MDMA memory loss study + good commentary

Another study, with some intriguing results, substantiating the link between MDMA and long-term memory damage. The abstract is below. Also, Arch Gen Pysch also included a critique by McCann, Ricaurte, and crew, which does a nice job in my view of covering the problems and major unresolved issues with MDMA neurotoxicity.
A couple highlights:

• Marijuana as always remains an important confounding factor in these studies. Despite what you may have heard, heavy marijuana use does impair cognitive ability, though much if not all of the impairment seems to be reversible upon cessation of marijuana use.
• The observed impairment in memory was large (47 vs 60 on their test, the RAVLT) and statistically very significant. (P<0.001) Users (n=22) had taken on average about 500 pills over 5.5 years, or about two pills/week for 5.5 years. The level of memory impairment was well correlated with the total amount of MDMA taken (p<0.05, r=-0.3), but not with abstinence from MDMA nor beta-CIT binding in the brain. This suggests that the regeneration of damaged serotonergic nerves after heavy MDMA use does not alleviate MDMA-induced memory problems.
• Studying the cortex may be the wrong way to go. It is important to study other brain regions, eg the hippocampus, to see if MDMA neurotoxicity here is better correlated with functional damage.
• The serotonin transporter loss seen in humans in newer studies using radioligands/SPECT appears to be on the order of 10%, whereas animals given similar regimins of MDMA have 80%-95% loss. McCann et al suggest this could be due to inaccuracies of the radiolabel in the region being scanned. I can't judge this claim but it seems reasonable, and they cite some independent research in support of it.

Full article at AMA's website.Commentary at AMA's webiste.
Article Abstract
Reneman L, Lavalaye J, Schmand B, de Wolff FA, van den Brink W, den Heeten GJ, Booij J.
Cortical Serotonin Transporter Density and Verbal Memory in Individuals Who Stopped Using 3,4-Methylenedioxymethamphetamine (MDMA or "Ecstasy"); Preliminary Findings
Arch Gen Psych 58(10):901-906 (Oct 2001).
Background Although the popular drug 3,4-methylenedioxymethamphetamine (MDMA or "ecstasy") has been shown to damage brain serotonin (5-HT) neurons in animals, the fate and functional consequences of 5-HT neurons after MDMA injury are not known in humans. We investigated the long-term effects of MDMA use on cortical 5-HT neurons in humans and memory function, because brain 5-HT has been implicated in memory function.
Methods Twenty-two recent MDMA users, 16 ex-MDMA users who had stopped using MDMA for more than 1 year, and 13 control subjects. The effects of MDMA use on cortical 5-HT neurons was studied by means of single-photon emission computed tomography with iodine 123–labeled 2-carbomethoxy-3-(4-iodophenyl) tropane ([123I]-CIT) by quantification of brain 5-HT transporter densities. Verbal memory performance was assessed with the Rey Auditory Verbal Learning Test.
Results Mean cortical [123I]-CIT-labeled 5-HT transporter density was significantly lower in recent MDMA users than in controls (1.17 vs 1.28 [–9%]) but not in ex-MDMA users (1.24 vs 1.28 [-3%]). Recent and ex-MDMA users recalled significantly fewer words than did controls on the immediate recall (47.0 and 48.0 vs 60.0, respectively; P = .001) as well as the delayed recall (9.8 and 10.1 vs 13.1, respectively; P = .003). Greater use of MDMA was associated with greater impairment in immediate verbal memory. However, memory performance was not associated with [123I]-CIT binding to cortical 5-HT transporters or duration of abstinence from MDMA.
Conclusion The present study suggests that, while the neurotoxic effects of MDMA on 5-HT neurons in the human cortex may be reversible, the effects of MDMA on memory function may be long-lasting.
Commentary text
McCann UD, Ricaurte GA, Molliver ME
"Ecstasy" and Serotonin Neurotoxicity: New Findings Raise More Questions.
Arch Gen Psych 58(10):907-908 (Oct 2001)
THE ARTICLE by Reneman et al1 in this issue of the ARCHIVES is timely and provocative and highlights several areas of controversy in the fields of substance abuse, drug-induced neurotoxic effects, and neuroimaging. The authors present evidence that the illicit recreational drug 3,4-methylenedioxymethamphetamine (MDMA, or "ecstasy") may cause persistent cognitive deficits2-9 and that these deficits are related to the extent of previous MDMA use. Based on single-photon emission computed tomography (SPECT) imaging with 123I-labeled 2-carbomethoxy-3-(4-iodophenyl)tropane (-CIT), they conclude that MDMA causes neurotoxic injury to cortical serotonin (5-HT) axon terminals that may be reversible. This is the first study to evaluate a separate cohort of previous MDMA users who have abstained from use for longer than 1 year, and thus has the potential to provide information regarding long-term effects of exposure to MDMA.
With respect to cognitive effects of MDMA, there is some question regarding the role of concomitant marijuana use in the cognitive deficits observed in MDMA users. Reports by Rodgers10 and Croft et al11 indicate that marijuana is an important confounding factor in studies of cognitive function in MDMA users, with marijuana use predicting much of the observed cognitive decline. This is a nettlesome problem, since most MDMA users also use marijuana. In the present study, although the "ex-MDMA" group of subjects abstained from MDMA for at least 1 year, they continued substantial marijuana use, potentially accounting for diminished cognitive performance. Nevertheless, the present findings of impaired verbal memory are cause for concern and underscore the need for longitudinal studies in MDMA users to evaluate the persistence of functional deficits and to determine whether tardive adverse effects emerge. Clearly, future studies aimed at elucidating cognitive effects of MDMA use will need to control for concomitant marijuana use. Moreover, the relationship between MDMA-induced serotonin neurotoxicity and cognitive deficits in MDMA users needs further investigation and will require assessment of several brain regions (eg, hippocampus) in addition to cerebral cortex.
Data obtained using SPECT with the radioligand [123I]-CIT insert themselves into the ongoing debate regarding the suitability of this method for measuring serotonin transporter (SERT) sites in cerebral cortex. Several laboratories12-14 have demonstrated the capability of [123I]-CIT for measurement of brainstem and mid brain SERT. However, as noted by Heinz and Jones,15 there is little evidence that SPECT with [123I]-CIT can accurately measure specific binding to cortical SERT sites, with studies in a nonhuman primate demonstrating no change in the level of cortical [123I]-CIT binding following administration of the serotonin reuptake inhibitor, citalopram.12 Nevertheless, one other research group16 has also reported reductions in the binding of [123I]-CIT in occipital cortex of MDMA users. To address the question of whether MDMA-induced brain serotonin injury in cerebral cortex can be detected by SPECT with [123I]-CIT, studies using large nonhuman primates, such as baboons, with similar cortical SERT distributions and densities as humans are required, ideally before and after MDMA treatment.
An important issue regarding SPECT sensitivity deserves clarification. The authors note in the "Comment" section that reductions in cortical [123I]-CIT in MDMA users were on the order of 9%, while previous binding studies in nonhuman primates17 given doses of MDMA similar to those used by humans exhibit far greater deficits in cortical 5-HT axonal markers, ranging from 83% to 95%. This disparity leads the authors to conclude that MDMA is less toxic toward humans than primates. However, it has not been established that reduced binding of [123I]-CIT corresponds directly with decreased 5-HT axonal markers measured using in vitro tissue samples. Notably, SPECT studies in MDMA-treated monkeys using [123I]INQUIP,18 a SERT (5-HT transporter) radioligand with similar cortical/cerebellar binding ratios to [123I]-CIT, found that MDMA-treated monkeys exhibited cortical reductions of less than 5% by SPECT, while in vitro measures indicated reductions greater than 75%. These values reveal a large disparity between SERT levels determined by SPECT and by direct in vitro tissue samples. This difference between 2 dissimilar methods raises questions regarding the sensitivity of SPECT for detecting SERT in cerebral cortex, as recognized by the authors. Thus, until the sensitivity and accuracy of SPECT with [123I]-CIT for measuring SERT density in neocortex is established, conclusions regarding the degree of MDMA-induced cortical damage may be premature.
While the results of Reneman and coauthors and others suggest that MDMA can produce cognitive impairment (memory loss), several important questions concerning causality and mechanisms remain unresolved. Does decreased binding of -CIT measured with SPECT reliably indicate a loss of SERT in neocortex? Does the decrease in SERT result from frank axonal degeneration or from reduced SERT expression in surviving axons? Histologic studies in animals demonstrate that MDMA causes extensive loss of 5-HT axons in numerous brain regions,17, 19, 20 and a similar pattern of neurotoxic effects likely occurs in humans. An unresolved issue is whether the 5-HT axonal damage in neocortex is directly responsible for cognitive changes. Moreover, the regional localization of 5-HT axon loss that produces cognitive decline should be determined since denervation of hippocampus may cause memory loss, while the neocortical changes might be unrelated.
The observation that [123I]-CIT binding in cortex returns to normal in subjects who were abstinent for 1 year leads the authors to conclude that MDMA-induced damage to cortical 5-HT axon terminals may be reversible. This interpretation must be viewed with caution since, as mentioned earlier, the sensitivity of [123I]-CIT for measuring cortical SERT reliably is not established. Moreover, axonal regeneration in the adult brain may lead to abnormal, dysfunctional circuitry. The notion of spontaneous recovery has important public health implications since drug users may be led to believe that MDMA-induced damage can be reversed merely by abstaining from MDMA. However, experimental studies report that MDMA's neurotoxic effects on serotonin neurons in primates are extremely long-lasting and may be permanent.17, 21
The findings by Reneman and colleagues suggest several future directions for research in MDMA neurotoxicity. As noted previously, preclinical studies in nonhuman primates are essential for the interpretation of findings in humans. For example, it is important to establish that SPECT with [123I]-CIT is capable of reliably detecting MDMA-induced serotonin neurotoxicity in primate neocortex, and if so, the sensitivity of the method should be determined in multiple regions of cerebral cortex. Similarly, abnormal magnetic resonance spectroscopy, positron emission tomography, or functional magnetic resonance imaging data from human MDMA users can best be interpreted with reference to data from nonhuman primates that exhibit documented neurotoxic injury. A similar approach could be applied to functional consequences of MDMA use, including cognitive and neuroendocrine changes seen in MDMA users. Such studies will help clarify the relationship between neurotoxic injury and functional deficits. Another direction for clinical research in MDMA is to determine the long-term effects of MDMA use. These studies should address questions of potential recovery, as well as the possibility that tardive effects of MDMA on serotonin neurons may become manifest with age. Additionally, as knowledge grows regarding the mechanisms of MDMA-induced neurotoxicity, it may be possible to define genetic risk factors for the development of toxicity or methods for preventing injury and promoting recovery in those who have sustained damage.
In conclusion, the article by Reneman and colleagues raises several important questions. Fortunately, the tools for addressing many of these questions are available, and will undoubtedly be used to shed light on many issues regarding the neurotoxic effects of MDMA and its functional consequences in humans. Meanwhile, it is urgent to focus public attention on the current results indicating that MDMA may cause long-term damage and dysfunction in the human brain.
[ 15 December 2001: Message edited by: zorn ]

Of course, the study subjects are the same as from the Lancet journal article from December 1st. You see, it takes a couple of months of massaging the data in order to figure out which subjects to leave in to make the data look even nicer.

What I would like to see:

• A study using MDMA users who utilize harm reduction techniques
• A study from Ricaurte that doesn't use the same 25 or so e-tards that he found
• A new hypothesis: when there is debate, it generally means your hypothesis is wrong
• A control group with altered circadian rhythms

I would suggest the following hypothesis:
Generally, MDMA users use the drug at times that greatly fuck up circadian rhythms. Circadian rhythms ultimately control hormone release, including release of estrogen and testosterone. Both of these hormones have been shown to alter SERT and 5-HT receptor levels. MDMA may not be directly responsible for lowered SERT and 5-HT receptor levels; the lowered levels may be due to altered circadian rhythms which may be altering hormone levels. Circadian rhythms are much more important than most clowns in medicine can admit, especially if their minds are set on some other course of action that gets them grants.

By the title of the article... it says "preliminary results"... then, look at the authors... not too hard to figure out. Best way to make it definite is to actually ask for the raw data from the authors-- if they have nothing to hide, they will hand it over with a note stating,"If you find anything useful, please tell us, and if you want to publish, please acknowledge us."
Standard procedure in science: publish prelim results with subset; then gather more data (really, just look for a few more study subjects); massage data: look for odd things b/w anything possible [b/w male/female popular, easy way to get published], subtracting and/or adding subjects until p=<0.05, et cetera... publish same data, with different 'interesting' findings highlighted depending on what journal you want it published in.
Also, in each of the journals you publish in, have a 'counterpoint' (really, a thinly vailed attempt at one) from someone [*coughRicaurtecough*] that 'strengthens' your argument [your hypothesis] by using childish reverse psychology.
All clear?

2 tabs a week for 5.5 years is a shitload of e. I'd like to see a study where subjects used no more than 1 a fortnight, or even better, 1 per month.

babydoc-
Many of them do include a group of moderate users, 50 tabs of E seems to have become a commonly used cutoff. None of the papers I have read have found any evidence of abnormalities in these users. Thank God.
PoMo-
You seem to have an overly cynical misunderstanding of how research is conducted. Are you at a university? If so, I suggest you talk with some professors there about the way it is done. There are also always published guidelines detailing research misconduct, which include something like "manipulation of experiments or of statistical or analytical procedures such that the published results differ significantly from those that would normally result from the application of the methods reported by the investigator." Though misconduct somtimes occurs, it is generally rarer than you might think, for a number of reasons not the least of which is that results must be reproducible. If you massage some results into existence, you had better be damn sure you're right; a charge of misconduct can destroy your career.
A couple more points:

• The primary difference between preliminary results and real articles are that preliminary results are not peer-reviewed. Journals will normally send submitted articles to unaffiliated scientists, who will review the paper and suggest changes before it is approved and published. This process can easily take over a year; if there's some compelling reason for it get out -- public health here -- you can skip the review process (though it is still reviewed by the journal staff) and get it published preliminarily.
• Publication of the same results in multiple places is not allowed, except in very special cases. Journals will not accept a paper if it has been published elsewhere; some will not even publish it if you talk too much about your results. Now, you can indeed publish different results from the same group of subjects. For example, this paper (which I found did use the same subjects) reported on RAVLT memory scores vs SERT levels; whereas the Lancet article reported on dose/sex vs SERT levels. Neither article included the others' results!
• Subject selection is not nearly as plastic as you seem to believe, for obvious reasons. If you notice, papers contain a methods section with a list of the objective selection criteria used. Inconsistent criteria are a no-no likely to alert the reviewers of the paper; furthermore, they are typically established before the study and statistical analysis begins. There is this little thing called a "research proposal," and you don't get to include in it "we will play around with the data to see which critera give the best results, and then use those." Similarly data-mining is very much frowned upon.
• Perhaps you should read some commentaries? Often they are fairly brutal. And no, you don't get to ask your friends to write them... the journal solicits people to do this. Ricaurte, as you pointed out, is often asked because he has been working in the field for a long, long time and is responsible for a number of major results, eg he is a "respected figure".
• "when there is debate, it usually means your hypothesis is wrong"
  What? Are you serious? When a hypothesis is proposed, it has by nature not been investigated, and there will be debate; this is healthy. If you notice the majority of the debate in the literature has shifted to delineating the risk factors and nature of MDMA-induced damage, rather than arguing about its existence.

Finally, what is with the circadian rhythms thing? It's sketchy as hell, and wouldn't explain how in EVERY published study, damage to SERTs/memory/whatever corresponds to the amount of MDMA use, is unchanged with MDMA abstinence, is uncorrelated with other drug consumption, etc. These are signs that all shout MDMA neurotoxicity, which has been DIRECTLY OBSERVED in animals. It's almost criminal at this point to continue to pretend it's not an issue.
-Zorn

for shit sake, when is someone going to do a conclusive study. several studies are done every year and every study is flawed in one or more ways. the flaws are always noted and detailed in the review. yet none of the successive studies make corrections. we continually get results that are inaccurate for the same reasons.
when will they get their shit together and do just one really good study? it will save a lot of time, money, and confusion.
i hope its just a coincidence that these poorly designed studies are helping the governments war on drugs by creating confusion about this big bad drug MDMA.
im not disputing the fact(?) that MDMA causes neurotoxicity. i just wish that the extent it effects cognitive function could be defined acurately.
SirLSD

this might be silly, but im wondering when they say that they noted a reduction of receptors, how do they know that?
did they scan the same subjects before they started mdma use.
and 2 tabs a week for 5.5 years? seems unlikely? real ecstacy, 240mg a weekend, every weekend? where did they find these people?
i realise my questions seem defensive, but its not intended to be that way.
are the people doing mdma research for therapy(the new fda approved one) gonna do these scans and asses memory loss after treatment etc?

Directly observed neurotoxicity in other animals is from large doses. There is no argument here, large doses of just about anything are neurotoxic.
Altered circadian rhythms cause quite some effects, zorn, and the effects will persist until the clocks are properly phase-shifted back (which may not be as easy as just "getting back on schedule). Interestingly, one of the first effects of altered circadian rhythms is on the MEMORY.
I can forgive you for being young and naive, if you forgive me for being old and cynical. The way I describe science as *actually* working may seem a bit cynical, but I've seen it happen, and I continue to see it happen. Science is not as objective as you seem to think it is. Societal attitudes shape people's understanding of the data... data exist not in a vacuum, but in a grey area of subjective interpretation.
Zorn, look into the history of ulcers and H. pylori for an example of how a simple paradigm shift can take over a decade to gain acceptance. If you think illicit drug research exists in a more objective environment than that of ulcers, you are sorely mistaken.

I would like to see MORE STUDY PARTICPANTS goddammit! I mean, I suppose the results are somewhat siginificant, but there doesn't seem to be HUGE discrepancy between the control and the MDMA users (though I'm by no means denying the existance of the discrepancy)...still, I would like to know more specifically HOW BADLY MDMA users' memory has been impaired -- and more test subjects are needed for that.
As for circadian rhythms, PoMo, that is a good point I had not considered before. I know that circadian rhythms and early puberty have been theoretically linked, and I would not be suprised to see alteration of these rhythms affecting cognitive function as well. However, I do think it is rather presumptuous to assume that all MDMA users have radically abnormal circadian rhythms, as it would be rather foolish to assume that ALL people who roll are "ravers".

sirlsd-
<laughing> I wish someone would do the perfect study, too. But sadly there is rarely (never) enough known to control for all possible confounders, or eliminate every source of possible error. You proceed slowly from different avenues, and slowly the evidence accumulates, results become better, etc.
djremix-
They found these people in the Netherlands via a harm-reductiona agency. You're right, studies like this can't rule out preexisting differences in heavy users, though the correlation between MDMA use/impairment and the use of controls help this. These people take 2.2 tabs, there is no way to know if it was real MDMA/how much besides what the subjects say. There have been longitudinal studies done, which found the same markers of SERT reduction, correlated with the number of E pills taken between the two tests.
fairnymph-
Yup, me too... though the errors bars are not too bad on these tests. I'm not familiar enough with the RAVLT to say what the difference means, but I'd guess it's moderately bad.
pomo-
To be completely technical, neurotoxicity hasn't *techincally* been directly observed from MDMA... it's only been seen via silver staining, which is believed to mark neurotoxicity 100% of the time. Nearly every observation is somewhat indirect. However at high doses in humans, both very clear markers of serotonergic damage and functional impairment have been seen correlated to MDMA use, and this is consistent with theory. There's pretty good proof, and there's plenty more... anyways, I don't understand, did you not just say a few days ago:

Pomo said:I am not denying that heavy doses of mdma over long periods causes damage. Certainly does.
So why insinuate that the authors must be dishonestly manipulating their data to get these results?
The example of H. pylori and ulcers is, I think, and excellent one. New discoveries, slow buildup of evidence, originally confusing results, and eventual understanding. Nowhere here is the blatant misconduct and fabrication you seem to be implying above. I should also point out that serious MDMA research has been going on since the early 80s, and it took a similar amount of time until neurotoxicity evidence built up to a compelling level.
As for misconduct, yes, it does occur. The ISIS protocol, for example, or Summerlin's skin graft via inkpen come to mind. But it's not that common. In funding audits, I think, only like 1% of projects are found to contain misconduct. Sure, it's still a problem, and one that needs to be addressed; but the point is you can trust the vast majority of articles. Saying "oh, they must have manipulated their data" might be a tiny possibility in a single article.... but assuming many different groups are all in cahoots coming up with bullshit data is a little ridiculous, don't you think?
And I agree that of course, data doesn't exist in a vacuum, and human preconceptions, hopes, etc enter into it. But the whole scientific process is designed to minimize this as much as possible. When you read an article, certain explanations may be preferred, sure, and you should consider this. But with MDMA? Cmon... the evidence doesn't fit anything else well, at all. And importantly, it's VERY reasonable theoretically that it's causing the damage. There are just too many leaps of faith you must make to justify that it's not MDMA, and zero data supporting that conclusion, despite looking -- it's clear from reading the literature that a number of the researchers didn't want to find damage -- that you've gotta pull out Occam and move on.
The points I pointed out above still apply; those are facts, not my personal judgements. If you want to see if someone played with subject selection, look at their proposal, and if they changed it, ask why. Either they have a good reason, or they are out-and-out lying.... in which case they might have as well just fabricated their data to begin with.
As far as circadian rhythms go, well, very few things seem to cause this type of serotonergic misfunction. I would expect that any adverse consequence of circadian disruption have been well-characterized: do they match those seen here? Are there other consequences not seen here? And the rats/primates weren't being sleep deprived, nor are the ex-users likely to, any more (less actually) than the raver controls used here. It just doesn't jibe.
Hell, I have seen science working too, so don't get snippy with me. Not biomedical science, but science nonetheless. Big egos are useful: someone is bound to disagree with you on general principles (I've been known to do this plenty) , and he's gonna point out every flaw in your work he can think of, just to make sure you can take it.
Finally, I have trouble believing you're that old and experienced... from the tone of your posts and the way you seem to enjoy arguing here, I'd guess you're fairly young.... most older scientists get enough of that thing arguing with intelligent colleagues, and become frustrated arguing with less-informed people.... they've got other things to do in their free time. Anyways, how old are you? As I said before, I'm 22.

Click to expand...

sure, i cant expect them to control all possible confounders, or eliminate every source of possible error. but if i were funding the study i would bitch slap them for not allowing for cannabis use, and for the small number of subjects in the study! come on, its so simple its rediculous that they didnt. dont you agree??
SirLSD

zorn, I didn't say I was experienced and old-- I said old and cynical... (how old I am isn't really the point)...
Arguing from an extreme brings out good arguments from the middle (i.e from you)... plus, it's fun... and, it's not that time consuming for me (though I do have better things to do, it is a few minutes a day for fun)...
You won't find too much about circadian rhythms quite yet... give it a few years-- trust me, I ain't wrong too often on predictions...
Interestingly, even bio-clock genes between insects and mammals have enormous homology... kind of points out their evolutionary importance, no?

In response to PoMo:
A study using MDMA users who utilize harm reduction techniques
Useful for more precise data. But not helpful in letting the average ecstasy user who does little as far as harm reduction in letting them know what's going to happen.
A study from Ricaurte that doesn't use the same 25 or so e-tards that he found
A new hypothesis: when there is debate, it generally means your hypothesis is wrong
As far as any debate involving Ricaurte's studies, citing the re-use of the 25 individuals would suggest the debate is caused by inconclusive results caused by an incompetant doctor preforming the test.