Effects of ecstasy on cooperative behaviour and perception of trustworthiness: A naturalistic study

MDMA (ecstasy) is reported to produce stimulant and pro-social effects, including increased feelings of friendliness, closeness and empathy. Previous clinical research has suggested that these empathogenic properties might augment psychological therapies such as those for post-traumatic stress disorder, and laboratory studies implicate serotonergic and oxytocinergic changes as mechanistic contributors to altered social cognition and affiliative behaviour. Experimental manipulations of serotonin and administration of oxytocin have been shown to change face perception, trust and economic cooperation in healthy volunteers, but no prior empirical study had objectively assessed social-cognitive mediators of MDMA's socially enhancing properties in a naturalistic recreational context. Stewart and colleagues set out to examine whether acute recreational ecstasy use is associated with changes in perceived trustworthiness of faces and in cooperative behaviour, and whether any such effects persist or reverse three days later. Using a naturalistic, independent-groups repeated-measures design, the investigators hypothesised that acute ecstasy use would increase face trustworthiness ratings and cooperative choices relative to non-using controls, and that sub-acute (day 3) effects would show a reversal consistent with post-use mood changes linked to serotonergic depletion.

A parallel-group, repeated-measures naturalistic study was conducted in which ecstasy users and non-using controls were assessed on the night of recreational drug use (day 0) and again three days later (day 3). A convenience sample of 39 healthy volunteers (majority university students) was recruited; the ecstasy group comprised 17 participants (12 male, five female) and the control group 22 participants (nine male, 13 female). All participants provided written informed consent and the study received University College London ethics approval. Participants were paid £10 after the final session. Testing took place in participants' or friends' homes on Friday or Saturday evenings. Ecstasy users self-administered their typical dose and testing began about 45 minutes later to approximate peak effects; controls were tested on parallel evenings without illicit drug use. Participants were asked not to use illicit drugs between day 0 and day 3, and alcohol use during testing was prohibited and recorded when present prior to the session. Assessments included a drug-use history questionnaire and trait measures on day 3: a shortened Empathy Quotient (EQ-S), the Hospital Anxiety and Depression Scale (HADS), and a spot-the-word task as an index of verbal intelligence. The primary social-cognitive measures were a trustworthiness face-rating task (66 neutral faces, 7-point scale) and a four-round cooperative behaviour battery modelled on economic games: a public-good (prisoner's dilemma-type) round, a dictator game, and an ultimatum game with allocator and decider roles. Subjective state was measured on 12 visual analogue scales (VAS) tapping effects typically associated with MDMA (0–10 scale) and a modified Beck Depression Inventory covering the preceding three days. Statistical analysis used repeated-measures ANOVA with group (ecstasy vs control) as the between-subjects factor and day (0 vs 3) as the within-subjects factor; post-hoc t-tests were Bonferroni-corrected. Pearson correlations used a significance threshold adjusted to p=0.01 to reduce Type I error. Analyses were performed in SPSS v20. The extracted text does not clearly report whether any additional covariate adjustments (beyond group and day) were applied in the main ANOVAs.

Sample characteristics: Ecstasy users and controls did not differ significantly in age, gender ratio, trait anxiety or depression. The ecstasy group scored higher on trait empathy than controls (t37 = -2.11, p=0.041) and lower on the spot-the-word verbal intelligence test (t37 = 2.20, p=0.034). Drug-use histories showed group differences in nicotine, cocaine and ecstasy use: more ecstasy participants had smoked cigarettes (p<0.001) and used cocaine (p=0.005) within the last two weeks, and more reported ever using ecstasy and using it within the last two weeks (both p≤0.002). On day 0, 12 ecstasy participants and seven controls reported alcohol consumption, with the ecstasy group consuming more units (3.38 ± 2.94) than controls (1.30 ± 2.73; U = 114.50, p=0.018). Trustworthiness ratings: A 2×2×2 ANOVA (group × day × face gender) found a significant interaction of drug and day (F1,37 = 5.79, p=0.021) and main effects of face gender (F1,37 = 127.63, p<0.001) and day (F1,37 = 13.83, p=0.001). Post-hoc tests showed that ecstasy users rated faces as more trustworthy on day 0 than on day 3 (t16 = 3.16, p<0.05); controls did not show this difference (p=0.17). Female faces were rated as more trustworthy overall, and ratings on day 3 were similar between groups (p=0.94). Co-operative behaviour tasks: In the public-good round no significant group or day effects were observed. In the dictator game there was a significant drug × day interaction (F1,37 = 16.51, p<0.001) and main effects of day (F1,37 = 10.60, p=0.002) and drug (F1,37 = 4.79, p=0.035). Post-hoc comparisons indicated ecstasy users contributed more than controls on day 0 (t37 = 3.64, p<0.01) and gave more on day 0 than on day 3 (t16 = 3.56, p<0.05); groups did not differ on day 3 (p=0.75) and controls showed no day-to-day change (p=0.30). For the ultimatum game, the team computed the difference between amount allocated and minimum amount accepted; a group × day interaction was significant (F1,37 = 6.65, p=0.014) with a trend for a main effect of day (F1,37 = 3.68, p=0.063). Ecstasy users showed a larger allocation-minus-acceptance difference on day 0 than on day 3 (t16 = 3.06, p<0.05) and a larger difference than controls on day 0 (t37 = 3.64, p<0.01); no between-group differences were present on day 3 (p=0.75). Subjective ratings and correlations: Acute subjective effects typical of ecstasy were reported on day 0 (increased euphoria, energy, jaw clenching and an overall ecstasy effect). A drug × day interaction for self-rated empathy (p=0.001) reflected higher empathy on day 0 than day 3 among ecstasy users; on day 3 ecstasy users rated themselves as less empathic than controls (p<0.01). Interactions were also found for distrust, antagonism, lack of compassion, sadness, desire to be with others, and hostility; several scales showed main effects of day and some (distrust, lack of compassion, anxiety) also showed main effects of drug. Ecstasy users reported greater desire to be with others on day 0 relative to controls but showed decreased compassion and increased distrust, antagonism, hostility and sadness on day 3 compared with day 0; controls showed no significant changes across days. BDI scores showed no significant effects. Within the ecstasy group, empathy on day 0 correlated positively with contribution to the public-good task (r = 0.76, p<0.001). A negative correlation was reported between BDI scores and the ultimatum allocation-minus-acceptance difference on day 3. No significant correlations were found between alcohol consumed on day 0 and task performance, nor between face trustworthiness ratings and self-reported trust on the VAS.

Stewart and colleagues report two principal findings: acutely, recreational ecstasy use was associated with increased perceived trustworthiness of neutral faces and increased generosity in individual-directed economic games (dictator and ultimatum), while no group differences were evident on these tasks three days later. The investigators interpret the day-0 increases as acute effects of the drug rather than stable group differences, noting that day-3 ratings converged between groups; they also acknowledge that pre-existing differences such as higher trait empathy in the ecstasy group could contribute partly but did not correlate with task performance. The authors consider serotonergic and oxytocinergic mechanisms as plausible mediators, citing prior evidence that MDMA elevates oxytocin and that oxytocin administration increases generosity and trust-related behaviour. They suggest the acute pattern—greater ratings of facial trustworthiness and more money allocated to an unknown individual—fits a model in which ecstasy increases social approach and interpersonal trust. The magnitude of increased generosity after ecstasy was compared with that found after intranasal oxytocin in previous work. Despite hypothesising a sub-acute reversal, the behavioural measures on day 3 did not show reduced trust or cooperation relative to controls; however, subjective mood measures did reveal negative changes on day 3 consistent with a come-down profile (reduced empathy/compassion and increased distrust, hostility and sadness). The authors propose these subjective changes may reflect transient serotonergic depletion and/or lifestyle factors associated with recreational use (sleep loss, late nights, appetite suppression), and they note that the pattern of subjective effects supports that participants ingested MDMA despite the study's naturalistic limitations. Key limitations acknowledged in the paper include the naturalistic design, which prevents control over dose and purity, and the absence of a neutral baseline session prior to day 0; the latter was avoided for ethical reasons. The sample comprised healthy recreational users without psychiatric diagnoses, so the findings cannot be straightforwardly generalised to clinical settings where MDMA is administered in controlled doses alongside psychotherapy. The authors highlight ecological validity as a strength and suggest their results help explain MDMA's recreational appeal and may illuminate psychological mechanisms—greater trust and cooperation—that could plausibly facilitate therapeutic alliance in psychotherapeutic contexts, while urging caution about generalisation.