This study aimed to explore how a 15-week smoking cessation program containing an exercise intervention and nicotine transdermal patch impacted the cardiovascular risk profile among initially sedentary female smokers. As in many previous treatment studies, both the attrition and relapse rates were high. Nearly 20% dropped out before their prequit visit, and about half of the participants who came to their prequit visit and made a quit attempt had relapsed by the end of the treatment. Clearly, there is a need for more effective treatments for female smokers.
The smoking cessation and exercise program of 15 weeks in duration had a favorable impact on several CVD risk factors. Independent of tobacco abstinence, improvement was seen in inflammation (WBC), prothrombotic factor (RBC), and cardiovascular fitness level (Vo2max). This suggests that even if complete abstinence is not achieved, reduction in tobacco exposure and increase in exercise can improve the cardiovascular risk profile. Regarding lipids, significant results were observed for total cholesterol and the total cholesterol/HDL-C ratio; a significant decrease was seen among abstainers, whereas a slight nonsignificant increase was detected among relapsers. For other lipids, no significant smoking status or time effects were found. Interestingly, heart rate was generally reduced among all, but this decrease was more profound among abstainers. As expected, there was a significant weight gain and BMI increase among all. We also observed a significant increase in Hb A1c.
Some earlier smoking cessation studies have shown a decrease in the inflammatory markers. In a study with NRT, the abstainers and smoking reducers had a greater reduction in WBC count after 6 months.43,44
In line with those earlier findings, we also observed a significant decrease in WBC count. Interestingly, the WBC count decreased in both abstainers and relapsers, and these results did not change even when controlled for nicotine intake or exercise adherence, suggesting that the combination of the cessation program that led to either complete abstinence or reduction in smoking had a beneficial effect on this biomarker. CRP has gained attention as a significant biomarker for cardiac events, and we did find an impact for tobacco abstinence status. There is clear evidence that lifetime tobacco exposure impacts postcessation CRP levels,45,46
although this biomarker may need more time to normalize than 3 months of abstinence, which was our follow-up. Additional reasons may be that although our smokers were sedentary, smoked an average of 16 cigarettes per day, and had relatively high BMI (mean 28.9), the average CRP at baseline (3.3) was not clearly elevated. It is also reflective of healthy middle-aged women with similar BMI and activity levels in the Boston area.47
Previous research has suggested that smoking reduction or cessation leads to decreased levels in the biomarkers of a hypercoagulable state or prothrombotic factors.7,48
We did not find strong evidence to support this among any of our biomarkers. It may be that our follow-up period was not long enough to detect these changes, as some of them were observed at 6 months postcessation.44
Among hemodynamic factors, HR decreased among the participants, but it was more profound among the abstainers than among those who relapsed. Our results support previous findings indicating HR normalization even if NRT is being used.49
Although no significant time or abstinence effects were observed for systolic and diastolic BP, exercise seemed to have important effects on both. Diastolic BP decreased almost significantly (p
=0.11), the average decrease being 2.1 units for abstainers and 0.5 units for those relapsed. More importantly, when adjusting for exercise adherence, which had a significant impact on diastolic BP, the time effect for BP change became clearly nonsignificant. This means that exercise adherence had an important role here. A clinically important issue is that earlier literature reported that smoking cessation was associated with an increased incidence of hypertension.50
In relation to this evidence, our decreasing trend of BP, partly explained by exercise effect, seems to be encouraging.
Previous research on smoking cessation and lipid profile has consistently found an increase in HDL-C and triglycerides.44
We observed some increase in HDL-C among abstainers and an increase among relapsers in triglycerides. These differences did not reach statistical significance. However, both the total cholesterol and, even more promising, the total/HDL-C ratio improved among abstainers only, even after adjustment for exercise adherence or nicotine intake. This suggests that both smoking cessation and exercise had a favorable impact.
Hb A1c is a good indicator of glucose metabolism, reflecting processes of glucose tolerance for several weeks. In our analysis, no statistically significant increase was found among either abstainers or those who relapsed. This is encouraging, as immediate effects of smoking cessation may increase the risk for type 2 diabetes by elevated glucose and triglycerides levels.51
Unfortunately, we did not measure fasting glucose; because fasting glucose is a more sensitive measure,51
it is possible that we could have seen changes at follow-up if we had that outcome measured.
We observed a significant increase in weight and BMI, but this weight increase was relatively small, actually only on average 1.3
kg for abstainers and 1.6
kg for those relapsed. This is clinically clearly less than the average weight increase of 3.8
kg reported in women after smoking cessation.28
When adjusting the time effect for total nicotine intake during the treatment, such as NRT adherence, the time effect disappeared but the abstinence effect became significant. This leads us to suggest that NRT adherence among the abstainers was an important confounder. As a matter of fact, there is evidence that NRT can prevent weight gain to some extent.52
A similar trend was evident with BMI, in that although the time effect was significant, the average BMI increase was only 0.5 units in abstainers and 0.6 among those who relapsed. Here again, adjusting for nicotine intake attenuated the time effect. Further, the effect of exercise adherence seems to be important because time effect was attenuated toward less significance when adjusting for exercise. This leads us to acknowledge the fact that muscle mass weighs more than fat mass. Thus, it is possible that those abstainers who had high adherence to NRT and exercise gained some weight that was mostly due to increase of muscle mass. Unfortunately, we did not measure muscle mass in this study. It remains a question of longer follow-up to see if sustained abstinence or continued exercise or their interaction is responsible for long-term development in weight and BMI after smoking cessation.
Chiolero et al.53
reviewed the relationships among smoking, body weight, and insulin resistance. According to these authors, nicotine increases energy expenditure in the short term and could reduce appetite, explaining why smokers tend to have lower body weight than do nonsmokers and why smoking cessation is frequently followed by weight gain. However, the review suggests that smokers who smoke high number of cigarettes daily actually tend to have greater body weight than do nonsmokers or smokers who smoke only a few cigarettes daily. This is likely to reflect a clustering of risky behaviors conducive to weight gain, such as a low degree of physical activity, poor diet, and smoking. Other factors, such as weight cycling, could also be involved. In the context of the worldwide obesity epidemic and a high prevalence of smoking, greater risk of central obesity and insulin resistance among smokers, particularly female smokers in menopause, is a matter of major concern. Moreover, the previously mentioned risk factors are of high relevance in the target group of this intervention, as all women participating in this trial had two major risk factors, daily smoking and a sedentary lifestyle.
There are several limitations to our study. At baseline, every participant was engaged in at least two health-compromising behaviors identified as CVD risk factors.54,55
Hence, we are dealing with a subgroup of relatively high-risk women in terms of their CVD risk profile, and our intervention results can be generalized only among adult women who initially are daily cigarette smokers and physically inactive. As this trial recruited women from the Boston community and surroundings, however, our sample quite accurately represented sedentary female smokers as well middle-aged women in general in that metropolitan area.43
We recognize that our participants were exposed to nicotine via NRT during abstinence and via tobacco if relapsed. One may ask if NRT contributed to the results where we did not observe improvement among abstainers compared to relapsers. This question may be justified by the fact that nicotine may increase oxidative stress.56
However, nicotine has a very complex pharmacology.57
Long-term use of nicotine as a medication could have adverse cardiovascular effects, such as promoting angiogenesis and arteriogenesis,58
inducing endothelial cell dysfunction, and possibly inducing insulin resistance,57
as well as promoting lipolysis, which could contribute to adverse lipid profiles.59
Although nicotine may adversely affect cardiac output and its related components, its role in atherosclerosis remains controversial. Studies have reported that nicotine alone has caused no change, a decrease, or an increase in nitric oxide levels and that levels commonly found in a smoker have only a minor effect on initiating or proliferating lesion formation.60
Moreover, in our trial, exposure to NRT was for 3 months, a relatively short time in comparison to years of tobacco exposure.
There were 54 participants who completed the baseline visit but did not complete the prequit assessment. Attrition is a critical problem in clinical trials for substance use and dependence; high attrition rates are commonplace and often expected.37,61–63
Pretreatment attrition among smokers has begun to receive attention only recently.61,63
Factors associated with early attrition include younger age, lower socioeconomic status, and history of psychotropic medication use. Low motivation, weight concerns, and affective vulnerabilities, such as depression, also have been reported to predict early attrition. We examined these factors in our earlier exercise studies among women. In our first trial, we found that of the modifiable factors, feelings of guilt and weight concerns predicted attrition.37,61
In the present trial, we conducted similar analysis and found that depression was one of the strongest predictors of an early dropout (unpublished observations). The nonmodifiable factors, such as race (minority), lower education, and number of children at home, remain significant in both studies.
Finally, we investigated only how tobacco abstinence impacted these select CVD biomarkers. We did adjust for treatment adherence and nicotine intake but, given the small sample size, did not examine, for example, BMI changes, caloric intake, or other prospective mediating or moderating factors. Similarly, we assessed the change only at the end of the treatment, that is, 3 months postcessation. Additional studies are needed to measure biomarkers at more frequent intervals, with longer follow-up assessing patterns of changes.
In summary, our results suggest that in as short a time as 3 months, abstinence from or reduced consumption of cigarettes combined with an exercise intervention can reduce immediate tobacco-induced cardiovascular damage measured by such cardiovascular biomarkers as cholesterol and HR. In addition, the current program resulted in relatively low weight gain and triglyceride increases and a significant cardiovascular fitness level increase. Although it may not be clear how exercise or abstinence affects one particular biomarker, it is clear that overall, the impact of this smoking cessation program was more favorable that that of many other reported programs. Given that there is an increasing number of reports where smoking cessation may temporarily elevate the risk for type 2 diabetes,51
even smaller changes can be clinically significant and serve as a motivator for the participants.64