Exercise-based smoking cessation interventions among women

Abstract

Although smoking rates are lower among women than men, women are less likely to quit smoking in cessation trials. This is in part due to their tendency to smoke to help prevent or mitigate negative mood/affect, depression and/or postcessation weight gain. Exercise helps to alleviate women’s fear of postcessation weight gain and reduces their cessation-related mood symptoms, making it a theoretically ideal smoking cessation intervention for women. In addition, short bouts of exercise decrease cigarette cravings and withdrawal symptoms among temporarily abstinent smokers. However, results from exercise-based smoking cessation interventions to date have been mostly nonsignificant. This paper describes the theoretical mechanisms (psychological, behavioral, physiological and neurobiological) and practical reasons underlying our belief that exercise-based smoking cessation interventions should not yet be abandoned despite their current paucity of supporting evidence. It also presents ideas for modifying future exercise-based smoking cessation interventions to increase adherence and, as a result, more accurately evaluate the effect of exercise on smoking cessation.

Tobacco use and insufficient physical activity are among the four leading behavioral causes of mortality worldwide [1]. Researchers estimate that 23.7% of adults across the globe were current smokers in 2010 [2]. Although the prevalence of smoking has steeply declined in most developed countries in recent decades, tobacco use still accounts for approximately 5.4 million deaths annually, making it the second leading cause of death worldwide [3]. In the USA, recent estimates suggest that 19.3% of adults, including 21.5% of men and 17.3% of women, currently smoke [4]. Whereas smoking has been recognized as a leading behavioral risk factor for decades, physical inactivity has emerged relatively recently as an increasingly global epidemic [5]. Approximately 31% of adults, including 28% of men and 34% of women, around the world were insufficiently active in 2008 [6]. As a result, approximately 5.3 of the 57 million annual deaths (9% of global premature mortality) are attributable to insufficient physical activity [7]. Physical inactivity is particularly prevalent in the Americas, where nearly 50% of women were reportedly insufficiently active in 2008 [1]. In the USA, less than one-third of the adult population reportedly engages in regular leisure-time physical activity [8]; women and longterm smokers are even less likely to exercise regularly [9]. Leading researchers predict that insufficient physical activity will overtake smoking to become the biggest public health problem of the 21st century [10].

Smoking

Although approximately 40% of daily smokers make at least one serious quit attempt each year [11], only a small percentage succeeds [12]. Self-help approaches, single or combined pharmacotherapies, and psychological/counseling interventions with or without pharmacotherapies are the most common aids used for smoking cessation. These methods are associated with short-term cessation, but long-term outcomes are poor, with success rates increased by only approximately 7–9% [13–15]. This suggests that mainstream cessation aids do not work well for many smokers, who may benefit from alternative therapies. Although a lower percentage of women than men smoke [4], women are less likely to successfully quit smoking in cessation trials [16]. Possible explanations for women’s lower likelihood of quitting include their tendency to smoke to help prevent or alleviate negative mood/affect, depression, and/or postcessation weight gain [17]. Interventions designed to address these concerns may increase odds of cessation among women [18–20].

Exercise

In addition to boosting physical health, exercise also improves mood and well-being [21]. Indeed, research has shown that regular, moderate intensity exercise may alleviate symptoms of depression and anxiety and improve quality of life [21–26]. Because women are particularly susceptible to mood disorders, exercise may be especially beneficial for them [27]. However, despite their reported desire to exercise, most women struggle to maintain a regular exercise regimen [28]. A variety of factors contribute to this discrepancy including a lack of motivation, time, access to facilities or equipment, energy, workout partner, and self-efficacy [29–34]. Moreover, sedentary individuals often experience discomfort (e.g., pain or breathing difficulties) during initial attempts at sustained exercise, further discouraging them from making exercise part of their daily routine [35]. A multipronged approach that addresses barriers to exercise at multiple levels of influence (i.e., individual, social, organizational, environmental and policy) may be necessary to help individuals adopt and maintain a regular exercise program [36,37]. Likewise, an emphasis on the immediate rewards/benefits of exercise (e.g., mood enhancement and energy boost), may resonate with and motivate individuals to exercise more so than a focus on the delayed, longer-term effects (e.g., reduced risk of chronic diseases) [38].

Exercise & smoking

Numerous studies in the past two decades have examined the immediate and long-term effects of exercise on smoking-related variables (e.g., nicotine cravings and withdrawal symptoms) and smoking cessation, respectively. The bodies of literature examining each of these specialized topics, with a particular focus on women-only studies, are summarized below.

Short-term effects of exercise on smoking

A 2007 systematic review of 14 studies [39] and a 2012 update including an additional 15 studies [40] examined the relationships between short bouts of exercise and one or more of the following outcomes: nicotine withdrawal symptoms, cigarette cravings, mood/affect and subsequent smoking behavior. In addition, a recent meta-analysis obtained individual participant data from the trials included in the two systematic reviews of acute studies and examined the variables ‘desire to smoke’ and ‘strength of desire’ to smoke [41]. The reviews and meta-analysis found that exercise reduces withdrawal and cravings, improves mood/affect, and/or reduces both desire and strength of desire to smoke among nicotine-dependent but temporarily abstinent (ranging from 30 min to 24 h) smokers [39–41]. A handful of the reviewed studies [42–44] were conducted on participants enrolled in a smoking cessation trial, who had been abstinent for periods ranging from approximately 8 h prior to beginning the cessation trial [45] to 1–10 weeks post-quit date [42–44] at the time of the acute study portion of the trial; however, the majority of the trials assessing acute effects enrolled current smokers who were not actively attempting to quit. Other effects of exercise in these studies included increased concentration and decreased negative effect, stress levels and depressive and anxiety symptoms. The studies’ protocols varied in terms of duration, intensity and type of exercise, most of which resulted in significant short-term effects. Few of the studies measured the effects of the interventions beyond 30 min post-intervention; therefore, the lasting effects of brief exercise are less certain. Altogether, the studies in these reviews indicated that short bouts of exercise decrease cigarette cravings among temporarily abstinent smokers, suggesting that frequent bouts of exercise may potentially help smokers quit [39–41].

Most of the 29 reviewed studies included both men and women, and no sex differences were identified in those that compared them. However, four of the trials [42–44], all of which were also substudies of cessation interventions, including two that were conducted on the same sample during different phases of the intervention [43], enrolled only women. Brief descriptions of the women-only studies included in these reviews are located in Table 1.

Table 1.

Summaries of the four women-only studies from two systematic reviews^† examining the immediate effects of acute exercise on nicotine cravings, withdrawal symptoms and mood/affect.

Study; country	Participants	Abstinence period	Intervention	Design	Measures	Results	Ref.
Bock et al. (1999); USA	Group 1 = 24 group 2 = 44 Both groups inactive; mean cigarettes = 20/day; mean age = 38 years	Throughout smoking cessation intervention trial participation	(a) 30–40 min, 60–85% HRR, aerobic activity (groups 1 and 2); (b) Equal contact passive; All groups (a1, a2 and b) were involved in an 11-week cessation trial	Within subjects (pre- and post-aerobic activity or equal contact passive control time)	PANAS; ESR; cravings	(a) Groups 1 and 2: reduced negative affect, nicotine withdrawal and cigarette cravings in all weeks (5–10) after quit date; No effect on positive affect	[44]
Harper et al. (2011) Study 1; Canada	n = 119; mean age = 41 years; mean cigarettes = 17/day	(a) 1 week (b) 7 weeks (c) 10 weeks (all postquit date)	20 min on choice of treadmill, rowing machine, stair climber or stationary bike (a) moderate intensity exercise (50–60% HRR); (b) vigorous intensity exercise (>70% HRR); (c) vigorous intensity exercise(>70% HRR)	Substudy of exercise plus NRT smoking cessation intervention trial; assessment pre and IP	Cravings measure (Shiffman–Jarvik Withdrawal Scale)	Significant reductions in craving following exercise at time points (a), (b) and (c); significant reductions in psychological withdrawal and sedation at time points (a) and (b)	[43]
Harper et al. (2011) Study 2; Canada	n = 58; mean age = 43 years; mean cigarettes = 18/day	1-week postquit date	20 min on choice of treadmill, rowing machine, stair climber or stationary bike; Moderate intensity exercise (50–60% HRR); Participants categorized into high EX-EXP and low EX-EXP and high EX-CRED and low EX-CRED	Substudy of exercise plus NRT smoking cessation intervention trial; ECQ: pre; withdrawal scale: pre and IP	Cravings measure (Shiffman–Jarvik Withdrawal Scale); ECQ	High EX-EXP > reduction in craving following exercise compared with low EX-EXP; high EX-CRED> reduction in craving following exercise compared with low EX-CRED	[43]
Williams et al. (2011); USA	n = 60; mean age = 42 years; mean cigarettes range = 11–20/day; mean FTND = 4.8	Each contact (three-times/week) throughout an 8-week smoking cessation program	(a) 50 min brisk walking on a treadmill, three-times per week; (b) 30 min watching videos on health and lifestyle issues, three-times per week	Between-subjects design RCT; assessed cravings and affect pre and IP; treatment at each session throughout an 8-week smoking cessation program	Cravings measure (desire to smoke); ADACL	Nonsignificant trend toward time × treatment interaction for cravings: (a) < (b); Significant time × treatment interaction for energy and tiredness in favor of (a); no effect on tension or calmness	[42]

^†Reviews are [39,41].

ADACL: The Activation–Deactivation Adjective Checklist; ECQ: Treatment expectancy and credibility questionnaire; ESR: Evening Symptom Report; FTND: Fagerström Test for Nicotine Dependence; High EX-CRED: High exercise credibility group; High EX-EXP: High exercise expectancy group; HRR: Heart rate reserve; IP: Immediately post; Low EX-CRED: Low exercise credibility group; Low EX-EXP: Low exercise expectancy group; NRT: Nicotine replacement therapy; PANAS: Positive and Negative Affect Schedule.

Long-term effects of exercise on smoking cessation

Long-term studies summarized in a previous systematic review

A recently updated systematic review summarized 15 unique randomized controlled smoking cessation trials that examined exercise as the primary or supplementary intervention [46]. Seven of these studies, which are summarized in Table 2, included exclusively female samples [47–53]. Results of these trials, which focused primarily on short- and longer-term (12–16 months) smoking cessation rates, were somewhat discouraging compared with those examining responses to acute exercise among temporarily abstinent smokers not intending to quit. The exercise intervention groups in two [47,51] of the seven women-only studies had significantly higher post-treatment cessation rates than their counterpart control groups. Of these two studies, one demonstrated a lasting benefit of exercise on cessation at 3- and 12-month follow-ups [47]. Another women-only study failed to show a between-group difference at post-treatment or 12-month follow-up but detected a significant difference at the 3-month follow-up and a significant relationship between exercise adherence and post-treatment smoking abstinence [48]. Furthermore, results from another three of the seven studies revealed trends toward higher post-intervention cessation rates among exercise group participants [49,50,52].

Table 2.

The seven women-only trials included in the Cochrane review ‘Exercise interventions for smoking cessation’^†.

Study; country	Participants	Interventions	Outcomes‡	Percentage abstinence‡					Ref.
Kinnunen et al. (2008); USA	n = 182; mean age = 39; mean CPD = 19; exercise <3-times a week	(a) Intervention 1: CV equipment, individual, facility, 40 min, 60–80% HR maximum (twice a week for 5 weeks, then once a week for 14 weeks) + CP (once a week for 19 weeks) + nicotine gum (b) Intervention 2: CP and nicotine gum as (a) + health education for the same number of sessions as for exercise in (a) (c) Control: CP and nicotine gum as (a)	Prolonged abstinence; validation: CO, cotinine Follow-up: (i) 1 week (ii) 1 month (iii) 4 months (ET) (iv) 12 months		(i)	(ii)	(iii/ET)	(iv)	[50]
				(a)	60	41	24	10
				(b)	54	39	23	12
				(c)	38	27	15	6
Marcus et al. (1991); USA	n = 20; mean age = 39; mean CPD = 28; exercise <once a week	(a) Intervention: CV equipment, group, facility, 30–45 min, 70–85% HR maximum, 30-times a week for 15 weeks + CP (twice a week for 4 weeks) (b) Control: CP only as (a)	7-day PPA; validation: saliva cotinine <10 ng/ml Follow-up: (i) 1 month (ii) 3 months (iii) 12 months		ET	(i)	(ii)	(iii)	[51]
				(a)	50	40	30	20
				(b)	0	0	0	0
Marcus et al. (1995); USA	n = 20; mean age = 38; mean CPD = 23; exercise <once a week	(a) Intervention: CV equipment, group, facility, 30–40 min, 60–85% HRR (three-times a week for 15 weeks) + CP (once a week for 12 weeks) (b) Control: CP as (a) + health education (three-times a week for 15 weeks)	7 day PPA; validation: saliva cotinine <10 ng/ml Follow-up: (i) 1 month (ii) 3 months (iii) 12 months		ET	(i)	(ii)	(iii)	[49]
				(a)	30	30	30	30
				(b)	10	10	10	10
Marcus et al. (1999); USA	n = 281; mean age = 40; mean CPD = 22; exercise <twice a week	(a) Intervention: CV equipment, group, supervised facility, 30–40 min, 60–85% HRR (three-times a week for 12 weeks) + CP (once a week for 12 weeks) (b) Control: CP as (a) + health education (three-times a week for 12 weeks)	Continuous abstinence; validation: saliva cotinine <10 ng/ml, CO < 8 ppm Follow-up: (i) 3 months (ii) 12 months		ET	(i)*	(ii)*		[47]
				(a)	19	16	12
				(b)	10	8	5
Marcus et al. (2005); USA	n = 217; mean age = 43; mean CPD = 21; exercise ≤90 min a week	(a) Intervention: CV various, group/individual, home/facility, 45 min, 45–59% HRR (facility: once/week for 8 weeks, goal: 165 min/week) + CP (once a week for 8 weeks) (b) Control: CP as (a) + health education (once a week for 8 weeks)	Continuous abstinence; validation: saliva cotinine <10 ng/ml, CO < 8 ppm Follow-up: (i) 3 months (ii) 12 months		ET	(i)	(ii)		[48]
				(a)	15	7	1
				(b)	11	4	1
Prapavessis et al. (2007); New Zealand	n = 142; mean age 38; exercise <twice a week; (excludes 21 pretreatment dropouts)	(a) Intervention 1: CV activity: various, group/facility, 45 min, 60–75% HRR (three-times a week for 12 weeks) + CP (three-times a week for 12 weeks); (b) Intervention 2: exercise as (a) + nicotine patches; (c) Intervention 3: cognitive behavioral cessation program three-times a week for 12 weeks (d) Intervention 4: as (c) + nicotine patches	Continuous abstinence; validation: saliva cotinine <10 ng/ml, CO < 10 ppm Follow-up: (i) 3 months (ii) 12 months		ET	(i)	(ii)		[52]
				(a)	49	20	9
				(b)	23	13	9
				(c)	16	9	4
				(d)	20	7	4
Russell et al. (1988); USA	n = 42; mean age = 28; mean CPD = 23	(a) Intervention 1: walk/jog: group/individual, facility/home, 20–30 min, 70–80% HR maximum, (three-times a week for 9 weeks) + CP (four-times a week for 1 week) (b) Control 1: CP as (a) + health education (once a week for 9 weeks) (c) Control 2: CP as (a)	Not specified; validation: CO Follow-up: (i) 1 month (ii) 4 months (iii) 16 months		ET	(i)	(ii)	(iii)	[53]
				(a + b + c)	83	73	49	34
				NS; data for separate groups not provided

^†Cochrane review: [46].

^‡Using the strictest definition of abstinence employed in each study.

^*Denotes significant differences between groups. (a) vs (b) at ET and (i) (p = 0.03 for both), and at (ii) (p = 0.05).

CO: Carbon monoxide; CP: Smoking cessation program; CPD: Cigarettes per day; CV: Cardiovascular; ET: End of treatment; HR: Heart rate; HRR: Heart rate reserve; NS: Not significant; PPA: Point prevalence abstinence.

Whereas most of the trials only compared exercise with a standard treatment control group, two of them [50,52] also incorporated a nicotine replacement therapy (NRT) comparison group. Results were null in one trial [50] but favored the combination of exercise plus NRT versus exercise alone post-treatment and at the 12-month follow-up in the other trial [52].

Intervention approaches varied widely on numerous variables including, but not limited to, the following: exercise setting; type of exercise prescribed; length of intervention; timing of exercise initiation; exercise intensity and length of follow-up period. None of these characteristics were associated with unequivocally better results. In effect, specific types of components to include in future intervention efforts could not be ascertained.

Despite the lack of convincing evidence produced by the reviewed studies, the authors of the review concluded that exercise has the potential to be an effective smoking cessation intervention [46]. They called for studies with improved designs, greater sample sizes and sufficient exercise intensities, among other recommendations.

Additional long-term studies not included in the previous review

Although the Cochrane review summarizing studies that examine the effect of exercise on smoking cessation is extraordinarily comprehensive, it excludes studies that do not meet the stringent, standardized requirements set by the Cochrane Collaboration [54]. An additional three completed women-only studies that did not meet the minimum 6-month follow-up time specified by the Cochrane review [55–57], two studies that were not yet published at the time of the 2012 review update but appear to meet Cochrane standards [58,59], and two ongoing women-only studies expected to meet Cochrane standards at their conclusion [60,61] were identified. These seven studies are summarized in Table 3. The five completed studies reported a mix of significant differences [56,59], nonsignificant trends favoring exercise over control groups [55,59] and no between-group differences [57,58].

Table 3.

Seven trials with women only, either not included in the Cochrane review ‘Exercise interventions for smoking cessation’^† or currently in progress.

Study; country	Participants	Interventions	Outcomes‡	Percentage abstinence‡					Ref.
Bock et al. (2012); USA	n = 55; mean age = 46; mean CPD = 16; mean FTND = 5; exercise ≤3 days a week moderate or ≤2 days a week vigorous prior to study entry	(a) Intervention: yoga: group, facility, 60 min (twice a week for 8 weeks) + CP (once a week for 8 weeks); (b) Control: CP as (a) + health education (twice a week for 8 weeks)	Continuous abstinence; validation: saliva cotinine <15 ng/ml Follow-up: (i) 3 months (ii) 6 months		ET*	(i)	(ii)		[59]
				(a)	41	22	19
				(b)	13	9	13
Chaney and Sheriff (2008); USA	n = 101; mean age = 47	(a) Intervention: circuit training: individual, facility, 30 min (three-times a week for 8 weeks) + CP (once/week for 8 weeks) + nicotine lozenges; (b) Control: CP as (a) + nicotine lozenges	Not specified No follow-up		ET*				[56]
				(a)	82
				(b)	59
Jung et al. (2010); Canada	Target sample size: n = 420 Eligible if: >10 CPD, ≤ two 30-min sessions of exercise in previous 6 months	All participants: CV equipment: group, facility, 45 min, 70–75% HR maximum (three-times a week for 14 weeks) + nicotine patches; (a) Intervention 1: exercise counselling (once a week for 7 weeks + telephone support); (b) Intervention 2: as (a) + smoking relapse prevention booklets; (c) Control 1: relapse prevention booklets + health education (once a week for 7 weeks + telephone messages); (d) Control 2: health education only as (c)	Continuous abstinence; Validation: saliva cotinine <10 ng/ml, CO < 10 ppm Follow-up: (i) 26 weeks (ii) 56 weeks	Ongoing trial					[60]
Ussher et al. (2012); UK	Target sample size: n = 866 Eligible if: 10–24 weeks pregnant, smoking at ≤ one CPD now ≤5 CPD before pregnancy, aged 16–50 years	(a) Intervention: treadmill walking, 30 min (14 sessions over 8 weeks) + CP (once a week for 6 weeks) (b) Control CP as (a)	Continuous abstinence; validation: saliva cotinine <10 ng/ml, CO < 8 ppm Follow-up: (i) end of pregnancy (ii) 6 months postpartum (self-report only)	Ongoing trial					[61]
Vickers et al. (2009); USA	n = 60; mean age = 41; mean CPD = 21; exercising <20 min/day on fewer than 3 days/week, depression score >16 on CES-D	(a) Intervention: exercise counselling (once a week for 10 weeks) + CP (once a week for 10 weeks) + nicotine patches; (b) Control: health education (once a week for 10 weeks) + CP as (a) + nicotine patches	7-day PPA; Validation: CO < 8 ppm, urine cotinine Follow-up: (i) 24 weeks		ET	(i)			[57]
				(a)	5	1
				(b)	7	1
Whiteley et al. (2012); USA	n = 330; mean age = 44; mean CPD = 17; ≤20 min of vigorous activity twice per week	(a) Intervention: CV various + resistance training: individual, facility, 40–65 min, 64–85% HR maximum (three-times a week for 12 weeks) + CP (once a week for 12 weeks). (b) Control: CP as (a) + health education (four sessions over 12 weeks)	Continuous abstinence; validation: saliva cotinine <10 ng/ml, CO < 10 ppm Follow-up: (i) 12 weeks (ii) 24 weeks (iii) 52 weeks		ET	(i)	(ii)	(iii)	[58]
				(a)	14	8	7	7
				(b)	14	9	5	4
Williams et al. (2010); USA	n = 60; mean age = 42; <60 min/week of routine exercise	(a) Intervention: treadmill walking: individual, facility, 50 min, 70% maximum HR (three-times a week for 8 weeks) + single session CP + nicotine patches (b) Control: CP as (a) + health education (three-times a week for 8 weeks)	Prolonged abstinence; validation: CO < 10 ppm Follow-up: (i) 1 month		ET	(i)			[55]
				(a)	35	18
				(b)	20	13

^†Cochrane review: [46].

^‡Using the strictest definition of abstinence employed in each study.

^*p = 0.03.

CES-D: Center for Epidemiologic Studies Depression Scale; CO: Carbon monoxide; CP: Smoking cessation program; CPD: Cigarettes per day; CV: Cardiovascular; ET: End of treatment; FTND: Fagerström Test for Nicotine Dependence; HR: Heart rate; PPA: Point prevalence abstinence.

Overall summary of long-term studies

Clearly, the long-term studies on this topic vary considerably. We considered creating a graph or another more straightforward summary depicting the studies’ smoking cessation outcomes. However, we concluded that the heterogeneity of the studies’ results would make a direct comparison misleading. In addition to a wide variety of exercise prescriptions (e.g., type, duration and intensity) and additional intervention components (e.g., group-based smoking cessation treatment or nicotine patch), the studies and their smoking cessation results were heterogeneous in terms of the following critical variables:

• Sample sizes varied substantially. Although weighted averages could be calculated to account for differences in sample size at baseline, sample sizes used to calculate post-intervention and follow-up outcomes (e.g., intent-to-treat, completers only or not reported, among others) were too sizable to be properly accounted for in a direct comparison of studies;

• Length of treatment (and therefore timing of post-intervention) and number and timing of follow-ups varied considerably;

• Standards for counting as ‘quit’ (e.g., past 24 h, past 7 days or continuous abstinence since last assessment, among others) varied across studies and even within studies at different time points;

• Likewise, stringency of verifying self-reported smoking status (e.g., no objective verification, carbon monoxide monitoring or cotinine) varied across studies and time points;

• Some studies had multiple intervention and/or control groups, which could typically be collapsed to create overall treatment group outcomes. However, the studies varied in terms of what they considered closer to an intervention or control group, making them difficult to differentiate. For example, Prapavessis et al. labeled all four of the treatments to which participants were randomized in their trial as intervention groups even though only two of them included an exercise component [52], and Kinnunen et al. labeled two treatments as intervention groups even though only one included an exercise component [50]. By contrast, one of the two treatments Russell et al. [53] labeled as a control group mirrored the additional intervention groups without an exercise component described by Prapavessis [52] and Kinnunen [50].

Considerations/limitations of previous research

In summary, the literature on exercise-based smoking cessation interventions for women is, to date, inconclusive. In addition to being comprised of relatively few studies, the intervention literature is limited by study characteristics such as high attrition, low adherence and inadequate sample sizes. Furthermore, the variability of exercise (e.g., frequency, intensity, duration and type) and supplementary behavioral smoking cessation programs utilized in the studies makes pooling studies to assess their collective efficacy challenging if not impossible. Despite the lack of convincing evidence that exercise could be a promising smoking cessation intervention, the evidence supporting the efficacy of acute exercise for short-term reductions in cigarette cravings, desire to smoke, strength of desire to smoke and negative mood/affect, among other positive outcomes, is well-established. This evidence, combined with the theoretical support for exercise as a smoking cessation intervention, drives the continued investigation in to this potentially effective treatment. The purpose of this paper is not to systematically and thoroughly review the literature on exercise and smoking cessation among women, but rather to present evidence for and theories underlying the reasons we believe the area should not yet be abandoned. Although other papers reviewing the literature on smoking and exercise have been published, they have not focused specifically on women nor discussed in detail the theoretical reasons for using exercise as a treatment for smoking cessation.

Discussion

In light of an increasing focus on evidence-based treatments, why are we advocating for continued inquiry into exercise as a smoking cessation treatment for women despite a clear lack of convincing evidence? We present our case in the following sections.

Theoretical mechanisms

Psychological

A large and growing body of scientific research supports a positive relationship between physical exercise and psychological well-being [62–65], with particular attention to depression and mood/affect [21,64,66–69] and anxiety [64,67,70,71]. Like the exercise and smoking literature, the exercise and mental health literature falls under two general umbrellas: acute and chronic effects. Unlike the exercise and smoking literature, however, evidence is strong for both the acute and chronic effects of exercise on positive emotional well-being [21,66–68,70,72–74]. Ongoing research on the acute effects of exercise on mood/affect is investigating the optimal dose (e.g., type, duration and intensity) for maximal mood enhancement [74], which appears to vary on an individual basis [73]. Epidemiological research reveals a significant relationship between regular physical activity and better mental health, and multiple randomized controlled trials have demonstrated significant effects of regular exercise on depression [68,69] and anxiety [70], although a causal relationship has been questioned [75].

Smoking is associated with both poorer mental health and insufficient physical activity [76,77]. Relative to their nonsmoking counterparts, women smokers face higher rates of depression, anxiety and other mental health disorders [78], and they engage in less physical activity [77]. A major barrier to smoking cessation identified by a large percentage of smokers is negative mood/affect, which is increased by nicotine withdrawal in smokers [42,44]. Furthermore, smoking is more prevalent among individuals with mood disorders, who reportedly smoke to help alleviate their depressive symptoms [17,20]. Thus, one of the primary mechanisms through which exercise may be helpful as a treatment for smoking cessation is its ability to help alleviate the negative mood/affect symptoms associated with nicotine withdrawal as well as the chronic mental health problems (e.g., depression and anxiety) that are particularly prevalent among women smokers.

Behavioral

Smoking is a behavior perpetuated and reinforced by various environmental cues, psychological responses and real or anticipated rewards. Although smoking cessation targets an overall pattern of behavior, the quitting process involves avoiding one cigarette at a time when a craving arises and the automatic behavior (i.e., smoking) is evoked. Because individuals typically smoke in numerous contexts, cigarettes become associated with a multitude of places, activities and times of day, among others. These associations increase the difficulty of cessation owing to the constant environmental reminders smokers face. Therefore, a behavioral substitute that smokers can engage in instead of smoking a cigarette may help them quit. To rival the addictiveness of cigarettes, the behavior must be immediately rewarding, habit-forming, easily accessible and sustainable. Research examining the acute effects of brief exercise bouts on cigarette cravings, desire to smoke, strength of desire, withdrawal symptoms and mood/affect among temporarily abstinent smokers suggests that exercise works well as a substitute for smoking in terms of its effects on these variables. Thus, it appears to be immediately rewarding. Another body of research indicates that exercise is habit-forming for many individuals and may even be addictive for a small percentage of the population [79,80]. Basic but effective types of exercise such as walking, resistance training using one’s own body weight and even chair-based activities can be done essentially anywhere and at any time, making exercise as easily accessible as smoking (and perhaps more accessible in some circumstances, such as in places where smoking is limited or banned). Finally, exercise is safe and sustainable over time for most people [81,82].

Thus, engaging in short bouts of exercise instead of smoking when cravings arise appears to be a theoretically viable behavioral substitute for smoking. Research indicates that the health benefits of exercise are essentially equal [83] regardless of whether exercise is obtained in one prolonged session (e.g., 20–30+ min) or multiple shorter sessions (e.g., 5–15 min each). In addition, a systematic review revealed no consistent differences in adherence rates between groups engaging in continuous versus intermittent bouts of exercise [84]. Exercise intervention research indicates that even individuals who are apparently extremely motivated to begin an exercise program (assumed by their decision to join an exercise intervention research study) have trouble maintaining exercise over time. This lack of longevity may be exaggerated in smokers who are attempting to change two major behaviors (smoking and exercise) simultaneously. Therefore, exercise-based smoking cessation interventions may want to consider: gradually increasing the duration of each exercise bout to minimize initial discomfort and, in effect, help avoid early dropout; and incorporating both short, intermittent bouts of exercise when cravings arise to prevent acute lapses as well as sustained, regular bouts of exercise to help build a strong foundation supporting an overall healthy lifestyle.

Neurobiological

Neuroimaging research demonstrates that exercise activates brain reward pathways similar to those activated by substances of abuse (e.g., nicotine or stimulant drugs), suggesting that exercise may become addictive for some individuals [85,86]. Although a small percentage of the population develops an extreme, unhealthy exercise addiction due in part to exercise’s ability to evoke this reward pathway [79,80], the majority of individuals benefit from this phenomenon by exercising regularly (in a nonpathological pattern) in anticipation of this internally rewarding experience [87]. Indeed, the effects of exercise on multiple structures and systems in the brain may help individuals maintain regular exercise habits by facilitating the desirable behavioral, psychological and physiological effects associated with exercise [88,89]. Research examining the processes underlying exercise-induced euphoria has determined that increased activity of certain neurotransmitters is at least partially responsible [80]. β-endorphins and the neurotransmitters epinephrine, norepinephrine, serotonin and dopamine play verifiable roles in the innately rewarding aspects of exercise [90–93]. These neurotransmitters, especially dopamine, have also been implicated in nicotine addiction [94,95]. Recent neuroplasticity research suggests that the effect of exercise on the serotonergic system may partially explain the link between exercise and stress-related psychiatric disorders such as depression and anxiety [96]. Similarly, recent neuroimaging research suggests that exercise evokes hippocampal neurogenesis and, in effect, counteracts the impact of stress and reduces stress-related disorders such as depression [97]. Of all the mechanisms underlying the proposed incorporation of exercise into smoking cessation treatment, the neurobiological has garnered the most recent interest among animal and human researchers alike, likely due to advances in knowledge, technological equipment and other resources available to study this aspect of the relationship between exercise and the reward system. Although this topic is undoubtedly scientifically interesting, understanding the precise neurobiological effects of exercise is not absolutely necessary to advancing the exercise/smoking cessation field.

Physiological

Unlike a number of other smoking cessation strategies, exercise has a unique potential to reduce many of the negative physiological effects of smoking. Smoking independently increases the risk of a host of different diseases, many of which continue even after cessation. To date, research shows that exercise may be able to reduce some of these risks and the severity of the damage caused by smoking. For example, exercise has been shown to prevent or diminish the harmful effects of smoking on lung function. Recent studies have indicated that aerobic exercise, resistance training or a combined program of both aerobic exercise and resistance training can have significant positive effects on multiple indicators of lung function in current or former smokers [98,99]. Other smoking-induced risks that could potentially be reduced with exercise include Type 2 diabetes, systemic inflammation (i.e., elevated C-reactive protein levels), dyslipidemia, arterial stiffness and low bone mineral density. More specifically, combined aerobic and resistance training has been shown to lower glycosylated hemoglobin (HbA1c), an indicator of Type 2 diabetes severity [100]; daily brisk walking is known to reduce arterial stiffness and improve endothelium-dependent dilation [101]; regular exercise training and improved fitness are associated with reductions in systemic inflammation [102]; aerobic exercise and resistance training have both been shown to beneficially impact cholesterol and triglyceride levels [103,104]; and weight-bearing physical activity is widely supported and promoted as an effective way to increase bone mineral density, reduce fracture risk, and possibly stimulate a systemic response (i.e., increase bone turnover) [105–108].

Exercise & a healthy lifestyle

In addition to the aforementioned theoretical mechanisms, smoking cessation may be enhanced by exercise through its relationship with other health-enhancing behaviors. For example, regular exercise is a healthy behavior that is compatible with and often one component of a generally healthy lifestyle, including other positive health behaviors such as a healthy diet, good sleep hygiene, moderate alcohol use (or complete abstinence if alcohol abuse is a problem) and abstinence from illicit drugs. Although some cigarette smokers also engage in these and other healthy behaviors, smokers are more likely to be sedentary, regularly consume unhealthy foods, drink alcohol excessively and use illicit drugs, among other unhealthy lifestyle behaviors [109]. Because nicotine is a stimulant, smoking may interfere with sleep quality, especially when done too close to bedtime [110]. Research shows that the taste of cigarettes is less palatable when nutritious foods such as fruit and vegetables, noncaffeinated beverages and dairy products are consumed, and more palatable with meat products, caffeinated beverages and alcohol [111]. In effect, smokers tend to eat less nutritious foods that make cigarettes taste better [112]. Smoking makes exercise more difficult because it raises resting heart rate and blood pressure (causing the heart to work harder), interferes with circulation (the carbon monoxide in cigarettes attaches itself to hemoglobin more easily than oxygen does, which reduces the amount of oxygen available to the tissues) and increases the risk of blood clots by making the blood more ‘sticky’ [113]. In turn, the increased difficulty likely contributes to the lower rates of regular exercise among smokers. On the other hand, engaging in a positive health behavior such as exercise can lead to increased feelings of well-being, vitality, energy and motivation to adopt an overall healthier lifestyle [114], which is less compatible with smoking. As a result, all of these positive health behaviors may combine to increase the odds of initial smoking cessation as well as short- and long-term abstinence maintenance.

Dose–response effects

Perhaps the most compelling reason for exercise is that many exercise-based smoking cessation studies conducted to date have reported a dose–response relationship between exercise and smoking cessation and/or reduction. Specifically, increasingly higher rates of adherence to the exercise prescription are associated with increased odds of quitting and/or greater reductions in daily smoking rates [46]. Thus, evidence suggests that, just as medications only work if taken as prescribed, exercise only works if the dose is high enough. Unfortunately, minimum and optimal exercise doses have not yet been established [46].

Why women?

Although both men and women smokers desiring to quit may benefit from exercising to support the cessation process, women have received more attention in this area for a few central reasons. As mentioned previously, although fewer women than men smoke [4], women are less likely to quit in smoking cessation clinical trials [16]. Although the reasons for this sex disparity are likely complicated, research suggests that mainstream cessation options (e.g., NRT) do not address many of the primary reasons why women smoke [16,17]. Perhaps women are more affected by the behavioral and psychological aspects of smoking. In particular, women appear to have a greater sensitivity to the mood/affect regulation and the real or perceived weight control effects of smoking [16,17]. Thus, women attempting to quit may quickly relapse when they experience symptoms such as negative mood/affect (related to nicotine withdrawal), increased appetite and weight concerns [16,17].

Indeed, women are particularly susceptible to the smoking/affect regulation cycle because they experience higher rates of mood disorders and are more likely than men to report smoking for mood regulation [16,17]. The mounting evidence indicating that acute exercise reduces negative mood/affect and regular exercise helps prevent and/or treat depressive episodes suggests that exercise may be an ideal smoking cessation treatment, particularly for women who are prone to negative mood/depression and smoke at least in part to regulate their mood [42]. Likewise, recent evidence indicates that postcessation weight gain may be even higher than previously estimated [115], somewhat legitimizing this previously labeled ‘exaggerated’ fear among women smokers. However, when combined with a healthy diet, exercise helps with long-term weight control [116]; both of these positive health behaviors are associated with greater odds of smoking cessation and maintenance [117,118]. Unfortunately, rates of regular exercise among women, particularly women smokers, are extremely low [28]. This group may require substantial amounts of motivation and resources to begin and maintain an exercise program [30,34], but they stand to benefit tremendously from exercise because of their numerous risk factors. Addressing barriers to exercise at multiple levels of influence may be particularly important for women, whose motivation to exercise is often enhanced by social support, group-based exercise and other factors beyond those within the individual level of influence [36,37]. Mounting evidence suggests that policy changes may be necessary to evoke long-term changes in physical activity levels in our increasingly sedentary environment [6,119].

In line with recent research, emphasizing the positive acute effects of exercise (e.g., immediate improvements in mood/affect, anxiety, and nicotine cravings) may be more effective than focusing on the long-term effects of exercise (e.g., lower risk of chronic diseases and premature mortality) [38]. Because smokers are especially prone to choosing immediate rewards over long-term consequences, rebranding exercise to them as a behavior with instantaneous benefits similar to those evoked by smoking [39–41] may increase their motivation to try exercising instead when they desire to smoke.

Special consideration: pregnancy/postpartum

In high-income countries, smoking during pregnancy is a serious public health concern, as it is the major preventable cause of poor health outcomes for women and their babies. In nonpregnant smokers, NRT, bupropion and varenicline are effective pharmacological smoking cessation treatments [13–15]. However, the safety and efficacy of NRT during pregnancy is not yet known [120], many pregnant women are reluctant to use NRT [121] and other smoking cessation medications are contraindicated during pregnancy [122]. Behavioral support can increase smoking cessation rates during pregnancy [123]; identifying other effective nonpharmacological adjunctive interventions to support cessation during pregnancy is critical.

A number of compelling arguments for testing physical activity as a smoking cessation aid during pregnancy have been identified. Moderate intensity exercise is recommended during pregnancy [124,125] and may represent an attractive and healthy option for pregnant smokers who are reluctant to use NRT [121], wish to use exercise as an adjunct to NRT, and/or fear postcessation weight gain or postpartum weight and tone issues [126]. Pregnant smokers express interest in using exercise to help them quit [127], and two pilot studies have demonstrated the feasibility of an exercise intervention, combining supervised exercise and physical activity consultations, as an aid to smoking cessation during pregnancy [128]. These pilot studies led to a currently in-progress full-scale trial of exercise for smoking cessation in pregnancy and postpartum (Table 3) [61].

Qualitative findings from the Commit to Quit in the YMCAs study

Focus groups were conducted with women who had recently participated in the Commit to Quit in the YMCAs study [58] in order to learn more about their experiences with the trial, which examined the effect of smoking cessation cognitive behavioral therapy (CBT) plus exercise versus CBT plus wellness contact control. The focus group facilitators enquired about the women’s perceived facilitators and barriers to the exercise program, as well as their opinions about the intervention program in general. Women in the CBT plus exercise condition expressed their enthusiasm about the idea of exercise but reported that their perceived barriers to exercise frequently outweighed their perceived benefits; for example, they reported that travelling to the YMCA to exercise was not always convenient, and they had trouble finding the dedicated time to exercise for the >40 min that were prescribed in the study. They also reported that they did not fully appreciate the link between exercise and smoking cessation because the two topics were not explicitly combined in the study, which purposely separated the two issues in order to keep the exercise and control groups as similar as possible on everything other than the exercise prescription and access to the YMCA facility. They also expressed a desire to exercise with other participants rather than alone because they would feel more motivated and supported by the social aspect. Although they desired to exercise while quitting smoking to help alleviate/prevent weight gain associated with cessation, most did not adhere to the exercise protocol well enough to reap that potential benefit. Other comments were inconsistent, with some women preferring: a more (or less) intense exercise prescription; working out regularly with a personal trainer or not having one at all; engaging in supervised or unsupervised exercise; and incorporating strength training in addition to aerobic exercise or not. Overall, the focus groups suggested that a greater degree of tailoring the exercise program to each individual’s preferences may help when designing future iterations of exercise-based smoking cessation interventions.

Conclusion

Overall, the evidence suggesting that acute bouts of exercise reduce cigarette cravings and withdrawal symptoms among temporarily abstinent smokers is strong [39,41]. However, the evidence supporting the role of regular exercise in the context of smoking cessation treatment is relatively weak [46]. Whereas the vast majority of studies investigating the acute effects of exercise have included men and women, a relatively high percentage of exercise-based smoking cessation studies have enrolled only women [47–53,55–61]. Researchers in this specialized field recognize that women often smoke and quit smoking for different reasons than men [16,17]. These interventions have targeted women more frequently than men largely because exercise helps alleviate women’s fear of postcessation weight gain [19] and reduces their cessation-related mood symptoms [42], two major barriers to smoking cessation identified by women smokers [16,17].

Future perspective

We have presented multiple theoretical and practical reasons suggesting that exercise-based smoking cessation programs should work well for women. Common sense says that it should work. At face value, exercise appears to be an ideal smoking cessation tool for women. Unfortunately, to date, science has not supported it. Why not? Undoubtedly, the primary reason is that participants enrolled in previous exercise-based smoking cessation trials have not received the dose of exercise prescribed to them (i.e., they have not adhered to the exercise prescription). This problem is not unique to smoking cessation trials; indeed, low adherence rates plague exercise interventions targeting a wide variety of populations. Unfortunately, it prevents researchers from discovering whether or not exercise would be an effective treatment if it was actually used.

So where does the field go from here? Does the burden fall on behavioral scientists to design interventions that entice participants to exercise more? Does it fall on participants to more actively participate in the studies in which they enroll? Should we completely abandon this idea?

Because of the multitude of reasons presented in this review, we advocate continuing to investigate the role of exercise in smoking cessation. However, we believe that interventions should incorporate acute bouts of exercise in response to immediate cigarette cravings in order to capitalize on the numerous benefits consistently found in the acute effects studies [39,41]. Furthermore, we believe exercise should be tailored not only to each woman’s baseline level of fitness but also to her lifestyle and exercise preferences (e.g., walking outdoors vs in a gym, with a group vs alone and aerobic vs strength, among others.). The wide variety of preferences expressed by the women in the focus groups described earlier suggest that working individually with each woman to design her unique exercise program may increase adherence. Methods used elsewhere for increasing exercise adherence, such as pedometers and telephone check-ins, also need to be more routinely incorporated. More drastic measures may ultimately be necessary; for example, financial incentives have been shown to be effective for both smoking cessation [129] and physical activity [130,131] adherence, and such incentives may be necessary to motivate women smokers to exercise. These types of effectiveness studies would ideally be implemented after efficacy has been established in multiple tightly controlled trials, but tightly controlled studies with this population have proven extremely difficult to accomplish. Therefore, we believe that a radically different approach may need to be taken in order to advance the field.

Executive summary.

• Tobacco use and insufficient physical activity are among the four leading behavioral causes of mortality worldwide.

• Women are less likely to successfully quit smoking in cessation trials due, at least in part, to their tendency to smoke to help prevent or alleviate negative mood/affect, depression and/or postcessation weight gain.

• Exercise, which may also alleviate these symptoms, may increase the odds of cessation among women.

• Numerous studies have demonstrated that exercise reduces withdrawal and cravings, improves mood/affect and/or reduces desire to smoke among temporarily abstinent smokers.

• The literature on exercise-based smoking cessation interventions for women is more limited, variable and inconclusive.

• The purpose of this paper is to present evidence for and theories underlying the reasons that we believe exercise-based smoking cessation interventions should not yet be abandoned despite the currently limited evidence supporting them.

• Theoretical mechanisms supporting exercise-based smoking cessation interventions include psychological, behavioral, neurobiological and physiological.

• Other arguments supporting the case for exercise include the role of exercise in overall healthy lifestyle adoption, evidence of a dose–response effect in previous studies and the ability of exercise to address the unique reasons many women continue to smoke.

• Future exercise-based smoking cessation studies should focus on increasing exercise adherence in order to more accurately evaluate the efficacy of exercise.