A randomized controlled trial of 3,4-methylenedioxymethamphetamine (MDMA) and fear extinction retention in healthy adults

Posttraumatic stress disorder (PTSD) is a disabling condition for which trauma-focused psychotherapies such as prolonged exposure (PE) are effective but not universally successful. Maples-Keller and colleagues note growing interest in MDMA-assisted psychotherapy as a potential treatment for PTSD; mechanistic work from animal models suggests that MDMA given before extinction training can enhance long-term extinction retention, an effect linked to amygdala and medial prefrontal cortical activity and increased BDNF expression in rodents. Extinction learning and extinction retention (the within-session reduction of conditioned fear and the later retrieval of that learning, respectively) are widely used translational models for fear-related symptoms of PTSD and are theorised to underlie PE's therapeutic effects. This randomised, placebo-controlled human trial tested whether acute MDMA administration would enhance extinction learning or, primarily, extinction retention in healthy adults. The investigators hypothesised that participants who received 100 mg MDMA prior to extinction training would show reduced fear-potentiated startle during an extinction retention test 48 hours later compared with placebo-treated participants. This study therefore aimed to translate preclinical findings into a human experimental fear-conditioning paradigm and to assess safety and tolerability of acute MDMA in this context.

This was a parallel-group, randomised, placebo-controlled trial conducted at Emory University between March 2018 and July 2020. Thirty-four healthy adults aged 21–55 were randomised 1:1 to a single oral dose of 100 mg MDMA or matching placebo; all randomised participants completed the protocol. Randomisation used permuted blocks prepared by an independent biostatistician and was maintained by an unblinded study physician; participants and all other study staff were blinded to assignment. The protocol received institutional review board approval and adverse events and vital signs were monitored throughout. Eligibility required prior recreational MDMA use without adverse experience, negative drug screen at key visits, and standard medical and psychiatric exclusions (for example, lifetime bipolar disorder or primary psychotic disorder, recent substance use disorder, pregnancy, significant cardiac or neurological illness). Key exclusions also included more than 10 lifetime MDMA uses in 10 years or any MDMA use in the past 6 months. Participants completed a baseline visit with psychiatric assessment (MINI interview), medical screening, and the fear acquisition session. The experimental timeline comprised three visits: Visit 1—fear acquisition; Visit 2—extinction training conducted 2 hours after dosing (aligned to MDMA Tmax) with physiological monitoring for 6 hours post-dose; and Visit 3—extinction retention testing 48 hours after Visit 2. The behavioural paradigm was a well-validated fear‑potentiated startle task measured with orbicularis oculi EMG. The unconditioned stimulus (US) was a 250 ms 140 psi airblast to the larynx and conditioned stimuli (CS+ and CS-) were coloured shapes presented on a monitor. Acquisition involved CS habituation followed by three acquisition blocks with reinforced CS+ and non‑reinforced CS- trials and noise-alone (NA) trials. Extinction training presented the same CSs without US across four extinction blocks at Visit 2. Extinction retention at Visit 3 comprised four trials of each type presented without the US. Primary outcome was fear‑potentiated startle during extinction retention, operationalised as the increase in startle magnitude when a CS was presented compared with baseline acoustic startle (NA). Repeated measures ANOVA (RM‑ANOVA) was used to test acquisition, discrimination (CS+ vs CS-), extinction training, and extinction retention effects with group (MDMA vs placebo) as the between‑subjects factor. Given individual variability in extinction, the investigators also conducted post hoc categorical analyses classifying participants as "retainers" (no increase from end of extinction training to retention) versus "non‑retainers" and compared proportions across groups using chi‑square tests.

Thirty‑four participants were randomised and completed Visits 2 and 3. Groups did not differ on reported demographic variables. During acquisition both groups learned the conditioned fear: RM‑ANOVA showed significant block × trial type interactions and main effects indicating greater startle to the reinforced CS+ across acquisition blocks. Separate analyses within each group confirmed higher startle to CS+ versus NA (both p < 0.001), and there were no significant differences between MDMA and placebo groups during acquisition or in CS+/CS- discrimination. During extinction training (administered with drug on board), baseline startle (NA) did not differ between groups. The full sample demonstrated significant within‑session reductions in fear‑potentiated startle to the CS+ across extinction blocks (F(3,96) = 21.50, p < 0.001), with no significant group main effect or group × block interaction (group F(1,32) = 1.31, p = 0.26), indicating comparable extinction learning under MDMA and placebo. Extinction retention was the primary endpoint. As a group, participants showed significant spontaneous recovery (a return of fear) from the end of extinction training to the retention session; the session effect was significant (F(1,32) = 26.87, p < 0.001) but there was no significant session × group interaction or overall main effect of group on the continuous retention outcome. In exploratory categorical analyses, however, a greater proportion of participants in the MDMA arm met the investigators' predefined criterion for "retainer" (no increase in fear‑potentiated startle from end of extinction to retention): 6 of 17 in the MDMA group versus 0 of 17 in the placebo group (χ2 = 7.29, p = 0.007). Within the MDMA group, retainers showed stable extinction with no spontaneous recovery, whereas non‑retainers exhibited spontaneous recovery (RM‑ANOVA block × group interaction, F(1,15) = 9.34, p = 0.008). Safety and tolerability were acceptable in this sample. No serious adverse events occurred and no participants withdrew for harms. The most common adverse events in the MDMA group were transient elevated blood pressure (3/17), headache (2/17), anxiety (3/17), and tachycardia (2/17); similar events were rare or absent in placebo. Repeated measures analyses of physiological data across eight timepoints during Visit 2 revealed significant time × group interactions for pulse (F(1,31) = 51.37, p < 0.001), systolic blood pressure (F(1,31) = 78.73, p < 0.001), diastolic blood pressure (F(1,31) = 18.69, p < 0.001), and body temperature (F(1,30) = 14.62, p = 0.001), with transient increases in the MDMA group that remained within normal ranges.

Maples‑Keller and colleagues interpret the findings as showing that acute MDMA given before extinction training does not impair within‑session extinction learning and, at the group level, does not produce a clear enhancement of extinction retention in healthy adults using this paradigm. Successful acquisition and extinction across the sample indicate the experimental procedures are feasible under MDMA administration, and physiological and subjective effects did not prevent engagement with the task. The primary hypothesis—that MDMA would enhance extinction retention—was not supported in continuous group comparisons; most participants in both arms showed spontaneous recovery at the retention test, which is a common outcome in this paradigm. The authors highlight an exploratory finding: a significantly greater proportion of participants in the MDMA group (6/17) met a strict criterion for retaining extinction (no increase from end‑of‑extinction levels) versus none in the placebo group. They propose this may indicate MDMA facilitates extinction retention in a subset of individuals and suggest future research should examine individual difference moderators such as hormonal, genomic, or experiential factors. The discussion also situates the results within neurobiological hypotheses, noting overlap between MDMA's effects on amygdala and hippocampal perfusion and circuits implicated in extinction and PE therapy, and proposing that MDMA might act via extinction retention or via effects on reconsolidation depending on timing of administration. Key limitations acknowledged by the investigators include the small, highly selected healthy sample that limits generalisability, challenges maintaining blind with an inert placebo, and the specific timing of dosing: MDMA was given about 2 hours before extinction to align with Tmax in humans whereas many animal studies administered MDMA earlier relative to extinction. The authors recommend follow‑up studies with larger and more representative or clinical samples (notably PTSD patients with documented extinction deficits), varied timing to probe reconsolidation effects, neuroimaging to examine circuit mechanisms, and comparisons with stimulant drugs to probe pharmacological specificity. Overall, the authors conclude that while the primary hypothesis was not confirmed, the exploratory retainer result and safety data support further translational and clinical investigation of MDMA's impact on fear extinction processes.