Oxytocin receptor gene variations and socio-emotional effects of MDMA: A pooled analysis of controlled studies in healthy subjects

MDMA (3,4-methylenedioxymethamphetamine) produces empathogenic and prosocial subjective effects and increases peripheral oxytocin. Previous human and animal studies have implicated oxytocin in social cognition and prosocial behaviour, and animal work shows that blocking oxytocin receptors can prevent MDMA-induced prosocial effects. However, human evidence for oxytocin as a mediator is mixed: some studies report correlations between MDMA-induced oxytocin release and prosocial feelings, while others do not, and direct pharmacological blockade in humans is limited by blood–brain barrier issues. Genetic variability in the oxytocin receptor gene (OXTR) has been associated with social traits in prior work, suggesting that single-nucleotide polymorphisms (SNPs) in OXTR might moderate individual responses to MDMA. Vizeli and colleagues set out to test whether three common OXTR SNPs (rs53576, rs1042778, rs2254298) modulate the socio-emotional, empathic, and prosocial effects of a single 125 mg oral MDMA dose in healthy volunteers. The study used pooled data from eight similar double-blind, placebo-controlled, crossover Phase I studies to explore genotype-by-drug interactions on subjective ratings, emotion recognition, empathy tasks, economic prosociality measures, and plasma oxytocin and MDMA concentrations. The investigation was framed as exploratory and sought both to characterise potential moderating effects and to attempt replication of previously reported rs53576 findings.

This pooled analysis combined eight Phase I, double-blind, placebo-controlled, crossover studies conducted between 2009 and 2014, using a common MDMA-alone versus placebo comparison. In total 136 healthy subjects of European descent were recruited; genotyping and complete data were available for 132 subjects (64 men, 68 women), aged 18–44 years (mean ± SD = 24.8 ± 4 years). Most studies used two MDMA sessions per subject (125 mg each), but only MDMA-alone and placebo sessions were analysed here. Washout between doses was ≥7 days. Exclusion criteria included psychiatric or significant physical illness and substantial illicit drug history; urine screens excluded recent illicit drug use. MDMA was administered orally as a single 125 mg dose (mean dose 1.9 ± 0.3 mg/kg). Genotyping of OXTR SNPs rs53576, rs1042778, and rs2254298 was performed from whole-blood DNA using TaqMan assays; linkage disequilibrium analyses found no LD among these SNPs. Subjective effects were assessed repeatedly up to 6 h post-dose by Visual Analog Scales (VAS), including global items (“any drug effect”, “closeness to others”) and social items (“trust”, “want to be hugged”, etc.); some VAS social items were assessed in a subset of 53 subjects. Emotion recognition was measured at 90 min using the Facial Emotion Recognition Task (FERT) in 69 subjects. Empathy was assessed with the Multifaceted Empathy Test (MET) 90–180 min post-dose (n=69). Prosocial behaviour was measured with the Social Value Orientation (SVO) slider task 3–4 h post-dose (primary SVO data n=69; inequality-aversion index n=33). Plasma oxytocin was sampled at baseline and 2 h post-dose, and MDMA concentrations were measured across multiple time points up to 6 h. Subjective VAS responses were summarised as area under the effect–time curve from 0 to 6 h (AUEC6). Repeated-measures ANOVAs tested drug effects and one-way ANOVAs tested genotype group effects on MDMA-induced changes (MDMA minus placebo). Analyses incorporated MDMA plasma AUC6 and, where available, oxytocin change at 2 h as covariates; sex was included in additional models to check confounding. Multiple comparisons were addressed using the Nyholt correction, yielding an effective number of independent variables (Veff) of 28.96 and a corrected significance threshold of p < 0.0017. For rs2254298, AA and AG genotypes were pooled due to the rarity of AA.

Genotyping and sample characteristics matched expected Caucasian allele frequencies (rare allele frequencies: rs53576 A = 35%, rs1042778 T = 38%, rs2254298 A = 11%) and no linkage disequilibrium was detected among the three SNPs. MDMA produced robust subjective, behavioural, and neuroendocrine effects versus placebo across the pooled sample. Subjective effects: MDMA significantly increased AUEC6 on the VASs for "any drug effect" (F1,131 = 544, p < 0.001) and "closeness" (F1,131 = 57, p < 0.001). In the subset with social VASs (n = 53), MDMA increased "trust" (F1,52 = 33, p < 0.001), "want to be hugged" (F1,52 = 6.6, p < 0.05), "want to hug" (F1,52 = 7.6, p < 0.01), and "want to be with others" (F1,52 = 20, p < 0.001), and decreased "want to be alone" (F1,52 = 21, p < 0.001). Regarding OXTR moderation, rs1042778 TT homozygotes showed larger MDMA-induced increases in "trust" (F1,49 = 14, p < 0.001), "want to be hugged" (F1,49 = 5.3, p < 0.05), and "want to be with others" (F1,49 = 6.5, p < 0.05), and a larger reduction in "want to be alone" (F1,49 = 4.5, p < 0.05) compared with G-allele carriers. After Nyholt correction for multiple testing, only the rs1042778 effect on "trust" remained statistically significant. The rs53576 and rs2254298 SNPs did not significantly moderate VAS outcomes in the analysed samples. The rs1042778 effect on trust was observed in a small TT subgroup (reported as n = 5 in one place) versus larger GG/GT groups. Emotion recognition and empathy: On the FERT (n = 69), MDMA impaired recognition accuracy for fearful (F1,68 = 47, p < 0.001), sad (F1,68 = 14, p < 0.001), and angry (F1,68 = 16, p < 0.001) faces versus placebo. None of the three OXTR SNPs moderated these MDMA effects. On the MET (n = 69), MDMA increased explicit emotional empathy for positive stimuli (F1,68 = 7.6, p < 0.01); OXTR variants did not alter MET outcomes. Prosociality (SVO): MDMA produced a trend toward an increased SVO angle (greater prosociality) compared with placebo (F1,68 = 3.1, p = 0.08). MDMA reduced the inequality-aversion index in subjects with a prosocial orientation (F1,32 = 9.3, p < 0.01), indicating a shift toward greater inequality aversion. Some genotype interactions were reported: rs53576 AA subjects showed a larger MDMA-induced reduction in inequality-aversion than G-allele carriers (F1,31 = 9.4, p < 0.01), and rs1042778 GG subjects showed a larger reduction than T-allele carriers (F1,29 = 5.6, p < 0.05). However, these SVO-related genotype effects did not survive correction for multiple comparisons, and the inequality-aversion analyses were limited to a reduced subsample (n = 33). Plasma concentrations: MDMA significantly increased plasma oxytocin at 2 h versus placebo (placebo: 19 ± 39 pg/mL; MDMA: 74 ± 70 pg/mL; F1,99 = 62, p < 0.001). MDMA peak concentration and AUC6 across 132 subjects were 224 ± 48 ng/mL and 950 ± 207 ng×h/mL, respectively. Oxytocin and MDMA plasma concentrations did not differ between OXTR genotype groups. The authors also note evidence of acute tolerance to some subjective MDMA responses over time.

Vizeli and colleagues interpret their principal finding as preliminary evidence that OXTR genetic variation, specifically rs1042778, can modulate some subjective prosocial effects of MDMA, most notably increases in self-reported trust. They note this provides indirect support for a role of oxytocin in MDMA's subjective effects, consistent with animal studies where oxytocin mediates MDMA-induced prosocial behaviour. At the same time, the investigators failed to replicate an earlier report that rs53576 moderates MDMA-induced sociability, and they discuss methodological differences that may explain discrepancies: prior work combined several VAS items into a sociability component and used different MDMA dosing regimens (dose-dependent effects were previously observed). The authors emphasise that several of the genotype-related findings (including those on the SVO inequality-aversion index) did not survive correction for multiple comparisons and were based on small subsamples, limiting confidence. They also highlight that MDMA acts on multiple systems—serotonin, norepinephrine, dopamine, vasopressin, and corticosteroids—and that interactions among these mediators could influence social outcomes. Additional limitations acknowledged include the exploratory nature of the analysis, incomplete ascertainment of all outcome measures across the full cohort (reducing sample sizes for some tests), assessment of only three OXTR SNPs rather than broader haplotypes, and potential cultural or early-life environmental moderators of OXTR effects. The authors therefore urge caution in interpretation and call for replication in larger, well-powered studies. In terms of implications, the investigators suggest that if confirmed, OXTR genotyping might be relevant for understanding individual differences in response to MDMA and could have relevance for MDMA-assisted psychotherapy; they recommend that therapeutic MDMA trials consider genotyping OXTR variants and genes involved in MDMA metabolism.

The study concludes that the OXTR rs1042778 SNP—but not rs53576 or rs2254298—was associated with altered MDMA-induced feelings of trust, while rs53576 effects reported previously could not be replicated. After correction for multiple comparisons, OXTR SNPs did not moderate the overall subjective MDMA effect or feelings of "closeness to others" in the full sample. The authors emphasise that the results are preliminary and should be interpreted cautiously because of multiple comparisons and small genotype subgroup sizes.