Characterization of Childhood Trauma, Hippocampal Mediation and Cannabis Use in a Large Dataset of Psychosis and Non-Psychosis Individuals

Abstract

STRUCTURED ABSTRACTImportanceCannabis (CA) and childhood trauma (CHT) independently increase the risk of earlier development of psychosis. The interaction between CA and CHT in relation to psychosis risk and its association with endocannabinoid receptor rich brain regions such as the hippocampus (HP), remains to be elucidated.ObjectiveTo determine whether CA use and CHT interact to increase risk of developing psychosis and/or lower age of psychosis onset (AO) through mediation by the HP changes and genetic risk, as measured by schizophrenia polygene scores (SZ-PGRS).DesignThis is a cross-sectional, case-control study.SettingMulticenter study in 5 metropolitan US regions.Participants1288 participants, including 493 healthy controls (HC); 210 participants affected by bipolar disorder type 1 with psychosis; 281 by schizoaffective disorder; and 304 with schizophrenia diagnosed using the DSM IV-TR criteria. The HC subjects had no DSM Axis 1 disorder and no first-degree relative with psychotic illness.Intervention(s) (for clinical trials) or Exposure(s) (for observational studies)CHT was assessed using the self-reported Childhood Trauma Questionnaire (CTQ); CA use, abuse, and dependence were assessed by self-reports and trained clinical interviewers. Neuroimaging, symptomatology, and cognition were also assessed.Main Outcome(s) and Measure(s)The relation between age of psychosis onset i.e., time of first appearance of symptoms, measures of CHT, and CA use mediated by the HP, cognition and the SZ-PGRS.ResultsIn survival analysis, low CTQ and CA acted synergistically to lower AO. At high CHT or abuse/dependence CA, CHT or CA were individually sufficient to affect AO. CHT relation with AO is mediated by the HP volume in CA users before AO. Higher SZ-PGRS in probands versus HC, was driven by probands with CA use before AO. Cognition was inversely related to CHT, and positively with higher AO.Conclusions and RelevanceCA use and CHT interact to predict lower age of psychosis onset when moderate. Severe CHT is associated with lower age of psychosis onset irrespective of exposure to cannabis; while abuse/dependence levels of CA use are also sufficient to affect age at psychosis onset irrespective of CHT exposure. CHT effect on AO is mediated by the HP in probands with CA use before AO.Key PointsQuestionAre there interactive effect of cannabis use (CA) and childhood trauma (CHT) on the onset age of psychosis (AO) and symptomatology? Are these effects mediated by the hippocampus (HP)? Does the schizophrenia polygenic risk score (SZ-PGRS) influence how these factors impact AO?FindingsIn this multicenter study including 1288 participants with/out psychosis and with/out CA, high CHT or high CA were independently correlated with AO, these factors interacted to further lower AO. The HP was affected by CHT and mediated CHT effects on AO only in probands that used CA before AO. The PGRS was higher in probands, particularly in probands with CA use before AO.Meaningand CHT. In youth with less severe CHT and moderate levels of CA, CHT and CA may interact to predict speed psychosis onset. In the context of a “three hit” model, the SZ-PGRS in probands with CA before AO might represent a first hit that together with CHT (2nd) and CA (third hit) induces earlier onset of psychotic symptomatology. Reducing CA may be a promising strategy for preventing or delaying onset of psychosis, especially those at genetic risk and with a history of CA.