Currently, >40.2 million individuals over the age of 65 years reside in the USA (36). These individuals are at an increased risk of sustaining an osteoporotic fracture because of the bone loss that accompanies the aging process (16,30). Participation in physical activity is advocated for older individuals because of its protective effects on bone mineral density (BMD) (7,14). Numerous intervention studies have reported positive effects of exercise on BMD (1,7,10,12–14,19,21,23,25,26,30,31,34,35). Most of these studies have either been on younger individuals or have been short-term exercise interventions in the elderly. At present, the effects of long-term elite highly competitive exercise on BMD in older individuals are unknown.
Bone remodeling increases in response to mechanical stress (28,30). The mechanical stress associated with exercise is delivered via 2 methods: the longitudinal loading of the skeleton, such as during weight-bearing exercise, and the tensile pull of the muscles on the bone (28). However, the beneficial effects of exercise on BMD may be mitigated by the fact that the greater body mass of heavier individuals has a protective effect against bone loss because of the concomitant increases in longitudinal loading of the skeleton from the larger body mass and the greater muscle forces required to move a larger person (2,3,14,29). Previous data confirm that elite highly competitive elderly athletes have less body mass than do healthy age-matched controls (18).
Muscle strength has been studied as a predictor of BMD (10,12,17,21,22,27,34,35). Much of the research in this area has been performed on adolescents or on young adults (1,10,12,21,26,27). Conroy et al. (10) reported that muscle strength may explain 30–65% of the variance in BMD in elite young weightlifters. Others have reported that the relationship between muscle strength and BMD is stronger among sedentary than among active individuals (1,12,21,27). Investigators who have studied elderly individuals also report a positive association between muscle strength and BMD (2,3,13,14,24,27,31,33–35,37), elucidating the importance of exercise for an aging population.
The National Senior Games Association is an American nonprofit organization dedicated to motivating older adults to lead healthy lifestyles. The National Senior Games, also known as the “Senior Olympics,” is the largest multisport event in the world for senior athletes. All events are divided into 5-year age categories so that the participants compete against other athletes of approximately the same age. Approximately 250,000 individuals participate in the community-level games. The competitors who make it to the National (i.e., American) Senior Games represent the upper 5% of older competitive athletes in the USA. The National Senior Olympians comprise a unique study cohort because of their history of success at a highly competitive level over a number of years. Study of these senior athletes has the potential to provide more information on the benefits of long-term competitive exercise than what can be learned from short-term exercise interventions. Examining these unique athletes can provide an insight into the effects of long-term highly intense exercise on the mitigation of the musculoskeletal decline associated with aging.
Various aspects of BMD and muscle strength have been examined previously in this cohort of senior athletes who participated in this study. Stone et al. (32) reported that these senior athletes had a lower bodyweight, body mass index, percent body fat, and greater lean mass compared with those of a group of healthy controls. Additionally, in women, age and weight significantly contributed to BMD, but athlete status (i.e., athlete or control) did not (32). Velez et al. (38) reported significantly greater total body BMD in runners compared with that in controls, whereas those of senior swimmers did not differ from controls. Additionally, we reported that these senior athletes have greater knee extension and flexion strength than did healthy aged-matched controls (18). Given that the total body BMD was greater in these senior runners, and that the senior athletes displayed greater strength of the knee musculature, one may assume that the greater muscle strength is associated with greater BMD. However, the correlation between BMD and strength in these unique highly competitive athletes has not been established.
The goals of our project were to examine the association between (a) participation in highly competitive exercise and site-specific BMD and (b) the relationship between knee extensor and flexor strength and BMD, while controlling for age, bodyweight, sex, and vitamin D and calcium intake, in a cohort of individuals aged 65 years or older. We hypothesized that (a) participation in highly competitive athletics would be positively related to total body and regional BMD and (b) maximum knee flexion and extension strength would be predictive of total hip BMD when age, group (i.e., elite athlete or control), bodyweight, sex, and dietary considerations are used as covariates.
Experimental Approach to the Problem
The rationale for this study is as follows: given that this group of senior Olympians have demonstrated a greater total body BMD than did controls and given that the athletes displayed greater strength of the knee musculature, our 2 primary research objectives were to examine whether (a) participation in highly competitive elite exercise provides a protective effect against bone demineralization associated with aging at 4 sites in the body and (b) if their greater thigh muscle strength (i.e., knee extensors and flexors) is predictive of BMD in this study cohort. We included known covariates to BMD in our statistical model, namely, age, sex, bodyweight, and dietary vitamin D and calcium. We recruited 104 senior athletes from the National Senior Games and 79 healthy community dwelling seniors to participate. Our independent predictor variables include the following: age, weight, sex, dietary calcium and vitamin D, group (i.e., senior Olympian or control), knee extensor peak torque, and knee flexor peak torque. Our dependent variables included the following: total body BMD, and BMD at the forearm, lumbar spine, and hip. A stepwise multivariate linear regression analysis was performed to determine the ability of our independent variables to predict BMD. Step 1 examined the ability of age, sex, bodyweight, and dietary calcium and vitamin D to predict BMD at the 4 sites. Step 2 examined the ability of participation in highly competitive elite exercise to further predict site-specific BMD. Step 3 involved the ability of knee extensor strength to further predict site-specific BMD. In Step 4, knee extensor strength was removed from the multivariate regression analysis and knee flexor strength was included. This study will provide information for athletes, coaches, and clinicians on whether long-term intense exercise is protective against typical BMD losses that accompany aging, and whether knee muscle strength is indicative of site-specific BMD in the elderly.
A group of elite senior athletes (n = 104) and a control group of healthy community dwelling seniors (n = 79) participated in this study. All the subjects were 65 years or older. The subjects in the athletically competitive group were participants of the 2005 National Senior Games (aka the “Senior Olympics”). They competed in 1 of the following sports: running events >400 m (n = 44; 28 men, 16 women), cycling events >5 km (n = 17; 11 men and 6 women), and any swimming event (n = 43; 25 men and 18 women). These events were chosen because data from this study were collected simultaneously with another study, an aim of which was to investigate the magnitude of lower extremity loading (i.e., high, medium, and low impact sports) on bone density (37). All the subjects signed an informed consent document approved by the university Institutional Review Board.
Considering the age range of our participants, disease or injury could have been strong confounding factors. Therefore, we recruited as healthy a subject population as possible in both the athlete and control groups. The control group consisted of healthy individuals whose activity levels ranged from very sedentary to active in sports other than running, cycling, and swimming (e.g., softball and tennis). However, none of the controls were competitively active in these or any other sports. Subject demographics are provided in Table 1. Controls were significantly older and heavier (p < 0.05) than the athletes were.
The subjects were excluded from the study if they had a history of any of the following: chronic obstructive pulmonary disease, myocardial infarction, or coronary artery disease; cerebral vascular accident or a history of transient ischemic attacks; rheumatoid arthritis, gout, or osteoarthritis severe enough to limit activity; lower extremity joint replacement, use of a cane or walker; history of osteoporosis or bisphosphonate therapy; current use of antidepressant drugs or any drugs that may interfere with neurological, musculoskeletal, or cognitive function; history of insulin-dependent diabetes mellitus or neurological or rheumatological disorders that might interfere with sensory input; fracture, ligament reconstruction, or sprain within the past 12 months; and any other disease, injury, or disorder that may affect strength.
The athletes were recruited from the registration area of the 2005 Summer National Senior Games, which were held in Pittsburgh, PA, USA. Community dwelling healthy control subjects were recruited through the University of Pittsburgh's Claude D. Pepper Older Americans Independence Center, a federally funded research center focused on keeping elders independently living and functioning. We did not control for the activity level of this group; therefore, their self-reported activity levels ranged from sedentary to active (i.e., exercising for >3 d·wk−1 for >1 h·d−1). However, none were active in running, cycling, or swimming, and none competed in the Senior Olympics, even at the local level. This group represented healthy but not competitively active individuals.
All the subjects completed the same experimental protocol. All the procedures were approved by the University of Pittsburgh Institutional Review Board before data collection. Each subject participated in 1 data collection session that lasted for approximately 3 hours. Data collection occurred at the Clinical and Translational Research Center in the University of Pittsburgh Medical Center Montefiore Hospital.
Because of the strong inverse association between quadriceps strength and morbidity and mortality (33), we assessed the strength of the thigh musculature. All measurements were taken on the subject's left side. Isometric knee extension and flexion torque of the left leg were measured with a tension and compression load cell (model 3132; Lebow Products Inc., Troy, MI, USA). Data were collected at 100 Hz. The load cell was attached to an adjustable bar, which was attached to a metal bar secured to a custom designed aluminum chair. This design allowed for adjustments to be made for height and leg length differences of the subjects. Pilot testing in our laboratory has determined that the output of the chair has an intraclass correlation coefficient (ICC) of 0.907 with that of a Biodex dynamometer. This strength assessment device has been used in previously published research (18).
The subjects were placed in a comfortably seated position on the “strength chair” and secured using canvas straps to minimize extraneous body movements. The hip was positioned at 90° of flexion, and the knee was positioned in 45° of flexion. The load cell was attached to the adjustable bar so that the load cell could be positioned just proximal to the malleoli. The distance from the knee to the adjustable bar was recorded to calculate torque. The subjects were asked to position their arms either crossed at the chest level or resting on their lap to avoid bracing or pulling on the chair.
After gravity effect torque was calculated, the subjects performed 3 repetitions of maximal isometric knee extension or flexion lasting 5 seconds each. The order in which flexion or extension was tested was counterbalanced to avoid the influence of any learning effect on the data. Thirty seconds of rest was provided between contractions. Similarly, the subjects performed 3 repetitions, each lasting 5 seconds, of maximal isometric knee flexion or extension. Thirty seconds of rest was again provided between contractions.
The subjects were instructed to continue breathing during the tests and to not hold their breath to prevent against doing the Valsalva maneuver during the tests. The subjects were also instructed that they could stop the test at any time. During the test, the subjects were instructed to begin and were encouraged to “push” for the extension trials and to “pull” for the flexion trials. The test administrator loudly repeated the word “push” or “pull” each second for a total of 5 seconds.
Data were analyzed using Matlab 7.1 (Mathworks Inc., Natick, MA, USA). Raw data were filtered with a fourth-order low-pass Butterworth filter with a cutoff frequency of 25 Hz. This was chosen as the cutoff frequency based on the residual analysis described by Winter et al. (40). The torque was calculated as the product of the force in the load cell times the distance from the knee to the adjustable bar. Peak torque was recorded as the maximum torque value for each trial of knee extension and flexion. The 3 trials were averaged to yield representative values for each subject. Isometric peak torque has an ICC of >0.89 (4).
Bone density measurements were made using a Hologic QDR-4500A (Hologic Inc., Bedford, MA, USA) dual-energy x-ray absorptiometer (DXA). A licensed DXA technician performed the scans. The BMD was obtained for each of the following body regions: total hip, lumbar spine, distal radius, and total body. For the total hip and the forearm, measurements were obtained on the left side only because for the majority of the USA population, it is their nondominant side. In our study, 89 of the athletes declared themselves to be right leg dominant, 3 left leg dominant, and remainder stated that they had no preference. Seventy of the controls stated that they were right leg dominant.
The participants completed food frequency questionnaires to evaluate calcium and vitamin D intake in their diet and vitamin supplements (11). These questionnaires were administered orally to each subject by an investigator. The subjects were also asked to complete a physical activity questionnaire to determine lifetime and current physical activity (15). The questionnaire was handed to the subjects, and they were instructed to complete it before leaving the testing center. From this questionnaire, we confirmed that the senior athletes had been competitively active for at least 20 years. The activity levels of the healthy controls ranged from sedentary to active but not competitively active.
After the data were collected, independent t-tests were performed to compare senior athletes with controls on the demographic variables of age, height, and weight (α = 0.05). Before analysis of the experimental strength and BMD data, an exploratory statistical analysis was conducted to determine whether the statistical assumptions were fulfilled for the planned analysis. Descriptive statistics were obtained on each of the variables.
For each of the 4 BMD sites assessed, multivariate linear regression analysis was performed in 4 steps (α = 0.10). Because age (years), bodyweight (N), sex (male, female), dietary calcium (milligrams), and dietary Vitamin D (international units) are all known to be related to BMD, these factors were first entered into the model simultaneously as a block to determine their relationship with BMD at the hip, spine, forearm, and total body. Second, the variable of group (i.e., elite senior athlete or healthy control) was added, and the change in R2 at each of the BMD sites was noted.
In the third step in the regression analysis, knee extension peak torque (Newton meters) was added to the model. The amount of variance explained by this variable (R2) was noted. Then, knee extension peak torque was removed from the model. Finally, knee flexion peak torque (Newton meters) was added to the original model to determine the amount of variance explained by that variable. The 2 strength variables were not entered into the model together because of the high correlation between them (R = 0.66). The alpha level was set to 0.10 to achieve a statistical power of 80%. Peak torque was analyzed in units of Newton meters and not normalized to bodyweight. Bodyweight is known to be positively correlated to the BMD (2,3). Because of this, bodyweight itself was a predictor in the regression model.
Strength data on 7 subjects were not included in the analysis because of equipment malfunction. Three subjects (athlete group, 2 male, 1 female) were outliers because of the extreme level of their strength. These subjects were removed from the multivariate regression analysis because very high Mahalanobis distance scores indicated that the subjects were multivariate outliers and would have an undue influence on the statistical results of the study.
In the first stage of the regression model, age, sex, bodyweight, dietary calcium and dietary vitamin D were entered as a block. At each of the sites assessed, these factors were highly correlated to BMD (p < 0.001, Table 2). The block explained roughly 24, 53, 25, and 34% of the variance in BMD at the hip, radius, spine, and total body, respectively. Group (i.e., elite senior athlete or healthy control) was entered into the statistical model in the second stage of the regression analysis. Group did not add significantly to the model at any of the 4 sites assessed (Table 2).
Knee extension peak torque (Newton meters) was added to the model in step 3 of the multivariate regression analysis. Its inclusion significantly added to the model for hip BMD by increasing the R2 by 3.8% (p = 0.06). However, knee extension peak torque did not contribute significantly to the model at the forearm, spine, or total body (Table 2). In step 4, knee extension peak torque was removed, and knee flexion peak torque (Newton meters) was added to the model. Knee flexion peak torque was not found to be a significant contributor to BMD at the hip, forearm, spine, or total body (Table 2). Visual representations of the association of BMD at each of the 4 sites to knee extension and flexion peak torque are provided in Figures 1–4.
The influence of elite competitive exercise in seniors and thigh muscle strength on BMD at the hip, forearm, spine, and total body was examined in this study. We determined that, as a block, age, sex, bodyweight, and dietary vitamin D, and calcium explained a significant amount of the variance in the BMD in each of the 4 body regions assessed. Participation in elite highly competitive exercise did not affect the BMD at any of the regional sites or in the total body. Additionally, knee extension strength explained 3.8% of the variance in hip BMD, but knee flexion strength did not have an influence at any of the sites.
When entered as a block into the statistical model, the factors of age, bodyweight, sex, and dietary calcium and vitamin D were significantly correlated to BMD. This finding was expected, because numerous other researchers have reported similar findings (3,20,29,34,35). Strong evidence supports the contribution of these factors to BMD. These factors were considered in the first step of the model so that we could examine any additional contribution of exercise level and thigh muscle strength.
Both cross-sectional and intervention studies have reported a positive correlation between physical activity and BMD (5,6,8,13,19,30,31,37). Also, we have previously reported that the senior Olympians who were runners had a greater total body BMD compared with that of controls (38) and that the senior athletes demonstrated greater knee extension and flexion strength than did these controls (18). As such, we expected to find that “group,” or whether the person was an elite senior athlete or a control, would be related to BMD. Our results do not support this. We believe that there are several reasons for this finding. Our control group was not necessarily sedentary. The activity level of the control group was in the typical range of a healthy person over the age of 65 years, ranging from sedentary to active in sports other than running, cycling, and swimming. However, the control group was not competitively active, nor did they participate in the level of training needed for competitive activities. Additionally, bodyweight is highly correlated to BMD (2,3,14,29). The senior athletes in this study weighed significantly less by approximately 7 kg when compared to the control group. The lower bodyweight may mitigate any positive effects of exercise in this population. Additionally, we did not assess endocrine factors in this study. Given that the senior Olympians have been life-long athletes, hormonal effects on BMD may have confounded the study results.
Knee extension strength was related to total hip BMD but not to BMD at the other sites assessed. This relationship is most likely because of the close anatomical association between the quadriceps and the hip. For example, the highest activation of the quadriceps during walking and running is immediately after heel contact, when the quadriceps are eccentrically controlling knee flexion during the landing (9). Similarly, the compressive forces at the hip during locomotion are at a peak after heel contact when the leg is supporting the bodyweight and absorbing the inertial forces associated with impacting the ground (39). Knee flexion strength was not related to hip BMD, however. The results of other studies support this finding (12).
Others researchers have reported a positive association between quadriceps strength and BMD. Eickhoff et al. (12) reported that knee extensor strength, but not flexor strength, was predictive of femoral neck BMD in sedentary young women. Our results are similar, although we measured older men and women, some of whom were competitively active. Taaffe et al. (35) stated that the portion of BMD variance explained by strength (11–21%) is low. In our sample population, we found the amount of variance explained by knee extension peak torque to be 3.8%. Taaffe et al. (34) also noted that the relationship of dynamic strength to BMD in older individuals is generally not site specific.
The association between knee extension strength and hip BMD in this study was not as strong as that reported by others. This weaker association may be because of the design of our statistical model. In our study, the data of the athletes and the controls were included in the prediction model together. Others have analyzed athlete and control data separately and found a relationship in the nonathletes only (24). Therefore, for the purpose of this discussion, separate analyses were run on the data of the athletes and controls. However, no new information was gained. The results were the same, except that when the data were separated into 2 groups, the sample size of each is much smaller than in the combined data set, and quadriceps strength was no longer predictive of BMD in either group.
In some of the research published by others, strength alone was correlated to BMD, and no other variables that are known to affect BMD were considered in the statistical model. In this study, we chose to determine the relationship of strength to BMD after variables already known to strongly influence BMD were considered in the model. Ribom et al. (27) performed 1 univariate regression analysis to correlate knee extension strength with total body BMD in women, and a second regression analysis to correlate knee flexion strength with total body BMD. They reported significant correlations with R2 being 0.27 (p < 0.001) for each (27). If we had simply correlated knee extension strength to BMD and no confounding variables were considered, the R-value would be 0.341 (p < 0.001). Similarly, Ribom et al. (27) reported that correlation of flexion peak torque and hip BMD is 0.199 (p = 0.010). However, we do not consider this model to be accurate because strength is not independent of age, bodyweight, and sex. We feel that it is important to consider these confounding variables in the statistical model.
Madsen et al. (17), like us, chose to account for age and bodyweight first, and then include quadriceps strength in their prediction of forearm and tibial BMD. They reported that quadriceps strength was associated with tibial BMD but not forearm BMD (17). These associations are not surprising considering the close anatomical connection between the quadriceps and the tibia and the distant anatomical connection between the quadriceps and the forearm. We also did not find a relationship between quadriceps strength and forearm BMD.
There are several limitations to this study. The study is cross-sectional in design, and therefore, the influence of the addition of a physical activity regimen on BMD within a single person cannot be assessed. The sample size of 183 individuals was relatively small for a study of this nature. Because we did not personally administer the physical activity questionnaire to our participants, but rather they were given the questionnaire and told to complete it before they left the testing center, the results may not be as complete and accurate as they would have been if the questionnaire was personally administered to each subject. Therefore, these data were not analyzed in detail. Rather, the questionnaires were used to confirm that the senior athletes exercised for at least the past 20 years, and that the controls did not participate in highly competitive exercise. The lifetime physical activity of the study participants is unknown. The control group was healthy and did not have many of the health complications commonly seen in individuals over age 65 years.
After the completion of the study, we realized that our activity surveys were incomplete, particularly in regard to resistance training. Therefore, we cannot truly quantify the amount of resistance training done by the subjects. We accept that this is a major limitation of the study as resistance training can have a large influence on BMD (10,12,14,27). However, of the 104 senior Olympians, 50 definitively stated that they either did resistance training (n = 32) or that they did not (n = 18), whereas the other 54 subjects did not answer that question. For the purpose of this discussion, we performed a post hoc analysis of variance on BMD at each of the 4 sites in those senior athletes who reported resistance training and those who reported that they did not resistance train. Sex (male, female) and resistance training (yes, no) were the independent variables in this analysis and BMD was the dependent variable (α = 0.05). The results of this analysis are shown in Table 3. The BMD at the hip, radius, spine, and total body BMD were not different among those senior athletes who performed resistance training and those who did not, although sex was a significant factor in each analysis (p < 0.05). The sex by resistance training interaction was also not statistically significant. Although we agree that resistance training can have a very large impact on BMD, we did not see a difference in BMD between those athletes who resistance trained and those who did not. We accept that the lack of complete resistance training information is a limitation of the study and it may confound the results, but we believe that our results are valid given that resistance training was not significant in this study subsample.
Because we were trying to get as many senior Olympians as possible to participate in our study, we scheduled their testing sessions at their convenience. We did not control for nutrition or hydration status at the time of testing. To our knowledge, most of the athletes requested to be tested on a noncompetition day. We do not believe that we tested any athletes immediately after competition. Nutrition and hydration may therefore be confounding factors, although we do not believe that they are.
Master athletes may be considered an ideal model of aging because of their long-term participation in high-intensity exercise. Most individuals would not be amenable to performing the amount, intensity, and long-term duration of exercise training reported by senior athletes of this level. Therefore, a training intervention of this magnitude would not be feasible. This study examined the influence of highly competitive exercise, and the training it entails, on BMD in healthy elders. This intense training for the National Senior Games may be considered a best-case scenario for a training intervention. Our ultimate goal in this line of research is to determine the extent to which intense exercise training in the elderly improves physical function and ultimately reduces or delays frailty and morbidity.
Although exercise is likely to protect against the bone demineralization that accompanies the aging process, this study revealed that participation in long-term highly competitive exercise does not necessarily protect the senior athletes from age-associated bone loss. Numerous factors are known to contribute to BMD: age, sex, bodyweight, dietary considerations, exercise participation particularly because it relates to skeletal loading, strength, etc. Bodyweight, sex, age, and dietary calcium and vitamin D were better able to predict BMD than participation in elite competitive exercise. This may be because the senior Olympians weighed less than the control group did, and bodyweight is strongly correlated to BMD. Furthermore, quadriceps strength is only marginally predictive of hip BMD and not predictive of wrist, spine, or whole-body BMD. Senior Olympians must be aware that although exercise does improve BMD, bone health is multifactorial. Being a senior Olympian was not automatically protective of BMD and athletes should therefore have BMD tests at the same intervals as their noncompetitively active counterparts.
The authors would like to acknowledge the University of Pittsburgh Claude D. Pepper Older Americans Independence Center for allowing them to recruit our healthy control subjects through its research registry (Grant # P30 AG024827-01AG). Additionally, this study was performed with the permission of the National Senior Games Association. The authors would also thank Megan Miller, CCRC, Karen Vujevich, CRNP, and Julie Wagner, PAC, MPA, for their assistance during data collection, and Dr. Susan Sereika, of the University of Pittsburgh School of Nursing for her assistance in performing the statistical analyses. Results of this study do not constitute endorsement by the National Strength and Conditioning Association. The authors of the study do not have relationships with companies or manufacturers who will benefit from the results of this study.
1. Alfredson H, Nordstrom P, Lorentzon R. Total and regional bone mass in female soccer players. Calcif Tissue Int 59: 438–442, 1996.
2. Aloia JF, Vaswani A, Ma R, Flaster E. To what extent is bone mass determined by fat-free or fat mass? Am J Clin Nutr 61: 1110–1114, 1995.
3. Blain H, Vuillemin A, Teissier A, Hanesse B, Guillemin F, Jeandel C. Influence of muscle strength and body weight and composition on regional bone mineral density in healthy women aged 60 years and over. Gerontology 47: 207–212, 2001.
4. Callaghan MJ, McCarthy CJ, Oldham JA. Electromyographic fatigue characteristics of the quadriceps in patellofemoral pain syndrome. Man Ther 6: 27–33, 2001.
5. Cauley JA. The determinants of fracture in men. J Musculoskelet Neuronal Interact 2: 220–221, 2002.
6. Cauley JA, Fullman RL, Stone KL, Zmuda JM, Bauer DC, Barrett-Connor E, Ensrud K, Lau EM, Orwoll ES. Factors associated with the lumbar spine and proximal femur bone mineral density in older men. Osteoporos Int 16: 1525–1537, 2005.
7. Cauley JA, Lui L-Y, Barnes D, Ensrud K, Zmuda J, Hillier T, Hochberg M, Schwartz A, Yaffe K, Cummings S, Newman A. Successful skeletal aging: A marker of low fracture risk and longevity. The Study of Osteoporotic Fractures (SOF). J Bone Miner Res 24: 134–143, 2009.
8. Cauley JA, Lui LY, Ensrud KE, Zmuda JM, Stone KL, Hochberg MC, Cummings SR. Bone mineral density and the risk of incident nonspinal fractures in black and white women. JAMA 293: 2102–2108, 2005.
9. Ciccotti M, Kerlan R, Perry J, Pink M. An electromyographic analysis of the knee during functional activities I: The normal profile. Am J Sports Med 22: 645–650, 1994.
10. Conroy BP, Kraemer WJ, Maresh CM, Fleck SJ, Stone MH, Fry AC, Miller PD, Dalsky GP. Bone-mineral density in elite junior Olympic weightlifters. Med Sci Sports Exerc 25: 1103–1109, 1993.
11. Dawson-Hughes B, Jacques P, Shipp C. Dietary calcium intake and bone loss from the spine in healthy postmenopausal women. Am J Clin Nutr 46: 685–687, 1987.
12. Eickhoff JA, Molczyk L, Gallagher JC, De Jong S. Influence of isotonic, isometric and isokinetic muscle strength on bone mineral density of the spine and femur in young women. Bone Miner 20: 201–209, 1993.
13. Jacobson PC, Beaver W, Grubb SA, Taft TN, Talmage RV. Bone density in women: College athletes and older athletic women. J Orthop Res 2: 328–332, 1984.
14. Kohrt WM, Bloomfield SA, Little KD, Nelson ME, Yingling VR. American College of Sports Medicine Position Stand: Physical activity and bone health. Med Sci Sports Exerc 36: 1985–1996, 2004.
15. Kriska A, Knowler W, LaPorte R, Drash A, Wing R, Blair S, Bennett P, Kuller L. Development of questionnaire to examine relationship of physical activity and diabetes in Pima Indians. Diabetes Care 13: 401–411, 1990.
16. Krolner B, Pors Nielsen S. Bone mineral content of the lumbar spine in normal and osteoporotic women: cross-sectional and longitudinal studies. Clin Sci (Lond) 62: 329–336, 1982.
17. Madsen OR, Schaadt O, Bliddal H, Egsmose C, Sylvest J. Relationship between quadriceps strength and bone mineral density of the proximal tibia and distal forearm in women. J Bone Miner Res 8: 1439–1444, 1993.
18. McCrory J, Salacinski A, Hunt S, Greenspan S. Thigh muscle strength in senior athletes and healthy controls. J Strength Cond Res 23: 2430–2436, 2009.
19. Mussolino ME, Looker AC, Orwoll ES. Jogging and bone mineral density in men: Results from NHANES III. Am J Public Health 91: 1056–1059, 2001.
20. Nguyen TV, Eisman JA. Assessment of significant change in BMD: A new approach. J Bone Miner Res 15: 369–372, 2000.
21. Nordstrom P, Nordstrom G, Thorsen K, Lorentzon R. Local bone mineral density, muscle strength, and exercise in adolescent boys: A comparative study of two groups with different muscle strength and exercise levels. Calcif Tissue Int 58: 402–408, 1996.
22. Nordstrom P, Thorsen K, Bergstrom E, Lorentzon R. High bone mass and altered relationships between bone mass, muscle strength, and body constitution in adolescent boys on a high level of physical activity. Bone 19: 189–195, 1996.
23. Peterson SE, Peterson MD, Raymond G, Gilligan C, Checovich MM, Smith EL. Muscular strength and bone density with weight training in middle-aged women. Med Sci Sports Exerc 23: 499–504, 1991.
24. Pettersson U, Nordstrom P, Lorentzon R. A comparison of bone mineral density and muscle strength in young male adults with different exercise level. Calcif Tissue Int 64: 490–498, 1999.
25. Pocock N, Eisman J, Gwinn T, Sambrook P, Kelly P, Freund J, Yeates M. Muscle strength, physical fitness, and weight but not age predict femoral neck bone mass. J Bone Miner Res 4: 441–448, 1989.
26. Rector RS, Rogers R, Ruebel M, Widzer MO, Hinton PS. Lean body mass and weight-bearing activity in the prediction of bone mineral density in physically active men. J Strength Cond Res 23: 427–435, 2009.
27. Ribom E, Ljunggren O, Piehl-Aulin K, Ljunghall S, Bratteby LE, Samuelson G, Mallmin H. Muscle strength correlates with total body bone mineral density in young women but not in men. Scand J Med Sci Sports 14: 24–29, 2004.
28. Rubin C, Lanyon C. Regulation of bone formation by applied dynamic loads. J Bone Joint Surg Am 66: 397–402, 1984.
29. Segal NA, Torner JC, Yang M, Curtis JR, Felson DT, Nevitt MC. Muscle mass is more strongly related to hip bone mineral density than is quadriceps strength or lower activity level in adults over age 50 years. J Clin Densitom 11: 503–510, 2008.
30. Snow-Harter C, Marcus R. Exercise, bone mineral density, and osteoporosis. Exerc Sport Sci Rev 19: 351–388, 1991.
31. Snow CM, Shaw JM, Winters KM, Witzke KA. Long-term exercise using weighted vests prevents hip bone loss in postmenopausal women. J Gerontol A Biol Sci Med Sci 55: M489–M491, 2000.
32. Stone B, Perera S, Velez N, Zhang A, Miller M, Greenspan S. Predicting skeletal integrity in master athletes and non-athlete controls. In: Proceedings of Annual Meeting of the American Geriatrics Society. Seattle, WA: American Geriatric Society, 2007. pp. S196–197.
33. Swallow EB, Reyes D, Hopkinson NS, Man WD, Porcher R, Cetti EJ, Moore AJ, Moxham J, Polkey MI. Quadriceps strength predicts mortality in patients with moderate to severe chronic obstructive pulmonary disease. Thorax 62: 115–120, 2007.
34. Taaffe DR, Luz Villa M, Delay R, Marcus R. Maximal muscle strength of elderly women is not influenced by oestrogen status. Age Ageing 24: 329–333, 1995.
35. Taaffe DR, Pruitt L, Lewis B, Marcus R. Dynamic muscle strength as a predictor of bone mineral density in elderly women. J Sports Med Phys Fitness 35: 136–142, 1995.
36. US_Census_Bureau_website, 2010 US Census, U.D.o.C. Population Estimates Program. Washington, DC: United States Federal Government, 2010.
37. Velez N, Zhang A, Miller M, Perera S, Greenspan S. The effect of high impact exercise on skeletal integrity in master athletes. J Am Geriatr Soc 54: s193–s194, 2006.
38. Velez NF, Zhang A, Stone B, Perera S, Miller M, Greenspan SL. The effect of moderate impact exercise on skeletal integrity in master athletes. Osteoporos Int 19: 1457–1464, 2008.
39. Winter D. Kinetics: Forces and moments of force. In: Biomechanics and Motor Control of Human Movement (3rd ed.). Hoboken, NJ: John Wiley & Sons, Inc., 2005. pp. 86–117.
40. Winter DA, Sidwall HG, Hobson DA. Measurement and reduction of noise in kinematics of locomotion. J Biomech 7: 157–159, 1974.