Trauma under psychedelics: how psychoactive substances impact trauma processing

The Hamas-led attack in southern Israel on October 7, 2023 was one of the deadliest terror attacks in history, resulting in 1,182 fatalities, over 4,000 wounded, and 251 taken hostage1. The Nova festival, an 18-hour rave held in the Gaza Envelope region, suffered the highest casualties in the attack, with over 370 killed. For the Nova attendees, the highlight of the all-night party was sunrise, and many of them reported taking psychoactive substances timed to take effect at dawn. Less than half an hour after sunrise, the first rockets came in sight. As a result, survivors endured prolonged exposure to acute, life-threatening traumatic events, many while under the influence of psychoactive substances. These tragic circumstances created an unprecedented opportunity to examine the effect of peritraumatic exposure to psychoactive substances on short- and long-term impact of severe, life-threatening trauma. From an estimated population of 3,710 Nova survivors (66% male; 76% 18-24 years old), we identified 1,239 eligible individuals. Of these, 923 (74.5%) completed the study questionnaire by February 21, 2024, of whom a total of 107 (11.6%) did not meet the DSM-5 Criterion A for post-traumatic stress disorder (PTSD), as they were not directly exposed, and 44 (4.8%) did not complete all questions, leading to their exclusion. This resulted in an analytic cohort of 772 survivors (487 males; mean age: 26.96±6.55 years). The study was approved by the University of Haifa Ethics Committee, and all participants provided informed consent. We systematically collected data on exposure to psychoactive substances and peritraumatic experiences. Primary outcome measures included the PTSD Checklist for DSM-5 (PCL-5) with cutoff score set at 33, and the Kessler Psychological Distress Scale (K6) with cutoff score set at 13. Secondary outcomes included a 0-100 metric of perceived substance helpfulness, sense of control, and feelings of isolation during trauma exposure. Post-traumatic processing measures included self-perceived social interactions, social support, feelings of guilt, and sleep quality. Seventy-two percent (N=556) of survivors reported being under the influence of psychoactive substances during the attack, with most (79.1%) consuming them within the three hours prior to the attack. Due to the prevalence of polysubstance use and in order to isolate the effects of individual substances, analyses focused on participants that were under the influence of a single substance: hallucinogens (psilocybin or lysergic acid, LSD; N=84); 3,4-methylenedioxymethamphetamine (MDMA) (N=99); or cannabis and/or alcohol (N=68). No substance was used by 216 participants. Group differences were analyzed using a linear regression model with substance groups as independent variables, and a multivariate model with age, sex and time from event as covariates. Only models with the lowest Bayesian information criterion (BIC) are presented. Additional results and polysubstance analyses are presented in the supplementary information. Significant between-groups differences were found on the metric of perceived substance helpfulness during the traumatic event (F2,247=6.14, p=0.003, R2=0.05). Specifically, individuals in the MDMA (62.6±21.7, β=12.4, p=0.001) and hallucinogens (61.5±28.3, β=11.3, p=0.004) groups perceived more substance helpfulness during the traumatic event, as compared to the cannabis/alcohol group (50.2±20.6). This finding is unlikely to be explained by differential scope and impact of the traumatic event, as we found those to be similar across groups (all p values >0.15). Anecdotal reports suggest that survivors who were under the influence of MDMA during the trauma experienced reduced sensations of fear and threat as the event unfolded. PTSD symptom severity scores differed significantly between groups (F3,229=4.8, p=0.003, R2=0.06), with significantly higher PCL-5 scores in the cannabis/alcohol group (48.3±12.8) compared to the no-use group (39.8±14.9, β=8.5, p=0.006). Notably, mean PCL-5 scores across groups were high (41.3±15.3), with all four groups exceeding the clinical cutoff score of 33 (all p values