Auditory event-related potentials (P3) and cognitive performance in recreational ecstasy polydrug users: evidence from a 12-month longitudinal study.

Abstract

There is important preclinical evidence of the long-lasting neurotoxic and selective effects of ecstasy (MDMA) on serotonin systems in nonhuman primates. In humans, long-term recreational use of ecstasy has been mainly associated with memory impairment.
The first aim of our study was to evaluate the cognitive and electrophysiological long-term alterations associated with lifetime ecstasy use within a sample of ecstasy polydrug users along a 1-year follow-up. Our second aim was to explore the relationship between specific cognitive functions and P300 (P3) event-related potentials (ERPs) in ecstasy users.
We conducted auditory P3 latency and amplitude and administered a battery of cognitive tests to three groups of subjects: 14 current ecstasy polydrug users, 13 current cannabis users, and 22 controls free of illicit drugs in two evaluations during 1 year.
We found significant differences between ecstasy users and controls on cognitive measures of word fluency, processing speed, and memory recognition after 1-year follow-up. We found no significant differences between ecstasy and cannabis users or cannabis users and controls on cognitive tests. Lifetime ecstasy use was associated with poorer memory recognition. No group differences were shown on P3 latency or amplitude. Significant correlations emerged between P3 latency and cannabis lifetime use (higher cannabis use was related to faster latency, showing a paradoxical effect) but not with ecstasy exposure.
Our findings provide evidence of mild long-term cognitive deficits among ecstasy polydrug users. Both ecstasy use and the dynamic interaction between ecstasy and cannabis effects may account for these deficits. No significant P3 alterations were found in ecstasy users.

Full text

Page 1

ORIGINAL INVESTIGATION
Auditory event-related potentials (P3) and cognitive
performance in recreational ecstasy polydrug users: evidence
from a 12-month longitudinal study
Susana de Sola & Thais Tarancón &
Jordi Peña-Casanova & Josep María Espadaler &
Klaus Langohr & Sandra Poudevida & Magí Farré &
Antonio Verdejo-García & Rafael de la Torre
Received: 1 February 2008 /Accepted: 28 May 2008 /Published online: 26 June 2008
# Springer-Verlag 2008
Abstract
Rationale There is important preclinical evidence of the
long-lasting neurotoxic and selective effects of ecstasy
(MDMA) on serotonin systems in nonhuman primates. In
humans, long-term recreational use of ecstasy has been
mainly associated with memory impairment.
Objective The first aim of our study was to evaluate the
cognitive and electrophysiological long-term alterations
associated with lifetime ecstasy use within a sample of
ecstasy polydrug users along a 1-year follow-up. Our
second aim was to explore the relationship between specific
cognitive functions and P300 (P3) event-related potentials
(ERPs) in ecstasy users.
Materials and methods We conducted auditory P3 latency
and amplitude and administered a battery of cognitive tests
to three groups of subjects: 14 current ecstasy polydrug
users, 13 current cannabis users, and 22 controls free of
illicit drugs in two evaluations during 1 year.
Results We found significant differences between ecstasy
users and controls on cognitive measures of word fluency,
processing speed, and memory recognition after 1-year
follow-up. We found no significant differences between
ecstasy and cannabis users or cannabis users and controls
Psychopharmacology (2008) 200:425–437
DOI 10.1007/s00213-008-1217-5
S. de Sola:J. Peña-Casanova (*)
Behavioral Neurology and Dementia Section,
Neurology Department, Hospital del Mar,
Universitat Autònoma de Barcelona,
Passeig Maritim, 25–29,
08003 Barcelona, Spain
e-mail: jpcasanova@imas.imim.es
S. de Sola:T. Tarancón:J. Peña-Casanova:J. M. Espadaler:
K. Langohr:M. Farré:A. Verdejo-García:R. de la Torre
Neuropsychopharmacology Programme,
Institut Municipal d’Investigació Mèdica,
Parc de Recerca Biomèdica,
Barcelona, Spain
T. Tarancón:J. M. Espadaler
Neurophysiology Section, Neurology Department,
Hospital del Mar,
Barcelona, Spain
K. Langohr:S. Poudevida:M. Farré:A. Verdejo-García:
R. de la Torre
Unitat de Recerca en Farmacologia,
Institut Municipal d’Investigació Mèdica,
Barcelona, Spain
M. Farré
UDIMAS, Universitat Autònoma de Barcelona,
Bellaterra, Spain
R. de la Torre
CEXS, Universitat Pompeu Fabra,
Barcelona, Spain
K. Langohr
Departament d’Estadística i Investigació Operativa,
Universitat Politècnica de Catalunya,
Barcelona, Spain
A. Verdejo-García
Institute of Neuroscience, Universidad de Granada,
Granada, Spain

Page 2

on cognitive tests. Lifetime ecstasy use was associated with
poorer memory recognition. No group differences were
shown on P3 latency or amplitude. Significant correlations
emerged between P3 latency and cannabis lifetime use
(higher cannabis use was related to faster latency, showing
a paradoxical effect) but not with ecstasy exposure.
Conclusions Our findings provide evidence of mild long-
term cognitive deficits among ecstasy polydrug users. Both
ecstasy use and the dynamic interaction between ecstasy
and cannabis effects may account for these deficits. No
significant P3 alterations were found in ecstasy users.
Keywords Cognitiveimpairment.Cannabis.Ecstasy.
MDMA.P300.Event-relatedpotentials.ERP
Introduction
Ecstasy or 3,4-methylenedioxymethamphetamine (MDMA)
is a ring-substituted amphetamine derivative, structurally
related to the hallucinogenic compound mescaline, which
has been tentatively classified as entactogen drug for its
main psychoactive profile that distinguishes it from
classical hallucinogens and stimulants (Green et al. 2003).
Ecstasy and other amphetamine type stimulants are nowa-
days the second most commonly used illicit drugs after
cannabis (OEDT Annual Report 2007; UNODC 2007).
Epidemiological surveys refer that almost 8.6 million
people have taken ecstasy at least once in their life with
Europe accounting approximately for 36% of all ecstasy
users worldwide (UNODC 2007). The popularity of ecstasy
use may be explained by its highly desirable psychological
and behavioral effects, which include loss of inhibitions,
euphoric state, reduced anxiety, and emotional openness
and empathy (Morgan 2000). Frequent long-term effects
related to prolonged ingestion are the development of
tolerance, impaired concentration, depression, and in-
creased openness to others (Verheyden et al. 2003).
MDMA affects peripheral and central nervous system
functions acting mainly on the serotonergic system, block-
ing 5-TH reuptake, inducing 5-TH release, and to lesser
extent, also causing DA and NE release (Lyles and Cadet
2003). Animal studies evidence that MDMA is neurotoxic
(Green et al. 2003) and can induce persistent alterations in
the brain serotonin system when given in high doses. In
nonhuman primates, sensitivity to its neurotoxic effects has
shown to be more pronounced, resulting in higher rates of
5-HT depletion with smaller doses of MDMA, lower 5-HT
levels, reduced serotonergic axon density, and persisting
abnormal reinnervation patterns for as long as 7 years
posttreatment (Hatzidimitriou et al. 1999; Scheffel et al.
1998). Up to date, the generation of free radicals by
MDMA metabolites, in addition to the associated oxidative
stress and membrane damage, has been pointed as the
possible causal agents for long-term MDMA-induced
neurodegeneration (Green et al. 2003). Considering the
important limitations of the applicability of animal studies
to humans (de la Torre and Farre 2004), a relevant question
is whether similar changes may occur in humans and their
potential impact on cognition. There is growing consensus
that MDMA might be selectively neurotoxic to serotonergic
modulation in humans (Morgan 2000; Parrott 2006):
studies have shown reduced cerebrospinal fluid 5-HIAA
levels, significant decrease in 5-HT binding, and altered
cerebral glucose metabolism in MDMA users compared to
nonusers (Green et al. 2003; Lyles and Cadet 2003;
McCann et al. 2005), and there is limited evidence of
partial recovery of these alterations (Thomasius et al. 2006).
Neuropsychological studies have revealed a broad range
of cognitive deficits associated with MDMA use, including
attention, learning, memory, and executive function alter-
ations (Kalechstein et al. 2007; Lamers et al. 2006;
Reneman et al. 2006; Zakzanis et al. 2007). Furthermore,
most of these studies have demonstrated dose-related
cumulative effects of lifetime ecstasy use on cognitive
functioning (Morgan 2000; Parrott 2006). In an attempt to
assess the durability of these alterations, recent longitudinal
studies have observed that ecstasy use-related deficits are
persistent but do not deteriorate further with time (Gouzoulis-
Mayfrank et al. 2005; Reneman et al. 2006; Thomasius et al.
2006). Furthermore, there is mixed evidence of partial
recovery or stability of deficits after prolonged ecstasy
abstinence (i.e., the pattern is different across different cases),
whereas continuation of ecstasy use is associated with further
deterioration of memory functioning (Zakzanis and Campbell
2006).
The observed cognitive deficits are likely related to
alterations in other neurobiological indices, such as elec-
trophysiological activity. Several electrophysiological stud-
ies have been conducted in ecstasy users attempting to find
sensitive measures of central serotonin dysfunction. Studies
on dipole source analysis have reported that ecstasy users,
compared to controls, have increased intensity dependence
of the tangential N1/P2 source activity with higher stimulus
intensity (Tuchtenhagen et al. 2000; Croft et al. 2001a).
Furthermore, there is an association between the intensity
dependence of the auditory-evoked potentials and several
measures of ecstasy use (Croft et al. 2001a; Daumann et al.
2006), although it is unclear if this association is stable
across time (Daumann et al. 2006). Other electrophysio-
logical measures such as P300 (P3), which has been
extensively studied in research on substance use could also
be impaired. P3 is considered a measure of cortical activity
on complex processing and is normally assessed by means
of its two main components: amplitude (μV) and latency
(ms), although other subcomponents have also been
426Psychopharmacology (2008) 200:425–437

Page 3

proposed (Polich 2007). The P3 is believed to underlie the
neural mechanisms required to switch the mental set to
make appropriate responses to changing demands. In this
complex process, attention and memory interaction seem to
play a key role. P3 amplitude has been associated with
memory function and with efficient allocation of attentional
resources during information processing. On the other
hand, P3 latency is considered a measure of stimulus
evaluation speed and has been associated with mental
processing speed (for review, see Linden 2005; Polich
2007). Research has linked a broad network of brain
regions to the P3 generation (Horn et al. 2003; Linden
2005) and there is evidence that P3 components reflect
indirect modulating effects from dopamine, acetylcholine,
noradrenaline, or serotonin neuromodulators (Hansenne
2000; Polich and Criado 2006).
The P3 component has shown functional significance in
a range of substance use disorders, and it has been proposed
as a biological marker predisposing to substance use
disorders and externalizing psychopathology during ado-
lescence (Iacono et al. 2002; Iacano and McGue 2006;
Yoon et al. 2006). However, to date, only few studies have
attempted to assess the proposed neurotoxic effects of
ecstasy use using P3 brain event-related potentials (Casco
et al. 2005; Gamma et al. 2005; Mejias et al. 2005). Casco
et al. (2005) found smaller P3 amplitude in heavy and
moderate ecstasy users. Furthermore, Mejias et al. (2005)
observed longer P3 latencies in MDMA heavy users when
detecting target stimuli; the larger latencies were associated
with lower levels of cognitive processing. In contrast with
previous studies, Gamma et al. (2005) concluded that
similar neuroelectric mechanisms were observed in ecstasy
polydrug users compared to drug controls after adjusting
for relevant confounds, failing to show disturbed brain
functioning in MDMA users. Overall, the available evi-
dence is mixed, tentatively suggesting that ecstasy use
effects are associated with subtle but significant neuro-
electric changes in humans. Nevertheless, these could
reflect preexisting traits that might predispose to drug use
(Gouzoulis-Mayfrank and Daumann 2006). In this study,
we aimed to investigate if the P3 component is specifically
associated with ecstasy use and with cognitive deficits
linked to ecstasy use. We believe that the P3 may be
specifically associated with ecstasy use in light of the
evidence of the neuromodulatory effects of ecstasy on
serotonin neural function and its dose-related association
with memory and attention deficits (i.e., the main cognitive
correlates of the P3).
This study compared current ecstasy polydrug users,
current cannabis users (both with high–moderate drug use),
and drug-naïve controls on cognitive performance and ERP
P3 over a period of 12 months. The cannabis group was
included because MDMA users very frequently coabuse
cannabis, which also have important effects on cognition
and electrophysiological activity (Solowij et al. 1995). The
neuropsychological battery included tests measuring atten-
tion, memory, executive function, and processing speed. P3
response was elicited using an auditory oddball paradigm
and was assessed by means of its two main components:
latency and amplitude. The aim of this study was to find
neuropsychological and electrophysiological evidence re-
lated with recreational long-term ecstasy use over the
course of 1 year. In addition, we attempted to explore the
relationship between P3 main components with specific
cognitive domains and lifetime MDMA consumption in the
ecstasy group. Based on previous findings, we hypothe-
sized that: (1) Ecstasy polydrug users would have poorer
neurocognitive performance compared to cannabis users
and drug-free controls; (2) Ecstasy polydrug users would
have longer P3 latency and/or reduced P3 amplitude
compared to cannabis users and drug-free controls; (3)
Cognitive performance and P3 response within the ecstasy
group would remain stable or deteriorate over the course of
1 year; (4) Significant associations would emerge between
P3 indices and memory function in ecstasy polydrug users.
Materials and methods
Participants
Forty-nine participants were recruited: 14 ecstasy polydrug
users, 13 cannabis users, and 22 drug-naïve controls. All
subjects were selected randomly from an ongoing 2-year
follow-up study whose results are reported elsewhere (de
Sola et al., in press). All participants were healthy, self-
reporting an adequate functioning within their social and
professional context, and were comparable in terms of years
of education and socioeconomic status. As for drug users,
subjects were not seeking for drug abuse treatment.
Participants were recruited through several sources:
‘word of mouth’ notices in the local area, advertisement
in the local university, and advertisement in a local NGO
(Energy Control) specialized in providing harm reduction
guidelines among drug users. All participants were
screened in a telephone interview to determine drug history.
Subjects with neurological, relevant medical disease, and
current psychiatric histories were excluded.
Inclusion criteria for the ecstasy group were: current
ecstasy use with a minimum intake of more than five
occasions during lifetime and at least one occasion during
the last year. For the cannabis group, inclusion criteria were
daily cannabis use or having used cannabis more than 25
times during lifetime. A number of exclusion criteria were
applied on the different groups. For the cannabis group,
current history of regular use of other illegal psychotropic
Psychopharmacology (2008) 200:425–437427

Page 4

drugs during the last year, as well as past use of illegal
drugs for more than five occasions. For nonusers, current
history of use of any illegal drugs during the past year and
past use of illegal drugs in more than five occasions.
Alcohol and nicotine were allowed. As for the ecstasy
group, because it was impossible to recruit exclusive
ecstasy users, it was decided to include ecstasy users with
moderate use of other illicit drugs, being ecstasy the main
psychostimulant drug abused.
Test procedure
This study was approved by and conducted in accordance
with the local ethics committee (CEIC-IMAS) and with the
ethical standards laid down in the Declaration of Helsinki.
Upon arrival to the research center (Institut Municipal
d’Investigació Mèdica [IMIM]), subjects were informed of
the ensuing protocol and gave their written informed
consent before participating in the study.
Testing consisted of drug history, medical checking, drug
screens, biochemical analyses, psychiatric, neuropsychologi-
cal, and neurophysiological assessments. After completing
testing, all subjects were economically compensated with 150
euros for their participation. All participants underwent a
neuropsychological assessment session of 90 min. They
completed all tasks in a quiet comfortable room. Two
neuropsychological evaluations were carried during the 12-
monthfollow-upperiod:thebaselinecognitivetesting(t1)and
a follow-up assessment 12 months later (t2). After completing
the neuropsychological testing, participants got an appoint-
ment for the ERP assessment. The electrophysiological study
included the following measures in addition to ERPs: blink
reflex and accommodation, which will be reported elsewhere.
InthecurrentERPstudy,twoelectrophysiologicalevaluations
were carried during the 12-month follow-up period: baseline
measures at ERP study onset (t1) and 12 months later (t2).
Fifty-three subjects were initially recruited for the ERP study.
All subjects had normal audition. From this initial sample,
ERPs could not be recorded in two subjects (one from the
ecstasy group and one from the cannabis group) due to
capillary problems (dreadlocks hair style), whereas two
controls were excluded from the first evaluation (t1) due to
relevant clinical abnormalities in their somatosensory-evoked
potentials.From the 49subjectsthatstartedthe ERPstudy (14
MDMA, 13 cannabis, and 22 drug-naïve controls), after
12 months, 41 participants remained in the study: 11 ecstasy
polydrug users (three drop-outs), nine cannabis users (four
drop-outs), and 21 nonusers (one drop-out). These eight
subjects left the study because they lost their interest and
missed the follow-up assessment. The loss of the follow-up
examination of these participants is not expected to have a
significant impact on cognitive and ERP results since the
number of subjects lost to follow-up was small.
Abstinence period was not strictly delimitated, although
all participants were requested to observe a minimum 72 h
abstinence period. Due to this fact, urine drug screens were
carried out by immunoassay (CEDIA, Microgenics) in all
subjects prior to neurophysiological and neuropsychologi-
cal testing in order to avoid acute effects. Drug classes
screened included: cannabis, ecstasy, cocaine, and amphet-
amine/methamphetamine. Drug screens were performed
also in hair samples by segmental analysis (last month
and previous 6 months) for the same drug classes in order
to verify self-reported drug consumption history following
methods previously published (Pichini et al. 2006; Pujadas
et al. 2003). This procedure allowed us to reliably classify
participants into the different subgroups according to the
pattern of drug use (ecstasy/cannabis vs. cannabis).
Neuropsychological tests
Premorbid intelligence estimation
Vocabulary subtest (from the WAIS-III, Spanish version;
Wechsler 1997a) Subjects are required to provide defini-
tions to 33 words with increasing difficulty. The accuracy
of word knowledge provides a good estimation of pre-
morbid IQ, and the test was used to control for preexisting
differences in general cognitive capacity between groups
which could be clinically relevant.
Attention measures
California computerized assessment package (CalCAP;
Miller 1990) We used a selection of subtests included in
the standard version, which consisted of a series of simple
and choice reaction time measures administered by a
computer. All tasks are designed to be self-explanatory.
Several cognitive domains were assessed: basal measure of
reaction time (simple RT), recognition memory (choice RT-
digits), and divided attention (sequential RT2).
Executive functions
Tower of London (TOL; Shallice 1982) This test requires
the movement of three different colored balls across three
different size pegs in order to duplicate a goal configura-
tion. Movements must follow strict rules. Scores for
planning time (latency), solution time, and total number of
moves needed to complete the configuration are obtained,
providing a good measure of planning ability.
Wordfluency(RamierandHécaen1970; Benton and Hamsher
1976) Subjects are asked to generate as many words as
428 Psychopharmacology (2008) 200:425–437

Page 5

possible in 1 min belonging to the specified category of
“animals.” High scores indicate greater verbal fluency.
Symbol digit modalities test (Smith 1973) This task
involves the conversion of meaningless geometric designs
into written Arabic number responses within 90 s. Answers
must follow the correspondence shown in a visual key. The
test involves the integration of visual, motor, and executive
functions and provides a good measure of processing speed.
Memory and learning
California verbal learning test II (Delis et al. 2000) The
test consists on a list of 16 words which is read out to the
examinee in five occasions. Then the subject must repeat as
many words as remembered each time and is required a short-
term recall after interference list in a free and cued recall trial.
After 20 min, a free and cued recall is requested again. Last
task consists in the recognition of the original list among
another extensive list words. Measures of total learning,
immediate and delayed recall, and recognition are obtained.
Rey complex figure test (Rey 1941, 1959) This test involves
a copy trial of a complex figure stimulus, followed by a
3-min immediate recall and a 30-min delayed recall where
subjects are requested to redraw the memorized complex
stimulus. Scores in immediate and delayed recalls provide a
good measure of visuospatial memory performance.
Corsi block tapping subtest (from the WAIS-III, Wechsler
1997b) Subjects must memorize and reproduce a sequence
of spatial locations demonstrated by the examiner. The
examineehastotouch theprearrangedwoodenblocksfirstin
the same order, and backwards later on. The length of the
sequence increases with the subject’s success. Direct
sequence measures attention span, while backward sequence
is predominantly a measure of visual working memory span.
Letter number sequencing subtest (from the Wechsler
Memory Scale III, Wechsler 1997b) Subjects are required
to listen to a series of numbers and letters with randomized
presentation, then repeat them back, giving numbers first
following numerical order, and afterwards letters alphabet-
ically arranged. The length of the series increases with the
subject’s success. Higher scores provide a good measure of
verbal working memory span.
ERP recordings, stimulation, and procedure
ERPs were recorded with 32 electrodes EEG International
System mounted in an electrode Quick-Cap. Electrode
placement followed the standard 10#20 International
System locations and impedance was kept below 10 K.
Recordings were made with a linked ear lobe reference.
Amplification parameters were: gain of 30,000, band-pass
of 0.1–100 Hz, and low pass filter of 30 Hz. Artifact
rejection was performed. Subjects were seated in a
comfortable upright chair in a dedicated room with a
constant low level of illumination and environmental noise.
P300 ERPs (or p3b component) were obtained by averag-
ing the signal. Two grand average ERP P3 components
were assessed: P3 latency and P3 amplitude (according to
its maximum topographical peak amplitude, central or
parietal). P300 component was not only manually identified
mainly on the basis of its latency range but also according
to its amplitude.
ERPs were evoked using an auditory oddball paradigm,
which consisted of a simple discrimination task involving
sustained attention by means of two different stimulus: a
frequent standard tone and an infrequent “target” one (tone
burst and tone pips, respectively) of 90 dB each. Both ears
were simultaneously stimulated. Subjects had to memorize
and count the infrequent stimuli (target) in two consecutive
different runs with ERPs being recorded each time in order
to obtain the best latency value from each participant.
Statistical analysis
The comparison of the three groups (ecstasy, cannabis, and
control) with respect to the sociodemographic characteristics
anddrugconsumption atbaselinewas carriedout bymeans of
one-wayanalysis ofvariance(F test) for continuous variables
and Fisher’s exact test for categorical variables. To check
whether drug consumption at baseline and after 12 months
differed significantly, the t test for paired observations was
applied. To compare the neuropsychological test results and
the P3 components (amplitude and latency) among these
groups at baseline (t1), analysis of covariance (ANCOVA)
models were used adjusting for gender and premorbid
intelligence. Spearman’s correlation coefficient was used to
measure the association between lifetime drug consumption,
neuropsychological test results, and P3 components.
Longitudinal analyses were performed for all tests scores
using the linear mixed model, which accounted for the
introduced dependency among the data due to repeated
measures at baseline and after 12 months. Additionally to
the variables considered in the ANCOVA models, the linear
mixed models included the variable ‘time of evaluation’
(baseline and 12 months). In those cases where significant
differences were found, post hoc pairwise comparisons
between categories of drug consumption were carried out
using the Tukey test for repeated measures.
Statistical analyses were performed using the statistical
software packages SPSS, version 15.0, and R, in particular,
Psychopharmacology (2008) 200:425–437429

Page 6

the R libraries ‘nlme’ (Pinheiro et al. 2007) and ‘multcomp’
(Hothorn et al. 2007). Test results were considered to be
statistically significant if the resulting p value was less than
0.05. Due to the limited sample size, p values were not
corrected for multiple comparisons.
No drug group segmentation in lifetime consumption
categories was applied due to the small sample size of both
ecstasy and cannabis group. This would have compromised
the statistical power and hence the validity of any result.
Even though we talk of the ecstasy, cannabis, and control
group, we are aware of the fact that the subjects of the first
group also consumed cannabis.
Results
Demographic and drug use data
Sociodemographics of the participants are provided in
Table 1. No significant differences were observed regarding
age or gender among the three groups. There were no
significant differences concerning years of education,
although the proportion of individuals with a university
degree or studying at a university was significantly lower in
the ecstasy group (71.4%) compared to the cannabis
(75.0%) and control (100%) groups (p=0.015). Significant
differences emerged for employment status (p=0.022) with
ecstasy group showing a minor rate of students (14.3%) and
a mayor percentage of employed (64.3%) and unemployed
(21.7%) compared to the controls (with 63.6%, 27.3%, and
9.1%, respectively). Scores on the vocabulary test (WAIS-
III) were worse in the ecstasy group (p=0.017) at baseline,
although this group difference did not persist after
12 months (p=0.4).
Table 2 displays the drug use patterns from participants,
including alcohol, tobacco, cannabis, and ecstasy consump-
tion. At baseline, the ecstasy group showed a mean total
ecstasy lifetime consumption of 207.4 tablets (SD=151.0)
and of 7.0 (SD=2.4) in terms of years on average, having
started first use at the age of 18.0 (SD=4.0). The mean
frequency of consumption during the 6 months prior to
baseline testing was 6.8 days (SD=3.7). Concerning
alcohol consumption, on average, controls showed a
significant less frequency of consumption prior to baseline
evaluation (days of use in the last 6 months) in comparison
to ecstasy users (p=0.019). Groups did not differ signifi-
cantly in terms of age at onset of alcohol use or duration of
use in years. As for smoking, there were more current
smokers in the ecstasy group (p=0.005) and again ecstasy
users showed an earlier age of onset of use at 14.3 years
(SD=1.7; p=0.005) than the other two groups. Comparing
ecstasy and cannabis users in terms of cannabis consump-
tion, no significant differences emerged between the groups
although the ecstasy group showed a longer duration of use
in years and higher daily consumption. The age at onset of
use was also very similar between both groups. Frequency
of cannabis use in days during the last 6 months was
borderline significant (p=0.057), despite ecstasy users
showing a more frequent consumption of cannabis with
109.0 days (SD=78.6) than the cannabis group with an
average of 44.6 days (SD=65.3).
The consumption of other drugs was recorded but not
presented in Table 2. Among the 14 ecstasy users, previous
to baseline evaluation, 8 (57.1%) subjects had used cocaine
at least once, 8 (57.1%) speed, 2 (14.3%) LSD, 5 (35.7%)
ketamine, 4 (28.6%) GHB, 2 (14.3%) sedatives, and 2
(14.3%) heroin. As for the cannabis and control groups, no
one had used these drugs before.
No significant changes were observed in drug consump-
tion patterns in both user groups after 12 months compared
to baseline intake. That is, no statistical changes emerged in
the 11 ecstasy users and nine cannabis users that remained
Table 1 Sociodemographic characteristics at baseline
Ecstasy (n=14)Cannabis (n=13) Control (n=22)F (df)a
p value
Age
Males, n (%)
Years of education
University degreeb, n (%)
Employment status, n (%)
Student
Employed
Unemployed
Vocabulary WAIS-III (index score)
25.2 (3.3)
6 (42.9)
15.4 (3.1)
10 (71.4)
25.1 (2.9)
5 (38.5)
16.2 (2.1)
9 (75.0)
24.3 (3.0)
7 (31.8)
16.8 (2.0)
22 (100)
0.458 (2) 0.636
0.808
0.268
0.015
0.022
1.359 (2)
2 (14.3)
9 (64.3)
3 (21.7)
12.2 (2.3)
6 (50.0)
6 (50.0)
14 (63.6)
6 (27.3)
2 (9.1)
14.7 (2.3) 13.9 (2.7)4.485 (2)0.017c
Results are presented as the mean (standard deviation) for continuous variables and absolute frequency (relative frequency) for categorical
variables. p values result from either F test (age, years of education, WAIS-III) or Fisher’s exact test
aValue of F statistic (degrees of freedom); only in case of F test
bIncluding students
cSignificant differences between MDMA users and control subjects
430 Psychopharmacology (2008) 200:425–437

Page 7

in the study concerning ecstasy, cannabis, tobacco, and
alcohol consumption. Only the control group showed a
significant change with an increase in the alcohol frequency
of use (22.83 vs. 44.33 days during the past 6 months)
between baseline and 12-month follow-up (p=0.05). At the
12-month follow-up, reported consumption of other drugs
during the 1-year interval was the following: among the 11
ecstasy users, 6 (54.5%) subjects had used cocaine at least
once, 6 (54.5%) speed, 1 (9.1%) LSD, 2 (18.2%) ketamine,
4 (36.4%) GHB, 1 (9.1%) sedatives, and 2 (18.2%) heroin.
As for both other groups, no one had used these drugs
during the 12 months.
Neurocognitive performance: cross-sectional
and longitudinal analysis
The results of the neuropsychological tests for the groups
are shown in Table 3. ANCOVA revealed no significant
overall group differences at baseline with respect to
neuropsychological performance. All three groups per-
formed within the normal range at baseline according to
published tests norms, although ecstasy users showed
poorer results compared to cannabis and control groups in
measures of attention, memory, and executive functions,
but did not differ at a significant level.
The results of the longitudinal analysis are also shown in
Table 3. After 1 year of follow-up, scores among all three
groups remained within the normal range. However, an
overall trend to poorer cognitive functioning is observed in
ecstasy users. Linear mixed models revealed significant
overall group differences in neuropsychological perfor-
mance. Pairwise comparisons indicate that these differences
are found between ecstasy users and controls but not
between ecstasy and cannabis groups. Ecstasy users
showed significantly poorer results than drug-naïve controls
with respect to semantic word fluency (p=0.021), verbal
memory recognition (p=0.018), and information processing
speed (p=0.048). Differences were also observed in verbal
memory delayed recall, although this variable failed to
reach statistical significance (p=0.055).
Univariate correlation analyses were carried out between
cognitive measures with ecstasy and cannabis intake (lifetime
consumptionandrecentfrequencyofuse)inordertoexplorea
link between these variables. Only a negative correlation
emerged between lifetime ecstasy use with performance in
verbal memory recognition (p=0.027) but not with cannabis
exposure, suggesting a poorer performance in verbal
recognition associated to lifetime intake of ecstasy tablets.
Electrophysiological P3 measures: cross-sectional
and longitudinal analysis
The regression models applied did not reveal significant
overall group differences for P3 amplitude or P3 latency,
Table 2 Drug use characteristics at baseline
Ecstasy (n=14) Cannabis (n=13)Control (n=22)F (df)a
p value
Alcohol
Age at onset of use
Duration of use (years)
Total use in the last 6 months (days)
Tobacco
Current smokers, n (%)
Age at onset of use
Duration of use (years)
Cigarettes per day
Cannabis
Age at onset of use
Duration of use (years)
Total use in the last 6 months (days)
Joints per day
Ecstasy
Age at onset of use
Duration of use (years)
Lifetime consumption (tablets)
Total consumption in the last 6 months (days)
14.8 (3.5)
10.2 (2.7)
63.0 (43.6)
15.2 (1.2)
8.6 (2.4)
41.0 (54.4)
16.5 (1.4)
7.9 (2.9)
22.2 (15.6)
2.333 (2)
2.608 (2)
4.438 (2)
0.111
0.087
0.019b
12 (85.7)
14.3 (1.7)
9.4 (2.4)
9.7 (7.7)
5 (38.5)
15.0 (2.2)
10.3 (2.6)
7.1 (8.1)
7 (31.8)
18.3 (1.0)
7.3 (1.9)
3.6 (3.2)
0.005
0.005c
0.201
0.194
7.834 (2)
1.805 (2)
1.774 (2)
16.2 (3.1)
8.8 (2.0)
109.0 (78.6)
1.7 (1.1)
17.0 (1.3)
7.0 (2.6)
44.6 (65.3)
1.1 (1.1)
0.577 (1)
3.363 (1)
4.119 (1)
2.026 (1)
0.457
0.082
0.057
0.167
18.0 (4.0)
7.0 (2.4)
207.4 (151.0)
6.8 (3.7)
Results are presented as the mean (standard deviation). p values result from either t or F test. Proportions of current smokers are compared by
means of Fisher’s exact test
aValue of F/t statistic (degrees of freedom); only in the case of F/t test
bSignificant differences among ecstasy users and control subjects
cAge at onset significantly higher among control subjects
Psychopharmacology (2008) 200:425–437431

Page 8

Table 3 Neuropsychological test scores as a function of drug consumption group
Drug consumption group
Baseline evaluation
Follow-up evaluation
Ecstasy
Cannabis
Nonusers
F (df)a
p value
Ecstasy
Cannabis
Nonusers
F (df)a
p value
Attention
CalCAP: simple RT
349.3 (59.3)
318.7 (51.2)
323.0 (50.2)
0.591 (2)
0.559
345.5 (72.2)
331.4 (45.2)
336.6 (49.2)
0.787 (2)
0.462
CalCAP: choice RT-digits
421.9 (53.9)
389.8 (32.1)
396.5 (39.6)
1.862 (2)
0.169
414.8 (52.9)
387.0 (43.0)
402.4 (37.0)
1.669 (2)
0.201
CalCAP: sequential RT2
564.6 (90.3)
509.0 (116.3)
507.1 (95.0)
0.873 (2)
0.426
559.8 (82.8)
495.1 (98.4)
507.3 (118.8)
1.710 (2)
0.194
Working memory
Corsi block visual span backwards
8.8 (0.8)
8.4 (1.2)
9.1 (1.5)
0.627 (2)
0.539
8.9 (1.1)
9.8 (1.8)
9.4 (1.3)
0.446 (2)
0.643
Letter number sequencing
12.8 (1.9)
12.2 (1.6)
12.7 (1.8)
0.362 (2)
0.698
13.5 (4.1)
12.1 (1.5)
13.2 (2.5)
0.377 (2)
0.688
Memory
CVLT: total A1–A5
57.5 (7.7)
58.0 (8.7)
61.6 (7.1)
0.347 (2)
0.710
61.9 (6.9)
63.8 (5.8)
64.9 (6.6)
0.620 (2)
0.543
CVLT: immediate recall
13.4 (2.1)
13.7 (1.1)
13.8 (1.9)
0.026 (2)
0.974
13.4 (2.3)
15.3 (1.0)
14.9 (1.6)
1.369 (2)
0.266
CVLT: delayed recall
11.9 (4.8)
14.2 (1.4)
14.3 (1.5)
1.887 (2)
0.168
14.2 (2.3)
15.6 (0.5)
15.2 (1.4)
3.122 (2)
0.055
CVLT: total recognition
15.2 (1.0)
15.1 (1.4)
15.7 (0.7)
1.018 (2)
0.372
15.0 (1.2)
15.8 (0.5)
15.9 (0.2)
4.431 (2)
0.018b
ROCFT: immediate recall
24.2 (5.4)
25.4 (4.7)
28.1 (4.2)
1.319 (2)
0.279
25.7 (3.4)
28.5 (2.2)
27.2 (4.6)
1.657 (2)
0.204
ROCFT: delayed recall
23.9 (5.6)
26.9 (4.4)
27.1 (3.8)
0.797 (2)
0.458
25.2 (2.4)
28.3 (4.1)
26.8 (5.0)
2.086 (2)
0.138
Executive functions
Semantic word fluency
23.9 (6.2)
27.3 (5.4)
27.7 (5.5)
0.958 (2)
0.393
22.8 (4.8)
24.9 (5.0)
28.6 (5.2)
4.259 (2)
0.021b
ToL: total movements
66.3 (7.1)
74.9 (9.3)
70.6 (11.1)
2.128 (2)
0.133
64.6 (6.5)
64.0 (5.7)
64.9 (9.9)
2.222 (2)
0.122
ToL: initiation time
53.0 (25.4)
49.4 (26.2)
52.4 (24.2)
0.122 (2)
0.886
79.8 (29.0)
72.0 (28.7)
58.7 (34.1)
0.079 (2)
0.924
Mental processing speed
SDMT: total correct
64.8 (12.8)
70.6 (13.1)
72.3 (8.6)
0.253 (2)
0.778
64.5 (8.9)
71.4 (13.9)
75.0 (8.3)
3.287 (2)
0.048b
SDMT: total incorrect
0.8 (1.1)
0.7 (1.3)
0.5 (1.1)
0.078 (2)
0.925
1.0 (1.2)
0.8 (1.2)
0.4 (0.7)
1.022 (2)
0.369
P300
LP300
306.8 (18.8)
306.3 (20.5)
301.4 (18.7)
0.928 (2)
0.404
311.6 (15.0)
308.8 (20.4)
306.0 (20.4)
0.353 (2)
0.705
AP300
8.1 (2.4)
9.5 (2.7)
10.3 (2.5)
1.155 (2)
0.325
8.8 (2.7)
10.1 (2.8)
9.9 (3.1)
0.755 (2)
0.477
Results are presented as the mean (standard deviation). p values result from ANCOVA models (baseline evaluation) adjusting for gender and premorbid intelligence and linear mixed models
(baseline until follow-up) adjusting for gender, evaluation time (baseline, follow-up), and premorbid intelligence, respectively
aValue of F statistic (degrees of freedom)
bSignificant differences between ecstasy users and control subjects (p=0.008 [CVLT: total recognition]; p=0.01 [semantic word fluency] and p=0.05 [SDMT]); differences between the cannabis
group and both MDMA users and controls are not significant (p>0.25 in all cases)
432Psychopharmacology (2008) 200:425–437

Page 9

neither at baseline nor after 12 months of follow-up study.
In accordance with the neuropsychological findings, all
three groups had indices within the normal range for both
P3 components.
Correlation of electrophysiological measures
with neurocognitive measures: baseline and follow-up
First, univariate correlation analyses were carried out within
ecstasy users to assess the relationship between the P3 main
components and neurocognitive functions in this group.
Second, we conducted correlation analyses between the P3
main components and ecstasy and cannabis lifetime intake
to explore the link between these variables. Here, we used
the ecstasy group for the ecstasy use–P3 correlations and
both the ecstasy and cannabis groups for the cannabis use–
P3 correlations.
Analyses between P3 main components and cognitive
functions in ecstasy users revealed significant links with P3
latency but not with amplitude. Quite strong associations
were found at baseline between P3 latency and verbal
working memory, word fluency, divided attention, and
verbal memory recognition. Negative correlations emerged
between P3 latency and verbal working memory (r=
−0.674, p=0.016) and semantic word fluency (r=-0.629,
p=0.028), whereas positive associations were found with
divided attention (r=0.732, p=0.007) and memory recog-
nition for verbal information (r=0.657, p=0.039), both in
the unexpected direction. Therefore, improved neural
processing speed would be related to a higher achievement
in executive functions, whereas increased processing timing
would be associated with higher efficiency in divided
attention and memory recognition. After 12 months of
follow-up, no significant correlations were found.
Univariate correlation analyses carried out between drug
use and P3 response failed to reveal significant associations
between ecstasy consumption and P3 main components at
baseline or after 12 months. Despite this fact, significant
associations were found for lifetime cannabis use and P3
latency at baseline (r=−0.464, p=0.003) and with borderline
significance for the follow-up period (r=−0.322, p=0.052).
These associations emerged with an unexpected trend,
relating greater long-term cannabis exposure with improved
neural processing speed. Nonetheless, a marginally signifi-
cant correlation between P3 amplitude and cannabis use was
also observed at baseline (r=−0.301, p=0.059), linking
higher lifetime cannabis use with lower P3 amplitude.
Discussion
The aims of the study were (1) to assess changes in
cognition and P3 response in current ecstasy polydrug
users, current cannabis users, and drug-naïve controls over
the course of 1 year and (2) to explore the associations
between P3 indices, cognitive performance, and lifetime
ecstasy use. Results showed significant differences between
ecstasy users and drug-free controls on neurocognitive
measures of semantic word fluency, processing speed, and
verbal memory recognition at 1-year follow-up. These
results remained significant after statistically controlling
for gender and premorbid intelligence. Furthermore, life-
time ecstasy use was negatively correlated with perfor-
mance in verbal memory recognition. However, we found
no significant differences between the ecstasy and cannabis
groups, so that the observed effects cannot be entirely
attributed to ecstasy use. Ecstasy polydrug users showed
longer P3 latency and reduced P3 amplitude than both other
groups, remaining stable along the whole study, but these
differences were negligible and nonsignificant. According
to the second aim of the study, significant correlations were
found within the ecstasy group for P3 latency with working
memory, memory recognition, word fluency, and divided
attention. We found no correlations between P3 amplitude
and cognition in ecstasy users. P3 main components were
related to cannabis cumulative use but not to ecstasy
consumption.
These results are consistent with extensive evidence of
poorer cognitive performance in tests of processing speed,
memory, and executive functions associated with heavy
ecstasy use. Impaired recognition and delayed recall are
among the more common memory deficits reported in
ecstasy users (Quednow et al. 2006) with lifetime use being
significantly associated with poorer verbal memory perfor-
mance (Parrott 2006). Moreover, long-term prospective
memory problems are also self-reported by heavy and
moderate ecstasy users related with the extent of ecstasy
use (Rodgers et al. 2003) and with nondrug factors such as
prolonged dancing or overheat self-perception (Parrott et al.
2006). In addition, word fluency and syllogistic reasoning
deficits indicate that ecstasy polydrug use is associated with
impaired higher-order executive functioning, particularly
related to poorer access to long-term memory (Fox et al.
2002; Montgomery et al. 2005) and working memory
processes (Fisk et al. 2005), respectively. Slower processing
speed in ecstasy users has been also reported in previous
studies (Croft et al. 2001b; Wareing et al. 2007) with
negative implications over several executive functions. This
latter deficit has been more closely related to MDMA use
than cannabis intake (Croft et al. 2001b). The fact that no
significant group effect was found at baseline (t1) may well
respond to the reduced sample size, since our previous
study in a larger sample of 94 participants (de Sola et al., in
press) showed fluency deficits in ecstasy users at baseline.
In this previous study, we also showed that ecstasy users
performed poorer than controls on tests of word fluency,
Psychopharmacology (2008) 200:425–437433

Page 10

working memory, and processing speed at 2-year follow-
up. Furthermore, there was a significant association
between ecstasy lifetime use and episodic memory, which
was also observed in the present study. Although we are
aware that the limited sample size is an important limitation
of this study, results from both studies agree to demonstrate
subtle cognitive deficits in ecstasy users that remain static
or further deteriorate with time. The fact that cognitive
alterations were observable after 1 and 2 years in our two
studies is consistent with previous findings, indicating that
cognitive alterations tend to persist (Gouzoulis-Mayfrank et
al. 2005; Reneman et al. 2006; Thomasius et al. 2006) or
further deteriorate across time (Zakzanis and Young 2001;
Zakzanis and Campbell 2006). However, we should note
that these alterations were subtle and subclinical, since
ecstasy users performed all tests within the normal range.
P3 ERPs elicited in all three groups fell within the
normal range across the whole study, although ecstasy
polydrug users showed a nonsignificant pattern of reduced
P3 amplitude and increased latency when compared with
the control groups. Thus, the present findings failed to
reflect neurotoxic changes induced by prolonged recrea-
tional ecstasy use on this neurophysiological index. These
results are consistent with those of Gamma et al. (2005)
who failed to show significant P3 differences between
ecstasy users and drug-free controls after covariating for
relevant confounding variables (age, education, and canna-
bis use), which were also controlled in the present study.
However, our results stand in contrast with those of Casco
et al. (2005) who found reduced P3 amplitude in ecstasy
users and Mejias et al. (2005) who found delayed latencies
for the P3b component in their ecstasy group. Several
differences related to the ecstasy group characteristics and
task design could account for discrepancies between both
studies. Noteworthy, Casco et al. examined ecstasy users
abstinent for at least 6 months, whereas Mejías et al.
assessed a mixed group including participants that were
actually taking the drug and participants abstinent for an
undetermined period of time. In contrast, the present study
analyzed performance in current users with minimum
controlled abstinence duration of 72 h. Furthermore, we
tried to account for polydrug use incorporating a cannabis
control group and to carefully assess cooccurring drug use
using detailed toxicological analyses, which is an asset
contributing to supply more reliable data about the role of
ecstasy use. Thus, our results are more representative of
nonacute residual or long-term effects of ecstasy polydrug
use. Furthermore, the type of task used was markedly
different among these three studies. Casco et al. used a
visual attention discrimination task and found robust effects
for the frontal and especially the occipital waves, which
were not recorded in the present study. In contrast, Mejias
et al. used a more ecologically sound visual test containing
emotional facial expressions, which may introduce factors
related to emotional appraisal in the P3 evaluation.
Nonetheless, more studies are warranted to consistently
establish the electrophysiological correlates of ecstasy use.
Specifically, although the oddball task is a sound index of
attentional orientation, it would be interesting to further
explore ERP components during memory or executive
function tasks that are impaired at the cognitive level.
Another related aim of the study was to explore the
relationship between P3 cortical activity and specific
cognitive domains in ecstasy users. Significant associations
emerged between P3 latency and working memory, verbal
memory recognition, word fluency, and attention with a
quite strong power of association. No association was
found with P3 amplitude. Several studies support the
negative relationship between P3 latency and cognitive
performance, reporting faster P3 latency in the prefrontal
cortex related to better working memory (Hansell et al.
2005) or increasing P3 latency linked to normal ageing and
slow down of cognitive processing (Knott et al. 2004).
Therefore, P3 latency as a measure of cognitive efficiency
and mental processing speed (Polich and Criado 2006;
Polich 2007) would improve memory subprocesses such as
short maintenance and updating and may account for a
major efficiency in this domain. This notion could certainly
be made extensive to the association relating P3 latency
with word fluency due to the finding that probably faster
neural processing timing (required to evaluate a target
stimuli) may enhance other interconnected processes such
as access to long-term memory storage. Consistent with this
view, working memory, semantic word fluency, and P3
latency would be related measures. In contrast, association
between P3 latency with memory recognition emerged but
not with the expected trend (faster latency was associated
with poorer recognition). This finding is counterintuitive
and may reflect a statistical artifact. Otherwise, this could
tentatively explain the poorer memory recognition of the
ecstasy group. In that case, it would reflect poorer cognitive
efficiency associated with increased neural processing
speed that leads to inaccurate endorsement of false alarms
(a “lure” effect). Divided attention was also found to be
positively associated with increased processing timing in
ecstasy group. Again, this finding could respond to the
engagement of compensatory attentional processes that
could slow down mental processing speed. Overall, the
relationships shown between P3 latency, memory, attention,
and executive functions could partly respond to the
widespread distribution of brain areas involved both in P3
generation (Wager and Smith 2003; Nebel et al. 2005;
Baldo et al. 2006) and these cognitive abilities. As such, the
latter associations may tentatively suggest that these
specific cognitive domains and P3 ERPs could be mediated
by common neural substrates.
434 Psychopharmacology (2008) 200:425–437

Page 11

We found no correlation between the P3 main compo-
nents and lifetime ecstasy use, similar to Gamma et al.
(2005). In contrast, correlation analyses highlighted the
contribution of cannabis long-term cumulative consumption
(number of joints) to P3 latency, although the direction of
the relationship was unexpected: higher use, faster latency.
This finding indicates that cannabis use is not an overriding
negative factor for P3 integrity in this sample of ecstasy and
cannabis users. In fact, previous studies indicate that the
couse of cannabis and ecstasy may result on dynamic or
synergistic effects of both drugs. It has been suggested that
cannabis couse may attenuate ecstasy-related neurotoxicity
due to cannabinoids’ antioxidant and hypothermic proper-
ties. This hypothesis is supported by the fact that the degree
of “neurotoxicity” observed in ecstasy polydrug users in
PETand SPECT studies is often weaker to that predicted by
animal studies with pure MDMA treatment (Parrott et al.
2007). Nevertheless, neuroprotective effects appear to be
only partial; protective doses seem much higher than
recreational doses and protective effects are still unknown
in cannabinoid-tolerant animals and, therefore, in frequent
cannabis users (Sala and Braida 2005). Turning back to
cognition, the majority of studies report that couse of both
drugs contribute to cognitive impairment, although the
nature of this association remains unclear (Parrott et al.
2007). More research is thus warranted to clarify this issue.
Besides the mentioned methodological limitations, there
are further factors inherent to this field of research in
humans which allow alternative explanations of our find-
ings, such as inaccurate retrospective self-reports of ecstasy
use or uncertainties related to the precise chemical
composition of ecstasy tablets. Reduced arousal and
restlessness associated with ecstasy user’s lifestyle and
heavy cannabis use withdrawal can also affect cognitive
and ERP measures. In this sense, the 72-h abstinence period
could have not, being in certain cases, long enough to rule
out residual effects of the substances used. Moreover, our
study is retrospective, therefore, we cannot rule out the
possibility of premorbid group differences in intellectual
skills or cognitive and brain functioning. However, IQ
cannot explain the poorer cognitive performance reported in
ecstasy users, since differences with healthy controls were
not clinically relevant and became nonsignificant at the end
of the study, and group differences remained significant
after adjusting for this variable. Possible directions for
future research on long-term ecstasy effects could focus on
prospective designs with large cohorts of young nonusers
belonging to risk groups of drug abuse, during wide life
periods (Gouzoulis-Mayfrank and Daumann 2006). This
kind of study would supply more reliable data on several
key issues: the neurotoxic impact on developmental stages,
behavioral and cognitive changes associated to drug history,
drug interactions, probable trend of chronic impairment or
reversibility in case of abstinence, and functional impact on
every daily life.
In conclusion, our results demonstrate several cognitive
subclinical deficits in chronic ecstasy polydrug users than
drug-naive controls after 1 year relating to memory, executive
functions, and information processing speed. Findings remain
significant after adjusting for gender and IQ. Cognitive
deficitsshowastablebutnotprogressivetrendafter12months
inecstasyusers.Thethree groupsdid nodiffer significantly in
P3 response, suggesting no impairment in neuroelectric brain
activation in ecstasy users. This finding was further supported
by the fact that P3 response remained stable and within the
normal range along the whole study period, although ecstasy
users showed consistently poorer latency and amplitude.
Ecstasy cumulative use was related to lower performance in
memory recognition whereas cannabis consumption was
linked to P3 latency. The overall findings may be explained
by ecstasy or ecstasy/cannabis-induced neurotoxic impair-
ment on cognitive function, but not in brain activation,
although preexisting cognitive differences cannot be ruled
out. The present group differences seem to have limited
consequences insofar as poorer cognitive performance by
ecstasy users suggests subclinical impairment due to the fact
that the results keep within standard norms and because
neuroelectricbrainactivationissimilarinallthreegroupswith
measures standing within the normal range.
Acknowledgements
Investigaciones Sanitarias (FIS-00/00777) and Plan Nacional Sobre
Drogas (INT/2012/2002) Spain, the project “Neurotossicita’ a Lungo
Termine dell’ Ecstasy” from Istituto Superiore di Sanità, Rome, Italy,
and Generalitat de Catalunya (2001SGR00407, SGR 2005).
This study was supported in part by Fondo de
Conflict of interests
interest.
The authors declare the absence of conflicts of
References
Baldo JV, Schwartz S, Wilkins D, Dronkers NF (2006) Role of frontal
versus temporal cortex in verbal fluency as revealed by voxel-based
lesion symptom mapping. J Int Neuropsychol Soc 12:896–900
Benton AL, Hamsher K deS (1976) Multilingual Aphasia Examination:
Manual of instruction. Iowa City: University of Iowa
CascoC,ForcellaM,BerettaG,GriecoA,CampanaG(2005)Long-term
effects of MDMA (ecstasy) on the human central nervous system
revealed by visual evoked potentials. Addict Biol 10:187–195
Croft RJ, Klugman A, Baldeweg T, Gruzelier JG (2001a) Electro-
physiological evidence of serotonergic impairment in long-term
MDMA (“Ecstasy”) users. Am J Psychiatr 158:1687–1692
Croft RJ, Mackay AJ, Mills AT, Gruzelier JG (2001b) The relative
contributions of ecstasy and cannabis to cognitive impairment.
Psychopharmacology 153:373–379
Daumann J, Till B, Fischermann T, Rezk M, Gouzoulis-Mayfrank E
(2006) Intensity dependence of auditory evoked dipole source
activity in polydrug ecstasy users: evidence from an 18 months
longitudinal study. J Psychopharmacol 20:236–244
Psychopharmacology (2008) 200:425–437 435

Page 12

de la Torre R, Farre M (2004) Neurotoxicity of MDMA (ecstasy): the
limitations of scaling from animals to humans. Trends Pharmacol
Sci 25:505–508
Delis DC, Kramer JH, Kaplan E, Ober BA (2000) California Verbal
LearningTest-SecondEdition(CVLT-II).SanAntonio:Psychological
Corporation
European Monitoring Centre for Drugs and Drug Addiction (2007)
OEDT Annual Report- The State of the Drugs Problem in
Europe, pp. 48–55
Fox HC, McLean A, Turner JJ, Parrott AC, Rogers R, Sahakian BJ
(2002) Neuropsychological evidence of a relatively selective
profile of temporal dysfunction in drug-free MDMA (“ecstasy”)
polydrug users. Psychopharmacology 162:203–214
Gamma A, Brandeis D, Brandeis R, Vollenweider FX (2005) The P3
in ‘ecstasy’ polydrug users during response inhibition and
execution. J Psychopharmacol 19:504–512
Gouzoulis-Mayfrank E, Daumann J (2006) Neurotoxicity of methyl-
enedioxyamphetamines (MDMA; ecstasy) in humans: how
strong is the evidence for persistent brain damage? Addiction
101:348–361
Gouzoulis-Mayfrank E, Fischermann T, Rezk M, Thimm B, Hensen
G, Daumann J (2005) Memory performance in polyvalent
MDMA (ecstasy) users who continue or discontinue MDMA
use. Drug Alcohol Depend 78:317–323
Green AR, Mechan AO, Elliott JM, O’Shea E, Colado MI (2003) The
pharmacology and clinical pharmacology of 3,4-methylenediox-
ymethamphetamine (MDMA, “ecstasy”). Pharmacol Rev
55:463–508
Fisk JE, Montgomery C, Wareing M, Murphy PN (2005) Reasoning
deficits in ecstasy (MDMA) polydrug users. Psychopharmacol-
ogy 181:550–559
Hansell NK, Wright MJ, Luciano M, Geffen GM, Geffen LB, Martin
NG (2005) Genetic covariation between event-related potential
(ERP) and behavioral non-ERP measures of working-memory,
processing speed, and IQ. Behav Genet 35:695–706
Hansenne M (2000) The p300 cognitive event-related potential. I.
theoretical and psychobiologic perspectives. Neurophysiol Clin
30:191–210
Hatzidimitriou G, McCann UD, Ricaurte GA (1999) Altered serotonin
innervation patterns in the forebrain of monkeys treated with (+/−)
3,4-methylenedioxymethamphetamine seven years previously:
factors influencing abnormal recovery. J Neurosci 19:5096–5107
Horn H, Syed N, Lanfermann H, Maurer K, Dierks T (2003) Cerebral
networks linked to the event-related potential P300. Eur Arch
Psychiatry Clin Neurosci 253:154–159
Hothorn T, Bretz F, Westfall P with contributions by Richard M.
Heiberger (2007) multcomp: simultaneous inference for general
linear hypotheses. R package version 0.992-6.
Iacono WG, McGue M (2006) Association between P3 event-related
brain potential amplitude and adolescent problem behavior.
Psychophysiology 43:465–469
Iacono WG, Carlson SR, Malone SM, McGue M (2002) P3 event-
related potential amplitude and the risk for disinhibitory disorders
in adolescent boys. Arch Gen Psychiatry 59:750–757
Kalechstein AD, De La Garza R 2nd, Mahoney JJ 3rd, Fantegrossi
WE, Newton TF (2007) MDMA use and neurocognition: a meta-
analytic review. Psychopharmacology 189:531–537
Knott V, Millar A, Dulude L, Bradford L, Alwahhabi F, Lau T, Shea
C, Wiens A (2004) Event-related potentials in young and elderly
adults during a visual spatial working memory task. Clin EEG
Neurosci 35:185–192
Lamers CT, Bechara A, Rizzo M, Ramaekers JG (2006) Cognitive
function and mood in MDMA/THC users, THC users and non-
drug using controls. J Psychopharmacol 20:302–311
Linden DE (2005) The p300: where in the brain is it produced and
what does it tell us? Neurosci 11:563–576
Lyles J, Cadet JL (2003) Methylenedioxymethamphetamine (MDMA,
ecstasy) neurotoxicity: cellular and molecular mechanisms. Brain
Res Brain Res Rev 42:155–168
McCann UD, Szabo Z, Seckin E, Rosenblatt P, Mathews WB, Ravert
HT, Dannals RF, Ricaurte GA (2005) Quantitative PET studies of
the serotonin transporter in MDMA users and controls using
[11C]McN5652 and [11C]DASB. Neuropsychopharmacology
30:1741–1750
Mejias S, Rossignol M, Debatisse D, Streel E, Servais L, Guerit JM,
Philippot P, Campanella S (2005) Event-related potentials (ERPs)
in ecstasy (MDMA) users during a visual oddball task. Biol
Psychol 69:333–352
Miller EN (1990) California Computerized Assessment Battery (Cal-
CAP): Manual. Los Angeles: Norland Software
Montgomery C, Fisk JE, Newcombe R, Murphy PN (2005) The
differential effects of ecstasy/polydrug use on executive compo-
nents: shifting, inhibition, updating and access to semantic
memory. Psychopharmacology 182:262–276
MorganMJ (2000) Ecstasy (MDMA): a review of its possible persistent
psychological effects. Psychopharmacology 152:230–248
Nebel K, Wiese H, Stude P, de Greiff A, Diener HC, Keidel M (2005)
On the neural basis of focused and divided attention. Brain Res
Cogn Brain Res 25:760–776
Parrott AC (2006) MDMA in humans: factors which affect the neuro-
psychobiological profiles of recreational ecstasy users, the integra-
tive role of bioenergetic stress. J Psychopharmacol 20:147–163
Parrott AC, Rodgers J, Buchanan T, Ling J, Heffernan T, Scholey AB
(2006) Dancing hot on Ecstasy: physical activity and thermal
comfort ratings are associated with the memory and other
psychobiological problems reported by recreational MDMA
users. Hum Psychopharmacol 21:285–298
Parrott AC, Milani RM, Gouzoulis-Mayfrank E, Daumann J (2007)
Cannabis and ecstasy/MDMA (3,4-methylenedioxymethamphet-
amine): an analysis of their neuropsychobiological interactions in
recreational users. J Neural Transm 114:959–968
Pichini S, Poudevida S, Pujadas M, Menoyo E, Pacifici R, Farré M, de
la Torre R (2006) Assessment of chronic exposure to MDMA in
a group of consumers by segmental hair analysis. Ther Drug
Monit 28:106–109
Pinheiro J, Bates D, DebRoy S, Sarkar D the R Core team (2007) nlme:
linearandnonlinearmixedeffectsmodels.Rpackageversion3.1-86
Polich J (2007) Updating P300: an integrative theory of P3a and P3b.
Clin Neurophysiol 118:2128–2148
Polich J, Criado JR (2006) Neuropsychology and neuropharmacology
of P3a and P3b. Int J Psychophysiol 60:172–185
Pujadas M, Pichini S, Poudevida S, Menoyo E, Zuccaro P, Farré M, de
la Torre R (2003) Development and validation of a gas
chromatography-mass spectrometry assay for hair analysis of
amphetamine, methamphetamine and methylenedioxy derivatives.
J Chromatogr B Analyt Technol Biomed Life Sci 798:249–255
Quednow BB, Jessen F, Kuhn KU, Maier W,Daum I, Wagner M (2006)
Memory deficits in abstinent MDMA (ecstasy) users: neuropsy-
chological evidence of frontal dysfunction. J Psychopharmacol
20:373–384
Ramier AM, Hécaen H (1970) Rôle respectif des atteintes frontales et
de la latéralisation lésionnelle dans les déficits de la fluence
verbale. Revue Neurologique 123:17–22
Reneman L, Schilt T, de Win MM, Booij J, Schmand B, van den
Brink W, Bakker O (2006) Memory function and serotonin
transporter promoter gene polymorphism in ecstasy (MDMA)
users. J Psychopharmacol 20:389–399
Rey A (1941) L´examen psychologique dans le cas d´encéphalopathie
traumatique. Archives de Psychologie 28:286–340
Rey A (1959) Test de copie et de reproduction de mémoire de figures
géométriques complexes. Les éditions du Centre de Psychologie
appliqué. Paris
436Psychopharmacology (2008) 200:425–437

Page 13

Rodgers J, Buchanan T, Scholey AB, Heffernan TM, Ling J, Parrott
AC (2003) Patterns of drug use and the influence of gender on
self-reports of memory ability in ecstasy users: a web-based
study. J Psychopharmacol 17:389–396
Sala M, Braida D (2005) Endocannabinoids and 3,4-methylenediox-
ymethamphetamine (MDMA) interaction. Pharmacol Biochem
Behav 81:407–416
Scheffel U, Szabo Z, Mathews WB, Finley PA, Dannals RF, Ravert HT,
Szabo K, Yuan J, Ricaurte GA (1998) In vivo detection of short-
and long-term MDMA neurotoxicity—a positron emission to-
mography study in the living baboon brain. Synapse 29:183–192
Shallice T (1982) Specific impairments of planning. Philosophical
Transactions of the Royal Society of London 298:199–209
Smith A (1973) Symbol Digit Modalities Test (SDMT). Manual. Los
Angeles: Western Psychological Services
Solowij N, Michie PT, Fox AM (1995) Differential impairments of
selective attention due to frequency and duration of cannabis use.
Biol Psychiatry 37:731–739
Thomasius R, Zaplova P, Petersen K, Buchert R, Andresen B, Wartberg
L, Nebeling B, Schmoldt A (2006) Mood, cognition and serotonin
transporter availability in current and former ecstasy (MDMA)
users: the longitudinal perspective. J Psychopharmacol 20:211–225
Tuchtenhagen F, Daumann J, Norra C, Gobbelé R, Becker S, Pelz S,
Sass H, Buchner H, Gouzoulis-Mayfrank E (2000) High intensity
dependence of auditory evoked dipole source activity indicates
decreased serotonergic activity in abstinent ecstasy (MDMA)
users. Neuropsychopharmacology 22:608–617
United Nations Office on Drugs and Crime (UNODC) (2006) World
Drug Report 1:143–152
Verheyden SL, Henry JA, Curran HV (2003) Acute, sub-acute and long-
termsubjectiveconsequencesof‘ecstasy’(MDMA)consumptionin
430 regular users. Hum Psychopharmacol 18:507–517
Wager TD, Smith EE (2003) Neuroimaging studies of working
memory: ameta-analysis. CognAffectBehavNeurosci 3:255–274
Wareing M, Fisk JE, Montgomery C, Murphy PN, Chandler MD
(2007) Information processing speed in ecstasy (MDMA) users.
Hum Psychopharmacol 22:81–88
Wechsler D (1977a) Wechsler Adult Intelligence Scale-III. San
Antonio: The Psychological Corporation
Wechsler D (1977b) Wechsler Memory Scale. Third edition manual.
San Antonio: The Psychological Corporation
Yoon HH, Iacono WG, Malone SM, McGue M (2006) Using the brain
P300 response to identify novel phenotypes reflecting genetic
vulnerability for adolescent substance misuse. Addict Behav
31:1067–1087
Zakzanis KK, Young DA (2001) Memory impairment in abstinent
MDMA (“ecstasy”) users: a longitudinal investigation. Neurolo-
gy 56:966–969
Zakzanis KK, Campbell Z (2006) Memory impairment in now
abstinent MDMA users and continued users: a longitudinal
follow-up. Neurology 66:740–741
Zakzanis KK, Campbell Z, Jovanovski D (2007) The neuropsychol-
ogy of ecstasy (MDMA) use: a quantitative review. Hum
Psychopharmacol 22:427–435
Psychopharmacology (2008) 200:425–437437