Additional comments on Rodgers et al. 2003

Ilsa Jerome PhD

Note that a report by Parrott and colleagues published in 2002 also listed a sample 0f 282 subjects, suggesting that the sample is the same as the one discussed in this paper. In 2001, Rodgers and colleagues also published a report that relied on data from an on-line survey of Ecstasy users; it is unclear what relationship the previous report bears to this one.

Previous studies of Ecstasy users have demonstrated that perceived and actual memory are not equivalent. In one study (Fox et al. 2001), self-reported problems attributed to Ecstasy use were not correlated with actual performance on measures of memory.

There is already a growing literature examining memory in current and former Ecstasy users (see MDMA literature Review and MDMA literature updates). It is unclear how to treat assessments of perceived memory in the context of this literature, but the results cannot be treated as measuring the same thing as those studies that assessed memory itself.

Even if a person's perceptions about his or her memory tells us something about how well the person's memory actually is, perception is liable to be influenced by additional factors unrelated to cognitive function, including beliefs about the effects of drug use on memory. Given the widespread media coverage of research findings of reduced memory in Ecstasy users, most respondents are probably well aware of these findings, and they may use this information as a "guide" to answering survey questions when they are uncertain. The influence of these beliefs upon self-reported memory are accentuated when respondents who do not report memory difficulties are systematically removed from the sample, as was done here.

Mariana Cherner PhD

It is important to note the literature is very mixed regarding the relationship of self-reported cognitive complaints and actual neuropsychological (NP) performance. Complaints are often more related to depressive symptoms than to actual cognitive test performance, although some have found complaints to be associated with NP scores.

see e.g.,

No relationship

Some relationship:

Higher depressive symptoms have been reported among MDMA users (which might be preexisting and/or related to use), and this could be associated with perception of cognitive (and other psychosocial) problems. It is also possible that substance users are generally more vigilant about their cognitive functioning and thus report more problems. In either case, self-selection bias among those who answer surveys or participate in such research continues to be a big problem for interpretation of findings.

That they did find decent correlations between reported level of MDMA and cannabis use and these memory complaints (after removing 84 respondents, of course) is more interesting but possibly still vulnerable to the issues above. For cannabis at least, meta-analyses of the literature suggest trivial long-term effects.


In our research group, we found better NP performance among methamphetamine dependent study participants who also smoked pot.

View the full text (PDF format) of this paper, and read or return to the critique of the Rodgers et al paper.

Visit PubMed to learn more about the referneces cited by Cherner.