Does Urge To Drink Predict After Treatment?

Damaris J. Rohsenow, Ph.D., and Peter M. Monti, Ph.D.

The urge to drink, also often referred to as craving, is an emotional state in which a person is motivated to seek and use alcohol. In abstinent alcoholics, this urge may contribute to the risk of relapse. Researchers have developed several models—including the conditioned withdrawal model, conditioned appetitive motivational model, social learning model, and information-processing model—to describe the role of urges in relapse. Several studies have evaluated the role of urges in predicting alcoholism treatment outcome and relapse. Some findings indicate that the degree of urge an alcoholic experiences when confronted with a simulated high-risk situation at the end of alcoholism treatment can predict subsequent drinking. Other studies, however, show inconsistent results regarding the role of urges in predicting treatment outcome. Overall, the study results suggest that urges do not necessarily increase the risk of relapse but may actually protect some drinkers against further drinking. KEY WORDS: AOD (alcohol and other drug) craving; AODD (alcohol and other drug dependence) relapse; treatment outcome; AOD abstinence; scientific model; AOD withdrawal syndrome; motivation; social learning theory; CNS (central nervous system) information processing; alcohol cue; coping skills; literature review

esearchers and clinicians have to some laboratory studies, only mod- between craving and relapse. It also long considered the urge to erate (although statistically significant) presents some methods for assessing Rdrink (also commonly called correlations exist between urges and craving as well as the results of studies craving) a key cause of relapse follow- alcohol consumption (Tiffany 1990). ing alcoholism treatment. Accordingly, Finally, clinicians have noted that in craving has been a focus of many treat- many patients, relapse is not necessarily DAMARIS J. ROHSENOW, PH.D., is a ment approaches, a strong concern preceded by urges. research professor of community health in some medication development pro- Few studies actually have investigated and research director of the Addictive grams, and a central aspect of various the role of craving in relapse among Behavior Research Laboratory at the theoretical formulations of alcoholism abstinent alcoholics. Consequently, Center for Alcohol and Studies, and relapse. Some researchers, however, investigations into the role of craving Brown University, and is a career research have raised doubts about the extent to have become an important focus of scientist at the Veterans Affairs Medical which craving plays a role in drinking many research programs that include Center, Providence, Rhode Island. after treatment. Several factors con- not only alcoholics but also smokers tribute to the reevaluation of craving’s and -dependent patients. Those PETER M. MONTI, PH.D., is a professor role in relapse. For example, some analyses should help provide a better of psychiatry and human behavior studies have demonstrated that alcohol understanding of the role of urges in and associate director of the Center for consumption does not necessarily lead both treatment and treatment outcome Alcohol and Addiction Studies, Brown to loss of control over drinking among (Monti et al. in press). University, and is a senior career research recovering alcoholics (Nathan and This article reviews some of the scientist at the Veterans Affairs Medical Lisman 1976). Furthermore, according models used to explain the relationship Center, Providence, Rhode Island.

Vol. 23, No. 3, 1999 225 investigating the relationship between ior (i.e., how much alcohol an animal • Alcoholics in treatment can expect urges and relapse to drinking. Finally, the drinks voluntarily). According to this that alcohol will have pleasurable article discusses the implications of those approach, an animal that drinks more effects but nonetheless still intend results. Most of the theories presented alcohol is considered to have a greater not to drink. here refer to drug abuse in general; the urge to drink. In humans—particu- review of empirical studies, however, is larly in treatment populations—this • Alcoholics can expect that alcohol limited to studies of alcoholics. approach is not appropriate, however, will have negative effects (e.g., make because a discrepancy exists between them more depressed) and yet still the urge to drink and actual alcohol intend to drink. Definition of Craving consumption. This discrepancy results or the Urge To Drink from the conflict between two compet- • Alcoholics can experience urges to ing motivations: (1) the motivation drink without any intention to drink. The terms “craving” and “urge to drink” to drink (e.g., when thinking about often are used interchangeably in the alcohol’s pleasurable effects) and (2) Again, researchers and clinicians are literature. However, consistent with the concurrent motivation not to drink primarily interested in analyzing and Kozlowski and colleagues (1989), the (e.g., when recalling severe negative enhancing these distinctions. The goal authors of this article prefer using the consequences of drinking). In fact, a of their investigations is to identify sit- terms “urge to drink” and “desire to recovering alcoholic’s ability to refrain uations in which abstinent alcoholics drink” in both research and practice, from drinking despite urges to drink may have urges or positive expectations because the meaning of the term “crav- is of paramount interest to clinicians, regarding alcohol’s effects but neverthe- ing” frequently is ambiguous or incon- and the strengthening of that ability less still intend not to drink. sistent. For example, many researchers, is a key treatment focus. Some of the debate about the role clinicians, and alcoholics use the term Similarly, urges are not the same of urges in drinking behavior may stem “craving” only to refer to an intense as psychophysiological reactions (e.g., from insufficient distinctions being desire to drink, whereas others use the changes in heart rate, blood pressure, made between the role of urges in cur- term to refer to urges with a wide range skin conductance, and salivation) in rent drinkers as opposed to abstinent of intensities. Furthermore, researchers situations that presumably elicit urges. alcoholics, in whom urges may increase do not yet know whether even mild Most of these responses are general the risk for relapse. A current drinker urges or only strong urges contribute to measures of arousal, rather than specific may experience frequent strong urges relapse. Consequently, investigators must indicators of urges, and therefore do as a result of recent drinking. In this study the effects of the entire range of not correlate well with urges.1 Salivation situation, the ability of urges to predict motivations for drinking, a concept that in response to images or objects associ- future drinking independent of recent is better served by the terminology “urge ated with alcohol use (i.e., alcohol drinking is unclear. Alternatively, a cur- to drink.” Therefore, this terminology cues) probably reflects conditioned rent drinker may not experience more is used for the remainder of this article. learning processes—that is, responses than a slight urge, because whenever Urge to drink (also called “desire to that are learned when two stimuli (one the drinker feels any desire for a drink, drink,” “need to drink,” “want to drink,” that is neutral and one that causes the he or she can have one with minimal and “missing drinking” in the treatment response) are paired repeatedly. How- delay. Conversely, urges in abstinent and research literature) is generally con- ever, salivating in response to cues is alcoholics may rise to high levels because ceptualized as an emotional state that is also not the same as an urge, because drinking is prevented (either voluntar- characterized by the motivation to seek people often are not aware of these ily or involuntarily). These urges may and use alcohol (Baker et al. 1987). reactions and thus do not report urges be particularly strong in situations that This motivation, which can be associ- when salivating (Monti et al. 1993b). had been previously associated with ated with either positive or negative Therefore, whereas urges are inherently drinking. This article investigates the emotions (e.g., anticipation of alcohol’s conscious processes, most psychophysi- ability of urges to predict subsequent positive effects or frustration over prob- ological changes are unconscious reac- drinking only in abstinent alcoholics, lems at work), is thought to be predic- tions and may reflect processes differ- not in current drinkers. tive of drinking. ent from urges. Urges also differ from intentions to Models of the Role Characteristics of the UrgeTo Drink drink or expectancies about the effects of drinking. Some recent studies have of Urges in Relapse Urges are inherently a self-reported blurred the distinctions between those phenomenon—that is, a person must variables, assessing all three as though Formalized models of the role of urges in some way describe his or her desire they were aspects of a single construct. can serve three major functions: (1) to to drink. Some researchers who work Clinical observation, however, demon- shape current thinking about urges, with animal models assess craving in strates that the three variables are dis- (2) to help investigators develop research terms of alcohol- or drug-seeking behav- tinct from each other, as follows: designs, and (3) to guide clinical

226 Alcohol Research & Health Does Urge To Drink Predict Relapse?

choices. Numerous models have yet users respond to both sets of cues with unable to cope with a high-risk situation addressed urges, including the condi- increased arousal (Rohsenow et al. 1990). (i.e., a situation that is stressful or in tioned withdrawal model, the condi- which alcohol use has occurred previ- tioned appetitive motivational model, The Conditioned Appetitive ously) in any way other than drinking. the social learning model, and the Motivational Model The social learning model also posits information-processing model. The that either conditioned stimuli or stress- following sections briefly review these In contrast to the conditioned with- ful situations can result in increased four models, all of which have been drawal model, which suggests that urges to drink, especially when com- used to guide research. For more exten- urges develop to avoid the unpleasant bined with positive expectancies about sive discussions of the models and the consequences of withdrawal, the condi- alcohol’s effects. These urges, however, evidence for and against them, readers tioned appetitive motivational model do not necessarily result in alcohol use, are referred to other review articles (see, for example, Stewart et al. 1984) because other factors (e.g., expected (Niaura et al. 1988; Rohsenow et al. proposes that urges result from the desire negative consequences of drinking, the 1994). Because many predictions from for alcohol’s pleasant effects. According drinker’s ability to cope with stress in these models are not mutually exclu- to this model, emotional states charac- other ways, and the drinker’s belief in sive, researchers and clinicians cannot terized by the motivation to consume his or her own ability to handle the sit- easily support or refute any model more alcohol (i.e., appetitive motiva- uation) affect a person’s decision or unequivocally. tional states) become conditioned to intention to drink. Thus, the social stimuli (e.g., the sight of liquor bottles) learning model considers urge as only The Conditioned Withdrawal and settings (e.g., bars) associated with one of many factors that can contribute Model the positive effects of drinking. Sub- to relapse and which is necessary but sequent reexposure to the drinking- not sufficient for relapse to occur. This model suggests that when a per- related stimuli is assumed to induce those Many predictions made by the social son repeatedly undergoes withdrawal appetitive motivational states as well as learning model have been supported in episodes, a conditioning process occurs. positive mood and increases in alcohol- experimental and clinical studies Through this process, the unpleasant related thoughts or urges to drink. Thus, (Niaura et al. 1988). In fact, the social symptoms of withdrawal (e.g., tremors, both in this model and in the condi- learning model has provided the most agitation, and anxiety) become associ- tioned withdrawal model, urge accom- guidance to clinicians because of the ated with stimuli present during the panies learned reactions and may lead wealth of variables (e.g., expectancies, withdrawal episodes (e.g., treatment to drinking. conditioned responses, coping skills, settings, such as a hospital or therapist’s More experimental support exists and urges) that it incorporates. For office) (see, for example, Ludwig et al. for the conditioned appetitive motiva- example, coping-skills-training meth- 1974). As a result, exposure to those tional model than for the conditioned ods have been derived directly from stimuli will induce a mild withdrawal withdrawal model (Niaura et al. 1988). this model (e.g., Monti et al. 1989). syndrome, which, in turn, will result in For example, reactions elicited in response urges to drink to relieve withdrawal. to all types of drinking- or drug-related The Information-Processing Model Research findings and clinical obser- stimuli resemble the arousal associated vation, however, do not provide strong with drug-seeking behavior (e.g., Instead of implicating conditioning support for this model. For example, decreased skin resistance), whereas the processes, the information-processing patients usually experience few urges conditioned withdrawal model would model proposes that much alcohol- in treatment settings; in contrast, they predict different reactions across various seeking behavior in alcoholics is con- experience strong urges in settings in types of substances (Rohsenow et al. trolled by automatic overlearned which drinking has taken place previously. 1990). Furthermore the increased sali- processes2 rather than by conscious Furthermore, responses to withdrawal- vation that occurs when drinkers see thoughts (Tiffany 1990; also see the associated stimuli do not consistently alcohol resembles preparation for con- article in this issue by Tiffany, pp. resemble withdrawal symptoms. For sumption (i.e., is an appetitive response). 215–224). These automatic processes example, the model would predict that occur in the presence of stimuli associ- drug users should respond to the sight The Social Learning Model ated with past drinking and may occur of opiates with increased arousal and without awareness or without any urge to the sight of cocaine with decreased According to the social learning model on the part of the drinker. In contrast, arousal (because opiate withdrawal (see, for example, Abrams and Niaura urges result from conscious processes, causes increased arousal and cocaine 1987), relapse risk is associated with such as problem-solving thoughts and withdrawal causes decreased arousal), numerous learned factors. These factors

include conditioned responses, positive 2 expectancies about alcohol’s effects, Automatic overlearned processes are behaviors that 1This lack of correlation can be expected if the are repeated so frequently that they are performed information-processing theory of urges, which is ineffective coping responses, and the with little awareness and effort (e.g., eating, dress- discussed later in this article, is valid. alcoholic’s expectation that he or she is ing, or driving a car).

Vol. 23, No. 3, 1999 227 increased attention or awareness, that Role-play assessment presents alco- consistency and validity and have exhib- are initiated when alcohol use is blocked, holics with 8 to 10 scenarios describing ited no gender differences in three sam- either involuntarily (e.g., when a drinker’s situations that pose a high risk for relapse ples of alcoholics (Monti et al. 1993a). favorite bar is closed) or voluntarily among alcoholics—for example, feeling Furthermore, the urge ratings in cue- (e.g., when an alcoholic decides to quit angry and frustrated after being unem- reactivity assessment are highly reliable drinking). ployed for 1 month and yet being across repeated trials (e.g., Monti et al. According to the information- ineligible for unemployment benefits 1993b). Also, some studies found that processing model, relapse can occur either (Monti et al. 1993a). After each situation alcoholics reacted more strongly to through conscious processes, including is presented, the alcoholics describe alcohol-related cues than did nonalco- urges, or through automatic drug-seeking holics (Monti et al. 1987). Approximately processes and in the absence of urges. 65 percent of the alcoholics experienced Thus, urges are not necessary for relapse an increased urge to drink and approxi- to occur but, instead, may result from mately 70 percent exhibited increased the person’s attempt to abstain from Alcoholics salivation when exposed to alcohol cues drinking in a high-risk situation. In that undergoing (Rohsenow et al. 1992, 1994). In addi- situation, urges reflect the desire to seek tion, other studies have demonstrated, as alcohol despite the conflicting wish to treatment tend expected, that the urge to drink is largely avoid alcohol use. Some experimental independent of salivation and other support exists for this model (Rohsenow to experience only psychophysiological measures (Niaura et al. 1994), which is discussed in the infrequent et al. 1988; Rohsenow et al. 1992). section “Urges as Predictors of Outcome,” Patients’ responses during cue- p. 229. and mild urges. reactivity or role-play assessment have been found to correlate with various concurrent behaviors relevant to relapse Assessing Urges risk, thereby supporting the validity of how they would handle the situation. these assessment approaches. For example, When assessing the role of urges in Trained judges later evaluate those urge in response to cues or to role-play relapse, researchers must consider the responses for the degree of coping skills situations is greater among alcoholics context in which alcoholics are queried exhibited. The alcoholics also rate (on a with greater alcohol dependence than about their urges to drink. For exam- scale from 0 to 10) their urge to drink among alcoholics with lesser alcohol ple, alcoholics undergoing treatment and their confidence that they can han- dependence (Monti et al. 1993b; tend to experience only infrequent and dle such a situation without drinking. Rohsenow et al. 1992). mild urges, possibly because they are One role-play assessment measure used Some of the relationships between not exposed to alcohol-related stimuli in prediction studies is the Alcohol- urges and other measures also suggest that lead to urges. Consequently, urges Specific Role Play Test (ASRPT) (Monti ways through which these reactions reported during treatment may not et al. 1990). may increase relapse risk. For example, predict drinking outcomes (Monti et Cue-reactivity assessment evaluates alcoholics who reacted early in treat- al. in press). Furthermore, a patient’s urges while asking alcoholics to hold ment to alcohol cues with a greater risk of relapse likely is not constant but and smell a glass of the alcoholic beverage urge to drink also performed less skill- varies across several days or even over that they most often drink. This type fully in the ASRPT during treatment, the course of a given day. In addition, of evaluation mimics real-life situations suggesting that reactions to alcohol certain situations or events—such as in which alcoholics are exposed to alco- cues disrupt the ability to use coping an argument with a family member, hol but know that they must not drink skills in high-risk situations (Monti et a celebration, or seeing friends drink it. In addition to rating their urge to al. 1993a). In another study, alcoholics at a party—may be more likely to drink, the patients rate the amount of were asked to push a button in response provoke a relapse and thus represent attention they pay to the sight and to hearing a sound while they were an “Achilles’ heel” for the recovering smell of the drink and to their thoughts simultaneously exposed to either alcohol- alcoholic. Consequently, to obtain about alcohol as well as their sensory related or alcohol-unrelated cues meaningful information about the awareness (e.g., salivation or heart (Sayette et al. 1994). In that study, the relationship between urges and relapse pounding). In addition, the investigator risk, researchers should assess urges collects information on salivation and 3Reliability means that if the same person is tested in contexts designed to provoke other psychophysiological measures several times, the results will be consistent across the responses similar to those that may (e.g., heart rate and skin conductance). tests (test-retest reliability) or if a concept (such as be experienced in real-life, high-risk Both role-play and cue-reactivity urge) is tested using a number of items, the person situations. Two methods frequently assessment have excellent reliability and will respond consistently across the items (internal 3 consistency). Validity means that the test has been used in urge assessment are role play validity. For example, the urge ratings shown to measure effectively the variable it is sup- and cue reactivity. in the role-play test have good internal posed to measure.

228 Alcohol Research & Health Does Urge to Drink Predict Relapse?

alcoholics’ reaction times for pushing frequency of cravings, significantly pre- drinking after treatment, whereas addi- the button were significantly longer when dicted resumption of drinking during tional variance may be associated with they were exposed to alcohol-related the subsequent 2-month period. One other unconscious and conscious processes. cues than when they were exposed to shortcoming of this study, however, was Unconscious automatic processes may non-alcohol-related cues. These findings that not all participants were abstinent be reflected by psychophysiological indicate that the presence of alcohol cues at the time the urges were assessed. reactions, such as increased salivation slows information processing by divert- Consequently, if drinking results in or heart rate. ing the alcoholics’ attention. Both the increased urges, the results may be A conscious process that may influ- study by Monti and colleagues (1993a) somewhat tautological. ence the risk of relapse in addition to and the study by Sayette and colleagues One study used the urge to drink urges is the drinker’s attention in the (1994) suggest mechanisms through and other responses to the ASRPT at presence of alcohol cues. The informa- which stronger reactions to cues may the end of residential treatment to predict tion-processing model suggests that place alcoholics at greater risk for relapse. drinking during a 6-month followup attempts to avoid drinking in the presence period (Monti et al. 1990). Using sta- of alcohol cues can increase a drinker’s tistical methods, the various self-report attention to those cues and to his or Urges as Predictors measures (i.e., urge to drink, anxiety, her reactions to the cues (Tiffany 1990). of Outcome difficulty coping without drinking, and This hypothesis regarding the role of self-rated skill in coping) then were attention is extended by social learning Several studies using a variety of methods averaged across the situations presented models suggesting that awareness of have assessed the role of urges in pre- during the ASRPT. These analyses found danger or of one’s own reactions is central dicting treatment outcome and relapse. that urge to drink alone significantly to self-regulation of behavior (Bandura For example, Marlatt and Gordon predicted drinking quantity over the 1977) and may be crucial in mobilizing (1985) asked alcoholics retrospectively followup period, accounting for 26 coping responses (Rohsenow et al. 1994). about the situations in which they had percent of the variance in the alcohol Four cue-exposure studies have investi- relapsed and reported that urge was amount consumed. The urge to drink gated the ability of automatic processes rarely the primary cause of relapse in induced during a procedure designed (e.g., salivation) and nonautomatic pro- those situations. The ability of such an to produce stress similarly predicted cesses, such as urges and attention, to pre- approach to determine the role of urges drinking quantity during followup. The dict outcome after treatment, as follows: in relapse is limited, however. For strongest predictor of subsequent drink- example, memory becomes less reliable ing, however, was the urge to drink that • Monti and colleagues (1993c) used and more selective over time, making persisted through the 3-minute recovery cue-elicited urges and salivation in retrospective information increasingly period after each role-play test. This the first few days of treatment to pre- inaccurate. Also, Marlatt and Gordon variable accounted for 58 percent of dict outcome in a small sample of (1985) likely underrepresented the impor- the variance in drinking quantity and alcoholics undergoing treatment in tance of urges, because the researchers 41 percent of the variance in drinking a Veterans’ Affairs (VA) facility. In categorized the relapse situations as frequency during the followup period. that study, higher urge ratings predicted resulting from urges or craving only if These findings indicate that the degree less frequent drinking. Furthermore, no additional possible causes (e.g., con- of urge at the end of treatment as salivation did not predict drinking.5 flicts with other people) occurred. assessed in a role-play test using simu- Consequently, both laboratory studies lated high-risk situations can predict • Drummond and Glautier (1994) and ambulatory monitoring techniques subsequent drinking. Furthermore, also found that higher urge ratings (e.g., use of small portable computers lasting urges that persist even after the at the end of treatment predicted a into which participants enter information high-risk situation is over may be par- longer interval before relapse to immediately after a situation occurs) offer ticularly predictive of outcome. heavy drinking. Conversely, psy- advantages over retrospective accounts chophysiological arousal during that for assessing the relevance of urges. Urges During Cue-Reactivity session predicted more rapid relapse. Other researchers have used a prospec- Assessment as Predictors of Outcome tive approach to elucidate the role of • Cooney and colleagues (1997) urges. For example, Miller and colleagues Several studies evaluating the role of noted that the urge to drink after (1996) asked alcoholics to rate their urges in response to alcohol beverage “urges” and “cravings” 4 months after cues as predictors of drinking outcome 4 treatment.4 The investigators then used have generated somewhat mixed results. The terms “urges” and “cravings” were not specifi- cally defined in that study and their meaning there- this information to predict relapse or Those analyses have allowed researchers fore may not correspond to the definitions used in resumption of drinking after at least 4 to test some of the predictions from the the present article. days of abstinence during the next 2 information-processing model of relapse. 5Because the study used only a small sample, its months. The study found that the fre- According to that model, urges to drink statistical power to detect the effects of salivation quency of reported urges, but not the account for only some of the variance in was limited.

Vol. 23, No. 3, 1999 229 exposure only to alcohol cues was alcohol cues or to their own reactions For example, many alcoholics report not predictive of outcome. Instead, may be less aware of the riskiness of little difficulty refusing a drink during increased desire to drink predicted a these situations and therefore may be coping-skills training (e.g., Monti et al. shorter time to the first drink only controlled to a greater extent by their 1989). This finding may explain why among alcoholics who were exposed automatic overlearned behaviors (as the urge to drink during alcohol cue both to alcohol cues and to a proce- indicated by increased salivation). exposure may be unrelated to future dure designed to induce negative Through this mechanism, lack of drinking or may even protect against mood. attention may increase a recovering later drinking. Complex emotional sit- alcoholic’s likelihood of drinking in uations (e.g., problems at work or con- • Rohsenow and colleagues (1994) high-risk situations. flicts with family members), in con- studied a large group of alcoholics In contrast to salivation and atten- trast, may be more difficult to cope being treated at a VA facility to deter- tional variables, the results regarding with effectively. Thus, studies using mine the ability of urge to drink, the role of urge in drinking outcome negative mood induction or the com- salivation, and self-reported atten- plex social situations represented in the tional variables in the first few days ASRPT found that the urge to drink in of treatment to predict drinking those situations predicted later drink- outcomes (e.g., frequency of drink- Urges may not ing. (Cooney et al. 1997; Monti et al. ing during followup). In that study, 1990). According to the information- urge to drink did not predict out- necessarily increase processing and social learning models, come. However, salivation was asso- the risk of drinking those complex situations may over- ciated with treatment outcome, in whelm an alcoholic’s coping resources that participants who salivated more but may actually or the nonautomatic mental efforts drank more often during followup. needed to prevent drug-seeking behav- In addition, the amount of atten- protect against ior. As a result, more intense urges to tion the alcoholics reported paying drinking under drink in these complex situations to either the alcohol cues or to their increase the likelihood of drinking. sensory reactions to the cues signifi- certain circumstances cantly predicted outcome. Thus, alcoholics who reported paying more Clinical Implications attention to either the cues or to their were inconsistent, with urge predicting reactions to the cues drank signifi- less drinking in some studies and more The research findings regarding the cantly less during the followup period drinking in others. Thus, although a role of urges described in the previous than did alcoholics who reported positive correlation exists between urge section have several clinical implica- paying less attention. to drink and attention to alcohol cues tions. Most important, the results indi- (i.e., the greater a drinker’s urge, the more cate that urges may not necessarily The ability of both salivary reactions attention he or she will pay to alcohol increase the risk of drinking but may to alcohol cues and attentional variables cues) (Monti et al. in press; Rohsenow actually protect against drinking under to predict outcome is consistent with et al. 1994), the two variables differ in certain circumstances (e.g., in drinkers the information-processing model. Thus, their ability to predict outcome. who are aware of alcohol cues and of salivation may reflect the automatic One explanation for this inconsis- their reactions to those cues). Many learning processes related to drinking tency may be that newly abstinent alco- alcoholics try to prevent themselves (Rohsenow et al. 1994), whereas atten- holics who are in treatment generally from experiencing urges, possibly tion is considered an aspect of nonau- experience conflict: On the one hand, because they consider them to be a risk tomatic thought processes and is believed they are motivated to drink to experi- for relapse. As a result, however, some to correlate with attempts to inhibit ence alcohol’s effects; on the other of those alcoholics may develop an desires to seek drugs. Accordingly, peo- hand, they are motivated not to drink overconfidence in their ability to cope ple who pay greater attention to alco- for a variety of reasons. As a result of with high-risk situations or may under- hol cues probably inhibit their desire this conflict, the urge to drink among estimate the riskiness of some situations. to seek alcohol, possibly because greater newly abstinent alcoholics may consist Monti and colleagues (1993b) attention to high-risk situations or to of two components: (1) motivation to found that alcoholics who salivated rel- one’s reactions to those situations facili- drink and (2) awareness of the danger atively more in response to alcohol cues tates mobilization of coping skills, there- of relapse. The variance in urge to (i.e., who were at greater risk for drink- by helping to handle the situation more drink that results from awareness of ing) only reported an increased urge to effectively (Rohsenow et al. 1994). danger should induce the alcoholic to drink if they also were aware of salivat- This greater use of coping skills, in initiate coping skills that reduce the ing more. These findings suggest that turn, predicts less drinking. Conversely, risk of drinking. However, some situa- the experience of urges may require alcoholics who pay less attention to tions are easier to cope with than others. some self-awareness. Consequently, for

230 Alcohol Research & Health Does Urge To Drink Predict Relapse?

alcoholics in high-risk situations who Future Directions In conclusion, the urge to drink may experience psychophysiological among recovering alcoholics is a construct reactions, an increased awareness of their Investigators are currently pursuing of considerable interest to researchers as reactions in such situations (including several lines of research to further clar- well as clinicians that is finally being the urge to drink) and of the potential ify the role of urges in recovery after rigorously studied. The studies reviewed danger of relapse may protect them alcoholism treatment. For example, in this article demonstrate that even against drinking by prompting them to some researchers are investigating the inherently subjective variables, such as mobilize their coping resources. In highly effects of naltrexone—a medication the urge to drink, can contribute con- complex situations that pose too great a that was approved for alcoholism treat- siderably to the understanding of the challenge for coping, however, urges ment in 1994—on the urge to drink as factors that determine relapse and may signal an increased risk of relapse. well as the relationship between these recovery when measured with method- Within this conceptualization of the effects and outcome (i.e., amount and ological rigor and in meaningful con- role of urges in the risk of relapse, several frequency of drinking). In one con- texts. Such analyses can be heuristic in factors appear to increase an alcoholic’s trolled laboratory study, naltrexone guiding clinical practice. probability of drinking when experiencing reduced the number of alcoholics who the urge to drink (Monti et al. in press): experienced any urge to drink; how- ever, the medication did not affect the References • Factors that increase the motivation magnitude of the urges to drink among to drink, such as positive expectancies ABRAMS, D.B., AND NIAURA, R.S. Social learning those alcoholics who reported experi- theory. In: Blane, H.T., and Leonard, K.E., eds. about alcohol, negative emotions, and encing such urges (Monti et al. in press). Psychological Theories of Drinking and Alcoholism. certain physiological states (e.g., low Researchers currently do not know New York: Guilford Press, 1987. pp. 131–178. levels of certain chemicals in the brain) whether urge reduction is the primary BAKER, T.B.; MORSE, E.; AND SHERMAN, J.E. The mechanism through which naltrexone motivation to use drugs: A psychobiological analy- • Factors that decrease the awareness improves drinking outcome, but that sis of urges. In: Rivers, C., ed. The Nebraska of danger, such as overconfidence or issue is under investigation. Symposium on Motivation: Alcohol Use and Abuse. maladaptive beliefs about the riski- Lincoln, NE: University of Nebraska Press, 1987. A second interesting line of research pp. 257–323. ness of a situation as well as physio- involves the extent to which changes in logical states that decrease general BANDURA, A. Social Learning Theory. Englewood cue-elicited urges are associated with awareness (e.g., overtiredness) Cliffs, NJ: Prentice Hall, 1977. changes in the levels of a certain brain chemical (i.e., the neurotransmitter COONEY, N.L.; LITT, M.D.; MORSE, P.A.; BAUER, • Factors that decrease the effective- L.O.; AND GAUPP, L. Alcohol cue reactivity, nega- ) in the brain regions that ness of coping, such as inadequate tive-mood reactivity, and relapse in treated alco- coping skills, highly complex situa- mediate the rewarding effects of alco- holic men. Journal of Abnormal Psychology 106: tions, failure to believe in one’s abil- hol (Monti et al. in press). Many 243–250, 1997. researchers think that higher dopamine ity to cope effectively, and physio- DRUMMOND, D.C., AND GLAUTIER, S. A con- logical states that impair cognitive levels in those brain regions are related trolled trial of cue exposure treatment in alcohol abilities (e.g., exhaustion or illness). to greater urges, and that medications dependence. Journal of Consulting and Clinical which reduce dopamine levels in those Psychology 62:809–817, 1994. Clinicians can help alcoholics recog- areas could prevent urges and thus be KOZLOWSKI, L.T.; MANN, R.E.; WILKINSON, D.A.; nize that urges are a danger sign that useful in alcoholism treatment. AND POULOS, C.X. “Cravings” are ambiguous: Ask In a third line of research, investiga- about urges or desires. Addictive Behaviors 14: should not be ignored but should be 443–445, 1989. considered an indication that coping tors are assessing the effects of extended resources must be mobilized. Further- cue-exposure treatment (both by itself LUDWIG, A.M.; WIKLER, A.; AND STARK, L.H. The and in combination with either training first drink: Psychobiological aspects of craving. more, certain alcoholism treatment Archives of General Psychiatry 30:539–547, 1974. approaches, such as coping-skills train- for coping with urges or with commu- ing, can strengthen an alcoholic’s abil- nication skills training) on drinking MARLATT, G.A., AND GORDON, J.R. Relapse Prevention. New York: Guilford Press, 1985. ity to cope with urges and risky situa- outcome. This treatment approach was tions as well as increase confidence in piloted in a small study that suggested MILLER, W.R.; WESTERBERG, V.S.; HARRIS, R.J.; his or her own ability to apply coping such combinations of cue-exposure AND TONIGAN, J.S. What predicts relapse? Prospective testing of antecedent models. Addiction skills (Monti et al. 1989). Certain urge- treatment with coping-skills training 91(Suppl.):S155–S171, 1996. specific coping skills in particular— could improve treatment outcome such as coping with urges by thinking (Monti et al. 1993c). Another study MONTI, P.M.; BINKOFF, J.A.; ABRAMS, D.B.; ZWICK, W.R.; NIRENBERG, T.D.; AND LIEPMAN, about the positive consequences of suggested that cue-exposure treatment M.R. Reactivity of alcoholics and nonalcoholics to staying sober and the negative conse- even without any skills training could drinking cues. Journal of Abnormal Psychology 96: quences of drinking—are associated reduce future drinking (Drummond et 122–126, 1987. with less frequent drinking after treat- al. 1994). The results of a subsequent MONTI, P.M.; ABRAMS, D.B.; KADDEN, R.; AND ment (Monti et al. 1993c). larger trial will be released soon. COONEY, N. Treating Alcohol Dependence: A Coping

Vol. 23, No. 3, 1999 231 Skills Training Guide. New York: Guilford Press, treatment for male alcoholics: A preliminary inves- COLBY, S.M. Cue-elicited urge to drink and saliva- 1989. tigation. Journal of Consulting and Clinical Psychology tion in alcoholics: Relationship to individual differ- 61:1011–1019, 1993c. ences. Advances in Behavior Research and Therapy MONTI, P.M.; ABRAMS, D.B.; BINKOFF, J.A.; 14:195–210, 1992. ZWICK, W.R.; LIEPMAN, M.R.; NIRENBERG, T.D.; MONTI, P.M.; ROHSENOW, D.J.; AND HUTCHISON, AND ROHSENOW, D.J. Communication skills train- K.E. Toward bridging the gap between biological, ROHSENOW, D.J.; MONTI, P.M.; RUBONIS, A.V.; psychobiological, and psychosocial models of alco- ing, communication skills training with family, and SIROTA, A.D.; NIAURA, R.S.; COLBY, S.M.; hol craving. Addiction, in press. cognitive behavioral mood management training WUNSCHEL, S.M.; AND ABRAMS, D.B. Cue reactiv- for alcoholics. Journal of Studies on Alcohol 51: NATHAN, P.E., AND LISMAN, S.A. Behavioral and ity as a predictor of drinking among male alcoholics. 263–270, 1990. motivational patterns of chronic alcoholics. In: Journal of Consulting and Clinical Psychology 62: MONTI, P.M.; ROHSENOW, D.J.; ABRAMS, D.B.; Tarter, R.E., and Sugerman, A.A., eds. Alcoholism: 620–626, 1994. ZWICK, W.R.; BINKOFF, J.A.; MUNROE, S.M.; Interdisciplinary Approaches to an Enduring Problem. SAYETTE, M.A.; MONTI, P.M.; ROHSENOW, D.J.; FINGERET, A.L.; NIRENBERG, T.D.; LIEPMAN, Reading, MA: Addison-Wesley, 1976. pp. 479–522. GULLIVER, S.B.; COLBY, S.M.; SIROTA, A.D.; M.R.; PEDRAZA, M.; KADDEN, R.M.; AND COONEY, NIAURA, R.S.; ROHSENOW, D.J.; BINKOFF, J.A.; NIAURA, R.; AND ABRAMS, D.B. The effects of cue N.L. Development of a behavior analytically derived MONTI, P.M.; ABRAMS, D.A.; AND PEDRAZA, M. alcohol-specific role-play assessment instrument. exposure on reaction time in male alcoholics. The relevance of cue reactivity to understanding Journal of Studies on Alcohol 55:629–633, 1994. Journal of Studies on Alcohol 54:710–721, 1993a. alcohol and relapse. Journal of Abnormal Psychology 97:133–152, 1988. MONTI, P.M.; ROHSENOW, D.J.; RUBONIS, A.V.; STEWART, J.; DEWIT, H.; AND EIKELBOOM, R. The role of unconditioned and conditioned drug effects NIAURA, R.S.; SIROTA, A.D.; COLBY, S.M.; AND ROHSENOW, D.J.; NIAURA, R.S.; CHILDRESS, A.R.; in the self-administration of opiates and stimulants. ABRAMS, D.B. Alcohol cue reactivity: Effects of ABRAMS, D.B.; AND MONTI, P.M. Cue reactivity in detoxification and extended exposure. Journal of addictive behaviors: Theoretical and treatment Psychological Review 91:251–268, 1984. Studies on Alcohol 54:235–249, 1993b. implications. International Journal of the TIFFANY, S.T. A cognitive model of drug urges and 25(7A & 8A):957–993, 1990. MONTI, P.M.; ROHSENOW, D.J.; RUBONIS, A.; drug use behavior: Role of automatic and non- NIAURA, R.; SIROTA, A.; COLBY, S.; GODDARD, P.; ROHSENOW, D.J.; MONTI, P.M.; ABRAMS, D.B.; automatic processes. Psychological Review 97: AND ABRAMS, D.B. Cue exposure with coping skills RUBONIS, A.V.; NIAURA, R.S.; SIROTA, A.D.; AND 147–168, 1990.

Are women more vulnerable to alcohol’s effects?

he answer to this and other questions can be found in Alcohol Alert, the quarterly bulletin pub- lished by the National Institute on Alcohol Abuse and Alcoholism. Alcohol Alert provides timely Tinformation on alcohol research and treatment. Each issue addresses a specific topic in alcohol research and summarizes critical findings in a brief, four-page, easy-to-read format. Our most recent issue—Are Women More Vulnerable To Alcohol’s Effects (No. 46)—summarizes recent research on the consequences of drinking in women.

Forthcoming issues include the following: • Imaging and Alcoholism (No. 47), which examines the use of imaging in alcohol research and the potential role this technology may play in monitoring alcohol’s effects.

For a free subscription to Alcohol Alert, write to: National Institute on Alcohol Abuse and Alcoholism, Attn.: Alcohol Alert, Publications Distribution Center, P.O. Box 10686, Rockville, MD 20849Ð0686. Fax: (202) 842Ð0418. Alcohol Alert is available on the World Wide Web at http://www.niaaa.nih.gov

232 Alcohol Research & Health