Journal Logo

Basic Sciences: Original Investigations

Exercise, smoking cessation, and short-term changes in serum lipids in women: a preliminary investigation


Author Information
Medicine& Science in Sports & Exercise: September 1998 - Volume 30 - Issue 9 - p 1414-1418
  • Free


Cigarette smokers have higher concentrations of triglycerides, total and very LDL (VLDL) cholesterols, and lower concentrations of HDL cholesterol (HDL-C) and its major apoprotein, apo A-I, than nonsmokers (5). These differences are dose-related, with heavier smokers having higher triglyceride and lower HDL-C concentrations than light smokers. Changes in the lipoprotein profile occur after smoking is first taken up (6) and reverse with smoking cessation (13,22).

Several prospective studies are in general agreement that HDL-C, HDL-C2, and the major apoprotein attached to HDL, apo A-I, increase after smoking cessation (2,9,12-14,16-19,22,23), although there are exceptions (25). These studies also indicate that significant increases in HDL-C are apparent as soon as 2-6 wk after quitting smoking (22), and some studies have demonstrated durable increases up to 6 yr postcessation (9), although most of the benefit seems to accrue within the first 2 yr after quitting smoking (18).

There is some controversy concerning possible mechanism(s) that may account for elevated HDL-C in smokers compared with nonsmokers and the decrease observed upon cessation. In one of the first prospective investigations, HDL-C increases after 6 wk of abstinence were positively correlated with increases in saturated fat consumption during smoking cessation and increased lipoprotein lipase activity (22). In a follow-up study, these investigators controlled diet, including fat intake, 2 wk before and after smoking cessation, and failed to observe an increase in HDL-C (15). By contrast, many other studies assessed and controlled for changes in diet (e.g., fat and alcohol consumption) and body weight after quitting smoking and found that HDL-C levels increased nonetheless (9,14,16-18,23), so it remains unclear whether changes in diet and body weight after quitting smoking account for changes in HDL-C. In addition, it is unclear whether there exist gender differences in HDL-C response to smoking cessation. Most studies conducted separately in men or women have shown similar increases; however, two studies demonstrated an increase only among male smokers (12,16), whereas another study showed an increase only among women smokers(14). Therefore, effects of smoking cessation on HDL-C should be examined separately for men and women.

Increased physical activity also increases HDL-C and decreases triglycerides (26,30), mediated in part through fitness effects stimulating lipoprotein lipase activity and changes in body weight (27). Cross-sectional studies have demonstrated that smoking attenuates the salubrious effects of exercise on HDL-C(20,21). No study to our knowledge has prospectively investigated the combined effects of smoking cessation and exercise training on serum lipids. Consequently, we examined the short-term (12 wk) effects of smoking cessation with or without exercise training on serum lipids. We hypothesized that smoking cessation and short-term exercise training would have additive effects on concentrations of HDL-C. We additionally assessed diet and body weight changes as a function of smoking cessation to examine whether these changes could account for effects of quitting smoking, and possible additional effects of exercise training, on HDL-C.


Subjects. Twenty-eight healthy women smokers who had smoked for at least 3 yr were recruited via newspaper advertisements. Eighteen of these women had successfully quit smoking by the end of treatment (nine in each of the exercise and control groups), as verified by alveolar expired carbon monoxide concentrations <8 ppm. We therefore analyzed data from this sample of 18 women to examine exercise effects among subjects who had successfully quit smoking. To be eligible, subjects were required to exercise once per week or less; to be free of coronary heart disease (CHD), substance abuse, and orthopedic problems that could limit exercise training; and to not use medications that could affect serum lipids, including oral contraceptives or exogenous estrogens. Mean age was 38 (± SD 7) yr, and mean smoking rate was 19 (±8) cigarettes per day. Body mass index (BMI: kg−1·height m2) averaged 26.8 (±8). Exercise training and control groups did not differ on these characteristics(Table 1).

Mean (±SD) values and difference scores (95% CI) for BMI, CO, V˙O2 peak, and dietary variables over time according to treatment group.

Procedure. After obtaining informed consent according to ACSM guidelines, subjects underwent maximal exercise testing on a bicycle ergometer before treatment. A bicycle ergometer was chosen for two reasons: (1) as many women gain weight when they stop smoking, we wanted to use equipment that was non-weight bearing; and (2) our fitness training facility contained mostly bicycle ergometers; therefore we wanted to be sure to test our subjects on the same type of equipment that they would use in their training. This test was used to screen participants, to determine fitness level at baseline, and to calculate the exercise prescription. Subjects were instructed not to smoke or eat for 1 h and to avoid caffeine for 4 h and alcoholic beverages for 12 h before exercise testing. All exercise tests were performed on a Monark 818E cycle ergometer. A Marquette case 12 exercise electrocardiographic recorder was used for continuous ECG monitoring. The testing protocol consisted of 2-min stages beginning with a zero load warm-up and followed by 25-W increases until the subject was unable to maintain 50 rpm. Subjects were given a 2-min active recovery pedaling at zero load, followed by sitting at rest until heart rate (HR) was within 15% of baseline or until 10 min had elapsed. Measurement of functional capacity expressed as estimated peak V˙O2 was used to determine if participation in the exercise condition resulted in a training effect. After testing, subjects were randomly assigned to a smoking cessation program plus exercise training or to smoking cessation plus contact control. Exercise testing was repeated at the end of treatment.

The smoking cessation program for both groups consisted of twelve 1-h behavioral modification sessions held over 12 wk (11). Sessions for the two groups were presented separately, and the two groups had no interaction during the course of the study. Exercise training consisted of three supervised sessions per week for 12 wk. Subjects in the Exercise condition began the 12-wk exercise program 3 wk before the quit date of the smoking cessation treatment. This permitted subjects to gradually adjust to the physical demands of the exercise program before significantly changing their smoking behavior. The exercise prescription was calculated using the peak HR achieved on the baseline exercise test. The HR reserve method was used to determine a HR range of 60-85%. This method was chosen because it corresponds to a training intensity of approximately 60-85% of functional capacity, which is considered vigorous exercise by the American College of Sports Medicine. Subjects were instructed to work at a perceived exertion range of somewhat hard to hard, which also corresponds to a training intensity of 60-85%. The exercise intensity and duration were gradually increased to allow subjects an opportunity to adapt to the exercise and to decrease the chance of injury. The exercise session consisted of a 5-min warm-up, 30-40 min of aerobic activity, and a 5-min cool down with stretching. A variety of exercise modes were used including treadmills, rowers, stationary bikes, a recumbent step machine, and a cross country ski machine. To ensure training specificity, subjects were required to spend 50% of their aerobic time on a stationary bike. The sessions were supervised by an exercise specialist who verified and documented the HRs and perceived exertion levels obtained during each exercise session. Subjects were required to attend three exercise sessions per week. Subjects missing an exercise session were telephoned by the exercise specialist, encouraged to continue, and scheduled for a replacement session.

Contact control consisted of three health education lectures/discussions per week for 12 wk, with each session lasting 45 min. At each session, subjects provided expired alveolar breath samples for analysis of carbon monoxide concentrations (Vitalograph, Lexington, KY) to assess their smoking intensity. Subjects were instructed to quit smoking at 4 wk after the initial behavioral treatment session.

Dietary variables were assessed before and after treatment using a food-frequency questionnaire (3). Food-frequency data were analyzed using software developed by the National Cancer Institute (Dietsys v. 3.0). The analysis estimates daily intake of calories and the percent of calories consumed as fat, carbohydrates, protein, and alcohol.

Serum samples drawn after an overnight fast were obtained during the weeks before and following treatment. All samples were assayed at the same time to minimize inter- and intra-assay variability. Total cholesterol and triglycerides were assayed with enzymatic methods (1,4). HDL-C was estimated after precipitation of lower density lipoproteins (29). The major subclasses of HDL-C, HDL-C2, and HDL-C3 were determined using the Gidez technique (10). LDL cholesterol was estimated according to the Friedewald equation (8). Total plasma protein concentration was used to estimate changes in plasma volume (Biuret method on a temperature-controlled Gilford Impact 400 autoanalyzer). Between assay coefficients of variation were 0.88% for total cholesterol, 2.83% for triglycerides, 2.3% for HDL-C, 7.7% for HDL-C3, and 7.0% for HDL-C2.

Statistical analyses. Values obtained at posttreatment were contrasted with baseline scores via paired t-tests within each of the exercise and contact control groups. Difference scores were computed by subtracting baseline from post-treatment values, and 95% confidence intervals(CI) were calculated for the difference scores. To control type I errors, p values of 0.01 or less were considered significant. Tables identify p values between 0.05 and 0.01 for the benefit of the reader. Spearman rank order correlations were computed between changes in smoking rate, cardiac output (CO), BMI, total protein, V˙O2 peak, dietary variables, and lipid values. Because of the unique and preliminary nature of the study, we were not able to compute power a priori to detect possible differences in HDL-C as a function of exercise training during smoking cessation.


Attendance and training. Contact control subjects attended 85% of the smoking cessation sessions and 92% of the contact sessions. The nine exercise subjects attended 85% of the smoking cessation sessions and 88% of the training sessions and exercised at an average training HR of 73% of maximum.

BMI and smoking. There was a slight but nonsignificant increase in BMI within groups across the two assessment intervals (Table 1). Carbon monoxide concentrations fell significantly and similarly within groups from pre- to-posttreatment (Table 1), reflecting abstinence. At posttreatment, 8 of 9 subjects in the exercise groups and 6 of 9 subjects in the contact control group reported abstinence of at least 7 d.

Fitness and dietary variables. V˙O2 peak increased significantly in the exercise as compared with the contact control group, demonstrating that the exercise program was sufficiently rigorous to improve fitness (Table 1). No changes in dietary variables were observed within the exercise group. However, among subjects in the contact control group, total caloric intake increased significantly after smoking cessation (Table 1) without a change in caloric distribution.

Lipids, lipoproteins, and plasma volume. HDL-C2 increased in the exercise group, whereas changes in total HDL and total protein only approached significance (Table 2). In the contact controls, increases were noted in total HDL and subfractions, but only for HDL-C3 did the change approach significance. There were no significant changes in total cholesterol, LDL-C, and triglycerides in either group. Significant effects noted above were not altered when total protein was included as a covariate in separate analyses.

Mean (±SD) values and difference scores (95% CI) for lipids, lipoproteins, and total protein over time according to treatment group.

Groups did not differ significantly at baseline for all variables shown in Tables 1 and 2.

Correlational analyses. Across all subjects, Spearman rank order correlation coefficients were computed between changes in CO levels, V˙O2 peak, and lipids and lipoproteins. The correlation between change in CO and change in HDL-C was in the expected direction but did not achieve statistical significance (r (16) = −0.20, 1 tail, P= 0.44). Change in V˙O2 peak correlated significantly with change in triglycerides(r (14) = −0.59, P = 0.017) and almost significantly with change in HDL-C (r (14) = 0.48, P = 0.06).


Among women in the exercise treatment group, concentrations of HDL-C increased by 9.6 mg·dL−1 after 12 wk of smoking cessation treatment and 12 wk of exercise. By contrast, HDL-C increased 5 mg·dL−1 in the contact control group, despite similar reductions in expired alveolar carbon monoxide concentrations in both groups. The increase in HDL-C among the women who exercised was due primarily to a statistically significant increase in HDL-C2. The change in HDL-C in the contact control group (mean increase 5 mg·dL−1) was not statistically significant but was of similar magnitude to that observed in previous smoking cessation studies with women (13,19,23). The results indicate that the addition of exercise to smoking cessation has an additive effect on increasing HDL-C2. Some studies demonstrate that HDL-C2 may confer greater antiatherogenic benefit (24); the combination of smoking cessation and exercise may therefore be more beneficial to patients than smoking cessation alone.

The increase in HDL-C2 observed in the exercise group was accompanied by a decrease(albeit nonsignificant) in triglycerides. Moreover, across all subjects change in V˙O2 peak correlated directly with change in HDL-C and inversely with change in triglycerides after smoking cessation. This indicates that smoking cessation and/or exercise may influence HDL-C indirectly through its influence on triglyceride metabolism. Studies have shown that smoking increases plasma triglycerides (5). Other studies indicate that lipoprotein lipase (LPL) activity is suppressed in smokers(7), which, in turn, could contribute to elevations in triglyceride concentrations and decreases in HDL-C. Smoking cessation could increase LPL activity as does physical activity (27). Thus, the changes observed in the HDLs in the present study may be linked to triglyceride metabolism. Alternatively, smoking cessation may influence HDL-C metabolism through a variety of mechanisms(7), including increased production of apo AI and AII (12).

Changes in lipids and lipoproteins were not accounted for by changes in BMI, dietary, or hemoconcentration variables (17). Previous studies in men have indicated that postcessation increases in HDL-C were associated with increased dietary fat intake (15,19,22). However, our results are consistent with other studies in women showing no association (13,14). Changes in BMI and alcohol intake in the present study were small, were equivalent between groups, and were uncorrelated with changes in serum lipids. Thus, these variables do not explain the change in serum lipids. Changes in expired carbon monoxide concentration were insignificantly associated with changes in HDL-C overall. This is inconsistent with other research showing positive associations between other indicators of cigarette smoke intake, e.g., thyocyanate, and serum lipids (9,12). However, small sample size and limited variability in CO levels after quitting may have limited our ability to detect an association.

It is unlikely that the changes observed in HDL-C and HDL-C2 in the exercise group are solely the result of exercise training. A meta-analysis of 66 training studies showed that, among women, the increase in HDL-C after training is small and not significant (28). It is possible, however, that exercise training augmented any change in HDL-C2 produced by smoking cessation. Further controlled studies, however, which vary presence and absence of exercise crossed with smoking and nonsmoking status, are needed to completely evaluate the combined effects of both exercise and smoking cessation on serum lipids.

Several limitations should be mentioned. The relatively small sample size may have limited ability to detect significant effects, especially in terms of mediating mechanisms, because of low statistical power. Moreover, this study was conducted only in women. Additional comparative studies are needed to clarify whether similar results would be observed in men. However, we believe the significant change in HDL-C2 among the smokers who exercised, above and beyond the change observed among the smokers who did not exercise, warrants further investigation.


1. Allain, C. C., L. S. Poon, C. S. G. Chan, W. Richmond, and P. C. Fu. Enzymatic determination of total cholesterol. Clin. Chem. 20:470-475, 1974.
2.Allen, S. S., D. Hatsukami, J. Gorsline, and The Transdermal Nicotine Study Group. Cholesterol changes in smoking cessation using the transdermal nicotine system.Prev. Med. 23:190-196, 1994.
3. Block, G., A. M. Hartman, C. M. Dresser, M. D. Carrol, J. Gannon, and L. Gardner. A data-based approach to diet questionnaire design and testing. Am. J. Epidemiol. 124:453-469, 1986.
4. Buccolo, G., and H. David. Quantitative determination of serum triglycerides by the use of enzymes. Clin. Chem. 19:476-482, 1973.
5. Craig, W. Y., G. E. Palomaki, and J. E. Haddow. Cigarette smoking and serum lipid and lipoprotein concentrations: an analysis of published data. Br. Med. J. 298:784-788, 1989.
6. Dwyer, J. H., G. E. Rieger-Ndakorerwa, N. K. Semmer, et al. Low-level cigarette smoking and longitudinal change in serum cholesterol among adolescents: the Berlin-Bremen study. J. Am. Med. Assoc. 259:2857-2862, 1988.
7. Freeman, D. J., and C. J. Packard. Smoking and plasma lipoprotein metabolism. Clin. Sci. 89:333-342, 1995.
8. Friedewald, W. T., R. I. Levy, and D. S. Fredrickson. Estimation of the concentration of LDL cholesterol in plasma without use of the preparative ultracentrifuge. Clin. Chem. 18:499-502, 1972.
9. Gerace, T. A., J. Hollis, J. K. Ockene, and K. Svendsen. Smoking cessation and change in diastolic blood pressure(BP), body weight, and plasma lipids. Prev. Med. 20:602-620, 1991.
10. Gidez, L. I., G. J. Miller, M. Burstein, S. Slagle, and H. A. Elder. Separation and quantitation of subclasses of human plasma HDL by a simple precipitation procedure. J. Lipid Res. 23:1206-1223, 1982.
11. Marcus, B. H., A. E. Albrecht, R. S. Niaura, et al. Exercise enhances the maintenance of smoking cessation in women. Addict. Behav. 20:87-92, 1995.
12. Masarei, J. R. L., I. B. Puddey, R. Vandongen, L. J. Beilin, and W. J. Lynch. Effect of smoking cessation on serum apolipoprotein A-I and A-II concentrations. Pathology 23:98-102, 1991.
13. Moffatt, R. J. Effects of cessation of smoking on serum lipids and HDL cholesterol. Arteriosclerosis 74:85-89, 1988.
14.Moffatt, R. J., A. Stamford, S. G. Owens, and L. F. Chitwood. Cessation from cigarette smoking: lipoprotein changes in men and women. J. Smoking Relat. Dis. 3:11-19, 1992.
15. Quensel, M., A. Soderstrom, C. D. Agardh, and P. Nilsson-Ehle. HDL concentrations after cessation of smoking.Atherosclerosis 75:189-193, 1989.
16. Rabkin, S. W. Effect of cigarette smoking cessation on risk factors for coronary atherosclerosis: A control clinical trial. Atherosclerosis 53:173-184, 1984.
17. Richard, F., N. Mareacaux, J. Dallongeville, et al. Effect of smoking cessation on lipoprotein A-I and lipoprotein A-I:A-II levels. Metabolism 46:711-715, 1997.
18. Shennan, N. M., M. Seed, and V. Wynn. Variation in serum lipid and lipoprotein levels associated with changes in smoking behaviour in non-obese males. Atherosclerosis 58:17-25, 1985.
19. Stamford, B. A., S. Matter, R. D. Fell, and P. Papanek. Effects of smoking cessation on weight gain, metabolic rate, caloric consumption, and blood lipids. Am. J. Clin. Nutr. 43:486-494, 1986.
20.Stamford, B. A., S. Matter, R. D. Fell, S. Sady, M. K. Cresanta, and P. Papanek. Cigarette smoking, physical activity, and alcohol consumption: relationship to blood lipids and lipoproteins in premenopausal females. Metabolism 33:585-590, 1984.
21. Stamford, B. A., S. Matter, R. D. Fell, S. Sady, P. Papanek, and M. Cresanta. Cigarette smoking, exercise and HDL cholesterol.Atherosclerosis 52:73-83, 1984.
22. Stubbe, I., J. Eskilsson, and P. Nilsson-Ehle. High-density lipoprotein concentrations increase after stopping smoking. Br. Med. J. 284:1511-1513, 1982.
23. Swank, A., K. Pickering, and R. D. Fell. Short-term changes in high-density lipoprotein cholesterol consequent to a seven-week smoking cessation program in women. J. Cardiopulmonary Rehabil. 11:353-359, 1991.
24. Tall, A. R. Plasma high density lipoproteins: metabolism and relationship to atherogenesis. J. Clin. Invest. 86:379-384, 1990.
25. Terres, W., P. Becker, and A. Rosenberg. Changes in cardiovascular risk profile during the cessation of smoking. Am. J. Med. 97:242-249, 1994.
26. Thompson P. D., E. M. Cullinane, S. P. Sady, et al. Modest changes in high-density lipoprotein concentration and metabolism with prolonged exercise training. Circulation 78:25-34, 1988.
27. Thompson, P. D., E. M. Cullinane, S. P. Sady, M. Flynn, C. Chevert, and P. N. Herbert. HDL metabolism in endurance athletes and sedentary men. Circulation 84:140-152, 1991.
28. Tran, Z. V., A. Weltman, G. V. Glass, and D. P. Mood. The effects of exercise on blood lipids and lipoproteins: a meta-analysis of studies. Med. Sci. Sports Exerc. 15:393-402, 1983.
29. Warnick, G. R., and J. J. Alberts. A comprehensive evaluation of the heparin-manganese precipitation procedure for estimating HDL cholesterol.J. Lipid Res. 19:65, 1978.
30. Wood, P. D., and M. L. Stefanik. Exercise, fitness and atherosclerosis. In: Exercise, Fitness, and Health: A Consensus of Current Knowledge, C. Bouchard, R. J. Shepard, T. Stephens, et al. (Eds.). Champaign, IL: Human Kinetics, 1990, pp. 499-523.


©1998The American College of Sports Medicine