Implications of MDMA use for prospective memory function and substance use patterns in an Australian sample: A web‐based pilot study

Abstract

The use of amphetamine type stimulants, particularly MDMA, is a global concern. Little research has been conducted on the association between MDMA use and everyday memory function—prospective memory. Twenty‐five MDMA users, 37 cannabis users, and 43 illicit substance‐naïve controls were assessed on their substance use history and reported prospective memory performance as measured by the Prospective Memory Questionnaire (PMQ) using a web‐based survey. There were significant differences between MDMA users and controls and cannabis users and controls on long‐term episodic subscale of the PMQ. However, given the high prevalence of cannabis co‐use by MDMA users, it was not possible to determine if MDMA use alone is associated with prospective memory performance. The substance use patterns of the sample were evaluated. Alcohol was the most used substance followed by tobacco, cannabis, and MDMA. The incidence of polydrug use was high, with all illicit substance use reporting having used at least two substances in their lifetime. The present study supports previous research into prospective memory deficits associated with substance use, and provides a basis for future research, particularly for elucidation of prospective memory deficits specific to MDMA use and further evaluation of substance use patterns.

Key words:

• biological correlates
• cognitive processes
• consciousness
• hypnosis
• individual differences
• intelligence
• neural bases of motivated behaviors
• neural networks
• neuropsychology and other biological issues
• neuroscience
• personality processes and individual differences
• psychophysiology

INTRODUCTION

Abuse of amphetamine‐type stimulants are fast becoming a global concern of epidemic proportions, with some 30 million people affected by their use worldwide (United Nations Office on Crime and Drugs, 2003a, 2003b). Similar usage trends have been reported in Australian studies (Black et al., 2007; Degenhardt, Copland, & Dillon, 2005; Simon & Mattick, 2002). Of the amphetamine‐type stimulants, MDMA (3,4‐methylenedioxymethamphetamine or ‘ecstasy’) is the most concerning due to its increasing popularity among many recreational drug users, with the United Nations Office on Drugs and Crime (2003b) estimating that some 8 million people worldwide have used the substance. Consequently, the effects of MDMA use on cognition, memory function and behavioural indices are well reported (Rodgers et al., 2001, 2003, 2006; Zakzanis, Young, & Campbell, 2003). Researchers report that underlying neurobiological damage of serotonergic systems within the frontal cortex and hippocampus have been implicated with MDMA abuse history (Parrott, 2000; Parrott & Lasky, 1998; Reneman et al., 2006; Rodgers et al., 2001, 2006; Zakzanis et al., 2003). Since studies have shown that these systems are associated with memory function (Burgess, Scott, & Frith, 2003; Hannon, Adams, Harrington, Fries‐Dias & Gipson, 1995), then memory dysfunction may be one of the changes associated with MDMA use, which could be used to quantify the extent of damage or change (Montgomery & Fisk, 2007).

The ability to form, maintain, and execute delayed intentions is an integral memory activity in everyday life (Burgess et al., 2003; Okuda et al., 2007). The cognitive function driving such a memory process is referred to as ‘prospective memory’, a distinct form of episodic memory that is believed to rely on central executive processes (Graf & Uttl, 2001; Hannon, Adams, Harrington, Fries‐Dias, & Gipson, 1995). Simply, this reported prospective memory is the process of remembering to perform an action at a future point in time or, ‘remembering to remember’ (Hannon et al., 1995; Rodgers et al., 2001; Zakzanis et al., 2003). The Prospective Memory Questionnaire (PMQ) is a subjective measure of prospective memory function that allows for assessment of ‘real‐world’ memory experiences (Rodgers et al., 2001).

The PMQ includes three subscales measuring different components of prospective memory: long‐term episodic (LTE), short‐term habitual (STH), and internally cued (IC) prospective memory (Hannon et al., 1995). Additionally, there is a fourth subscale, the techniques to remember scale (TR), which measures the employment of strategies to aid memory. Heffernan, Ling, and Scholey (2001) conducted a laboratory‐based study, comparing a sample of MDMA users to a sample of MDMA‐free control subjects on reported prospective memory performance as measured by the PMQ. Participants gave self‐reported substance use histories for cannabis, tobacco, alcohol, and cocaine in addition to MDMA use, and the results indicated no significant difference between the groups in the effects of these other substances on PMQ scores. However, MDMA users reported significantly poorer performance than MDMA‐free controls on each of the PMQ memory subscales: LTE, STH, and IC prospective memory. The TR subscale revealed no significant differences. Heffernan, Ling, et al. suggested that although MDMA users suffer prospective memory compromises, they are either not aware of these compromises or do not compensate for these deficits by employing strategies.

These findings were replicated by a second study by Heffernan, Jarvis, Rodgers, Scholey, and Ling (2001), except MDMA users and controls displayed no significant difference on the IC prospective memory subscale. This study also used an assessment of central executive function in addition to the PMQ in a different sample of MDMA users and MDMA‐free controls. MDMA users also exhibited significantly poorer performance on the measure of central executive function, leading the authors to conclude that reported prospective memory and central executive functioning are linked. The nature of that link was not further elucidated. Yet Montgomery and Fisk (2007) found no significant differences between MDMA polydrug users and MDMA‐free controls on all four of the PMQ subscales. Although MDMA users scored poorer on the LTE and IC PMQ subscales, regression analyses revealed that cannabis use was the most important predictor of reported prospective memory deficits in MDMA users. It was suggested that reported prospective memory deficits apparently related to MDMA use may in fact be the result of MDMA and cannabis use, or cannabis use alone.

Several similar studies investigating the effects of MDMA on memory performance have been conducted over the World Wide Web (Rodgers et al., 2001, 2003, 2006), even though Buchanan et al. (2005) highlighted potential lack of reliability when compared with pen‐and‐paper methods. However, web‐based methodologies are generally popular because of their capacity to recruit large sample sizes; they are also ideal for research where subject anonymity is paramount and face‐to‐face testing is problematic (Rodgers et al., 2001). Rodgers et al. (2001) explored the effects of MDMA and cannabis use on prospective memory function using a web‐based self‐report survey. MDMA and cannabis were found to differentially affect reported prospective memory as measured by the PMQ. Specifically, cannabis use was associated with poorer performance on the short‐term and IC subscales of the PMQ, reflecting deficits associated with current scenarios. MDMA users demonstrated significant impairments on the long‐term subscale of the PMQ, indicating deficits in storage and retrieval. The researchers suggested that these findings imply that underlying mechanisms that affect prospective memory performance differ for cannabis and MDMA users. It was also observed that cannabis users adopt fewer TR (and less frequency). Although they suffer from everyday memory complications, cannabis users do not, or cannot adopt effective strategies/techniques to aid in memory. In another web‐based study (Rodgers et al. 2003), MDMA users were found to report more difficulties in long‐term prospective memory (as measured by the long term episodic subscale of the PMQ) than users of other substances or controls. A consistent feature of past MDMA prospective memory research is the propensity for many of the substance‐using participants to be ‘polydrug’ users (Gouzoulis‐Mayfrank & Daumann, 2006; Winstock, Griffiths, & Stewart, 2001), but, particularly, cannabis (Montgomery & Fisk, 2007).

The current web‐based study attempted to investigate the pattern of substance use in a small pilot Australian cohort with three aims in light of the UK studies (Rodgers et al., 2001, 2006). The first aim was to investigate the effect of MDMA on reported prospective function as measured by the PMQ. The second aim was to address the effects of MDMA and cannabis on prospective memory by comparing MDMA users with MDMA naïve cannabis users. Furthermore, the current study also aimed to investigate differences in prospective memory performance between these illicit substance using groups and a control group of illicit substance naïve participants (i.e., users of alcohol and/or tobacco, or no history of substance use).

METHOD

Participants

One hundred and sixteen self‐reported HIV‐seronegative participants ranging from 19 to 66 years of age completed the survey. However, seven participants reported being under the influence of a substance (including alcohol, methamphetamine, MDMA and other amphetamines) at the time of completing the survey, and thus were excluded from subsequent analyses. There were 25 participants who were MDMA users (13 males, M = 27.28 years, standard deviation (SD) = 6.30), 37 who were classed as MDMA‐free cannabis users (9 males, M = 27.70 years, SD = 7.65) and 43 control participants who were naïve to illicit substance use (12 males, M = 27.72 years, SD = 11.20). Four participants (1 male, M = 38.00 years, SD = 18.09) did not meet criteria for any of these categories; however, their data was retained for investigation into substance use patterns.

Materials

The ‘Opinio’ web‐based survey creation and management software was used to create a survey for online access. All of the survey instruments utilised in the current study were presented as interactive web forms. The online survey was hosted by Swinburne University of Technology, and was accessible using a single web address (URL). A short questionnaire was included to record basic demographic details, including age, gender and education level. The study was given ethical approval by the University Human Research Experimentation Committee.

A questionnaire was created to measure substance history and was loosely based on the UEL Recreational Drug Use Questionnaire (Parrott, Sisk, & Turner, 2000). Questions regarding alcohol, tobacco, cannabis, methamphetamine, MDMA, amphetamines (other than methamphetamine and MDMA), cocaine, LSD, magic mushrooms, barbiturates and benzodiazepines, opiates, anabolic steroids, solvents and volatile substances, and excess prescription or over the counter medication use were included. For each substance type, participants were required to indicate if they have used the substance, the age of first use, lifetime use, if they are abstinent from use, the duration of abstinence, and their average and maximum doses of the substance. Participants responded to lifetime use by indicating how many times they had used the particular substance in their lifetime: 1–2 times, 3–9 times, 10–99 times, or more than 100 times. These intervals were classified as ‘experimental’, ‘light’, ‘moderate’, and ‘heavy’ lifetime use, respectively. Participants were only required to specify the type of excess prescription or over‐the‐counter medication they had used to excess; questions regarding abstinence, dose, and age of first use were omitted. Similarly, participants were not required to specify average or maximum dose for solvents or volatile substances. The questionnaire also required participants to indicate if they were under the influence of any substances at the time of testing, and, if applicable, to specify the substance, or substances, which were influencing them. For each question, participants were provided with a ‘prefer not to answer’ option. Those reporting to be under the influence were not included in the study.

Prospective memory performance was measured using the Prospective Memory Questionnaire (PMQ), a valid and reliable self report measure of prospective memory function (Hannon et al., 1995).¹ The PMQ consists of three subscales measuring different aspects of prospective memory: 14 items measure LTE prospective memory (such as ‘I forgot to run an errand I meant to do’), 14 items measure STH prospective memory (items such as ‘I forgot to comb my hair in the morning’), and 10 items measure IC prospective memory (items such as ‘I started to do something, and then forgot what it was I wanted to do’).

Responses are recorded on a 9‐point Likert scale, where 1 indicates the memory failure never occurring, and 9 indicates the memory failure occurring frequently. Higher scores on the PMQ subscales indicate higher incidences of memory failures or poorer prospective memory performance. For each of the 14 items, an average response is calculated. The PMQ contains a fourth subscale, the TR scale, which measures the frequency of strategies/techniques used to aid prospective remembering. This subscale also consists of 14 items (e.g., ‘I write myself reminder notes’) measured on a 9‐point Likert scale, where 1 indicates the technique is never used, and 9 indicates the technique is frequently used (4 or more times a month). For each item on the PMQ, participants are given the option to respond ‘not applicable’, which translates to a score of zero on that item.

Procedure

Participant recruitment included emails to personal and support contacts, posts on web discussion forums, announcements in Internet chat rooms, displaying the survey link on substance support websites, notices at the home institution, and general word of mouth.²

Upon accessing the survey, participants were presented with an informed consent statement that explained the purpose and aims of the current study. Participants were instructed that by submitting a completed survey, they were indicating informed consent to participate in the study. After proceeding with the survey by clicking ‘next’, respondents were presented with each of the survey instruments and relevant instructions for completion sequentially. For each instrument, participants responded to questions by selecting a radio button, typing in a text box or selecting an option from a drop down box. At no point during the survey were participants required to provide details that would make them identifiable to the researchers.

The IP address (URL identity) for each respondent was logged to prevent multiple submissions (this method does not threaten participants' anonymity). Although some respondents may share computers (e.g., the household computer) making this method vulnerable to deletion of valid data, it was deemed prudent to prevent multiple submissions in this manner. The ‘Opinio’ survey software was not set up to distinguish different demographics from same IP. The software also allowed scrutiny of IP addresses so that only Australian data were used.

RESULTS

Results were analysed using the statistical package SPSS for Windows, version 15 and 16 (IBM Corporation, Somers, NY, USA). For the purpose of classification for this study only, participants were classed as MDMA users if they had used the substance at least experimentally (1–2 times), cannabis users were identified as those who had used cannabis at least experimentally but had not used MDMA, and participants who had not used any illicit substances were classed as controls (this included substance‐naïve participants). Polydrug use was classified as at least experimental use for more than one substance (illicit and licit substances) in the participant's lifetime.

Table-1 displays the substance using characteristics for each substance use group: controls, MDMA users, and cannabis users, together with statistics reported by Black et al. (2007) for Australian user population (n = 909) so as to ascertain the demographic similarities between our cohort and that from previous Australian data. All MDMA users and cannabis users were classed as polydrug users. MDMA users have the most varied substance use backgrounds, with at least one participant reporting use for each other substance category. Cannabis was co‐used by almost all MDMA users. Cannabis users exhibited far less marked polydrug use, with polydrug use typically confined to co‐use of alcohol and tobacco. While nearly half of the control group reported polydrug use (alcohol and tobacco), there were two participants who were naïve to substance use.

Table 1 Substance use information for controls, MDMA users and cannabis users

Substance	Controls (n = 43)	MDMA users (n = 25)	Cannabis users (n = 37)
	M = 27.72 years	M = 27.28 years	M = 27.70 years
	SD = 11.20	SD = 6.30	SD = 7.65
Polydrug users	20	25 (23.8%)	37 (35.2%)
Alcohol (64%*)	41	25	36
Tobacco (94%*)	19	23	34
Cannabis (81%*)	–	24	37
Methamphetamine (74 %*)	–	6	–
MDMA (23%*)	–	25	–
Other amphetamines	–	9	–
Cocaine (22%*)	–	9	–
LSD	–	11	1
Magic mushrooms	–	5	1
Barbiturates/benzodiazepines (46%*)	–	9	3
Opiates	–	3	–
Anabolic steroids	–	1	–
Solvents and volatile substances	–	2	–
Excess prescription or over the counter medication	2	6	1

Figures suggest that the numbers of MDMA users in this pilot study are consistent with national figures, whereas the cannabis users group was not. Also, MDMA users tend to use many substances.

* Drug use among the national sample of Australian users (Black et al., 2007).

The PMQ subscale scores for each group are detailed in Table-2. To assess whether there were any differences between groups for the four components of the PMQ, four separate one‐way between groups analyses of variance (ANOVAs) were performed, with the prospective memory subscales as dependent variables. It was revealed that there were no significant differences between the groups on the STH, IC and techniques to remember (TR) subscales. However there was a significant difference between the groups on the LTE subscale, F_{2, 102} = 5.95, p < .001. Post hoc Tukey HSD comparisons indicated that there were significant differences on LTE prospective memory failures between the control group and the MDMA users (p < .05), and between the control group and cannabis users (p < .05). Therefore, MDMA and cannabis users have significantly more LTE prospective memory failures than control participants.

Table 2 PMQ subscale scores for controls, MDMA users, and cannabis users

PMQ subscale	Controls		MDMA users		Cannabis users
	M	SD	M	SD	M	SD
Long‐term episodic (LTE)	1.60	0.71	2.17*	1.02	2.14*	0.96
Short‐term habitual (STH)	1.08	0.50	1.21	0.31	1.11	0.37
Internally cued (IC)	1.93	1.21	2.14	1.32	2.17	1.11
Techniques to remember (TR)	3.11	1.52	3.01	1.51	3.48	1.66

* p < .05.

A multiple analysis of variance (MANOVA) was also performed, further supporting the separate ANOVA findings. Tests between subjects effects revealed a significant difference for the LTE subscale (F_{2, 105} = 3.120, p < .05) (R² = 0.038) for polydrug users and (F_{2, 105} = 3.246, p < .05) (R² = 0.059) for substance group (MDMA, cannabis, others). No significant findings were found with drug frequency (data not reported here). However, a pairwise correlational analysis did reveal a significant (two‐tailed) negative correlation (−0.835, p < .05) for age of drug first use for MDMA and the LTE score. A gender‐related correlational analysis also demonstrated a positive correlation with MDMA (0.232, p < .05) and LSD (0.198, p < .05) usage for polydrug user females compared with polydrug males.

Patterns of substance use in the population surveyed were explored by assessing the number of participants within a variety of categories. Table-3 details the features of the substance use patterns, including the number of users for each substance group (by gender), age of first use, and the number of abstinent users. The number of abstinent users for each group did not affect the outcomes of the statistical analyses. As we had recruited participants from sites associated with support, a higher than normal number of participants who are seriously motivated to quit were recruited. Participants were classed as a user of a substance if they had used the substance at least experimentally. Of the 109 participants surveyed, 86 reported polydrug use. Conversely, a very low number of participants (two) were classed as being substance naïve—that is, they had not used any form of substances, licit or illicit.

Table 3 Patterns of substance use including age of first use, number of users, and number of abstinent users

	Users (n)			Age of first use (years)		Abstinent users (n)
	Male	Female	Total	M	SD
Polydrug users	27	59	86	–	–	–
Substance naive	1	1	2	–	–	–
Alcohol	34	72	106	14.92	2.26	6*
Tobacco	26	52	78	15.35	2.73	35*
Cannabis	22	39	61	17.15	2.30	33*
Methamphetamine	4	2	6	21.17	2.48	3*
MDMA	13	12	25	20.16	2.01	11*
Other amphetamines	10	10	20	19.70	2.56	9*
Cocaine	5	1	9	24.67	4.72	6
LSD	7	5	12	20.50	3.07	10
Magic mushrooms	4	2	6	20.83	3.31	5
Barbiturates/Benzodiazepines	5	11	16	21.07	3.60	9*
Opiates	1	2	3	23.00	5.57	2
Anabolic steroids	1	0	1	19.00	0	1
Solvents and volatile substances	1	1	2	17.50	6.36	2
Excess prescription or over the counter medication	2	9	11	–	–	–

The significant correlations for PMQ score‐long term episodic (LTE) are in bold. The number of abstinent users may be associated with the recruitment support websites targeted.

* 5 alcohol, 2 tobacco, 1 cannabis, 1 methamphetamine, 2 MDMA, 2 other amphetamines and 3 barbiturates/benzodiazepines users specified ‘prefer not to answer’ for the abstinence question.

Each substance group had been used by at least one participant. The most commonly used substance was alcohol, followed by tobacco, cannabis, and MDMA. Alcohol was also the substance with the smallest number of abstinent users. For the less common illicit substances, the mean age of first use ranged from approximately 19 years to 25 years. Alcohol, tobacco, and cannabis were typically first used at much younger ages, between approximately 15 and 17 years. Furthermore, the mean age of first use for solvents and volatile substances was also quite young (approximately 17 years). Approximately 80% of all substance users reported polydrug use; this included 100% of illicit substance users (see Table-1). More females than males reported polydrug use, however, with the exceptions of alcohol, tobacco, cannabis, barbiturates and benzodiazepines, and excess medication. More males reported the use of methamphetamine, MDMA, cocaine, LSD, and magic mushrooms.

DISCUSSION

MDMA users and cannabis users did not show any significant differences in overall reported prospective memory performance. However, there were significant differences on LTE prospective memory when compared with the illicit‐substance naïve controls. Exploration of substance use patterns revealed that alcohol, tobacco, cannabis, and MDMA are the most commonly used substances. Disturbingly, each substance group had been used by at least one participant, and all illicit substance users reported polydrug use. Apart from the female polydrug use pattern, these overall trends were consistent with those reported as national trends (Black et al., 2007). For this study, it must be acknowledged from the outset that the trends and significances reported are limited because the state or condition that the participants were in could not be verified.

However, the findings of the current study are consistent with those of Rodgers et al. (2003), with MDMA users reporting significantly more long‐term reported episodic prospective memory failures than substance naïve controls. Gender and age of first usage correlated with this LTE score. However, unlike the Rodgers et al. (2001) study, the current study extends upon these findings by identifying cannabis users as also demonstrating significantly more LTE failures than illicit substance naïve controls. Additionally, there were no significant differences between MDMA and cannabis users on any of the components of prospective memory. However, in view of the fact that the MDMA user group was contaminated with 24 out of the 25 MDMA users also reporting cannabis use, the results may indicate that cannabis use is the underlying issue that affects LTE memory in MDMA users. This speculation is supported by Montgomery and Fisk (2007), who suggest that prospective memory deficits could arise due to cannabis use alone or the combination of MDMA and cannabis use. However, without testing cannabis naïve MDMA users, it is impossible to conclude that MDMA alone does not also have a negative effect on long term episodic prospective memory.

A confounding factor for all substance research is the tendency for illicit substance users to engage in polydrug use, which was certainly the case in the current sample (Rodgers et al., 2001). This makes interpretation of the findings problematic, as the varying patterns of substance use complicate the ability to determine exactly what substance, or substances, are having an effect. This issue is further complicated by the fact that, of the control group, only 2 participants reported never using any type of substance. The remaining 41 controls reported either using alcohol or tobacco, and nearly half had used both of these substances. This is problematic, as impairments in prospective memory function have also been demonstrated in association with the use of these licit substances (Heffernan & Bartholomew, 2006). It was found that teenagers classed as excessive alcohol users reported more errors in all three PMQ memory subscales, yet there were no significant differences for the TR subscale. Heffernan et al. (2005) investigated the relationship between cigarette smoking and prospective memory function, and reported that cigarette smokers performed significantly worse than non‐cigarette smokers on long‐term prospective memory (as measured by the LTE memory subscale of the PMQ). This difference was evident even when controlling for the use of other substances and the number of strategies use to aid memory (as measured by the TR subscale). This suggests that the control group in the current study may not be an accurate population of participants with ‘normal’ prospective function, unaffected by the use of substances. One way to address this problem would be to assess substance naïve, MDMA‐only substance users, alcohol‐only users, and cannabis‐only users, etc., and employing much larger cohorts.

There are several concerns for any study using self‐report measures of prospective memory such as the PMQ. First, there is the ‘memory paradox’, which states that those who suffer memory impairments cannot effectively report cognitive slips (Zakzanis et al., 2003). Second, there is the issue of comorbid conditions, such as learning disabilities, decision‐making deficits, and concentration difficulties that may complicate neurocognitive function in individuals who are substance dependent (Bechara & Martin, 2004; Gonzalez, Bechara, & Martin, 2007; Rodgers et al., 2006; Rogers et al., 1999). Finally, there is the concern that although participants reported not being under the influence of a substance at the time of testing, they may have been going through varying wash‐out phases (variable times taken to eliminate the substance (s) from the body) which could be quite be typically 1–2 days (sometimes even a week), and the symptoms can include physical exhaustion and negative effects associated with serotonin depletion (Heffernan, Jarvis, et al., 2001). It is possible that some participants could have been experiencing a wash‐out period from recent MDMA use, which may have affected their performance on self‐report measures.

Studies employing web‐based designs are unable to make qualitative observations of concentration deficits and test for the current influence of a substance.

Employing methodologies that utilise laboratory based objective measures of prospective memory may address these issues. Implementing the Rivermead Behavioural Memory Test, Zakzanis et al. (2003) report MDMA users to be impaired on two objective measures of prospective memory. MDMA users have memory deficits associated with event‐ and time‐based prospective memory tasks. Similarly, Rendell, Gray, Henry, and Tolan (2007) performed a study using an objective measure of prospective memory called ‘Virtual Week’, which closely mimics prospective memory tasks that occur in everyday life. MDMA users were found to be significantly impaired on the task, irrespective of the type of prospective memory demands. This would also control for another confounding issues associated with motivation and emotional state of the user.

The overall substance use patterns within the current sample were consistent with national trends. There was a high instance of polydrug usage, and high numbers of users of alcohol, tobacco, and cannabis. Every substance use category had at least one reported user, and, interestingly, just two participants reported never using a substance of any type. In general, age of first use was similar for the more common substances (alcohol, tobacco, and cannabis), while most of the illicit substances exhibited later, but similar, ages of first use. The age of MDMA first use correlated with the performance on the LTE—poorer LTE scores. One may speculate that neurodevelopmental brain changes associated with the early use of the illicit substance may have a major impact, similarly to a history of trauma and life history (Cook, Ciorciari, Varker, & Devilly, 2009), and the underlying reason why they continue using substances. Further research is necessary to address this.

In conclusion, the current quasi‐experimental study aimed to investigate the effect of MDMA on reported prospective memory function compared with cannabis using and control participants. It was found that MDMA users, cannabis users, and control participants did not differ on the STH, IC and TR subscales of the PMQ. However, significant differences were evident for LTE prospective memory between controls and MDMA users and controls and cannabis users were noted. However, given the high incidence of cannabis co‐use in the MDMA user group and potential pre‐existing conditions in different groups, it was not possible to determine whether MDMA use alone contributed to deficits in long‐term reported episodic prospective memory scores. It may also be possible that individuals with poor PMQ scores or frontal lobe dysfunction would be more likely to abuse or navigate towards abuse. Other important factors such as motivational factors and occupational factors need to be considered in future studies. Indeed, in order to address this issue, there are several directions for future research, including neuroimaging studies to determine underlying neurological mechanisms associated with the effects of MDMA on prospective memory function and cognition, as well as correlational studies for investigating substance use patterns in much larger cohorts longitudinally.

Notes

1. Adapted for online use with permission from Roseann Hannon.

2. The second best recruitment method reported by Black et al. (2007), National Drug Trends.