Is Ecstasy an “Empathogen”? Effects of ±3,4-Methylenedioxymethamphetamine on Prosocial Feelings and Identification of Emotional States in Others

MDMA (±3,4-methylenedioxymethamphetamine, 'ecstasy') is commonly reported to produce so-called "empathogenic" effects, including increased empathy and prosocial feelings. Such subjective effects are cited both as reasons for recreational use and as a rationale for using MDMA as a psychotherapeutic adjunct. Although several controlled laboratory studies have demonstrated MDMA-induced increases in self-reported sociability-related states (for example friendliness, closeness, and talkativeness), there has been little direct evidence on whether MDMA alters overt social-cognitive behaviour in humans, such as the ability to identify emotions in others. Bedi and colleagues designed a within-subject, double-blind study to characterise MDMA's effects on two domains: socially relevant mood states and social cognition operationalised as emotion-recognition performance from faces, the eye region, and vocal cues. They tested two doses of MDMA (0.75 mg/kg and 1.5 mg/kg), included methamphetamine (20 mg) as a stimulant comparator, and hypothesised that MDMA would alter both self-reported sociability/empathy and accuracy on emotion-identification tasks. The authors noted that changes could reflect either improved sensitivity to others' emotions (consistent with enhanced empathy) or reduced sensitivity to threat-related signals (potentially increasing social approach behaviour).

This was a four-session, within-participant, double-blind, randomised study. Twenty-one healthy volunteers who reported at least two prior occasions of MDMA/ecstasy use were recruited after screening for psychiatric, medical and pregnancy-related exclusions. Sessions were spaced at least five days apart and participants were required to abstain from other recreational drugs and alcohol for specified intervals prior to each session; compliance was verified with urine, saliva and breathalyser screens. Female participants provided a pregnancy test on each session. At each session participants received one of four oral capsules in randomized order: MDMA 0.75 mg/kg, MDMA 1.5 mg/kg, methamphetamine (METH) 20 mg, or placebo. Drug administration occurred at noon; baseline cardiovascular and subjective measures were recorded, and cognitive tasks were initiated 65 minutes after capsule ingestion to coincide with anticipated peak drug effects. Participants remained under observation in a laboratory room with minimal social interaction for at least 4.5 hours following dosing. Subjective measures comprised Visual Analogue Scales (VAS) for adjectives including stimulated, loving, sociable, playful and lonely, administered at multiple timepoints up to 240 minutes post-dose, and the Profile of Mood States (POMS), administered at baseline, 90 and 240 minutes, with a particular focus on the Friendliness subscale. Cognitive assessments included three emotion-recognition tasks completed once per session during peak effects: a Facial Emotional Recognition (FER) task using Ekman faces morphed from neutral to prototype in 10% increments (four basic emotions: anger, fear, happiness, sadness; 500 ms presentation per face; primary outcome accuracy), the Reading the Mind in the Eyes (Eyes) test (36 eye-region images of complex emotional states; accuracy), and the DANVA-2 Adult Paralanguage vocal affect task (24 audio clips conveying happy, sad, angry or fearful tone at two intensity levels; accuracy). Statistical analyses used repeated-measures ANOVA for primary subjective and cognitive outcomes, with Bonferroni corrections for multiple comparisons. For FER and DANVA-2, two-way repeated-measures ANOVAs assessed drug-by-emotion interactions; significant interactions led to simple main effects testing with adjusted alpha levels. When normality assumptions were violated for some measures (VAS Loving, VAS Lonely, POMS Friendliness, DANVA scores), non-parametric tests (Friedman's ANOVA and Wilcoxon signed-rank tests) were applied; VAS Lonely is reported using non-parametric statistics. Mauchley's test of sphericity and Greenhouse-Geisser or Huynh-Feldt corrections were applied as appropriate. Session order was tested and found non-significant, and analyses were conducted in SPSS 17.0.

Twenty-one participants (mean age 24.4 years, SD = 4.9; 9 female) completed the study. Most were White (17), with the remainder Asian (2), African-American (1) or mixed race (1). Mean age at first ecstasy use was 19.8 years (SD = 2.7) and lifetime ecstasy use averaged 15.0 occasions (SD = 23.1); 13 participants had used ecstasy fewer than 10 times. Use of other substances in the prior month was common (alcohol for all, cigarette smoking in 12 participants, cannabis in 16), and most reported lifetime use of stimulants and hallucinogens. Subjective effects: Drug condition affected all five pre-specified subjective measures. MDMA 1.5 mg/kg significantly increased self-reported loving feelings on the VAS relative to placebo (F(3,57) = 5.04, p = 0.004, η2 = 0.21). POMS Friendliness was increased by MDMA 1.5 mg/kg versus both placebo and MDMA 0.75 mg/kg (F(2.7,51) = 6.81, p = 0.001, η2 = 0.26). MDMA 0.75 mg/kg unexpectedly increased reported loneliness (χ2(3) = 11.94, p = 0.008, r = 0.54; non-parametric). Both MDMA 1.5 mg/kg and methamphetamine increased VAS Playfulness (F(3,57) = 8.57, p < 0.001, η2 = 0.31), whereas only methamphetamine significantly increased VAS Sociability relative to placebo (F(2.5,47.7) = 3.06, p < 0.05, η2 = 0.14). Emotion-recognition performance: There was a significant interaction between drug condition and emotion type on FER accuracy (F(5.2,104.6) = 3.23, p = 0.008). Follow-up analyses revealed a specific effect on fear recognition: MDMA 1.5 mg/kg reduced accuracy for fearful facial expressions relative to placebo (simple main effect for fear: F(2.8,56.0) = 4.45, p = 0.008, η2 = 0.18). After correction for multiple comparisons, no significant drug effects were observed for identification of angry, happy, neutral, or sad faces. Neither MDMA dose nor methamphetamine altered accuracy on the Eyes test or the DANVA-2 vocal task; there was no drug-by-emotion interaction on DANVA-2 scores. Exploratory analyses indicated that MDMA 1.5 mg/kg increased the tendency to misclassify non-neutral facial expressions as neutral compared with placebo (p < 0.05, uncorrected). Re-analyses based on participants' guesses about which drug they had received found no significant relationship between expectancy and the measures sensitive to drug effects.

Bedi and colleagues interpret their findings as demonstrating that MDMA at 1.5 mg/kg produces measurable changes in social cognition by reducing the accuracy of recognising fearful facial expressions, while concurrently increasing subjective feelings commonly labelled "empathogenic" (loving feelings and friendliness). The authors argue that this dissociation—enhanced prosocial affect alongside reduced sensitivity to threat-related facial cues—may reflect an increase in social approach behaviour rather than an enhancement of empathic perception per se. Reduced detection of others' negative or threat-related emotions could plausibly lower social avoidance and thereby facilitate interpersonal approach. The study also found that some subjective prosocial effects were not unique to MDMA: methamphetamine increased playfulness and, uniquely among the tested drugs, increased self-rated sociability. The authors acknowledge that expectancies could influence subjective reports but report that participants' guesses about drug identity were not associated with the measured outcomes. They therefore suggest that some prosocial properties may be shared across psychostimulants, and that methamphetamine's prosocial potential may be underappreciated. Several limitations are acknowledged. The behavioural tasks may not capture all socially relevant effects of MDMA: static facial photos may lack ecological validity compared with dynamic stimuli, tasks did not include matched non-social stimuli to test specificity, MDMA doses were weight-adjusted while methamphetamine was not (although participants were within a relatively narrow BMI range), and all participants had prior experience with MDMA so results may not generalise to MDMA-naive individuals. The authors propose further research directions, including studies probing the role of oxytocin—given evidence that MDMA elevates plasma oxytocin and that oxytocin mediates prosocial behaviour in animals—and investigations into the cognitive mechanisms underlying increased misclassification of subtle emotional cues as neutral. They also note potential individual differences, such as baseline trait sociability, that might moderate MDMA's effects. Finally, the investigators discuss implications for recreational and therapeutic contexts. Recreationally, they warn that MDMA's combination of increased interpersonal warmth and reduced sensitivity to negative social signals could increase social risk-taking while intoxicated. Clinically, reduced perception of fear or negative social cues might assist engagement with traumatic material in MDMA-assisted psychotherapy, but the authors stress that mechanistic work is needed to inform treatment design and to optimise any therapeutic application.