Purpose: Neuropsychological, cognitive function: to investigate whether changes in behavioral impulsivity associated with regular ecstasy use persist after prolonged abstinence from ecstasy, and to investigate associations between the use of other drugs and changes in behavior and cognitive function. Specific Hypotheses Tested - (1) that current ecstasy users and former ecstasy users would exhibit greater numbers of psychiatric symptoms and greater selective deficits on tests of cognitive function when compared with polydrug users who had never used ecstasy and non-user controls, (2) that former ecstasy users (abstinent for at least 6 months) would exhibit fewer psychiatric symptoms than would current ecstasy users, but that (3) former ecstasy users would continue to have selectively impaired performance on tests of cognitive function when compared with polydrug users or non-drug user controls, and (4) there would be dose-response relationships between the degree of impairment on cognitive performance and parameters of past ecstasy use.

Design: Retrospective (non-experimental) between-subjects design, with drug use (current ecstasy use, previous ecstasy use, polydrug use and no drug use) as a between-subjects variable, and with the effects of different drugs also examined via regression analysis. All participants underwent tests of cognitive performance and behavioral impulsivity.

Subjects: 18 current ecstasy users, 15 former ecstasy users, 16 polydrug-no ecstasy users, and 15 non-drug using controls, probably residing in the Swansea (Wales) area, and recruited via snowball sampling. Matching - Age, education level, estimated verbal IQ, approximately matched on gender.

Criteria for Inclusion - Current ecstasy Users - Had used ecstasy on at least 20 separate occasions. Former ecstasy users - Had used ecstasy on at least 20 occasions, but had abstained from further use for at least 6 months before the study day, with abstinence verified by self-report only. Polydrug Users - No past or current use of ecstasy, but drug use pattern that is otherwise similar to that of ecstasy users. No-drug controls - No prior history of ecstasy use and no use of other drugs save alcohol and nicotine. All Groups - No past or current major medical or psychiatric illnesses, not pregnant, and no evidence of alcohol or substance abuse. Abstinence from alcohol for 10 h before study day, abstinence from cannabis for 24 h and from all other psychoactive drugs for at least 1 wk before study day, with abstinence verified via self-report only.

Drug Use Parameters - Ecstasy Users - (Ecstasy use figures averaged across men and women, data presented separately by gender in paper). On average, current ecstasy users reported taking an average of 303 +/- 267 tablets over a lifetime (20-?600), and that the average dose per occasion was 2.2 +/- 0.8 tablets. Current ecstasy users reported taking ecstasy on an average of 1.5 +/- 1.7 times a month (assessed as number of times in previous month), and that they had used ecstasy for an average of 52.8 +/- 22.8 months. The average time since last use and study day was 28.35 +/- 22.4 days for current users. Average maximum dose of ecstasy used on 1 occasion, in tablets, was 5.05 +/- 2.65. Former ecstasy users reported taking 512.73 +/- 714.4 tablets over a lifetime, with average dose per occasion reported as 1.67 +/- 0.73 tablets. Average frequency of dose per month not reported (abstinent for at least 6 mo). Average duration of use for former users, in months, was 49.08 +/- 144,6 months. The average time since last use and study day was 785.4 +/- 606.2 days for former users. Average maximum dose of ecstasy used on 1 occasion, in tablets, was 3.17 +/- 1.79 tablets. Other Drug Use (use per year reflects usage in year prior to study) - Cannabis was used by 17/18 current ecstasy users, who smoked 519 +/- 794 joints in a year, with a reported duration of use of 5.4 +/-4 years (64.8 +/- 48 months). 11/15 former ecstasy users used cannabis, smoking 785 +/- 1484 joints in a year, and had used it for 7.2 +/- 3.5 years (86.4 +/- 42 months). 14/16 polydrug users used 590 +/- 1442 joints in a year, having used it for 5.3 +/- 2.93 years (70.8 +/- 35.16 months). Psilocybin ("magic mushrooms") used by 11/18 current ecstasy users, who consumed 75.3 +/- 90.8 mushrooms in past year, and had used mushrooms for 39.5 +/- 52.2 months. 4/15 former ecstasy users had used magic mushrooms, taking 30.8 +/- 57.4 in past year, reporting use for 27.2 +/- 31 months. 3/16 polydrug users reported using magic mushrooms, reporting use of 30 +/- 70.4 in a year, and duration of use at 16.1 +/- 34.4 months. Cocaine was used by 11/18 current ecstasy users, reporting 2.14 +/- 3.4 grams consumed in past year, with duration of use reported at 27.1 +/- 37.4 months. 3/15 former ecstasy users used cocaine, consuming 0.15 +/- 0.3 grams in past year, and duration of use reported at 18.7 +/- 25 months. 4/16 polydrug users used cocaine, consuming 2.56 +/- 6.1 grams in past year, with duration of use reported at 4.7 +/- 8.4 months. Amphetamine was used by 9/18 current ecstasy users, who reported consuming 29.9 +/- 56.4 grams in past year, and reporting having used it for 42.4 +/- 45.5 months. 3/15 former ecstasy users reported using amphetamine, consuming 4.6 +/- 15.4 grams in past year, reporting duration of use of 45.3 +/- 28.8 months. 4/16 polydrug users used amphetamine, consuming 2.6 +/- 6.5 grams in last year, and reporting duration of use of 26 /= 25.1 months. LSD was used by 8/18 current ecstasy users, who had taken 8.5 +/- 18.2 trips in past year, and with a reported duration of use of 53.3 +/- 59 months. 3/15 former ecstasy users had used LSD, taking 2.71 +/- 6.9 trips in past year, reporting duration of use of 40.1 +/- 32.8 months. 4/16 polydrug users used LSD, taking 0.4 +/- 0.8 trips in last year, with reported duration of use of 16.7 +/- 31.7 months. Poppers were used by 5/18 current ecstasy users, who used 10.4 +/- 38.6 hits in past year, and reported duration of use of 3.1 +/- 4.3 months. 2/15 former ecstasy users had used poppers, taking 1.7 +/- 6.2 hits in past year, with reported duration of use of 1.2 +/- 2.1 months. 1/16 polydrug user reported using poppers, taking 15.6 +/- 62.3 hits in last year, with reported duration of use of 0.9 +/- 2.6 months. 5/18 current ecstasy users reported using 3 to 15 benzodiazepine tablets in the past year, and 3/18 reported taking 1 to 10 doses of ketamine in past year; 1/15 former user reported using benzodiazepines. Non-drug users smoked significantly fewer cigarettes than members of all other groups, and former ecstasy users consumed significantly less alcohol. Men consumed more alcohol and more ecstasy than women, but there were no other differences in substance use.

Group Demographics and Matched Variables - Authors matched participants for age, education level and estimated verbal IQ, and approximately matched (no significant differences) on gender across groups. Authors also state groups matched by height and weight, but this data is not presented. Gender, as M/F ratio - current ecstasy users = 9/9, former ecstasy users = 4/11, polydrug-no ecstasy users = 8/8, and non-user (no drugs) controls = 6/9. Age, in years - Average age of current users was 23.4 +/- 2.2, for former ecstasy users = 24.7 +/- 2.5, for polydrug-no ecstasy use = 22.1 +/- 3.3, and for non-users = 22.4 +/- 4.1 Education Level - Current ecstasy users, approximately 15.3 years (2.61 +/- 0.7), former ecstasy users, approximately 15.5 years (2.67 +/- 0.6), for polydrug users, approximately 16 years, (3 +/- 0): no-drug, approximately 15.7 years (2.73 +/- 0.6). Education level 1 = Basic high school exam, 2 = Advanced high school exam, 3 = University degree. Estimated Verbal IQ - As estimated through NART score, for current ecstasy users = 116.6 +/- 6.8, for former ecstasy users, 114.2 +/- 4.4, for polydrug-no ecstasy, 116.4 +/- 6.6, and for non-user controls, 114.5 +/- 5.9

Measures: Estimated Verbal IQ - Assessed via NART.

Psychological Function - The GHQ, the 90 item SCL-90-R and the IVE (Impulsiveness Venturesomeness Empathy) were used to measure psychological health and personality.

Impulsivity - Trait impulsivity measured through IVE (personality measure, self report). Behavioral measure: Impulsivity measured via another CANTAB test, the Matching Familiar Figures Test (MFF20). Subjects scored on time to first response, the first alternative indicated and number of errors before correct response. Latencies and errors analyzed in this study.

Memory - Immediate and delayed recall measured via Rivermead Behavioural Memory Test (RBMT), with subjects writing down as much as they could recall from audiotaped story immediately after presentation and again 40-50 minutes after presentation, with subjects performing unrelated tasks between the measures. Recall scored by number of ideas correctly recalled.

Executive Function - Assessed via Controlled Oral Word Association task (COWA). Stroop task, Serial Subtracted Sevens task (SSS), digit cancellation task and Trail Making Task. COWA similar to functional fluency; generate words beginning with consonant in 60 s and members of specific category (e.g. "fruit") in 90 s. Digit cancellation task requires marking of target digit, with target randomly placed within 2 digit numbers in array.

Analyses: Psychological Function - A one-way analysis of variance (ANOVA) performed on IVE, GHQ and SCL-90-R scores, with drug use (current ecstasy, former ecstasy, no ecstasy, no drugs) serving as a between-group factor, and p set at 0.05. (Same method used to compare demographics, NART score, drug use history), with Kruskal-Wallis test used when items (as drug use) violated homogeneity.

Memory - A 2-way between subjects/repeated measures ANOVA was performed on immediate and delayed RBMT-Story scores, with time (immediate versus delayed) serving as repeated (within-subjects) measure and group (current ecstasy, former ecstasy, no ecstasy, no drugs) as between-subjects factor, with p. = 0.05. Planned comparisons used to compare means across all 4 groups, with p. = 0.01. Relationships between drug use parameters and cognitive performance - Pearson's product moment was used to correlate all dependent variables with drug-use parameters.

Regression - A stepwise linear regression was used to discover which drug use parameters were most predictive of psychological function and impairment on tests of cognitive function. Number of doses of cannabis, amphetamine, cocaine, poppers, LSD, magic mushrooms entered into regression, as were average and maximum ecstasy dose per use and total lifetime ecstasy consumption.

Analysis of covariance - Drug-use parameters were used as covariates, and analyses performed to examine whether differences between drug-use groups would remain if a particular drug use parameter were covaried out of analysis.

Results - Significant Differences Found: Psychological Function - Men scored higher on IVE venturesomeness, and women scored higher on empathy, across all four groups. Current and former ecstasy users had higher total SCL-90-R (Severity) and symptoms scores when compared with polydrug users and with non-user controls. Current ecstasy users had higher scores on 8 of 10 SCL-90-R subscales (including obsessive compulsive, depression, phobic anxiety, paranoid ideation, inter-personal sensitivity, somatization and altered appetite/sleep). Former ecstasy users had higher positive symptoms, somatization, obsessive compulsive, anxiety, inter-personal sensitivity and altered appetite/sleep scores when compared with controls; when compared with polydrug users, former ecstasy users only had higher positive symptom, inter-personal sensitivity and altered appetite/sleep scores.

Executive Function - While there were no overall group differences in verbal fluency, current ecstasy users generated fewer category words than former users, polydrug users or non-drug controls. Though there no group differences in overall Trail Making Test performance, current and former ecstasy users made more errors on TMT (B) than did polydrug users and non-drug user controls. Current ecstasy users made a greater number of errors on the SSS task compared with polydrug users and controls, and former ecstasy users made more errors than non-drug user controls.

Impulsivity - Both current and former ecstasy users produced faster responses to the MFF when compared with polydrug users or controls, and current and former ecstasy users had a greater number of errors compared with polydrug user and non-drug user controls. Composite "impulsivity" scores were higher for current and former ecstasy users than polydrug users or controls.

Memory - Immediate recall scores higher than delayed recall across all groups. Current and former ecstasy users had lower immediate and delayed recall scores on the RBMT story test when compared with non-drug user controls. Former ecstasy users, but not current ecstasy users, did not perform as well as polydrug user controls on the RBMT immediate and delayed recall.

Correlations - GHQ scores were highly correlated with SCL-90-R scores, and SCL-90-R scores highly intercorrelated. As expected, number of errors on MFF (behavioral impulsivity) increased as latencies decreased.

Relationships between drug-use parameters and measures - Estimated lifetime ecstasy use was positively correlated with average dose per use and maximum dose per use for all ecstasy users (current and former). Average dose per use and maximum dose per use were positively correlated. Average LSD trips taken in a year also correlated with both average dose per use and maximum dose per use. Immediate and delayed recall scores for RBMT-Story were negatively correlated with lifetime ecstasy use for current ecstasy users, but not former ecstasy users.

Regression analysis - Number of joints smoked in past year predicted SCL-90-R overall symptom severity scores, and SCL-90-R anxiety, depression, obsessive compulsive, interpersonal sensitivity, hostility, and altered appetite/sleep scores. Somatisation and psychoticism scores were predicted by cannabis use in past year plus number of poppers used in last year. Phobic anxiety scores predicted by amount of cannabis used in prior year plus amount of amphetamines used in last year, and paranoid ideation score predicted by cannabis use in the last year plus amount of cocaine used in the past year. Immediate and delayed recall scores were predicted by lifetime ecstasy use. All MFF scores (errors, latency and "I" score) were associated/predicted by average dose of ecstasy per use. Performance on RBMT-Story immediate recall and delayed recall were predicted by overall lifetime ecstasy use. SSS performance was correlated with maximum dose of ecstasy per use. Errors on the Trail Making Test were associated with number of LSD trips and number of magic mushrooms used in past year.

Analysis of Covariance - When cannabis use was covaried out, only somatization, obsessive compulsive, anxiety, phobic anxiety, altered appetite/sleep and severity scores were still significant. Both LSD use in the past year and number of poppers used in the last year were significant covariates for most SCL-90-R scores, whereas use of cocaine and amphetamine in the past year were covariates only for paranoid ideation and phobic anxiety. When examining tests of behavioral impulsivity and cognitive function, only psilocybin use in the past year was a significant covariate for the Trail Making Test; there were no longer any group differences in TMT performance after amount of psilocybin used in past year was covaried out. Group differences between ecstasy users and both types of control (polydrug control and non-drug use) remained even after use of other drugs covaried.

Results - No Significant Differences: Psychological Function - Drug use groups did not differ in any of their IVE scores. There were no differences across drug use groups for any scores on the GHQ or IVE, though there was a trend for ecstasy users to have higher impulsivity scores than non-user controls. Ecstasy users did not have higher SCL-90-R scores on the hostility or psychoticism subscales. When compared with polydrug users, former ecstasy users did not differ on many SCL-90-R scores (including phobic anxiety, depression, obsessive compulsive, and anxiety scores.

Executive Function - There were no statistically significant differences across drug use groups on Stroop task, SSS task, Trail Making Test, verbal fluency (COWA) task or digit cancellation task. (Ecstasy users generated fewer category words compared with all other groups).

Correlations - There were no positive correlations between performance on the MFF and performance on other tests of cognitive function. Performance on the RBMT-Story, SSS, verbal (category) fluency, Trail Making Test (B) and MFF were not correlated with responses on the GHQ. None of the tests of cognitive function were correlated with any of the SCL-90-R scores, or with IVE scores. Relationship between Drug Use Parameters and Measures - There were no associations between average dose per use and performance on RBMT-Story or on MFF test for either current or former ecstasy users. There were no relationships between duration of ecstasy use or time since last use on performance on RBMT-Story or MFF for either current or former ecstasy users. No relationships were found between maximum dose per use and performance on RBMT-Story or on MFF test. While lifetime ecstasy use was negatively correlated with RBMT-story immediate and delayed recall for current ecstasy users, thee was no correlation between lifetime ecstasy use and either recall score for former ecstasy users.

Regression analysis - Average dose of ecstasy per use, duration of use and maximum dose per use were not associated with performance on the RBMT-Story task. Neither lifetime ecstasy use nor duration of use or maximum dose per use predicted any of the MFF scores. SSS scores were not predicted by lifetime ecstasy use, average dose per use or duration of use. None of the measures of ecstasy consumption (total lifetime use, average dose per use, duration or maximum dose per use) predicted scores on SCL-90-R (including total severity score or any sub-scale scores). None of the drug use parameters predicted performance on Stroop test, digit cancellation test, or either verbal (consonant) or verbal (category) fluency test. None of the ecstasy use parameters were associated with performance on the Trails Making Test.

Analysis of Covariance - There were no group differences in SCL-90-R scores when cannabis use, LSD use, cocaine use and popper use were all covaried with analysis. However, use of other drugs were not significant covariates in most tests of impulsivity and cognitive performance.

Overall Effects: Both current and former ecstasy users had higher scores on one measure of psychiatric symptoms (SCL-90-R), but not on another measure of psychological health (GHQ). The greatest differences were between current users and both polydrug user and non-drug user controls, and with former ecstasy users displaying fewer psychiatric symptoms than current users. A stepwise regression and analysis of variance suggest that cannabis use and use of other drugs, such as LSD and amphetamines, were more strongly associated with increased psychiatric symptoms than ecstasy use. Ecstasy use was not associated with higher impulsivity scores on the IVE. Nevertheless, both current and former ecstasy users were likely to respond more quickly to the MFF trials than non-ecstasy users, and to make a higher number of errors. Both current and former ecstasy users demonstrated signs of impaired immediate and delayed verbal memory, as measured via RBMT-Story. Current ecstasy users, but not former users, made a greater number of errors on the SSS test and listed fewer category members in the verbal fluency test. Impaired performance on measures of immediate and delayed recall was associated with number of tablets used in a lifetime, whereas increased impulsivity was associated with average ecstasy dose size and increased errors on the SSS were associated with maximum dose size. Use of psilocybin in the prior year still seemed to make an independent contribution to performance on the trail making test, with increased use associated with impaired performance. The first hypothesis was only partially confirmed; current and former ecstasy did have higher psychiatric symptoms than non-drug controls, but a large part of the variance in psychiatric symptoms can be explained by use of other drugs, such as cannabis, amphetamines and LSD. Both current and former ecstasy users did score below both sets of controls on some measures of cognitive function (RBMT-Story), but not on others (digit cancellation, verbal fluency.) The second and third hypotheses were confirmed, in that former ecstasy users did display lower psychiatric symptoms and distress than did current ecstasy users, but the difference between current and former ecstasy users on measures of behavioral impulsivity and verbal memory were not statistically significant. The fourth hypothesis was confirmed, (there were dose-response relationships between impairment on some tests of memory and executive function and different ecstasy use parameters) but its vagueness makes it liable to be confirmed.

Comments: The study described in this paper is one of several attempts to distinguish the effects of ecstasy on psychiatric function and cognitive performance from the effects of other drugs by employing drug user controls as well as drug-naive controls. Other researchers have tended to focus on cannabis user comparison groups (Croft et al. 2001; Gouzoulis-Mayfrank et al. 2000; Rodgers 2000). Previous reports by these researchers have employed polydrug user controls rather than cannabis user controls (Morgan 1998; Morgan 1999), but this is the first attempt at examining the effects of polydrug use in general, and the effects of other drugs on psychological function and cognitive performance. Like other recently published papers (e.g. Daumann et al. 2001; Fox et al. 2001b; Morgan et al. 2001; Parrott et al. 2001b), these study findings suggest psychiatric symptoms in ecstasy users cannot be used as an indicator of serotonergic neurotoxicity. Furthermore, the results of this study suggest that indices of self-reported psychiatric problems cannot be used to demonstrate neurotoxicity, since drugs such as cannabis and LSD mediated self-reported psychiatric symptoms, yet are not known to be neurotoxic (i.e., do not produce gross alterations in cellular morphology or increases in indicators of cell damage). This study shares the limitations of retrospective studies, including non-random assignment and the possibility of reverse causality, and verification of abstinence from drugs was done by self-report only. In addition, lifetime usage figures were used to calculate predictive values for ecstasy, but not for other drugs (as amphetamine or cocaine).