Aftereffects

The aftereffects of ecstasy use (typically at a weekend dance) have been investigated by assessing mood and cognitive function a few days after drug consumption. Early studies found a lowering of mood in both novice and experienced ecstasy users measured 2 or 5 days after use, a phenomenon that has sometimes been termed the “midweek blues.”27–29 Verheyden et al subsequently reported greater depressive symptomatology in female compared to male ecstasy users.30 In a survey of over 400 regular ecstasy users in the UK, 80% also reported impaired concentration between drug-taking sessions.31 Moreover, a controlled laboratory study of ecstasy users given MDMA during the evening and then kept awake all night found significant memory impairment the following morning that exceeded the memory impairment produced by sleep deprivation alone.32 In a later section, we will discuss evidence for persistent cognitive deficits that have been associated with repeated ecstasy use over time.

Tolerance

Many psychoactive drugs produce tolerance (ie, diminished responsiveness) upon repeated exposure. Several kinds of tolerance can occur, most notably pharmacokinetic or drug disposition tolerance due to an increased rate of drug inactivation, and pharmacodynamic tolerance due to reduced cellular sensitivity to the drug. Animal studies involving repeated MDMA administration have provided some evidence for tolerance development, although the opposite effect, which is called sensitization or reverse tolerance, occurred in some cases.15 Differences in dosing regimen and/or specific outcome measures likely account for these discrepant findings. Regular ecstasy users often report diminished responsiveness to the drug and a consequent need for dose escalation, both of which imply the development of tolerance. In a different approach, Farré et al52 and Peiró et al53 studied the pharmacodynamic and pharmacokinetic effects of administering two doses of MDMA to human subjects at an interval of either 2 or 24 hours. Both studies found a higher than expected plasma concentration of MDMA after the second dose, which is a consequence of the drug inhibiting its own metabolism. In contrast, the subjective and physiological effects produced by the second dose were lower than expected based on the plasma concentration. These results suggest that tolerance to MDMA can develop quite rapidly, even potentially within a single session of drug use.

Mood changes

The ecstasy literature is replete with reported associations between use of this substance and abnormal mood changes. Two Internet studies of ecstasy users in the UK found a high incidence of self-reported mood fluctuations, feelings of depression, increased anxiety, and heightened impulsivity.16,54 These effects were more pronounced in heavy compared to novice users. Numerous other studies have provided more systematic data by comparing ecstasy polydrug users with polydrug users who were not taking ecstasy and/or with controls who were not taking any illicit drugs.20 Two reviews of this literature that were published in 2005 and 2007 focused on depressive symptomatology assessed by means of the Hamilton Rating Scale for Depression, the Beck Depression Inventory, the Symptom Checklist 90, or the Center for Epidemiological Studies Depression Scale.55,56 Both reviews noted a mix of results, in which some studies found a statistically significant increase in depressive symptomatology in the ecstasy users compared to the reference group, whereas other studies did not. A meta-analysis performed in conjunction with one of the reviews found a statistically significant overall association of ecstasy use with increased depression, although the effect size was relatively small.56

Despite the apparent association between ecstasy use and depressed mood, there are several reasons why it is premature to conclude that exposure to MDMA causes such a mood change. First, many of the cited studies failed to control adequately for the influence of polydrug use. Indeed, there are several reports suggesting that use of other substances (including cannabis) as well as environmental factors are at least as important as ecstasy itself in the heightened depression and anxiety seen in ecstasy users.57–60 Second, in many studies, the selection of subjects may not have yielded a representative sample of ecstasy users. Third, the possibility must be considered that depressed mood precedes rather than follows the onset of ecstasy use in many individuals. The plausibility of this notion is supported by the mood-elevating effects of MDMA discussed earlier. Moreover, three different longitudinal studies found evidence that mood disturbances are often present before the onset of ecstasy use.61–63 One study, conducted in The Netherlands, found that children (mean age of approximately 10 years) with elevated ratings of anxiety or depression were twice as likely to have used ecstasy when surveyed 14 years later than children without such a rating.61 Two other studies, one conducted in Ohio, USA and the other in Munich, Germany, found that the first occurrence of a clinical psychiatric disorder, including major depression, commonly preceded rather than followed the onset of ecstasy use.62,63 These findings raise the intriguing possibility that at least some ecstasy users may be taking the drug for the purpose of “self-medication.”

Neuropsychological deficits

A large number of studies, described in this section, have compared neuropsychological functioning in current ecstasy users (polydrug users in almost all cases) with non-ecstasy polydrug users or controls that do not consume any illicit substances. In these studies, the subjects are requested to abstain from drug use for at least some minimum number of days so as to avoid confounding by acute drug effects. We will begin this section by focusing on assessments of memory, after which we will consider other aspects of neuropsychological function.

Ecstasy users have been tested on several different kinds of memory, including short-term memory (unprocessed information stored for a brief period of time, eg, immediate word recall in the Rey Auditory Verbal Learning Test); verbal working memory (briefly stored verbal information subjected to processing, eg, n-back task); visuospatial memory (working memory of visuospatial information, eg, Rey-Osterrieth Complex Figure Test); and long-term memory (information held in a longer-term store, eg, delayed recall tasks). Because the individual studies in this area have typically suffered from relatively small sample sizes, have used differing criteria for subject inclusion in the ecstasy-using and control groups, have tested the subjects on a variety of different tasks, and have yielded conflicting findings, a number of investigators have performed meta-analyses to determine the overall direction and strength of these findings. A 2007 meta-analysis of short-term, long-term, verbal, and visual memory by Laws and Kokkalis64 found statistically significant deficits in all of these types of memory except for visual memory. Effect sizes were generally moderate or large, which supports the contention that regular ecstasy use is associated with a broad range of memory deficits. A more refined 2010 meta-analysis by Nulsen et al65 differentiated between tests of short-term and working memory (verbal and visuospatial in both cases) and found that the ecstasy users performed more poorly in all memory domains. Results were significant regardless of whether the control group was composed of non-ecstasy polydrug users or individuals who had not been exposed to any illicit drugs. Moreover, estimated lifetime ecstasy consumption was related to the effect size in working (but not short-term) memory. An even more recent (2012) meta-analysis by Murphy et al66 focused on visuospatial memory tasks. Significant deficits in ecstasy users were found for tasks requiring memory of the spatial distribution of stimulus items, tasks requiring figure recognition, and tasks requiring production or reproduction of figures. In contrast to the results of Nulsen et al,65 estimated lifetime ecstasy consumption did not predict effect sizes in the analysis of visuospatial memory performance. It should be noted that, at least with respect to long-term memory, memory deficits in ecstasy users are more clearly seen in high-complexity than in low-complexity tasks.67 If this is also true for other memory domains, it may help explain some of the conflicting findings in the literature.

Fisk and Sharp68 proposed that working memory consists of a central executive function along with four subcomponents which they termed “updating,” “attention shifting,” “inhibition,” and “access to long-term memory.” A recent meta-analysis comparing ecstasy users to polydrug-using controls on these four subcomponents found significant ecstasy-associated deficits in updating, attention shifting, and access to long-term memory.69 Effect sizes were generally moderate for the three significant subcomponents, whereas the inhibition subcomponent was not significantly affected. Meta-analyses that examined aspects of cognitive function other than memory found significant impairment in attention and concentration, verbal comprehension, processing speed, and motor/psychomotor speed.70,71 Together with the reviews of memory performance, these findings suggest that regular ecstasy users suffer from widespread problems across a wide range of cognitive domains.

Despite the positive findings from a number of different analyses, the literature on cognitive deficits in ecstasy users suffers from some of the same interpretive problems as the literature on mood changes. For example, Gouzoulis-Mayfrank and Daumann19 discuss the issue that, even if ecstasy users are compared to drug users who do not consume ecstasy, the latter often have a more moderate pattern of drug consumption than the ecstasy-using group. Cannabis is particularly problematic in studies of cognitive function, given the evidence for significant cognitive deficits in heavy cannabis users.72 Comorbidity of psychopathology with ecstasy use may additionally contribute to the cognitive impairment observed in some studies.73 Finally, the typical use of cross-sectional studies again makes it impossible, in such cases, to ascertain whether cognitive differences between the ecstasy-using and control group(s) preceded or followed the onset of ecstasy use. Two prospective studies have addressed this issue by recruiting new ecstasy users, assessing their cognitive function at baseline, and then retesting the subjects from 1 to 3 years later.74,75 Interestingly, in both cases, the users did not differ from controls at baseline but did exhibit significant deficits in immediate and delayed verbal recall memory. These findings point up the limitations inherent in cross-sectional studies in which baseline mood or cognitive function prior to the onset of ecstasy use is not known. Importantly, the results suggest that, at least for some ecstasy users, repeated exposure to the drug leads to later impairment in certain cognitive domains.

Longitudinal studies of up to 2 years in length have also asked whether cognitive function keeps declining with continued ecstasy use, and whether cessation of use leads to a recovery of cognitive function. In general, such studies have found neither further deterioration in users nor recovery of function in subjects who discontinued ecstasy use76–78 (see Zakzanis and Campbell79 for an exception to these findings). Because in these studies the ecstasy users already differed from controls at the time of initial testing, the results are difficult to interpret. There may have been preexisting cognitive differences that were unchanged by ecstasy use, or the typical use patterns of the subjects in the first three studies may have caused an asymptotic decline in cognitive performance that neither got worse over time nor recovered upon 2 years of abstinence.

Neurotoxicity

The abovementioned evidence for mood changes and neuropsychological deficits in ecstasy users raises the question of how MDMA might be acting within the brain to cause these effects. The answer given most often centers around numerous findings obtained from experimental animals, mainly rats and nonhuman primates (eg, squirrel monkeys), for a toxic effect of MDMA on the serotonergic system. The first evidence for MDMA neurotoxicity was published in the mid-1980s, and these and numerous follow-up studies showed that high doses of MDMA (or MDA) lead to a long-lasting depletion of 5-HT, reduced 5-HT uptake and SERT binding sites, and decreased activity of tryptophan hydroxylase in forebrain areas such as the neocortex, hippocampus, and striatum (see Green et al33). Time-course studies demonstrated a biphasic action of MDMA consisting of the initial 5-HT depletion discussed earlier, a short-term recovery of 5-HT levels seen at 1–2 days following dosing, and then a secondary, more long-lasting depletion measured at 1–2 weeks and beyond.33 Histological examination of brain tissues from treated rats and monkeys consistently showed a persistent reduction in the density of axons immunoreactive for 5-HT or for SERT (Figure 1). Moreover, staining of brain tissues obtained at earlier time points revealed swollen axons and a positive reaction for silver staining, both of which are consistent with the notion of drug-induced axonal degeneration.33,81

Open in a separate window

Figure 1

Photomicrograph of serotonin transporter-immunoreactive axons in the upper layers of occipital cortex of a control rat (A) compared to a 3,4-methylenedioxymethamphetamine (MDMA)-treated rat (B).

Notes: The animals received twice-daily injections of either saline or MDMA (10 mg/kg per injection) from the first to the fourth day of life and then were killed at 9 months of age (see Meyer et al80 for experimental details). These results illustrate the long-lasting nature of MDMA-induced serotonergic deficits when the drug is administered early in development.

The latter observations coupled with the apparent loss of serotonergic axons stained with immunohistochemical methods quickly gave rise to the hypothesis that high doses of MDMA cause a degeneration of rostral forebrain serotonergic fibers.33,81 Nevertheless, other experimental approaches suggest that MDMA may not be causing long-lasting physical damage to the serotonergic fibers; rather, there is evidence that serotonergic neurons are responding to the drug by dramatically reducing expression of the genes for SERT and for tryptophan hydroxylase.82,83 According to this downregulation hypothesis, long-lasting reductions in gene expression can explain both the persistent depletion of 5-HT (because tryptophan hydroxylase activity is too low to sustain normal levels of this neurotransmitter) and the seeming disappearance of fibers that are immunoreactive for 5-HT or SERT (ie, the “lost” fibers are still present but are difficult to visualize due to the substantial decrease in both 5-HT and SERT protein; see Biezonski and Meyer83 for a discussion of the neurodegeneration versus downregulation hypothesis of MDMA neurotoxicity). Regardless of which hypothesis is proven correct, two important points should be noted. First, proponents of the neurodegeneration hypothesis focus on the apparent loss of serotonergic fibers in the forebrain; the cell bodies of the serotonergic neurons in the midbrain raphe nuclei appear to be spared (ie, MDMA is generally not thought to cause serotonergic nerve cell death). Second, even if forebrain serotonergic fibers are not physically damaged, as postulated by the downregulation hypothesis, long-lasting deficits in 5-HT levels and in SERT expression would still cause severe dysfunction of the serotonergic system. In this broader sense, MDMA is a serotonergic neurotoxin according to both hypotheses.

Long-term studies of MDMA-treated animals indicate that a reappearance of serotonergic fibers and 5-HT reuptake sites occurs over time; however, recovery may take many months following extremely high doses of the drug.33 Moreover, some rats and monkeys exhibited abnormal fiber patterns after recovery had occurred,33 and, in one highly cited paper from Hatzidimitriou et al, forebrain 5-HT-immunoreactive fibers were still abnormal in monkeys at 7 years after dosing.84 Such findings are remarkable in showing the potential persistence of MDMA-related disturbances in serotonergic function.

Several factors influence the extent and the characteristics of MDMA-induced neurotoxicity. The first factor is dose of the drug. Not surprisingly, the greatest degree of 5-HT depletion is produced by high (and typically multiple) doses of MDMA (eg, four injections of 10 mg/kg given to rats at 2-hour intervals in a single day). However, serotonergic deficits have been reported following single doses at lower levels. A recent example comes from Mueller et al,85 who reported that single oral doses as low as 5.7 mg/kg given to squirrel monkeys produced statistically significant reductions in 5-HT concentrations and SERT binding in many forebrain areas when measured 1 week after drug administration. Do and Schenk86 also found that intravenous self-administration of a relatively low dose of MDMA (0.5–1.0 mg/kg per infusion) by rats eventually led to 30%–35% reductions in forebrain 5-HT, though the cumulative amount of drug taken by the animals over time was 315 mg/kg. Second, there are important species differences in MDMA neurotoxicity. For example, in laboratory mice, MDMA exerts much greater deleterious effects on the dopamine system than on the serotonergic system.87 Whereas rats do exhibit some dopaminergic effects at high MDMA doses, there is relatively little actual depletion of dopamine compared to that found in mice.87,88 The reason for this species difference has not yet been determined. Third, ambient temperature can modulate the extent of MDMA-induced neurotoxicity. As mentioned earlier, at the typical temperatures found in a laboratory animal colony room, a human residence, or a dance club, MDMA elevates core body (and brain) temperature. The significance of this hyperthermic effect can be seen by the fact that, in rats maintained in a slightly cooler environment, MDMA administration led to a hypothermic instead of a hyperthermic response, and the hypothermic animals failed to exhibit reductions in brain 5-HT.89 Finally, prior exposure to low-to-moderate doses of MDMA can blunt or even prevent the neurotoxic effects of a subsequent high-dose treatment regimen.90,91 It is not yet known whether this effect, which has been termed “MDMA preconditioning,” applies to human recreational ecstasy users.

Substantial effort has been devoted to elucidating the cellular and molecular mechanisms of MDMA neurotoxicity. One of the important findings to emerge from this work is that MDMA must be administered systemically for neurotoxicity to occur. That is, microinjection of the drug directly into the brain fails to produce 5-HT depletion despite causing acute release of 5-HT.92 This finding has been interpreted by some researchers to mean that MDMA must be subjected to peripheral metabolism (eg, in the liver) to produce neurotoxicity.93,94 Consistent with that view, neurotoxic metabolites of MDMA have been detected in rat brain and in human urine after MDMA ingestion. 93,94 Other studies, however, do not support the neurotoxic metabolite hypothesis.95,96 Regardless of the exact nature of the toxic agent, it is believed that this substance is carried into the serotonergic neurons by way of the 5-HT uptake system, because blockade of SERT with an SSRI attenuates or even blocks MDMA-induced neurotoxicity.97 Once inside the serotonergic neurons, the toxic agent is thought to exert its effects by increasing the formation of oxygen-containing free radicals (ie, reactive oxygen species [ROS]), thereby producing oxidative stress on the cells. The central concept that MDMA-induced neurotoxicity is mediated by ROS is supported by studies demonstrating the formation of such species within the brains of MDMA-treated animals and the ability of antioxidants and free radical scavengers to block the drug’s neurotoxic effects.98,99 An alternate neurotoxicity model hypothesizes that DA levels are increased within the serotonergic neurons due to a combination of direct DA uptake via SERT and uptake of tyrosine with subsequent conversion to the DA precursor dihydroxyphenylalanine (DOPA) and then to DA.98,99 According to this model, metabolism of the abnormally accumulated DA by monoamine oxidase B (MAO-B) within the serotonergic cells generates hydrogen peroxide, which can then undergo further chemical reactions to produce ROS and oxidative stress. On the other hand, studies by Yuan et al100 have yielded evidence against the DA hypothesis. To summarize, oxidative stress is likely to play a key role in MDMA-induced neurotoxicity; however, the cellular and molecular events that lead to ROS formation are not yet fully understood.

Of critical importance is whether repeated use of ecstasy (or even a single dose) exerts neurotoxic effects in the human brain. Some investigators have questioned the relevance of animal (particularly rat) studies of MDMA neurotoxicity because the doses typically administered to the animals are much higher than those taken by most recreational ecstasy users and because of important differences in MDMA metabolism between rodents and humans.101–103 The latter issue limits the potential accuracy of using interspecies dose scaling to translate rodent MDMA doses to the amounts estimated to be consumed by recreational ecstasy users. On the other hand, a combination of biochemical and neuroimaging findings from the human ecstasy literature support the contention that “heavy” users may suffer from serotonergic deficits as well as other neurobiological abnormalities. Early studies found reduced levels of 5-hydroxy-indoleacetic acid (5-HIAA) in the cerebrospinal fluid of ecstasy users as well as blunted hormonal responses to various serotonergic drugs.33 More recently, Kish et al104,105 reported on a single case study of a 26-year-old man who used extremely large amounts of ecstasy (along with some cocaine and heroin) for several years before dying of an apparent drug overdose. Biochemical analyses of postmortem brain tissues revealed massive reductions in 5-HT and 5-HIAA levels in both the cortex and striatum compared to control tissues. These changes were accompanied by substantial decreases in SERT and tryptophan hydroxylase protein levels, which are further indications of serotonergic neurotoxicity in this subject.

In addition to the important information gleaned from cerebrospinal fluid and postmortem tissue analyses, researchers have made extensive use of modern neuroimaging methods (including positron emission tomography [PET], single photon emission computed tomography [SPECT], magnetic resonance imaging [MRI], diffusion tensor imaging [DTI], and magnetic resonance spectroscopy [MRS]) to probe markers of the serotonergic system along with other markers of brain structure, function, and chemistry. Such studies have consistently shown that SERT binding in various cortical as well as subcortical areas is significantly reduced in ecstasy/polydrug users compared to matched controls.78,106–111 This effect could reflect either loss of forebrain serotonergic axons (as proposed by the neurodegeneration hypothesis of MDMA neurotoxicity in experimental animals81) or an MDMA-induced downregulation of SERT expression by the serotonergic neurons (as proposed by Biezonski and Meyer82,83). Some (though not all) of these studies found an inverse relationship between SERT binding and amount of ecstasy consumed, and some also found evidence for recovery of binding following abstinence from the drug. A second common finding concerns changes in cortical 5-HT₂ receptor binding by PET or SPECT imaging. Binding was typically found to be significantly upregulated in ecstasy users,106,107,109,111,112 although a few studies found a downregulation instead.106–109 Considering the typical regulation of G protein-coupled receptors (like the 5-HT_2A receptor) found in animal studies, it is possible that initial ecstasy use causes a receptor downregulation due to excessive 5-HT release, whereas the 5-HT depletion thought to occur with chronic heavy ecstasy use leads to a compensatory receptor upregulation. Further studies are needed to substantiate this hypothesis. Importantly, there may be a causal relationship between the mood and cognitive changes observed in heavy ecstasy users and the dysregulated serotonergic transmission implied by the above mentioned 5-HT depletion and altered SERT and 5-HT_2A receptor expression, as well as potentially other serotonergic effects yet to be identified. Such a relationship has been proposed by Parrott113 and other investigators106,107,109 and is summarized in Table 2.

The use of structural and functional MRI, DTI, and MRS has revealed additional differences between ecstasy users and control subjects under some conditions. Although the results of this work cannot be detailed here due to space limitations, interested readers are referred to the appropriate references.106,107,109,110,114–117

Because animal studies of MDMA neurotoxicity have typically used large and/or repeated drug doses, we may ask whether one or a few modest doses of ecstasy are capable of exerting neurotoxic effects in users. This question has been addressed by several prospective studies of new ecstasy users participating in the Netherlands XTC Toxicity (NeXT) study. The results thus far have failed to show any serotonergic deficits in these low-dose users; however, other abnormalities were found related to brain vasculature and white matter structure. 114,115 Thus, the jury is still out on whether damaging effects can be produced by consuming even a few ecstasy tablets.