Drug misuse is a major health issue with wide-ranging economic and social implications. In Australia, tobacco use is responsible for 9% of the nation’s disease burden, while alcohol and illicit drug use account for 5.1% and 1.8% respectively (AIHW, 2016). The total economic and social cost of these drugs is estimated at more than $55 billion per annum (AIHW, 2011). Nonetheless, high proportions of Australians 14 years or older report recent (preceding 12 months) drug use: 14.9% are current smokers, 77.5% indicate consuming alcohol, and 15.6% have used illicit drugs (AIHW, 2017). While relatively few of these individuals will transition to dependent use, the factors influencing this transition, as well as those that may have a mitigating impact, remain unclear. Numerous studies have consistently reported the association between various cognitive factors and dependent drug use. For instance, compared to healthy controls, individuals dependent on various substances exhibit difficulties on tasks that assess executive function, particularly those related to decision-making and inhibitory control (Bickel et al., 2012; MacKillop et al., 2011; Smith et al., 2014; Verdejo-García et al., 2008). Alcohol and substance dependent individuals additionally show significant impairments across tasks that assess various types of learning and memory (Lundqvist, 2005; Potvin et al., 2014; Stavro et al., 2013). At the neural level, attention is focused on the interplay between bottom-up dopamine-mediated reward circuitry, and top-down prefrontal regions involved in self-control (Goldstein & Volkow, 2011; Verdejo-García et al., 2006). Substance dependent individuals show dysfunction across both systems (Goldstein & Volkow, 2011; Verdejo-García et al., 2006; Verdejo-García & Bechara, 2009). The cognitive deficits evident among substance dependent individuals are considered both a cause and consequence of drug use (De Wit, 2009). Longitudinal studies involving both children and adolescents have shown how executive function performance in early life predicts illicit drug use and the onset of problematic alcohol consumption, independent of other risk factors (Audrain-McGovern et al., 2009; Fernie et al., 2013; Kollins, 2003; Nigg et al., 2006; Verdejo-García et al., 2008). At the same time, drug taking appears to exacerbate cognitive issues related to memory, learning and executive function (Colzato et al., 2007; De Wit, 2009; Fillmore & Rush, 2002; Mendez et al., 2010; Verdejo-García et al., 2008). Given these findings, studies concerned with identifying non-dependent drug users at risk of transitioning to dependency often focus on intake patterns and their relationship to (usually single) assessments of executive function, memory and learning; such studies tend, however, to yield inconsistent results (Fernie et al., 2010; Field et al., 2007; Henges & Marczinski, 2012; MacKillop et al., 2011; Moreno et al., 2012; Murphy & Garavan, 2011; Poulton et al., 2016; Rossiter et al., 2012; Smith & Mattick, 2013). This may be due to a number of factors. For instance, intake is frequently recorded using retrospective summary surveys, which can underestimate consumption and are limited in their ability to capture within- person variability and change over time (Piasecki et al., 2007; Poulton et al., 2018; Runyan et al., 2013). At the sub-clinical level, it is possible variability of intake is linked to performance – or, indeed, variability of performance – on cognitive tasks. To investigate this proposition, we are utilising mobile-based ecological momentary assessment (EMA) – which involves using a smartphone app to collect data repeatedly, in close proximity to the event, and in the natural environment (Shiffman, 2009; Trull & Ebner-Priemer, 2013) – to assess both intake and cognition. We hope this enables a more detailed exploration of the in-the-moment cognitive antecedents and consequences of intake behaviour. EMA techniques additionally provide a unique opportunity to assess how individualized feedback about intake and cognition delivered in real-time might alter the trajectory of licit and illicit drug use over time. Early brief interventions have already been identified as a potentially effective means of curbing intake among non-dependent drug users, particularly when they are individualized, though a number of difficulties have been noted (Bernstein et al., 2009; Das et al., 2016; Riper et al., 2009; Tait & Hulse, 2003). Namely, there are limited professionals to deliver these interventions; it can be difficult contacting users; and, implementation and delivery costs can be high (Riper et al., 2009). Interventions delivered via smartphone apps potentially address these limitations.
Research Questions / Hypotheses
This is an ongoing study that aims to investigate whether individual differences in cognitive function are related to licit and illicit drug use in the general population. Specifically, we seek to assess if cognitive factors – related to memory, learning, and executive function – are associated with particular patterns of drug use, including occasional, social and binge intake behaviours. We additionally hope to determine if these cognitive factors can predict changes in drug use over time and if feedback about cognition and/or intake precipitates reduced drug-taking behaviour. The current study aimed to assess whether personalised impulsivity feedback would augment the benefit of alcohol-related feedback. It also aimed to validate whether differences in impulsivity can be observed between groups of binge and non-binge drinkers.
Participants included 169 first-year university students at Time 1. Of these, 146 completed all baseline measures and were randomly allocated to one of three conditions: no feedback, alcohol use feedback only, and impulsivity combined with alcohol use feedback. At Time 2, 103 participants completed all follow-up measures (8 weeks post-intervention).
Participants were requested to take part in two assessments 8 weeks apart. They were recruited remotely. At the commencement of each assessment period, participants were required to complete online baseline (Time 1) or subsequent (Time 2) surveys. They were then required to complete a cognitive stop-signal task online. Feedback condition participants were provided with feedback about their alcohol intake and/or cognition at the conclusion of each assessment period. Control condition participants received no feedback about their intake or cognition.
This is an ongoing study. Key variables of interest are binge drinking, total alcohol consumption, choice impulsivity, and response inhibition. Choice impulsivity is indexed by the Monetary Choice Questionnaire (MCQ) overall k-score; higher scores are indicative of greater impulsive decision making. Response inhibition is indexed by stop-signal reaction time, which is derived from Stop-Signal Task (SST) data. Longer stop signal reaction times are indicative of reduced response inhibition. Prior to primary hypothesis testing, the three groups will be compared on key demographic, alcohol use, and impulsivity variables measured at Time 1. A one-way analysis of variance (ANOVA) will be conducted to test the groups are matched on continuous variables, with a chi-square test of independence used for gender. The primary hypotheses will be tested using a repeated measures ANOVA for each alcohol use variable, with three feedback conditions (No Feedback, Alcohol Feedback, and Alcohol + Impulsivity Feedback) and two time periods (Time 1 and Time 2). Effect sizes will be computed using partial eta squared and interpreted according to Cohen's guidelines: .01 = small, .06 = moderate, and .14 = large effect (Cohen, 1988).
We expect binge drinking and total alcohol consumption to be reduced at Time 2 across all groups as recording alcohol consumption has been found to reduce alcohol intake over time. We expect participants in the Alcohol Feedback group to have reduced drinking to a greater extent than those in the No Feedback group, and those in the Alcohol + Impulsivity Feedback group to have reduced drinking to a greater extent than those in any other group. Data from S1 will be analysed and form part of 4th year theses in 2022. Data will continue to be collected in S2 with a view to publishing results early in 2023.