Purpose: Self-reported cognitive function; to examine whether ecstasy use increases occurrences of self-reported impairment in prospective memory.

Design: Retrospective (non-experimental) 2-group between subjects design, with drug use (ecstasy use versus no ecstasy use) serving as a between-subjects factor, and with all participants completing measurements of self-reported prospective memory. Roughly half of the participants in each group completed a measure of social desirability and tendency to lie.

Subjects: 46 ecstasy users and 46 non-ecstasy users probably residing in the Newcastle-upon-Tyne (England) area, and recruited via snowball sampling technique. Matching -Groups approximately matched age and cannabis use.

Criteria for Inclusion, Ecstasy Users - Not explicitly stated beyond self-reported use of ecstasy, but apparently use of ecstasy at least once (or more) per month. Non-ecstasy users - No self-reported use of ecstasy during lifetime, but use of other drugs permitted. Both Groups - Abstinence from cannabis use for at least 3 days prior to study day and abstinence from ecstasy and alcohol for 24 h prior to study day, with abstinence verified through self-report only.

Drug Use Parameters - No information is provided concerning lifetime ecstasy use (in tablets or occasion), but average dose per use is 1 tablet. No information is provided concerning duration of use. Average frequency of ecstasy use was 10 occasions per month, range 6-30 times a month. Time since last use, in days, is not provided, though all participants reported abstinence from ecstasy for at least 24 h prior to study day. Other drug use - 35/46 ecstasy users had used cannabis at least 1 time per week, with an average of 2 joints a session for 96 months. 15/46 non-ecstasy users smoked cannabis at least 1 time per week, with an average of 1.5 joints per session, and for a duration of 72 months. 14/46 ecstasy users had used cocaine on 4 to 8 occasions during a lifetime, and 6/46 non-ecstasy users reported using cocaine on 1 to 4 occasions during a lifetime. 42/46 ecstasy users had consumed an average of 20 units of alcohol per week, with duration of use of 96-120 months. 34 non-ecstasy users drank an average of 19 units of alcohol per week, with duration of use reported at 108-120 months. Neither ecstasy users nor non-users reported having used LSD.

Group Demographics and Matched Variables - The authors do not explicitly state any matched characteristics, though it appears an attempt was made to match for age, and use of substances other than ecstasy. Results of analyses indicated participants' ages did not significantly differ. Gender, as M/F ratio, for ecstasy users = 28/18, and for non-ecstasy users = 17/29. Age - The average age of ecstasy users was 24.6 years (range = 18-43 years), and average age of non-ecstasy users was 26.1 years (range = 18-40).

Education - No information is provided concerning educational attainment in either population.

Measures: Memory - Prospective memory (memory to engage in behaviors at some future date, planning) assessed via the Prospective Memory Questionnaire (PMQ), a self-report questionnaire using Likert-type scale to indicate likelihood of occurrence in set period of time. The PMQ has 3 scales; short-term habitual prospective memory (e.g. forgetting to turn off alarm clock after waking), long-term episodic prospective memory (e.g. forgetting to pass on a message to someone) and internally cued prospective memory (e.g. forgetting what one intended to say mid-sentence). In addition, PMQ has "memory strategies" scale.

Socially Desirable Response/Faking - Assessed via the Lies sub-scale of the Eyesenck Personality Questionnaire (EPQ). The instrument completed by 24/46 ecstasy users and 24/46 non-ecstasy users.

Analyses: PMQ "Strategies" and Lies Score - A 1-way between-subjects analysis of variance (ANOVA) was conducted on the PMQ-Strategies scale and the Lies scale of the EPQ, with drug use (ecstasy user versus non-ecstasy user) serving as the between-subjects variable. P value was unstated but probably set at p. = 0.05.

PMQ Scales - An analysis of covariance (ANCOVA) was performed on PMQ scores (except for "Strategies), with drug use (ecstasy user versus non-ecstasy user) serving as a between-subjects factor, and with past alcohol, cannabis and cocaine use serving as covariates. Value for p. not stated explicitly, but was probably set at p. = 0.05.

Results - Significant Differences Found: Prospective Memory - Significant differences between the responses of ecstasy users and non-ecstasy users were found on all PMQ memory scales (long-term, short-term and internally cued). In all cases, ecstasy users reported more instances of prospective memory failure than non-ecstasy users.

Results - No Significant Differences: Memory - Ecstasy users and non-ecstasy users did not have significantly different PMQ strategy scores, suggesting that the two groups used a similar number of mnemonic strategies. Socially Desirable Response/Tendency to Lie - There were no significant differences between ecstasy users and non-ecstasy users on Lies scale scores, indicating that both groups did not differ in tendency to provide socially desirable responses or to "fake."

Overall Effects: When compared on a measure of self-reported failures in prospective memory, or remembering to act at some future date, ecstasy users and non-ecstasy users similar in age differed on number of self-reported failures in prospective memory. An analysis that accounted for differences in use of other substances, such as alcohol, cannabis and cocaine, ecstasy users reported significantly higher occurrences of forgetting to perform a habitual or frequent behavior, forgetting to carry out an action (non-habitual) in the future and forgetting internally cued actions. Ecstasy users were neither more nor less liable to lie or generate socially desirable responses in general than were non-ecstasy users, at least as measured by the EPQ Lies scale.

Comments: This study is reported as one of 3 studies, and is intended as a follow-up to an earlier study published by the same author (Heffernan et al. 2001a). It is one of the first reports examining the effects of ecstasy use on prospective memory. One major weakness of all studies relying on the PMQ is that it assesses self-reported memory failures rather than actual prospective memory failures. Other studies indicate that self-reported cognitive failures may not reflect actual cognitive failures (e.g. Croft et al. 2001; Fox et al. 2001). Furthermore, a previous study employing a complete assessment of memory found impairment in story recall and global memory, but not prospective memory, after continued ecstasy use (Zakzanis et al. 2000). In addition, an on-line study published in the same issue of this journal and co-authored by Heffernan suggests that only long-term self-reported prospective memory is strongly associated with past ecstasy use (Rodgers et al. 2001), while cannabis use, and not ecstasy use, may be associated with impairment on at least one PMQ scale (short-term). It is notable that the authors strove to control for differences in use of other drugs via analysis of covariance, so that results can be more definitely associated with ecstasy use. However, the possibility of gender differences in prospective memory remains a viable alternative hypothesis for the study findings, since most ecstasy users in this study are male, and most non-ecstasy users are female, and analyses did not examine or attempt to account for gender effects. Additionally, abstinence from drug use prior to study day was verified by self-report alone rather than through urinary analysis or other means of verification (hair analysis). A full account of drug-use patterns, including duration of use, total lifetime consumption and time since last use is not provided, depriving the authors of the opportunity to associate parameters of ecstasy use with PMQ scores. The limitations inherent in retrospective studies apply to this study, including self-selection and difficulties inferring causality.