Purpose: Cognitive function, neuropsychological; to investigate the relationship between self-reported problems attributed to ecstasy use and performance on tests of cognitive function. Specific hypotheses - Individuals who report experiencing ecstasy-related problems would be more sensitive to ecstasy-associated impairment on tests of cognitive function (H1) and there might also be an association between lifetime ecstasy dosage and impaired cognitive function.

Design: Retrospective (non-experimental) between-subjects design, with self-reported ecstasy related problems serving as one factor (present versus absent) and overall lifetime exposure serving as another factor (no exposure, "low" lifetime dose, "medium" lifetime dose and "high" lifetime dose, with participants categorized after study entry and not selected for use characteristics except in the case of non-user controls). All participants underwent tests of cognitive function and assessment of everyday life events.

Subjects: 20 non-users, 20 "problem" users and 20 "no problem" users residing in England (most probably London area). Participants recruited through magazine advertisements or through "snowball" technique. Groups matched for estimated verbal IQ (via NART) and approximately matched for age and gender. Criteria for Inclusion - "Problem" Ecstasy users - Marking statement indicating ecstasy use and self-reported experience of problems attributed to ecstasy use. "No Problem" Ecstasy users - Marking statement indicating ecstasy use and reporting absence of problems attributed to ecstasy use. Both "problem" and "no problem" users indicated having used ecstasy at least once in their lives. Non-users - Reported never having used ecstasy, though use of other substances permitted. All groups - Abstinence from use of all drugs for 2 wks prior to study day, with abstinence verified via self-report only.

Drug Use Parameters - Ecstasy Use - "Problem" ecstasy users reported having taken 372.3 +/- 663.3 tablets over a lifetime (range not reported), and "no problem" users reported using 356.9 +/- 339.8 tablets over a lifetime. No information is provided concerning frequency of use, in tablets per month. For "problem" users, time since last use, in days, was 234 +/- 345 days, range not provided, and for "no problem" users, time since last use was reported at 75 +/- 162 days, range not provided. Average duration of use was for "problem" users was 62.6 +/- 33.8 months, and duration of use, in months, was 65.2 +/- 28.1 months. Average dose per use for "problem" ecstasy users was 1.9 +/- 0.7 tablets, and for "no problem" users, 2.9 +/- 3.2 tablets. Largest number of tablets ever taken per use, for "problem" users was 4.7 +/- 2.6 tablets, and for "no problem" users, 7.8 +/- 7.3 tablets per use. Other Drug Use - Presented in percentage of participants using (x/20). Cannabis used by 100% of all three groups (non-users, "problem" ecstasy users and "no problem" ecstasy users). Amphetamines used by 55% non-users, 100% "problem" users and 95% "no problem" users. Cocaine, used by 25% non-users, 90% of "problem" users and 89% of "no problem" users. LSD, used by 20% of non-users, 95% of "problem" users and 89% of "no problem" users. Barbiturates, used by 15% of non-users, 35% of "problem" users, and 26% "no problem" users. Opiates, used by 10% non-users, 50% of "problem" users and 42% "no problem" users. Psilocybian mushrooms, used by 20% non-users, 85% "problem" users, and 63% "no problem" users. Solvents, used by none of the non-users, 40% of "problem" users and 16% of "no problem" users. Nicotine, used by 60% non-users, 65% "problem" users and 79% "no problem" users. Alcohol, used by 95% non-users, 80% "problem" users and 74% "no problem" users. Drug use parameters for three dosage categories (in brief) Lifetime dose, in tablets, for low use = 38.31 +/- 21.1, for medium = 236.4 +/- 116.5 tablets, and for high 918.2 +/- 730.5 tablets. No information on frequency of use. Duration of use, in months, for low = 46.4 +/- 24.1 mo, for medium = 80.9 +/- 32.6 mo, for high = 64.5 +/- 25.1. Number of tablets taken per use, for low = 1.8 +/- 0.9, for medium = 2.2 +/- 1.1, and for high = 3.7 +/- 3.8. Time since last use, in days, for low = 168 +/- 364 days, for medium = 159 +/- 219 days, for high = 84 +/- 177 days. Maximum dose per use, low = 3.6 +/- 1.5 tablets, for medium = 5.1 +/- 2 tablets, and for high = 10.9 +/- 8.7 tablets. Analysis found significantly duration in use for medium dose users compared to low and high dose users, and that high dose users had taken significantly larger maximum doses than low or medium dose users. Other drug use by dosage category (summary) - All (11/11) high does users had used cannabis, cocaine, amphetamine, and LSD at least once, 36% had used barbiturates, 64% had used opiates, 91% had used psilocybin / mushrooms, 18% had used solvents, 82% had used nicotine, 73% alcohol. All (14/14) medium dose users had used cannabis and amphetamines, and 93% had used cocaine and LSD, 57% had used barbiturates, 64% had used opiates, 86% had used psilocybin mushrooms, 50% had used solvents, 64% had used nicotine and 79 had used alcohol. All (14/14) low dose users had used cannabis, 93% had used amphetamines, 71% had used cocaine, 86% had used LSD, none had used barbiturates, 14% had used opiates, 50% had used mushrooms, 14% had used solvents, 71% nicotine and 79% had used alcohol. Generally, high dose users have tried a greater number of drugs than medium or low drug users, and medium dose users have tried more drugs than low dose users.

Group Demographics and Matched Variables - Authors matched non-users, "problem" users and "no problem" users on estimated verbal IQ, and approximately matched these groups on age and gender (no significant differences found).

Gender, as M/F Ratio - Non-users. 6/14, "problem" users, 12/8, "no problem" users, 9/11. Age - Average age of non-users was 23.3 +/- 6.5, for "problem" users," 27.4 +/- 4.5 years, and for "no problem" users, 26.2 +/- 5 years. Education - Reported as years of education after age 11 and estimated for American system: For non-users, average years of education obtained was 9.4 +/- 2.0 (estimated 16.4 years overall), for "problem" users, 7.7 +/- 2.8 (estimated 14.7 overall), and for "no problem" users, 8 +/- 2.7 (15 years overall). Estimated Verbal IQ - (Estimated from National Adult Reading Test, NART). Estimated verbal IQ was 108.7 +/- 6.2, for "problem" users, 107.9 +/- 4.6, and for "no problem" users, 111.6 +/- 6.3 Demographics and dosage groups - Gender, as M/F ratio, Non-users = 6/14, low = 5/9, medium = 9/5, high = 6/5. Non-users spent more time in education than high dose users (9.4 +/- 2 (approx. 16 years) vs. 6.6 +/- 3.3 (approx. 13-14 years); low = 8.3 +/- 2.3 (approx. 15 years), medium = 7.9 +/- 2.6 (approximately 14-15 years). Ages ascend from control to high use (23.3 (NU), 25.7 (L), 26.9 (M), 28.0 (H). Estimated verbal IQ amongst dosage groups better matched than for "problem," "no problem" groups (L = 109.7 +/- 5.5, M = 109.9 +/- 4.5, H = 109.3 +/- 7.9

Measures: Everyday Life Events/Psychological function - Measured via Uplifts, Hassles, Stresses and Cognitive Failures questionnaire (UHSC), a self-report measure of positive and negative life events and everyday cognitive failures. Reaction time - Assessed by visual RT task, with response latency recorded for targets presented on computer monitor. Memory, Verbal - Immediate and delayed verbal memory were assessed via Immediate and Delayed Prose Recall, a task similar to RBMT-Story Recall and WMS-Logical Memory task, with participants recalling passage both immediately and upon surprise test of recall 1.5 h later. Also measured by list recall task, with 24 item lists organized into categories, scored as recall - intrusions. Spatial working memory - Assessed through computerized task designed at University of East London; participants compare two consecutively presented targets (houses with lit and unlit windows), indicating two lit windows in common. Executive Function - Measured by computerized version of Wisconsin Card Sort (WCST) and by manual Tower of London (TOL) task.

Analyses: Demographics and Drug Use - Because data did not satisfy parametric requirements, Kruskal-Wallace comparisons were employed. Differences in demographics and drug use parameters studied with one-way ANOVAs (comparing across dosage categories). Chi-square was used to assess differences in distribution between dosage categories (non-user control, low, medium, high) and self-reported problems (non-user control, present, or absent). Low use = 1-100 tablets taken in lifetime, medium use = 100-500 tablets taken in lifetime, and high use = 500 or more tablets taken over lifetime.

Everyday Life Events and Cognitive Function - Two separate ANOVAs were performed on each assessment or test scores except for measures of immediate and delayed recall. Self-reported ecstasy-related problems as a factor in one analysis (non-user control, present or absent) and ecstasy use category (non-user control, low, medium or high) serving as a factor in the second analysis. Immediate and delayed recall analyzed via repeated measures ANOVA, with time of measurement (immediate, delayed) serving as a repeated measure, and with self-reported problems serving as a between-subjects factor in one analysis and drug dosage category as a between-subjects factor in a second analysis. P was set at 0.01 and post-hoc comparisons were made with Duncan's range.

Results - Significant Differences Found: Demographics and Drug Use - See "Drug Use Parameters" and "Demographics" sections above." There were significant differences in years of education obtained across dosage categories and controls, but not between "problem" and "no problem" users.

Problem/No Problem comparisons - Everyday Life Events and Psychological Function - "Problem" ecstasy users reported more cognitive failures than "no problem" users (no information on comparisons between ecstasy users and non-users on this measure). Cognitive Function - RT - "Problem" ecstasy users had slower reaction times than non-users. Executive Function - "No problem" ecstasy users had longer planning time than either non-users or "problem" ecstasy users ("no problem" users > non-users, "problem" users). Memory - Both "problem" and "no problem" users had lower scores on the spatial working memory test. Dosage category analysis - Memory - Both high and medium ecstasy users had a greater number of errors on the spatial working memory task than non-user controls. Executive Function - High users had significantly longer planning times than did low users or controls on the TOL task.

Results - No Significant Differences: Demographics and Drug Use - See "Drug Use Parameters" and "Demographics" sections above."

Problem/No problem comparisons - Everyday Life Events and Psychological Function - Non-users, "no problem" ecstasy users and "problem" ecstasy users reported the same degree of positive events, negative events and stressful events (uplifts, hassles and stresses), or comparisons on these subscales were not performed. Cognitive Function - RT- "Problem" and "no problem" ecstasy users did not have significantly different reaction times (though it was slowest in "problem" users). Memory - There were no significant differences between non-users, "problem" users and "no problem" users on performance on immediate or delayed verbal recall as measured through prose recall or as measured through list recall. Executive Function - There no significant differences in performance on WCST for non-users, "problem users" and "no problem" users, and there were no differences in all other TOL scores for non-users, "problem" users and "no problem" users (solution time, number of correct solutions and number of trials completed). Everyday Life Events and Psychological Function - Low, medium and high ecstasy users did not differ from each other or from non-users on all UHSC scores, including that for self-reported cognitive failures. RT - There were no significant differences in reaction time for low, medium and high ecstasy users. Memory - Non-users, low, medium and high users did not differ on immediate and delayed prose recall or on either immediate or delayed list recall. Executive Function - Non-users, low users, medium users and high users did not significantly differ from each other on WCST performance. None of the groups (non-users, low, medium or high users) differed on TOL solution time, number of errors or trials completed.

Overall Effects: Self-reported problems attributed to ecstasy use, including self-reported cognitive difficulties, did not predict performance on most measures of memory and executive function. Ecstasy users who reported experiencing problems had slower reaction times on a visual RT task than did non-users. However, longer planning time on the TOL (a measure of executive function) was recorded for ecstasy users who reported no problems after ecstasy use, and both "problem" and "no problem" users made more errors on a test of working spatial memory. There were no differences in immediate or verbal recall or on WCST performance. When ecstasy users were placed into three different dosage categories (low, medium, high), high and medium ecstasy users made a greater number of errors on a test of spatial working memory when compared with controls. (Low use participants apparently did not make as many errors as the higher use participants, but did not make as few errors as to show performance significantly different from the higher dose ecstasy users). High dose users needed more planning time for the TOL task than either low dose users or non-user controls. There were no significant differences in performance on the WCST, either across self-reported problems or across dosage category, and ecstasy users performed as well as controls on this test of executive function. Ecstasy use, self-reported problems attributed to ecstasy use and dosage category all failed to predict performance on tests of immediate and delayed verbal memory, whether tested with a prose recall task or with a matching (i.e. organized by category) list-learning task. The authors' first hypothesis was not confirmed; there were only two significant differences between test performance in "problem" ecstasy users, "no problem" ecstasy users and non-user controls. The second hypothesis was partially confirmed; this hypothesis was not well stated, but higher ecstasy use (versus non-use) and extent of ecstasy use (in lifetime dosage) was associated with impaired performance on a test of spatial working memory.

Comments: This retrospective study comparing ecstasy users with non-users is notable for explicitly comparing self-reported ecstasy related problems with performance on tests of cognitive function. It is possible that people reporting problems attributed to ecstasy are strongly influenced by their recall of tasks where reaction time was important, rather than more complex tasks, as variations in RT was correlated with presence or absence of self-reported problems. Impaired performance on a test of spatial working memory, but not verbal memory, stands in contrast to findings reported in other studies (e.g. Bolla et al, 1998; Morgan, 1999; Rodgers, 2000). At least in the case of list-learning tasks, other papers have employed word lists that were not organized by category, whereas those used by Fox et al were, suggesting other assessments of memory may have inadvertently measured organization or executive function. This study shares the limitations of many other retrospective studies, including small sample size, non-random assignment, reliance on self-report alone for verifying abstinence from substance use prior to testing and some poorly matched groups. For instance, it still appears that number of other substances used increased with ecstasy dosage category, so that impaired cognitive function may still be related to use of other drugs. The authors themselves point out that this study, and most others, are not able to rule out effects produced by use of other drugs.