Before menopause, women are at a decreased risk for cardiovascular disease (CVD) compared to men, with men having between 3.5 and 4.5 times the risk of women. Within 10 yr following menopause, the risk of CVD increases in women to a level similar to that seen in men (3,15).
Following menopause, significant changes occur in plasma lipids (14,20,28) and fibrinogen (FIB) (13,16,21,25). Hormone replacement, whether unopposed estrogen or a combination of estrogen and progestin, can have significant but mixed effects on CVD risk factors. Estrogen replacement in postmenopausal women is associated with a decrease in total cholesterol (TC) and an increase in HDL cholesterol (HDL) and triglycerides (TG). Depending on the dosage, the addition of progestin has been postulated to decrease both HDL and TG (19). The use of unopposed estrogen, or a combination of estrogen and progestin, also decreases the level of FIB (16,22).
Regular aerobic exercise and improvements in cardiorespiratory fitness increase HDL and decrease TG while having no significant effect on TC (11). Increased aerobic exercise and improvements in cardiorespiratory fitness are also associated with decreases in FIB (12). Most research, however, on the association between exercise or fitness and these CVD risk factors has been done on men, and it is unclear if the benefits of improved cardiorespiratory fitness extend to women.
Both cardiorespiratory fitness and hormone replacement are thought to have a significant effect on CVD risk factors in postmenopausal women. To date, no research has simultaneously examined the relation of cardiorespiratory fitness to the CVD risk factors of TC, HDL, total cholesterol/HDL cholesterol ratio (TC/HDL), TG, and FIB, while controlling for hormone replacement, differentiating between unopposed estrogen and a combination of estrogen and progestin.
The purpose of this study was to determine if higher levels of cardiorespiratory fitness are associated with beneficial levels of TC, HDL, TC/HDL, TG, and FIB in a group of nonsmoking, postmenopausal women. Second, we examined the relation between fitness and these CVD risk factors after controlling for hormone status and other possible confounding factors, including age, year of testing, blood glucose, and body mass index (BMI).
Study participants. The study participants were postmenopausal women who completed a preventive medical physical examination at the Cooper Clinic in Dallas, Texas between 1987 and 1995. Records of participants were included in our study if they reached at least 85% of their age-predicted maximum heart rate on their treadmill test and were not limited by factors that could not be explained by their cardiorespiratory fitness, such as an orthopedic injury. Participants were excluded if they were current smokers or had quit smoking within the previous 2 yr. In addition, all participants were free of known heart disease or cancer. Following the above exclusion criteria, 283 women were available for data analysis.
Data collection. All data were abstracted from the original Aerobics Center Longitudinal Study (ACLS) data files. Most of the data are in a computer data base in which patient information and testing results are entered following their physical examination. Information not included in the computer data base (hormone status, medication use, and surgical or natural menopause) was abstracted from patient charts.
The clinical examination was performed after participants gave their written informed consent. The examination followed an overnight fast of at least 12 h and included a personal and family health history, a physical examination, a questionnaire on demographic characteristics and health habits, anthropemetry, resting ECG, blood chemistry analyses, blood pressure measurement, and a maximal exercise test using a standard protocol (2). Height and weight were measured on a standard physician's scale and stadiometer. BMI was calculated as the weight in kilograms divided by the square of height in meters. Blood pressures were measured by ausculatory methods with a mercury sphygmomanometer. Cardiorespiratory fitness was determined by total treadmill time to exhaustion. The protocol used in collecting these data has been shown to have a correlation with measured V˙O2max of 0.9 in women (23). Serum blood samples were analyzed by automated techniques in a laboratory participating in the Centers for Disease Control and Prevention Lipid Standardization Program.
Blood levels of fibrinogen were not measured on all subjects at the beginning of the data collection period; therefore, the number of participants who had measures of fibrinogen was fewer than for the other outcome variables (N = 159 of 283, or 56%).
Data analysis. All data were entered into SPSS 7.0 (27). Double entry was performed on 10% of the sample to check accuracy, and no discrepancies were found. Descriptive statistics were run on all variables to check for outliers. There was one outlier for the TC/HDL (21.5) and three outliers for FIB (465, 439, and 435 mg·dL−1); these subjects were removed from all relevant analyses because inclusion of these values skewed the data, violating the normal curve assumption necessary for linear regression.
Cardiorespiratory fitness, as determined by total treadmill time to exhaustion, was used as the main predictor variable for each of the CVD risk factors. Plasma levels of TC, HDL, TC/HDL, TG, and FIB were the CVD outcome variables measured. Participants were divided into quintiles of cardiorespiratory fitness based on their total treadmill time. Chi square analyses and t-tests were used to determine any differences between the upper and lower fitness groups. Spearman correlations were computed to determine if a linear relation existed between fitness and each of the five outcome variables. Linear regression was then used to determine the amount of variance in the CVD outcome variables that could be accounted for by cardiorespiratory fitness. Use of hormones, along with age, year of testing, blood glucose, and BMI were used as control variables to determine if an independent relationship existed between fitness and the CVD outcome variables. Other variables were not controlled for statistically owing to their lack of contribution to outcome measures or the lack of significant difference found between the upper and lower fitness levels (Table 1). Significance for all tests was set at a P value of ≤0.05 and adjusted for multiple comparisons using the Bonferoni technique.
Study sample characteristics are shown for all subjects in Table 1. Participants were divided into age-adjusted quintiles based on the entire ACLS population (5-7). The subset of women in our study was more fit than the ACLS population as a whole. Instead of the expected 20% in each quintile, 131 (46%) of the sample fell into the highest fitness quintile. The main reason for this higher level of fitness is likely owing to the use of only healthy nonsmokers in the present research. Previous research has divided the participant treadmill times into tertiles, comparing the lower 20%, the middle 40%, and the upper 40% of fitness. Almost 50% of the subjects in this study were in the highest fitness quintile, and so this group of highly fit women were compared with the remainder of the subjects in our study (Table 1).
t-tests on the means of continuous variables showed significant differences between the upper and lower fitness levels for each of the outcome variables, along with significant differences in age, total treadmill time, height, weight, BMI, and percent fat based on hydrostatic and skinfold measures (P < 0.01). No significant differences were seen for blood glucose, amount of alcohol intake, years of education, or number of pregnancies. According to Chi square analysis, no significant differences existed between fitness levels on any of the categorical variables of medication use, history of diabetes, hormone replacement, or history of a hysterectomy. These variables were therefore not included in further analyses.
To examine the effect of hormone replacement on the five CVD outcome variables, an ANOVA was performed for the three HRT groups (Table 2). The subjects taking no hormone replacement had a significantly lower HDL and a higher TC/HDL than those taking either unopposed estrogen or a combination of estrogen and progestin (P ≤ 0.001). There were no other significant differences. To assess if there was an interaction between hormone replacement, cardiorespiratory fitness, and the CVD risk factors, a two way ANOVA was performed. There was a significant interaction between hormone replacement and the upper and lower fitness levels for TC/HDL (F = 4.05, df = 5,281, x¯ = 4.0, P ≤ 0.05) (Fig. 1). Higher fitness was associated with significantly higher TC/HDL in those who did not take hormone replacement (4.3 vs 3.2) and those who took unopposed estrogen (3.5 vs 3.0), but not in the group that took estrogen and progestin (3.3 vs 3.0). No other two-way interactions were found, although a trend did occur with FIB (P = 0.10).
Spearman correlations showed that total treadmill time was related linearly to each of the CVD outcome variables (Table 3). Regression analyses were then performed to determine the degree of variance in the outcome measures that could be accounted for by cardiorespiratory fitness. Cardiorespiratory fitness significantly predicted all CVD outcome variables (Table 4).
As seen in Table 4, cardiorespiratory fitness was significantly related to each of the outcome variables, independent of the effects of hormone replacement, age, year of testing, and blood glucose (P < 0.05). Although blood glucose was not significantly different between the upper and lower fitness groups (Table 1), blood glucose was controlled for owing to the association between blood glucose and diabetes, which effects blood lipid levels. Because of the strong association between cardiorespiratory fitness and BMI (r2 = 0.19), regression analysis was performed on each outcome variable with and without BMI as a control. After controlling for BMI, in addition to hormone replacement, age, year of testing, and blood glucose, cardiorespiratory fitness continues to exert an independent effect on the TC/HDL and TG (Table 4).
To date, only two studies have examined the combined effect of exercise and hormone replacement on blood lipids in postmenopausal women (4,17) (Tables 5 and 6), whereas no research has examined the combined effect on FIB. Although both previous studies examined exercise and hormone replacement, the amount of exercise and the type of hormone replacement varied. Our study is cross-sectional, whereas the previous two studies were intervention studies. However, our study is the first to examine the effect of cardiorespiratory fitness on blood lipid and fibrinogen levels while controlling for both unopposed estrogen or a combination of estrogen and progestin.
It is generally accepted that TC will not be reduced with an increase in exercise or with higher levels of fitness (11). In addition, research on postmenopausal women has failed to show this relation (8,26,29), although two studies have shown a significant decrease in total cholesterol with exercise training (4,17). In our study, higher cardiorespiratory fitness was significantly associated with a lower TC (P < 0.001, Table 4). This decrease in TC continued to be significant after controlling for hormone replacement, age, year of testing, and blood glucose level (P < 0.005, Table 4). Further study of cardiorespiratory fitness and total cholesterol, with diet as a covariate, would help determine if part of the difference is due to a better diet in more fit individuals.
HDL is expected to increase with increasing amounts of exercise and increased cardiorespiratory fitness (11). Crosssectional studies of postmenopausal women tend to support this association between exercise and HDL (8,9). Follow-up studies of exercise and HDL have not shown the same positive association (4,8,17). The use of an objective measure, such as cardiorespiratory fitness, probably led to the significant association found in our analyses (Table 4). A previous study of pre- and postmenopausal women did not show a significant improvement in HDL in those who started a jogging program compared with a control group; however, changes in endurance capacity were significantly associated with HDL (30). This discrepancy between the effects of cardiorespiratory fitness and exercise on HDL could be due to the amount and intensity of exercise performed. A study of older men and women found that exercise was only associated with an increase in HDL if the exercise was performed at greater than 80% of maximum heart rate (26).
Both TC and HDL are risk factors for CVD, and therefore the ratio of TC/HDL has been regarded as a significant indicator of CVD risk (18). Because cardiorespiratory fitness was significantly related to both TC and HDL, it was not surprising that the TC/HDL was therefore significantly related to cardiorespiratory fitness in our study (P < 0.001, Table 4). Cardiorespiratory fitness accounted for 12% of the variance in the TC/HDL (Table 5).
The association between TG and CVD has been debated; however, the evidence seems to point toward a positive association, especially in women (1). Previous research on the association between exercise and TG in postmenopausal women has shown mixed results, with some studies showing decreased TG with higher amounts of physical activity (17,29), whereas others have not demonstrated this relationship (4,8). Our study, however, indicates that higher levels of cardiorespiratory fitness are independently related to lower TG above the effect of age, year of testing, hormone status, and blood glucose level (Table 4).
Exercise and increased cardiorespiratory fitness is thought to decrease FIB (12). Prior research specific to postmenopausal women has yielded mixed results with both a negative association (24) and no association (29). Our results support the hypothesis that increased cardiorespiratory fitness is associated with lower FIB (P < 0.005, Table 5), with this association holding after controlling for hormone use, age, year of testing, and blood glucose (P < 0.05, Table 5).
The significant association found in our research between cardiorespiratory fitness and each of the measured CVD risk factors is likely due to the use of an objective measure of cardiorespiratory fitness. Other studies have relied on self-reported physical activity, which may fail to accurately reflect overall activity, whereas measured cardiorespiratory fitness accounts for both structured and unstructured exercise.
Hormone replacement is considered to have a strong influence on each of these CVD risk factors (16,19,22). Our research does point to a significantly higher HDL, and therefore a lower TC/HDL, in those who take either unopposed estrogen or a combination of estrogen and progestin compared with those who do not take hormone replacement. TC, TG, and FIB were not related significantly to the use of hormones, although TG tended to increase in those taking unopposed estrogen, whereas the addition of progestin tended to decrease the TG back toward what is seen in those not taking hormone replacement (Table 2). Our results indicate that cardiorespiratory fitness is more important than hormone replacement in respect to blood lipid and FIB levels, in that cardiorespiratory fitness was significantly related to the five CVD risk factors beyond the effect of hormone replacement (Table 5).
Cardiorespiratory fitness continued to be significantly related to each of the CVD risk factors until after controlling for BMI (Table 4). The drop in significance after controlling for BMI is expected, as cardiorespiratory fitness is biologically related to body mass, with a decrease in body mass and fat being one of the mechanisms through which exercise and fitness exerts its effect (10). The fact that the TC/HDL ratio and TG were still significantly related to cardiorespiratory fitness, above and beyond the effects of BMI and the other control variables, indicates an even stronger relation for these two variables.
In summary, higher levels of cardiorespiratory fitness, as determined by maximal treadmill testing, were associated significantly with a lower TC, TC/HDL, TG, and FIB and an increase in HDL in this group of nonsmoking, postmenopausal women on univariate analysis. Cardiorespiratory fitness was still associated significantly with each of the outcome variables after controlling for hormone replacement, age, year of testing, and blood glucose. The significant association with the TC/HDL and TG continued to hold even after controlling for BMI. The current research provides strong evidence linking improved cardiorespiratory fitness with an improved CVD risk profile in nonsmoking, postmenopausal women, regardless of their hormonal status.
1. Austin, M. A., and J. E. Hokanson. Epidemiology of triglycerides, small dense low-density lipoprotein, and lipoprotein (a) as risk factors for coronary heart disease. Med. Clin. North Am.
2. Balke, B., and R. W. Ware. An experimental study of physical fitness in air force personnel. US Armed Forces Med. J.
3. Barrett-Connor, E. Heart disease in women. Fertil. Steril.
4. Binder, E. F., S. J. Birge, and W. M. Kohrt. Effects of endurance exercise and hormone replacement
therapy on serum lipids in older women. J. Am. Geriatr. Soc.
5. Blair, S. N., J. B. Kampert, H. W. Kohl, C. E. Barlow, C. A. Macera, R. S. Paffenbarger, and L. W. Gibbons. Influences of cardiorespiratory fitness and other precursors on cardiovascular disease and all-cause mortality in men and women. JAMA
6. Blair, S. N., H. W. Kohl, C. E. Barlow, R. S. Paffenbarger, L. W. Gibbons, and C. A. Macera. Changes in physical fitness and all-cause mortality: a prospective study of healthy and unhealthy men. JAMA
7. Blair, S. N., H. W. Kohl, R. S. Paffenbarger, D. G. Clark, K. H. Cooper, and L. W. Gibbons. Physical fitness and all-cause mortality. JAMA
8. Cauley, J. A., A. M. Kriska, R. E. Laporte, R. B. Sandler, and G. Pambianco. A 2-year randomized exercise trial in older women: effects on HDL cholesterol. Atherosclerosis
9. Cauley, J. A., R. E. Laporte, R. B. Sandler, T. J. Orchard, C. W. Slemenda, and A. M. Petrini. The relationship of physical activity to high density lipoprotein cholesterol in postmenopausal women. J. Chronic Dis.
10. Dipietro, L. Physical activity, body weight, and adiposity: an epidemiologic perspective. Exerc. Sport Sci. Rev.
11. Durstine, J. L., and W. L. Haskell. Effects of exercise training on plasma lipids and lipoproteins. Exerc. Sport Sci. Rev.
12. Ernst, E. Regular exercise reduces fibrinogen
levels: a review of longitudinal studies. Br. J. Sports Med.
13. Folsom, A. R., K. K. Wu, C. E. Davis, M. G. Conlan, P. D. Sorlie, and M. Szklo. Population correlates of plasma fibrinogen
and factor VII, putative cardiovascular risk factors. Atherosclerosis
14. Jensen, J., L. Nilas, and C. Christiansen. Influence of menopause on serum lipids and lipoproteins. Maturitas
15. Kafonek, S. D. Postmenopausal hormone replacement
therapy and cardiovascular risk reduction: a review. Drugs
16. Lee, A. J., G. D. O. Lowe, W. C. S, Smith, and H. Tunstall-Pedoe. Plasma fibrinogen
in women: relationships with oral contraception, the menopause and hormone replacement
therapy. Br. J. Haematol.
17. Lindheim, S. R., M. Notelovitz, E. B. Feldman, S. Larsen, F. Y. Khan, and R. A. Lobo. The independent effects of exercise and estrogen
on lipids and lipoproteins in postmenopausal women. Obstet. Gynecol.
18. Linn, S., R. Fulwood, M. Carroll, J. G. Brook, C. Johnson, W. D. Kalsbeek, and B. M. Rifkind. Serum total cholesterol/HDL cholesterol ratios in U.S. white and black adults by selected demographic and socioeconomic variables (NHANES II). Am. J. Public Health
19. Lobo, R. A. Clinical Review 27: effects of hormonal replacement on lipids and lipoproteins in postmenopausal women. J. Clin. Endocrinol. Metab.
20. Matthews, K. A., E. N. Meilahn, L. H. Kuller, S. F. Kelsey, A. W. Caggiula, and R. R. Wing. Menopause and coronary heart disease. N. Engl. J. Med.
21. Meade, T. W., A. P. Haines, J. D. Imeson, Y. Stirling, and S. G. Thompson. Menopausal status and haemostatic variables. Lancet
22. Nabulsi, A. A., A. R. Folsom, A. White, W. Patsch, G. Haeiss, K. K. Wu, and M. Szklo. Association of hormone-replacement therapy with various cardiovascular risk factors in postmenopausal women. N. Engl. J. Med.
23. Pollock, M. L., R. L. Bohannon, K. H. Cooper, J. J. Ayres, A. Ward, S. R. White, and A. C. Linnerud. A comparative analysis of four protocols for maximal treadmill stress testing. Am. Heart J.
24. Rankinen, T., R. Rauramaa, S. Vaisanen, I. Penttila, S. Saarikoski, J. Tuomilehto, and A. Nissinen. Inverse relationship between physical activity and plasma fibrinogen
in postmenopausal women. Atherosclerosis
25. Scarabin, P., G. Plu-Bureau, L. Bara, C. Bonithon-Kopp, L. Guize, and M. M. Samama. Haemostatic variables and menopausal status: influence of hormone replacement therapy. Thromb. Haemost.
26. Seals, D. R., J. M. Hagberg, B. F. Hurley, A. A. Ehsani, and J. O. Holloszy. Effects of endurance training on glucose tolerance and plasma lipid levels in older men and women. JAMA
27. SPSS for Windows. Release 7.0 (Dec. 19, 1995). Copyright SPSS Inc., 1989-1995.
28. Stevenson, J. C., D. Crook, and I. F. Godsland. Influence of age and menopause on serum lipids and lipoproteins in healthy women. Atherosclerosis
29. Stevenson, E. T., K. P. Davy, and D. R. Seals. Hemostatic, metabolic, and androgenic risk factors for coronary heart disease in physically active and less active postmenopausal women. Arterioscler. Thromb. Vasc. Biol.
30. Suter, E., and B. Marti. Little effect of long term, self monitored exercise on serum lipid levels in middle aged women. J. Sports Med. Phys. Fitness
Keywords:© Williams & Wilkins 1998. All Rights Reserved.
HORMONE REPLACEMENT; ESTROGEN; PROGESTIN; BLOOD LIPIDS; FIBRINOGEN