Medscape www.medscape.com

To Print: Click your browser's PRINT button.
NOTE: To view the article with Web enhancements, go to:
http://www.medscape.com/viewarticle/450776

 

 

It Is Never Too Late: Change in Physical Activity Fosters Change in Cardiovascular Risk Factors in Middle-Aged Women

Jane F. Owens, DrPH, Karen A. Matthews, PhD, Katri Räikkönen, PhD, Lewis H. Kuller, MD, DrPH

Prev Cardiol 6(1):22-28, 2003. © 2003 Le Jacq Communications, Inc.

Posted 04/01/2003

Abstract and Introduction

Abstract

The purpose of the study was to determine the effect of physical activity, particularly change in physical activity over time, on cardiovascular risk factors in women. The 520 women in this analysis are part of an ongoing epidemiologic investigation of the effects of menopause on risk for cardiovascular disease; the investigation spans almost 20 years. The findings show that on average, physically active women have healthier risk factor profiles over time, and that as women change their activity level, their risk factor profiles change as well. Thus, for middle-aged women going through the menopausal transition, it is never too late to reduce their cardiovascular risk by increasing their activity level.

Introduction

Low levels of physical activity are associated with higher rates of both all-cause[1-4] and cardiovascular mortality.[4-7] Reports generally support similar findings in men[1] and women.[1,8] Physical activity is associated with lower incidence rates of stroke,[9,10] coronary heart disease (CHD),[11-13] diabetes,[14,15] hypertension,[16] osteoporosis,[17] and cancer.[18-21] Whether or not physical activity directly causes lower morbidity and mortality or is mediated by a more salubrious risk factor profile has not been determined, but most studies support a strong relationship between activity and risk factors for these diseases.

Cross-sectional studies consistently show a relationship between physical activity and a healthier cardiovascular risk factor profile. In general, physically active persons have lower systolic blood pressure (SBP), diastolic blood pressure (DBP),[22] triglycerides, and total and low-density lipoprotein cholesterol (LDL-C) levels and higher high-density lipoprotein cholesterol (HDL-C) levels.[23] In addition, physical activity is inversely related to serum insulin levels, suggesting increased insulin sensitivity among the more active.[24] Lower body weight and body mass index (BMI), as well as less body fat and a lower waist-to-hip ratio (WHR), are also characteristic of more active persons.[22,25] In general, similar patterns of risk factor/physical activity associations are found in men and women.

Prospective studies of the relationship between physical activity and risk factor change are much fewer in number. A review of clinical trials of physical activity and blood pressure concluded that increasing physical activity had the capacity to lower blood pressure in both hypertensive and normotensive persons.[26] In observational longitudinal studies, moderate physical activity reduced the risk of developing non-insulin-dependent diabetes mellitus (NIDDM).[14] In a large, population-based, observational study,[25] increasing physical activity over a 7-year period of time was associated with improved lipid and lipoprotein risk factor profiles and BMI in both men and women. In an observational study[27] of middle-aged women, decreasing physical activity over a 3-year period was associated with a greater decline in HDL-C.

Cross-sectional studies of physical activity and risk factors are unable to allow us to comment about cause and effect due to the potential for selection bias. Prospective observational studies traditionally have one measure of physical activity and one or more outcome measures evaluated at some time point removed from the initial measurement of activity. These studies do not take into account the well known variability in physical activity over the time. Thus, they are unable to capture concomitant changes between physical activity and risk factors for cardiovascular disease. The study described here has the advantage of multiple measurements of both physical activity and risk factors across almost two decades. Thus, it allows for the study of both the static relationship of physical activity and risk factors as well as the dynamic relationship between risk factor change and physical activity change in women. We hypothesize that more active women will have healthier risk factor profiles throughout the follow-up period, and that as women change their level of physical activity, there are concomitant changes in their risk factor profiles.

 

Materials and Methods

Sample

Participants in this study were recruited beginning in 1983 for a longitudinal study of biologic and behavioral changes associated with natural menopause. At intake, all women (n=541) were premenopausal and between the ages of 42 and 50. Of the 541 women recruited to participate, 520 had sufficient data points for analysis in this study of physical activity. The women were recruited using driver license lists where home addresses were within an easy commute to the University of Pittsburgh. Eligibility criteria included premenopausal hormone status (menstrual bleeding in the preceding 3 months), being normotensive (DBP <100 mm Hg), no use of medications known to influence biologic risk factors, and no use of psychotropic or hormone medications. In general, the women were well educated, largely employed volunteers who were free of any chronic disease. All women were evaluated at study entry and then approximately 3-4 years later. The women were evaluated when they were postmenopausal (no menstrual bleeding for 12 months) and then approximately every 2 years after the first postmenopausal visit. After approximately 16 years of follow-up, seven women had died and 51 had withdrawn from the study.

Physical Activity Measurement

Physical activity was measured using the Paffenbarger Physical Activity Questionnaire. This assessment device, originally used in a longitudinal study of college alumni,[28] permits a calculation of kilocalories per week in expended energy. The calculation takes into account the energy expended in climbing flights of stairs, walking city blocks, and participating in sports or recreational activities. The questionnaire is now widely used in studies where assessment of physical activity is required.

Other Risk Factor Measurements

Blood pressures were taken in a clinic setting at the University of Pittsburgh. Two readings were taken by registered nurses using a random-zero muddler sphygmomanometer. The average of the two readings was used as the final reading. Weight and height were measured by a nutritionist or other clinic personnel. When participants were wearing normal street clothes, weight was measured to the 1/2 pound on a balance scale, and height was measured to the nearest 1/2 inch. Approximately two thirds of the way through the initial recruitment period, measurements of waist and hip were added to the protocol. These measurements were taken using a standard tape measure by clinic personnel. Waist was measured at the narrowest point mid-torso, and hip at the point of maximal protuberance of the gluteus maximus. Participants were queried about various health habits and personality and psychosocial factors. The number of cigarettes smoked per day was determined by interview or questionnaire at each time point. Alcohol intake was determined by questionnaire and was computed to reflect total grams of alcohol per day from all sources.

All blood samples were drawn when subjects were in a fasting state. Serum determinations of lipid and lipoprotein measurements were done in the Heinz Nutrition Laboratory in the Graduate School of Public Health at the University of Pittsburgh, a Centers for Disease Control standardization laboratory. Total cholesterol was measured using the enzymatic method[29] and HDL-C and HDL3-C by the precipitation method.[30] HDL2-C was calculated by subtracting HDL3-C from HDL-C. Triglycerides were measured enzymatically,[31] and LDL-C estimated by the Friedewald equation.[32] Plasma insulin levels were determined by radioimmunoassay, and glucose was measured with the Abbott glucose UV test (Abbott Laboratories, Abbott Park, IL) (Yellow Springs glucose analyzer [Yellow Springs Instruments, Yellow Springs, OH]).

Risk factor assessments were done on the following schedule: 1) premenopausal visit; 2) approximately 3-4 years later; 3) after amenorrhea for 12 consecutive months; and 4) approximately 2-3 years apart thereafter. Physical activity was assessed at each visit. This scheme allowed for a potential of nine separate assessments of risk factors and physical activity, depending upon how quickly women entered the menopause.

Physical activity data were available on 540 women at the premenopausal evaluation. At the 3-4 year follow-up, and at the following postmenopausal evaluations, physical activity and risk factor data were available on varying numbers of women, depending on the numbers who had been postmenopausal for that length of time. An average participant produced between 4 and 5 (range, 2-8) concurrent, repeated measurements of physical activity and risk factors after an average of 10.5 years (range, 1.0-13.9) in the study. Concurrent data were available on 503 women for blood pressure, 530 women for waist, height, and weight measurements, 499 for lipid and lipoprotein determinations, and 496 and 499 for insulin and glucose levels, respectively. These women formed the sample for the analyses of change in physical activity and in cardiovascular risk factors over time.

Statistical Analyses

To test whether changes in physical activity co-occur with any potential changes in risk factors over time, multilevel random coefficient regression analyses were carried out. Repeated risk factor measurements served as dependent variables (each factor tested in a separate model), and repeated measurements of physical activity served as a time-varying predictor variable for these models. This differentiated between-person (cross-sectional) and within-person (longitudinal) effects in these models, and also took into account the fact that persons have varying numbers of observations available for analysis. The physical activity data were person-centered, i.e., transformed to deviations from personal means, to account for between-person confounding in the within-person associations.

We expected changes in physical activity over time to be associated on a within-subject basis with concurrent changes over time in the risk factors. We also expected that those with higher average levels of physical activity across the study period would exhibit a more salubrious risk factor profile relative to those with a lower average level of physical activity. We used the maximum likelihood estimation method, and set up a variance component covariation matrix and a spatial, residual covariance matrix. In the longitudinal analyses of change, time to each assessment point from the study entry was used as a covariate. In addition, analyses were conducted with weight and alcohol consumption in the model, and among never-smokers only. The same strategy was used in the between-person cross-sectional analyses. Log transformations were conducted where appropriate.

 

Results

Premenopausal Risk Factor Levels

Table I shows the mean levels of risk factors for the women in our analysis at study entry. The number of women varied because of missing data, as changes were made in the protocol over time. SBP and DBP are low due to the exclusionary criteria established for the study. The average kilocalorie expenditure for 1 week was 1423 kcal (SD, 1640 kcal). Forty-seven percent of the women reported recreational sport activity in the preceding week. The majority of the activity was at a moderate level, and the most prevalent activity was walking.

Change in Risk Factors Across Time

Over time, the women, on average, reported an increase in their energy expended. Mean change scores for kilocalories and risk factors at selected time points are shown in Table II. The time points shown are approximately 3 years after study entry, and at assessments made at 1, 5, and 8 years after menopause. Also shown in Table II are the coefficients of variation, which indicate the extent of variability in the measurement across all available data points. The variability in kilocalories expended is evidence of the fact that over time, women experience changes in their physical activity levels. However, the changes in terms of kilocalories are, on average, rather modest, and are consistent with the amount of change in activity for the first time that risk factor change and activity change were evaluated in this group.[27]

Average and Concurrent Associations Between Physical Activity and Risk Factors

In the between-person analyses shown in Table III, women who were, on average, more active across the study period had lower average fasting insulin, triglyceride, blood pressure, waist circumference, and weight levels compared to women who were, on average, less active across the study period. In addition, more active women had higher HDL-C and HDL2-C levels (all, p<0.002). When weight was controlled, the association between higher activity and lower triglycerides became marginally significant (p=0.08). All other associations remained significant (all, p<0.02). Adjustment for alcohol intake or restriction to only nonsmokers at all time points did not yield different results. These results are consistent with the findings at study entry.[23]

In Table IV, the mean level of risk factors measured across all available time points is shown by tertile of kilocalories, similarly measured across all available time points. A test for linear trend showed that SBP (p=0.06), DBP (p=0.01), and fasting insulin (p=0.001) decreased across tertiles of physical activity. Both HDL-C and HDL2-C increased significantly across activity groups (all, p<0.003). In terms of the anthropometric measurements, weight, BMI, and waist circumference showed a significant decrease across the activity groups (all, p<0.001). This serves to further illustrate the trend for a more salubrious risk factor profile across time in more physically active women.

Within-person associations between kilocalories expended weekly and risk factors are also shown in Table III. Results show that as a woman increases her level of physical activity, her levels of fasting insulin and triglycerides, BMI, and weight and waist circumference decrease significantly (all, p<0.03). There was a tendency (p=0.06) for SBP to decrease as activity levels increased, but lipid and lipoprotein levels were unaffected by change in activity. Results were unaffected after statistical controls for alcohol consumption and smoking.

 

Discussion

This study confirms that physically active women have risk factor profiles that are associated with lower cardiovascular morbidity and mortality. Furthermore, it shows that over time, middle-aged women who increase their physical activity levels have concomitant improvements in their risk factor profiles. This gives support to the hypothesis that physical activity is effective in preserving and improving the health and well-being of middle-aged women.

Over multiple cross-sectional measurements across the 16 years of observation in this study, higher activity levels were associated with lower blood pressure, a healthier lipid profile, lower fasting insulin levels, and lower weight, BMI, and waist measurement. Similar relationships were seen in the sample at study entry, when all the women were premenopausal and 47 years of age on average, and over time these relationships remained unchanged.

Furthermore, in the longitudinal analyses, when women increased their activity, their risk factor profiles improved; decreasing activity was associated with deleterious changes in risk factor profiles. Having multiple measurements available for analyses lends confidence to the hypothesis that changing physical activity can lower risk for disease. Kaplan and colleagues[4] employed a similar strategy in analyzing the effect of activity and all-cause and cardiovascular disease mortality. They had three time points available for analyses, and found that the effects of physical activity were much stronger when all time points were used than when one initial assessment of activity was used as a predictor.

In this study, there was a strong relationship between weight and physical activity in both the longitudinal and cross-sectional analyses. Given the known association between weight and blood pressure, and the observed relationship between activity and weight, being active may play an important role in reducing the cardiovascular and cerebrovascular sequelae of elevated blood pressure.

The relationships between lipids and lipoproteins and physical activity seen in other studies were not confirmed in the within-person analyses in this study. However, there was a very interesting finding of a relationship between triglycerides and activity, controlling for age and weight, that has not been noted before. Elevated triglyceride levels have been related to coronary artery disease in some studies,[33] although the evidence is not as strong as the risk associated with total cholesterol, LDL-C, or HDL-C. Nonetheless, in studies where attempts have been made to identify ways to elucidate the risk basis of elevated triglycerides, there is good evidence that there is greater risk for cardiovascular disease[34,35] that is largely independent of age, weight, and alcohol intake.

Greater WHR and higher BMI have been identified as risk factors for heart disease. It has also been shown that waist circumference is a risk factor for cardiovascular disease and for NIDDM.[36] Waist circumference is a measure of central adiposity, and has been shown to be superior to WHR in dual-energy x-ray absorptiometric,[37] magnetic resonance imaging,[38] or computerized tomographic[39] determination of central adiposity. In this study, waist circumference was significantly smaller in women who, over time, increased their physical activity level, as well as in more physically active women in the cross-sectional analyses. The cross-sectional relationship is independent of the women's weight.

Physical activity level and increase in activity level over time are significantly related to fasting insulin levels, even when age, weight, and alcohol intake are controlled for in the analyses. Elevated insulin levels are a marker for decreased insulin sensitivity, a predictor of NIDDM. There is considerable interest in physical activity as a means of preventing NIDDM, and there are supportive studies, both cross-sectional and prospective. Obesity is a major risk factor for NIDDM, so interventions designed to reduce or prevent overweight or obesity have been of major interest in this field. Unfortunately, dietary interventions and exercise programs are fraught with adherence problems that make them of somewhat limited value. Therefore, it is of considerable interest to note, in this study, the strong relationship between a moderately active lifestyle and a lower insulin level in women, which is independent of age and weight.

The sample recruited for this study initially excluded women who had any chronic disease, as indicated by self-report or use of certain medications. Therefore, they represented a healthy segment of the population. Furthermore, they were, in general, well educated and primarily Caucasian. These attributes limit the ability to extrapolate our findings to other less healthy, less educated, and non-Caucasian populations. Studies of physical activity and cardiovascular health should be undertaken with specific intent to target non-Caucasian and high-risk women to determine if activity is similarly beneficial in these groups.

It should also be noted that this study was not designed to explicitly measure the effects of physical activity on the metabolic syndrome. However, many of the risk factors examined in this sample are covariates with change in physical activity: triglycerides, insulin, weight, and waist circumference are a part of a cluster of risk factors that, taken together, are thought to indicate an altered metabolic syndrome that is associated with greater cardiovascular risk.[40] To our knowledge, this is the first time this set of risk factors has been reported to change over time with physical activity in middle-aged women.

Taken together, the findings of this study show that the cardiovascular risk factors of elevated blood pressure, serum triglycerides, and insulin, and central adiposity -- all part of a frequently seen clustering of metabolic abnormalities -- are sensitive to the influence of changing physical activity in middle-aged women. This study also provides evidence that over time, women are able to maintain or even increase their physical activity as they progress through middle age, dispelling the myth that aging is necessarily associated with an overall decline in activity level. The potential for moderate physical activity to have a role in modifying this clustering of metabolic risk factors is exciting and worthy of further attention.

 

Tables

Table I.



Table II.



Table III.



Table IV.

References

  1. Lee IM, Paffenbarger Jr RS, et al. Physical activity, physical fitness and longevity. Aging Clin Exp Res. 1996;9:1-2.
  2. Kushi LH, Fee RM, Folsom AR, et al. Physical activity and mortality in postmenopausal women. JAMA. 1997;277:1287-1292.
  3. Kannel WB, Sorlie P. Some health benefits of physical activity: the Framingham Study. Arch Intern Med. 1979;139:857-861.
  4. Kaplan GA, Strawbridge WJ, Cohen RD, et al. Natural history of leisure-time physical activity and its correlates: associations with mortality from all causes and cardiovascular disease over 28 years. Am J Epidemiol. 1996;144:793-797.
  5. Gartside PS, Wang P, Glueck CJ. Prospective assessment of coronary heart disease risk factors: the NHANES I Epidemiologic Follow-up Study (NHEFS) 16-year follow-up. J Am Coll Nutr. 1998;17:263-269.
  6. Haapanen N, Miilunpalo S, Vuori I, et al. Characteristics of leisure-time physical activity associated with decreased risk of premature all-cause and cardiovascular disease mortality in middle-aged men. Am J Epidemiol. 1996;143:870-880.
  7. Hakim AA, Curb JD, Petrovitch H, et al. Effects of walking on coronary heart disease in elderly men: the Honolulu Heart Program. Circulation. 1999;100:9-13.
  8. Lissner L, Bengtsson C, Bjorkelund C. Physical activity levels and changes in relation to longevity: a prospective study of Swedish women. Am J Epidemiol. 1996;143:54-62.
  9. Gillum RF, Mussouno ME, Ingram DD. Physical activity and stroke incidence in women and men: the NHANES I Epidemiologic Follow-up Study. Am J Epidemiol. 1996;143: 860-869.
  10. Sacco RL, Gan R, Boden-Albala B, et al. Leisure-time physical activity and ischemic stroke risk: the Northern Manhattan Stroke Study. Stroke. 1998;29:380-387.
  11. Lakka TA, VenaLainen JM, Rauramaa R, et al. Relation of leisure-time physical activity and cardiorespiratory fitness to the risk of accute myocardial infarction in men. N Engl J Med. 1994;300:1549-1554.
  12. Manson JE, Hu FB, Rich-Edwards JW, et al. A prospective study of walking as compared with vigorous exercise in the prevention of coronary heart disease in women. N Engl J Med. 1999;341:650-658.
  13. Folsom AR, Arnett DK, Hutchinson RG. Physical activity and incidence of coronary heart disease in middle-aged women and men. Med Sci Sports. 1997;29:901-909.
  14. Lynch J, Helmrich SP, Lakka TA, et al. Moderately intense physical activities and high levels of cardiorespiratory fitness reduce the risk of non-insulin-dependent diabetes mellitus in middle-aged men. Arch Intern Med. 1996;156:1307-1314.
  15. Manson JE, Rimm EB, Stampfer MJ, et al. Physical activity and incidence of non-insulin-dependent diabetes mellitus in women. Lancet. 1991;338:774-778.
  16. Paffenbarger Jr RS, Kampert JB, Lee IM. Physical activity and health of college men: longitudinal observations. Int J Sports Med. 1997;18:S200-S203.
  17. Greendale GA, Barrett-Connor E, Edelstein S, et al. Lifetime leisure exercise and osteoporosis. Am J Epidemiol. 1995;141:951-959.
  18. Rosengren A, Wilhelmsen L. Physical activity protects against coronary death and deaths from all causes in middle-aged men. Evidence from a 20-year follow-up of the primary prevention study in Goteborg. Ann Epidemiol. 1997;7:69-75.
  19. Martinez ME, Giovannucci E, Spiegelman D. Physical activity, body size and colorectal cancer in women. Am J Epidemiol. 1996;143:S73.
  20. Slattery ML, Edwards SL, Boucher KM, et al. Lifestyle and colon cancer: an assessment of factors associated with risk. Am J Epidemiol. 1999;150:869-877.
  21. Colditz GA, Coakley E. Weight, weight gain, activity, and major illnesses: the Nurses' Health Study. Int J Sports Med. 1997;18:S162-S170.
  22. Pols MA, Peters PHM, Twisk JWR, et al. Physical activity and cardiovascular disease risk profile in women. Am J Epidemiol. 1997;146:322-329.
  23. Owens JF, Matthews KA, Wing RR, et al. Physical activity and cardiovascular risk: a cross-sectional study of middle-aged premenopausal women. Prev Med. 1990;19:147-157.
  24. Greendale GA, Bodio-Dunn L, Ingles S, et al. Leisure, home and occupational physical activity and cardiovascular risk factors in postmenopausal women. Arch Intern Med. 1996;156:418-424.
  25. Thune I, Njolstad I, Lochen ML, et al. Physical activity improves the metabolic risk profiles in men and women: the Tromso Study. Arch Intern Med. 1998;158:1633-1640.
  26. Arroll B, Beaglehole R. Does physical activity lower blood pressure: a critical review of the clinical trials. J Clin Epidemiol. 1992;45:439-447.
  27. Owens JF, Matthews KA, Wing RR. Can physical activity mitigate the effects of aging in middle-aged women? Circulation. 1992;85:1265-1270.
  28. Paffenbarger Jr RS, Wing AL, Hyde RT. Physical activity as an index of heart attack risk in college alumni. Am J Epidemiol. 1978;108:161-175.
  29. Allain CC, Poon LS, Chan CSG, et al. Enzymatic determination of total serum cholesterol. Clin Chem. 1974;20: 470-475.
  30. Warnick GR, Albers JJ. A comprehensive evaluation of the heparin-manganese precipitation procedure for estimating high-density lipoprotein cholesterol. J Lipid Res 1978;19: 65-76.
  31. Bucolo G, David H. Quantitative determination of serum triglycerides by use of enzymes. Clin Chem. 1973;19:476-482.
  32. Friedewald WT, Levy RI, Frederickson DS. Estimation of the concentration of LDC-C in plasma without the use of the preparative ultracentrifuge. Clin Chem. 1972;18:499-502.
  33. Castelli WP. Epidemiology of triglycerides: a view from Framingham. Am J Cardiol. 1992;70:3H-9H.
  34. Rainwater DL, Mitchell BD, Comuzzie AG, et al. Associations among 5-year changes in weight, physical activity, and cardiovascular disease risk factors in Mexican Americans. Am J Epidemiol. 2000;152:974-982.
  35. Stampfer MJ, Krauss RM, Ma J, et al. A prospective study of triglyceride level, low-density lipoprotein particle diameter, and risk of myocardial infarction. JAMA. 1996;276:882-888.
  36. Lean MEJ, Han TS, Seidell JC. Impairment of health and quality of life in people with large waist circumference. Lancet. 1998;351:853-856.
  37. Taylor RW, Keil D, Gold EJ, et al. Body mass index, waist girth, and waist-to-hip ratio as indexes of total and regional adiposity in women: evaluation using receiver operating characteristic curves. Am J Clin Nutr. 1998;67:44-49.
  38. Conway JM, Chanetsa FF, Wang P. Intraabdominal adipose tissue and anthropometric surrogates in African American women with upper-and lower-body obesity. Am J Clin Nutr. 1997;66:1345-1351.
  39. Räikkönen K, Matthews KA, Kuller LH. Anthropometric and psychosocial determinants of visceral obesity in healthy postmenopausal women. Int J Obes. 1999;23:775-782.
  40. Sakkinen, PA, Wahl P, Cushman M, et al. Clustering of procoagulation, inflammation, and fibrinolysis variables with metabolic factors in insulin resistance syndrome. Am J Epidemiol. 2000;152:897-907.
Funding Information

This research was supported by the Healthy Women Study (NIH grant HL28266) and the Pittsburgh Mind-Body Center (NIH grants HL65111 and HL65112).

Reprint Address

Address for correspondence: Jane F. Owens, DrPH, Department of Psychiatry, University of Pittsburgh, 3811 O'Hara Street, Pittsburgh, PA 15213



Jane F. Owens, DrPH,1,2 Karen A. Matthews, PhD,1 Katri Räikkönen, PhD,3 Lewis H. Kuller, MD, DrPH2

University of Pittsburgh, Departments of Psychiatry1 and Epidemiology,2 Pittsburgh, PA; and the University of Helsinki and the Finnish Academy, Helsinki, Finland3