Page 1 of 9http://spider.apa.org/ftdocs/ccp/1997/october/ccp655878.html 10/12/2000Cue Exposure in Moderation Drinking A Comparison With Cognitive—Behavior Therapy Thiagarajan SitharthanWestern Sydney Area Drug and Alcohol Services and Westmead Hospital Gomathi SitharthanDepartment of Behavioural Sciences , Faculty of Health Sciences University of Sydney Michael J. HoughDepartment of Behavioural Sciences , Faculty of Health Sciences University of Sydney David J. KavanaghDepartment of Psychiatry University of Queensland ABSTRACTTo date, the published controlled trials on exposure to alcohol cues have had an abstinence treatment goal. A modification of cue exposure (CE) for moderation drinking, which incorporated priming doses of alcohol, could train participants to stop drinking after 2 to 3 drinks. This study examined the effects of modified CE within sessions, combined with directed homework practice. Nondependent problem drinkers who requested a moderation drinking goal were randomly allocated to modified CE or standard cognitive—behavior therapy (CBT) for alcohol abuse. Both interventions were delivered in 6 90-min group sessions. Eighty-one percent of eligible participants completed treatment and follow-up assessment. Over 6 months, CE produced significantly greater reductions than CBT in participants' reports of drinking frequency and consumption on each occasion. No pretreatment variables significantly predicted outcome. The modified CE procedure appears viable for nondependent drinkers who want to adopt a moderate drinking goal. This research was supported by a Cumberland Research Grant. We thank Keith Carter for his assistance in advertising the details of this trial to the media and for recruiting participants. We also thank the editor and two anonymous referees for their helpful comments on earlier versions of the article. Correspondence may be addressed to Thiagarajan Sitharthan, Western Sydney Area Drug and Alcohol Services and Westmead Hospital, 4A Fleet Street, North Parramatta, New South Wales, Australia, 2151. Received: July 11, 1996 Revised: December 11, 1996 Accepted: April 9, 1997 Cognitive—behavior therapy (CBT) for problem drinking involves teaching clients to identify high-risk situations and develop appropriate behavioral and cognitive coping strategies to deal with them ( Marlatt & Gordon, 1985 ). Although avoidance strategies are part of the CBT package, proponents of the exposure model ( Rankin, Hodgson, & Stockwell, 1983 ) postulate that avoidance does not Journal of Consulting and Clinical Psychology © 1997 by the American Psychological Association October 1997 Vol. 65, No. 5, 878 -882 For personal use only--not for distribution.Page 2 of 9http://spider.apa.org/ftdocs/ccp/1997/october/ccp655878.html 10/12/2000prepare clients to cope effectively with cravings or excessive desires to drink. Because drinking cues are thought to acquire potency through conditioning, extinction would require repeated exposure to the alcohol cues with response prevention ( Rankin et al., 1983 ). Several case studies applying cue exposure (CE) treatment appeared promising ( Blakey & Baker, 1980 ; Hodgson & Rankin, 1976 ), but only recently have randomized controlled trials evaluating the effectiveness of CE training become available ( Drummond & Glautier, 1994 ; Monti et al., 1993 ). In the Monti et al. (1993) trial, no significant difference was observed between CE and standard hospital treatment in the first 3 months after the intervention. However, the CE group reported drinking less than participants in standard treatment during the second 3 months, and they also reported a higher percentage of abstinence days. Monti et al. ascribed this improvement to coping skills training that was offered as part of the CE intervention. Drummond and Glautier (1994) found that CE did not result in more participants being abstinent over the subsequent 6 months than did relaxation training, nor did the CE participants take longer to have their first drink after treatment. However, the CE group consumed less alcohol during follow-up and took longer to relapse. These two trials had focused on inpatient male populations and an abstinence goal. As intensive inpatient treatments are no more effective than less intensive treatment in outpatient settings ( Institute of Medicine, 1990 ), we decided to evaluate the effectiveness of CE in an outpatient setting. We offered a moderation drinking goal, which keys into the demands of nondependent problem drinkers ( Sobell & Sobell, 1987 ). Because the critical task for moderate drinkers is to control intake during a drinking session, in contrast to previously published trials, we included priming doses of alcohol (cf. Hodgson & Rankin, 1976 ). We administered CE in groups rather than in the more expensive individual format used in the previous trials and included both men and women in the study. Standard CBT was used as the control condition. This set a very high standard for judging the effectiveness of CE, because CBT has established efficacy as a treatment for alcohol abuse ( Miller, Leckman, Delaney, & Tinkcom, 1992 ). This study, therefore, represented the first randomized controlled trial of CE with a moderate drinking goal and the first to incorporate priming doses of alcohol in the CE procedure. CE was compared with standard CBT over a 6-month period. Method Participants Potential participants were recruited through media announcements similar to those in a previous investigation by Sitharthan and Kavanagh (1990) . Inclusion criteria for the study were as follows: (a) age of at least 18 years, (b) acknowledged problems with control of alcohol use, (c) a score of 10 or more on the Impaired Control Questionnaire (ICQ; Stockwell, Sitharthan, McGrath, & Lang, 1994 ), (d) a stated wish to participate in a moderation drinking program, (e) no current medical or major psychiatric condition at examination (including pregnancy and physical dependence on alcohol), and (f) a score of less than 30 (i.e., excluding high alcohol dependence) on the Severity of Alcohol Dependence Questionnaire–Form C (SADQ-C; Stockwell et al., 1994 ). The latter criterion was included because participants with high dependence scores are more likely to experience withdrawal and may have coexisting medical conditions such as liver disorders. People with significant alcohol dependence have poorer outcomes from moderation drinking than do those with lower dependencePage 3