Regular exercise has been observed to have a beneficial effect on various risk factors that are of major, increasing public health concern throughout the world; these include osteoporosis, cardiovascular disease (CVD), and type 2 diabetes. However, designing an optimal exercise program for preventing different diseases at the same time is difficult, or even impossible. Consequently, there are two alternative approaches: a general exercise regimen targeted to prevent risk factors of several diseases, or a specialized exercise regimen dedicated to optimize preventive characteristics of risk factors of an individual disease.
Osteoporosis and osteoporotic fractures are inevitably increasing problems. Exercise programs have been shown to be beneficial for bones by maximizing the peak bone mass and reducing the subsequent bone loss (30). Skeletal adaptation has been suggested to be threshold driven, and impact-type exercise has been found to be the most efficient (5,29,30). However, osteoporotic fractures, especially hip fractures, are generally associated with falls. Therefore, one of the major risk factors for falls and fractures is the decaying physical performance capacity (12). Several different types of exercise have been found to improve muscle strength, balance, coordination and mobility, all of which may help in preventing falls and fall-induced fractures (2,15).
Exercise also has been shown to be associated with a reduced risk of CVD and type 2 diabetes by affecting favorably circulating concentrations of lipids and lipoproteins, glucose metabolism, and visceral fat (13,16). Recently, it has been suggested that osteoporosis and CVD may share some etiologic factors (32), and some previous evidence exists that stepping and jumping exercises may have positive effects on cardiorespiratory fitness and serum lipid levels (14,22,26). However, such information is limited, and no population-based trials exist. In this population-based, randomized exercise trial with continuous measurement of physical activity, we evaluated the effect of a high-impact exercise regimen, which previously had been found to be osteogenic (27-29), on physical performance and glucose and lipid profiles in premenopausal women. Our hypothesis was that the chosen impact exercise regimen would increase maximal oxygen uptake and isometric leg strength and would have a positive effect on lipoprotein levels and glucose metabolism.
SUBJECTS AND METHODS
The study design was a population-based, randomized controlled trial; it has been described in more detail elsewhere (27). The study subjects were randomly obtained from the cohort of all 35- to 40-yr-old women residing in the city of Oulu, Finland in March 2002 (N = 5161). The sample size of 120 participants, with an equal dropout rate of 10 subjects per group, was assessed in advance to detect a 3% difference between the exercise and control groups in bone mineral density, with a significance level of 5% and a power of 80%. This enabled us to detect differences of 0.10 mM in cholesterol levels between the groups, with standard deviations of 0.18 mM (16) and power of 80%. The names, addresses, and dates of birth of the women in the cohort were obtained from the National Population Register of Finland. Thereafter, the names were randomly sequenced, and the women were contacted in this random order. The recruited subjects were randomly assigned to an exercise group (N = 60) or a control group (N = 60). The exclusion criteria for the exercise intervention were any functional limitation or chronic disease that might have limited the training and testing of the cardiovascular, musculoskeletal, and respiratory systems; any disease or medication known to affect bone metabolism; pregnancy; breast feeding; and current or previous participation in impact-type exercise. The study protocol was approved by the local institutional ethics committee, and all participants gave their informed written consent. The procedure of the study was in accordance with the Declaration of Helsinki.
Questionnaires and measurements.
A self-administered health questionnaire was mailed to all contacted women. Information concerning lifestyle factors, including physical activity and dietary factors, was acquired (27). Clinical examinations and all measurements were performed at baseline and at 12 months. Height, weight, and hip and waist circumference were measured in the clinical examination, and the body mass index (BMI) was calculated.
Muscle and fat cross-sectional areas (CSA) were assessed with a spiral quantitative computed tomography scanner (Siemens Somatom Emotion, Siemens GmbH, Munich, Germany) at midthigh (50% of the estimated bone length from the distal endplate of the femur) and at the proximal calf (67% from the distal endplate of the tibia), using the Geanie 2.1 computed tomography analysis software (BonAlyse, Jyväskylä, Finland). The measurement procedure and reproducibility of the method have been reported previously in more detail (11).
Maximal oxygen uptake (V˙O2max) was determined directly using a step protocol consisting of 3 min of progressive uphill walking and jogging to a voluntary maximum on a treadmill (Telineyhtymä Oy, Kotka, Finland). The test protocol was modified from that of Oja (21) and consisted of the following: warm-up, 2-4 min (3 km·h−1, 2° incline); first step, 1-3 min (3.0 km·h−1, 5°); second step, 3-6 min (4 km·h−1, 5°); third step, 6-9 min (5km·h−1, 5°); fourth step, 9-12 min (5 km·h−1), 6°); fifth step, 12-15 min (6 km·h−1, 6°); sixth step, 15-18 min (7km·h−1, 6°); seventh step, 18-21 min (8 km·h−1, 6°); eighth step, 21-24 min (9 km·h−1, 6°); ninth step, 24-27min (10 km·h−1, 6°); tenth step 27-30 min (11 km·h−1, 6°). Oxygen consumption was measured with a Medikro 919 gas analyzer (Medikro Oy, Kuopio, Finland) throughout the test, and there was continuous ECG monitoring during the test. The maximum isometric leg-extensor strength and the simple reaction time were assessed bilaterally with a computerized strain-gauge dynamometer (Digitest, Muurame, Finland) in the seated position, with the knee set at 108°. The grip strength was measured with a hand dynamometer (Digitest). In all strength measurements, the best result of the three consecutive recordings was used in the final analyses. Calibration of the dynamometer was carried out daily before the measurements. The speed-strength of the leg extensor muscles was measured during a countermovement (CM) jump and static jump on a contact mat (Powertimer 1.0, Newtest Ltd., Oulu, Finland) connected to a digital timer (± 0.001 s). The height of the vertical jump was calculated from the measured flight time.
Venous blood samples were collected using a standardized procedure from all the subjects at baseline, at 6 months, and at 12 months. After an overnight fast and rest, a blood sample was drawn at 8:00 to 11:00 a.m. to obtain fasting values for insulin (CV < 7.5%) (ELISA kit from Dako A/S, Glostrup, Denmark), glucose (CV: 0.8%), total cholesterol (CV: 0.9%), high-density lipoprotein (HDL-C) cholesterol (CV: 0.9%), and triglycerides (TG) (CV: 1.0%) (Konelab, Thermo Electron Ltd., Vantaa, Finland). Serum was separated by centrifugation and was stored at −70°C. All samples were analyzed at the end of the study. Low-density lipoprotein cholesterol (LDL-C) was calculated (4).
Exercise training protocol.
The training sessions were carried out three times a week for 12 months. All training sessions were supervised by a physiotherapist and were done with the accompaniment of music. The training regimen was based on a pilot study and on the previous literature. Each workout lasted 60 min, including a 10-min warm-up, a 40-min high-impact training session, and a 10-min cool-down and stretching period. The warm-up period included walking and running on the spot, with and without arm movements and knee bends. The high-impact period included step patterns, stamping, jumping, running, and walking. After 3 months of training, a one-step bench (height, 10 cm) (Reebok, Lancaster, UK) was used to enhance the impact effect, and after 6 months, two or three benches were used. The cool-down consisted mainly of stretching. The programs were modified bimonthly and became progressively more demanding, including higher jumps and drops. Additionally, the participants were asked to train for 10 min daily at home, following a specially designed program that consisted of exercise patterns similar to those in the supervised sessions. The home program also was modified seasonally. The women in the control group were asked to continue their normal daily life and to maintain their current physical activity during the 12 months.
Measurement of physical activity.
All subjects were asked to carry an accelerometer-based human-body movement monitor (Newtest Ltd) daily for 12 months, for individual quantification of their daily physical activity. The monitor records the vertical acceleration peaks and classifies them according to peak acceleration. It was worn on a belt close to the right iliac crest. The subjects were instructed to wear the monitor during all waking hours. The collected data were transferred into a server computer every other week. The acceleration of gravity was subtracted from the results; thus, 0g equaled standing. The number of daily impacts was recorded at 33 acceleration levels, from 0.3 to 9.9g (g = acceleration of gravity: 9.81 m·s−2). Individual daily average values were calculated for the analyses. Acceleration levels attained in exercise patterns used in our exercise training regimen were measured during our preliminary study. The average peak accelerations (range of individual means) in the vertical direction were 0.8g (0.7-1.2g), 3.2g (2.0-4.6g), and 4.2g (2.9-5.2g) in walking at 5 km·h−1, running at 9 km·h−1, and running at 13 km·h−1, respectively, measured on a treadmill. The average peak values for aerobic stepping, lateral jumping, CM jumps, jumps without CM, and drop jumps from 40 cm were 1.2g (0.9-2.2g), 2.0g (1.1-3.3g), 4.4g (1.9-6.5g), 4.6g (2.1-6.4g), and 5.6g (3.8-9.9g), respectively. The physical activity measurement procedure and the validation data have been reported previously (10,29).
The data were analyzed using the SPSS statistical package (SPSS 12.0 for Windows, SPSS Inc. Chicago, IL). All subjects with baseline and end-point measurements were included in the analyses according to their group assignment. The individual average number of daily accelerations was normalized by the mean values of the controls and was analyzed on five acceleration levels to describe exercise volumes at different loading levels: 0.3-1.0g (e.g., walking), 1.1-2.4g (e.g., stepping), 2.5-3.8g (e.g., jogging), 3.9-5.3g (e.g., running, jumping), and 5.4-9.2g (e.g., jumping) (29). Multiple stepwise regression analysis was used to quantify the association between intensity of exercise and change in bone metabolism and to determine confounding factors, which were used in subsequent analyses. The variables used in the stepwise regression analyses were the number of impacts at each acceleration level, weight change, baseline variable value, previous physical activity, smoking, and baseline weight and height. The within-group and between-group changes in end-point variables during the trial were analyzed with the repeated-measures analysis of covariance (ANCOVA). Subgroup analyses were performed with the repeated-measures ANCOVA in quartiles of the measured number of impacts at different loading levels within the pooled groups (Table 1) and in quartiles according to the number of exercise sessions attended within the exercise group. Quartiles were derived separately for each acceleration level. In all tests, P < 0.05 was considered statistically significant.
Thirty-nine women (65%) in the training group and 41 women (68%) in the control group completed the trial. Complete data from the physical performance and metabolism measurements and impact loading recordings from 65 subjects (30 in the control group and 35 in the exercise group) were available. The effect of the intervention on bone mineral density, bone geometry, compliance, reasons for dropouts, and distribution of daily impacts have been reported previously (27,28). There were no significant differences in any baseline variables between the dropouts and the subjects who completed the study.
The changes in anthropometrics and physical performance in the exercise and control groups are presented in Table 2. During the trial, there was a mean weight loss of 0.7 kg in the training group, whereas the control group gained 0.9 kg (P = 0.09). The exercise group demonstrated a significant decrease compared with the control group in waist circumference (−1.1 vs 0.9 cm; P = 0.05) and hip circumference (−1.0 vs 1.1 cm; P = 0.04). In addition, exercise training led to increases in muscle CSA in the midthigh (0.9 vs 0.2 cm2; P = 0.15) and calf (0.9 vs −0.5 cm2; P = 0.02) compared with the control group. In pooled groups, the changes in the midthigh and calf muscle CSA were positively associated with the measured number of impacts exceeding 1.1g (r = 0.36-0.43, P < 0.01-0.001 and r = 0.28-0.36, P < 0.05-0.01, respectively). In regression analyses, the measured number of impacts explained 19% of the change in the midthigh muscle CSA and 13% of the change in the proximal calf muscle CSA; this was the most important predictor of change.
In subgroup analyses, there was a significant difference in thigh and calf muscle CSA between the highest and lowest quartiles of activity at each intensity level exceeding 1.1g in pooled groups (Fig. 1). In the exercise group, those who participated in more than 66 exercise sessions during the 12 months showed increases in thigh (3.6 vs −2.4 cm2; P = 0.001) and calf muscle CSA (2.3 vs −1.0cm2; P < 0.001) compared with the subjects in the lowest quartile (< 20 sessions).
The maximal oxygen uptake increased significantly more in the exercise group than in the control group (6.2 vs 3.1 mL·kg−1·min−1; P = 0.008). The women in the exercise group also improved their CM (2.3 vs −0.3 cm; P < 0.001) and static jump (1.4 vs −0.3cm; P = 0.004) heights significantly more than controls. Both exercise and control groups significantly improved their leg strength (351 N, P < 0.001 and 289 N, P < 0.001, respectively), but there was no difference in the observed changes between the groups (P = 0.69). In pooled groups, the changes in the V˙O2max and CM jumping height were positively associated with the measured number of impacts exceeding 3.9g (r = 0.33-0.34, P < 0.05-0.01 and r = 0.39-0.45, P < 0.01-0.001, respectively). The measured number of impacts accounted for 29% of the variation in CM jump height change and 12% of the variation in V˙O2max change. Additionally, low initial values in V˙O2max and static jump height predicted greater improvements in response within the exercise group.
In the subgroup analyses of the pooled groups, the highest activity quartile showed the most significant increment in V˙O2max and in static and CM jump height compared with the lowest quartile at each intensity level exceeding 1.1g (Fig. 2). Within the exercise group, the subjects who most frequently attended the supervised sessions (> 66 exercise sessions) improved their CM jump height (4.2 vs 0.4 cm; P < 0.001), static jump height (2.7 vs −0.5; P = 0.009), and V˙O2max (12.7 vs 2.1 mL·kg−1·min−1; P < 0.001) significantly more than the subjects in the lowest activity quartile (< 20 sessions).
Lipid and glucose metabolism.
Serum lipid and glucose metabolism measurements are presented in Table 3. The exercise group demonstrated a minor reduction in LDL-C concentration (−0.2 mM; P = 0.02) during the 12 months. Glucose concentrations increased significantly in both the exercise and control groups. In efficacy analyses, cholesterol and LDL-C decreased significantly more in the subjects in the highest activity quartile compared with the lowest activity quartile of measured exercise volumes at every intensity level exceeding 1.1g. The difference in the changes was −0.5 mM (95% CI = −0.8 to −0.2 mM; P = 0.005) in LDL-C concentrations (Fig. 3). Baseline values were the most significant determinants of the changes in lipids within the exercise group, explaining 21, 13, and 27% of the changes in TG, cholesterol, and HDL-C, respectively (Table 4). However, the measured number of impacts was the only significant determinant of the change in LDL-C, explaining 19% of the change.
The exercise regimen initially designed to load weight-bearing bones demonstrated positive effects on extraskeletal risk factors of osteoporotic fractures and physical performance and also on some of the main risk factors of CVD. The most distinct improvements were found in the explosive power properties and muscle mass of the lower limbs and oxygen uptake. Total daily physical activity was associated with changes in waist and hip circumference and changes in lipid and lipoprotein levels, implying a reduced risk of CVD. The changes seemed to have a dose-response relationship with the measured quantity and intensity of exercise.
The role of physical activity and exercise (especially resistance and high-impact activities) in the prevention of osteoporosis is commonly accepted (20). Several high-impact intervention trials have been successful in improving bone mass (30), whereas resistance training has been shown to be most effective for muscle strength and mass (19). High-impact exercise also has been found to improve muscular performance, dynamic balance, and oxygen uptake in healthy premenopausal women (7) and in early postmenopausal women (25,26). The 12-month treatment effect of 13% in lower-limb plyometric force found here is higher than previously reported (7); this might be partly explained by the slight differences in the exercise training regimen. We found no treatment effect in the isometric leg muscle strength, which is in line with the results of Heinonen et al. (7), even though definite improvements were found in both groups in our study. Most of the subjects were unaccustomed to testing procedures, and regardless of careful instructions, increases found in isometric leg strength may partly be explained by the learning effect, even though exercise training clearly improved muscle CSA. Altogether, high-impact training with rapid movements may improve neuromuscular control, reaction ability, and coordination, in addition to improved speed-strength and muscle performance-all important risk factors in falling, fracture, and disability.
CVD is the leading cause of death in women (23). The number of cardiovascular deaths is expected to almost double from 13.1 million in 1990 to 24.8 million in 2020 (24). Lipid and lipoprotein levels, blood pressure, cardiorespiratory fitness, obesity, diabetes, and smoking are the main modifiable risk factors of CVD (24). Our study had favorable effects on several of these main risk factors. Exercise improved cardiorespiratory fitness and had a 12% treatment effect on maximal oxygen uptake; this is consistent with previous osteoporosis-prevention exercise interventions (7,14,26). However, there was also a minor increase in the maximal oxygen uptake in the control group; this increase may be attributable to a learning effect together with the intervention effect. Previously maximal oxygen uptake has been found to be the strongest predictor of CVD-related mortality in the Finnish population (18). Lipid and lipoprotein (especially LDL-C) concentrations are also major determinants of vascular events (24). A drug-induced decrease of 1 mM in the LDL-C concentration has been reported to lead to a 25% reduction in relative risk of CVD events (6). The 12-month exercise period induced a difference of 0.5 mM in the change in LDL-C concentration between the most and least active subjects. The difference is clinically significant and demonstrates the potential of exercise in the reduction of CVD risk. Although exercise had been reported to have favorable effects on various factors associated with enhanced glucose homeostasis (1,17), our study has demonstrated only minor effects on glucose metabolism. This may be partly attributable to evident degradation of glucose during storage. Exercise-induced metabolic changes were inversely related to initial metabolic parameters, indicating that exercise could be more beneficial for the subjects at the highest risk of impaired glucose metabolism.
We observed a dose-response relationship between measured daily physical activity and physical fitness and lipoprotein changes. Improvements were found in quite a wide intensity range (including even light exercise types, such as slow aerobic stepping) when the repetition number was high. Walking and other very-low-intensity exercise was not associated with changes, whereas the greatest exercise effects between the most active and least active subjects were found in moderate- to high-intensity levels of exercise. The amount of impacts in the most active subjects was 1500 per day exceeding 1g-equivalent to, for instance, 1.5 km of jogging or 15 min of step aerobics. Physical fitness changes also were associated with compliance to supervised training sessions, but this association was not observed in lipids. This may reflect the importance of total daily activity levels rather than specified training regimens in relation to lipids. Our findings are complementary to the previous studies suggesting that the amount of exercise is more important than its intensity for improving lipoprotein profiles and insulin sensitivity (9,16). Yet, increases in either the intensity or the amount of exercise may yield additional benefits in terms of cardiorespiratory fitness (3).
This study also had limitations. Previously, accelerometer data has been found to have moderate to good association with energy expenditure (8). When comparing our results with the previous findings, it must be noted that our accelerometer-based measurement device was designed and validated to analyze the quantity and intensity of mechanical loading, that it has not been validated to measure energy expenditure or oxygen uptake, and that it may underestimate the effects of nonimpact exercise and the increased energy costs of upper-body movements. Nevertheless, it provides a measured estimate of the amount and intensity of exercise. Recently published recommendations (31) for the use of accelerometers in clinical studies also should be considered in future studies. The exercise regimen in this study was designed to optimize the effect of exercise on the weight-bearing skeleton, and the aim was to find the optimal quantity and quality of exercise for preventing osteoporosis. Thus, the exercise training regimen may not have been optimal for cardiovascular outcomes. However, our study has revealed that an exercise program designed for bone health may also have a favorable effect on several cardiovascular risk factors and aerobic capacity.
In conclusion, our results suggest that an exercise program targeted at weight-bearing bones also has as favorable effect on physical performance and cardiorespiratory risk factors with a moderate amount of exercise at a wide intensity range. The response is dose dependent, but it is also inversely related to the baseline levels of the outcome variables. In addition to the prevention of osteoporosis, impact exercise may be recommended for prevention of CVD.
We thank Ms. Minna Tervo, the physiotherapist in our study team, for supervising the training and testing of our trial participants. This work was supported by the National Technology Agency of Finland; the Juho Vainio Foundation, Helsinki, Finland; the Instrumentarium Research Foundation, Helsinki, Finland; the Research Foundation of the Institutes of Sports, Helsinki, Finland; The Finnish Cultural Foundation, Helsinki, Finland; the Finnish Foundation for Sports Research, Helsinki, Finland; Newtest Ltd., Oulu, Finland; CCC Group, Oulunsalo, Finland; and Fastrax Ltd., Vantaa, Finland. The corporate financiers had no control on the conduct or the publication of the study.
Dr. Vainionpää, Mr. Kaikkonen, and Prof. Knip have no conflicts of interests; Prof. Leppäluoto, Prof. Jämsä, and Dr. Korpelainen have a patent application with Newtest Ltd. Prof. Jämsä is also a minor shareholder of Newtest Ltd.
1. Bassuk, S. S., and J. E. Manson. Epidemiological evidence for the role of physical activity in reducing risk of type 2 diabetes and cardiovascular disease. J. Appl. Physiol.
2. Borer, K. T. Physical activity in the prevention and amelioration of osteoporosis in women: interaction of mechanical, hormonal and dietary factors. Sports Med.
3. Duscha, B. D., C. A. Slentz, J. L. Johnson, et al. Effects of exercise training amount and intensity on peak oxygen consumption in middle-age men and women at risk for cardiovascular disease. Chest
4. Friedewald, W. T., R. I. Levy, and D. S. Fredrickson. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin. Chem.
5. Frost, H. M. Bone's mechanostat: a 2003 update. Anat. Rec.
6. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20 536 high-risk individuals: a randomised placebo-controlled trial. Lancet
7. Heinonen, A., P. Kannus, H. Sievänen, et al. Randomised controlled trial of effect of high-impact exercise on selected risk factors for osteoporotic fractures. Lancet
8. Hendelman, D., K. Miller, C. Baggett, E. Debold, and P. Freedson. Validity of accelerometry for the assessment of moderate intensity physical activity in the field. Med. Sci. Sports Exerc.
9. Houmard, J. A., C. J. Tanner, C. A. Slentz, B. D. Duscha, J. S. McCartney, and W. E. Kraus. Effect of the volume and intensity of exercise training on insulin sensitivity. J. Appl. Physiol.
10. Jämsä, T., A. Vainionpää, R. Korpelainen, E. Vihriälä, and J. Leppäluoto. Effect of daily physical activity on proximal femur. Clin. Biomech. (Bristol, Avon)
21: 1-7, 2006.
11. Jämsä, T., O. Parviainen, I. Salmela, and P. Pulkkinen. Assesment of lower limb cross-sectional geometry using CT image analysis. Technol. Health Care
12. Kannus, P., H. Sievänen, M. Palvanen, T. Järvinen, and J. Parkkari. Prevention of falls and consequent injuries in elderly people. Lancet
13. Kelley, G. A., K. S. Kelley, and Z. V. Tran. Aerobic exercise and lipids and lipoproteins
in women: a meta-analysis of randomized controlled trials. J. Womens Health (Larchmt.)
14. Kemmler, W., D. Lauber, J. Weineck, J. Hensen, W. Kalender, and K. Engelke. Benefits of 2 years of intense exercise on bone density, physical fitness, and blood lipids in early postmenopausal osteopenic women: results of the Erlangen Fitness Osteoporosis Prevention Study (EFOPS). Arch. Intern. Med.
15. Korpelainen, R., S. Keinänen-Kiukaanniemi, J. Heikkinen, K. Väänänen, and J. Korpelainen. Effect of exercise on extraskeletal risk factors for hip fractures in elderly women with low BMD: a population-based randomized controlled trial. J. Bone Miner. Res.
16. Kraus, W. E., J. A. Houmard, B. D. Duscha, et al. Effects of the amount and intensity of exercise on plasma lipoproteins. N. Engl. J. Med.
17. LaMonte, M. J., S. N. Blair, and T. S. Church. Physical activity and diabetes prevention. J. Appl. Physiol.
18. Laukkanen, J., T. A. Lakka, R. Rauramaa, et al. Cardiovascular fitness as a predictor of mortality in men. Arch. Intern. Med.
19. Lohman, T., S. Going, R. Pamenter, et al. Effects of resistance training on regional and total bone mineral density in premenopausal women: a randomized prospective study. J. Bone Miner. Res.
20. NIH Consensus Development Panel on Osteoporosis Prevention. Osteoporosis prevention, diagnosis, and therapy. JAMA
21. Oja, P. Intensity and frequency of physical conditioning and determinants of the cardiovascular response of middle-aged men at rest and during exercise [dissertation]. State College, PA: Pennsylvania State University, 1973.
22. Olson, M. S., H. N. Williford, D. L. Blessing, and R. Greathouse. The cardiovascular and metabolic effects of bench stepping exercise in females. Med. Sci. Sports Exerc.
23. Peden, M., K. McGee, and E. Krug (Eds.). Injury: a Leading Cause of the Global Burden of Disease, 2000
. Geneva, Switzerland: World Health Organization, pp. 5-16, 2002.
24. Poulter, N. Global risk of cardiovascular disease. Heart
89(Suppl. 2):ii2-ii5, 2003.
25. Sipilä, S., D. R. Taaffe, S. Cheng, J. Puolakka, J. Toivanen, and H. Suominen. Effects of hormone replacement therapy and high-impact physical exercise on skeletal muscle in post-menopausal women: a randomized placebo-controlled study. Clin. Sci. (Lond.)
26. Uusi-Rasi, K., P. Kannus, S. Cheng, et al. Effect of alendronate and exercise on bone and physical performance of postmenopausal women: a randomized controlled trial. Bone
27. Vainionpää, A., R. Korpelainen, J. Leppäluoto, and T. Jämsä. Effects of high-impact exercise on bone mineral density: a randomized controlled trial in premenopausal women. Osteoporos. Int.
28. Vainionpää, A., R. K. Korpelainen, H. S. Sievänen, E. Vihriälä, J. Leppäluoto, and T. Jämsä. Effect of impact exercise and its intensity on bone geometry at weight-bearing tibia and femur. Bone
29. Vainionpää, A., R. Korpelainen, E. Vihriälä, A. Rinta-Paavola, J. Leppäluoto, and T. Jämsä. Intensity of exercise is associated with bone density change in premenopausal women. Osteoporos. Int.
30. Wallace, B. A., and R. G. Cumming. Systematic review of randomized trials of the effect of exercise on bone mass in pre- and postmenopausal women. Calcif. Tissue Int.
31. Ward, D. S., K. R. Evenson, A. Vaughn, A. B. Rodgers, and R. P. Troiano. Accelerometer
use in physical activity: best practices and research recommendations. Med. Sci. Sports Exerc.
32. Whitney, C., D. E. Warburton, J. Frohlich, S. Y. Chan, H. McKay, and K. Khan. Are cardiovascular disease and osteoporosis directly linked? Sports Med.