Australians have high rates of alcohol consumption especially among younger populations [
1]. In particular, younger populations frequently engage in high-risk single-session alcohol drinking, frequently termed ‘binge’ drinking. A recent survey revealed that two out of five young drinkers regularly engage in binge drinking putting them at significantly increased risk of alcohol-related injury [
1]. In addition, 31.7% of Australians aged 18 and19 years and 26.9% of Australians aged 20 to 29 years are more likely to engage in risky drinking patterns leading to an increased vulnerability to alcohol-related harm over their lifetime when compared with older age groups. In addition, excessive alcohol consumption has had a negative impact on Australian economic and social outcomes [
2]. For example, treatment of alcohol-related illnesses and injuries is estimated to cost 14.352 billion Australian dollars per year [
3]. Excessive alcohol consumption is also related to maladaptive health outcomes such as diabetes, heart disease, asthma, risk of cancer, mortality, and mental illness [
1,
4].
The problem of excessive alcohol consumption is particularly exacerbated among student populations. Rates of high-risk alcohol consumption patterns such as ‘binge’ drinking are often higher in University and college student samples compared with non-student samples of the same age [
5-
7]. Research has shown that undergraduate students who consume excessive alcohol are vulnerable to profound acute and chronic harms, such as drink driving, substance abuse, alcohol-related injuries, violence, and alcohol dependence [
4,
8,
9]. Studies have also demonstrated that undergraduate students who drink excessively are more likely to perform poorly in their studies [
10,
11] and subsequently drop out from university [
12].
A potential solution to reduce prevalence of excessive alcohol consumption among undergraduate students is the development and implementation of behavioral interventions based on social psychological theory and models of behavior change. Many interventions adopting these kinds of approaches have been developed and applied to reduce alcohol consumption [
13-
17]. Despite the relative success of these interventions, two problems exist. First, many of the interventions adopt multiple theories and, as a consequence, multiple intervention techniques. Although the resulting intervention may have demonstrated efficacy in changing the primary behavioral outcome variable, and passed ‘fit for purpose’ on that basis, the intervention itself is not revealing in terms of the precise mechanism by which the intervention exerts its effects and does not permit the isolation of the individual techniques so that their relative independent effectiveness on the behavioral outcome can be ascertained [
18,
19]. This means that the interventions are comparatively silent on exactly what components of the intervention ‘work’ in terms of changing behavior and ‘how’ the components exert their effects [
20]. Recent research on the components of behavioral interventions has called for better intervention designs to isolate the individual components or
techniques that are effective in bringing about a change in behavioral outcomes and to explain the mechanisms behind the effects through appropriate mediator variables [
21-
24]. As a consequence, interventions need to adopt factorial designs examining individual intervention components in isolation, and in synergy, to ensure that the independent and interactive effects of the intervention techniques can be isolated and to include measures of the psychological components that are proposed to mediate the effects of the intervention on behavioral outcomes [
18,
19].
Taking these considerations into account, the aim of the present study is to develop a theory-based intervention that includes multiple intervention techniques from two different social psychological theories of behavior change in a randomized-controlled factorial design to reduce alcohol consumption in undergraduate students. The intervention will include appropriate theory-based mediator variables of target components associated with the intervention techniques to explain the processes behind the proposed intervention effects. The intervention will make a unique contribution to knowledge by not only demonstrating significant reduction in the primary and secondary outcome variables, namely undergraduate alcohol consumption and binge drinking occasions, but demonstrate in independent and interactive effects of the intervention techniques on the outcome and the appropriate mediator variables. The intervention will be driven by two emerging theoretical approaches relevant to individuals’ self-regulation of behavior, namely, the mental simulation approach and the strength or resource-depletion model of self-control. In the following sections, we outline these theoretical approaches, highlight their relevance to reducing alcohol consumption, and introduce the intervention components and the psychological constructs that the components are hypothesized to change and, therefore, provide a mechanistic explanation of the effectiveness of the intervention.
Imagery, mental simulations, and behavioral engagement
There has been a recent resurgence of interest in the efficacy of visualization, rehearsal and imagery strategies in promoting behavior change [
25,
26]. The effectiveness of imagery-based strategies stems from the tenets of Bandura’s [
27,
28] social cognitive theory. According to the theory, imagery, particularly imagining or visualizing the self engaging in a desired behavior, leads to behavior change because it provides a form of a ‘self-model’ or vicarious experience, which is central to the development of self-efficacy. Self-efficacy is, therefore, one potential mechanism by which such imagery strategies affect a change in behavior. Another mechanism by which imagery-based techniques may promote behavioral engagement is through promoting greater importance of the goal [
25,
26] and promoting better accessibility of cues to action and goals [
29,
30].
One class of imagery-based intervention techniques is
mental simulations, which are defined as imagining and rehearsing future events. There are two kinds of mental simulations: outcome simulation and process simulation. Outcome simulation involves imagining attainment of a targeted goal while process simulation requires imagining and rehearsing the steps required to achieve the goal [
31]. Mental simulations have been shown to be effective in evoking behavior change in diverse contexts such as studying for exams [
31], fruit consumption [
30], intention to buy a product [
32], anxiety reduction [
33], and alcohol consumption [
34]. The effect sizes of mental simulation interventions is generally small-to-medium, with a medium effect size reported for the research on alcohol consumption of employees, the context that is most closely aligned to that of the current study [
34]. The effect size is comparable with the effect sizes of intervention techniques designed to change intentions [
35] and self-efficacy [
36] in research on health behavior change.
A key approach to understanding intervention effectiveness is to identify the psychological variables that mediate the effects of interventions on behavioral outcomes [
37,
38]. Researchers have turned to social cognitive [
39-
46] and integrated models of motivation [
18,
47-
56] in order to identify the key mediators of interventions. Research has suggested that imagery-based and mental simulation intervention techniques in health behavior exert their effects through changes in motivation [
30,
57], intentions and attitudes [
33,
58] and planning [
31], although there are few studies that have conducted formal mediator analyses. Knauper et al. [
30] revealed that the effectiveness of a mental imagery intervention was mediated by motivation and Pham and Taylor [
31] demonstrated that planning was a key mediator of the effect of mental simulations on studying behavior, which is highly salient given increased recent interest in planning interventions in health contexts [
59]. Interestingly, no studies have found self-efficacy to be a mediator of the effects of imagery on behavior despite hypothesizing it as a theoretically-relevant mediator, measuring the construct and including it as a mediator in analyses [
30,
33,
57]. Research in other domains, such as injury prevention, have found effects for imagery interventions on self-efficacy and behavioral outcomes, implicating it in the process by which imagining processes and outcomes may effect behavior change [
60,
61]. Overall, research in the health domain has presented consistent evidence to support the effectiveness of imagery-based interventions such as mental simulations on health related behavior. On the proposed mechanisms, however, the evidence is less conclusive with motivation and self-efficacy as identified as possible mechanisms for mental simulation effects.
The ‘Strength’ or ‘Resource Depletion’ Model of Self-control
Self-control is another construct that has been identified as an important factor associated with the regulation of health-related behavior [
62-
64]. Self-control is defined as the capacity to control or regulate impulses, temptations, or ‘dominant’ responses and to overcome well-learned, ingrained habitual actions for some goal-directed alternative [
65-
72]. Much of the research on self-control in health domains has focused on trait conceptualizations of self-control and has demonstrated that good’ self-control is associated with numerous adaptive health-related behaviors and outcomes [
62,
73-
77].
An alternative perspective on self-control is offered by the ‘strength’ or resource depletion model which conceptualizes self-control as limited resource that permits individuals to engage in acts of self-control, but only for a finite period after which the resource becomes depleted leading to impaired self-control capacity unless an individual is able to rest and recover [
64,
78-
80]. The state of reduced self-control capacity or ‘strength’ is known as
ego-depletion. Research adopting the model has typically adopted an experimental procedure, known as the dual task paradigm, to test model effects [
78,
81]. The paradigm requires individuals to engage in two consecutive tasks, for experimental group participants both tasks require self-control while for control group participants only the second task requires self-control. To the extent that experimental participants’ performance on the second self-control task is impaired, we have sharp confirmation of the ego-depletion effect. Research has supported the ego-depletion effect across multiple studies and has shown that the depletion effect occurs for tasks in multiple domains of self-control indicating that the resource is a unitary, generalized effect rather one that is confined to particular tasks [
63,
70,
81-
83].
An important additional hypothesis derived from the strength model is the training effect. A growing number of studies have demonstrated that repeated practice on self-control tasks improves regulatory capacity. According to the strength model, engaging in tasks that demand self-control on a regular basis can improve self-control capacity by ‘building up’ additional resources that can be made available or by making the application of the resource more efficient [
84]. Research has demonstrated that regular practice on self-control tasks in laboratory and field settings leads to better performance on self-control tasks in the laboratory [
84-
86] as well as health-related behaviors requiring self-control [
87-
90] including alcohol consumption [
91-
93]. The strength of self-control training effects has been shown to be of medium effect size which is comparable to other interventions such as those targeting intention [
35] and self-efficacy [
36]. For example, a meta-analysis of self-control training on self-control task performance found a medium-sized effect [
81] and a recent meta-analysis of response-inhibition training, using similar tasks to those used in the current study, on health behavior found a small-to-medium sized effect [
94].
Training studies have demonstrated that the practice of self-control promotes behavior change in a number of contexts, providing indication of the generalized, unitary nature of self-control resources [
85,
87,
93]. However, a key unresolved issue is the mechanism that drives the direction and allocation of self-control resources to increase behavioral enactment. It is unlikely that individuals will commit self-control resources toward engaging in behaviors for which they have no motivation. The resources would more likely be allocated elsewhere, such as toward enacting behaviors that they are motivated to perform. This gives rise to the possibility is that behavior change will be more effective if participants can direct their self-control efforts towards a particular target behavior for which they are highly motivated [
95,
96]. Interventions might, therefore, be more effective if means to increase motivation toward behaviors could be delivered alongside the practice on self-control tasks in a factorial design giving participants the opportunity and drive to direct their self-control resources toward that specific behavior. Instilling increased motivation may, therefore, be effective in focusing individuals on directing their self-control resources towards specific target behavior.
The current research
The purpose of the current research is to develop a brief theory-based intervention adopting imagery-based motivational and self-control training behavior-change components and a randomized controlled factorial-design that will lead to a reduction in alcohol consumption among undergraduate students, an at-risk population, over a four-week period. We expect that the influence of motivational and self-control components on promoting alcohol reduction will interact, such that individuals who are both motivated and provided with training to enhance their self-regulatory capacity will exhibit the greatest reduction in their alcohol consumption. Why would training self-control improve an individual’s capacity to reduce their future alcohol consumption? Our position is that training individuals to inhibit well-learned, ingrained responses with little cognitive control or conscious thought (i.e., so called ‘automatic’ or ‘habitual’ behaviors) will have abroad impact on behaviors that are dependent on such automatic processes [
69,
72]. This proposition is consistent with recent trends in theory on behavioral enactment which indicate that actions are controlled by two systems: reflective and impulsive [
97]. The reflective system is a deliberative pathway to action in which individuals decide on a course of action as a result of conscious consideration of the costs, benefits, consequences and outcomes of the action. Such a system is controlled, slow, reasoned, and conscious, often termed a ‘cool’ system by some theorists [
98]. This is contrasted with the impulsive system which is a more spontaneous, automatic pathway to action in which individuals act in response to well-learned cues or heuristics that require little conscious involvement or deliberation. The impulsive system is fast, non-conscious, and automated and often referred to as a ‘hot’ system [
98].
As actions controlled by the impulsive system often occur outside the individual’s awareness and in responses to well-learned cue-response pairings, it is difficult to override and change such actions i.e. to break the cue-response link. It often takes considerable self-control or capacity to inhibit responses to overcome the automated link. From the perspective of the strength model, self-control, that is the capacity to inhibit the automated response, is a limited resource permitting individuals to inhibit their responses for a finite period until the resource becomes depleted limiting subsequent capacity for inhibition. An important feature of the self-control ‘resource’ is that it is ‘domain general’ i.e. it is a generalizable resource that enables individuals to control their behavior in multiple domains. This has been shown in many studies in which individuals engaged in a task that requires them to inhibit their self-control in one domain results in impaired response inhibition in another. Importantly in the current context, exerting self-control on laboratory-based tasks that require the inhibition of responses leads to reduced capacity to inhibit responses when presented with tempting behaviors in health-related domains such as resisting alcohol in social drinkers or tempting foods in people with low eating restraint. Consistent with this line of research, studies have revealed that training on tasks that require self-control i.e. the ability to inhibit responses will improve response inhibition capacity and provide individuals to inhibit cue-response pairings more effectively. And this effect also appears to be domain general, consistent with the strength model [
78] and the reflective-impulsive model [
97]. Training on self-control tasks which require regular inhibition of a pre-potent response is, therefore, hypothesised to improve generalized capacity for self-control as shown in previous studies [
84,
88]. With individuals who have sufficient motivation to reduce their alcohol intake, improving response inhibition capacity is expected to improve capacity to inhibit the temptation to drink. This may be relevant in situations where they may be tempted to drink more than usual. They can therefore bring their improved generalized capacity to bear on reducing their alcohol consumption is situations where they may be tempted to drink to excess. Their additional response inhibition capability afforded to them by virtue of the training would provide sufficient capacity to override the automatic, cue-driven response to stimuli to drink alcohol.
We therefore propose that self-control training will only have a substantive effect on behavioral outcomes if individuals are motivated to change their behavior. Participants whose self-control capacity has been trained, but have little or no motivation to reduce their alcohol consumption are less likely to direct the increased self-control capacity gleaned from training toward that particular behavior. They may direct their efforts toward behaviors to which they are more motivated instead. We therefore expect that the combined manipulation of self-control training and motivational imagery-based intervention components, namely, mental simulations, to be more effective in changing behavior than either of the components alone. Importantly, the factorial design adopted in the present study permits the evaluation of the independent and synergistic effects of each of the intervention components on alcohol consumption. In addition, we expect the intervention to be highly acceptable for use in public health promotion campaigns due to its low response burden and highly practical, accessible, and cost-effective means of delivering the intervention.
We will administer outcome measures of two forms of alcohol consumption in the current research: total alcohol consumption and frequency of ‘binge’ drinking. Our primary outcome variable in the current research is overall self-reported alcohol consumption by undergraduate students. We cannot expect students to curb their alcohol drinking altogether, so the target outcome identified for participants in the current study will be keeping alcohol consumption within the guideline limits specified by the Australia National Health and Medical Research Council (NHMRC). The NHMRC guideline limits on safe alcohol consumption are 14 standard drinks (each standard drink is equivalent to 12.5 ml of pure alcohol) per week. The guideline limits will be clearly outlined to participants in advance of the research commencing and our purpose was to ensure that students did not exceed this limit as it is associated with long-term (chronic) harm [
4]. We do, however, recognise that the guidelines reflect limits aimed at reducing chronic harm and that other patterns of alcohol consumption prevalent in students may also present a serious threat to health. For example, the consumption of 14 standard drinks on a single occasion, once per week may mean that an individual’s consumption falls within NHMRC overall guideline limits, but would constitute increased risk of acute harm. High-risk single-session alcohol consumption, also known as ‘binge’ drinking, defined as consuming more than 4 standard drinks on a single drinking occasion by the NHMRC [
4], is related to substantially increased health risks such as unintentional injury, increased probability of unplanned and unprotected sexual intercourse, and risk of being involved in violence and social disorder [
9,
99,
100]. Given that students are more likely to engage in binge drinking than their non-student peers [
10,
101], binge drinking is considered an important risk factor for alcohol-related harm specific to this population. We will, therefore, include a secondary outcome variable, frequency of occasions of binge drinking, self-reported by participants to account for this risky pattern of alcohol consumption that is likely to be endemic in this population. The variable is defined as the number of single occasions in which an individual’s alcohol consumption exceeded 4 standard drinks [
4].
The intervention will comprise two components, motivational and self-control training. The motivational component of the intervention will comprise a process and outcome mental simulation manipulation in which participants will be required to visualize the steps they need to take in order to reduce keep their alcohol consumption within guideline limits in the next four weeks and the outcomes they will achieve. This task has been adopted in previous studies and the standardized protocol adopted in these studies will be used [
34]. Participants not allocated to receive the motivational intervention component will receive an irrelevant visualization task. The self-control training component will require participants to engage in an online Stroop color-naming task delivered online either by smartphone or personal computer for the duration of the subsequent four-week period. The Stroop task was developed and piloted previously and has been shown to effectively enhance self-control capacity over a four-week training period [
85]. Participants allocated to the self-control training condition will receive a ‘challenging’ version of the Stroop task while participants in the control condition will receive an ‘easy’ version, which is not expected to lead to any substantive improvements on self-control capacity.