Effects of MDMA on socioemotional feelings, authenticity, and autobiographical disclosure in healthy volunteers in a controlled setting

Design and participants. Baggott and colleagues employed a double-blind, placebo-controlled, within-subject crossover design with gender balance. Twelve healthy, MDMA-experienced volunteers (six male, six female), aged 21–40 years (mean 29±2 years, SEM), completed two experimental sessions separated by at least one week. Inclusion required medical clearance by a licensed physician; exclusion criteria included current DSM-IV dependence (except nicotine or caffeine), significant medical or psychiatric conditions, recent interacting medications, and BMI outside 18–30 kg/m2. Participants had a mean of 24±7 prior MDMA experiences (range 5–75). Intervention, setting and blinding. Each participant received one oral active dose of MDMA (1.5 mg/kg as the hydrochloride salt) and one placebo across sessions; dose selection was based on prior clinical work. Sessions occurred after an overnight hospital stay in a redecorated clinical room intended to resemble a living room or therapist's office. Participants were discharged at least 6 h after administration and returned for a brief 24 h follow-up. To support blinding, consent allowed for the possibility of receiving one or two active doses, although the study used a single active and a single placebo session per participant. Autobiographical memory task and narrative analysis. In a pre-session screening, participants generated candidate autobiographical events in four emotion categories (fear, safe, sad, joy) and rated episodes on intrusion symptoms, affect (PANAS), recency, vividness, involvement, consequence, and confidence. Two matched episodes per category were randomly assigned to the two drug conditions. During each drug session, beginning 1.5–2 h post-administration, participants spent 5 minutes describing each of four memories (20 minutes total), with Joy always presented last. After each description they completed visual analogue scale (VAS) ratings about how upsetting, comfortable, relived, and well remembered or understood the episode was. Speech was digitally recorded, transcribed, and analysed with the Linguistic Inquiry and Word Count (LIWC) program; 43 LIWC categories previously used by the group were examined. Self-report measures and timing. Time-course VAS items (0–100 scale) measured general drug effects (e.g. any drug effect, liking, high), stimulant/sedative sensations (e.g. anxious, stimulated, drunk) and affective/social items (e.g. loving, kind). These were collected pre-dose and at 1, 2, 3, 4, 5 and 6 h post-dose; peak baseline-subtracted responses were analysed. Social anxiety was assessed with a modified Brief Fear of Negative Evaluation-revised (BFNE) with altered Likert anchors to capture state changes; this was given before and at 1.5, 2, 2.5 and 24 h post-dose. State authenticity was measured using a modified 45-item Authenticity Inventory (items and instructions adjusted to index current feelings rather than trait), administered at ~2.5 h post-dose. Interpersonal functioning was assessed at the same time using a shortened 32-item Interpersonal Adjectives Scale-Revised (IASR), focusing on Dominance/Agency and Nurturance/Communion dimensions. Statistical analysis. The investigators used mixed-effects models with participant as a random effect and fixed effects for drug condition, emotional memory category, and their interaction for analyses of VAS, autobiographical ratings, and LIWC outcomes. Peak baseline-subtracted VAS scores were modelled for drug effects. For LIWC, each category was modelled separately with emotion and condition predictors. Power calculations were based on a prior effect size (Cohen's d≈0.93) for a "Closeness to others" VAS item, indicating n=12 would give ~80% power at p=0.05 for that item; replication of speech findings had lower power (~72%). The extracted text notes some missing self-report data (computer failure and version control errors) and that exploratory gender interaction tests were underpowered and not reported in detail.

Participant and task characteristics. All 12 participants completed the protocol. The autobiographical episodes selected for the two sessions did not differ significantly on recency or impact measures. Typical content per emotion category was described: Fear involved violence, accidents, arrest or illness threats; Joy involved relationship changes, achievements, childbirth; Sad involved loss and separation; Safe involved returns from danger, stability or good news. Experience of remembering. Participants reported greater comfort discussing emotional memories on MDMA. A mixed-effects model found a main effect of drug condition (F1,80=5.30, p=0.024), with participants reporting on average 10.4±4.5 VAS units greater comfort on MDMA compared to placebo. Emotion category also influenced outcomes: Joy memories were easier to recount than Sad, Fear, or Safe (p-values 0.006, 0.005 and 0.015 respectively). There were no significant condition-by-emotion interactions. MDMA did not significantly change self-reported ability to remember, describe, understand, or re-live emotions associated with the memories. Main effects of emotional category were observed for reliving (F3,77=3.47, p=0.020) and for how upsetting talking about the memory was (F3,77=11.8, p<0.001). Narrative content (LIWC) and word count. MDMA reduced overall word count from 651.8±42.33 words after placebo to 592.3±45.98 words after MDMA (F1,107=9.46, p=0.003). In LIWC analyses, MDMA increased use of present-tense verbs (F1,77=8.22, p=0.005), words of assent (e.g. agree, okay, yes; F1,77=5.07, p=0.027), and family-related words (e.g. daughter, husband; F1,77=4.51, p=0.037). The authors note that these results would not survive correction for multiple comparisons across 43 categories. Emotional memory category significantly affected 14 of the 43 LIWC categories, largely those related to affect and social language (affective processes, positive and negative emotion, anger, anxiety, sadness, 'feel', social processes), with F-values for these effects ranging from 2.86 to 50.1 (p-values 0.042 to <0.001). Visual analogue measures. MDMA produced robust increases in general drug-effect VAS items. Mixed-effects models showed condition effects on peak baseline-subtracted scores for Any drug effect (F1,11=82.35, p<0.001), Good drug effect (F1,11=103.18, p<0.001), Drug liking (F1,11=95.12, p<0.001), and High (F1,11=82.08, p<0.001). Both stimulant-like and sedative-like sensations increased, and MDMA increased peak ratings of Anxious. Social-emotional VAS items Love and Kind showed significant increases: Loving (F1,10=6.53, p=0.029) and Kind (F1,10=7.24, p=0.023). No significant condition effects were detected for Adventurous, Amused, Closeness to others, Contented, Insightful, Proud, Passionate, Self-conscious, or Trusting. Social anxiety, authenticity and interpersonal functioning. MDMA reduced state social anxiety as measured by the modified BFNE: MDMA decreased maximum magnitude change from baseline BFNE scores (F1,10=7.7, p=0.019). State authenticity increased on MDMA (total authenticity score main effect: F1,10=12.07, p=0.006). On the IASR, MDMA produced a rightward shift on the Nurturance/Communion dimension (increased affiliative feelings). A mixed-effects model found condition predicted higher Nurturance (F1,11=5.52, p=0.039), and the Gregarious subscale increased (F1,11=8.49, p=0.014). No significant changes were found on the Dominance/Agency dimension. Missing data and subgroup analyses. Some VAS data were missing due to a computer failure and version control errors affecting the Self-conscious item for a few participants. Exploratory analyses including gender and condition-by-gender interactions were underpowered and did not reveal significant gender effects according to the extracted text.

Baggott and colleagues interpret their findings as evidence that MDMA produces a prosocial syndrome in healthy volunteers that increases state authenticity, reduces social anxiety (fear of negative evaluation), and enhances comfort disclosing emotional autobiographical material, even while producing both stimulant-like and sedative-like subjective effects and sometimes increasing general anxiety. The pattern—greater affiliative feelings alongside increased self-reported anxiety—led the authors to argue against a simple general anxiolytic account of MDMA. Instead, they frame the results as consistent with a tend-and-befriend style stress response, in which anxiety coexists with caregiving and affiliative behaviours; the authors note rodent data showing decreased social aggression and increased social interaction under MDMA despite anxiety-like behaviours on other tasks. The investigators suggest these socioemotional effects could help explain why MDMA might aid psychotherapy by improving the therapeutic alliance and facilitating discussion of stressful autobiographical material. However, they did not find evidence that MDMA increased reported insight or understanding of memories, and they offer several possible reasons: such effects may have been absent, the memories had been recently rehearsed during screening and so were already well understood, or the sample size lacked power to detect subtle changes. Speech-content findings are described as only partially consistent with prior work. Whereas earlier studies reported increases in positive emotion or other specific categories, this study found increased present-tense use, assent words, and family-related words; the authors caution that these speech results would not survive correction for multiple comparisons and may reflect task differences (describing a loved one versus describing mixed autobiographical memories), limited power, or type II and type I errors. On authenticity, the authors emphasise that MDMA increased state feelings of being "one's true self," a construct linked to reduced self-censorship and greater wellbeing; they note this may distinguish MDMA from classical psychedelics, which can produce both self-insight and depersonalisation, but they also acknowledge that positive mood can itself increase state authenticity and comparisons with stimulants were not made in this study. Key limitations acknowledged by the authors include the small sample size, truncation of narratives at 5 minutes which may have truncated salient material, imperfect matching of autobiographical memories (notably temporal duration and narrative complexity), the fact that some "safe" memories included initial fear, and the modification of validated instruments (BFNE and Authenticity Inventory) to measure state rather than trait. The study did not include an active stimulant comparator to test specificity of MDMA effects. Finally, the authors discuss broader measurement challenges in MDMA research: general drug-effect measures are very sensitive but non-specific, while socioemotional measures can be insensitive, highly variable, and context-dependent. In conclusion, the investigators describe their findings as preliminary but suggestive: MDMA decreased social anxiety, increased affiliative feelings and authenticity, and enhanced comfort in disclosing autobiographical material, occurring alongside mixed stimulant and sedative subjective effects. They state these effects are consistent with the idea that MDMA constitutes a distinctive pharmacological class with psychotherapeutic potential, while emphasising the need for replication and further research.