Introduction
The majority of human studies pertaining to MDMA or ecstasy continue to be of people reporting repeated, and often regular, ecstasy use, henceforth referred to as "ecstasy users." This literature is extensively reviewed in the IB and in all previous updates of the IB (Baggott et al. 2001; Jerome and Baggott 2003; Jerome 2004), and in other reviews (Cole and Sumnall 2003A; Green et al. 2003; Parrott 2004), including a meta-analysis of ten studies of memory in ecstasy users (Verbaten 2003). A few researchers have examined the acute effects of ecstasy in naturalistic settings (e.g. Brookhuis et al. 2004; Curran et al. 2004), and some researchers have studied sub-acute effects of MDMA (Curran et al. 2004; O'Regan et al. 2004). However, most research in this area consists of comparisons between samples of ecstasy users and people who report no or very little lifetime use of ecstasy. These studies have assessed brain structure and function, mood or psychological well-being, cognitive function and a number of other variables, such as neuroendocrine profile. Most studies seek to detect possible indicators of MDMA neurotoxicity in humans and any relationships between at least one indirect indicator of reduced serotonin function and changes in mood or cognitive function.

As stated in the IB and in all subsequent updates, studies in ecstasy users can be treated as a conservative estimate of the upper limits of risk involved in human clinical trials of MDMA. It is expected that participants enrolled in clinical trials will experience far less risks than people who repeatedly use illicit ecstasy of unknown purity and strength in uncontrolled conditions. Nevertheless, this review will examine and use studies in ecstasy users in estimating risks and benefits of participating in clinical trials of MDMA. Nearly all studies of ecstasy users are either retrospective or longitudinal, with the majority being retrospective (see Baggott et al. 2001; Jerome and Baggott 2003; Jerome 2004). Longitudinal studies have assessed people who chose to self-administer ecstasy prior to study enrollment (Daumann et al. 2004A; DAumann et al. 2004E; Zakzanis and Young 2001) or who choose to self-administer ecstasy during the course of the study (Lieb et al. 2002). To date, there remains only one genuinely prospective study of the long-term effects of MDMA in humans (Ludewig et al. 2003; Vollenweider et al. 2000). Data from this study has been presented at conferences, and there are plans for the publication of at least one report from this study (Gamma, personal communication). The researchers failed to detect reduced serotonin transporter sites or impaired cognitive function three weeks after administering 1.5 to 1.7 mg/kg MDMA.

The difficulties inherent in the majority of studies of ecstasy users have already been discussed in the IB and previous updates to the IB, and elsewhere (Baggott et al. 2001; Cole and Sumnall 2003; Green 2003; Jerome and Baggott 2003; Jerome 2004) and so will not be discussed in greater detail. Because retrospective studies make comparisons between volunteers who chose to self-administer ecstasy and volunteers who chose not do so, they are unable to eliminate the possible impact of pre-existing differences between these groups, including differences that may have led to the decision to self- administer ecstasy repeatedly. As well, a majority of these studies use small sample sizes, and up until very recently, samples were often poorly matched on self-reported use of other substances. Recent studies have used better strategies, such as employing both polydrug using and drug-naïve controls (Thomasius et al. 2003), or seeking out both ecstasy users and controls reporting very moderate use of other substances (Halpern et al. 2004). A recent investigation employing hair analysis in a sample of people who believed they had only taken ecstasy (Kalasinsky et al. 2004) suggests that even in these cases, people might have unwittingly exposed themselves to other compounds, such as MDA and amphetamines. Kalasinsky and colleagues examined hair samples from 21 Toronto- area ecstasy users. They detected MDMA in most hair samples, but they also detected the related compound MDA in the hair of 19 of 21 participants (90%) and amphetamine or methamphetamine in 12 participants (57%). These findings are significant because some researchers have hypothesized that MDA and amphetamine pose an equal or greater risk of neurotoxicity than MDMA. Because of the aforementioned methodological flaws and because a large number of studies only include individuals who repeatedly use ecstasy in uncontrolled settings with a lifetime exposure far exceeding levels in clinical trials, studies in ecstasy users do not provide a good model for predicting effects in clinical trials of MDMA.

Research published subsequent to March 2004 continues to support our initial risk/benefit analysis, and does not increase or decrease estimated risks of a few exposures to MDMA. Recent studies of psychological well-being in ecstasy users lend even more support to the existence of an association between polysubstance use, especially when intense, and self- reported psychological problems, and not a unique or specific link between ecstasy use and psychological difficulties. However, data in this area was inconclusive at the time of the IB, and so this does not reflect a great change in estimated risk. Likewise, recent studies continue to find impairment in memory and executive function (planning and decision making) reported in previous studies. Earlier and more recent studies continue to detect impairment in the same areas of cognitive function, including verbal and visual memory and various facets of executive function. The only potentially novel findings reviewed here are of a relationship between ecstasy use and polysubstance use, and detection of a positive association between intensity of ecstasy use and intensity of substance use (Butler and Montgomery 2004; Milani et al. 2004; Sumnall et al. 2004A). These findings are significant because they suggest that when researchers did not control for polysubstance use, at least some of the effects they attributed to ecstasy use may have been at least partly due to use of use of other substances. Beyond such potential qualifications in interpreting previous studies, reports appearing after the most recent update of the literature review do not alter our analysis of the risks and benefits of participating in clinical trials of MDMA.

Studies in Naturalistic Settings and Retrospective Studies
Investigators have used retrospective studies (e.g. Cohen 1995; Peroutka et al. 1988; Liester et al. 1992; Solowij and Hall 1992) to investigate the acute subjective effects of ecstasy. Some researchers have turned to naturalistic or on-site investigations (e.g. Curran et al. 2004; van Wijngaart et al. 2001), or they have assessed effects of self-administered ecstasy in the laboratory (Brookhuis et al. 2004). Findings from naturalistic and retrospective studies have generally matched findings from clinical trials, with differences discussed in the initial literature review (Baggott et al. 2001). Some researchers have studied specific effects or behaviors, such as ecstasy effects on skills related to driving motor vehicles (Brookhuis et al. 2004). In this study, discussed in the last most recent update to the IB, the researchers found impaired performance an hour after self-administration of ecstasy, and even greater impairment three to five hours later after additional consumption of ecstasy and other substances.

One naturalistic study was published subsequent to the last previous update of the literature review (Hoshi et al. 2004), and researchers in England presented data from a study in a naturalistic setting at the Proceedings of the Physiological Society (Wolff et al. 2004). Study findings in both cases are novel but not strongly related to the risks or benefits of MDMA in humans. In addition to these naturalistic studies, a retrospective study of ecstasy users in Brazil that was published in 2003 has been located and reviewed here (de Almeida et al. 2003).

Researchers assessed changes in ecstasy users' accuracy in recognizing facial expressions immediately after drug self-administration and four days afterwards (Hoshi et al. 2004). In this naturalistic study, participants matched a series of faces, presented via computer, with the appropriate emotion label (anger, disgust, fear, happiness, sadness or surprise). Hoshi and colleagues tested participants at a nightclub and again four days later. Participants were 16 people reporting use of ecstasy at the nightclub (on "Day 0") and 21 people who reported they had not used ecstasy that evening. The researchers also assessed mood, self-reported subjective effects, and pulse rate. Since seven of the 21 controls had taken ecstasy previously, this study is not a comparison of ecstasy users with non-users. Hoshi and colleagues found that people who had just taken ecstasy were more accurate than controls in detecting facial expressions of fear on Day 0, and less accurate than controls in recognizing expressions of fear on Day 4. Ecstasy intoxication was associated with less self-reported aggression and sedation, while ecstasy was sub-acutely associated with increased self-reported aggression, but the authors failed to detect any changes in depressive symptoms, as measured via Beck Depression Inventory (BDI) on either testing day. As expected from clinical trials, ecstasy intoxication was associated with elevated pulse rate. Self-reported subjective effects of ecstasy, such as increased energy and teeth-grinding, were similar to those reported in clinical trials and retrospective studies, and there was a trend for ecstasy users to report greater openness to others than controls on Day 0, and similar levels of openness to others on Day 4. Self- reported subjective effects, changes in mood and physiological effects were similar to those reported in other studies, and sub-acute effects seen on Day 4 were also similar to findings in clinical trials and retrospective studies (e.g. Liechti et al. 2001; Solowij and Hall 1992; Vollenweider et al. 1998). Findings concerning changes in sensitivity to spotting fearful expressions in others await future investigation. It is possible that shifts in attention to facial expressions are one of the factors involved in producing the entactogenic effects of MDMA, such an increase in closeness to others. On the basis of laboratory research altering tryptophan levels, the authors propose that decreased ability to spot expressions of fear four days later might be indicative of lower brain serotonin levels, but this finding is also preliminary. Though this research team often re-assesses participants seven days after drug use (see Curran and Travill 1998), they did not report assessment of this sample seven days later. It is notable that ten of 16 ecstasy users in this study reported some use of cocaine on Day 0, while only one control reported cocaine use, raising the possibility that at least some of the effects seen on either day are due to combined use of cocaine and ecstasy. It should also be noted that the researchers did not test tablets, urine or hair for presence of MDMA on Day 0. However, the effects reported by participants suggest that most of them had taken MDMA or a related entactogen. Another team of researchers in England assessed pulse rate, blood pressure, plasma levels of cortisol, arginine vasopressin and oxytocin, plasma sodium levels and urinary osmolality (number of particles in urine, an indirect measure of electrolyte content) in 51 people attending a nightclub (Wolff et al. 2004, data presented at the Proceedings of the Physiological Society). The researchers measured pulse rate immediately prior to visiting the club and an unspecified number of hours later, after clubbing. Twenty-one out of 31 participants with a positive urinary drug screen tested positive for MDMA. Only one participant reported using both alcohol and ecstasy. Like Hoshi et al, Wolff and colleagues reported elevated pulse rate in ecstasy users (though the abstract does not directly compare ecstasy user pulse rate with pulse rate people who had not used ecstasy). The researchers did not present all data on blood pressure in this abstract, but stated that all of the five participants who had blood pressure considered hypertensive had taken MDMA during the evening. Though Wolff and colleagues reported on various changes in plasma neuroendocrine hormone concentrations, the only statistically significant finding was for higher plasma oxytocin in people who had taken MDMA when compared with people who had not taken MDMA. The other non-significant findings were of elevated cortisol in all participants that was greater in people who had taken MDMA, and elevated vasopressin in people who had taken MDMA. Plasma sodium was significantly reduced immediately after MDMA, but not after using alcohol or other substances. Urinary osmolality was higher in people who had taken MDMA when compared with controls, but this difference was not statistically significant. Somewhat surprisingly, plasma sodium levels in this sample were not correlated with vasopressin levels. Wolff and colleagues compare sodium levels from the naturalistic study with levels detected in a human trial of 40 mg MDMA (Henry et al. 1999; Forsling et al. 2001). The researchers consider the data presented here and in their previous studies as a partial explanation for hyponatremia seen after ecstasy use in uncontrolled settings. Study findings do not suggest that the previous controlled study underestimated or overestimated risk of hyponatremia after MDMA. It should be noted that when hyponatremia appears after ecstasy use in uncontrolled settings, this rare but serious complication probably arises from the combined effects of excessive water consumption, vigorous exercise without electrolyte replacement, and direct pharmacological effects of MDMA or related compounds. Participants in research studies will not be exercising vigorously, and researchers investigating MDMA-assisted therapy further reduce risk of hyponatremia by restricting liquid consumption and providing participants with electrolyte-containing beverages. In addition to providing information related to risk of hyponatremia, these findings hint at a possible role for oxytocin in some of the physiological and subjective effects of MDMA. However, the researchers did not assess relationships between drug effects and plasma oxytocin levels.

Researchers in Brazil interviewed 52 regular (habitual) ecstasy users living in Sao Paolo (de Almeida et al. 2003). This study collected information about drug use parameters and habits, self-reported reasons for using ecstasy, attitudes toward ecstasy use, and desirability of the acute effects of ecstasy. Like samples of European and North American ecstasy users, Brazilian ecstasy users tended to restrict use to weekends, with the majority of participants taking ecstasy one or more times a month but no more than once a week. Most participants used the drug for dancing or socializing. Though de Almeida and colleagues did not make direct comparisons between their sample and samples of ecstasy users outside Brazil, it seems that Brazilian ecstasy users might be slightly older on average and more likely to have completed a college degree than ecstasy users in Europe or North America. Thirty-five (67.3%) participants reported that ecstasy effects were mostly positive, while 17 participants stated that effects were both positive and negative. Positive effects were similar to those reported in clinical trials and retrospective studies, and included euphoria (happiness, being at ease), increased energy, feeling tender, and increased closeness to others. Negative effects were also similar to side effects reported in retrospective studies of ecstasy users and controlled trials of MDMA, and included dry mouth, perceived elevation in heart rate and insomnia. Hence it appears that Brazilian ecstasy users take ecstasy in settings similar to those favored by European or North American ecstasy users, and that they report experiencing similar positive and negative drug effects.

Data on physiological effects gathered in both of the naturalistic studies described above are similar to findings in clinical trials and retrospective reports, and include elevated pulse and blood pressure. Subjective effects and changes in mood assessed by Hoshi and colleagues are also similar to effects seen in previous naturalistic studies performed by the same group of researchers (Curran et al. 2004) and others (Brookhuis et al. 2004), and seen in clinical trials (for example Cami et al. 2000; Grob et al. 1996; Harris et al. 2002; Lamers et al. 2003; Liechti et al. 2001; Tancer and Johanson 2003). Each naturalistic study presented some new data on the effects of ecstasy. However, degree of accuracy in detecting fearful expressions does not directly affect estimated risks or benefits of MDMA, and changes in plasma oxytocin in ecstasy users do not significantly increase estimated risk of hyponatremia for people taking part in clinical trials of MDMA. Both reports may stimulate further basic research concerning MDMA effects on social cognition (thinking about actual or imagined others) and on neuroendocrine effects of MDMA. A retrospective study of ecstasy users in Brazil suggests that people in disparate cultures report that ecstasy produces similar positive and negative effects.

Mood and Psychological Well-Being
Findings supporting a link between ecstasy use and decline in psychological well-being, changes in mood or impulsivity were inconclusive when first reviewed in 2001 (Baggott et al. 2001). At that time, some studies reporting significant findings (e.g. Parrott et al. 2000) while others did not (e.g. Verkes et al. 2000). While there continue to be studies that find differences between ecstasy users and controls in these areas, an increasing number of them report that these differences, when apparent, are not uniquely related to ecstasy use. As early as 2001, Daumann and colleagues found that self-reported psychological problems were more closely associated with cannabis use than ecstasy use (Daumann et al. 2001). Findings from a more recent longitudinal study by the same research team support this conclusion (Daumann et al. 2004). One study reported that former ecstasy users, but not moderate or heavy current users, scored higher on measures of depression (de Win et al. 2004), while another study found that while both current and former ecstasy users had higher depression scores than non-drug users, only current ecstasy users had higher depression scores than polydrug user controls (Roiser and Sahakian 2003). A third study detected higher anxiety scores in both current and former users, but only when compared with non-drug users, and not when compared with polydrug users (Thomasius et al. 2003). Recent studies also continued to find that changes in mood and psychological well-being were more strongly related to substance use generally than to ecstasy use (Bond et al. 2004; Daumann et al. 2004; Dafters et al. 2004). The bulk of recent research findings suggest a complex relationship between ecstasy use, drug use and changes in psychological well-being.

To date, four studies published between March 2004 and January 2005 examined mood, impulsivity or psychological well-being in ecstasy users (Butler and Montgomery 2004; Milani et al. 2004; Singer et al. 2004; Sumnall et al. 2004A). These studies tended to report significant differences in mood between people reporting ecstasy use and people not reporting any lifetime use. However, in every study, it appeared that presence and intensity of polysubstance use was equally or more strongly associated with decline in psychological well-being or increased impulsivity than ecstasy use. Two studies suggest that heavy ecstasy users are more affected than light users (Butler and Montgomery 2004; Milani et al. 2004). It should be noted that two studies found an association between intensity of ecstasy use and intensity of polysubstance use. None of the studies provide support for an increased estimate of risk of decline in psychological well-being after a limited number of exposures to MDMA, and all studies offer at least preliminary evidence that earlier findings of an association between ecstasy use and psychological problems are at least partially due to the failure to match groups for presence and intensity of substance use.

While three of four studies compared ecstasy users with non-ecstasy users, none of the studies employed matched samples or selection of samples on the basis of drug use characteristics, and in all cases participants were divided on the basis of responses to drug use questionnaires. Study participants include university undergraduates (Butler and Montgomery 2004), dance event attendees or club-goers (Singer et al. 2004; Sumnall et al. 2004A), and young residents of selected English and Italian cities (Milani et al. 2004). In most cases, participants were below the age of 30. Three of four studies focused on mood or psychological problems (Milani et al. 2004; Singer et al. 2004; Sumnall et al. 2004A), while the fourth examined impulsivity (Butler and Montgomery 2004).

Anxiety
A study comparing 42 young ecstasy users (9 tablets per year, lifetime consumption not reported) and 58 non-ecstasy recruited from the same neighborhoods in Cleveland, Ohio, failed to find increased anxiety scores on the Brief Symptom Inventory (BSI), a measure similar to, but shorter than, the more commonly used Symptom Checklist 90 (SCL90) (Singer et al. 2004). In a study of 768 young people residing in London, Manchester (UK), Rome and Padua, Milani and colleagues also failed to find higher self-reported anxiety in either light or heavy ecstasy users when compared with non-ecstasy users (Milani et al. 2004). These researchers divided their sample into six groups on the basis of drug use, and made comparisons across all groups. Light ecstasy users in this study reported a lifetime consumption of 7.26 ± 6.79 occasions, with an average dose per occasion of 1.3 ± 0.9 tablets, while heavy users reported a lifetime consumption of 321 ± 368 occasions, and an average dose of 3.2 ± 1.5 tablets. Milani and colleagues detected an interaction between gender and drug use, wherein women reporting use of alcohol and tobacco reported more anxiety, depression and somatization (feelings of physical discomfort or illness) than men with the same drug use histories. Lastly, Sumnall and colleagues reported that lifetime and weekly ecstasy use was associated with higher scores on one measure of anxiety (the Beck Anxiety Inventory, or BAI), in a sample of 100 dance event attendees probably residing in or near Liverpool, England, but not with higher scores on another anxiety measure (Sumnall et al. 2004A). However, these researchers also found that higher BAI scores were also associated with greater lifetime amphetamine use, units of alcohol consumed per week and frequency of use for a number of substances.

Depression and Symptoms of Depression
Three studies assessed self-reported depressed mood or depression symptoms (Milani et al. 2004; Singer et al. 2004; Sumnall et al. 2004A), and of those three, only one (Sumnall et al. 2004A) detected a significant association between ecstasy use and increased depressive symptoms. In their assessment of young residents of London, Manchester, Rome and Padua, Milani et al. failed to find associations between ecstasy use and increased levels of depressive symptoms. Both light (lifetime consumption of 7.26 ± 6.79 occasions) and heavy (lifetime consumption on 321 ± 368 occasions) ecstasy users did not have higher depression scores than people reporting no drug use, little drug use, or polydrug use without ecstasy use (Milani et al. 2004). However, as was the case with anxiety in this study, the researchers detected an interaction between gender and drug use history, with women who reported cannabis use or light ecstasy use having higher depression scores than men with the same drug use histories. In their comparison of young Cleveland-area residents, Singer and colleagues found only a trend for ecstasy users to have higher depression scores than non-ecstasy users, and these findings only became apparent when the researchers employed a more lenient test of significance. These researchers found that ecstasy users in their sample were more likely than non- users to report experiencing childhood physical abuse or neglect and emotional neglect, raising the possibility that pre-existing factors related to increased rates of childhood abuse might also have affected both depressive symptoms and patterns of drug use.

Lastly, Sumnall and colleagues found that frequency of ecstasy use, but not lifetime ecstasy use, was associated with higher depression scores in their sample of 100 Liverpool-area dance event attendees. People in this sample who drank more units of alcohol per week and who drank more frequently also reported more symptoms of depression.

Aggression and Hostility
Two studies that assessed aggression or hostility in ecstasy users failed to detect increased levels of either psychological problem (Milani et al. 2004; Singer et al. 2004). Milani and colleagues assessed hostility with a modified version of the SCL90R, and Singer and colleagues assessed hostility with the BSI. As well, ecstasy users assessed by Singer and colleagues did not report experiencing any increased aggression or delinquency-related problems when compared with non-ecstasy users. Previous research has sometimes found increased aggression or aggressive response to threat in ecstasy users (see Gerra et al. 2001), while other researchers have found that anger is more accessible for both polydrug users and ecstasy users when compared with population norms (Bond et al. 2004). Recent study findings do not provide evidence for a link between ecstasy use and increased anger, aggression or hostility.

Impulsivity and Risk-Taking
Only one study published subsequent to the last update of the IB directly assessed impulsivity in ecstasy users (Butler and Montgomery 2004). In this study, 254 undergraduates completed a questionnaire on drug use history and patterns of use, and the Impulsivity, Venturesomeness and Empathy scale (IVE), a measure of self-reported trait impulsivity, and 249 of the 254 participants completed the Bets16, a behavioral measure of risk taking and impulsivity designed by the researchers. Participants completing the Bets16 evaluated pairs of bets consisting of one small but guaranteed win and one larger but less certain win. 28 undergraduates in this sample reported using ecstasy on fewer than 20 occasions, 18 reported use on more than 20 occasions, 116 did not report using any illicit substance, 55 reported cannabis use only, and 37 reported polydrug use without use of ecstasy. Butler and associates found that all participants who had used at least two illicit substances had higher Impulsiveness and Venturesomeness scores. Polydrug users, light ecstasy users and heavy ecstasy users also had higher novelty-seeking scores on the Tridimensional Personality Questionnaire (TPQ) than non-drug and cannabis users. Heavy ecstasy users were significantly more likely to favor high-risk bets than non-drug users, but moderate ecstasy users were no more likely than other groups to favor risky bets. Taken together, these findings suggest that, as one might expect, people reporting polydrug use are more impulsive (acting without thinking) and venturesome (consciously or intentionally taking risks) than people who do not report polydrug use. Butler and colleagues also detected an association between heavy ecstasy use and increased risk- taking. It should be noted that heavy ecstasy users in this sample also reported using higher maximum and average doses of ecstasy per use (1 versus 1.55 tablet average dose, and 1 versus 3.3 tablets maximum dose). A number of recent studies found an association between intensity of and scope and intensity of polydrug use (Milani et al. 2004; Scholey et al. 2004; Sumnall et al. 2004A). Hence it is possible that lifetime ecstasy dose, size of dose per use, or use of other drugs may all contribute to increased risk-taking in heavy ecstasy users.

Though not referred to as a measure of impulsivity per se, Singer and colleagues reported that ecstasy users were more likely than non-ecstasy users to engage in risky sexual behavior, such as unprotected sex (Singer et al. 2004). Ecstasy users in this sample of Cleveland-area residents also differed from non-users in several respects, including use of other drugs, more self-reported life problems, and history of childhood abuse, raising the possibility that risky sexual behavior might be related to these factors as well as or in addition to ecstasy use. However, increased likelihood of sexual risk-taking may be similar in nature to the increased likelihood of making or preferring risky bets seen in undergraduates reporting heavy ecstasy use. When examined together, both studies suggest an association between ecstasy use and impulsivity, but neither strongly suggests a unique relationship.

Other Psychological Problems and Traits
In their study of 100 Liverpool-area dance event or nightclub attendees, Sumnall and colleagues assessed symptoms of obsessive-compulsive disorder (OCD) with the Padua Inventory-Revised (PI-R) and dissociative symptoms with the Dissociative Experience Scale (DES). Obsessive compulsive symptoms include frequent washing or checking, and dissociative experiences refer to cases where people feel as if they or their surroundings are not real. Ecstasy use was unrelated to reporting obsessive-compulsive symptoms, but frequency of ecstasy and amphetamine use were both associated with increased levels of dissociation. Sumnall and colleagues note that these experiences of dissociation did not trouble study participants, possibly because scale items resemble drug effects (Sumnall et al. 2004A).

Self-reported life problems, including those related to health, friends, family, school, and vocation (current or planned career or work) were assessed by Singer and colleagues (2004). The 42 ecstasy users taking part in this study (no lifetime consumption listed; 9.9 tablets in last year) indicated higher rates of problems with family, peers, vocation, social life and leisure, but they did not report having more health-related problems or increased delinquency-related problems. Ecstasy users were also more likely than non-users to have taken at least one other substance. Taken together, these findings suggest that at least in some cases, regular ecstasy use is associated with and perhaps preceded by relational and other life problems. Study findings do not conclusively demonstrate a unique link between ecstasy use and these life problems.

Concluding Remarks
Similar to previous research examining mood, personality and psychological well-being in ecstasy users and controls, research findings published between March 2004 and January 2005 remain inconclusive but tend to offer only tenuous support for a unique link between ecstasy use and decline in psychological well-being. Recent research found links between ecstasy use and anxiety and depression, but these studies did not an association between ecstasy use and aggression or hostility. One study found that heavy ecstasy use was associated with greater risk-taking, but that polydrug use, with or without ecstasy use, was associated with self-reported impulsivity and novelty-seeking. Since all of the studies used retrospective designs, none of the findings demonstrate a causal relationship between repeated ecstasy use and psychological problems, and at least some data, like that of Singer and colleagues (2004) concerning increased childhood abuse or neglect in ecstasy users, suggest that at least some differences may be due partially or wholly to pre- existing factors.

Relationships between polydrug use and ecstasy use were found in all four studies. Three of four studies indicated that ecstasy users were more likely than non-users to have used other substances besides ecstasy, and three of four studies also found that intensity of ecstasy use was associated with intensity of substance use. This means that samples of ecstasy users are more likely to contain polysubstance users than samples of non-users. An association between presence and intensity of polysubstance use and presence and intensity of ecstasy use also suggests that findings from previous studies comparing ecstasy users and non-ecstasy using controls could have been affected by other variables, such as factors leading to or supporting polydrug use.

Findings from recent research studies are generally similar to previously reported findings and continue to find that relationships between use of any one drug and a specific psychological trait, problem or behavior are complex. Very few findings implicate a unique association between ecstasy use and increased depression, anxiety, hostility, impulsivity or other psychological problems. If anything, these and other recent reports suggest that the original analysis offered in the IB overestimated the risk of decline in psychological well-being for repeated ecstasy users and for participants in clinical trials of MDMA. Relying on both earlier and more recent research studies, it appears that participants in human trials of MDMA face a very minimal risk of decline in psychological well-being.

Cognition
Researchers have investigated cognitive function in ecstasy users for over fifteen years (see Baggott et al. 2001; Krystal et al. 1992). Most studies report finding subtle but detectable impairments in executive function and memory in ecstasy users, though these results are not found in all studies. Findings concerning cognitive function in ecstasy users have been reviewed in the IB and in each successive update of the IB (Baggott et al. 2001; Jerome and Baggott 2003: Jerome 2004), and are addressed in other publications as well (Cole and Sumnall 2003A). A number of recent studies suggest that moderate ecstasy use is not associated with impaired cognition (Gouzoulis-Mayfrank 2003; Halpern et al. 2004), and two studies found that former ecstasy users fared worse than current users on some measures of memory (Curran et al. 2003; Thomasius et al. 2003). Recent research continues to detect impaired memory and executive function in ecstasy users, but does support an estimation of more than minimal impairment in cognitive function for participants in clinical trials of MDMA.

Five studies of cognitive function in ecstasy users and controls were published between March 2004 and January 2005 (Daumann et al. 2004B; Daumann et al. 2004E; Daumann et al. 2004D; Fisk et al. 2004; Wareing et al. 2004A), and one study examining the contribution of various drugs to cognitive function within a sample of polydrug users (Verdejo-Garcia et al. 2004). Comparisons of ecstasy users and controls were performed by two research teams, one in Germany (Daumann et al. 2004A; Daumann et al. 2004B; Daumann et al. 2004C) and one in England (Fisk et al. 2004; Wareing et al. 2004A). All of these studies compared the performance of ecstasy users and controls on memory tasks (Daumann et al. 2004A; Daumann et al. 2004B; Daumann et al. 2004C; Fisk et al. 2004; Wareing et al. 2004A), and two studies compared ecstasy users and controls on measures of executive function (Fisk et al. 2004; Wareing et al. 2004A). The researchers examining a sample of people enrolled in a substance abuse program measured executive function, but not memory (Verdejo-Garcia et al. 2004). Additionally, the study of self-reported and behavioral impulsivity described earlier (Butler and Montgomery 2004) could be considered a measure of executive function.

The focus in all research reviewed here was on assessing specific cognitive domains rather than on comprehensive assessments of cognitive function. Studies compared ecstasy users with cannabis users (Daumann et al. 2004B; Daumann et al. 2004C), and light polydrug users (Daumann et al. 2004D; Fisk et al. 2004; Wareing et al. 2004A), and made comparisons between continuing versus abstinent ecstasy users roughly eighteen months after a first assessment (Daumann et al. 2004D). None of the studies employed non-drug using controls. Four of five studies used retrospective designs (Daumann et al. 2004B; Daumann et al. 2004C; Fisk et al. 2004; Wareing et al. 2004A), and one was longitudinal (Daumann et al. 2004D). Though one study made comparisons over time, none of the studies measured cognitive function before and after onset of ecstasy use, rendering all studies retrospective at this level of comparison.

As has been the case with previous reviews, findings from these studies are sometimes contradictory, but suggest that ecstasy users have subtle impairments in some areas of executive function and memory, while but not in other areas of memory or executive function. Ecstasy users had lower scores in specific areas of executive function (Fisk et al. 2004; Verdejo-Garcia et al. 2004). Likewise, ecstasy users had lower scores on specific measures of memory, such as measures of visual memory (Daumann et al. 2004B) or computation span (Fisk et al. 2004; Wareing et al. 2004A), but showed little to no differences on measures of working memory (Daumann et al. 2004B; Daumann et al. 2004D; Fisk et al. 2004; Wareing et al. 2004A). A longitudinal study failed to find an association between continued ecstasy use and impaired memory (Daumann et al. 2004D) despite detecting an association between continued use and changes in brain function. The within-subjects study found an association between intensity of ecstasy use and impaired working memory (Verdejo-Garcia et al. 2004), while intensity of cocaine and cannabis use were associated with impaired executive function. None of the studies produced findings that would lead to a change in the estimated risk for study participants in trials of MDMA offered in the IB (Baggott et al. 2001). As was true of research examined in previous reviews, findings supporting impaired memory in ecstasy users continue to be stronger than findings of impaired executive function.

Executive Function
Executive function refers to skills related to planning and decision-making, and is sometimes associated with impulsivity-related behaviors such as response inhibition. Approximately a quarter to a half of studies of executive function in ecstasy users have detected some impairment in this area (Baggott et al. 2001; Jerome and Baggott 2002; Jerome 2004). It is possible that mixed results in this area continues to encourage further investigations into executive function. An increasing number of researchers have concluded that executive function is not unitary and may consist of at least three to four separate domains (Alting von Geusau et al. 2004; Fisk et al. 2004; Verdejo-Garcia et al. 2004). These domains are shifting or set shift, updating and inhibition. Some researchers, such as Verdejo-Garcia and colleagues, also consider working memory to be a component of executive function.

Both 42 current ecstasy users and 17 former ecstasy users who reported being abstinent for at least six months had lower scores than 31 moderate polydrug users on computation span, a measure of number-related memory and executive function, though all three groups scored similarly on a measure of verbal recall and executive function (Wareing et al. 2004A). Current users in this study reported a lifetime consumption of 553 ± 681 tablets and had last taken ecstasy 21 ± 25 days before testing, and former ecstasy users reported a lifetime consumption of 385 ± 362 tablets, and had last taken ecstasy 781 ± 616 days previous to testing. Analyses controlling for amount of alcohol, cannabis and tobacco used in the last three months continued to find lower computation span scores in current and former ecstasy users. In a study employing the same "span" tasks and a random generation task (Fisk et al. 2004), the same team of researchers reported that 44 ecstasy users with a lifetime consumption of 377 ± 343 tablets and who were abstinent from ecstasy for 76 ± 192 days had lower computation span scores than 59 moderate polydrug users. However, Fisk and colleagues failed to find any differences between ecstasy users and moderate polydrug users on the random item generation task, a commonly used measure of executive function.

Behaving impulsively can be considered a component of planning or decision-making, and so may be viewed as an aspect of executive function, possibly related to "inhibition" in the model proposed by Miyake (as cited in Fisk et al. 2004). As described in "Mood and Psychological Well-Being", undergraduates who used ecstasy on 20 or more occasions were more likely to favor risky bets than undergraduates reporting no illicit substance use (Butler and Montgomery 2004). Moderate ecstasy users, who had used ecstasy on fewer than 20 occasions, were not more likely to make risky bets than non- drug users. It is notable that all polydrug using individuals in this sample had higher scores on a measure of impulsivity and venturesomeness, but that only heavy ecstasy were more likely to favor risky bets. Study findings suggest an association between intensity of ecstasy use and impulsivity, but they also raise issues as to whether reduced inhibition in ecstasy users may be partially related to polydrug use or pre-existing factors leading to polydrug use.

Researchers in Spain investigated possible relationships between intensity of substance use and performance on tests of executive function and working memory (conceived of as a component of executive function) in 38 detoxified polydrug users enrolled in a rehabilitation program (Verdejo-Garcia et al. 2004). Intensity of use was defined as average dose per use multiplied by frequency of use per month multiplied by chronicity of use. The researchers did not provide figures for lifetime ecstasy consumption, but stated that average dose per use times frequency of use was 6.47 tablets, and duration of use was 0.8 years. The researchers failed to find any associations between intensity of ecstasy use and measures of cognitive flexibility or response inhibition, but found a relationship between intensity of ecstasy use and working memory, described below. Generalizing from these study findings should be approached with caution, as the sample was restricted to people being treated for substance abuse or dependence, and the researchers did not make any cross-group comparisons.

Studies that found ecstasy users had difficulties with one or more component of executive function did not agree on the identity of the component. In a study reviewed in the most recent update to the IB, Alting von Geusau and colleagues found that men reporting ecstasy use were impaired on "cognitive flexibility" or shifting when compared with men who did not use ecstasy (Alting von Geusau et al. 2004), though female ecstasy users were no more likely than female controls to exhibit this difficulty. These findings are somewhat similar to those reached in a study that did not divide executive function into components (Halpern et al. 2004), except that Halpern and colleagues found that both men and women who used ecstasy, particularly those reporting use on 50 or more occasions, had lower scores on a measure of shifting. By contrast, Fisk and colleagues found ecstasy users were impaired on information updating (Fisk et al. 2004). Finally, Verdejo-Garcia and colleagues found an association between intensity of ecstasy use and impaired working memory (Verdejo-Garcia et al. 2004). Except for the use of Digit Span for working memory, these studies did not share any measures in common, so it is possible that differences in measures might lie behind differences in results. Additionally, and as noted above, one of the studies did not compare across groups but instead made within-group comparisons. These findings may also reflect the need for a clearer working definition of executive function.

When examined alongside previous research, the most recent publications continue to present findings supporting a complex relationship between repeated ecstasy use and impairments in executive function. There is enough evidence supporting such a relationship, yet some studies fail to detect it, and others suggest that only certain aspects of executive function are affected by ecstasy use while others are not.

Working Memory
For the most part, research reviewed in the IB and both previous updates to the IB have found lower scores on measures of working memory in at least some groups of ecstasy users (Baggott et al. 2001; Jerome and Baggott 2002; Jerome 2003). However, some later publications have qualified these findings (see for instance Alting von Geusau et al. 2004; Curran et al. 2003; Halpern et al. 2004). It is notable that in the last update to the IB, researchers using Digit Span as a test of working memory were less likely to detect differences between ecstasy users and controls than researchers using other measures, such as the N-back task or Spatial Span (see Daumann et al. 2003; Hanson and Luciana 2004; Wareing et al. 2004B). Despite these clarifications, the majority of previous studies detected at least some impaired working memory in repeated ecstasy users.

Three studies examined working memory in ecstasy users (Daumann et al. 2004D; Verdejo-Garcia et al. 2004; Wareing et al. 2004A), with only two making between-group comparisons. Two of three studies did not find impaired working memory in ecstasy users, while the third reported an association between intensity of ecstasy use and impaired working memory.

Daumann and colleagues assessed working memory in a longitudinal study of ecstasy users, comparing people who reported continued ecstasy use eighteen months after baseline with people who reported abstinence from ecstasy after baseline (Daumann et al. 2004D). At baseline, both groups performed similarly on the N-back task, a measure requiring participants to delay response to a target until after observing a specified number ("N") of successive non-target presentations. Eighteen months after baseline, the nine current or "continuing" ecstasy users, who reported a lifetime consumption of 185 ± 202 tablets, did not make significantly more errors than the eight former ecstasy users (lifetime consumption = 243.8 ± 271.9 tablets). Current users responded more slowly than abstinent users on the most difficult level of the task, but they responded more quickly on less difficult levels. In another study using Digit Span and Word Span, Wareing and colleagues failed to detect significant differences in working memory task performance in 42 current ecstasy users, 17 former ecstasy users, and 31 moderate polydrug users (Wareing et al. 2004A). Yet in their study of 38 substance-free polydrug users enrolled in a drug rehabilitation clinic, Verdejo-Garcia et al. detected an association between intensity of ecstasy use and lower Digit Span scores (Verdejo-Garcia et al. 2004). Considered in the light of recent findings of an association between intensity of ecstasy use and intensity of polydrug use (e.g. Milani et al. 2004; Sumnall et al. 2004A), the association that Verdejo-Garcia and colleagues detected may be due at least in part either to intensity of polysubstance use or to pre-existing factors leading to intense substance use. When taken together, findings from these three studies offer only weak support for a relationship between repeated ecstasy use and impaired working memory.

Verbal and Visual Memory
Researchers have assessed memory in ecstasy users perhaps since the first studies in ecstasy users appeared (e.g. Gouzoulis-Mayfrank et al. 2000; Krystal et al. 1992; Morgan 1999; Rodgers et al. 2000), with lower scores detected in ecstasy users when compared with controls. One of the few longitudinal studies of ecstasy users found lower scores a year after continued ecstasy use (Zakzanis and Young 2001), with prose recall especially affected. A discussion of these studies and their limitations can be found in the IB (Baggott et al. 2001) and elsewhere, as noted earlier in this review. Researchers continue to find that ecstasy users perform less well on measures of memory than non-drug using or cannabis-using controls. Other studies have found that cannabis use is more closely associated with reductions in memory than ecstasy use (Dafters et al. 2004; Simon et al. 2002), while still others have found lower verbal memory scores in former, and not current, ecstasy users (Curran et al. 2003; Thomasius et al. 2003). A quantitative analysis of ten research studies detected an effect of ecstasy use on immediate recall, and an effect of cannabis use on delayed recall (Verbaten 2003). While the two studies that employed the strictest matching of ecstasy users and controls on use of other drugs failed to detect impaired verbal memory in current users (Thomasius et al. 2003; Halpern et al. 2004), a number of studies have found this difference, suggesting that repeated use of ecstasy is associated with impaired memory, especially impaired verbal recall.

Two studies published between March 2004 and January 2005 assessed memory in ecstasy users and controls (Daumann et al. 2004B: Daumann et al. 2004C). Both were performed by the same team of researchers in Germany. One study assessed both verbal and visual memory (Daumann et al. 2000D), and the longitudinal study described earlier assessed memory with a requiring participants to learn associations between images (faces) and verbal information (professions) (Daumann et al. 2004). Both the English and Spanish research teams assessed working memory only (Fisk et al. 2004; Verdejo-Garcia et al. 2004; Wareing et al. 2004A).

Researchers compared learning and memory task performance in 12 ecstasy users and 12 controls matched on cannabis use (Daumann et al. 2004B). Daumann and colleagues found that ecstasy users reporting a lifetime consumption of 201.7 ± 224.2 tablets who had used ecstasy 51.6 ± 56.4 days previous to testing performed similarly to 12 cannabis user controls (Daumann et al. 2004B). In another study comparing ecstasy users with cannabis users (Daumann et al. 2004C), 13 Ecstasy users reporting a lifetime consumption of 324.5 ± 416.63 tablets and 13 cannabis user controls performed the LGT- 3, a measure of memory consisting of two verbal memory tasks (pair association recall, prose recall) and two visual memory tasks (visual association task, spatial association). Ecstasy users in this study scored significantly lower than cannabis user controls on immediate recall on the visual association task, and showed a trend for scoring lower on the three memory tasks. The same study found that lifetime ecstasy consumption and average ecstasy dose per use were inversely associated with visual association scores, implying a link between intensity of ecstasy use and impaired visual memory. These findings are somewhat surprising, since in most previous studies, repeated use of ecstasy is more strongly linked with impaired verbal recall than visual recall (Baggott et al. 2001). However, some studies published after the initial review have found that ecstasy users score lower on measures of visual recall (e.g. Fox et al. 2002; Gouzoulis-Mayfrank et al. 2003). These studies continue to support the possibility that repeated exposure to ecstasy poses a risk to verbal or visual memory.

Other Functional Domains
Only one of six studies published subsequent to the most recent update of the IB examined a cognitive functional domain beyond executive function and memory (Verdejo-Garcia et al. 2004). Verdejo-Garcia and colleagues administered a test of analogical reasoning to people in a substance abuse clinic (Verdejo-Garcia et al. 2004). They found that intensity of ecstasy use was associated with impaired analogical reasoning, which involves recognizing similarities between one problem or relationship and applying that information to the analogous problem. Some researchers have reported detecting impaired reasoning in ecstasy users (McCann et al. 1999; Gouzoulis-Mayfrank et al. 2000), while others have failed to find general differences in reasoning (Bhattachary and Powell 2001; Dafters et al. 2003), and many studies match participants on the basis of estimated verbal IQ (see for example Curran et al. 2003; Morgan 1999; Reneman et al. 2001B). It is also notable that rather than sampling from dance event attendees, undergraduates, or through word of mouth, Verdejo-Garcia and colleagues assessed people enrolled in a drug abuse treatment clinic. Hence it is possible that findings from this sample may not be readily generalizable to ecstasy users not diagnosed with a substance abuse problem. There are very few findings assessing reasoning apart from IQ in ecstasy users. Given the paucity of study findings relating specifically to reasoning in ecstasy users and the unusual nature of this sample, it is difficult to interpret these findings.

Concluding Remarks
Recent research on cognitive function in ecstasy users continued to find selective impairments in memory and executive function. Recent study findings offer a stronger case for impairment in verbal and visual recall than for impairment in working memory or executive function, but impairments in all three areas are detected in at least one study. Because most researchers employed retrospective study designs, these findings do not offer clear support for a causal relationship between ecstasy use and impaired memory or executive function. However, it is notable that studies that matched controls and ecstasy users on use of other substances still found that ecstasy users have lower scores on measures of memory. These findings are not cause for revising the estimated risk of impaired memory from exposure to MDMA in controlled settings. As noted in previous reviews, evidence supports the existence of such a risk, but an examination of study findings also suggests that this risk is minimal for people enrolled in clinical trials of MDMA.

Functional and Structural Imaging Studies
After noting that signs of damage to brain serotonin axons appeared in animals after repeated doses of MDMA, researchers have used various imaging methods to detect similar effects in humans. Early studies used PET to measure number of serotonin transporter sites with radioactive compounds, or ligands (McCann et al. 1998; Reneman et al. 2001A; Semple et al. 1999). Other studies have sought to detect signs of brain injury or stress by measuring levels of compounds associated with neuronal injury (Chang et al. 1999; Obergriesser et al. 2001; Reneman et al. 2002B), or by measuring numbers of 5HT2A receptors with radioactively labeled 5HT2A receptor antagonists (Reneman et al. 2002A). Previous imaging studies have been reviewed in the IB and in subsequent updates to the IB (Baggott et al. 2001; Jerome and Baggott 2003; Jerome 2004). To date, PET imaging studies have consistently found fewer serotonin transporter sites in the brains of current ecstasy users, while results of studies measuring substances associated with brain injury have been less consistent (Chang et al. 1999; Obergriesser et al. 2001; Reneman et al. 2001C; Reneman et al. 2002B). Earlier reviews have also noted a dissociation between findings from imaging studies and findings from studies of cognitive function in ecstasy users. For instance, Reneman and colleagues found lower levels of serotonin transporter in women reporting ecstasy use, and not in men, while this research team never detected gender differences in cognitive function (compare Reneman et al. 2001A and Reneman et al. 2001B). Imaging studies reviewed in the two most recent updates to the IB indicate a risk of reduced serotonin system function after repeated ecstasy use that appears to resolve after prolonged abstinence (Buchert et al. 2003; Reneman et al. 2001A).

Four imaging studies comparing the brains of ecstasy users with those of controls have been published between March 2004 and January 2005. These include two structural imaging studies (Buchert et al. 2004; Daumann et al. 2004C), and two functional magnetic resonance imaging (fMRI) studies, both performed by the same research team (Daumann et al. 2004B; Daumann et al. 2004D). Imaging and assessments of cognitive function were made in all four samples, though one research team chose to publish study results for the two assessments separately (Buchert et al. 2003; Buchert et al. 2004; Thomasius et al. 2003). One structural imaging study used voxel-based PET (Buchert et al. 2004) and the other used MRS, and each of the fMRI studies measured brain function during different types of cognitive task. For the most part, findings reported in these studies were similar to at least some previously reported findings.

Buchert and colleagues used voxel-based imaging to re-examine PET scans made with the radioligand McN5652 in 30 current ecstasy users, 29 former ecstasy users, 29 polydrug user controls and 29 non-drug user controls (Buchert et al. 2004). Previously, they had published an analysis of these scans using regions of interest (Buchert et al. 2003). Current ecstasy users in this study reported a lifetime consumption of 831 ± 1269 tablets, and were abstinent from ecstasy for 25 ± 15 days, former users reported a lifetime consumption of 793 ± 677 tablets and were abstinent for 520 ± 486 days, and samples were matched for gender and use of other substances. As was true of their initial analysis (Buchert et al. 2003), the researchers found that current ecstasy users, but not former ecstasy users, had lower numbers of serotonin transporter sites. The re-analysis detected lower serotonin transporter sites in a wider number of areas than detected in the previous study, including cingulate as well as caudate and thalamus. When scans of current users were compared with scans of same-gender polydrug users, the researchers discovered that female ecstasy users had fewer serotonin transporter sites than polydrug using women, while male ecstasy users had very few areas of lower serotonin transporter sites when compared with polydrug using men. The researchers also found that ecstasy using women had wider areas with lower numbers of serotonin transporter sites when compared with polydrug using women. These findings are similar to results of a PET imaging study that used a different ligand, Beta-CIT (Reneman et al. 2001A). Buchert and colleagues' voxel-based analysis found slightly lower numbers of serotonin transporter sites in ecstasy users than their initial analysis (Buchert et al. 2003), but they still found comparably small changes in serotonin transporter sites when compared with the first study performed with the ligand McN5652 (McCann et al. 1998). It is notable that to date, very few studies of cognitive function in ecstasy users have detected any interactions between gender and ecstasy use, and when detected, the findings are of greater impairments in men, and not women (Alting von Geusau et al. 2004; Bolla et al. 1998). There are probably many reasons for this dissociation brain serotonin transporter levels and cognitive functioning in ecstasy users, including imprecision in both types of measurement. However, such a mismatch between the two types of finding indicates that assessments of memory or executive function cannot serve as an indirect measure of brain serotonin transporter sites.

Daumann and colleagues measured levels of compounds associated with neurons in order to detect neuronal injury in the brains of 13 ecstasy users (lifetime consumption of 324.54 ± 416.63 tablets) and 13 cannabis user controls with proton magnetic resonance spectroscopy (1 H-MRS) (Daumann et al. 2004C). Participants in this study also completed measures of verbal and visual memory, as described above in "Cognitive Function." Daumann and colleagues expected to find lower ratios of one compound (N- acetylaspartate, or NA) to another (creatine, or Cr) in the hippocampus. The researchers detected a trend for ecstasy users to have lower left hippocampal NA/Cr ratios, but this trend did not reach significance. Levels of NA and choline (Cho) in frontal brain areas were found to be similar across the two groups. Lifetime consumption of ecstasy and average dose per use were associated with impaired immediate visual recall (see "Verbal and Visual Memory.") However, the researchers failed to detect an association between any of the ecstasy use parameters they measured and reduced hippocampal NA/Cr ratios. Because ecstasy users in this study were more likely than controls to report amphetamine use, it is possible that lower NA/Cr ratios are associated with use of this psychostimulant. An earlier publication using MRS failed to find any signs of decreased NA in the brains of ecstasy users (Chang et al.1999) while another study detected decreased frontal cortex NA in ecstasy users (Reneman et al. 2002B). Variance in intensity of ecstasy use and use of other substances across samples, as well general variance, might at least partially explain mixed results from MRS studies. The findings of Daumann and colleagues support a dissociation between results of their imaging study and assessment of cognitive function, so that measures of NA/Cr ratios cannot serve as indirect evidence of functional impairment.

As well as conducting the MRS study described above, Daumann and colleagues also performed two functional imaging studies (Daumann et al. 2004B; Daumann et al. 2004D). In one study, brain function was compared in ecstasy users and cannabis user controls while they performed a learning and memory task, and a control task that did not involve learning or memory (Daumann et al. 2004B). Twelve ecstasy users reporting a lifetime consumption of 201.7 ± 224.2 tablets had lower left hippocampal activity when retrieving associations than cannabis users, and they did not show the bilateral increase in hippocampal activity usually seen during retrieval (Daumann et al. 2004B). However, there were no detectable associations between parameters of ecstasy use, such as lifetime consumption or average dose per use, and differences in brain activity. These findings share similarities with the MRS findings described earlier, as both affect the left hippocampus and neither is associated with impaired cognitive function. In another functional imaging study, Daumann and colleagues measured the brains of ecstasy users and controls while they performed the N-back task described above in "Working Memory." When performing the most difficult block of the n-back task, nine current users reporting abstinent from ecstasy for 42.5 ± 32.8 days prior to testing showed a greater increase in parietal brain activity when compared with 8 former ecstasy users who reported being abstinent over a year ago (Daumann et al. 2004D). Average dose per use across the entire period of this longitudinal study was associated with greater parietal activity during the N-back task in current ecstasy users. Daumann and colleagues interpreted their findings as demonstrating that ecstasy use produced long-term effects that grew with continued use and that did not vanish after cessation of use. However, these long-term effects were not linked to changes in working memory. It is interesting that both functional imaging studies failed to find associations between changes in brain activity and scores on memory or working memory tasks. These findings suggest that cognitive function cannot be treated as an indirect measure of differences in patterns of brain function. Since functional imaging studies were not performed alongside scans of serotonin transporter sites, it is difficult to establish whether differences in brain activity are associated with reduced serotonin system function.

Concluding Remarks
It appears that studies comparing current ecstasy users with former users and polydrug users continue to find a slight but significant decrease in numbers of serotonin transporter sites in current, but not former, ecstasy users. To date, two studies performed on different samples and using different ligands reported that women who used ecstasy had a greater reduction in serotonin transporter sites than men (Buchert et al. 2004; Reneman et al. 2001A). The most recent investigation of compounds associated with neuronal injury (Daumann et al. 2004C) adds to the collection of inconclusive findings in this area. One functional imaging study detected differences in the brain activity of ecstasy users and cannabis users during working memory tasks, but found very few differences in task performance (Daumann et al. 2004B). Likewise, the other functional imaging study found differences in brain activity in current and former ecstasy users (Daumann et al. 2004D) without finding impaired task performance in current users. Taken together, these study findings continue to suggest that repeated use of ecstasy may produce changes in the serotonin system or in brain function. However, the apparent dissociation between imaging data and scores on measures of memory raise questions about the relationship between these measures. Participants in all of the imaging studies reported extensive use of ecstasy. Moderate users do not seem to exhibit the same changes in serotonin system function (Reneman et al. 2001A) or in cognitive function (Gouzoulis-Mayfrank et al. 2003; Halpern et al. 2004). This suggests that the risk of incurring either damage to serotonin axons or changes in brain function is minimal after a small number of exposures to MDMA during a clinical trial.

Overall Conclusion
A thorough examination of studies in ecstasy users published in the last year did not provide any cause for revising the estimated risks or benefits to study participants in clinical trials of MDMA. As was the case of research reviewed in 2003, recent studies offer stronger support for impaired memory than for decline in psychological well-being in ecstasy users. Imaging studies continue to offer inconclusive support for long-term effects of ecstasy on brain structure and function, with data from PET scans detecting reduced numbers of serotonin transporter sites, particularly in women, and functional imaging studies finding differences in brain activity that are not reflective of changed task performance. None of the recently reported study findings increase the degree of estimated risk, and findings failing to support a unique link between ecstasy use and decline in psychological well-being do not significantly reduce this risk. Since lifetime ecstasy consumption in these studies far exceeds the number of doses employed in clinical trials and since at least two recent studies indicate that moderate ecstasy users do not show reduced serotonin transporters or impaired memory or executive function (Gouzoulis-Mayfrank et al. 2004; Halpern et al. 2004; Reneman et al. 2001B), study participants in trials of MDMA should experience minimal risk of these changes. It remains true that study participants in clinical trials of MDMA may face several risks related to psychological well-being, cognitive function and brain function, but that these risks remain minimal.