Force platform measurements.
The athletes performed two 60-m sprint trials on an indoor track equipped with a 9.4-m-long force platform (TR-test, Finland and Kistler, Switzerland; natural frequency ≥170 Hz and sampling frequency 1 kHz) (see text and figure, Supplemental Digital Content 1 and 2, http://links.lww.com/MSS/A177, and http://links.lww.com/MSS/A178 for description of the measurements). The force platform data were analyzed for maximum instantaneous loading rate (LRmax; peak slope of the vertical force-time curve in the initial braking phase) and maximum resultant GRF in the braking (Frbrake-max) and push-off (Frpush-max) phases of contact. The parameters were determined for the dominant leg (average score of three to six contacts). Missing data on sprint performance were due to mild leg muscle strain (two subjects) and technical problems with the force platform (one subject).
After the sprint tests, the athletes performed a reactive hopping test on the same force platform. The test involved a series of two-legged rhythmical vertical hops over a period of 10 s (14–17 hops per test), keeping the legs extended and hands on the hips. The subjects were instructed to hop as high as possible while minimizing the contact time. Two successful 10-s tests were recorded for each subject, and the vertical GRF data of better test (average score of six hops) were further analyzed for hopping height, LRmax, maximum vertical GRF in the braking (Fzbrake-max) and push-off (Fzpush-max) phases of contact, and average mechanical power in the braking (Pbrake-ave) and push-off (Ppush-ave) phases. Missing values in hopping performance were due to mild leg muscle strain (two subjects), technical problems (two subjects), and unwillingness to participate owing to concern about possible back injury (one subject).
Maximal isometric voluntary contraction force of the knee extensor muscles (MVCKE) was measured in the dominant leg by a David 200 dynamometer (David Fitness and Medical Ltd., David Health Solutions, Helsinki, Finland). The subject was in a sitting position with knee and hip angles of 90° and 110°, respectively, and on a verbal command exerted maximal force for a period of approximately 4.0 s. Three trials were recorded, and the trial with the highest force value was used for the final analysis.
The cross-sectional area of lower leg muscles (MCSALL) was determined from the pQCT images taken at the midshaft level of the dominant leg. The thickness of individual knee extensor (KE) muscles (vastus lateralis (VL), vastus medialis, vastus intermedius (VIM), and rectus femoris (RF)) was measured on the dominant leg by B-mode ultrasonography (SSD-1400, Tokyo, Japan) with the subject in a supine position and the leg relaxed. The measurements for the VL, VIM, and RF were taken at 50% (for VIM two anatomic sites: under VL and RF) and for the vastus medialis at 30% of the distance between the lateral condyle of the femur and greater trochanter. The sum of the KE muscle thicknesses (MThKE) was used in the subsequent analysis as an indicator of quadriceps femoris muscle size. The interday coefficient of variation of the muscle thickness measurements using this method has previously been shown to be 2.5% (16). Earlier studies also indicate that ultrasonographically assessed muscle thickness is a good estimate for muscle mass (31).
Laboratory tests were run on fasting (12 h, approximately 8 h sleep) blood samples drawn from an antecubital vein in the morning (between 07:00 and 08:00 a.m.). Specimens were centrifuged (3500 rpm, 4°C for 10 min) and frozen at −75°C until assayed. Measurement of serum concentrations of total T, total E2, and SHBG were carried out by Immulite chemiluminescent method (Diagnostic Products Corporation, Los Angeles, CA). The samples for total T and SHBG were run in singlicate, whereas the total E2 was assayed in duplicate with the mean value used for the data analysis. The intraassay coefficients of variation were 5.5% for T (n = 25), 18.3% for E2 (n = 81), and 2.5% for SHBG (n = 30). Serum concentrations of bioavailable T (Bio T) and bioavailable E2 (Bio E2) were calculated from the concentrations of total T, total E2, and SHBG and assuming a constant albumin concentration of 4.3 g·dL−1, using the equations described by Mazer (20). In addition, the free androgen index (FAI) and free estrogen index (FEI) were calculated as total T/SHBG and total E2/SHBG, respectively. Hormone data are missing for three subjects. One subject misunderstood the instructions or was not willing to participate in the laboratory test, and the data for two subjects were excluded because of outlier data values.
The physical, training, sport-specific loading, muscle, and hormonal characteristics of the groups were compared with the one-way ANOVAs. Comparisons of the mean values of the bone parameters among the groups were made using ANCOVA with body mass and height as covariates. The Sidak method was used as the post hoc test in ANOVA and ANCOVA. Partial correlations controlling for the effect of age, body mass, and height were performed to examine the relationship between the bone properties and other measured parameters in the athletes. Stepwise multiple regression analyses were used to find out the combination of independent predictors that explained the most variance in the bone parameters. The regression analyses were conducted without the training parameters, and age, body mass, and height were entered into all the analyses. Results are given as mean ± SD and as the mean and 95% confidence intervals (CI). The statistical analyses were done by the PASW 18 statistical program. A P < 0.05 was regarded as statistically significant in all the analyses.
Physical and training characteristics.
In the athletes, age group differences were observed in body mass and height, whereas body mass index (BMI) and body fat did not differ between the groups (Table 1). Body mass, height, and BMI in the reference subjects (31–45 yr) were similar to those in the youngest athlete group (40–49 yr). In the athletes, the number of strength training hours was lower in the 60- to 69- versus 40- to 49-yr-old athletes, whereas no group differences existed for training frequency, training hours, or training years. The training frequency and training hours of the reference group were lower than those in the athlete groups.
Sprint and hopping performance.
In sprint running, maximum velocity (Vmax), Frbrake-max, and Frpush-max differed between the age groups (Table 2). LRmax was lower in the 60- to 69- versus 50- to 59-yr-old athletes but showed no consistent differences with advancing decades of age. In the hopping test, differences existed among the age groups in LRmax, Fzbrake-max, Fzpush-max, Pbrake-ave, and Ppush-ave. Hopping height ranged from 30.4 ± 4.2 cm in the youngest to 17.0 ± 4.4 cm in the oldest athlete group (P < 0.001, data not shown).
Muscle strength and structure.
The MVCKE did not differ between age groups (Table 2). The analysis of muscle structure by ultrasound showed that the age group 40–49 yr had greater MThKE than the other age groups, whereas MThKE of the 50- to 59-yr-old athletes was greater than that of the oldest athletes. Age-related differences in MCSALL from pQCT images at the midtibia were also observed (Table 2).
No significant age group differences were observed in serum concentration of total T and Bio T, whereas FAI of the oldest group was lower than that of the youngest group (Table 2). SHBG of the oldest athletes was significantly higher than that of the 40- to 49- and 60- to 69-yr-old athletes. Bio E2 and FEI did not vary between groups.
Owing to their significant correlations with the bone measures, body mass and height were controlled for in the main analyses (see Table, Supplemental Digital Content 2, for the absolute bone values). The athletes had greater values in most bone parameters compared with referents (Fig. 1). The differences were most noticeable for BMCtot of the distal tibia and BMCtot, CSAtot, and Imax of the tibial midshaft with significantly higher values (11%–48%) in the athletes in all age groups than that in reference group. The vBMDco of the tibial midshaft was similar in all groups.
The results of polar distribution of bone mineral mass of the tibial midshaft are shown in Figure 2B. The youngest and oldest athletes demonstrated higher (23%–37%) bone mass in the P-M, P (young athletes) and A-L sectors compared with referents.
When bone values were compared between the different age groups in the athletes, only vBMDtrab of the distal tibia was lower (12.3%, P < 0.05) in the oldest than youngest group (Figs. 1 and 2B). In the partial correlation analysis (controlled for body mass and height), age was negatively associated with vBMDtrab (r = −0.25, P < 0.05) and positively with CSAtot (r = 0.33, P < 0.01) of the distal tibia (Table 3). Periosteal circumference of the midshaft did not correlate significantly with age (r = 0.10, data not shown).
Association of bone parameters with loading, muscle, hormone, and training characteristics.
After controlling for age, body mass, and height, most loading-related characteristics, KE muscle thickness, and hormone concentrations correlated with the bone parameters (Table 3). None of the correlations between training and bone parameters achieved significance (partial r = −0.17 to 0.23, P ≥ 0.06, data not shown).
Multivariate regression models explained 12%–67% (on average, 47.5%) of the variance in the dependent variables (Table 4). Body mass and hopping Pbrake-ave were consistently the strongest predictors of the bone measures. Age entered the models for BMCtot, CSAtot, and BSIcomp of the distal tibia and BMCtot and CSAtot of the midtibia and explained 4%–10% of the total variance. Of the muscle parameters, MThKE had a small independent effect on BMCtot and BSIcomp at the distal site and was the only significant predictor of vBMDco at the midtibia. Some hormone parameters contributed to all the models of the distal tibia and models of Imin and Imax of the midtibia.
The sprint athletes demonstrated superior bone strength measures compared with active younger referents, which is in line with previous masters athlete studies (38,40,41). However, the present study extends previous findings by showing that at midshaft, the difference between athletes and reference subjects was most marked in bending strength in the anteroposterior plane as estimated by Imax, with even the oldest athletes having 48% higher values than reference group (Fig. 1). Our study also suggests bone adaptations to training are more evident in structural (CSAtot, Thco) than densitometric (vBMD) parameters, particularly at midtibia. Furthermore, analysis of bone mass distribution at the midshaft revealed larger cortical mass in the posteromedial, posterior (youngest group), and anterolateral directions in a subset of the youngest and oldest athlete groups compared with the reference group (Fig. 2B) that may largely contribute to the direction-specific increase in bone bending strength. Comparable direction-specific adaptations in geometry have also been reported in younger athletes (28) and physically active middle-age and older men (2).
In athletes, the average trabecular vBMD of the distal tibia decreased progressively approximately 12% from the youngest to oldest group to the level of reference group (Fig. 1). The pattern of early trabecular bone loss, beginning already in the third or fourth decade, has also been shown for masters track athletes (39,41) and for untrained people (30). Increased CSAtot of the distal tibia with age (11%), which is likely to reflect bone apposition in periosteal surface, is also noteworthy in the present and previous studies (39,41). The bone parameters of the midtibia, however, showed no significant age group differences in this study. The absence of differences in the structural parameters between athlete age groups suggests that unlike the distal tibia, shaft of tibia seems not to undergo significant aging-related periosteal apposition and endocortical resorption. Our results of bone shaft are similar to those observed in tibia and radius in male masters cyclists (39) and in playing arm radius in middle-aged female tennis players (23).
Contribution of current sport-specific loading characteristics.
Our data suggest that running speed itself is a weak predictor of bone properties, whereas certain GRF parameters have significant predictive value. Because bone remodeling can be strongly dependent on the strain rate (22), one could assume that the high rate of direct impact loading to the leg, as characterized by the ground contact (heel strike) of sprinting, provides a positive osteogenic effect. However, we found that maximum loading rate and braking force of sprinting were not associated and even showed a negative correlation with the bone parameters (Table 3). Conversely, maximum push-off force in sprinting, although smaller in magnitude than the braking forces, was positively associated with BMCtot, CSAtot, and BSIcomp at the distal tibia. The reason for the different associations between the push-off compared with braking phase GRF of running and bone properties is unclear. Studies estimating the loading of the distal tibia at slower running speeds have indicated that the peak compressive forces in the tibia (approximately 10 BW) occur at the start of the push-off phase and are produced primarily by muscle activity (32). Thus, one possibility is that in running, propulsive GRF better reflects actively induced muscle forces that can place the largest load on the skeleton (10).
However, in hopping, maximum loading rate as well as braking and push-off forces correlated positively with bone properties. The explanation for the positive association of the hopping but not sprint braking force with bone properties could be the powerful influence of active eccentric muscle forces (or less confounding influence of passive impact forces) in vertical hopping due to a “pure” forefoot landing strategy with no heel contact. If true, then eccentric muscle contraction might also be capable of accounting for part of the adaptive responses. This would seem logical because the highest attainable muscle force values are obtained during rapid eccentric contraction due to force–velocity property. However, a major finding in our biomechanical data is that mechanical power in the eccentric phase of hopping was the strongest overall predictor of bone characteristics. The high predictive value of Pbrake-ave could be interpreted to mean that the optimal stimulus for bone response is not maximum force but rather a combination of high force and high velocity that could be maximized in eccentric phase of impact exercises.
Thus, our findings point out that one of the key unanswered questions concerning training-induced bone adaptation is the relative importance of passive impact forces versus sustained forces. In the study by Nikander et al. (26), it was deduced that training producing sustained high muscle work without excessive vertical impacts (slalom skiing) is capable of producing similar or even higher osteogenic effect at the distal tibia than training characterized by very high passive impact loads (triple jump). Our study indicated that the high sustained GRF during the contact was more evident in the youngest runners (see Supplemental Digital Content 2), which is in line with the view that the high sustained loads with rapidly changing vector could be more important than high passive loads for the trabecular bone of distal tibia. In addition, we found that the trabecular density of distal tibia was the only bone measure that differed with age. This might be partially ascribed to vector differences between younger and older runners for the sustained sprint GRF, because the peak GRF was quite similar.
Influence of muscle, body size, and training characteristics.
Thigh muscles are important force/power producers in sprinting and jumping actions translating into increased GRF. In addition, bone loading during lower extremity weightlifting exercises could be related to strength level of thigh muscles. Thus, strong thighs can have an indirect influence on tibial bone loading and adaptation (28). However, isometric KE muscle force did not correlate with any of the examined bone parameters. Given that MThKE showed a positive relationship with BMCtot and BSIcomp of the distal tibia and was the only predictor of vBMDco in the multivariate analysis suggests, however, that the KE muscle group may play a significant role in tibial bone adaptations. The weak or nonsignificant associations between MVCKE and the bone parameters could be partially explained by the isometric strength test method, which is unspecific concerning the dynamic loading pattern of sprint training the subjects are accustomed to. It is noteworthy, however, that MCSALL showed no positive relationship with the bone properties.
It should be emphasized that body mass was a strong predictor of bone parameters in the regression analysis. Our finding is supported by some previous studies in younger athletes reporting that body size can explain up to one-half of the total variance in bone strength (25). Bone may not significantly benefit from increased body mass due to direct static loading (6). Instead, the effect of body mass could reflect differences in muscle mass (strength) that affects the forces exerted on bone during exercise (6).
The training background of the athletes varied in all the age groups. Some of the athletes have been participating in sport activities since childhood or adolescence, whereas some subjects had started training during adulthood or middle age. However, no association was observed between training years and bone parameters. This result agrees with previous findings in speed–power athletes (27,35) and could be interpreted to mean that the good bone status of these athletes not only is the result of impact sports participation before adulthood but can also be highly promoted by training in middle and older age (23). The current weekly training measures showed also no relationship with bone traits. This could reflect small differences in the type of loading among the athletes in the same sport, which is difficult to detect by questionnaire. Based on previous studies, it could also indicate the reaching of a ceiling (approximately 6 h·wk−1) where any additional training is ineffective (11).
In our study, estradiol levels (total E2, bio E2, and FEI) were associated negatively with CSAtot at the distal site (Table 3). In addition, FEI correlated positively with vBMDtrab and total E2 was found to have an independent effect on vBMDtrab in the multivariate analysis. These data are in line with reports in sedentary men (13) and suggest that higher E2 levels may promote trabecular BMD and possibly affect bone structure by inhibiting periosteal bone formation and/or endocortical resorption. Interestingly, serum levels of T showed an inverse correlation with some bone parameters and were independent negative predictors also in the multivariate models. Although a negative correlation between total T and areal BMD has been described in some studies (33), many other studies have reported a lack of association for untrained older men (18,36) and masters athletes (35). The explanation for our finding is unknown. However, there is a complex interrelationship between hormones, and one possibility is that in our athletes with normal testosterone levels, the true independent effect of T on bone is masked by the stronger influence of other hormones. For example, total T correlated positively with SHBG (r = 0.48, not shown), which affects the bioavailability of both E2 and T, and this could partly explain the negative relationship between trabecular BMD and total T (18). Taken together, these results suggest that in addition to biomechanical loading, sex hormone levels may have a significant influence on the conservation of tibial bone properties, especially at the distal site rich in trabecular bone.
In the present study, we had a good sample of continuously trained sprinters representing the highest national or international performance level. Theoretically, in these elite athletes, near-maximum adaptive capacity to sprint exercise has been reached, and therefore, the physiological changes reflect the aging process itself (17). The major strength of the investigation is the various outcomes that were evaluated. Not only did we evaluate the bone properties in detail, we had opportunity to investigate sport-specific GRF, muscle, and sex hormone characteristics that may explain variation in bone properties. However, the study has certain important limitations. A concern of this study is that the results of athletes were obtained with a cross-sectional design, instead of longitudinal design, and may have been affected not only by training and aging per se but also by genetic and constitutional factors. In addition, a self-selection bias, i.e., those individuals with superior bone properties are more inclined to choose an athletic lifestyle, cannot be completely ruled out. However, given that the differences were most marked in the direction-specific bone strength and structural parameters, we believe that the athlete versus nonathlete differences were mainly due to training with self-selection bias playing minor role. In support of this hypothesis, recent findings of middle-age monozygotic twin pairs (i.e., controlled for the influences of genes and childhood family environment) indicate that higher physical activity levels during adulthood increases or maintains bone quality in a site- and direction-specific manner (19). One limitation is the lack of older control groups that complicates the interpretation of the data. However, we believe that, as with defining the referents for osteoporotic persons (42), the use of carefully selected younger reference group can provide an adequate control to support the conclusions. It should be emphasized that the fact that the young referents are representative of healthy active individuals rather than sedentary population strengthens the results of the effectiveness of high-impact training on bone health during aging. Another limitation is that the hormonal data were derived from a single fasting sample. Given the inherent variability (circadial, day-to-day) of sex steroids, multiple time point measurements would have lent more accuracy to the detection of hormonal influences on bone properties.
Systematically trained masters sprinters showed greater bone strength, particularly bending rigidity in anteroposterior plane, in comparison with younger physically active reference group. The good bone strength of aging athletes seem to be related to improved geometrical structure (cross-sectional area, cortical thickness, distribution of bone mass in anteroposterior direction), whereas smaller difference was observed in bone density measures. Among athletes, current sport-specific GRF characteristics, body mass and height, KE muscle size, hormonal status, and age all had independent predictive value for bone parameters. Nevertheless, mechanical power in the eccentric phase of the hopping and body mass appeared to be the most important predictors of the bone measures. These data highlight the complexity of age- and training-related differences in bone properties and the importance of more detailed analyses of bone geometrical structure as well as predictive biomechanical and hormonal factors to better understand the mechanisms mediating bone health during aging. There is a need for additional research to know if habitual sprint training is able to maintain bone properties also at clinically relevant skeletal sites (femoral neck, lumbar spine) and does this type of intensive training involve any risks for the aging musculoskeletal system. Female masters athletes, especially those entering menopause, are important to include in the future studies.
Support for this study was provided by the Academy of Finland (250683), Ministry of Education and Culture, National Graduate School of Musculoskeletal Disorders and Biomaterials, Finnish Cultural Foundation, and Juho Vainio Foundation.
The authors thank Timo Annala, Merja Hoffren, Leena Kiviaho, Kaisa Malmberg, Tuovi Nykänen, Risto Puurtinen, Sarianna Sipilä, and Tuuli Suominen for valuable technical assistance with the data collection and analysis and all the athletes participating in this study.
All authors state that they have no conflicts of interest.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Aagaard P, Magnusson PS, Larsson B, Kjaer M, Krustrup P. Mechanical muscle function, morphology, and fiber type in lifelong trained elderly. Med Sci Sports Exerc
. 2007; 39 (11): 1989–96.
2. Bailey CA, Kukuljan S, Daly RM. Effects of lifetime loading history on cortical bone density and its distribution in middle-aged and older men. Bone
. 2010; 47 (3): 673–80.
3. Bennell KL, Malcolm SA, Khan KM, et al.. Bone mass and bone turnover in power athletes, endurance athletes, and controls: a 12-month longitudinal study. Bone
. 1997; 20 (5): 477–84.
4. Burr DB, Milgrom C, Fyhrie D, et al.. In vivo measurement of human tibial strains during vigorous activity. Bone
. 1996; 18 (5): 405–10.
5. Cheng S, Sipila S, Taaffe DR, Puolakka J, Suominen H. Change in bone mass distribution induced by hormone replacement therapy and high-impact physical exercise in post-menopausal women. Bone
. 2002; 31 (1): 126–35.
6. Frost HM. On our age-related bone loss: insights from a new paradigm. J Bone Miner Res
. 1997; 12: 1539–46.
7. Heinonen A, Kannus P, Sievanen H, et al.. Randomised controlled trial of effect of high-impact exercise on selected risk factors for osteoporotic fractures. Lancet
. 1996; 348 (9038): 1343–7.
8. Heinonen A, Oja P, Kannus P, et al.. Bone mineral density in female athletes representing sports with different loading characteristics of the skeleton. Bone
. 1995; 17 (3): 197–203.
9. Heinonen A, Sievanen H, Kyrolainen H, Perttunen J, Kannus P. Mineral mass, size, and estimated mechanical strength of triple jumpers’ lower limb. Bone
. 2001; 29 (3): 279–85.
10. Judex S, Carlson KJ. Is bone’s response to mechanical signals dominated by gravitational loading? Med Sci Sports Exerc
. 2009; 41 (11): 2037–43.
11. Karlsson MK, Magnusson H, Karlsson C, Seeman E. The duration of exercise as a regulator of bone mass. Bone
. 2001; 28 (1): 128–32.
12. Kaufman JM, Vermeulen A. The decline of androgen levels in elderly men and its clinical and therapeutic implications. Endocr Rev
. 2005; 26 (6): 833–76.
13. Khosla S, Melton LJ 3rd, Robb RA, et al.. Relationship of volumetric BMD and structural parameters at different skeletal sites to sex steroid levels in men. J Bone Miner Res
. 2005; 20 (5): 730–40.
14. Korhonen MT, Cristea A, Alen M, et al.. Aging, muscle fiber type, and contractile function in sprint-trained athletes. J Appl Physiol
. 2006; 101 (3): 906–17.
15. Korhonen MT, Mero AA, Alen M, et al.. Biomechanical and skeletal muscle determinants of maximum running speed with aging. Med Sci Sports Exerc
. 2009; 41 (4): 844–56.
16. Kubo K, Kanehisa H, Azuma K, et al.. Muscle architectural characteristics in young and elderly men and women. Int J Sports Med
. 2003; 24 (2): 125–30.
17. Lazarus NR, Harridge SD. Inherent ageing in humans: the case for studying master athletes. Scand J Med Sci Sports
. 2007; 17 (5): 461–3.
18. Legrand E, Hedde C, Gallois Y, et al.. Osteoporosis in men: a potential role for the sex hormone binding globulin. Bone
. 2001; 29 (1): 90–5.
19. Ma H, Leskinen T, Alen M, et al.. Long-term leisure time physical activity and properties of bone: a twin study. J Bone Miner Res
. 2009; 24 (8): 1427–33.
20. Mazer NA. A novel spreadsheet method for calculating the free serum concentrations of testosterone, dihydrotestosterone, estradiol, estrone and cortisol: with illustrative examples from male and female populations. Steroids
. 2009; 74 (6): 512–9.
21. Milgrom C, Finestone A, Levi Y, et al.. Do high impact exercises produce higher tibial strains than running? Br J Sports Med
. 2000; 34 (3): 195–9.
22. Mosley JR, Lanyon LE. Strain rate as a controlling influence on adaptive modeling in response to dynamic loading of the ulna in growing male rats. Bone
. 1998; 23 (4): 313–8.
23. Nara-Ashizawa N, Liu LJ, Higuchi T, et al.. Paradoxical adaptation of mature radius to unilateral use in tennis playing. Bone
. 2002; 30 (4): 619–23.
24. Nichols JF, Rauh MJ. Longitudinal changes in bone mineral density in male master cyclists and nonathletes. J Strength Cond Res
. 2011; 25 (3): 727–34.
25. Nikander R, Sievanen H, Heinonen A, Kannus P. Femoral neck structure in adult female athletes subjected to different loading modalities. J Bone Miner Res
. 2005; 20 (3): 520–8.
26. Nikander R, Sievanen H, Heinonen A, Karstila T, Kannus P. Load-specific differences in the structure of femoral neck and tibia between world-class moguls skiers and slalom skiers. Scand J Med Sci Sports
. 2008; 18: 145–53.
27. Nowak A, Straburzynska-Lupa A, Kusy K, et al.. Bone mineral density and bone turnover in male masters athletes aged 40–64. Aging Male
. 2010; 13 (2): 133–41.
28. Rantalainen T, Nikander R, Heinonen A, Suominen H, Sievanen H. Direction-specific diaphyseal geometry and mineral mass distribution of tibia and fibula: a pQCT study of female athletes representing different exercise loading types. Calcif Tissue Int
. 2010; 86 (6): 447–54.
29. Riggs BL, Khosla S, Melton LJ 3rd. Sex steroids and the construction and conservation of the adult skeleton. Endocr Rev
. 2002; 23 (3): 279–302.
30. Riggs BL, Melton LJ, Robb RA, et al.. A population-based assessment of rates of bone loss at multiple skeletal sites: evidence for substantial trabecular bone loss in young adult women and men. J Bone Miner Res
. 2008; 23 (2): 205–14.
31. Sanada K, Kearns CF, Midorikawa T, Abe T. Prediction and validation of total and regional skeletal muscle mass by ultrasound in Japanese adults. Eur J Appl Physiol
. 2006; 96 (1): 24–31.
32. Scott SH, Winter DA. Internal forces of chronic running injury sites. Med Sci Sports Exerc
. 1990; 22 (3): 357–69.
33. Slemenda CW, Longcope C, Zhou L, Hui SL, Peacock M, Johnston CC. Sex steroids and bone mass in older men. Positive associations with serum estrogens and negative associations with androgens. J Clin Invest
. 1997; 100 (7): 1755–9.
34. Suominen H. Bone mineral density and long term exercise. An overview of cross-sectional athlete studies. Sports Med
. 1993; 16 (5): 316–30.
35. Suominen H, Rahkila P. Bone mineral density of the calcaneus in 70- to 81-yr-old male athletes and a population sample. Med Sci Sports Exerc
. 1991; 23 (11): 1227–33.
36. Szulc P, Munoz F, Claustrat B, et al.. Bioavailable estradiol may be an important determinant of osteoporosis in men: the MINOS study. J Clin Endocrinol Metab
. 2001; 86 (1): 192–9.
37. Vainionpaa A, Korpelainen R, Leppaluoto J, Jamsa T. Effects of high-impact exercise on bone mineral density: a randomized controlled trial in premenopausal women. Osteoporos Int
. 2005; 16 (2): 191–7.
38. Welch JM, Rosen CJ. Older women track and field athletes have enhanced calcaneal stiffness. Osteoporos Int
. 2005; 16 (8): 871–8.
39. Wilks DC, Gilliver SF, Rittweger J. Forearm and tibial bone measures of distance- and sprint-trained master cyclists. Med Sci Sports Exerc
. 2009; 41 (3): 566–73.
40. Wilks DC, Winwood K, Gilliver SF, et al.. Bone mass and geometry of the tibia and the radius of master sprinters, middle and long distance runners, race-walkers and sedentary control participants: a pQCT study. Bone
. 2009; 45 (1): 91–7.
41. Wilks DC, Winwood K, Gilliver SF, et al.. Age-dependency in bone mass and geometry: a pQCT study on male and female master sprinters, middle and long distance runners, race-walkers and sedentary people. J Musculoskelet Neuronal Interact
. 2009; 9 (4): 236–46.
42. World Health Organization. Assessment of Fracture Risk and Its Application to Screening for Postmenopausal Osteoporosis. Technical Report Series 843
. Geneva: World Health Organization; 1994. p. 129.
GROUND REACTION FORCE; HIGH IMPACT; MASTERS ATHLETES; MECHANICAL LOADING; OSTEOPOROSIS; pQCT
Supplemental Digital Content
©2012The American College of Sports Medicine