Temptations to Smoke After Quitting A Comparison of Lapsers and Maintainers Saul Shiffman Department of Psychology University of Pittsburgh Maryann Gnys Department of Psychology University of Pittsburgh Thomas J. Richards Department of Psychology University of Pittsburgh Jean A. Paty Department of Psychology University of Pittsburgh Mary Hickcox Department of Psychology University of Pittsburgh Jon D. Kassel Department of Clinical and Health Psychology University of Florida ABSTRACT This study addresses whether characteristics of temptations to smoke differ for participants who quit smoking and maintain abstinence compared to those who quit and then lapse. Participants used hand-held computers to record temptations and were beeped at random for base-rate assessments. We used generalized estimating equations to compare 1,851 temptation episodes and 5,192 random assessments recorded by 151 participants (116 lapsers, 35 maintainers) over 1—23 days of abstinence. Compared to randomly sampled occasions, temptations were marked by greater negative affect, restlessness, attention disturbance, and exposure to smoking cues; participants were also more likely to be eating or drinking during temptations. Temptations reported by lapsers and maintainers did not differ in any respect, including their reported coping. The results highlight situational variance over individual differences. This research was supported by Grant DA06084 from the National Institute on Drug Abuse. We are grateful to Kenneth Perkins and Kathleen O'Connell for their helpful comments on drafts of this article. We are also grateful to Celeste Elash, Walter Perz, Yolanda Dibucci, and Stephanie Paton for their assistance. Correspondence may be addressed to Saul Shiffman, Department of Psychology, University of Pittsburgh, Bellefield Professional Building, 130 North Bellefield Avenue, Suite 510, Pittsburgh, Pennsylvania, 15260. Electronic mail may be sent to [email protected] Relapse continues to be the major problem in the treatment of addictive behaviors, including cigarette smoking ( Brownell, Marlatt, Lichtenstein, & Wilson, 1986 ). One approach to understanding relapse has been to focus on the situations in which lapses occur ( Baer & Lichtenstein, 1988 ; Brownell et al., 1986 ; Shiffman, 1989b ; Shiffman, Paty, Gnys, Kassel, & Hickcox, 1996 ; Sutton, 1992 ). This literature has generally demonstrated that lapse situations are characterized by negative affect, smoking cues, and alcohol consumption ( Shiffman, 1982 ; Shiffman et al., 1996 ; Sutton, 1992 ). To isolate the Health Psychology © 1996 by the American Psychological Association November 1996 Vol. 15, No. 6, 455-461 For personal use only--not for distribution. Page 1 of 1111/20/2000http://spider.apa.org/ftdocs/hea/1996/november/hea156455.htmlfactors that specifically precipitate resumption of smoking, investigators have often compared lapse episodes to temptation episodes–"close-call" situations in which the person was tempted but did not actually lapse ( O'Connell & Martin, 1987 ; Shiffman, 1982 ). These comparisons suggest that temptations are distinguished from lapses primarily by the performance of coping, which seems to prevent relapse ( Curry & Marlatt, 1985 ; Shiffman, 1984 ). However, comparisons between temptations and lapses have been marred by significant methodological flaws. First, data are typically retrospective. Second, accounts of lapses solicited from people who have relapsed are typically compared to reports of temptations solicited from those who have maintained abstinence. This confounds comparison of situations (temptations vs. lapses) with comparison of groups of people who differ in outcome (relapsers vs. successful maintainers). Moreover, these studies usually compare lapses that occur early in abstinence (because relapse is so rapid) with temptations that occur late in abstinence (because participants usually recall temptation episodes proximal to the follow-up), thus also confounding episode outcome with time (e.g., Bliss, Garvey, Heinold, & Hitchcock, 1989 ). We recently addressed this methodological conundrum by comparing lapses and temptations from the same lapsed smokers ( Shiffman et al., 1996 ). We found that lapses were indeed distinguished from temptations by a relative absence of coping, as well as by more severe negative affect, greater alcohol consumption, and a more conducive smoking environment (i.e., exposure to others smoking, cigarette availability, and lack of smoking restrictions). That analysis also compared temptations to randomly sampled moments–collected by having a hand-held computer "beep" participants at random–so as to control for base rates of various moods, activities, and so on ( Stone & Shiffman, 1994 ). Although these within-subjects comparisons circumvent the confusion between situational and population comparisons that plagued prior studies, they leave open a significant question: Are there differences in temptations between participants who lapse early in abstinence (lapsers) and those who do not (maintainers)? Participants who are able to maintain abstinence experience temptations to smoke, and it is plausible that these experiences may differ qualitatively and quantitatively from those experienced by participants who subsequently relapse. In a sense, our prior comparisons embodied what Shiffman (1989a) has called the episodic theory of relapse, that is, attributing relapse episodes entirely to situational factors and ignoring individual differences. In the present analyses from the same study ( Shiffman et al., 1996 ), we aim to evaluate an interactive model that considers interactions between situational factors and one specific individual difference between lapsers and maintainers. Thus, we compare the situations associated with temptation episodes in these two groups, while using as controls assessments administered at random intervals. The temptation experience of maintainers could differ from relapsers' experience in several ways. Abstainers may have milder temptations (e.g., less craving, less negative affect), reflecting an easier time quitting smoking and foreshadowing their ultimately positive outcome. It is also plausible that maintainers possess characteristics that enable them to forestall relapse in situations that would cause a lapser to smoke. In this case, the pool of sampled temptations for maintainers would contain temptations more severe than any reported
View Full Document
Unlocking...