Drug-induced social connection: both MDMA and methamphetamine increase feelings of connectedness during controlled dyadic conversations

Molla and colleagues frame MDMA as a stimulant-like compound with pronounced empathogenic and pro-social effects that may underlie both its recreational popularity and its emerging therapeutic utility, particularly as an adjunct in psychotherapy for post-traumatic stress disorder. The introduction notes that while animal and some human laboratory work have shown MDMA increases social approach, sociability and socioemotional processing, few controlled studies have examined how MDMA affects actual in-person social interactions, whether such effects are pharmacologically specific, or which physiological mechanisms (for example, oxytocin release) might mediate them. This paper reports two parallel within-subject, double-blind studies designed to address that gap. Study 1 tested the effects of MDMA (100 mg versus placebo; initial N = 18) and Study 2 tested methamphetamine (MA; 20 mg versus placebo; initial N = 19) on feelings of connectedness during a semi-structured 45-minute conversation with a previously unknown, same-sex partner. Salivary oxytocin and cardiovascular and subjective measures were also obtained to explore physiological correlates and the pharmacological specificity of any pro-social effects.

Two separate studies were run in parallel cohorts. Both used a within-subject, double-blind, placebo-controlled crossover design in which participants completed two laboratory sessions (drug and placebo) spaced at least 3–4 days apart. Study 1 administered MDMA 100 mg versus placebo; Study 2 administered methamphetamine 20 mg versus placebo. Capsules for active drugs and placebo were opaque and visually matched. Participants were informed their capsule might contain placebo, a stimulant, a sedative, or a hallucinogen to reduce expectancy effects. Healthy men and women aged 18–35 were recruited and screened with physical exam, ECG, psychiatric interview, medical and drug-use history and the SCL-90R. Exclusion criteria included cardiovascular abnormalities, current DSM-V diagnoses of substance dependence or mood, anxiety or psychotic disorders, past treatment for drug or alcohol use disorder, and pregnancy. Additional inclusion criteria were BMI 19–30, fluency in English and limited recent alcohol/caffeine consumption. For Study 1 only, lifetime MDMA use of 1–40 times was required. Written informed consent was obtained and procedures were approved by the institutional review board. Sessions ran 4.5 hours from morning. After arrival and abstinence verification by urinalysis and breath alcohol test, baseline subjective and cardiovascular measures were obtained. Participants ingested the capsule at 9:30 am and, at the expected peak window (70–115 minutes post-dose), engaged in a 45-minute semi-structured ‘‘small talk’’ conversation with a previously unknown, same-sex partner who was blind to drug condition and trained to interact naturally. Conversations were audiotaped. Subjective measures (Drug Effects Questionnaire, Visual Analog Scales for multiple adjectives, Profile of Mood States) and cardiovascular measures were taken repeatedly; a single saliva sample for oxytocin was collected at 120 minutes post-dose. At session end participants completed questionnaires assessing connection and closeness (Conversation Questionnaire, Connection During Conversations Scale (CDCS) with four subscales, and the Inclusion of Other in Self (IOS) single-item), and one-week online follow-up ratings of the two conversations and partners. Data were analysed separately for the two studies. Primary outcomes were the conversation rating questionnaires completed at end-of-session; analyses used mixed-model ANOVAs with drug condition as the within-subject factor and sex and session order as between-subject factors. Repeated subjective and cardiovascular measures were reduced to peak change from baseline and compared with paired t-tests. Pearson correlations assessed relationships between oxytocin (drug minus placebo) and closeness ratings. Significance was set at p < 0.05. The extracted text does not clearly report the final analysed sample sizes after exclusions for every analysis.

Two participants were excluded from analysis of Study 1 (one for strong negative feelings about a partner and one for socialising with a partner before follow-up); the initial sample sizes were N = 18 (MDMA study) and N = 19 (MA study). Demographically most participants were in their twenties with some college education and low-to-moderate prior drug use; Study 1 participants reported higher lifetime stimulant use consistent with the MDMA-use inclusion criterion. Primary social outcomes, end-of-session: In Study 1 (MDMA), there were no significant interactions with sex or order. Relative to placebo, MDMA significantly increased ratings of liking the conversation partner (F1,13 = 8.2, p = 0.01, ηp2 = 0.39), enjoyment of the conversation (F1,13 = 7.1, p = 0.02, ηp2 = 0.36) and meaningfulness of the conversation (F1,13 = 8.9, p = 0.01, ηp2 = 0.41). On the IOS there was a trend toward greater connection (F1,13 = 4.7, p = 0.05, ηp2 = 0.27). On the CDCS, MDMA increased several partner-related items (e.g., ‘‘They were interested in my thoughts and feelings’’, F1,13 = 7.7, p = 0.02, ηp2 = 0.37) and raised subscale scores for partner responsiveness (F1,13 = 15.4, p = 0.002, ηp2 = 0.54) and participant interest (F1,13 = 5.4, p = 0.04, ηp2 = 0.30). In Study 2 (MA), there were no systematic sex or order effects. Compared to placebo, MA increased ratings of conversational enjoyment (F1,15 = 13.8, p = 0.002, ηp2 = 0.48) and meaningfulness (F1,15 = 6.3, p = 0.02, ηp2 = 0.30), liking of the partner (main effect F1,15 = 15.2, p = 0.001, ηp2 = 0.50), perceived partner liking of the participant (F1,15 = 9.1, p = 0.009, ηp2 = 0.38) and closeness (F1,15 = 10.9, p = 0.005, ηp2 = 0.42). MA also increased IOS scores (F1,15 = 8.6, p = 0.01, ηp2 = 0.36) and several CDCS items and all four CDCS subscales, including shared reality (F1,15 = 7.2, p = 0.02, ηp2 = 0.32) and affective experience (F1,15 = 18.5, p = 0.001, ηp2 = 0.55). MA decreased ratings on several negative conversation items (for example ‘‘I thought they were boring’’, F1,15 = 14.7, p = 0.002, ηp2 = 0.49). Subjective drug effects and mood: On the DEQ both drugs increased ‘‘feel drug’’, ‘‘like drug’’, ‘‘high’’ and ‘‘want more’’. On VAS measures both MDMA and MA increased ratings of stimulated, insightful, sociable, loving and friendly. MDMA uniquely elevated feelings of trusting, appreciated, grateful and loved; MA uniquely increased feelings of being understood. On the POMS both drugs increased elation, friendliness and vigour; MDMA increased anxiety and confusion while MA reduced fatigue. End-of-session identification rates showed most participants identified active drugs as stimulants (MDMA: 75% identified MDMA as stimulant; MA: 74% identified MA as stimulant). Follow-up (one week): In Study 1 participants reported the MDMA conversation as more meaningful than the placebo conversation (F1,12 = 5.8, p = 0.03, ηp2 = 0.33). In Study 2 participants reported the MA conversation as more enjoyable (F1,15 = 22.6, p = 0.0003, ηp2 = 0.60), more meaningful (F1,12 = 12.9, p = 0.003, ηp2 = 0.46), and they reported greater liking and closeness to MA partners on follow-up. Ratings of partners and perceptions: After MDMA participants rated their partners as more physically attractive (F1,12 = 5.5, p = 0.04, ηp2 = 0.31) and warmer. Participants also reported that MDMA partners would perceive them as less competent but more warm. After MA participants reported their MA partners as warmer and believed MA partners perceived them as more intelligent, kind and warm. Physiological and oxytocin measures: Both drugs increased systolic and diastolic blood pressure and heart rate relative to placebo. Mean salivary oxytocin was higher after both MDMA and MA compared with placebo, but many samples were below the assay detection limit. In Study 1, 12 of 17 placebo samples were below detectability and 1 of 17 MDMA samples were below detectability; in Study 2, 18 of 19 placebo and 9 of 19 MA samples were below detectability. Undetectable samples were set to the assay lower limit (0.313 pg/mL). Oxytocin correlations: Drug-minus-placebo oxytocin changes correlated positively with drug-minus-placebo closeness ratings in the MDMA study (p = 0.03) but not in the MA study (p = 0.8). No significant correlations were found between oxytocin and heart rate for either drug. The extracted text does not provide confidence intervals for these correlations.

Molla and colleagues interpret their findings as demonstrating that MDMA robustly increases perceived connection and the meaningfulness of a brief in-person conversation with a stranger, and that methamphetamine produced a surprisingly similar increase in feelings of closeness and enjoyment. They highlight that these effects were evident during an actual dyadic interaction tested under double-blind conditions, extending prior work that mostly used indirect or computerised social tasks or self-report in group settings. The authors consider several possible mechanisms. For MDMA, they note prior evidence that the drug alters socioemotional processing — for example by changing recognition of emotional facial expressions, increasing neural responses to positive stimuli, and enhancing explicit and implicit empathy — and suggest these changes could make social encounters feel more rewarding. MDMA also produced a larger increase in salivary oxytocin and oxytocin changes correlated with closeness ratings after MDMA but not after MA, consistent with a serotonin–oxytocin pathway for some of MDMA's pro‑social effects. For MA, they propose that dopaminergic and noradrenergic actions might instead underlie increased attentiveness and reduced negative conversational aspects, thereby indirectly enhancing perceived connection. The authors candidly discuss limitations that constrain interpretation. The two studies were not directly comparable: only one dose of each drug was tested and participants in the MDMA study were required to have prior MDMA experience while most in the MA study were not. Salivary oxytocin measurement was particularly problematic — only a single post-dose sample was collected, many samples were below assay detection limits and no baseline oxytocin sample was obtained — making conclusions about oxytocin preliminary. Molla and colleagues also note that other measures of connectedness (including broader constructs like connectedness to self or the world) were not assessed and could be informative in future work. Finally, the authors outline implications without overstating them. They suggest enhanced patient–therapist connectedness might be one mechanism by which MDMA facilitates psychotherapy, that connectedness could serve as a measurable target in therapy protocols, and that findings warrant exploration of whether other drugs might also facilitate therapeutic relationships. They recommend future studies to compare drugs directly, test dose-response relationships, obtain more sensitive and repeated oxytocin measures, and investigate how different social contexts modulate drug effects. The procedure used here is presented as a useful model for studying pharmacological modulation of real-life social interaction.