Managing craving in the mind’s eye: How can positive and negative imagery influence cocaine craving?
Cognitive-behavioral approaches to addiction have long utilized mental imagery to help individuals manage cravings for alcohol and other drugs. In this study, researchers investigated how mental imagery might help people with severe cocaine use disorder manage acute craving episodes, and how different types of imagery might decrease or increase cocaine craving.
WHAT PROBLEM DOES THIS STUDY ADDRESS?
Craving is a diagnostic and cardinal feature of cocaine use disorder and is a key target of first-line treatments for cocaine use disorder such as Cognitive Behavioral Therapy (CBT). CBT promotes development of new skills and strategies that help individuals manage cravings, support addiction recovery, and prevent relapse.
Most CBT manuals include training around the use of imagery to manage craving. For instance, a patient might be encouraged to imagine the glass of wine they are craving as being caustic like bleach, or they may be coached to bring to mind positive images of waking relieved and refreshed the morning after not drinking without a hangover.
While cognitive-behavioral imagery strategies are commonly used in CBT, no previous study had explored how the use of imagery may help or hinder individuals’ efforts to manage cocaine cravings. Further, no previous study had explored in a granular way whether certain kinds of imagery, or the level of detail of the imagery, are better or worse than others for managing craving in this population. Greater understanding of the role mental imagery can play in managing craving would help to improve substance use disorder treatment. In this study, the authors explored how four kinds of imagery influenced cocaine craving in a sample of people with cocaine use disorder.
Based on the theory that negative/aversive memories or thoughts can exacerbate craving (because the mind seeks relief from these thoughts through its learned associations between substance use and emotional relief), and conversely that positive thoughts can reduce craving (perhaps through distraction or by reminding the individual of a future goal that can be achieved by resisting the urge to use a substance), the authors predicted the following: 1) Past and future-positive imagery would be associated with lower acute cocaine craving; 2) Past and future-negative imagery would be associated with greater acute cocaine craving; and 3) Higher imagery detail would be associated with lower acute craving after a positive imagery task and higher acute craving after a negative imagery task.
HOW WAS THIS STUDY CONDUCTED?
This was a randomized trial with 31 participants with cocaine use disorder (see below for more details) who were exposed to cocaine stimuli and instructed to bring to mind either, 1) a past-negative autobiographical image related to their cocaine use, 2) a past-positive autobiographical image from a time before they began to use cocaine, 3) a positive image of a desired future if they initiate recovery, or 4) a negative image of a feared future if their cocaine use disorder persists, to determine how these different kinds of imagery influenced cocaine craving immediately following the imagery exercise.
To personalize cue exposure intended to induce craving, participants attended an initial study intake session where they were shown cocaine paraphernalia and asked which items were most evocative to them. Objects included plastic bags used to hold drugs, pipes, lighters, scrapers, filters and unused, packet-sealed needles/syringes. At this session, participants also identified specific images/memories for each of the four imagery conditions. Then, a week later, participants completed a 2-hour laboratory session. Before this laboratory session, they were asked to abstain from cocaine, illicit drugs, and alcohol for as long as they could prior to the session.
At the beginning of the laboratory session, participants were randomized to one of the four conditions, then asked to rate their current past 24-hour craving using the Craving Experiences Questionnaire – strengths version. Then participants watched a video depicting the sequence of actions involved in preparing crack-cocaine for consumption (1-minute) and asked to hold cocaine related objects they had previously selected as being particularly evocative or triggering for them (2 minutes). Immediately following this, participants again reported their craving before being instructed to bring to mind the past/future, negative/positive mental image they had been randomly assigned to. They were told to close their eyes, and to describe the event, thoughts, and emotions in detail (5 minutes). Finally, participants again reported their craving. Afterwards, participants were offered food, asked to rest, and invited to listen to a relaxation tape or complete a guided relaxation exercise (30 minutes).
All participants met diagnostic criteria for cocaine use disorder (with crack-cocaine use), and 25 also met diagnostic criteria for opioid use disorder. The sample (N= 31) was 74% male and on average 42 years of age; 58% of participants identified as Black and 42% as White. The majority of participants (90%) had severe cocaine use disorder (i.e., 6-11 symptoms), with the remainder meeting diagnostic criteria for moderate cocaine use disorder (i.e., 4-5 symptoms), and 87% had used cocaine in the past 14 days. At baseline, on average, participants endorsed a Craving Experience Questionnaire score of 56 out of a total possible score of 110.
WHAT DID THIS STUDY FIND?
The craving cue exposure paradigm was successful for eliciting craving in most but not all participants.
Overall, craving ratings from pre- to post-cue exposure increased 15%, with three of the four conditions showing an expected increase in craving, suggesting the study’s craving elicitation manipulation worked. Notably however, the future-negative group actually showed an 8% decrease in craving from baseline to post cue exposure. It can’t be known why this was the case, however, this group did have fairly high session baseline craving scores (74/110), which represented 13% greater craving at baseline compared to the next highest group. They also had a high standard deviation of craving at session baseline suggesting outliers may have been affecting results.
Past-positive and future-positive thoughts were associated with absolute reductions in craving following cocaine cue exposure, but findings for past-negative and future-negative thoughts were not as expected.
As the authors predicted, both past-positive (positive memories before developing addiction) and future-positive (a positive image of a desired future in recovery) thoughts were associated with absolute reductions in craving following cocaine cue exposure; 27% and 25% reductions respectively. Notably however, contrary to the authors’ prediction, past-negative thoughts also reduced craving in absolute terms by around 19%, while future-negative thoughts did not markedly affect craving (craving increased slightly by 3%).
The authors also explored these changes using a statistical technique that allowed them to account for the smaller sample size in this study. Findings showed that the future-negative condition had significantly greater craving from the past-positive group following the imagery task (i.e., there was less than a 5% probability that the observed difference was a result of chance). Though the authors didn’t look at it in their analyses, it is likely the difference between the future-negative condition and future-positive condition was also significantly different based on the direction and raw magnitudes of effect.
Counter to the researchers’ prediction, there was no evidence that level of imagery detail (positive or negative) was related to craving.
WHAT ARE THE IMPLICATIONS OF THE STUDY FINDINGS?
In this study, exposure to cocaine stimuli evoked craving in most participants, though on average those in the future-negative condition (a negative image of a feared future if cocaine use disorder persists) endorsed decreases in craving from session baseline to immediately following cocaine cue exposure. In absolute terms, past-positive, past-negative, and future-positive imagery all appeared to reduce craving, while future-negative imagery increased craving slightly.
In the theoretical framework of this study the authors posited that negative imagery would increase craving. While this could be true under many circumstances, individuals commonly use negative imagery to reduce craving, and this is in fact a strategy commonly taught in CBT approaches for substance use disorder. For instance, thinking about something bad that happened, or could happen, because of substance use could buffer against subsequent substance use by reminding the individual about potential consequences of use based on their own past negative experiences. Individuals in treatment often find this to be an effective strategy for managing cravings associated with ‘euphoric recall’ of drug use in which one may be thinking only of the positive effects of drug use in that moment, and not thinking through the negative consequences that might ensue shortly thereafter (later that night or the following morning). This is often referred to as ‘thinking the drink (or drug) through’. The authors might have been picking up on this effect when they observed that past-negative imagery decreased craving. Consequently, this may present a flaw in the authors’ theory of how different types of emotionally-valenced memory affects current cravings.
From this perspective, it is less clear why the future-negative condition did not also experience reductions in craving. Given this group had higher session baseline craving than all other conditions (likely a result of the small sample size of the study), it could be that, on average, this group was simply experiencing higher craving to begin with, so regressed to the mean (a statistical phenomenon that commonly arises when a sample score on a variable is extreme and is thus likely at a future point to be closer to the mean).
Though it can’t be known from these data, it is possible that past-positive and future-positive imagery reduce craving through a similar pathway. It is plausible that past-positive imagery of a time before cocaine use might stand in for what individuals hope to get from recovery – i.e., a return to some of those positive experiences.
One challenge of addiction recovery is that it can be difficult to imagine positives in the more distant future, as the rewards of substance use are typically so immediate and evocative. A large literature on what is known as delay discounting shows that people with active addiction and in early recovery tend to favor immediate rewards, even if these rewards come at a longer term cost. It may be that imagery could help individuals in early recovery overcome this tendency to favor immediate rewards. At the same time, this cognitive tendency to favor immediate reward could also be neurophysiologically based due to post-acute withdrawal deficit phenomena making it difficult for some people to utilize imagery strategies when experiencing these strong cravings. There may be added benefit of combining imagery training with other neurocognitive strategies like working memory training that has been shown to reduce the degree to which individuals with addiction discount future rewards or utilize other cognitive strategies (e.g., cognitive distraction or behavioral evasion/safety-seeking).
As noted by the authors:
- Despite the fact the authors used a statistical technique to handle small sample sizes, their ability to detect statistical effects was still very limited. Related, although a randomized design was used, the small sample size means that randomization is highly unlikely to have evened out all pre-existing differences among groups across the studied levels of the independent variable. Indeed, although participants were randomized to conditions, those in the future-negative group appeared distinct from participants in the other conditions in terms of having higher session baseline craving. It is not clear how this may have affected the study’s findings. There are also likely to be unmeasured variables that differed among sub-groups.
- The longitudinal model was basic, and other baseline covariables like demographic factors were not included.
This study suggests that bringing to mind vivid images of past- and future-positive events, as well as past-negative images related to cocaine use might be a helpful strategy for managing acute craving for cocaine among individuals with cocaine use disorder. Though the sample size of this study was small and there were some differences between groups in craving at baseline that may have affected results in unknown ways, these findings are generally consistent with what is commonly observed in clinical practice, and it is possible these findings would generalize to craving for other substances such as alcohol or opioids.
- For individuals and families seeking recovery: Managing urges to use alcohol and other drugs can be one of the most challenging aspects of early recovery from substance use disorder. This study suggests that bringing to mind past- and future-positive, as well as past-negative images may be a helpful strategy for managing acute craving for cocaine among individuals with cocaine use disorder. Though this study focused on cocaine and cocaine use disorder, it is plausible these findings may generalize to craving for other substances such as alcohol or opioids. Therapists trained in cognitive-behavioral approaches will be well positioned to provide training in the use of imagery—as well as numerous other skills—to help manage craving.
- For treatment professionals and treatment systems: Managing urges to use alcohol and other drugs can be one of the most challenging aspects of early recovery from substance use disorder. This study suggests that bringing to mind past- and future-positive, as well as past-negative images may be a helpful strategy for managing acute craving for cocaine among individuals with cocaine use disorder. Though this study focused on cocaine and cocaine use disorder, it is plausible these findings may generalize to craving for other substances such as alcohol or opioids. While patients might be encouraged to use imagery to manage cravings, there is always the possibility that certain kinds of imagery may exacerbate cravings in some individuals. For this reason, ideally patients would be coached in the use of imagery by a therapist trained in CBT.
- For scientists: This study suggests that bringing to mind past- and future-positive, as well as past-negative images might be a helpful strategy for managing acute craving for cocaine among individuals with cocaine use disorder. The sample size of this study was very small and there were measurable differences between randomized groups in craving at baseline that may have affected results in unknown ways. Despite this design flaw, and some other flaws in the theoretical framework guiding the research, findings were generally consistent with what is commonly observed in clinical practice, and it is plausible these findings may generalize to craving management for other substances such as alcohol or opioids. While the use of imagery is not new to clinical practice, research has lagged behind. Larger studies are needed 1) to explore how different kinds of imagery can ameliorate and exacerbate craving, 2) to develop and test strategies that help individuals create vivid recovery-related imagery, 3) to examine the mechanisms through which imagery brings about reductions in craving, and 4) to test whether and which different types of craving management strategies may be optimal at different milestones in the phases of recovery.
- For policy makers: This study suggests that bringing to mind past- and future-positive, as well as past-negative images may be a helpful strategy for managing acute craving for cocaine among individuals with cocaine use disorder. The sample size of this study was very small and despite randomization (intended to equate groups on all confounding variables) there were nevertheless measurable differences between groups in craving directly after randomization at baseline that may have affected results in unknown ways. Despite this design flaw, and some potential flaws in the theoretical framework guiding the research findings were consistent with what is often observed in clinical practice, and it is plausible these findings could generalize to craving for other substances such as alcohol or opioids. While these kinds of craving management strategies are easily taught by qualified clinicians, access to such qualified clinicians remains a challenge.
Lowry, N., Marsden, J., Clydesdale, B., Eastwood, B., Havelka, E. M., & Goetz, C. (2021). Acute impact of self-guided mental imagery on craving in cocaine use disorder: A mixed-methods analysis of a randomized controlled trial. Addiction, 116(9), 2418-2430. doi:10.1111/add.15405