associations using a broad range of specific DSM-IV mental disorders, (b) the risk of experiencing the outcome due to prior comorbidity, and (c) the proportion of SAs that could be attributed to prior disorders. During a 10-year prospective study, data were gathered from 3021 community subjects, 14-24 years of age at baseline. DSM-IV disorders and SA were assessed with the Munich-Composite International Diagnostic Interview. Cox models with time-dependent covariates were used to estimate the temporal associations of prior mental disorders with subsequent first SA. Most prior mental disorders showed elevated risk for subsequent first SA. Highest risks were associated with posttraumatic stress disorder (PTSD), dysthymia, and nicotine dependence. Comorbidity elevated the risk for subsequent first SA, and the more disorders a subject had, the higher the risk for first SA. More than 90% of SAs in the exposed group could be attributed to PTSD, and over 30% of SAs in the total sample could be attributed to specific phobia. Several DSM-IV disorders increase the risk for first SA in adolescents and young adults. Several promising early intervention targets were observed, e.g., specific phobia, nicotine dependence, dysthymia, and whether a young person is burdened with comorbid mental disorders.

Prospective prediction of suicide attempts in community adolescents and young adults, using regression methods and machine learning.

BACKGROUND: The use of machine learning (ML) algorithms to study suicidality has recently been recommended. Our aim was to explore whether ML approaches have the potential to improve the prediction of suicide attempt (SA) risk. Using the epidemiological multiwave prospective-longitudinal Early Developmental Stages of Psychopathology (EDSP) data set, we compared four algorithms-logistic regression, lasso, ridge, and random forest-in predicting a future SA in a community sample of adolescents and young adults. METHODS: The EDSP Study prospectively assessed, over the course of 10 years, adolescents and young adults aged 14-24 years at baseline. Of 3021 subjects, 2797 were eligible for prospective analyses because they participated in at least one of the three follow-up assessments. Sixteen baseline predictors, all selected a priori from the literature, were used to predict follow-up SAs. Model performance was assessed using repeated nested 10-fold cross-validation. As the main measure of predictive performance we used the area under the curve (AUC). RESULTS: The mean AUCs of the four predictive models, logistic regression, lasso, ridge, and random forest, were 0.828, 0.826, 0.829, and 0.824, respectively. CONCLUSIONS: Based on our comparison, each algorithm performed equally well in distinguishing between a future SA case and a non-SA case in community adolescents and young adults. When choosing an algorithm, different considerations, however, such as ease of implementation, might in some instances lead to one algorithm being prioritized over another. Further research and replication studies are required in this regard.