Full length articleCurrent forms of inhibitory training produce no greater reduction in drinking than simple assessment: A preliminary study
Introduction
Inhibition is the ability to withhold, stop, or delay an inappropriate response (Barkley, 1997, Diamond, 2013); the cessation of an immediate response allows time for other important psychological processes to evaluate the situation, and select and execute a more appropriate response (Barkley, 1997). Deficits in inhibitory control feature prominently in new models of the development, maintenance, and relapse of substance use disorders (e.g., Hester et al., 2010, Jentsch and Pennington, 2014), and failures of control are implicated in DSM-5 criteria involving using a substance more, or more often, than intended, and consistently failed efforts to limit use (American Psychiatric Association, 2013). Inhibition deficits have been confirmed experimentally in users of a range of substances including not only alcohol dependence, but also heavy drinkers (reviewed in Smith et al., 2014).
If it is accepted that an inhibitory deficit is associated with undesirable and/or risky behaviours in these disorders, then the corollary is that training to improve this deficit may decrease these behaviours. Several studies have examined whether alcohol consumption in social drinkers can be reduced with training on an inhibitory task such as the Go/NoGo or Stop-Signal task. In several studies using a modified version of the Go/NoGo task (Bowley et al., 2013, Houben et al., 2011, Houben et al., 2012), the letters P and F, each 50%, were superimposed on images of beer and water. For half the participants, the beer image was paired with the Go stimulus (requiring a fast button press response, “Beer-Go” condition), while for the other half, the beer image was paired with the NoGo stimulus (requiring the response to be withheld, “Beer-NoGo” condition). Although the images were irrelevant to the task (i.e., the instructions focused on making or withholding responses to the letter stimuli), consistent pairing of the beer image with response inhibition should increase inhibitory control over beer stimuli (that is, should train direct associations between alcohol cues and stopping; Best et al., 2016, Bowditch et al., 2016). Indeed, participants in the Beer-NoGo condition decreased their consumption of alcohol in the week after compared to the week before inhibitory training (Houben et al., 2011, Houben et al., 2012; but see Bowley et al., 2013 for no effect). Some studies additionally use a bogus taste test to measure immediate alcohol consumption (for a review see Jones et al., 2016a). Here, participants are presented with a known amount of alcohol and asked to consume as much or as little as desired in order to rate the drink on several dimensions. The participant is not aware that the experimenter will later measure the amount of alcohol consumed. Training on the Beer-NoGo task is associated with a trend to reduced alcohol consumption in the taste test (Bowley et al., 2013, Houben et al., 2011). Inhibitory control training has also been studied in relation to other health behaviours such as food choices; across domains, the effect size for Go/NoGo tasks has been confirmed to be medium-sized and robust by two recent independently conducted meta-analytic reviews (Allom et al., 2016: 0.50; Jones et al., 2016b: 0.47).
The Stop-Signal task can also be used to assess inhibition; in this task, fast choice responses are required to two primary stimuli (e.g., respond with the left or right hand) and the occasional presentation of a stop-signal indicates the participant should interrupt the button press response (Logan and Cowan, 1984, Logan et al., 1984). In studies linking performance of a Stop-Signal task with subsequent alcohol consumption (Jones et al., 2011a, Jones et al., 2011b), participants were instructed to be especially restrained (i.e., successful inhibition was emphasized over fast responding) or disinhibited (i.e., fast responding was emphasized over successful inhibition). After training, participants in the restrained condition consumed less alcohol in the bogus taste test (Jones et al., 2011a, Jones et al., 2011b), but those studies did not examine changes in weekly alcohol consumption. A different variation on the Stop-Signal task was tested by Jones and Field (2013), in which alcohol-related or neutral pictures served as the Go stimuli, and, for different conditions, 90% of stop-signals occurred on alcohol trials (alcohol restraint condition) or on neutral picture trials (neutral restraint condition). A third group were instructed to ignore the stop-signal and respond to all pictures (disinhibited condition). In the bogus taste test, participants in the alcohol restraint condition drank less beer than the neutral and disinhibited conditions, which did not differ; weekly consumption was unaffected. Thus, inhibitory training appears to alter alcohol consumption measured both via an immediate taste test, and in standard drinks per week before and after the experimental session (although not all studies report reductions in weekly consumption: Bartsch et al., 2016). Recent meta-analyses have estimated the effect size for Stop-Signal tasks to be robust, albeit smaller than that for Go/NoGo tasks (Allom et al., 2016: 0.26; Jones et al., 2016b: 0.23). Two explanations are possible for the smaller effect: one is that, at least for the early versions of inhibitory training using the Stop-Signal task (Jones et al., 2011a, Jones et al., 2011b), the task does not associatively link alcohol with inhibition, despite alteration of associations being a principle of cognitive bias modification (MacLeod and Grafton, 2016). Secondly, Jones et al. (2016b) argued that the smaller effect in the Stop-Signal task is due to the fact that these tasks typically involve about 50% failed inhibitions, and that appetitive cues need to be reliably paired with successful inhibition in order for inhibitory training to reduce alcohol consumption (Jones et al., 2016b).
In the current study, we improve upon the previous research in three respects. The first relates to the control conditions to which the inhibitory training conditions are compared, and how the selection of the control condition may alter the results observed. For the Beer-NoGo task, consumption is often compared to the Beer-Go task, in which alcohol is paired with response execution (Bowley et al., 2013, Houben et al., 2011, Houben et al., 2012). It could be argued that such pairing of alcohol with fast responses may lead to impulsive responding and therefore greater alcohol consumption in the taste test. Thus, for the taste test, it is not clear whether differences between Beer-NoGo and Beer-Go conditions represent low consumption in the Beer-NoGo condition, or high consumption in the Beer-Go condition (or both). Indeed, the interaction between time and condition for weekly consumption is at least partly due to an increase in consumption for the Beer-Go condition in Houben et al. (2011), and a similar, although not significant, pattern was observed in Houben et al. (2012). Similarly, for Jones et al. (2011b), performance in the Restrained condition was compared with the Disinhibited condition, and thus it is difficult to interpret taste test differences between the conditions in that study also. In a subsequent study (Jones et al., 2011a), a Control condition was included which received the usual Stop-Signal task instructions (to balance speed and accuracy); results indicated that participants in the Restrained condition consumed less beer in the taste test than the Control and Disinhibited conditions, which did not differ. This is the clearest evidence of inhibitory training producing a reduction in alcohol consumption, yet the Stop-Signal task even with standard instructions still requires (and therefore trains) inhibition. In the current study we include a Control condition, involving a task with similar stimuli requiring attention, discrimination, and a motor response; however, Go stimuli are 25% of trials in the Control task. This means that the prepotent response (on NoGo trials, 75%) is to do nothing, and only activate a response occasionally. Thus, the Control task here cannot be said to require inhibition, but nor would it favour impulsive responding; furthermore, the task uses neutral (non-alcohol-related) stimuli. However, more importantly, our Control condition allows us to examine the effect of assessment alone on alcohol consumption. Known as subject reactivity, or the Hawthorne effect, changes in a behavior simply due to observation of that behavior were first described in 1933 (Mayo, 1933); demonstrations of reductions in alcohol consumption due to assessment have been noted since 1974 (Bartsch et al., 2016, Gallen, 1974, Kypri et al., 2007, McCambridge and Day, 2008, McCambridge and Kypri, 2011; see Clifford and Maisto, 2000, for a review). Not only have the Beer-NoGo and Restrained-Stop conditions been improperly compared to conditions which increase drinking, but the treatment effect of these interventions has so far been confounded with the assessment effect. Participants in the Control condition are expected to reduce their drinking in the week after compared to the week before taking part in the experiment (due to an effect of assessment); participants in conditions which perform an inhibitory training task must therefore decrease their consumption significantly more than those in the Control condition, in order for the inhibitory task to be considered an effective intervention.
Secondly, we also consider the effectiveness of inhibitory training relative to an established method of reducing consumption, namely a Brief Alcohol Intervention (BAI), which consists of questions about and motivational feedback concerning alcohol consumption. Meta-analytic reviews confirm BAIs are effective at reducing consumption (e.g., Bertholet et al., 2005); they are also effective within the specific target population of this study (i.e., university students; Kypri et al., 2009, Samson and Tanner-Smith, 2015). BAIs can easily reach large samples via the internet; although the inhibitory tasks above could theoretically be delivered online (Jones et al., 2014), in the studies cited above, participants have completed the sessions in the laboratory. In order to justify the extra time and effort associated with laboratory testing, inhibitory tasks should also be at least as effective as a BAI at reducing alcohol consumption. Indeed, Bowley et al. (2013) report that the BAI and Beer-NoGo conditions did not differ at taste test; participants in those conditions both consumed significantly less alcohol than those in the Beer-Go condition. In the current study, we include not only a carefully selected Control task but also a BAI condition.
Thirdly, we improve on previous research by more strongly linking alcohol with inhibition in a new version of inhibitory task. Evidence suggests that inhibitory deficits in heavy drinkers are greater when alcohol-related compared to neutral stimuli are used (e.g., Noël et al., 2007, Weafer and Fillmore, 2012), however, the studies using the basic version of the Stop-Signal task (Jones et al., 2011a, Jones et al., 2011b) do not attempt to link alcohol with inhibition, with participants discriminating between and responding to neutral letters X and O. The modified version of the Stop-Signal task (Jones and Field, 2013) has a stronger link between alcohol and inhibition, with 90% of inhibition trials occurring in the context of an alcohol image (for the alcohol restraint condition). However, in all three studies, the stop-signal indicating that inhibition is required is not alcohol-related, but rather, a neutral auditory tone. Similarly, studies using Beer-NoGo tasks have a consistent link between alcohol-related images and response inhibition (Bowley et al., 2013, Houben et al., 2011, Houben et al., 2012); however, since participants are instructed to respond according to the superimposed letter stimuli, it could be argued that they do not necessarily process the alcohol and water images. Best et al. (2016) have indeed recently argued that learning of stimulus-stop associations is most effective when images are directly relevant to the task. In the current study we test a new modified Stop-Signal task that directly links alcohol to the need for inhibition. In our new task, participants view images of water with the letters P and F superimposed, and respond with the left and right hand to those letters. The stop-signal was a change in the background image from water to beer. In this way, we combine the urgent need for inhibition afforded by the Stop-Signal task with alcohol-related images, and alcohol itself is the cue for inhibition.
Thus, in the current study we examine alcohol consumption associated with training on one of three inhibitory tasks: not only the Restrained-Stop and Beer-NoGo tasks as investigated previously (Bowley et al., 2013, Houben et al., 2011, Houben et al., 2012, Jones et al., 2011a, Jones et al., 2011b), but also a new Combined task where alcohol itself is the cue for inhibition. We also include a BAI condition as well as a carefully designed Control condition to test the efficacy of inhibitory training against the gold-standard and against non-specific means of reducing alcohol consumption. Participants were randomly assigned to one of these five conditions and completed two sessions one week apart in the laboratory; they were assessed on alcohol consumption in the laboratory (in a taste test administered after completing the inhibitory training, control task, or BAI), and outside the laboratory (consumption in the week before compared to after completion of the inhibitory training, control task, or BAI). For weekly consumption, we expected that alcohol consumption would reduce between sessions due to non-specific effects (i.e., a main effect of time), and additionally we expected that inhibitory training and the BAI would produce significantly greater reductions in consumption compared to the Control condition (i.e., significant time by condition interactions), with the largest reductions observed in the new Combined task. For taste test consumption, we expected that inhibitory training and the BAI would be associated with reduced alcohol consumption relative to Controls (i.e., condition main effects).
Section snippets
Participants
Participants were 114 adults recruited via advertisements on campus and from online research participation websites at the University of New South Wales and University of Wollongong (site of testing was included as a factor in preliminary analyses and not found to significantly alter the results). Participants were eligible to participate if they were aged 18–30, liked beer, consumed at least 4 standard drinks in the week prior to testing (1 Australian standard drink = 10 g alcohol), reported that
Results
Table 1 displays demographic information and behavioural task performance for each of the conditions. Participants assigned to each condition were well-matched on sex ratio (χ2 = 1.188, df = 4, p = 0.880; sex was included as a factor in preliminary analyses but was not found to significantly alter the results), proportion of right-handed participants (χ2 = 7.972, df = 4, p = 0.093), time of testing (F(4,109) = 0.474, p = 0.755), age (F(4,109) = 0.202, p = 0.937), BIS-11 score (F(4,109) = 0.617, p = 0.652) and AUDIT
Discussion
We aimed to test whether training on an established or new version of inhibitory task produced significant reductions in alcohol consumption, after accounting for the effect of simple assessment of drinking. Our primary outcome measures were alcohol consumption in the week after compared to the week before the inhibitory training/intervention, and in an immediate post-intervention bogus taste test.
The sample recruited was appropriate for the research question. The drinking habits of young adult
Conflict of interest
No conflict declared.
Contributors
JS, SJ, KH and MF were responsible for the study concept and design. JS and ND completed data collection. JS drafted the manuscript. All authors provided critical revision of the manuscript for important intellectual content, and approved the final version for publication.
Role of funding source
This study was funded by an Australian Rotary Health Postdoctoral Research Fellowship to JS. The National Drug and Alcohol Research Centre at the University of NSW is supported by funding from the Australian Government under the Substance Misuse Prevention and Service Improvements Grants Fund. Funding sources had no role in the study design, collection, analysis or interpretation of data, in the decision to publish, or in the writing of the report.
Acknowledgements
Thanks are due to Mr Tony Kemp for writing the stimulus presentation programs, and Mr Andrew Jones for his comments on an earlier draft.
References (62)
- et al.
The effects of inhibitory control training on alcohol consumption, implicit alcohol-related cognitions and brain electrical activity
Int. J. Psychophysiol.
(2013) - et al.
Resisting temptation: decreasing alcohol-related affect and drinking behavior by training response inhibition
Drug Alcohol Depend.
(2011) - et al.
Reward, interrupted: inhibitory control and its relevance to addictions
Neuropharmacology
(2014) - et al.
The effects of priming restrained versus disinhibited behaviour on alcohol-seeking in social drinkers
Drug Alcohol Depend.
(2011) - et al.
Inhibitory control training for appetitive behaviour change: a meta-analytic investigation of mechanisms of action and moderators of effectiveness
Appetite
(2016) - et al.
Anxiety-linked attentional bias and its modification: illustrating the importance of distinguishing processes and procedures in experimental psychopathology research
Behav. Res. Ther.
(2016) - et al.
Conflict and inhibition in the cued-Go/NoGo task
Clin. Neurophysiol.
(2011) - et al.
Electrophysiological activity underlying inhibitory control processes in normal adults
Neuropsychologia
(2006) - et al.
Deficits in behavioural inhibition in substance abuse and addiction: a meta-analysis
Drug Alcohol Depend.
(2014) - et al.
Does inhibitory control training improve health behaviour? A meta-analysis
Health Psychol. Rev.
(2016)
DSM-5. Diagnostic and Statistical Manual of Mental Disorders
AUDIT: The Alcohol Use Disorders Identification Test: Guidelines For Use In Primary Health Care
Behavioural inhibition, sustained attention, and executive functions: constructing a unifying theory of ADHD
Psychol. Bull.
The effect of non-specific response inhibition training on alcohol consumption: an intervention
J. Addict. Res. Ther.
Synthesizing standardized mean-change measures
Br. J. Math. Stat. Psychol.
Reduction of alcohol consumption by brief alcohol intervention in primary care: systematic review and meta-analysis
Arch. Intern. Med.
Should I stop or should I go? The role of associations and expectancies
J. Exp. Psychol. Hum. Percept. Perf.
Associatively mediated stopping: training stimulus-specific inhibitory control
Learn. Behav.
Power failure: why small sample size undermines the reliability of neuroscience
Nat. Rev. Neurosci.
Absence of evidence or evidence of absence: reflecting on therapeutic implementations of attentional bias modification
BMC Psychiatry
Subject reactivity effects and alcohol treatment outcome research
J. Stud. Alcohol
A power primer
Psychol. Bull.
The effectiveness of cognitive bias modification interventions for substance addictions: a meta-analysis
PLoS One
Strategies and mechanisms in nonselective and selective inhibitory motor control
J. Exp. Psychol. Hum. Percept. Perf.
Executive functions
Annu. Rev. Psychol.
Effects of noise letters upon the identification of a target letter in a nonsearch task
Percept. Psychophys.
Questionable classification of studies and choice of outcome measures
PLoS One
Toward an understanding of follow-up research with alcoholics
Psychol. Rep.
Exercising your brain: a review of human brain plasticity and training-induced learning
Psychol. Aging
Measuring individual differences in implicit cognition: the implicit association test
J. Pers. Soc. Psychol.
Cited by (16)
Inhibitory control training
2020, Cognition and Addiction: A Researcher's Guide from Mechanisms Towards InterventionsNeuroplastic and cognitive impairment in substance use disorders: a therapeutic potential of cognitive stimulation
2019, Neuroscience and Biobehavioral ReviewsCitation Excerpt :Again, positive results are generally found not only in cognitive performance (Alcorn et al., 2017; Snider et al., 2018), but also in the reduction of drug consumption (Back and Brady, 2010; Black and Mullan, 2015; Houben et al., 2011a, b; Smith et al., 2017) and craving (Kaag et al., 2018; May et al., 2010). Most studies were carried out in alcohol and nicotine users who were engaged in cognitive training programs focused on working memory (Houben et al., 2011b; Kaag et al., 2018; Snider et al., 2018), inhibitory control (Houben et al., 2011a; Smith et al., 2017), planning-ability (Black and Mullan, 2015), and visuospatial skills (May et al., 2010). In these interventions, benefits were found by comparing these subjects with other patients with SUD that engaged in standard treatment or on the waitlist (Table 2).
Assessment of alcohol intake: Retrospective measures versus a smartphone application
2018, Addictive BehaviorsCitation Excerpt :It is equally probable, however, that participants curbed their drinking in response to using CNLab-A, even though no feedback pertaining to intake or national drinking guidelines was provided. Indeed, reductions in alcohol consumption due only to measurement have been reported in other studies (McCambridge & Kypri, 2011; Smith, Dash, Johnstone, Houben, & Field, 2017). Future studies should seek to examine further how drinking and/or responding changes over time in response to app-based measurement; a multi-level modeling approach, as opposed to the repeated measures ANOVA employed in this study, might be a more pertinent means of shedding light on this phenomenon.
From cookies to carrots; the effect of inhibitory control training on children's snack selections
2018, AppetiteCitation Excerpt :Participants are trained to perform a motor response (keyboard press) when presented with a go signal and to inhibit this response when presented with a stop (or no-go) signal. By consistently pairing unhealthy stimuli (e.g., chocolate, alcohol) with motor inhibition, ICT has been found to reduce selection and consumption of unhealthy foods and beverages in adult populations (Veling, Aarts, & Stroebe, 2013; Allom et al., 2016; Jones et al., 2016; but see also; Smith, Dash, Johnstone, Houben, & Field, 2017) and has even led to weight loss in overweight young adults and community samples (Veling, van Koningsbruggen, Aarts, & Stroebe, 2014; Lawrence et al., 2015b). The mechanisms behind such training are thought to be twofold; firstly, consistently associating a stimulus with inhibition encourages the development of stimulus-stop associations, helping to automatize inhibitory control in response to that stimulus (Lenartowicz, Verbruggen, Logan, & Poldrack, 2011; Verbruggen & Logan, 2008; Verbruggen, Best, Bowditch, Stevens, & McLaren, 2014).