Ecstasy may damage brain cells, study finds

Ecstasy may damage brain cells, study finds
Last Updated: Tuesday, May 18, 2010 | 7:14 PM ET
CBC News

Commentary on paper is included below

The brains of ecstasy users show low levels of a certain protein, a finding that may explain why many feel they need to turn to higher doses of the drug.

In Tuesday’s issue of the journal Brain, Stephen Kish of the Centre for Addiction and Mental Health in Toronto and his colleagues confirmed that low levels of the serotonin transporter, or SERT, are found in part of the brain of chronic ecstasy users who typically take about two tablets twice a month.

The transporter is a protein responsible for regulating levels of serotonin, a neurotransmitter important for mood and impulse control.

“It is possible that the low levels of the transporter means that there’s actual brain damage, loss of serotonin neurons in the brain,” Kish said. “But we can’t prove that.”

Ecstasy interacts with the transporter to cause the release of serotonin, which may explain some of the “friendly effect” of increased sociability that users report, the researchers said.

Since the transporter is needed for ecstasy to act, the low transporter levels might also explain why most ecstasy users in the study reported the first dose is the best for getting high, but the effect begins to decline and higher doses are needed.

“The need for higher doses, possibly caused by low SERT, could well increase the risk of harm caused by this stimulant drug,” Kish said.

About 84 per cent of ecstasy users in the study showed some degree of drug tolerance. Participants were typically 26 years old when their brains were scanned and had been using the drug for four years when recruited for the study.

Using the same probe developed in Toronto, researchers at Johns Hopkins University in Baltimore also reported that SERT levels are low in the cerebral cortex of chronic ecstasy users.

For 12 years, scientists worldwide have struggled with conflicting brain-scanning results on how ecstasy might harm brain cells that use serotonin. The new independent replication makes the finding more believable, Kish said.

The study looked at 49 drug users and 50 control subjects. Hair analysis confirmed that ecstasy users actually used the drug, and an imaging probe measured SERT throughout the brain.

The hair analysis also showed many ecstasy users were also using methamphetamine, perhaps unknowingly. The meth itself might damage serotonin cells, the researchers said. MRI scans showed people who used both drugs showed a slightly thinner cerebral cortex.

But low SERT was found in both those who used ecstasy and meth and ecstasy alone.

Those with low SERT also showed a slight reduction in memory performance.


Kish SJ, Lerch J, Furukawa Y, Tong J, McCluskey T, Wilkins D, Houle S, Meyer J, Mundo E, Wilson AA, Rusjan PM, Saint-Cyr JA, Guttman M, Collins DL, Shapiro C, Warsh JJ, Boileau I (2010) Decreased cerebral cortical serotonin transporter binding in ecstasy users: a positron emission tomography/[11C]DASB and structural brain imaging study. Brain; Published online May, 2010.

Review and comments
Ilsa Jerome Ph.D

Kish and colleagues have made a recent addition to the literature measuring serotonin uptake sites via imaging in ecstasy users with a report presented in advance of publication in the journal Brain. The report describes findings from a retrospective study in 49 ecstasy users (28 men/21 women) and 50 cannabis using and non-drug using controls (25 men/25 women). Though the paper is chiefly an imaging study, the authors addressed a number of confounds through assessing participant psychological function, self-reported sleep habits, serotonin transporter genotype, levels of testosterone and estradiol, estimated IQ and education, and cognitive function was assessed with a number of tests, including learning and recalling a list of words (CVLT) and solving arithmetic problems given at a set pace via recording. The authors found evidence of reduced SERT presence in cortex and hippocampus, but not in the midbrain or subcortical areas. Neurotoxicity affecting serotonin axons is one of several possible explanations offered for their findings. The authors also reported finding a modest but detectable impairment in cognitive function in ecstasy users. They reported that people reporting average doses per use that were higher than the sample average were more likely to have problems with a verbal memory task. This report is one of a very few (e.g. Buchert et al. 2004; Semple et al. 1999) attempting to confirm self-reported drug use with drug detection in hair, a method permitting assessment of drug use over an extended period of time. Unlike previous studies, this one employs sizeable samples of ecstasy users and controls.

The study does not contain any novel findings, possibly with the exception of their structural imaging tests. Rather, their findings are in line with some recent imaging studies (as McCann et al. 2008) and not others (as Selvaraj et al. 2009). Some other findings reported, as finding no significant differences in reproductive hormone levels, have been reported in previous studies (Allott et al. 2008). Kish and colleagues found relationships between estimated SERT in areas of insula and hippocampus, and taking more than the usual typical ecstasy dose was associate with lower scores on the verbal memory test. Kish and colleagues indicated that there was a lot of between-group overlap in cognitive function and that differences in cognitive function in ecstasy users and controsl were modest.

The biggest problem with the study is failure to use an appropriately matched set of controls for use of other drugs; ecstasy users even used more cannabis than controls. Participants In addition, the study was retrospective rather than prospective, meaning that all participants were assessed after they had chosen to self-administer ecstasy. It is notable that ecstasy users had lower estimated IQ, intended as a measure of cognitive function prior to or beyond ecstasy use, and ecstasy users may have performed less well on tests of cognitive function owing to this differences. Hence while this study is well-designed at some points, the results do not allow firm conclusions about whether ecstasy use or some other factor produced a reduction in SERT sites or impaired memory.

The investigators recruited Toronto-area residents aged 18-40 years old via advertisements. All participants had to be without any major medical illness and without a history of psychiatric illness. However, ecstasy users who reported developing a psychiatric disorder after starting to use ecstasy could enroll in the study. Controls had to report no use of any illicit drugs save cannabis. Eight of 50 controls confirmed cannabis use, 18 self-reported some use). The study is chiefly an imaging study, comparing SERT using DASB. Ecstasy user participants were enrolled if they self-reported ecstasy use and if their report was confirmed via detection of MDMA in analysis of hair. On average, ecstasy users reported lifetime consumption of 206 +/- 31 pills, with cumulative use ranging from 4 to over 900 pills. Participants reported using an average of 2.2 tablets per session (r
ange = 0.3-10 tablets) and a maximum of 4.4 tablets (range = 1-20). Tine since last use averaged at 45.2 +/- 4.7 days, and ranged from 11-194 days).

Kish and colleagues measured a large number of parameters, including serotonin transporter genes, testosterone and estradiol levels, sleep, and screened on psychiatric disorders, excluding anyone with a history of psychiatric disorders but accepting ecstasy users whose diagnosis occurred after onset of use. The team failed to find any effects of ecstasy use on reproductive hormone levels (e.g. testosterone, estradiol), and they did not find any differences in variations of the gene for the SERT between ecstasy users and controls. Much of the paper details imaging findings, and the discussion lists the major finding as a mild to marked decrease in SERT binding in cerebral cortex and hippocampus, but not in SERT-rich striatum. Kish offers this could be due to decreased concentration of SERT in intact neurons or to decrease in SERT-containing neuron axons or to a premorbid difference. There was no relationship between SERT binding levels and depressive symptoms. The study also included an author-created questionnaire instrument assessing self-reported acute, subacute (short term) and long-term effects. (The data is as you would expect in the light of previous studies of acute and subacute effects, e.g. Huxster et al. 2008; Parrott and Lasky 1998; Pirona and Morgan 2010), though decreasing shyness is apparently the most common long-term effect listed (in 65% of respondents). IQ scores were at the normal range for the sample, though ecstasy users had lower overall IQ. Kish and colleagues found statistically significant performance on the CVLT but not two other memory tests. They found executive function test differences for Trails B, and PASAT (both trials). Using correlational analysis, found a that DASB assessment of SERT in hippocampus and insula related to CVLT short-delay). They also a relationship between reporting use of more than one or two tablets per use had greatest CVLT short-delay impairment. In correlational analysis of cognitive function task scores in ecstasy users, Kish and colleagues found that few other drugs were associated with performance differences, though use of methamphetamine was associated with more recall errors on the short-delay score for the CVLT, and use of cocaine was associated with slower performance on another measure, the Trails test. These findings are similar to recent findings of differences in ecstasy users. Differences are weaker when comparing ecstasy users with polydrug user controls (for example. Bedi and Redman 2008; Hanson and Luciana 2010).

Other researchers have done DASB studies and genotyping and even tried associating imaging results and cognitive function. In that sense, the paper is not reporting anything startling. What is impressive instead is the attempt to account for multiple potential confounds and especially the attempt to verify drug use through assessing hair and urine. Hence for instance all ecstasy users had to have detectable MDMA in their hair. As well, the study recruited relatively large numbers of participants and thus should be less susceptible to problems with small samples. This report offers stronger evidence for a reduction in SERT in ecstasy users than previous reports as it is more thorough in attempting to exclude potential confounds to imaging interpretation. However, because the study is retrospective and because ecstasy users reported far more extensive use of other substances than controls, there are several possible explanations for the differences in SERT site levels and cognitive function in the two samples.

Recently Kish and colleagues published a study chiefly assessing serotonin transporter sites in ecstasy users and controls who had had not used substances other than alcohol and cannabis, reporting modest but significant differences between the two groups. The study’s authors also assessed cognitive function (as memory or planning) and measured a number of confounds, as variations in serotonin transporter genes or hormone levels. The study is retrospective rather than prospective and results do not deviate much from current research findings, but the study is well-crafted and thorough. A review and commentary on the paper by MAPS Research Specialist Ilsa Jerome Ph.D. is included with the full text of the article.