Purpose: Neurocognitive; to investigate cognitive performance on selected neuropsychological tests sensitive to damage in the temporal or frontal (fronto-straital) regions (apparently tests of executive function and working memory).

Design: Retrospective (non-experimental) between-group design, with drug use (ecstasy user versus polydrug use-no ecstasy user) serving as a between-group variable, and with performance on each neuropsychological test analyzed separately. All participants completed a series of mostly computerized and mostly visual neuropsychological tests.

Subjects: 20 ecstasy users and 20 non-ecstasy users probably residing in or around London (England). All participants were recruited via advertisements in London-area magazines and via snowball sampling. Matched variables - Ecstasy user and polydrug-no ecstasy groups matched on gender, age and verbal intelligence, as assessed via National Adult Reading Test (NART).

Criteria for Inclusion - Ecstasy users - Criteria not stated, but apparently reporting use of ecstasy at least once in a lifetime, and perhaps as many as 10 times. Polydrug - no ecstasy - Reporting no use of ecstasy at any point in lifetime. Use of other licit and illicit drugs permitted. Both groups -Lack of major psychiatric or neurological illnesses as assessed via medical history, lack of alcohol dependence, and abstinence from psychoactive substances for at least 2 wks prior to study date, with abstinence verified via self-report only.

Drug Use Parameters - Ecstasy Use - Average lifetime ecstasy dose was 172 +/- 227.4 tablets (range = 10-1000 tablets), with no information provided on average dose per use. No information provided on frequency of use. Average duration of ecstasy use, in months, was 51.9 +/- 25.9 months (range = 3-122 months). No information provided on time since last use is reported, though abstinence from ecstasy for at least 14 days prior to study day was reported. Other drugs, Use of all other drugs reported as number of participants in each group (ecstasy users, polydrug-no ecstasy) and lifetime use for all but alcohol and nicotine). Alcohol was used by 14/20 ecstasy users on at least a weekly basis (8.7 +/- 9.5 units p/week), and 18/20 polydrug users drank alcohol at least weekly (10.7 +/- 15 units p/week). Nicotine was used by 16/20 ecstasy users (8.4 +/- 5.7 cigarettes p/day), and by 12/20 polydrug users (7+/- 8.5 cigarettes p/day). Cannabis was used by 20/20 ecstasy users 450 +/- 350.6 joints or times) and by 20/20 polydrug users (275.8 +/- 317.9 joints or times) . Amphetamines were used by 18/20 ecstasy users (101.4 +/- 231.5 times) and by 11/20 polydrug users (13.5 +/-21.1 times). Cocaine was used by 17/20 ecstasy users (30.5 +/- 39.8 times) and by 11/20 polydrug users (2 +/- 4.4 times). LSD was used by 10/20 ecstasy users (26.1 +/- 68.6 times) and by 5/20 polydrug users (0.5 +/-1.1 time). Psilocybian mushrooms had been used by 15/20 ecstasy users (4 +/- 6.7 times) and by 6/20 polydrug-users (0.4 +/-0.7 time). Benzodiazepines were used by 5/20 ecstasy users (2.6 +/- 7.0 times), but not by any polydrug users (i.e. 0/20). Opiates were used by 4/20 ecstasy users (5.4 +/- 22.3 times) and by 3/20 polydrug users (0.74 +/- 0.7 time). Solvents were used by 5/20 ecstasy users (4 +/- 22.3 times) and by 1 polydrug user (0.8 +/-3.4 times). Steroids were used by 1/20 ecstasy user (0.5 +/- 2.2 times) but by none of the polydrug users (0/20).

Group Demographics and Matched Variables - Authored matched groups on gender, age, and estimated verbal intelligence as assessed via NART. Gender - As M/F ratio - Ecstasy users = 10/10, polydrug = 8/12. Age - Average age of ecstasy users was 27.3 +/- 6.7 years (range = 18-40 years), and average age of polydrug users was 27.5 +/- 7.6 years (range = 19-48). Education - No information is provided on level of education attained by study participants in either group. Estimated Verbal Intelligence - Average NART score for ecstasy users was 100.3 +/- 6.3 (range = 90-115), and average NART score for polydrug users was 103 +/- 3.6 (range = 98-110).

Measures: Measures of Executive Function - FAS verbal fluency test, including "letter" and "semantic" conditions. Scored by number of words generated and in "letter" condition by "semantic" (category-related word) and "phonemic" scores (words sounding the same). One-Touch Tower of London (TOL) - (Participant indicates number of moves needed for one array to match another). IDED Attentional Shift - Wisconsin Card-Sort like task wherein participant must learn sorting cues and switch sorting rules in forced-choice discrimination). Measures of Working Memory - Spatial Working Memory - Participant searches for "hidden" item among boxes, and previously visited boxes and previously occupied boxes must be avoided, with 3 (practice only),4, 6 and 8-box trials (increasing difficulty). Scored for "between search" (returning to previously occupied box) and "within search" errors (returning to empty box) and search strategy. Pattern and Spatial Recognition - Participant required to select previously presented pattern from pair (previous/novel) (pattern recognition), and must select previously presented location of box in pair (previous/novel location) (spatial recognition). Correct response and response latency measured. Paired Associates learning - Participant required to indicate location of each of a number of patterns located in specific boxes, with trials using 2, 3, 6 and 8 patterns and locations. (Analyses only examined 6 and 8 pattern trials); number of presentations needed per trial, number of correct responses and number of errors recorded.

Measure of executive memory / impulsivity/working memory - Decision making task - Participant bets on color of box (red or blue) where counter is contained, with red/blue ratio varying in an array of 10 boxes. Deliberation time, number of points bet (from 1 to 100) and mean probability of choosing most likely outcome scored.

Measure of Impulsivity - Go/No Go Task - Participant presses space bar only for targets, while not responding to non-targets, with number of errors (missed targets, false alarms) measured.

Analyses: Measures of Executive Function - Scores on measures of executive function (FAS, IDED, and TOL analyzed via repeated analysis of variance (ANOVA) with drug use group (ecstasy user versus polydrug user) serving as a between-subjects factor and test scores serving as repeated measures. P value set at 0.05 for all analyses.

Measures of Working Memory - Scores on most measures of working memory (Paired Associates Learning, "within search" and "between search" scores of Spatial Working Memory) analyzed via repeated measures ANOVA, with drug use group (ecstasy user versus polydrug user) serving as a between-subjects factor and test scores serving as repeated measures. Pattern recognition, spatial recognition and the "search strategy" score for spatial working memory analyzed via independent sample t tests. P value set at p. = 0.05 for all analyses. Other Measures - Scores on Decision Making and Go/No Go Task were also analyzed via repeated measures ANOVA, with drug use group (ecstasy user versus polydrug user) serving as a between subject factor and test scores as repeated measures. Significance for all analyses set at 0.05.

Regressions - Number of words in "letter" condition regressed on "semantic strategy," "phonemic strategy," NART score, age and (presumably) ecstasy use.

Other Drug Use Parameters and Task Performance - Both "between search" and "within search" errors were examined using an analysis of covariance, with drug user (ecstasy use versus polydrug use) serving as a between-subjects factor and with use of amphetamine, cocaine and LSD serving as covariates.

Results - Significant Differences Found: Measures of Executive Function - Ecstasy users generated significantly fewer words in the "letter" condition of Word Fluency than did polydrug (no-ecstasy) users. On reversal trials in the IDED Attentional Shift task, ecstasy users made a greater number of errors for simple and compound reversals than polydrug users, but ecstasy users also made fewer errors on extradimensional reversals. Ecstasy users took longer to respond to difficult IDED trials. Participants in both groups performed better on easier (fewer moves) trials of the "one-touch" TOL than on more difficult trials.

Measures of Working Memory - Both "between search" and "within search" errors in Spatial Working Memory increased with increasing trial difficulty for members of both groups. Ecstasy users made more "between search" and "within search" errors than polydrug users, but only in the most difficult (8-box) trial. Ecstasy users made a greater number of errors than did polydrug users on the Pattern Recognition task (of Pattern and Spatial Recognition). On Paired Associates Learning, participants in both drug use groups needed a greater number of presentations and remembered fewer associations as trials increased in difficulty. (Ecstasy users were found to have made more errors on most difficult trial via post-hoc test).

Regressions - Regression analysis indicated that polydrug users made greater use of "semantic" word generation strategy than ecstasy users, but that word generation strategy was also significantly influenced by NART score and age.

Other Drug Use Parameters and Task Performance - Factoring out the effects of amphetamine, cocaine and LSD use did not alter the significantly greater number of "within search" and "between search" errors made by ecstasy users on the most difficult trial of Spatial Working Memory.

Results - No Significant Differences: Measures of Executive Function - There were no significant differences in word generation in the "semantic" condition of Word Fluency.

Ecstasy users and polydrug users made similar numbers of errors in the IDED task across all trials except on reversal trials. Latencies for IDED Attentional Shift task (time until response) were similar for ecstasy users and polydrug users on all but the most difficult trials. Ecstasy users and polydrug users did not perform significantly differently on the "one touch" TOL task. Measures of Working Memory - Ecstasy users and polydrug users made similar numbers of "between search" and "within search" errors on 4 and 6-box (less difficult and moderately difficult) trails on the Spatial Working Memory task. There were no significant differences in "search strategies" used by ecstasy users and polydrug users in the Spatial Working Memory task. Ecstasy users and polydrug users made similar numbers of errors on the Spatial Recognition task. Number of presentations needed and number of associates recalled on the Paired Associates Learning task were similar for ecstasy users and polydrug users.

Measures of Impulsivity - Ecstasy users and polydrug users did not perform significantly differently on the Go/No Go task.

Measures of Executive Function/Impulsivity/Working Memory - Both ecstasy users and polydrug users gambled with a lower number of points under "higher risk" conditions, and both gambled with a higher number of points on lower-risk conditions. Ecstasy users and polydrug users were equally likely to choose the most likely outcome for each probability.

Overall Effects: When compared with a sample of polydrug users matched on gender, age and use of some other drugs, a sample of 20 ecstasy using polydrug users scored significantly lower on one measure of executive function (word fluency) and on several measures of working memory (pattern recognition, spatial working memory). However, there was no single task in which ecstasy users performed less well than polydrug users. Word fluency scores of ecstasy users were lower in the "letter" condition, but not the "semantic" condition, and while ecstasy users did have lower pattern recognition scores than polydrug users, there were no significant differences in spatial recognition scores. Ecstasy users generally made more errors than polydrug users on a Wisconsin Card Sort-like task (IDED), but in some cases they made fewer errors than controls, and IDED performance differences only sometimes reached significance (simple and compound reversal trials). Ecstasy users made more "within search" and "between search" errors on an assessment of spatial working memory, but the difference was only apparent in the most difficult trial. While amphetamine, cocaine and LSD use (in volume) was greater in ecstasy users, factoring in possible effects from use of other drugs did not reduce or eliminate the differences in performance on the spatial working memory task. Both groups performed similarly on a measure of impulsivity or response inhibition (Go/No Go) and a measure of "decision making."

Comments: This is the third publication comparing polydrug users with and without ecstasy use on neurocognitive task performance authored by Fox and colleagues. There appears to be little, if any, overlap in samples appearing in each paper, with ecstasy users in this paper reporting somewhat lower overall (lifetime) ecstasy use than previous samples examined by Fox et al (Fox et al. 2001a; Fox et al. 2001b). It is notable that with the exception of the word fluency task, all other tasks employed in this study have a large visual component and many are largely visual in nature (spatial working memory, pattern and spatial recognition, the IDED task). Previous studies have tended to employ largely verbal tasks. It is possible that visual working memory relies on slightly different functions or processes than executive function, whereas verbal recall and executive function for verbal tasks may rely similar or related processes. Deficits in working memory and executive function reported in this study appear to be subtle and selective, and are often evident only during the most difficult trial of a task or in one condition, and not in easier trials. While the authors associate differences in task performance with reduced function in temporal versus frontal cortex, they do not explicitly identify each test with the corresponding brain area purportedly under assessment, and so it is difficult to assess claims concerning locality of ecstasy-induced effects. Like other retrospective studies comparing polydrug using samples, generalizability of study findings may be limited by sample size, retrospective study design and presence of polydrug use in both groups, with use of other drugs greater in ecstasy users than in polydrug users. In addition, drug history and abstinence from psychoactive substances is assessed via self-report only.