Share this article on:

Effects of Exercise in Overweight Japanese with Multiple Cardiovascular Risk Factors


Medicine & Science in Sports & Exercise: June 2007 - Volume 39 - Issue 6 - p 926-933
doi: 10.1249/mss.0b013e3180383d84
BASIC SCIENCES: Epidemiology

Purpose: The effects of exercise in subjects with multiple cardiovascular risk factors (RF) have never been tested in a large-scale randomized controlled trial. The favorable results from our observational study led to this randomized controlled trial.

Methods: The participants, recruited from a community, were overweight, with two of the following three risk factors: hypertension, hyperlipidemia, or glucose intolerance (N = 561; 44% male; mean age (SD) 67 (6); mean BMI 26.4 (2.0)). All participants received a standard health exam with counseling, followed by randomization. The intervention group (INT) was to exercise two to four times per week for 6 months at a fitness club.

Results: Dropout rate was 11% for INT and 10% for the controls. INT exercised an average of 2.6 times per week. Among primary outcome measures, there was a marginally significant between-group difference in changes in systolic blood pressure (intervention minus control, −2.46 mm Hg). Reductions in LDL cholesterol (−1.9 mg·dL−1) and hemoglobin A1c (−0.042%) were not significant. Among secondary outcomes, between-group differences were significantly greater (P < 0.05) for INT in mean body weight (−1.60 kg), waist circumference (−1.8 cm), triglyceride (−7 mg·dL−1), hsCRP (−0.063 mg·L−1), and estimated V˙O2peak (2.0 mL·kg−1·min−1). All directional mean RF changes consistently favored INT. A cardiovascular risk reduction estimated from the changes in RF was about 24%. A greater improvement was also found in health-related quality-of-life measures (SF-36) in INT. No difference was found in the adverse event rate.

Conclusions: Exercising an average of 2.6 times per week for 6 months produced a significant improvement in cardiovascular risk profile in subjects with multiple cardiovascular risk factors through cumulative results of modest yet pervasive changes in all conventional risk factors, without increased adverse effects.

1Sapporo Health Promotion Center, Sapporo, JAPAN; and 2Asai Hokusho University, Ebetsu, JAPAN

Address for correspondence: Hirotaka Nishjjima, MD, PhD, Sapporo Health Promotion Center, Minami-3 Nishi 11, Chuoku, Sapporo 060-0063, Japan; E-mail:

Submitted for publication September 2006.

Accepted for publication January 2007.

Cardiovascular disease (CVD) is a major cause of premature death and morbidity in developed and developing countries, with huge economic costs and negative impact on quality of life. On the other hand, the concept of CVD prevention through multiple risk factors has been established, and its practice has produced concrete results (23). Specific risk factors, singly or in combination, contribute to the development of atherosclerotic CVD. When multiple risk factors are involved, the cumulative burden has been shown to be substantial, even if any individual factor is not severely elevated (23). Such a syndrome of multiple risk factors is a major health problem globally (23).

Exercise seems to be an ideal, natural modality of therapy for this syndrome (25). The effects of exercise, however, have never been tested on this syndrome for its scope and magnitude by large-scale randomized controlled trials. To date, they have been investigated primarily in studies dealing with subjects with single risk factors, and the view of exercise as an effective multifactorial intervention seems to have been assumed from these separate studies (6,13,15,25). Most comprehensive intervention studies have employed various forms of exercise as one of the intervention components, not designed to test exercise per se, also employing medications as well (5).

To date, there have been only a few randomized controlled studies in men and women in which exercise per se was evaluated as a means of modifying multiple cardiovascular risk factors (14,24). King et al. (14) report one such study; their participants, however, were not recruited on the basis of CVD risk factors but, rather, on the basis of sedentary lifestyle, with relatively normal risk profiles. Surprisingly, there were no training-induced changes in CVD risk factors. The HERITAGE Family Study (12), although not a randomized controlled trial, addressed, for the first time, the usefulness of physical activity for a syndrome of multiple risk factors, acknowledging that randomized controlled trials should be undertaken. Stewart et al. (24) studied older men and women with systolic hypertension, 42% of whom satisfied the criteria for metabolic syndrome. Improvements in CVD risk factors were quite modest despite significantly improved peak oxygen uptake.

To the best of our knowledge, this is the first large-scale trial to test exercise on a syndrome of multiple risk factors. This study was derived from our 10 yr of experience in helping our clients train in our public health clinic/fitness club complex in Sapporo, Japan. The experience resulted in an observational study in which we observed significant improvements in risk factor profile for 396 such Japanese men and women who trained 1.8 times a week for a year, although their average risk factor levels were near normal (19). Encouraged by these results, we attempted to give our public exercise program a robust, evidence-based validation. We aimed at recruiting from the community participants with a higher level of risk factors in this randomized controlled trial.

The purpose of the present study was twofold: first, to obtain the unbiased scope and magnitude of exercise effects on a syndrome of multiple CVD risk factors, and second, to determine the extent to which exercise at a fitness club adds benefits to conventional lifestyle counseling to improve CVD risk factors. In Japan, conventional lifestyle counseling after a screening health check has been extensively practiced nationwide as the standard CVD preventive regimen.

Back to Top | Article Outline



The methods used in the present study have been described in detail elsewhere (10). Briefly, the Sapporo Fitness Club Trial (SFCT) was a randomized clinical trial that was conducted from April 2003 to October 2004, with trial periods defined as the first 6 months (from April 2003 to April 2004) and the second 6 months. The randomized design with testing of the null hypothesis was applied only to the first 6-month period and is the subject of the present report. The intervention group exercised at a fitness club, and the control group was asked to adhere to lifestyle recommendations; members of this group did not get memberships to a fitness club during the first 6 months. The second 6 months of the trial period were planned so that the control group could get the same benefits of experiencing exercise at a fitness facility as the intervention group.

The primary outcome measurements were defined as the following three major CVD risk factors: systolic blood pressure (SBP), low-density lipoprotein cholesterol (LDL-C), and hemoglobin A1c (HbA1c). Secondary outcome measurements included hsCRP and other biochemical measurements, waist circumference, estimated peak oxygen uptake (V˙O2peak), and health-related quality of life as assessed by the short-form health survey (SF-36) (7,26).

This trial was approved by the institutional review board of Sapporo Medical University, Sapporo, Japan.

Back to Top | Article Outline


The recruitment of participants was based on a database made available to the Sapporo Health Promotion Center by the National Health Insurance and National Pension Plan Section of the local government. The database contained the medical records from the annual health exams of the preceding year (from January to November 2002) of citizens of Sapporo aged 40 yr or older who subscribed to the National Health Insurance. National Health Insurance coverage is primarily for the self-employed, retired, or their dependents. These features also characterized our previous observational study population.

Inclusion criteria were an age of 40-89 yr and a body mass index (BMI) of 24.2-34.9 kg·m−2, with two or more of the following CVD risk factors: a resting SBP of 130-179 mm Hg, a fasting blood glucose level of 110-139 mg·dL−1 or HbA1c ≥ 5.8 when casual blood sugar was 140-199 mg·dL−1, and an LDL-C level of 120-219 mg·dL−1. The cutoff BMI value of 24.2 was chosen because it represents +10% obesity when a BMI of 22 is taken as normal-the most frequently cited normal value of BMI in Japan (16).

Exclusion criteria were a diastolic blood pressure (DBP) of 110 mm Hg or greater, a history of clinical heart disease or stroke, orthopedic problems that might interfere with exercise, abnormal exercise electrocardiogram (ECG) tests, and/or disqualification by the attending private physician.

The database included 10,421 individuals (18% of the total who had annual health exams in the period described earlier) who met the criteria, and a letter of invitation to participate in the study was sent to all of them. The participation fee was 5000 yen ($48), which included membership to a fitness club for 6 months. The decision to charge participants was made because we felt that this amount would act as a motivator for people to participate and complete the program, rather than discourage participation. The participants had to be able to commute to one of five designated fitness clubs (two public (ours) and three commercial); although this imposed a geographical limitation on prospective participants, we were limited by a fixed amount of funding. A total of 824 prospective applicants responded positively, and a letter of invitation to a group briefing session was then sent.

Figure 1 is a chart showing the flow of the study. Seven briefing sessions were held serially. At the end of each session, written informed consent was obtained, signifying tentative enrollment in the study, pending a screening examination. All screening medical examinations were performed at Sapporo Health Promotion Center, which offered morning or afternoon appointments in groups of four to six. After medical screening, 54 subjects were excluded, with the main reason being manifest heart disease (Fig. 1).



One week after the screening examination, eligible participants were required to visit our institution again in blocks for one-on-one lifestyle counseling sessions with public heath nurses, who informed each subject of the results of the screening tests and offered advice on lifestyle modification, including exercise at home, diet, and smoking cessation. Brochures on simple calisthenics and walking were provided. For participants who were on medication, a letter was written to their private physicians, asking whether they agreed to their patients' participation in our study, although at this point all those who met the study exclusion criteria had been excluded. They were asked to stay on the same medication. Those not on medication were asked to remain so unless they developed a symptom. Inability to enter the study because of disqualification by the subject's own physician (N = 12) was treated as an exclusion rather than a withdrawal.

After lifestyle counseling had been completed for all participants, randomization was carried out by having each participant draw one group or the other in a lottery. Randomization was block stratified with regard to fitness club, age group, and gender (four to six subjects per block).

All participants were instructed to keep a daily record of blood pressure self-monitored at home (if one owned a device), the number of minutes spent walking per day, and sets of calisthenics done at home. Participants submitted their records bimonthly, and in return they received letters of assessment and advice from the public health nurse.

Follow-up data-collection visits took place at 6 months after randomization. The scheduling of each individual's first examination (morning or afternoon) was adhered to in the follow-up examinations. The staff members administering and supervising exercise tests were blind to group assignment. Baseline and follow-up data collection each took about 6 months to complete because of limitations in human and other resources.

Back to Top | Article Outline

Measurement Methods

At baseline (screening) and 6 months, all subjects had outcome measurements (described in the Design section). In addition, at baseline, a chest x-ray film and a resting 12-lead electrocardiogram were taken. Blood pressure was measured in a sitting position using an automated oscillometric device. The average of three sequential measurements was entered into our records. A symptom-limited maximal incremental exercise was performed on an upright bicycle. The initial workload was set at 25 W, and the load was increased by 25 W every 2 min until the subject was unable to pedal above 50 rpm because of exhaustion. The peak V˙O2 was estimated on the basis of exercise time on the bicycle. This procedure had been validated previously at our institution during exercise while employing a breath-by-breath gas analyzer. Isokinetic muscle strength testing of knee extension and flexion was performed with an isokinetic dynamometer (Biodex system 3).

Questionnaires were given, addressing possible adverse effects, including unanticipated visits to the doctor's office for medical or orthopedic conditions, and the appearance of new musculoskeletal pains. Pain was also rated on the SF-36 bodily pain subscale.

Back to Top | Article Outline


All participants in the intervention group were required to attend eight individual exercise sessions with a certified fitness instructor, in addition to exercising a total of two to four times per week on their own. Bicycle exercise at a constant rate was employed as the primary aerobic exercise, and the workload was initially set at 40% of the estimated V˙O2peak from the maximal exercise test at baseline; then, the effort was made to increase the workload in two to three steps to 70% of the maximum with training effects. When increasing the load, the instructor in charge examined the individual's exercise log (which included exercise heart rate records), watched the individual participant exercise, and asked him or her about perceived exertion. A step increment of 5-10 W was usually used. The bicycle exercise duration was initially set at 20 min, and an effort was made to increase it in steps up to 40 min. A heart rate (HR) monitor (Polar Electro Oy, Finland) was worn by all participants during bicycle exercise to record HR and to estimate energy expenditure, as well as for safety reasons.

Various light resistance exercises were also added during later stages of the study, which included knee extension, leg press, leg abduction, and chest press. They were introduced for variety of exercise and also for initiating experience. They were not intended for rigorous training. Two sets of 20 repetitions were employed, the load of which was set so the individual's perceived rate of exertion for the muscle groups used would be mild to moderate. No systematic efforts to progressively increase the load were made. Before and after a session of aerobic and resistance exercises, stretching exercise was employed, totaling about 20 min.

The duration of each exercise session was set at 60-90 min, including warm-up and cool-down exercises; this duration was based on the average time spent exercising by our regular fitness club clients.

The control group was asked to follow lifestyle-modification advice.

Back to Top | Article Outline

Statistical Procedures

The required sample size was estimated to be a total of 300 (80% power with P = 0.01, two tailed). The P value was set at this level to keep it below 0.017 (0.05/3) after Bonferroni correction for the three primary outcome comparisons. The sample size was calculated from the observational study (19) of the clients in our routine health check/exercise program. For the sample size calculation, the data for 84 out of the original 396 subjects (the control and the exercise groups) in our observational study (6) were extracted on the basis of the same entry criteria as that of this trial. The effect size (SD) for SBP, HbA1c, and LDL cholesterol was 8 mm Hg (18.9), 0.15% (0.41), and 14 mg·dL−1 (32.3) or 0.36 mmol (0.84), respectively. Although the effect size was larger than the values reported in other randomized controlled trials (6,13,15), we chose to adopt the above because we had planned to employ approximately the same exercise protocol and mostly the same facilities and staff. However, we aimed to recruit more than 300 participants.

Analyses of secondary outcome measures were considered exploratory (1), and a P value of less than 0.05 was regarded as statistically significant. For the primary comparison of changes in each risk factor between the control and intervention groups, analysis of covariance (ANCOVA) was used, with the baseline value as a covariate. As secondary analyses, unadjusted comparisons and ANCOVA comparisons with age, gender, and club (public/private) as covariates were also performed. Triglyceride and hsCRP (multiplied by 1000) levels were log-transformed before analysis. Extreme hsCRP values identified by Tukey's method (17) were excluded from analysis (> 3.805 mg·L−1). Rates were analyzed by chi-square tests. All analyses were carried out using SPSS (version 11.0). The trial was analyzed by the intention-to-treat method. Values are expressed as mean (SD).

Back to Top | Article Outline


The baseline characteristics for all participants are shown in Table 1 by group. No imbalance between groups was observed. Although half of the participants reported exercising at least once per week, the predominant type of exercise was a light one such as walking.



At 6 months, the dropout rate was 11% for the intervention group and 10% for the control group. The intervention group exercised at fitness clubs an average of 58.3 (34) times (an average of 2.6 times, attendance rate for two or more times per week, 55%). The average time spent on bicycle exercise was 35.3 (11.1) min per session. Average maximal HR was 118 (14) bpm, average HR was 105 (12), and the average energy expenditure estimated from the HR monitor was 99 (52.2) kcal per bicycle session. The participants spent an average of 90 min on the fitness floor, including time for bicycle, light resistance, and stretch exercises, as well as the waiting time for training machines and breaks.

The analyses of the changes comparing the control and the intervention groups are summarized in Table 2. Among the three primary outcomes, the mean between-group difference in changes in SBP (−2.46 mm Hg) was marginally significant (P = 0.018, not reaching 0.017 corrected for three comparisons), indicating a trend towards greater reduction in the intervention group. The differences in the changes in LDL cholesterol (−1.9 mg·dL−1) and HbA1c (−0.042 mg·dL−1) were not statistically significant.



Among secondary outcomes, the between-group differences in changes indicated significant reductions in the intervention group in body weight (−1.60 kg), waist circumference (−1.8 cm), DBP (−1.4 mm Hg), casual triglyceride (median difference, −7 mg·dL−1), hsCRP (median difference, −0.063 mg·L−1), and white blood cell count (−0.23 cells per microliter), and nonsignificant differences in HDL cholesterol (0.43 mg·dL−1) and casual blood glucose (−1.4 mg·dL−1). The results of ANCOVA adjusted for age, sex, and club were not particularly contributory (data not shown).

Among physical function variables, between-group differences in changes indicated a significantly greater increase for the intervention group in estimated V˙O2peak (2.0 mL·kg−1·min−1) and the amount of bicycle exercise time (0.67 min). The significant reduction in resting HR changes (−3.3 bpm) was consistent with these findings.

The data shown in Table 2 indicate that the mean changes in all relevant risk factors favor the intervention group, regardless of P values. Figure 2 graphically summarizes this overall trend.



In the intervention group, an increase in estimated V˙O2peak was statistically significantly (P < 0.05), although only modestly, correlated with changes in multiple risk factors: r = −0.272 with waist circumference, −0.149 with SBP, −0.179 with LDL-C, 0.175 with HDL-C, and −0.161 with HbA1c. With body weight, the correlation was much greater (r = −0.451).

According to the bimonthly self-reports, there were no significant differences in daily minutes of walking between the intervention group and the control group (36 (21) vs 37 (23) min, P = 0.611). The numbers of calisthenics sessions done at home during a period of 6 months were not significantly different between two groups (126 (56) vs 133 (58), P = 0.205).

The mean SF-36 scores improved significantly more (P < 0.05) in the intervention group on three of the eight subscales: general health (intervention vs control: 64.3 to 69.5 vs 65.8 to 65.2), vitality (70.7 to 74.8 vs 73.0 to 70.1), and mental health (79.9 to 82.1 vs 79.6 to 77.6). The bodily pain subscale scores did not change significantly in either group.

The appearance of a new bout of musculoskeletal pain occurred in 20% of intervention participants and in 19% of control participants (P = 0.685). There were 69 (28%) unanticipated visits to the doctor's office in the intervention group and 78 (31%) in the control group (P = 0.425). Of these, 24 in the intervention group and 16 in the control group were orthopedic.

Back to Top | Article Outline


The present study, together with a limited number of similar previous studies (14,24), indicates that exercise-induced immediate changes in CVD risk factors in a syndrome of multiple risk factors are, on average, modest when they are rigorously evaluated by controlled trials. The results of the present randomized trial, however, are in marked contrast to those of our observational study (19); the exercise effect was greatly attenuated in this randomized controlled trial despite the higher level of baseline risk factors the raw data between-group comparison of differences in changes between the observational versus the randomized trial shows the following results: body weight (kg) changes, −1.1 versus −1.6; estimated peak V˙O2 changes, 1.8 versus 2.1; SBP changes (mm Hg), −4.0 versus −2.1; LDL cholesterol changes (mg·dL−1), −4.0 versus −2.1 (−0.1 vs −0.05 mM); and hemoglobin A1c changes (%), −0.14 versus −0.06. There were, however, significant differences (observational vs randomized) in sex (female percentage: 80 vs 56%), mean age (55 vs 67 yr), frequency of club visits (1.8 vs 2.6 visits per week), and duration of intervention (1 yr vs 6 months) between these two studies. Influences attributable to sex and age differences, analyzed using ANCOVA in both studies, did not explain the discrepancy between two studies. The frequency of club visits and the increase in estimated peak V˙O2 were greater in the randomized trial. The 1-yr follow-up of intervention subjects in the current trial who continued to use clubs showed that except for the increase in HDL cholesterol, the changes in risk factors after the first 6 months were marginal to small (not published). We believe that the bias possibly introduced in the observational study by the "healthy lifestyle effect" (27) may have been primarily responsible for the discrepancy between two studies.

The net impact of reductions in independent risk factors on CVD events or mortality is calculated as a multiplicative of each risk factor (11). Past studies with event endpoints indicate that the average extent of a reduction in one of any of the risk factors observed in the present study is small and can vary from a few to, at most, 10% (2,4,9,21,22). We selected five statistically significant (P < 0.05) risk factors that have been reported to be independent predictors from this study, and we attempted to estimate the net impact: waist circumference (2,21), diastolic blood pressure (4), triglyceride (9), hsCRP (22), and V˙O2peak (8,18) (correlation coefficients with each other were all under 0.3 in the present study). The net event reduction was roughly calculated as 0.92 × 0.98 × 0.91 × 0.96 × 0.96 = 0.76, which is not so modest an effect. Also, many cohort studies consistently show that the effect of exercise on disease prevention is far from modest, with a reduction of up to 50% (3,20).

The intervention group lost 1.6 kg more weight, on average, than the control group. The primary objective of this trial was to improve cardiovascular risk factors, not weight reduction per se. Dieticians were not involved. At the time of baseline lifestyle counseling, however, the importance of weight reduction was emphasized. Because the energy expenditure during exercise sessions was insufficient to explain the entirety of the lost weight, even after taking into account resistance training and stretches (at best, 1.5 times the bicycle energy expenditure), we speculate that the participants' voluntary dietary discretion likely contributed to the rest of the lost weight. This may simply reflect the fact that their participation in exercise sessions made them health conscious.

There are some limitations in this study. One is the short duration of the observation period. The observed benefits would not translate into the extrapolated benefits if it were not proven that such exercise could be sustained for a long time. Another limitation may be the relatively small volume and/or modest intensity of exercise employed in this study. Yet, the increase in estimated V˙O2peak in this study, 2.0 mL·kg−1·min−1, was between those reported in King et al.'s (14) and Stewart et al.'s (24) investigations. It was also slightly larger than that of our observational study (19). Furthermore, an increased frequency, duration, or intensity of exercise, particularly in this age group from a community, may jeopardize sustainability because of the increased risk of musculoskeletal problems and dropout. Another limitation was that there were no quantifiable recordings of nutritional intake or habitual activity other than time spent walking per day. The participants in this study were limited to National Health Insurance subscribers, who are generally in the middle- to lower-income bracket; nevertheless, they comprise 42% of people 40 yr or older.

Although exercise or fitness facilities may not be widely available or accessible in some countries, we believe that with economic and political stability, the merits of having such facilities for the general public would be appreciated. Furthermore, the results of the present study would be applicable to any form of aerobic exercise carried out and based on sound physiological principles, because there is general consensus that the overall benefits of exercise are not limited to any particular type of exercise.

To summarize, engaging in exercise at a fitness club an average of 2.6 times a week led to improved overall risk profiles for sedentary, middle- to older-aged overweight subjects with multiple cardiovascular risk factors; it also improved quality of life without increased untoward effects. The effect, if sustained, is estimated to be substantial in preventing future CVD. A greater use of inexpensive public exercise facilities across the nation would provide a safe exercise environment, regardless of seasons, for a larger population of people with cardiovascular risk factors.

The authors would like to thank Dr. Mitsuru Mori (Department of Public Health, Sapporo Medical University) for advice on the planning and implementation of this study. We also thank Yasuhiro Mouri (National Heath Insurance and National Pension Plan Section, Health and Welfare Bureau of Sapporo) for assistance with recruitment and advice on the study protocol. We are grateful to the following private fitness clubs for their collaboration in this project: Konami Sports Club Shinsapporo, Konami Sports Club Higashinaebo, and Sports Club Zip Hiragishi. We are indebted to the nurses, particularly Kumiko Fujita and Tomomi Yamase, exercise instructors, and management staff of Sapporo Health Promotion Center. We thank the citizens of Sapporo for their interest in and enthusiasm for this project and for their dedication to the spirit of science.

This study was sponsored by the National Health Insurance and National Pension Plan Section, Insurance and Public Health Department, Health and Welfare Bureau of Sapporo. The funding was originally allocated to the local government for health promotion projects by the Ministry of Health, Labor and Welfare of Japan. The sponsor had no role in data management, analysis, or manuscript writing.

Back to Top | Article Outline


1. Bender, R., and S. Lange. Adjusting for multiple testing-when and how? J. Clin. Epidemiol. 54:343-349, 2001.
2. Bigaard, J., K. Frederiksen, A. Tjonneland, et al. Waist circumference and body composition in relation to all-cause mortality in middle-aged men and women. Int. J. Obes. 29:778-784, 2005.
3. Blair, S. N., H. Kohl III, and R. S. Paffenbarger, Jr. Physical fitness and all-cause mortality: a prospective study of healthy men and women. JAMA 262:2395-2401, 1989.
4. Collins, R., R. Peto, S. MacMahon, et al. Blood pressure, stroke and coronary artery disease. Part 2, short-term reductions in blood pressure: overview of randomized drug trials in their epidemiological context. Lancet 335:827-838, 1990.
5. Ebrahim, S., and G. D. Smith. Systematic review of randomised controlled trials of multiple risk factor interventions for preventing coronary heart disease. BMJ 314:1666-1674, 1997.
6. Fagard, R. H. Exercise characteristics and the blood pressure response to dynamic physical training. Med. Sci. Sports Exerc. 33(6 Suppl.):S484-S492, 2001.
7. Fukuhara, S., Y. Suzukamo, S. Bito, and K. Kurosawa. Manual of SF-36, Japanese version 1.2. Tokyo: Public Health Research Foundation, 2001.
8. Gulati, M., D. K. Pandey, M. F. Arnsdorf, et al. Exercise capacity and the risk of death in women: the St James Women Take Heart Project. Circulation 108:1554-1559, 2003.
9. Hokanson, J. E., and M. A. Austin. Plasma triglyceride level is a risk factor for cardiovascular disease independent of high-density lipoprotein cholesterol level: a meta-analysis of population-based prospective studies. J. Cardiovasc. Risk 3:213-219, 1996.
10. Igarashi, K., K. Fujita, T. Yamase, et al. Sapporo Fitness Club Trial (SFCT)-design, recruitment and implementation of a randomized controlled trial to test the efficacy of exercise at a fitness club for the reduction of cardiovascular risk factors. Circ. J. 68:1199-1204, 2004.
11. Katz, M. H. Multivariable Analysis: A Practical Guide for Clinicians. Cambridge: Cambridge University Press, 1999.
12. Katzmarzyk, P. T., A. S. Leon, J. H. Wilmore, et al. Targeting the metabolic syndrome with exercise: evidence from the HERITAGE Family Study. Med. Sci. Sports Exerc. 35:1703-1709, 2003.
13. Kelley, D. E., and B. H. Goodpaster. Effects of exercise on glucose homeostasis in type 2 diabetes mellitus. Med. Sci. Sports Exerc. 33(6 Suppl.):S495-S501, 2001.
14. King, A. C., W. C. Haskell, C. B. Taylor, et al. Group- vs home-based exercise training in healthy older men and women. A community-based clinical trial. JAMA 266:1535-1542, 1991.
15. Leon, A. S., and O. A. Sanchez. Response of blood lipids to exercise training alone or combined with dietary intervention. Med. Sci. Sports Exerc. 33(6 Suppl.):S502-S515, 2001.
16. Matzuzawa, Y., K. Tokunaga, K. Kotani, et al. Simple estimation of ideal body weight from body mass index with the lowest morbidity. Diabetes Res. Clin. Pract. 10:S159-S164, 1990.
17. Munro, B. H. Statistcal Methods for Health Care Research. Philadelphia: Lippincott Williams and Wilkins, 2001.
18. Myers, J., M. Prakash, V. Froelicher, et al. Exercise capacity and mortality among men referred for exercise testing. N. Engl. J. Med. 346:793-801, 2002.
19. Nishijima, H., K. Fujita, K. Satake, K. Okita, K. Yonezawa, and A. Kitabatake. One-year changes in cardiovascular risk factors: lifestyle counseling vs. exercise at a fitness club [in Japanese]. J. Jpn. Assoc. Card. Rehabil. 8:189-194, 2003.
20. Paffenbarger, R. S. Jr., R. T. Hyde, A. L. Wing, and C. Hsieh. Physical activity, all-cause mortality, and longevity of college alumni. N. Engl. J. Med. 314:605-613, 1986.
21. Rexrode, K. M., V. J. Carey, C. H. Hennekens, et al. Abdominal adiposity and coronary heart disease in women. JAMA 280:1843-1848, 1998.
22. Ridker, P. High-sensitivity C-reactive protein: potential adjunct for global risk assessment in the primary prevention of cardiovascular disease. Circulation 103:1813-1818, 2001.
23. Smith, S. C. Jr., R. Jackson, T. A. Pearson, et al. Principles for national and regional guidelines on cardiovascular disease prevention: a scientific statement from the World Heart and Stroke Forum. Circulation 109:3112-3121, 2004.
24. Stewart, K. J., A. C. Bacher, K. Turner, et al. Exercise and risk factors associated with metabolic syndrome in older adults. Am. J. Prev. Med. 28:9-18, 2005.
25. Thompson, P. D., D. Buchner, I. L. Pina, et al. Exercise and physical activity in the prevention and treatment of atherosclerotic cardiovascular disease: a statement from the Council on Clinical Cardiology (Subcommittee on Exercise, Rehabilitation, and Prevention) and the Council on Nutrition, Physical Activity, and Metabolism (Subcommittee on Physical Activity). Circulation 107:3109-3116, 2003.
26. Ware, J. E. Jr., and C. D. Sherbourne. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med. Care 30:473-483, 1992.
27. Wynder, E. L., S. D. Stellman, and E. A. Zang. High fiber intake. Indicator of a healthy lifestyle. JAMA 275:486-487, 1996.


©2007The American College of Sports Medicine