Craving: Approaching Avoidance : Part 6

by National Institute of Health

(Page 6 of 7)

Avants and colleagues sought to examine the reactivity of addicts to drug cues using separate assessments of approach ("craving") and avoidance ("aversion") inclinations. The participants, who were on methadone maintenance for heroin addiction while undergoing treatment for cocaine dependence, viewed a videotape depicting persons using cocaine. The participants also were asked to handle their preferred type of cocaine paraphernalia. Before and after exposure to these cues, the participants were asked to use rating scales to respond to the following questions: "How much do you crave cocaine right now?" and "How much does the idea of using cocaine turn you off right now?" Results indicated that overall craving and aversion ratings were negatively correlated at baseline, but they were not significantly correlated after cocaine cue exposure, suggesting that these inclinations vary independently of one another.

Further analysis revealed the presence of four subsamples of patients demonstrating differing response patterns to the cues. One group showed an increase in craving and a decrease in aversion, indicating a clear shift toward the approach quadrant of the evaluative space. Another group showed an increase in craving but no decrease in aversion. Depending on the initial level of aversion, their craving may or may not have been balanced by an equivalent level of aversion, and thus they could be represented in either the approach or the ambivalence quadrant. A third group showed no increase in craving, but a decrease in aversion, indicating a shift away from the avoidance and toward either approach or indifference, depending on the initial level of craving. Finally, a group of "nonresponders" showed no increase in craving and no decrease in avoidance.

When analyzed together, these results indicated that levels of craving and aversion, constructs that can be readily mapped onto the dimensions of the evaluative space, could be altered by exposure to substance-related cues. Moreover, evidence suggests that these changes not only varied independently, but could also be influenced by individual differences. For example, further analyses revealed that individual differences in two areas distinguished nonresponders from those who showed the most pronounced shift toward the approach quadrant. Relative to nonresponders, those who reported an increase in craving and a decrease in aversion also perceived cocaine to be more reinforcing and less punishing, and they saw themselves as less able to avoid using cocaine in certain high-risk situations. Thus, the work of Avants and colleagues provided general support for a multidimensional, ambivalence model of reactivity to drug cues and helped establish the model's clinical relevance by noting that observed shifts appeared to be associated with variables such as perceptions of the net benefit of drug taking and of self-efficacy in coping with high-risk situations.

More direct evidence in support of this model stems from research conducted in our own laboratory. Using a large sample of undergraduate students, we measured separate approach and avoidance reactions to photographic stimuli depicting several kinds of consumable substances, including alcohol and cigarettes. Respondents with various patterns of routine usage of these substances were asked to view slides, responding after each one to the following questions: "How much do you want to consume the item pictured in the slide?" and "How much do you want to avoid consuming the item pictured in the slide?" Whereas ratings for approach and avoidance were significantly negatively correlated in abstainers from both categories (nonsmokers and nondrinkers), results for light and moderate alcohol drinkers and for occasional and daily smokers indicated no significant correlations between approach and avoidance. This independence of variation argues for the need to separate these two dimensions and for the potential of the competing inclinations to coexist.

Data from a college student sample also yielded some particularly interesting information on reactions to cigaretterelevant stimuli as a function of desire to change behavior. Participants were divided into three groups: nonsmokers, regular smokers not trying to quit, and regular smokers currently trying to quit. We then compared the participants' approach and avoidance reactions to cigarette stimuli. Results indicated that nonsmokers were characterized by a combination of low approach and high avoidance inclinations, representing the avoidance quadrant in the model. In contrast, regular smokers not trying to quit reported high approach and low avoidance inclinations, placing them in the approach quadrant in the model. Most interesting was the finding that smokers who were trying to quit reported high approach and high avoidance inclinations characteristic of the ambivalence quadrant. Thus, it was not the level of craving, but rather avoidance, that identified smokers who were ready to change.

In addition to evidence from research on reactions to drug cues, data from studies investigating memory processes are consistent with a multidimensional or ambivalence perspective. For instance, Leigh and Stacy reported that memory associations related to both reinforcing and punishing consequences of alcohol use can be activated by the same type of cognitive task. They examined associative memory and alcohol use and demonstrated that participants' histories of alcohol use (quantity and frequency of drinking) predicted their associative memory responses to both positive and negative outcomes of drinking. When given a list of positive and negative outcomes not specific to alcohol ("feeling good"; "forgetting problems"; and "being more social" versus "feeling sick," "being depressed," and "losing control"), heavy social drinkers generated significantly more alcohol-specific responses for both types of outcomes than did light social drinkers. This finding further supports the notion that approach or craving should be integrated with avoidance to reflect the multidimensional nature of responses that seem especially likely to accrue as repeated alcohol use strengthens associations in memory that link alcohol cognitions to both positive and negative consequences of drinking.

Strong memory associations to alcohol-related cues and behaviors are also central to Tiffany's influential cognitive processing model of craving and substance use. According to this model, if practiced regularly, drug use becomes automatized and, like other highly practiced skills, relies on strong memory associations for rapid and effortless execution.

Within the framework of the evaluative space, this response would be characteristic of the craving associated with the approach quadrant. However, Tiffany maintains that craving does not involve automatic processing. In his model, craving refers to a constellation of responses supported by nonautomatic, effortful cognitive processes activated only if the habitual sequence of drug use behaviors is blocked by limited access to the substance or by an intentional effort to curtail use. Implicit in at least the latter of these scenarios is the experience of ambivalence, operationalized in the model as the simultaneous activation of opposing response inclinations. To what extent can the two views be reconciled?

About the Author

NIH is the nation's medical research agency - making important medical discoveries that improve health and save lives. The National Institutes of Health (NIH), a part of the U.S. Department of Health and Human Services, is the primary Federal agency for conducting and supporting medical research.