Negative affect as a mediator of the relationship between vigorous-intensity exercise and smoking

Abstract

The present cross-sectional study evaluated whether people who engage in vigorous-intensity exercise are better able to regulate negative affective states, thereby changing core maintenance factors of smoking. Participants were a community sample of adults (n = 270) who completed self-report measures of physical activity, cigarette smoking, anxiety sensitivity, and negative affect. Consistent with hypothesis, vigorous-intensity exercise was related to lower levels of cigarette smoking, accounting for 10% of the variance in smoking. Additionally, negative affect mediated the relationship between vigorous-intensity physical activity and cigarette smoking, accounting for about 12% of this relation. Furthermore, these relationships were stronger for individuals with high anxiety sensitivity than for those with low anxiety sensitivity; including anxiety sensitivity as a moderator of the mediated relationship increased the amount of variance accounted for by negative affect to 17%. The findings are discussed in relation to developing further scientific insight into the mechanisms and pathways relevant to understanding the association among vigorous-intensity exercise, smoking, and emotional vulnerability.

Introduction

Growing evidence points to the role of negative affect in the maintenance of smoking and smoking cessation relapse. For example, when asked about triggers of smoking cessation relapse smokers consistently point to the experience of stress and negative affect (Brandon & Baker, 1991, Piper et al., 2004). These retrospective reports are complemented by prospective studies that indicate that negative affect is an important precipitating factor in smoking lapses and relapses. Specifically, ecological momentary assessments from 215 smokers collected during the two weeks before and four weeks after initiation of smoking cessation treatment indicate that abrupt increases in negative affect are associated with smoking lapses (Shiffman & Waters, 2004). Similarly, baseline negative affect and increases in negative affect during treatment have been shown to be the most reliable predictors of relapse in clinical trials of smoking cessation treatments (Covey et al., 1990, Hitsman et al., 1999, Kahler et al., 2002, Lerman et al., 2002, Niaura et al., 2001, Zelman et al., 1992, Zvolensky et al., 2008). Lastly, reducing negative affect during smoking cessation treatment has been shown to improve abstinence outcomes both with psychological interventions (e.g., Fergusson et al., 2003, Hall et al., 1994, Haas et al., 2004) and pharmacological interventions (e.g., Hughes et al., 2007, Prochazka et al., 2004, Richmond & Zwar, 2003). Collectively, these findings suggests that addressing negative affect in smokers may be important especially for those who are more prone to experience negative affect (Brown et al., 2001, Haas et al., 2004).

The relatively poor outcomes of standard smoking cessation treatments (Fiore, 2000, Hughes et al., 2004, Piasecki, 2006) combined with the observation that targeting negative affect during treatment may be critical to cessation success for many smokers (i.e., those who are prone to experience negative affect) provide justification for investigating the utility of exercise as an intervention for smoking cessation. Indeed, exercise is associated with reduced negative affect (Focht et al., 2007, Hassmen et al., 2000, Penedo & Dahn, 2005, Reed & Ones, 2006, Schlicht, 1994) and more importantly, exercise interventions have shown efficacy for the treatment of mood and anxiety problems (Smits et al., 2008, Stathopoulou et al., 2006, Broocks et al., 1998, Martinsen et al., 1989a, Martinsen et al., 1989b). Furthermore, cross-sectional surveys have consistently shown a negative relationship between physical activity levels and smoking (e.g. Boutelle et al., 2000, Boyle et al., 2000, Hu et al., 2002). Likewise, there is initial evidence from randomized controlled trials indicating that exercise interventions can decrease withdrawal symptoms and negative affect in smokers (Bock et al., 1999, Schneider et al., 2007, Taylor et al., 2007) as well as improve smoking cessation outcomes among adults receiving standard cessation treatments (cf. Ussher et al., 2008, Marcus et al., 1991, Marcus et al., 1995, Marcus et al., 1999, Martin et al., 1997). For example, Marcus and colleagues (1999) randomized 281 sedentary female smokers to either a 12-week cognitive-behavioral smoking cessation program with vigorous-intensity exercise (three sessions a week of 30 to 40 min at 60–85% of heart rate reserve), or a 12-week cognitive-behavioral smoking cessation program with contact control (three 45–60 minute health education sessions a week). All participants initiated the intervention three weeks prior to the quit date of the smoking cessation program. Results revealed that participants receiving the exercise intervention were more likely than participants in the control intervention to be continuously abstinent during the 8, 20, and 60 weeks following the quit date. Unfortunately, neither this study nor other studies in this area have investigated whether the association between exercise and reduced smoking is accounted for by reductions in negative affect. Evidence for this mediational hypothesis would help determine whether exercise is a viable option for smokers for whom negative affect operates prominently in the maintenance of smoking and smoking cessation relapse.

This study aimed to provide a preliminary test of the hypothesis that the association between exercise and smoking is, at least in part, accounted for by reduced negative affect. Using cross-sectional data, we examined self-reported negative affect as a mediator of the relationship between self-reported vigorous-intensity exercise levels and smoking. We chose to evaluate the relationship between vigorous-intensity exercise and smoking because there is evidence to suggest that the association between exercise and cigarette smoking is stronger for vigorous-intensity exercise relative to moderate- or low-intensity exercise (cf. Kaczynski, Manske, Mannell & Grewal, 2008). We also investigated the possibility that the strength of these meditational effects would vary as a function of anxiety sensitivity. Anxiety sensitivity, conceptualized as an emotional vulnerability variable, is a relatively stable trait (Peterson & Plehn, 1999, Weems et al., 2002) characterized by the fear of both anxiety and related autonomic arousal sensations (e.g., racing heart, sweating, nausea; Reiss, Peterson, Gursky & McNally, 1986). We selected anxiety sensitivity as a possible moderator of the hypothesized mechanism because of the increasing evidence that individuals with elevated levels of anxiety sensitivity, relative to persons with low levels of anxiety sensitivity, are more likely to smoke in response to negative affect (Brown et al., 2001, Brown et al., 2002, Novak et al., 2003, Zvolensky et al., 2006). Furthermore, smokers with higher levels of anxiety sensitivity are more likely to report negative affect reduction as a smoking outcome expectancy than smokers with lower levels of anxiety sensitivity (Brown et al., 2001, Zvolensky et al., 2007). Accordingly, negative affect reduction as a mechanism underlying the relationship between exercise and smoking may be more salient for individuals with high versus low anxiety sensitivity. We tested the following specific hypotheses: (1) vigorous-intensity exercise engagement would be associated with decreased smoking; (2) the relationship between vigorous-intensity exercise engagement and smoking would be partially mediated by negative affect; and (3) anxiety sensitivity would moderate these mediated relationships such that the mediational role of negative affect would be stronger for individuals with high levels of anxiety sensitivity relative to those with low levels of anxiety sensitivity. Based on the available evidence, we predicted that anxiety sensitivity would moderate the relationship between negative affect and smoking (i.e. the “b” path, see Fig. 3) as opposed to the relationship between exercise and negative affect (i.e. the “a” path, see Fig. 3).