Abstract
Background
Adolescence is a critical period for the emergence of psychopathology and risk behaviors, including psychotic symptoms and suicidal behavior. Literature has shown that psychotic experiences (PE) are associated with increased odds of suicidal ideation (SI) and attempts in young people. PE are normally regarded as subthreshold positive symptoms, and have been clustered in three domains: persecutory ideation (PI), bizarre experiences (BE) and perceptual abnormalities (PA). All of these domains have been linked to depression, and recent studies demonstrated that, in young people, perceptual abnormalities and persecutory ideation are associated with a higher risk of suicidality, instead, while bizarre experiences were not. Nevertheless, how specific PEs are associated to suicidal is not clearly understood, and the role of common risk factors in this link, such as depressive symptoms (DS), remains controversial. Thus, the aim of this study was to explore the underlying mechanisms linking specific PEs and SI.
Methods
1708 Chilean school adolescents aged 13–19 (M=15.68 + 1.67, women= 39%) answered an online screening aimed at detecting mental health symptoms between May and October 2019. The screening is composed of several questionnaires adapted for Chilean samples, including the Columbia-suicide Severity Rating Scale (CSSRS), the Patient Health Questionnaire (PHQ-9), and the Community Assessment for Psychic Experiences (CAPE-P15). Informed consent was granted by the parents and caregivers of all school adolescents. We conducted mediation analyses with the PROCESS statistical package, which runs regression-based mediations through a 5000 bootstrap resampling to estimate confidence intervals.
Results
Significant association were found among all variables in the study (SI-DS: r= .624, p<.001; PI-SI: r= .495, p<.001; PA-SI= .391, p<.001; PI-DS: r= .660, p<.001; PI-DS: r= .429, p<.001; PI-PA: r= .442, p<.001), except for BE with SI and SD. No demographic variables were correlated significantly to the dependent variable, thus these were controlled for in the mediation analyses. In the multiple mediation model, the link between abnormal perceptions and suicidal ideation is mediated by both persecutory ideation (PI) and depressive symptoms (DS) (b= .1278, 95% BCa CI [.1044, .1526]). Moreover, PI mediates the association between perceptual abnormalities (PA) and DS (b= .0494, 95% BCa CI [.0234, .0769], and DS mediates the link between PI and SI (b= .084, 95% BCa CI [.0596, .1077]). When the mediators were not included in the model, PA significantly predicted SI (b = .361, t = 17.55, p < 0.001). Two additional models were carried out using separate PA (auditory and visual hallucinations) as the predictor variables. In model 1, the link between visual PA and SI was mediated by both PI and DS (b= .1087, 95% BCa CI [.0874, .1326]). A significant indirect effect was also found in model 2, regarding auditory PA (b= .1175, 95% BCa CI [.0933, .1435]).
Discussion
Our findings highlights the underlying role of DS along with specific types of PEs, particularly with PI, as the pathway to SI. The consistency of the significant, yet small indirect effects in all three mediations, could indicate a relative robustness in the model. Our findings could suggest that depression does not influence suicidal behavior independently, but rather, it interacts with other psychopathological elements, such as PI, to influence the development of suicidal ideation. This supports literature stating that any aspect of the suicide continuum is the result of a complex interplay between numerous contributing psychosocial factors.
Psychotic symptoms in young people without psychotic illness: mechanisms and meaning
