Reduction in social anxiety after MDMA-assisted psychotherapy with autistic adults: a randomized, double-blind, placebo-controlled pilot study

Earlier research indicates that 3,4-methylenedioxymethamphetamine (MDMA) produces transient increases in social affiliation, reduces negative responses to social rejection, and promotes release of oxytocin, suggesting potential utility as an adjunct to psychotherapy for disorders of social fear. Danforth and colleagues note that autistic adults experience high rates of social anxiety disorder (SAD) and face limited benefit from standard treatments; biological differences in neurotransmitter systems among autistic people may also alter responses to conventional anxiolytics. Prior epidemiological reports of MDMA/ecstasy use by autistic adults further motivated investigation of MDMA-assisted psychotherapy for social anxiety in this population. This study set out to explore the feasibility and safety of MDMA-assisted psychotherapy for reducing social fear and avoidance in autistic adults with marked to very severe SAD. The trial was designed as a randomized, double-blind, placebo-controlled pilot study to assess change in social anxiety (primary outcome: Leibowitz Social Anxiety Scale, LSAS) from baseline to one month after the second experimental session, with additional follow-up at six months. The investigators emphasise that the intent was remediation of SAD symptoms, not to alter core features of autism itself.

This single-site exploratory Phase II trial used a randomized, double-blind, placebo-controlled design conducted between February 2014 and April 2017. Twelve autistic adults meeting inclusion criteria were enrolled and randomised 2:1 to receive MDMA (n = 8) or inactive placebo lactose (n = 4). Key eligibility requirements included age ≥21, self-reported MDMA-naïvety, physical health, psychological stability, and a baseline LSAS total score of 60 or greater, indicating marked to severe social anxiety; diagnostic instruments used included the ADOS-2 (Module 4) and SCID-I-RV. The trial protocol was approved by an institutional review board and conducted according to Good Clinical Practice. Intervention comprised psychosocial support plus two blinded 8-hour experimental sessions spaced approximately one month apart. Participants completed three 60–90 minute non-drug preparatory psychotherapy sessions before the first experimental session and three 60–90 minute integrative psychotherapy sessions after each experimental session. The psychotherapy package incorporated standardised mindfulness-based elements adapted from dialectical behaviour therapy to support emotion regulation and interpersonal skills. MDMA dosing followed a dose-finding strategy: four participants received 75 mg then 100 mg, and four received 100 mg then 125 mg; placebo capsules were identical in appearance and weight. Participants who received placebo were offered optional open-label MDMA sessions after unblinding (those data are not presented here). Assessments centred on the LSAS administered by a single blinded independent rater at baseline, 1 day, 2 weeks and 4 weeks after each experimental session, and at the 6-month visit. Secondary and exploratory measures included the BDI-II, STAI (Y-2), Perceived Stress Scale, Interpersonal Reactivity Index, Rosenberg Self-Esteem Scale, Toronto Alexithymia Scale (TAS-20), TASIT, and Emotion Regulation Questionnaire. Pharmacodynamic monitoring (blood pressure, heart rate, temperature) occurred pre-dose and hourly for 6–7 hours post-dose. Safety monitoring included adverse event (AE) capture at each visit, C-SSRS suicidality assessments, hourly subjective units of distress during sessions, and daily telephone checks for 7 days post-session. A consented study support partner accompanied participants home after sessions. For statistical analysis, data from the MDMA dose subgroups were combined. Independent samples t-tests compared change in LSAS total score from baseline to the primary endpoint (1 month after second experimental session) and from baseline to 6-month follow-up; alpha was set at 0.05 (two-tailed). Effect sizes were reported as Cohen's d for independent-groups pretest–post-test designs. One participant in the MDMA group had primary outcome data missing due to emerging medical-history information that rendered them ineligible; baseline scores were retained for intent-to-treat but missing data were not imputed. Analyses of secondary measures were described as exploratory and presented with descriptive statistics.

Twelve participants were randomised after screening 49 telephone respondents and 24 in-person assessments. Mean age was 31.3 years (SD 8.8); 83.3% identified as male and 16.7% as female (all females were randomised to MDMA). Comorbidities were common: 66.7% had a history of depression, 41.7% had generalized anxiety disorder, and 100% met criteria for SAD. Most participants had prior psychotherapy and many had previous pharmacological treatments. Primary outcome: Change in LSAS total score from baseline to the primary endpoint was significantly greater in the MDMA group than in the placebo group (t(9) = 2.451, P = 0.037; 95% CI for the mean difference 1.92 to 47.87). The placebo-subtracted Cohen's d was reported as 1.4 (95% CI -0.074 to 2.874). At the 6-month follow-up the MDMA group also showed a significantly larger decline from baseline compared with placebo (t(9) = 2.454, P = 0.036; 95% CI 1.92 to 47.01), with a placebo-subtracted Cohen's d of 1.1 (95% CI -0.31 to 2.53). Mean LSAS scores for the MDMA group were substantially lower at primary endpoint and changed little between the primary endpoint and 6-month follow-up, consistent with durability of effect; statistical comparison of primary endpoint to 6-month scores in the MDMA completer subgroup did not reach significance (t(6) = 1.117, P = 0.307). Using an alternate definition of clinical response as a ≥20-point LSAS reduction, 6/8 (75%) participants in the MDMA group met this threshold versus 2/4 (50%) in the placebo group. Secondary and exploratory measures showed descriptive changes that the authors report were generally similar between groups, but these analyses were not formally tested for significance. Pharmacodynamic outcomes reflected expected sympathomimetic effects of MDMA. Peak systolic blood pressure was significantly higher in the MDMA group (P = 0.021); mean peak pulse was also higher (P = 0.015) with a maximum observed pulse of 114 bpm. Peak body temperature was significantly elevated in the MDMA group (P < 0.001) with maximum observed temperature 37.7 °C; no participants reached SBP >180 mmHg or DBP >110 mmHg, and vitals returned to pre-drug levels by session end without clinical intervention. Safety and tolerability: No serious adverse events were reported. The most common spontaneously reported reactions during or shortly after experimental sessions were anxiety (75.0% MDMA vs 25.0% placebo) and difficulty concentrating (62.5% MDMA vs 25.0% placebo); fatigue, headache, and cold sensitivity were also more frequent in the MDMA group. One severe spontaneously reported reaction—a headache on day 1 post-drug—occurred in the MDMA group; otherwise AEs were rated mild or moderate. Suicidal ideation was reported by 25.0% of participants in each group but reflected pre-existing history in most cases and was assessed as low risk. Overall, the investigators report no acute cardiovascular or hyperthermia crises and no deleterious outcomes detected on long-term follow-up. Blinding fidelity: Participants, therapists, and the independent rater were blinded. Incorrect guesses of treatment assignment by participants occurred infrequently (one incorrect guess in eight placebo sessions, 12.5%); therapist misclassification rates were reported at similar low levels. One participant who received MDMA reported no subjective acute effects in-session despite laboratory confirmation of MDMA in plasma; the authors note possible influences such as prior SSRI exposure or other individual factors.

Danforth and colleagues interpret the findings as evidence that MDMA-assisted psychotherapy is feasible and can be safely administered in a controlled clinical setting for autistic adults with marked social anxiety. They highlight a significantly greater mean decline in LSAS scores for the MDMA group at both the primary endpoint and six-month follow-up, and report very large placebo-subtracted effect sizes (d = 1.4 at primary endpoint; d = 1.1 at six months). The authors note that treated participants typically moved down multiple severity categories on the LSAS and that subjective reports and semi-structured interview material aligned with quantitative reductions in social anxiety and avoidance. In considering mechanisms, the investigators connect observed therapeutic effects to MDMA's serotonergic and oxytocinergic actions and to prior findings of reduced amygdala reactivity under MDMA, positing that these neurobiological effects may have enabled participants to access previously inaccessible emotional states, engage with corrective experiences, and reveal latent social skills. The discussion emphasises that these clinical changes occurred without formal social skills training or psychoeducation. The authors acknowledge several limitations that temper interpretation. Chief among these are the small sample size and the consequent limits on generalisability, inability to assess dose–response relationships, and heterogeneity in baseline symptomatology that complicated interpretation of exploratory measures. They also identify challenges inherent to blinding in psychoactive drug trials, potential imprecision in autism diagnostic history for some participants, recruitment biases related to online outreach and possible self-selection, and the lack of objective confirmation of lifetime MDMA abstinence. Safety considerations are presented cautiously: although transient increases in anxiety and sympathomimetic vital signs were common, events were generally mild or moderate and resolved without serious medical consequences. Overall, the investigators conclude that the pilot data support further research in larger samples to evaluate efficacy, mechanisms, and dose optimisation of MDMA-assisted psychotherapy for SAD in autistic adults.

The authors conclude that this pilot trial established feasibility and safety of MDMA-assisted psychotherapy for social anxiety disorder in autistic adults. Changes in LSAS scores and subjective observations were consistent with the hypothesis that combining MDMA with structured preparatory and integrative psychotherapy can alleviate anxiety that interferes with social functioning. Danforth and colleagues recommend replication in larger trials to confirm efficacy and further evaluate this approach.