Page 1 of 12http://spider.apa.org/ftdocs/ccp/1999/august/ccp674555.html 8/30/2000Gender Differences in Smoking Cessation David W. WetterDepartment of Behavioral Science University of Texas M. D. Anderson Cancer Center Susan L. KenfordCenter for Tobacco Research and Intervention University of Wisconsin–Madison Stevens S. SmithCenter for Tobacco Research and Intervention University of Wisconsin–Madison Michael C. FioreCenter for Tobacco Research and Intervention University of Wisconsin–Madison Douglas E. JorenbyCenter for Tobacco Research and Intervention University of Wisconsin–Madison Timothy B. BakerCenter for Tobacco Research and Intervention University of Wisconsin–Madison ABSTRACTGender differences in smoking quit rates are frequently reported and are the subject of much speculation. This study examined the generalizability of gender differences in abstinence across study sites, treatments, and time of relapse, as well as potential mediators and moderators of gender effects. Participants were smokers who participated in 3 randomized clinical trials of the nicotine patch ( N = 632). Men had higher cessation rates than women at all follow-ups. The impact of gender on abstinence was unaffected by controlling for study site, treatment, or time of relapse. There was little evidence for mediation or moderation of this relation by any of a host of predictor variables. The magnitude and consistency of the gender differential, coupled with an inability to account for it, highlights a compelling need for additional research specifically aimed at elucidating the relation between gender and abstinence. This study was supported by a research grant from Elan Pharmaceutical Research Corporation, Gainesville, Georgia, and Athlone, Ireland. Correspondence may be addressed to David W. Wetter, Center for Health Studies, Group Health Cooperative, 1730 Minor Avenue, Suite 1600, Seattle, Washington, 98101-1448. Electronic mail may be sent to [email protected] Received: May 11, 1998 Revised: September 21, 1998 Accepted: October 27, 1998 Research on gender differences in smoking quit rates has suggested that men are more likely than women to successfully quit smoking. Population-based data indicate that the quit ratio (former smokers to ever smokers) is consistently higher for men (52%) than for women (47%; Centers for Disease Control andPrevention, 1994 ), and smoking prevalence curves within comparable birth cohorts show less quitting for Journal of Consulting and Clinical Psychology © 1999 by the American Psychological Association August 1999 Vol. 67, No. 4, 555-562 For personal use only--not for distribution.Page 2 of 12http://spider.apa.org/ftdocs/ccp/1999/august/ccp674555.html 8/30/2000women than for men ( Escobedo & Peddicord, 1996 ). Recent data from two large community intervention trials also suggest that women have lower rates of quitting than men ( Bjornson et al., 1995 ; CommunityIntervention Trial for Smoking Cessation Research Group, 1995 ). Finally, our own research on smokers seeking treatment reveals that men have higher cessation rates than do women ( Fiore et al., 1994 ; Wetteret al., 1994 ). These results are consistent with conclusions made almost 20 years ago in the Surgeon General's report that "women have more difficulty giving up smoking than men" ( U.S. Department of Healthand Human Services, 1980, p. 307 ). However, gender by itself is not a theoretically or clinically meaningful explanatory factor. Identifying the mechanisms through which gender influences smoking relapse, or identifying subgroups of men or women at particularly high risk for relapse, requires examining the association of gender and abstinence in conjunction with other variables. Mediator variables typically explain how or why a predictor variable influences an outcome variable ( Baron& Kenny, 1986 ). That is, a mediator represents a mechanism through which gender influences abstinence and provides a potential target of intervention. Moderator variables influence the strength or direction of the association between the predictor and dependent variable (i.e., there is an interaction between gender and the moderator variable when predicting abstinence; Baron & Kenny, 1986 ). Moderator variables are useful in identifying subgroups that might be particularly vulnerable to relapse. Gender differences in abstinence have been attributed to numerous factors (e.g., depression, nicotine dependence, withdrawal, and demographics). Perhaps the most common hypotheses concern the notion that negative affect, stress, and depression are especially determinant of outcomes among women ( Borrelli,Bock, King, Pinto, & Marcus, 1996 ). Depression is associated with a decreased likelihood of quitting smoking ( Glassman et al., 1990 ) and is more prevalent among women than among men ( Angst, 1992 ). Therefore, women may be less successful than men at quitting smoking because of the mediating effect of depression on cessation. Similar arguments have been advanced for negative affect and stress. Alternatively, depression, negative affect, and stress may moderate the relation of gender with abstinence, presumably by exerting a more pronounced effect among women than among men ( Borrelli et al., 1996 ). However, there is little research that explores how affective variables influence the relation of gender and outcome; and to the extent that such data exist, they are inconsistent, sometimes appearing to be a more important risk factor for men ( Glassman et al., 1990 ) and other times for women ( Covey, Hughes, Glassman, Blazer, & George,1994 ). Therefore, at present it is unclear whether the investigation of affective and stress variables can elucidate gender differences in cessation outcomes. Gender differences in abstinence have also been ascribed to nicotine-related variables ( Grunberg, Winders,& Wewers, 1991 ). For example, nicotine withdrawal and dependence measures predict abstinence, and there is some evidence that withdrawal severity may be greater among women ( Hatsukami, Skoog, Allen, &Bliss, 1995 ) or may interact with gender such that it is more strongly related to relapse among women than among men ( Gunn, 1986 ). With respect to nicotine dependence, women often display lower levels of dependence on both self-report and biochemical measures than do men (see Perkins, 1996 ). Thus, it is unlikely that nicotine dependence mediates the effects of gender on abstinence. However, in at least one study, dependence