Cycling is probably the most popular and certainly the oldest sport in machine-aided locomotion (27). One reason for its attractiveness lies in the greater speeds achievable compared with unaided means of locomotion, for example, running. Accordingly, trauma and subsequently fractures are common. Finch et al. (11) reported that cycling was one of the main reasons for sports injury presentation (10%) in Australian emergency departments. More importantly, cycling was found to produce some of the most severe injuries, of which ∼70% have been reported to occur in the upper and the lower limbs. The fracture prevalence amounted to 19% in adults (11), emphasizing the relevance for cyclists to achieve and maintain good upper and lower limb bone strength.
It is generally accepted that bones adapt to mechanical stimuli (12,22). Experimental evidence suggests that in particular bone strains of high magnitude and rate along with those from uncommon directions have the greatest osteogenic effects (39). As an apparent consequence, it has been repeatedly found that athletes, such as jumpers, volleyball players, and hurdlers, have stronger bones than controls at their loaded skeletal sites (18,20,30). Such associations are less clear in athletes participating within disciplines where strain rates or strains from uncommon directions play a minor role, as in cyclists. In fact, studies suggest that both young elite and master road cyclists have lower or comparable spine, hip, or femoral neck areal bone mineral density (aBMD) compared with controls (18,29,30,37,40). One might therefore conclude that the elevated risk of fractures in cyclists is partly attributable to a reduction in bone strength measures.
However, according to the mechanostat theory, bones adapt to the strain magnitude (12). Consequently, cycling should act as an osteogenic stimulus, particularly at faster speeds when the limbs are exposed to high muscle forces that possibly result in large bone strains. To the best of our knowledge, studies investigating bone characteristics of cyclists have involved distance riders, however, not sprint cyclists. Additionally, we are not aware of a study that has assessed the tibia, one of the predominantly loaded skeletal sites by peripheral quantitative computed tomography (pQCT). Most studies have assessed femoral neck or lumbar spine bone measures by dual x-ray absorptiometry (DXA) (18,29,30,37,40). Compared with pQCT, DXA has the limitation of estimating a three-dimensional bone structure by two dimensional measures, that is, the bone mineral content (BMC) and the projection area (1). Furthermore, trabecular and cortical bone tissue cannot be distinguished (21). The assessment of tibial bone measures of master sprint and distance-trained cyclists by pQCT therefore provides important and novel data.
On the assumption that the largest internal forces that our limb bones are subjected to arise from muscle forces (26,33) and that these peak muscle forces translate into peak strains(1, we hypothesized, first, in accordance with the mechanostat theory, that the osteogenic effect of muscle forces generated by competitive master cyclists will lead to an adaptation in bone mass and strength in spite of negligible strain rates and strains from uncommon directions. Second, we hypothesized that this postulated osteogenic effect would be affected only to a small degree by age. We tested these hypotheses by studying master athletes because in general they have been training vigorously and competing for many years.
Athletes were competitive male master cyclists or member of the Great Britain squad, recruited and tested at the World Master Track Championships in Manchester in September 2006. They included world record holders and several gold medalists in World Master Track Championships held between 2002 and 2006. Controls were members of the local university of the third age or employees of the Manchester Metropolitan University. They were mentally active, as evidenced by occupation or educational participation, but took little or no physical activity: less than 2 h of endurance exercise per week, no exhaustive and no resistive exercise. This was assessed by a telephone conversation along with a personal interview and a questionnaire. All participants gave written informed consent to participate in the study, which had been approved by the local ethics committee at Manchester Metropolitan University.
Training and competition history
Athletes completed a detailed questionnaire on their current and previous training and on their competition level. They were asked about their training sessions per week, years of active participation within organized sport and their disciplines, and starting age of competitive cycling as well as complementary running and resistance training. Resistance training was classified as training with heavy weights to increase maximum strength.
Anthropometrical and strength measurements.
Body mass and height of each participant was measured. Grip strength was assessed by a standard grip strength-meter. For familiarization with the testing equipment, one practice trial before testing was performed with both hands.
Bone mass and structural bone measures were assessed by peripheral quantitative computed tomography (pQCT; XCT2000; STRATEC Medizintechnik GmbH, Pforzheim, Germany). pQCT is a densitometry technique that offers the important advantages of conveniently assessing volumetric bone mineral density (BMD) and transversal geometric bone characteristics as well as distinguishing between cortical and trabecular bone (4,10). The latter in particular is of major importance because trabecular and cortical BMD (crtBMD) yield information about different bone characteristics: the former represents a correlate of the structural stiffness and strength of the trabecular network (e.g., compressive strength) (9), whereas cortical volumetric BMD provides information about the intrinsic stiffness of the solid bone substance (10). In addition, a previous study has observed that the greatest adaptations to exercise relate to the diaphyseal bone structure (15). Furthermore, pQCT measures provide information about various components that contribute to whole bone strength. The BMC and the cross-sectional area inform about the compressive strength of the bone on the organ level because the compressive strength depends on both measures (14). Additionally, the assessment of structural bone measures allows the calculation of the polar moment of resistance (RPol), which is a surrogate of bone strength in torsion and bending. The RPol, as used in this and most other pQCT-studies, is density weighted to consider both material and structural measures. It is calculated as:
where a is the area of the pixel, rc is the voxel position from the center, rmax is the maximal distance of the voxel to the bone center, crtBMDm is the measured crtBMD, and crtBMDn is the typical physiological crtBMD according to the manufacturer (1200 mg·cm−3; XCT2000 operator's manual; STRATEC Medizintechnik GmbH).
Following our standard procedures (33), scans of the tibia epiphysis, metaphysis, and diaphysis were taken at 4%, 14%, and 38% of the tibia length from its distal end. Forearm scans of the epiphysis and the diaphysis were taken at 4% and 60% of the ulna length from its distal end. The right side was scanned unless a fracture had been incurred on this side in the past. Daily phantom scans were performed according to the manufacturer's instructions for quality assurance. Only scans of high quality were used for the analysis (in total, one sprinter radius pQCT scan was not considered due to a movement artifact). Image analyses were performed using the XCT2000 integrated software version 5.40 D. Trabecular and cortical bone measures were determined using a segmentation threshold of 180 and 650 mg·cm−3, respectively. At the epiphysis, trabecular volumetric BMD and BMC (trbBMD and trbBMC, respectively) along with total cross-sectional area (totA) were analyzed. At both the metaphysis (tibia only) and the diaphysis, the following measures were analyzed: cortical BMD (crtBMD; corrected for partial volume effects as described in (34)), total BMC (totBMC) and cortical cross-sectional area (crtA), and RPol as well as endocortical and periosteal circumferences (endoC and periC, respectively).
Data processing and statistical analysis
Descriptive summary statistics were calculated using means and SD or, in the case of figures, ratios of the athletes' and controls' bone measures along with SEM. Differences in both anthropometric variables and bone measures between the groups (sprint cyclists, distance cyclists, and controls) were examined by ANOVA. A simple model was used, where bone measures were the dependent variables and group affiliation was the independent variable. This was because a more complex regression model using both resistance training and age as independent variables did not detect substantial influences of either independent variable on the bone measures. Simple contrasts were used within the ANOVA model to assess the nature of group differences.
To assess the resistance-training effects more precisely, a Student's t-test was carried out that compared bone measures of athletes who participated in resistance training and those who did not. This analysis was carried out for sprinters only because the sample size of distance cyclists participating in resistance training was too small to perform a meaningful statistical analysis and all controls were physically inactive.
To test for age relationships of bone measures, correlation analyses of reported bone measures were performed and Pearson's correlation coefficient (R) was obtained. Because we had a primary hypothesis, two-tailed tests were used for bone measures. One-tailed tests were applied for all other variables, for example anthropometry and handgrip force. For the regression, linear models were chosen because no nonlinear model increased R 2 by more than 0.03 (35). Other models tried were logarithmic, inverse, and polynomial. The regression equation was used to calculate annual percentage changes of bone measures based on the mean values observed at 35 yr. Significance was assumed if P < 0.05. Statistical analysis was performed using SPSS software for Mac OS X (SPSS Inc.®, Chicago, IL).
In total, 103 participants aged between 30 and 82 yr were included in the study. We aimed to have similar group sizes for both distance and sprint cyclists to that of the control group (n = 32), which was tested before the championships. Therefore, 71 athletes were included in the study. After the testing, athletes were classified as either sprint cyclists (all disciplines involving sprints, i.e., 200-m sprint, 500- and 750-m time trial, 1000 m, and points race) or distance cyclists (5000 m, 15.5 km, scratch race, and pursuit), depending on their self-rated best discipline. Fifty-two cyclists were categorized as sprinters and 19 as distance riders (Table 1). In spite of the lower sample size of distance riders, the statistical power of their tibial RPol, as the bone measure of greatest interest, amounted still to 59% at an alpha error level of 5%; for sprint cyclists, it amounted to 95% (8).
There were no group differences in age, height, and body mass, except that distance riders were older than sprinters (57 ± 10 and 50 ± 13 yr; P = 0.021), and distance athletes had a lower body mass than controls (74 ± 9 and 81 ± 14 kg; P = 0.043; Table 1). Ulna length, tibia length, and body mass were not correlated with age and were similar among groups, suggesting that body dimensions were unrelated to age and comparable among groups. Mean handgrip forces were 527 ± 77, 466 ± 64, and 465 ± 66 N in sprinters, distance riders, and controls, respectively (distance vs sprint cyclists: P = 0.006), being negatively correlated with age in sprinters (P = 0.001; R = −0.495; Table 1). Crank length was not correlated with age in either sprint or distance cyclists.
The starting age of a comparable training regime to the current one was on average 26 ± 15 yr in distance cyclists and 29 ± 16 yr in sprinters (P = 0.17). Six out of the 15 distance riders and 28 out of the 48 sprint cyclists undertook resistance training at the time of the study, with most of them strengthening both the upper and the lower body (Table 1). Few cyclists went running as a complementary training (nine sprinters and one distance cyclist), and all controls were physically inactive.
Group Differences of Bone Measures
The BMC, the cortical area, and the RPol of both the metaphysis and the diaphysis were 9-13% greater in sprint cyclists than in controls (all P values ≤0.002). Furthermore, the sprinters' diaphyseal periosteal circumference was 4% larger than the controls' (P = 0.005; Table 2, Fig. 1). Differences in tibia diaphyseal totBMC, crtA, and RPol of distance athletes compared with the control group amounted to 7-10% and to 3% for periC (P = 0.034, 0.031, 0.057, and 0.067). Sprint and distance athlete versus control group differences were detected neither for shaft crtBMD nor for epiphyseal bone measures (Table 2, Fig. 1).
Radius diaphyseal totBMC, crtA, and RPol values were 8-13% greater in sprinters compared with the controls, and the periC was 4% larger (all P values ≤ 0.016; Table 2, Fig. 2). The sprinters' epiphyseal trbBMC and totA values were 10% and 8% larger, respectively (P = 0.050 and 0.036; Table 2, Fig. 2). For these bone measures, no significant group differences were detected between distance athletes and controls (P > 0.40; Table 2, Fig. 2). No crtBMD group differences were found between any cycling group and the controls (Table 2).
The tibia diaphyseal RPol, crtA, and periC were 10%, 6.5%, and 3.6% larger, respectively, in 27 sprinters who participated in lower body resistance training compared with 21 one sprinters who did not (P = 0.040, 0.046, and 0.035). Diaphyseal totBMC was 5.5% larger in the resistance-training group, although the differences were not statistically significant (P = 0.081). No other resistance-training group differences were observed in the tibia, and none were observed in the radius (the latter with respect to upper body resistance training).
Age Dependencies of Bone Measures
In the athletes' tibiae, no statistically significant correlations were found between any of the analyzed epiphyseal, metaphyseal, and diaphyseal bone measures and age (except for the sprinters' metaphyseal crtBMD, where P < 0.001 and R = −0.56). Among controls, few moderate correlations between tibial bone measures and age were observed: correlations were positive for epiphyseal totA and metaphyseal endoC and negative for metaphyseal and diaphyseal crtBMD (P = 0.038, 0.030, 0.004, and 0.024, respectively; R 2 always ≤0.25).
In the radius, few correlations of bone measures and age were found for both sprinters and controls (Table 3): sprinters showed negative age correlations for epiphyseal BMD and mass (−0.5% per year in trbBMD and trbBMC, P = 0.004 and 0.006, respectively). In addition, both the controls' and the sprinters' crtBMD values were negatively correlated with age (−0.08% per year, P = 0.014 and −0.07% per year, P = 0.018; Table 3). When combining all of the data for cyclists and controls, in the absence of any systematic group differences, significant age correlations were observed only for endoC (P = 0.019) and trbBMD (P < 0.001; Fig. 3).
It was the primary aim of this study to compare both tibia and radius bone measures between sprint- and distance-trained cyclists and physically inactive controls. A secondary aim was to investigate possible correlations of those bone measures and age. In concordance with our hypothesis, we found sprint cyclists to have the greatest bone strength surrogates in both the radius and the tibia, followed by distance cyclists and controls. This result is remarkable in two ways: first, it clearly shows that cyclists, even in older age, do not have reduced BMD, as has been suggested formerly (29). Second, the increased forearm and tibial bone strength measures found particularly in sprint cyclists demonstrate that exercise-specific adaptations of bone can occur without impact forces, that is, loading patterns that involve a large rate of force development and hence large strain rates.
Greatest bone measure group differences were found in the diaphyses of both the tibia and the radius. For example, the sprinters' radius and tibia diaphyseal cross-sectional area, BMC, and RPol were 8-13% greater and their periosteal circumferences were 4% larger than the control groups'. Those bone strength measures were 7-10% larger in the tibia of distance cyclists compared with the controls. In contrast, neither tibia epiphyseal bone measures of sprint and distance cyclists nor radius epiphyseal bone measures of distance cyclists were found to be different to the controls' measures. It appears, therefore, that cycling-specific bone adaptations are particularly prominent in the diaphysis. This observation is confirmed by previous pQCT studies on athletes that report exercise having a great effect upon the diaphysis (6,15,19,23,36) and affecting the diaphysis rather than the epiphysis (19).
We found both radius diaphysis and epiphysis of sprint but not distance cyclists to have greater measures of strength against torsion, bending, and compression than the controls. This interesting finding suggests exercise-specific adaptations during sprint but not necessarily distance cycling or cycling-related training. Research on muscular forces in cyclists has focused mainly on the legs, and to the best of our knowledge, no studies exist on upper extremity forces in cyclists. It is obvious, however, that at the start of a sprint race, for which a cyclist is unseated, large arm forces are a prerequisite for successful performance. These large forces can then be used to counteract forces produced by the legs that are generated for propulsion. Studies on young elite track sprinters suggest peak pedal forces to be approximately 1.6 times body weight (7). The produced pedal force, which exceeds the body weight, needs to be counteracted by the muscular system to prevent the rider being pushed off the bicycle. Thus, a cyclist with a body mass of 80 kg or an equivalent body weight of 785 N would need to counteract 471 N when generating a peak pedal force of 1255 N (ignoring all external forces, e.g., friction). The proportion taken by the arms is not known but is likely to be a significant amount. Cycle manufacturers are aware of large forces acting on the handlebars of sprint bikes and use a resilient material for example steel for sprint bike handlebars and not aluminum or carbon fiber as used for handlebars of distance bikes.
The possibility that large arm muscle forces are involved in sprint cycling is furthermore reflected in our study by the approximately 10% larger observed grip strength in sprint cyclists compared with distance cyclists. In contrast, quick and forceful starting conditions are unimportant to distance cyclists who had a similar grip strength than the controls. Accordingly, distance cyclists also had radius bone measures comparable to the controls (Table 2). In that sense, our findings concur with the published literature that reported comparable aBMD values in the radius of road or distance cyclists and controls (18,37).
Strain, strain rate, and bone adaptation
Numerous authors agree that strain rate and magnitude both play a crucial role for the osteogenic response to loading (12,22,39). It has been argued that strain rate is paramount to bone canalicular fluid flow, which by many authors is believed to be an important mediator in the process of mechanotransduction (2,31,38). Admittedly, little is known regarding peak musculoskeletal forces during sprint cycling, and to the best of our knowledge no study has investigated bone strains and strain rates during sprint cycling. However, it is well known that during cycling, crank torque depicts a rather smooth time course, almost approaching a sinusoidal function (32). Hence, the strain rate is expected to be very low during cycling compared with running, where vertical ground reaction forces have been reported to be three to five times the body weight, and strain measurement studies suggest that the peak tibial strain lasts less than 0.2 s (5,17). Therefore, at first glance it may appear surprising that tibial surrogates of bone strength of the sprint cyclists reached values that could have been expected for runners and are clearly larger than those of the controls. For example, the average tibia diaphyseal RPol was 13% larger in sprint cyclists compared with the controls (Fig. 1). It has to be considered though that the cyclists' bones will adapt also to other activities that the cyclists may perform. For instance, a comparison of tibial bone measures between sprint cyclists who did and did not perform lower body resistance training revealed resistance-training participation to be associated with 5-8% larger diaphyseal bone measures, such as the RPol, the cross-sectional area, and the periosteal circumference. Interestingly, tibial bone measures of sprinters, who did not participate in resistance training, were comparable to those observed in the distance cyclists of this study. A general osteogenic effect of resistance training has been proposed previously (18,19). Because less than 30% of distance cyclists undertook resistance training, the sample size was too small to perform a meaningful statistical analysis. We therefore assumed that resistance training had little influence in this group and that the distance cyclists' larger tibia diaphyseal bone measures compared with the controls can be contributed to cycling or cycling-specific training. Hence, resistance training may enhance the osteogenic effect of cycling or cycling-specific training, but group differences in the tibia diaphysis of distance cyclists in particular cannot be explained by resistance training alone.
In contrast to the tibia, we found no differences in any radius measures between sprinters who did or did not participate in upper body resistance training. Therefore, we suggest that the largest strain magnitudes or strain rates in the radius occur during cycling or cycling-specific training (e.g., standing starts) rather than during resistance training. These are expected to lead to the large group differences of 9-14% in measures of both the radius epiphysis and the diaphysis of sprinters and controls. As discussed above, the considerable handgrip forces of sprinters compared with both distance cyclists and controls support this notion. It is likely that these postulated peak strains during the sprint cycle start occur in the absence of large strain rates.
Age dependencies of the tibia and the radius
Maximal cycling speed declines as a function of age. The required power at similar terrain and calm conditions increases with cycling speed (24), implying that the mechanical power output declines with age. On the other hand, cadence has an optimum (7,25) and crank length is unaffected by age (see Results). This implies that reduced pedal forces, rather than changes in angular velocity or mechanical advantage are responsible for the age-related decline in cycling speed. Nevertheless, we observed no correlations between age and any of the analyzed epiphyseal and diaphyseal measures in the tibia of athletes and only a few in the radius of sprint cyclists. In the latter, correlations followed a trend that was observed in the control group (Table 3). This is remarkable and emphasizes that a sport that has often been associated with poor bone strength surrogates is in fact related to average and even above-average bone measures up to old age.
Noteworthy, however, is that among sprint riders, radius trbBMD and trbBMC were predicted to decline by 0.5% per year. This may be due to a decline in peak forces during the start, reflected in slower acceleration, and is supported by the negative correlation of handgrip force and age in sprinters but not in distance cyclists. The positive associations of the endocortical circumferences and age in all groups are likely to be contributed to an age-related loss of muscle mass and strength in the upper limbs (28) (Fig. 3). This should result in lower forces acting on the bones at older as opposed to younger age. According to Frost's (12) generally accepted mechanostat theory, declining forces are expected to result in increased remodeling that usually causes bone loss close to the marrow, for example, at the endocortical circumferences of the diaphyses (13).
Nevertheless, the general lack of observed age correlations suggests that the assessed bone measures are equally well preserved with age in cyclists compared with controls, in spite of cycling being an exercise modality that is currently not advocated for osteogenic effects. These findings are in agreement with Beshgetoor et al. (3), who observed preservation of femur aBMD in female master cyclists (∼50 yr) and runners from baseline measurement to 18-month follow-up. Therefore, our results suggest that master cyclists do not suffer from bone deficiency and that the high fracture prevalence of cyclists is mostly attributable to trauma alone.
Summary and conclusion
In conclusion, although cycling is thought to be associated with low BMD and bone strength surrogates, we found those measures to be average or above average in both the tibia and the radius. In addition, very few bone measures were correlated with age, suggesting that cyclists maintain good surrogates of bone strength of the tibia in particular into old age.
The authors thank all participants for their interest to take part in this study. The authors appreciate the great support of the European Cycling Union and organizers of the Masters Track Championships 2006 in Manchester. Special thanks to Dr. Steve Peters for raising the idea for this study. Moreover, the authors would like to thank James Hartshorn, Manchester Metropolitan University, for critically reading the manuscript and Mark Chatfield, MRC Human Nutrition Research, for his statistical advice. The results of the present study do not constitute endorsement by ACSM.
1. Augat P, Gordon CL, Lang TF, Iida H, Genant HK. Accuracy of cortical and trabecular bone measurements with peripheral quantitative computed tomography (pQCT). Phys Med Biol
2. Bacabac RG, Smit TH, Mullender MG, Dijcks SJ, Van Loon JJ, Klein-Nulend J. Nitric oxide production by bone cells is fluid shear stress rate dependent. Biochem Biophys Res Commun
3. Beshgetoor D, Nichols JF, Rego I. Effect of training mode and calcium intake on bone mineral density in female master cyclist, runners, and non-athletes. Int J Sport Nutr Exerc Metab
4. Braun MJ, Meta MD, Schneider P, Reiners C. Clinical evaluation of a high-resolution new peripheral quantitative computerized tomography (pQCT) scanner for the bone densitometry at the lower limbs. Phys Med Biol
5. Burr DB, Milgrom C, Fyhrie D, et al. In vivo measurement of human tibial strains during vigorous activity. Bone
6. Daly RM, Bass SL. Lifetime sport and leisure activity participation is associated with greater bone size, quality and strength in older men. Osteoporos Int
7. Dorel S, Hautier CA, Rambaud O, et al. Torque and power-velocity relationships in cycling: relevance to track sprint performance in world-class cyclists. Int J Sports Med
9. Ebbesen EN, Thomsen JS, Mosekilde L. Non-destructive determination of iliac crest cancellous bone strength by pQCT. Bone
10. Ferretti J. Peripheral quantitative computed tomography for evaluating structural and mechanical properties of small bone. In: An YH, Draughn RA, editors. Mechanical Testing of Bone and the Bone-Implant Interface
. Boca Raton: CRS Press; 2000. p. 385-402.
11. Finch C, Valuri G, Ozanne-Smith J. Sport and active recreation injuries in Australia: evidence from emergency department presentations. Br J Sports Med
12. Frost HM. Bone "mass" and the "mechanostat": a proposal. Anat Rec
13. Frost HM. From Wolff's law to the Utah paradigm: insights about bone physiology and its clinical applications. Anat Rec
14. Gordon J. The Science of Structures and Materials
. New York: Scientific American Library, 1988. p. 1-217.
15. Haapasalo H, Kontulainen S, Sievanen H, Kannus P, Jarvinen M, Vuori I. Exercise-induced bone gain is due to enlargement in bone size without a change in volumetric bone density: a peripheral quantitative computed tomography study of the upper arms of male tennis players. Bone
16. Hamill J. Variations in ground reaction force parameters at different running speeds. Hum Mov Sci
17. Hamill J, Knutzen KM. Biomechanical Basis of Human Movement
. 2nd ed. Philadelphia: Lippincott Williams & Wilkins; 2003. p. 337-79.
18. Heinonen A, Oja P, Kannus P, Sievanen H, Manttari A, Vuori I. Bone mineral density of female athletes in different sports. Bone Miner
19. Heinonen A, Sievanen H, Kannus P, Oja P, Vuori I. Site-specific skeletal response to long-term weight training seems to be attributable to principal loading modality: a pQCT study of female weightlifters. Calcif Tissue Int
20. Heinonen A, Sievanen H, Kyrolainen H, Perttunen J, Kannus P. Mineral mass, size, and estimated mechanical strength of triple jumpers' lower limb. Bone
21. Khan K, McKay H, Kannus P, Bailey D, Wark J, Bennell K. Physical Activity and Bone Health
. Champaign: Human Kinetics; 2001. p. 39f.
22. Lanyon LE, Rubin CT. Static vs. dynamic loads as an influence on bone remodelling. J Biomech
23. Liu L, Maruno R, Mashimo T, et al. Effects of physical training on cortical bone at midtibia assessed by peripheral QCT. J Appl Physiol
24. Martin J, Cobb J. Body position and aerodynamics. In: Jeukendrup AE, editor. High-Performance Cycling
. Champaign: Human Kinetics; 2002. p. 105f.
25. Martin JC, Spirduso WW. Determinants of maximal cycling power: crank length, pedaling rate and pedal speed. Eur J Appl Physiol
26. Martin RB, Burr DB, Sharkey NA. Skeletal Tissue Mechanics
. New York: Springer; 1998. p. 225-71.
27. Minetti AE, Pinkerton J, Zamparo P. From bipedalism to bicyclism: evolution in energetics and biomechanics of historic bicycles. Proc Biol Sci
28. Narici MV, Bordini M, Cerretelli P. Effect of aging on human adductor pollicis muscle function. J Appl Physiol
29. Nichols JF, Palmer JE, Levy SS. Low bone mineral density in highly trained male master cyclists. Osteoporos Int
30. Nikander R, Sievanen H, Heinonen A, Kannus P. Femoral neck structure in adult female athletes subjected to different loading modalities. J Bone Miner Res
31. Owan I, Burr DB, Turner CH, et al. Mechanotransduction in bone: osteoblasts are more responsive to fluid forces than mechanical strain. Am J Physiol
. 1997;273(3 Pt 1):C810-5.
32. Pons DJ, Vaughan CL. Mechanics of Cycling. In: Vaughan CL. Biomechanics of Sport
: Boca Raton (FL): CRC Press; 1989. p. 289-315.
33. Rittweger J, Beller G, Ehrig J, et al. Bone-muscle strength indices for the human lower leg. Bone
34. Rittweger J, Michaelis I, Giehl M, Wusecke P, Felsenberg D. Adjusting for the partial volume effect in cortical bone analyses of pQCT images. J Musculoskelet Neuronal Interact
35. Runge M, Rittweger J, Russo CR, Schiessl H, Felsenberg D. Is muscle power output a key factor in the age-related decline in physical performance? A comparison of muscle cross section, chair-rising test and jumping power. Clin Physiol Funct Imaging
36. Sone T, Imai Y, Joo YI, Onodera S, Tomomitsu T, Fukunaga M. Side-to-side differences in cortical bone mineral density of tibiae in young male athletes. Bone
37. Stewart AD, Hannan J. Total and regional bone density in male runners, cyclists, and controls. Med Sci Sports Exerc
38. Turner CH. Functional determinants of bone structure: beyond Wolff's law of bone transformation. Bone
39. Turner CH. Three rules for bone adaptation to mechanical stimuli. Bone
40. Warner SE, Shaw JM, Dalsky GP. Bone mineral density of competitive male mountain and road cyclists. Bone
Keywords:©2009The American College of Sports Medicine
ATHLETES; BONE MINERAL DENSITY; BONE MASS; pQCT