Acute dose of MDMA (75 mg) impairs spatial memory for location but leaves contextual processing of visuospatial information unaffected

Kuypers and colleagues situate the study within a literature that links MDMA (Ecstasy) use to cognitive impairments, particularly in verbal memory, while noting mixed evidence for spatial memory deficits. Previous studies employed a variety of spatial tasks (spatial span, delayed matching-to-sample/recognition, and search-token paradigms), and the authors argue that task characteristics and polydrug use may have confounded between‑group findings. They suggest that tasks requiring only storage of location might be less vulnerable to MDMA than those requiring storage plus contextual processing, but existing categorisations of prior studies have not resolved the discrepant results. This paper reports a double-blind, placebo-controlled, three-way crossover human study that tests whether a single acute oral dose of MDMA (75 mg) affects spatial memory during intoxication and during a withdrawal phase about 26 hours later, and whether any effect depends on task complexity. Two spatial tasks were used: a simple spatial memory localisation task and a change blindness task that requires processing contextual visuospatial information. Methylphenidate 20 mg was included as an active comparator to probe whether dopaminergic mechanisms could explain any observed effects.

The investigators used a double-blind, placebo-controlled, three-way crossover design with treatments balanced across sessions and a minimum 14-day washout between periods. Testing occurred over two consecutive days in each treatment period: on day 1 participants received one of three treatments (MDMA 75 mg dissolved in bitter orange syrup and orange juice, methylphenidate 20 mg as a pill, or placebo), and on day 2 no drug was administered. Cognitive testing was performed during the intoxication phase (about 1.5 h post-dose) and during the withdrawal phase (about 25.5–26 h post-dose). Blood samples for pharmacokinetics were taken 1.5 h and 25.5 h after dosing. Eighteen recreational MDMA users (9 male, 9 female; age range 20–39, mean 26.22, SD 5.1) completed all three treatment periods. Twenty-eight volunteers began screening; three were excluded on medical grounds and seven dropped out (private reasons or failure to abstain between sessions). Lifetime MDMA use was light (<30 occasions) in 15 subjects and heavier (60–120 occasions) in three. Inclusion and exclusion criteria were medical screening, absence of major psychiatric/neurological conditions, and restrictions on recent drug, alcohol and caffeine use; urine and breath tests were used to verify abstinence on test days. Participants practised the tasks on a training day. The primary cognitive measures were two spatial tasks. The spatial memory localisation task comprised 75 trials in which a briefly presented target (500 ms) appeared at a random screen location; subjects relocated the cursor to the remembered position with response delays of 0, 2 or 4 s (dependent variables: localization error and reaction time). The change blindness task presented 100 traffic photographs with intermittent blank screens; 80% contained a change (half relevant to traffic, half irrelevant), and changes occurred centrally or peripherally relative to a fixation dot; dependent variables were number of correct detections and reaction time. Pharmacokinetic samples were analysed by LC-MS/MS; limit of quantification was 1 ng/ml for MDMA, MDA and ritalinic acid. Statistical analyses used SPSS 11.5. Repeated-measures general linear models (GLM) tested treatment (three levels) and day (two levels) as within-subject factors, with additional within-subject factors for response delay (three levels) in the spatial task and context/location of change (two levels each) in the change blindness task. Significant omnibus effects were followed by drug-versus-placebo contrasts. An additional multiple linear regression was run on individual trial data from the spatial task to test whether reaction time, drug, response delay or trial number independently correlated with localization error. The significance threshold was p=0.05.

Eighteen participants completed the protocol; no serious adverse reactions were reported. Plasma concentrations measured 1.5 h after dosing averaged 113.4 ng/ml (SD 37.4) for MDMA and 2.9 ng/ml (0.9) for MDA; at 25.5 h these values averaged 11.6 ng/ml (9.3) and 2.1 ng/ml (1.2), respectively. Ritalinic acid concentrations averaged 95.9 ng/ml (78.4) at 1.5 h and 22.9 ng/ml (7) at 25.5 h. For the spatial memory localisation task, a repeated-measures ANOVA showed a significant treatment-by-day interaction on localization error (F 2,34 = 3.46, p = 0.043) and a main effect of response delay (F 2,34 = 83.56, p < 0.001), reflecting greater localization error with longer delays. Drug-versus-placebo contrasts on each day indicated that MDMA significantly increased localization error on day 1 (intoxication) compared with placebo (F 1,17 = 10.86, p = 0.004), but no drug-related differences were apparent on day 2 (withdrawal). Reaction time in the spatial task showed main effects for drug-by-day (F 2,34 = 6.44, p = 0.004), day (F 1,17 = .46, p = 0.050 as reported), and response delay (F 2,34 = 3.42, p = 0.044). Contrasts showed that MDMA significantly reduced reaction time on day 1 relative to placebo (F 1,17 = 7.92, p = 0.012); again, no differences were observed on day 2. A trial-level multiple linear regression (total trials determined by subjects × treatments × trials per repetition) tested whether reaction time, drug (MDMA vs placebo), response delay or trial number predicted localization error. Results indicated that drug (MDMA) and response delay were positively correlated with localization error; together these two factors explained approximately 8% of the variance (R = 0.289, R2 = 0.083; F 3,2699 = 102.846, p < 0.001). Reaction time was not correlated with localization error, indicating that the increased inaccuracy under MDMA was not attributable to faster responding. Performance on the change blindness task was not affected by treatment. As expected from task manipulations, changes relevant to traffic were detected more frequently and faster than irrelevant changes, and centrally located changes were detected more often and sooner than peripheral ones; however, there were no significant interactions between treatment and content or location of change. Methylphenidate did not influence performance on either spatial task.

Kuypers and colleagues interpret the findings as evidence that a single acute dose of MDMA (75 mg) impairs short-term spatial memory for location during intoxication but does not affect contextual processing of visuospatial information in a more complex change-detection task. The impairment manifested as larger localization errors accompanied by faster reaction times on the spatial localisation task during intoxication; regression analysis indicated that faster responding did not account for the inaccuracy, and that MDMA and response delay together explained about 8% of the variance in localization error. Effects were transient and no drug-related impairments were observed during the withdrawal assessment about 26 hours after dosing. The absence of treatment effects on the change blindness paradigm led the authors to conclude that processing across multiple spatial locations or of contextual information remained intact under MDMA. They note, however, that this was the first drug trial to use a change blindness task and concede the alternative possibility that the task might be insensitive to drug effects; they argue against this by pointing out that the task did pick up the within‑task manipulations (relevance and location of changes), indicating sensitivity to experimental manipulations. Regarding mechanisms, the lack of effects from methylphenidate suggests that dopaminergic action is unlikely to account for the MDMA-related spatial deficit. The authors favour a serotonergic explanation: they relate their findings to prior human work showing increased localization error after a 5‑HT agent (fenfluramine) and to animal studies linking MDMA or serotonergic manipulation to hippocampal lesions and spatial memory impairment. These converging lines of evidence are presented as supporting a role for serotonin in the observed acute spatial memory deficit. The investigators acknowledge limitations that constrain generalisability. Participants were predominantly light recreational MDMA users and performed near ceiling on the spatial tasks under placebo and during withdrawal, which may reduce sensitivity to drug-related improvements from methylphenidate and limit inferences about long-term deficits in heavier users. They therefore caution that the present acute effects should not be taken to imply that persisting deficits cannot occur during abstinence in heavier or long-term users. The authors also note procedural factors such as extensive task familiarisation and the specific delay intervals employed as relevant to comparisons with prior studies that reported facilitation with dopaminergic agents.