Purpose: Neurofunctional: To examine possible differences in the intensity dependence of auditory evoked potentials in ecstasy users, cannabis users and non-users, with differences in auditory ERPs considered evidence of damage to 5-HT neurons.

Design: Non-experimental (retrospective) 3-group between subjects design comparing drug-free ecstasy users with cannabis user and non-drug user controls, with drug use (ecstasy + cannabis, cannabis only or control) serving as a between-subjects factor. All participants underwent EEG measures of auditory event related potentials (ERPs).

Subjects: 22 regular ecstasy users, 19 long-time cannabis users and 20 drug-na#&239;ve controls probably residing in the London (England) area. No information is provided on recruitment, but an earlier paper by the same authors states that participants were recruited via word of mouth, advertisements and through a local magazine. Matching - On gender, estimated IQ (via NART score) and approximately matched on age. Authors do not explicitly state that they have matched samples, but samples were tested for differences, and examining the data suggest attempted matching.

Criteria for Inclusion - Ecstasy Users - Having used ecstasy at least 20 times over lifetime. Cannabis users - Having used cannabis at least twice a week over at least 2 years, and having used ecstasy irregularly. Non-drug user controls - Never having used ecstasy and using cannabis only infrequently. All groups - Absence of self-reported neurological disturbances, and not taking any prescribed psychotropic medication. No requirement for abstinence from psychoactive drugs stated; previous paper by same authors required abstinence from all psychoactive substances for at least 48 h prior to study day, with abstinence verified via self-report only.

Drug Use Parameters, Ecstasy Use - Ecstasy users reported consuming 225.9 ± 185.5 tablets over a lifetime, with no information provided concerning average dose per use. Average duration of use, in months, was 54 ± 31.2 months, with average frequency of use (tablets per month) as being 2.9 ± 4.3 tablets per month. No information is provided concerning time since last use. Cannabis users reported using 2.6 ± 4.4 ecstasy tablets over a lifetime, with no information provided concerning average dose per use. Average duration of ecstasy use in cannabis users, in months, was 28.8 ± 54 months, with cannabis users reporting average frequency of ecstasy use as 0.07 ± 0.2 tablets per month. No information provided on time since last use. Cannabis use - Ecstasy users reported smoking 2.3 ± 2.8 joints per week. Cannabis users reported smoking 21.6 ± 17.2 joints per week. No other information provided on parameters of cannabis use.

Group Demographics and Matched Variables - The authors apparently sought to match participants in all three groups on the basis of gender, and estimated IQ (via NART score), and age, but significant differences were found in age between groups. Gender, as M/F ratio: Ecstasy users = 13/9, cannabis users = 13/6, non-drug users = 10/10. Age - Average age for ecstasy users was 27.1 ± 5.9 years, 34.2 ± 9.7 years for cannabis users, and 29.1 ± 8.4 years for non-drug user controls. It appears that cannabis users are significantly older than ecstasy users. Estimated IQ, as estimated via National Adult Reading Test (NART) score, average estimated IQ for ecstasy users = 112.8 ± 8.2, average estimated IQ for cannabis users = 114.9 ± 5.8 and for non-drug users = 114.7 ± 8.9.

Measures: Estimated Verbal Intelligence - Measured via NART. Auditory ERPs - EEG was recorded after binaural presentation of 1000 Hz tones, with tone intensity increasing from 60 dB by increments of 10 (60, 70, 80, 90 and 100 dB), with 1000 presentations. Five ERPs were created for each participant by averaging all epochs where stimuli of same intensity were presented. N1 (80-140 ms post-stimulus) and P2 (140-260 ms post-stimulus) were calculated. The N1/P2 complex was calculated for each participant, with N1/P2 complex defined as the difference between the N1 and P2 amplitudes. The slope of the relationship between N1/P2 complex and stimulus intensity was estimated for each participant by conducting least-squares linear regression with N1/P2 amplitude as criterion variable and stimulus intensity (in dB) as predictor. Analyses: Intensity Dependence Slopes - Compared across groups (ecstasy users, cannabis users and non-drug user controls) via the Kruskal-Wallis nonparametric test, with post hoc comparisons made via non-directional Mann-Whitney test, with Bonferroni correction.

Drug Use Parameters, Group Demographics and Intensity-Dependence Slopes - Only ERPs from ecstasy users analyzed via multiple regression, with total number of tablets consumed, frequency of ecstasy use, age and gender as predictors and with N1/P2 slope as dependent (criterion) variable. Total number of ecstasy tablets used was "normalized" via square root function and frequency of use was transformed via natural log. A correlational analysis was performed on transformed total number of tablets used and transformed frequency of use. Results - Significant Differences: Intensity Dependence Slopes - There were significant between-group differences in intensity dependence slopes, with ecstasy users > cannabis users, non-drug user controls.

Drug Use Parameters, Group Demographics and Intensity-Dependence Slopes - Total number of tablets used (transformed) significantly predicted intensity dependence slope, with a greater number of tablets associated with increased N1/P2 slope. Participants' gender was also significantly predicted intensity dependence slopes, but specificity of association not provided. Results - No Differences Found: Intensity Dependence Slopes - Cannabis users and non-drug user controls did not have significantly different intensity dependence slopes.

Drug Use Parameters, Group Demographics and Intensity-Dependence Slopes - Frequency of ecstasy use (transformed) did not significantly predict intensity dependence slopes. Participants' age did not significantly predict intensity dependence slopes. Transformed total number of ecstasy tablets used and transformed frequency of use were not significantly correlated with each other, indicating no strong association between frequency of use and lifetime (total number of tablets) ecstasy use.

Overall Effects: Differences in auditory ERP associated with stimulus intensity were measured in ecstasy users who also used cannabis, cannabis users reporting little or no ecstasy use and non-drug using controls who reported no use of either drug. Groups were matched on gender and estimated IQ, but only approximately on age. Ecstasy users had a significantly higher intensity dependence slope than either cannabis users or non-drug user controls, indicating more sensitivity to stimulus intensity. When age, gender, total number of tablets used in a lifetime, and frequency of use were all regressed on intensity ecstasy users' intensity-dependence slopes, it was found that gender and lifetime ecstasy use were related to intensity dependence slope, while age and frequency of use were not related to intensity dependence slopes. Greater intensity-dependence slopes were associated with greater number of ecstasy tablets taken in a lifetime, but not with frequency of ecstasy use. Total number of tablets taken across a lifetime was not associated with frequency of use (tablets taken per month).

Comments: This is the second of 2 reports finding differences in auditory ERPs in ecstasy users, as compared with cannabis user and non-drug user controls. An earlier study conducted by Tuchtenhagen and colleagues also found differences in ERP as intensity increased, though they used a somewhat different method of electrophysiological recording (Tuchtenhagen et al 2000). However, the work of Tuchtenhagen et al is not referenced in this paper. The authors of this paper claim, apparently on the basis of two studies, only one of them published, that intensity-dependence slope is a reliable measure of serotonergic function. While other aspects of the auditory ERP have been investigated in non-human primates given a neurotoxic regime of MDMA (Taffe et al, 2001), there are no studies examining intensity-dependence slope in non-human animals given MDMA in doses known to be neurotoxic. The authors also claim that they have demonstrated a causal link between ecstasy use and increased intensity dependence slope because lifetime ecstasy use predicted intensity dependence slope, but frequency of use (which the authors considered a measure of novelty-seeking or prior serotonergic dysfunction) did not do so, and lifetime ecstasy use was uncorrelated with frequency of ecstasy use. Both conclusions, and especially statements concerning causality in relation to a regression analysis, seem unwarranted. Regression analyses remain correlational in nature, and thus results of regression analyses cannot be used to derive claims of causality, even if steps have been taken to reduce the possible influence of other factors. As well, the possibility that changes in auditory ERP are due to residual drug effects cannot be ruled out, because authors have not reported or analyzed data pertaining to the time since last use for ecstasy or for cannabis.