University of Southern California Dissertations and Theses

Quantifying musculoskeletal load and adaptation: Biomechanical consideration

(USC Thesis Other)

Quantifying musculoskeletal load and adaptation: Biomechanical consideration

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content QUANTIFYING MUSCULOSKELETAL LOAD AND ADAPTATION: BIOMECHANICAL CONSIDERATION by Man-Ying Wang A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (BIOKINESIOLOGY) August 2002 Copyright 2002 Man-Ying Wang R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. UMI Number: 3094380 UMI UMI Microform 3094380 Copyright 2003 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES, CALIFORNIA 90089-1695 This dissertation, written by MAN-YING W ANG under the direction o f h e r dissertation committee, and approved by all its members, has been presented to and accepted by the Director of Graduate and Professional Programs, in partial fulfillment of the requirements for the degree o f D O C T O R O F P H IL O S O P H Y Director Date A u g u s t 6 , 2 0 0 2 Dissertatior/Committee. Chair R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. ACKNOWLEDGEMENTS This dissertation could not have been completed without the generous and competent help of many people. My wonderful advisor, Dr. George Salem, has guided me through my years to become an independent researcher. He is an outstanding mentor, not only to my academic work, but my life at USC. Dr. Salem is always so patient, supportive, and approachable. I am blessed to have him as my advisor. The idea o f this dissertation originated from a project of Dr. Robert Wiswell’s class, Physiological Correlates of Therapeutic Exercise. The project was then encouraged by Dr. Wiswell and Dr. Salem to continue as my dissertation project. It was a great experience designing and conducting this study which is an combination of my academic work and research experience. I am very indebted to Dr. Wiswell and Dr. Salem. In addition, I want to especially thank Dr. Kim O ’Connor for her excellent course, Scientific Basis of Skeletal Adaptation. With her guidance step by step, the course brought me to the world of “mechanical load and bone adaptation”, which lead to the birth of this project. I also want to thank Dr. Victoria Jaque for encouraging me to pursue a Ph.D. degree and serving as my committee member for both my Master’s thesis and this doctoral dissertation. I truly appreciated all the time and valuable suggestions she offered. Dr. Christopher Powers has provided me with important biomechanical interpretations of my project. His contributions are undoubted. I am also grateful to Dr. Stanley Azen for his guidance of statistical analysis. Moreover, I extremely appreciate Dr. Todd Schroeder and Hans Dreyer for their assistance in DEXA measurement. I thank Matt Sandusky for his help in biomechanical system setup. Further, Cliff Fornwalt is acknowledged for his assistance in developing the upper extremity biomechanical model. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. iii Through my six years at USC, I am extremely grateful to my best friends for their untiring support and unlimited encouragement. I am particularly indebted to Chia-Ting Su. His unconditional support and thoughtful company in the past decade made me courageous and helped me overcome the obstacles in my life. I am also thankful to Dr. Yuyu Chang for his enthusiastic and generous contribution to subject recruitment in my study. His warm encouragement and spiritual support is acknowledged. The most special appreciation is to my mother, Li-Hsueh Chen. I thank her for her constant support and belief in me over the years. In our life, there are always numerous people to appreciate. I want to thank to all of the people who have helped me in anyway, anywhere, and anytime during my study at USC. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. TABLE OF CONTENTS ACKNOWLEDGEMENTS...........................................................................................................................ii LIST OF TA BLES.......................................................................................................................................viii LIST OF FIG U R ES........................................................................................................................................ x ABSTRACT......................................................................................................................................................xi 1 . SPECIFIC AIMS............................................................................................................................... 1 2 . BACKGROUND AND SIGNIFICANCE....................................................................................4 2.1 Mechanical Loading and Its Transduction on B o n e.....................................................4 2.1.1 Mechanocoupling................................................................................................5 2.1.2 Biochemical coupling.........................................................................................6 2.1.3 Transmission of biochemical signal................................................................ 7 2.1.4 Effector response................................................................................................. 7 2.2 Impact Exercise and Bone Mass in Humans.................................................................10 3 . PRELIMINARY STUDIES.........................................................................................................18 3.1 Quantifying Exercise-related Musculoskeletal Loading in the Lower Extrem ity...........................................................................................................................18 3.1.1 Lower extremity weighted vest mechanics....................................................19 3.1.2 Weighted-vest intervention study.................................................................. 20 3.1.3 Bench stepping kinetics in young adu lts......................................................21 3.1.4 Bench stepping kinetics in older adults.........................................................22 3.1.5 Squatting exercises in older adults.................................................................23 3.2 Quantifying Exercise-related Musculoskeletal Loading in the Upper E xtrem ity.......................................................................................................................... 24 3.2.1 Pilot study I: Damping effects on loading forces........................................25 3.2.1.1 Introduction..................................................................................25 3.2.1.2 Study design and methods......................................................... 25 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. V 3.2.1.3 Statistical A nalysis......................................................................29 3.2.1.4 Results........................................................................................... 30 3.2.1.5 Conclusions.................................................................................. 31 3.2.2 Pilot study II: Distance effects on loading forces....................................... 34 3.2.2.1 Introduction.................................................................................. 34 3.2.2.2 Study design and methods.......................................................... 34 3.2.2.3 Statistical A nalysis...................................................................... 37 3.2.2.4 Results...........................................................................................37 3.2.2.5 Conclusions..................................................................................43 3.2.3 Pilot study III: Reliability of upper extremity kinem atics.......................... 44 3.2.3.1 Introduction..................................................................................44 3.2.3.2 Study design and m ethods.......................................................... 44 3.2.3.3 Statistical A nalysis.......................................................................46 3.2.3.4 Results...........................................................................................46 3.2.3.5 Conclusions..................................................................................47 3.2.4 Pilot study IV: Repeatability of DEXA measurement on forearm 47 3.2.4.1 Introduction..................................................................................47 3.2.4.2 Study design and m ethods......................................................... 48 3.2.4.3 Results...........................................................................................49 3.2.4.4 Conclusions..................................................................................49 4 . M ETHODS..................................................................................................................................... 50 4.1 O verview ........................................................................................................................... 50 4.2 Subjects...............................................................................................................................51 4.3 Exercise Intervention....................................................................................................... 52 4.4 Measurements....................................................................................................................53 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 . DATA AND STATISTICAL ANALYSIS................................................................................58 5.1 Data Analysis.................................................................................................................... 58 5.2 Statistical Analysis........................................................................................................... 60 6 . RESULTS......................................................................................................................................... 61 6.1 Subjects...............................................................................................................................61 6.2 Repeatability......................................................................................................................62 6.2.1 Reaction force....................................................................................................62 6.2.2 Upper extremity kinematics............................................................................ 62 6.3 Physical Activity Level and Hand Grip Strength........................................................ 63 6.3.1 Physical activity level...................................................................................... 63 6.3.2 Hand grip strength............................................................................................64 6.4 Bone Adaptations after 6 months o f DILE Intervention............................................ 64 6.5 Reaction Force Characteristics...................................................................................... 69 6.5.1 The influence of damping condition on reaction force characteristics ..70 6.6 Associations among Reaction Force Characteristics and Bone Adaptations 70 6.6.1 The influence o f damping condition on the associations among reaction force characteristics and bone adaptations...................................71 6.7 Joint Kinematics............................................................................................................... 74 7 . DISCUSSION...................................................................................................................................75 7.1 Summary.............................................................................................................................75 7.2 Participation and Adherence........................................................................................... 76 7.3 Reproducibility of Reaction Forces and Joint Kinematics.........................................77 7.4 Hand Grip Strength and Physical Activity Level........................................................ 77 7.5 Bone Adaptations after 6 months of DILE Intervention............................................ 78 7.6 Oral Contraceptives Effects............................................................................................ 81 7.7 Correlational A nalysis..................................................................................................... 83 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. vii 7.8 Conclusion and Further Direction..................................................................................87 8 . REFERENCES................................................................................................................................89 9. APPENDICES.................................................................................................................................98 A Medical History Questionnaire...................................................................................... 98 B Physical Activity Questionnaire.................................................................................. 104 C Exercise D iary................................................................................................................ 108 D Secondary Statistical Analysis......................................................................................113 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. viii LIST OF TABLES 1. Intraclass correlation coefficient for loading forces...................................................................30 2. Repeated-measures ANOVA results for the outcome variables............................................. 31 3. Post-hoc test results for average peak lo ad................................................................................. 32 4. Post-hoc test results for ultimate peak load................................................................................. 33 5. Intraclass correlation coefficient for loading forces at 3 loading distances..........................37 6. Intraclass correlation coefficient for loading forces at 3 loading distances......................... 38 7. ANOVA for average peak load..................................................................................................... 39 8. Post-hoc test results for the distance effect on average peak lo ad .......................................... 39 9. ANOVA for average loading rate.................................................................................................40 10. Post-hoc test results for the distance effect on average loading rate.......................................40 11. ANOVA for average im pulse....................................................................................................... 41 12. Post-hoc test results for the distance effect on average im pulse............................................. 41 13. Intraclass correlation coefficient for upper extremity kinematics at baseline and re test..................................................................................................................................................... 46 14. Intraclass correlation coefficient for upper extremity kinematics........................................... 47 15. Demographic characteristics .........................................................................................................62 16. Within-session intraclass correlation coefficient for reaction forces in Z direction at the baseline and final .................................................................................................................... 62 17. Within-session and between-session intraclass correlation coefficient for upper extremity kinematics at the baseline and final ..........................................................................63 18. Physical activity level in 3 different domains..............................................................................63 19. Hand grip strength on exercised and non-exercised arms at the baseline and fin al.............. 64 20. Radial BMD of exercised and non-exercised arms at the baseline and fin al..........................65 21. Radial BMD of exercised arm at damped and non-damped groups........................................ 66 22. Radial BMD of exercised arm at OC and non-OC groups........................................................ 68 23. Radial BMD of non-exercised arm at OC and non-OC groups............................................... 68 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 24. Radial BMD of exercised and non-exercised arms in non-oral contraceptives u sers..........69 25. Perpendicular reaction forces during the loading phase o f DILE ............................................. 69 26. Correlations between reaction force characteristics associated with DILE and changes in radial bone density......................................................................................................70 27. Correlations between reaction force characteristics associated with DILE and changes in radial bone density in the damped group................................................................71 28. Correlations between reaction force characteristics associated with DILE and changes in radial bone density in the non-damped group........................................................72 29. Upper extremity kinematics during the loading phase of D IL E ..............................................74 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. X LIST OF FIGURES 1. Loading distance...............................................................................................................................27 2. Loading distance at 20cm, 30cm, and 40cm ................................................................................36 3. Average peak loads at 3 different loading distances.................................................................. 42 4. Average loading rates at 3 different loading distances..............................................................42 5. Average impulse at 3 different loading distances.......................................................................43 6. Dynamic impact loading exercise..................................................................................................54 7. Loading force direction....................................................................................................................57 8. DR BMD Percent change of exercised arm at damped and non-damped groups................ 66 9. UD BMD Percent change of exercised arm at damped and non-damped groups................ 67 10. TOTAL BMD Percent change of exercised arm at damped and non-damped groups 67 11. Correlations between peak load and changes in BMD at D R ..................................................72 12. Correlations between peak load and changes in BMD at TOTAL............................................ 72 13. Correlations between impact load and changes in BMD at D R .............................................. 73 14. Correlations between impact load and changes in BMD at T O TA L..................................... 73 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. ABSTRACT Introduction ' . The purpose of this study was to investigate the effects of Dynamic Impact Loading Exercise (DILE) on radial bone mineral density (BMD) in healthy premenopausal women and to determine the relations among the mechanical characteristics of loading and the bone adaptations. M ethods' . Twenty-four women (25-45 yr.) completed DILE 36 cycles a day, 3 days a week for 6 months. The exercised arm was allocated randomly to either the dominant side or the non-dominant side with the contralateral side serving as the non-exercised control. Subjects were also randomly assigned into either the damped or the non-damped treatment arms in order to vary the mechanical loading characteristics of the activity program. Measurements, including anthropometries, physical activity levels, hand grip strength, radial BMD at distal third, ultradistal, and total distal 1/3 radius, upper extremity kinematics, and reaction forces were recorded at baseline and after 6 months. Statistics' . Reliability of reaction force and upper extremity kinematics were assessed by Intraclass Correlation Coefficients (ICCs). Repeated-measures ANOVA was conducted to assess changes in physical activity level and hand grip strength, as well as the exercise effects and damping effects on bone adaptations. The damping effects on reaction force characteristics and joint kinematics were examined by one-way ANOVA. One-tailed Pearson’s correlation analyses were used to investigate the relations between the reaction force characteristics including peak load, impact load, loading rate, and impulse and changes in BMD (ABM D). Results' . DILE demonstrated highly reproducible loading profiles and joint kinematics. The average adherence rate was 93.3% and no injuries were reported. Although there were no significant bone adaptations observed after 6 months, the correlations between ABM D and mechanical characteristics o f loading, particularly peak loads and impact loads, were statistically significant. The correlations between the reaction force characteristics and ABM D were higher in the non-damped group than those in the damped group. Conclusion: DILE appears to be a valid and reliable upper extremity musculoskeletal loading model using humans, and results of correlation analyses parallel findings of previous bone R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. adaptation models employing animals. Findings suggest that further research, investigating the associations between mechanical loading events and human bone health, is warranted. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1. SPECIFIC AIMS Introduction: To investigate the relations between external loading characteristics and bone adaptation, we developed an upper extremity dynamic impact-loading model. The model was an attempt to provide a practical and accessible osteoporosis prevention program and allowed investigators to characterize potential osteogenic loading stimuli. Previous studies in our laboratory were designed to characterize the relations between musculoskeletal loading events and adaptations in the lower extremity. Using an easily accessible, unobtrusive exercise intervention program (weighted vest use), we quantified the lower extremity joint kinetics and musculoskeletal adaptations associated with short-term program participation. The development of this new program contributed to 1) understanding of relations between musculoskeletal loading and bone adaptation in the upper extremity, 2) better elucidating the mechanisms of loading induced skeletal adaptation by eliminating the confounding factors associated with lower-extremity characterization (e.g. muscular contributions, postural influences and learning strategies), and 3) assessing the effectiveness of a practical and accessible osteoporosis intervention program. The distal forearm is a common site of osteoporotic fracture (62). Although research studies have assessed the influence of exercise on bone mass in the distal radius, none have characterized the relations between osteogenesis and mechanical loading factors (3, 13, 100). Animal studies have suggested that mechanical loading characteristics that are 1) dynamic, 2) of a high magnitude, 3) of a high loading rate, and 4) exhibit an unusual strain distribution are optimal for osteogenesis (59, 69, 89, 90, 112). The relations between the mechanical characteristics of loading events and bone adaptations, however, have not been established in humans. The upper-extremity dynamic-impact loading model allowed us to bridge the gap between animal and human models by incorporating analyses of both loading characteristics and skeletal adaptation. Dynamic Impact Loading Exercise (DILE) is easy to perform, and required no equipment, supervision or participant travel. Because it R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. is time-efficient and unobtrusive in nature, it may prompt long-term adherence. Finally, more reliable loading kinematics and kinetics were observed when comparing to other exercise regimens such as jumping, because the upper-extremity loading regimen was restricted and the loading position of each joint was controlled. Objectives: The purpose of this research study is to examine the effects of DILE on radial bone mineral density in healthy premenopausal women and to determine the relations among the mechanical characteristics of loading and the bone changes. Subjects performed DILE with different damping conditions, in order to vary the mechanical loading characteristics (i.e., loading rate and magnitude) of the activity program. Specific Aim I: To investigate the effects of a 6-month DILE on radial bone mineral density (BMD) in young healthy women. Hypothesis la: Radial BMD increases in the exercised arm would be statistically significantly greater than those o f the control arm after 6 months of exercise intervention. H ypothesis lb: Radial BMD changes after 6 months in the exercised arm would be statistically significantly greater in subjects assigned to the non-damped treatment group than in subjects assigned to the damped treatment group. Specific Aim II: To investigate the relations among the mechanical characteristics of the activity program and the changes in radial BMD. Hypothesis Ila: There would be statistically significant difference in loading rate and loading magnitude between damped and non-damped treatment conditions. Hypothesis lib : Changes in radial BMD after the 6-month exercise intervention would be statistically significantly correlated with the peak loading forces engendered during the exercise program. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3 Hypothesis lie: Changes in radial BMD after the 6-month exercise intervention would be statistically significantly correlated with the peak loading rates engendered during the exercise program. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2. BACKGROUND AND SIGNIFICANCE 4 2.1 MECHANICAL LOADING AND ITS TRANSDUCTION ON BONE W olff (124) proposed that mechanical stress was responsible for determining the internal architecture and external form of bone according to mathematical rules. Although many basic tenets o f W olff s law have been challenged (12, 26), the idea that bone mass is affected by mechanical stresses has been supported by numerous research studies (85, 87, 89, 108). W olffs law has been further expanded to suggest that changes in bone structure are regulated by a feedback system in which alterations in local mechanical signals drive bone cells to alter bone structure (9, 33-35). The most biologically relevant of these theories is Frost’s “mechanostat” theory (33, 34). This theory describes a normal or physiological window of mechanical usage, in which bone is at a state of normal turnover and bone balance is maintained. When the system is perturbed, bone will adjust its structure to restore the strain within its physiological window. According to this theory, mechanisms within the bone monitor its mechanical usage, and correct serious misfits among mass, structure, and mechanical usage. These mechanisms behave like a home thermostat, turning “ON” in response to an error and “OFF” when no strain error is present; hence the term mechanostat. When local mechanical signals in bone exceed the upper boundary of the physiological window, referred to us the minimum effective strain (MES), bone remodels to reduce the local strains below the MES. In contrast, when local mechanical signals in bone are less than the lower boundary of the physiological window, bone will remodel and will be resorbed until the local strains are increased. Moreover, if the mechanical loads on the skeleton are so large that they cause microdamage of bone, the bone strains will be pushed into a pathological overload, causing woven bone formation on bone surfaces (36). Nonmechanical factors such as hormones, nutrition, biochemical agents and diseases may also affect the response of bone to mechanical loading. Although the mechanisms that drive the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 mechanostat process are unknown, they likely include mechanical and biological components, such that a mechanical load on a bone can be detected by cells. The conversion of a biophysical force into a cellular response in bone, i.e., mechanotransduction, is summarized by Turner and Pavalko (1998). The process of mechanotransduction in bone can be divided into four distinct phases: (1) mechanocoupling, the transduction of mechanical force applied to the bone into a local mechanical signal perceived by a sensor cell; (2) biochemical coupling, the transduction of a local mechanical signal into a biochemical signal and, ultimately, gene expression; (3) transmission o f signal from the sensor cell to the effector cell (i.e., the cell that will actually form or remove bone); and (4) the effector cell response, the final tissue-level response. 2.1.1 Mechanocoupling Normal usage of long bones places bending and compressive forces on local bone tissues and causes local deformation (i.e., strain) (11, 88). Studies indicate that osteocytes and bone lining cells can sense local bone strains because they are appropriately located in the bone tissue for this function (109). During loading, the spatial gradients in strain caused by bending forces create pressure gradients within the canaliculae that drive extracellular fluid flow. This flow pumps nutrients to the osteocytes and creates fluid shear stresses on osteocytes and cell processes (121). Moreover, the stress-generated fluid flow causes electric fields in bone called “streaming potentials” (72, 73, 94, 95, 98), which may exert direct effects on bone cells. The rate o f change of bone strain (i.e., strain rate) plays an important role in mechanotransduction. Turner (1995) suggested that relatively large strains alone are not sufficient to activate bone cells. High strain rates and possibly stress-generated fluid flow are required to stimulate new bone formation (110, 112). It is still unknown, however, which type of stimulus, mechanical stretch, fluid flow, or streaming potentials, have the greatest osteogenic effect on bone cells. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 2.1.2 Biochemical Coupling Although several possible transduction pathways, including extracellular matrix-integrin- cytoskeletal axis and mechanosensitive ion channels, have been proposed in the literature (24, 25, 48, 49, 57, 70, 91, 101), the mechanism for the initial detection and conversion of mechanical force into a biochemical signal has not been determined. The actin cytoskeleton is formed by a series of proteins including vinculin, paxillin, talin, tensin, and a-actinin. It attaches to the extracellular matrix by binding to membrane spanning glycoproteins called integrins and connects the extracellular matrix with the nucleus and the cytoplasmic constituents of the cell (75). Because the integrins bind to the matrix proteins on a rigid substratum, tension evoked by mechanical strain causes changes in the cytoskeletal structure. This deformation is then transmitted to the nucleus, resulting in alterations of gene expression (24, 48, 101). Indeed, studies have illustrated the important role played by cytoskeletal structure in the regulation of cellular proliferation, differentiation, morphogenesis, and gene expression (48). Furthermore, integrins have been directly linked to mechanical strain induced cellular responses (97). This evidence suggests that the extracellular matrix-integrins-cytoskeletal axis serves as an important mechanism for the signal transduction of mechanical strain. Mechanical strain can also affect cells via activation of ion channels, and mechanosensitive channels are likely candidates for this coupling mechanism (25, 49, 57, 70, 91). Mechanosensitive pathways, including stretch activated, cation nonselective (SA-cat) channels (25), potassium-selective channels (70), phospholipase C pathways (49), and guanine nucleotide binding proteins (G proteins) (57) have been identified as primary mechanotransduction mechanisms. They respond differently to various mechanical signals but have a high degree of association with one another. These findings suggest that there are probably multiple transduction pathways, and that they may simultaneously be activated by mechanical loading. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 7 2.1.3 Transmission of Biochemical Signal Following mechanical transduction, a biochemical signal from the sensor cell must be propagated to the effector cell to increase osteogenic activity. Active osteoblasts located on the bone surface are the most likely candidates for sensor cells. They can detect mechanical strain and act as the effector cell that deposits new bone in response to altered strain stimuli (45). However, it is unlikely that stimulation of active osteoblasts alone is sufficient to substantially increases bone mass (113) because it has been shown that the majority of the bone surface is covered by nonproliferative lining cells (94%). Only 5% of the bone surface in the adult human are covered by active osteoblasts. The rest of the bone surface is made up by osteoclasts (1%) (74). In addition to osteoblasts, osteocytes and bone lining cells are also responsive to mechanical loading (105). These cells communicate with other bone cells through an extensive network of cellular processes connected at gap junctions (82). Therefore, it is possible that a biochemical signal can be propagated from nonproliferative, strain sensing cells (i.e., osteocytes and bone lining cells) to osteoprogenitor cells and osteoblasts through paracrine factors. Research studies have shown that mechanosensitive cells (osteoblasts, osteocytes and lining cells) respond to mechanical strain with increased levels of second messengers such as cAMP and inositol phosphates (IP3) (49, 96, 115). Moreover, endocrine factors such as insulin-like growth factor I (IGF-1), IGF-II and prostaglandin E2 (PGE2 ) increase after mechanical loading (31, 60, 66, 107). Increases in these factors are associated with increased osteogenic function in existing osteoblasts and/or increased production of osteoblasts through recruitment and differentiation of precursor cells (65). 2.1.4 Effector Response Research studies document that exogenous mechanical strains alter bone mass and morphometry (1, 89, 90, 123). Load type, strain magnitude, strain rate, and strain distribution, have all been identified as factors influencing osteogenesis (59, 69, 89, 90, 112). It has been shown that only dynamic loading effectively causes anabolic effects on bone (59). Lanyon and Rubin (1984), using a turkey R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. ulna m odel, dem onstrated that static loading produced no adaptive rem odeling and did not maintain bone mass. However, application of the same load with only 100 cycles per day induced bone formation and increased ulnar cross-sectional area. It also appears that bone mass increases with increasing magnitude o f load (90, 111) and the anabolic responses occur only if the mechanical loads surpass a threshold (20, 111). These findings are consistent with Frost’s minimum effective strain theory. Rubin and Lanyon (1985) showed a graded dose-response relationship between the peak strain magnitude and change in the mass of bone tissue present. They also reported that applications of mechanical strains greater than 1000 pstrain (100 cycles/day) were associated with little change in intra-cortical remodeling activity but substantial periosteal and endosteal new bone formation. Moreover, when the applied peak longitudinal strains were lower than 1000 pstrain, bone loss occurred by increased remodeling activity, endosteal resorption, and increased intra-cortical porosis. Turner (1994) demonstrated that lamellar bone formation on the endocortical surface of rat tibiae increased linearly with increasing load for applied dynamic bending loads above 40N. However, no increase in bone formation was observed for applied bending loads less than 40N. Strain rate of mechanical loading is an important factor for osteogenesis as well. Rate o f strain is a term used to describe the time over which strain develops following load application. Studies have indicated that higher loading rates will cause greater bone formation (112). Turner (1995) applied 54N bending loads (2Hz, 36 cycles/day) with various loading rates to rat tibiae. Bone formation and mineral apposition rates were measured at baseline and after 2 weeks. The results suggested a marked linear rise in both measures as rate of strain increased. O ’Conner (1982) also investigated the role of strain rate in new bone formation. Bending and compressive loads were applied intermittently at 0.5 Hz through implants permanently inserted into the radius and ulna of experimental sheep. The ratio between the maximum strain rate of the artificial loads and the maximum strain rate during walking, which had the greatest influence on every remodeling parameter, was used for analysis. The results showed that the variation in this ratio accounted for R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 68-81% of the variation in surface bone formation and 43% of the variation in intracortical remodeling. Only an additional 6-12% variation in surface bone formation could be explained by adding axial strain into the equation, suggesting that the effect of axial strain was less marked than that of strain rate. Also, the direction of bending and axial loading (tension or compression) appeared to have no effect on the course of the remodeling observed. Bone formation can be stimulated by submaximal strains if strain distribution is abnormal (89). Rubin and Lanyon (1984) demonstrated the anabolic effects of dynamic mechanical loading within physiological magnitudes and rates on a functionally isolated rooster ulna. The results showed that only 4 cycles a day (0.5Hz) of load was needed to maintain bone mass, when strains were applied orthogonal to the neutral axis. In addition, extensive subperiosteal and endosteal new bone formation was observed when bones were subjected to 36 cycles per day. However, further increases of loading cycles (360 and 1800 cycles/day) were not associated with greater new bone formation. The ability of the skeleton to respond to mechanical loading is greatly reduced with aging (86, 114). Turner (1986) suggested that older rats (19 months) are less responsive to mechanical stimuli on both the periosteal and endocortical surfaces of bone when comparing with younger rats (9 months). Thirty-four 19-month-old rats were subjected to bending loads varying from 30 to 64N on the tibiae. When the applied loads were greater than 40N, old rats showed lower periosteal bone formation (59%) than younger rats (100%), but the woven bone area or surface were similar in both age groups. This suggests that the periosteum of old rats had a higher threshold for activation by mechanical loading; however, after activation occurred, the cells had the same capacity to form woven bone as younger adult rats on the periosteal surface. On the endocortical surface, only a marginal increase in relative bone formation rate was observed when the applied load was higher than 64N in older rats (over 16-fold less than that reported for the younger adult rats). The relative R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 10 bone-forming surface in older rats was 5-fold less than that observed in younger rats. Furthermore, a mechanical threshold for lamellar bone formation of 1050 gsti'ain was calculated for the endocortical bone surface in the younger adult rats, whereas the old rats required over 1700 (istrain before bone formation was increased. This suggests that the older bones had a higher threshold for osteogenesis than the younger bones. Similar results were also reported by Rubin (1992). Again using the functionally isolated turkey ulna model, the effects of high but physiological loads (300 cycles/day) on the ulnae of 1-year-old and 3-year-old turkeys were examined. Following 8 weeks of loading, bone cross-sectional areas in the 1-year-old animal increased by 30.2% as compared with the intact contralateral control ulnae, whereas the cross-sectional areas in the older skeleton remained essentially unchanged (-3.3%). In addition, the osteon mean wall thickness and bone formation sigma (duration of one complete remodeling cycle) were significantly increased in the 3- year-old males. These data suggest that a physical signal which is clearly osteogenic in the young adult skeleton may not be acknowledged in older bone tissue. 2.2 IMPACT EXERCISE AND BONE MASS IN HUMANS Age-related loss of bone may result in the clinical condition referred to as osteoporosis (22). The problem is particularly prevalent in women, whose bone loss accelerates after menopause. A woman may lose up to 50% of her peak trabecular bone mass by the age of 90, increasing her risk of fracture (22, 64). Osteoporosis and associated fractures have become a worldwide epidemic, and the associated disabilities and medical costs are accentuated by the aging of populations (21). One of the most common sites associated with osteoporotic fracture is the distal forearm (i.e., Colles’ fracture) where the incidence of fractures rises sharply after 50 years of age in women (62). Research studies have demonstrated that relatively modest increases in bone density could substantially decrease one’s risk of fracture (15, 18). Unfortunately, previous studies investigating the effects of exercise on human bone mass showed inconsistent results (2, 23, 68, 79, 106). For R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 11 example, Smith (1981) showed significant 2.29% increase in forearm BMC after 3 years of light to moderate exercise in a chair. Dalsky (1988) reported significantly positive effects o f weight bearing exercise on lumbar BMC with increments of 5.2% after 9 months of intervention. Aloia (1978) and Nelson (1991) reported nonsignificant changes in bone mass after 1 year of physical activity. Cavanaugh (1988) demonstrated a failure to maintain bone mass (-5.6%) by a 1-year walking intervention. Although these discrepancies can be explained by the differences in age, hormonal status, study design and methodology, “inappropriate” exercise prescription may be the most significant limitation to maximizing osteogenesis. For example, many o f the exercise programs evaluated were designed for the maintenance of aerobic capacity or muscle strength (2, 19, 58). However, animal studies have shown that loading regimens characterized by dynamic (59), high load/strain (90), high loading rate (69, 112) and unusual strain distribution (89) are optimal for osteogenesis. Thus, we believe there is merit in designing an exercise program which parallels successful animal models and describes potentially osteogenic exercise in bone-loading terms. Several studies employing high-impact loading exercises, designed to optimize the strain characteristics described above, have been investigated by researchers in the lower extremities (6-8, 42, 46). Heinonen (1996) found that high-impact exercise that loaded bones with a rapidly rising force profile, in versatile movements, improved skeletal integrity, muscular performance, and dynamic balance in premenopausal women. Ninety-eight sedentary women aged 35-45 years participated in an 18-month high-impact exercise program which primarily employed multidirectional jump training. The results showed significantly increased bone mineral density at the weight-bearing sites, i.e., lumbar spine, femoral neck, distal femur, patella, proximal tibia, and calcaneus; however, no significant change at non-weight bearing sites such as the distal radius was found when compared to the controls. Similar results were reported by Bassey (1994). Here, 27 healthy premenopausal women were randomly assigned to either a control group that participated in low-impact exercise or a test group that participated in intermittent high-impact exercise. Bone R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 12 mineral density at the femur and lumbar spine were assessed at 0, 6, and 12 months. At 6 months, the test group (n = 14) demonstrated a significant increase in trochanteric bone density (3.4%), a result significantly different from the control group. In the second 6 months, the control group was crossed over to high-impact exercise (n = 7) and showed a significant increase in trochanteric bone density (4.1%), while the original group maintained its improvement relative to the baseline. Grove (1992) examined the effect of high-impact and low-impact exercise on lumbar bone mineral density in 15 healthy early postmenopausal women. Subjects in the high-impact exercise group performed activities such as jumping and running whereas subjects in the low-impact exercise group performed walking and dancing. After 1 year of intervention, the control non-exercise group experienced a significant linear decrease in bone mineral density. Both the low and high impact exercise groups prevented bone loss but results did not significantly differ between groups. Uncontrolled hormonal therapy o f subjects and small sample size in Grove’s study could account for the failure to show any difference in results between the two exercise groups. Bassey (1995) investigated the effects of heel- drop exercise characterized by a high rate of loading in forty-four healthy postmenopausal women. Bone mineral density at the lumbar spine, proximal femur and distal radius were assessed at 0, 6, and 12 months. Ground reaction forces associated with heel-drop were also recorded using a force plate. Subjects in the exercise group (n = 20) performed the “heel drop” 50 times a day, without shoes, on a hard surface at home. The control group (n = 24) performed flexibility exercises at home and low-impact exercises in class. No significant increases in bone mineral density after 12 months of exercise at any site in either group were found. However, when women more than 6 years postmenopausal were analyzed separately, the control group showed significantly decreased bone mineral density at the spine and Ward’s triangle, results which were significantly different from the exercise group. The ground reaction force data indicated that although the rate of loading during heel-drop was more than double that found in jumping, the peak forces did not exceed those found during jumping, jogging or even brisk walking. Insufficient peak loading forces associated with the heel-drop could account for the lack of osteogenesis. Also, the osteogenic effect of the exercise R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 13 could be masked by the high rate of bone loss during early menopause. Bassey (1998) reported that a vertical jumping exercise regime effectively increased femoral BMD in premenopausal women, but not postmenopausal women. The exercise consisted of 5 bouts of 10 vertical jumps, 6 days a week (mean jumping height = 8.5 cm). The biomechanics data indicated that the mean ground reaction forces produced by the jumping exercise was 3 times body weight in the premenopausal women and 4 times in the postmenopausal women. The loading rate during landing in older women was 4 times greater than that of young women. Although the period of the exercise intervention was only 5 months for the premenopausal women, there were significant increases in the BMD in lumbar spine, femoral neck and femoral trochanter. In contrast, no significant difference in femoral BMD between the exercise and control groups after 12 months or 18 months of exercise intervention were observed in postmenopausal women when the data were divided to either hormonal replacement therapy group or deplete group or combined. The lack of anabolic responses to exercise could be explained by the early stage of menopause in the older women. Moreover, it is also possible that older women were less responsive to external mechanical loading than young women even when the jumping exercise was performed with higher ground reaction forces and loading rates. Some studies using high-intensity weight training exercises demonstrated increases in human bone mass (52, 67); however, some divergent results have also been reported. Although the inconsistent results can be accounted for by the differences in exercise protocols, hormonal status, study designs, participant age and methodology (84, 104), these studies are unable to explain the relation between the external loads and bone adaptations alone, because other factors may have confounded the results. Research studies have suggested that heavy-resistance exercise both acutely and chronically influences hormonal profiles (e.g. growth hormone and testosterone), having anabolic effects on bone (39, 53-55, 122). It is well known that growth hormone has anabolic effects on bone and skeletal muscle directly, via GH receptors, or indirectly via insulin-like growth factors (IGF-1) (37, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 14 51, 61, 125). GH stimulates protein synthesis and lipolysis, increases gluconeogenesis and indirectly inhibits carbohydrate metabolism. Resistance exercise induces acute increases in growth hormone concentration (39, 53-55). Gotshalk (1997) demonstrated that serum concentrations of GH significantly increased for all postexercise time points (0, 5, 15, 30, 60 min) compared to pre exercise values in men (mean age = 25.4 ± 4.1) (39). Also, higher training volume (3 sets vs. 1 set of 10RM resistance training) induced higher serum elevations. Kraemer RR (1995) found a luteal- phase induced increase in GH in response to an acute bout o f resistance exercise in women (3 sets at 10RM, 2 min between exercises) (53). Kraemer WJ (1993) showed a significant increase in GH after high volume, short duration resistance exercise (10RM with 1 min rest) in women (54). No significant acute changes from resting concentration levels of serum total IGF-1 after the heavy resistance exercise were observed. This may be due to the short duration o f follow-up (2 hours), because the release of IGF-1 from hepatic sources peaks 3 to 9 hours following mRNA synthesis. It is also suggested that androgens including testosterone and androstenedione increase protein synthesis and have anabolic effects on muscle and bone (16, 119). Research studies clearly demonstrate that resistance exercise increases serum concentrations of testosterone in men (39, 55, 122). Kraemer WJ (1998) investigated the acute and chronic training effects of resistance exercise (3 sets o f 6-8RM/10-12RM, 2 min rest between sets, 2 d/wk for 8 weeks) on total serum testosterone, cortisol, GH and sex-hormone binding globulin (SHBG) levels in healthy untrained adults (55). The data indicated that the baseline levels of serum total testosterone significantly increased while serum cortisol concentrations significantly decreased after 8 weeks of resistance training in both men and women. These results suggest a homeostatic shift in testosterone secretion pattern, which is likely to be coupled to remodeling processes. Additionally, a reduction in the amount of cortisol was identified, which may enhance muscle hypertrophy by limiting the degradation of protein rather than increasing protein synthesis. Furthermore, there were significant acute increases in GH at 0, 6, and 8 weeks of training, but no significant lasting changes in GH due R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 15 to training effect for both men and women. These findings illustrate the important effects of heavy- resistance exercise on horm onal profiles; however, the training prescriptions investigated all required the subjects to travel to a laboratory or a gym. Exercise prescriptions using these techniques may not be applicable to certain populations whom are unwilling or have difficulty traveling. DILE appears to overcome these obstacles because it is relatively safe and easy to learn. It can be performed without travel, specialized equipment, and supervision. Most importantly, DILE is designed to impart impact forces to the forearm, not through a great amount of muscle activity like resistance exercise. Therefore, changes of hormonal profiles induced by muscle activity will minimally confound the effects of external mechanical loads on bone. Only five studies have examined exercise programs which specifically loaded the forearm (1, 3, 13, 47, 100). Adami (1999) investigated the effects of site-specific moderate physical exercise on the ultradistal radius in 250 postmenopausal women. The exercise program (70 min/session, 2 sessions/wk for 6 months) was designed to maximize stresses on the wrist, and included press-up, flexion on the arms in a prone position, playing volleyball either sitting or standing and lifting a 500g weight with the forearm. The cross-sectional area of trabecular bone at the ultradistal radius and the cortical bone at the proximal radius was assessed by pQCT at baseline and after 6 months. No changes in cross-sectional area of cortical bone, the BMC and the volumetric BMD were observed at the proximal radius. At the ultradistal radius, the total BMC did not change significantly. However, the cross-sectional area of cortical bone increased by 2.8%, which appeared to be due to periosteal apposition and corticalization of the trabecular tissue. Also, there were significant increases in volumetric density (2.2%) and BMC (3.1%) of cortical bone, whereas significant decreases in volumetric density (2.6%) and BMC (3.4%) were seen in trabecular bone. These results suggest that some exercises may reshape the bone segment under mechanical stress although the changes in bone mass were not obvious. Beverly (1989) reported a positive effect of maximum isometric exercise on bone mineral content at the forearm in postmenopausal women. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 16 Subjects (n = 69) were required to squeeze a tennis ball for 30 sec daily for a total of 6 weeks. The results indicated a 3.4% gain in bone mineral content of the exercised arm after the 6-week exercise regimen, but the bone gain was not maintained when measured after 6 months. Ayalon (1987) and Simlcin (1987) designed a dynamic loading exercise program for the forearm, which included tension, torsion, compression, and bending forces. Activities designed to impart these stresses included hanging from a ladder, arm wrestling, pulling and twisting. Fourteen women aged 53-74 years participated in this exercise program for a total of 5 months. After 5 months, the results showed that radial bone mass decreased by 2% in the control group whereas it increased significantly by 3.8% in the exercise group. Although these studies could not effectively provide evidence that physical loading was effective for bone formation, since the duration of the studies was short, they suggest that radial bone responds rapidly to external loads that are diverse and novel in nature. Heinonen (1996) examined the effects of 12-month unilateral high-resistance strength training and 8-month detraining on upper-extremity bone mass in young women. A total of 13 subjects trained their left upper extremities for elbow flexion and extension with dumbbells an average of 2.8 times per week (80% 1RM, 10 repetitions per set for 5 sets). Another 19 subjects served as controls. The measurements included BMC, BMD, bone width, cortical wall thickness and cross-sectional moment of inertia at proximal humerus, humeral shaft, radial shaft, ulnar shaft, and distal forearm for both upper extremities. Isometric strength tests of elbow flexion and extension were also employed. After 12 months, the training significantly increased the isometric elbow flexion strength by 14% and the isometric elbow extension strength by 21%. However, no significant differences in all of the bone measures were observed between the trained arms and the left arms in the control group, as well as the right arms in the trained subjects and the right arms in the control group. In addition, the changes in all o f the bone measures in the left limb did not differ significantly from those in the right limb for either group. After a detraining period of 8 months, all measured strength R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 17 values in the training group decreased in both limbs. The trends of all bone measures observed during the training periods remained mostly unchanged. A biomechanical model was also developed in this study to assess the strains associated with the strengthening program. The data indicated that the estimated loading-induced strains remained within customary (normal) loading and the change in strain level was only 15%, which may be not sufficient for osteogenesis. Furthermore, the loading fashion of the strengthening exercise probably did not produce a sufficiently novel pattern of strain distribution needed for bone adaptation. No impact exercise intervention studies, designed specifically for the distal forearm, have been reported. Moreover, the influences of osteogenic mechanical loading characteristics have rarely been investigated in humans. In this proposed investigation, we have designed a dynamic impact- loading exercise program that provides high loading rates, high loads and unusual strain distributions to the distal forearm. The loading regimen is restricted to limit the variation of loading forces; thus, more consistent loading kinematics and kinetics are expected when compared to other exercise regimens such as jumping. Furthermore, unlike traditional resistance exercise which involves a great amount of muscle activity, DILE is designed to impart impact forces to the forearm, rather than deliver the loading forces through muscular activity. Therefore, we believe that the effects of external mechanical loads on bone will be minimally confounded by factors such as muscle activity and muscle induced changes o f hormonal profiles. Finally, this upper extremity model will allow us to quantify the mechanical loading characteristics associated with the exercise using biomechanical analysis. Because of the novel nature of this loading regimen and the fact that young bone responds more favorably than older bone to the effects of mechanical loading (86, 114), this investigation seeks to assess the applicability of this intervention in young adults. If the intervention program proves successful, it may be repeated using older-adult participants. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3. PRELIMINARY STUDIES 18 Age-related osteoporosis increases the risk of fracture, disability and physical dependence (21, 22). It has been placed at $5.2 billion each year in the United States (76) and £615 million each year in the United Kingdom (50). Consequently, assessing the effects o f exercise intervention programs on the prescription of bone health is becoming increasingly important. These assessments require 1) quantification of physical loading of musculoskeletal regions associated with exercise participation, 2) quantification of the adaptations associated with exercise intervention, and 3) characterization of the relations between exercise-mediated musculoskeletal loading and musculoskeletal adaptation. Earlier studies conducted in our laboratory (USC Musculoskeletal Biomechanics Research Laboratory; MBRL) have examined practical and accessible intervention strategies for age- associated declines in musculoskeletal function. 3.1 QUANTIFYING EXERCISE-RELATED MUSCULOSKELETAL LOADING IN THE LOWER EXTREMITY We developed a physical activity intervention using a weighted vest and quantitatively assessed the effects of wearing a weighted vest on musculoskeletal adaptation in the lower extremities. The weighted vests are lightweight nylon garments with eight pockets (4 each in front and back). Each packet can be filled with weights ranging from 0.025 to 1.5 kg (maximum total weight, 12 kg) and the load o f weights can be distributed evenly in the front and the back. Vest weight is prescribed as a percent of body weight. An obvious advantage of the weighted-vest intervention is that it allows exercise to be done incidental to activities o f daily living via additional resistance applied by the weight of the vest. There is no need for equipment, supervision, or participant travel. Additionally, this intervention program may promote longer-term adherence, because of its unobtrusive nature. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 19 3.1.1 Lower Extremity Weighted Vest Mechanics Salem, G. J., W ang, M.-Y., Azen, S., Young, J. T., Marison, M., and Greendale, G. A. Lower- extremity kinetic response to activity program dosing in older adults. Journal o f Applied Biomechanics. 17:103-12,2001. In order to quantitatively evaluate the mechanical loading on the musculoskeletal system associated with wearing a weighted vest, comprehensive biomechanical analysis was employed. Biomechanical investigation is the only mechanism capable of characterizing musculoskeletal loading without invasive, surgical intervention. We determined the effects o f two doses of weighted vest use on lower-extremity gait kinetics in older adults (92). A motion analysis system (Vicon 370, Oxford Metrics, Oxford, UK) was used to quantify the peak lower-extremity joint moments and vertical peak ground reaction forces in 56 men and women volunteers (73.8 ± 6.9 years old) enrolled in a 6-month physical activity study. At the initial study visit, participants performed 6 walking trials (3 with-vest and 3 without-vest) at their “normal” pace. During the vest-wearing trials, participants wore a vest loaded with either 0% of body weight (BW) (N=19), 3% of BW (N=16), or 5% of BW (N=21). Two-factor repeated measures ANOVA was used to assess the overall differences in the mean values of the maximum peak extensor moments between trials performed with and without the vests. When significant differences in outcomes as a result of applying vest weight were identified, all pair-wise comparisons were done with alpha set at p = 0.017. With vest weights, maximum peak plantarflexion moments increased by 5.7% in the 5% BW group compared to the 0% BW group (p < 0.01). Compared to the 0% vest-weight group, knee extension moments increased by 13.8% w hen 5% B W w as applied (p < 0.01); a m arginally significant treatm ent effect was evident in the 3% BW group (p = 0.04). These findings suggest that changes in vest-mediated lower-extremity kinetics are joint specific and load dependent. This information may be used by R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 20 clinicians to appropriately prescribe vest weights to strengthen targeted muscle groups in older adults. 3.1.2 Weighted-Vest Intervention Study Greendale, G. A., Salem, G. J., Young, J. T., Damesyn, M., Marion, M., W ang, M.-Y., and Reuben, D. B. A randomized trial of weighted vest use in ambulatory older adults: strength, performance, and quality of life outcomes. Journal o f the American Geriatrics Society. 48:305-11, 2000 . We also conducted a 27-week, randomized, controlled, non-masked trial of wearing one of two doses of a weighted vest compared to no vest use (41). Sixty-two men and women volunteers (mean age 74 years) were randomly assigned to one of three treatments: no vest, 3%B W vest, or 5%B W vest use. Participants were asked to wear the vests for 2 hours daily, 4 days per week. No specific physical activities were prescribed. Bone turnover was assessed using serum osteocalcin (OC) and urine N-telopeptides (NTX). Muscular strength assessments included knee isokinetic, isometric, and endurance measures. Isokinetic and endurance muscle tests were performed at 60°/sec. Functional performance assessments included 8- and 50-foot walk tests at normal pace, five timed chair stands, timed stair climb, one-leg stand (eye open), and functional reach tests. Quality of life outcomes were also tested. After 27-week intervention, neither OC (p=0.36) nor NTX (p=0.48) changed significantly with treatment. No discernible effects of wearing a weighted vest on quality of life, strength or functional performance were found. Participants began the study with strikingly high (internal) health locus of control scores (an indication of beliefs that health is controlled by subject- self rather than other or chance). Fear of falling was virtually absent at baseline. Study retention was high: 90%, 80%, and 91% of the control, 3%, and 5% groups, respectively, attended the follow-up study visit. In both vest groups, walking (25% of vest-wearing minutes) and light housework (25%) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 21 were the leading activities reported. To produce musculoskeletal adaptations, the required stimulus is relative; it must be tailored to the condition of the participants at the start of intervention. The higher the individual’s baseline values, the greater the required stimulus. Therefore, it is possible that our volunteers were too fit at the outset to derive benefits from the vest stimulus prescribed in the study. We conclude, from the absence of any effect on strength or performance in this relatively fit study sample, that the vest intervention did not reach the lower limit of a training stimulus for the muscle groups and tasks assessed. The above studies were conducted to develop a weighted vest intervention program and to assess the effectiveness and applicability of the prescribed weighted vest use. The results of muscular function, including muscle strength and functional performance were inconsistent with previous data from our group (40, 93) and from other researchers (99). This may be due to the discrepancies in vest loads and physical activity program used among the studies. However, it is obvious that vest loads, up to 5%BW, have no effects on serum makers of bone adaptations, at least at the 27-week time point. Further research studies are required before we can effectively and safely prescribe maximally accessible, dose adjustable, adherent, weighted-vest activity programs that preserve musculoskeletal function of the lower extremity. 3.1.3 Bench Stepping Kinetics in Young Adults Shuler, C. J., Salem, G. J., W ang, M.-Y., and Jaque, S. V. Weighted-vest resistance and ground reaction forces associated with bench stepping exercise. Journal o f Strength and Conditioning R esearch (In R eview ). In order to assess the dynamic loading parameters associated with different bench-steeping activities, with and without a weighted vest, we investigated the peak vertical ground reaction forces R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 22 (PVGRFs) during bench step aerobics (117). Female volunteers (N= 19; 25.4 ± 2.5 years o f age) performed 4 stepping trials in random sequence; front unweighted (FS), front weighted (FSWV), lateral unweighted (LS), and lateral weighted (LSWV) bench stepping with and without a 5%BW weighted vest. The bench was 10 inch in height and the stepping exercises were performed at a rate of 20 cycles per minute. Repeated measures ANOVA revealed that LS produced 8.3% greater PVGRFs than FS (p = 0.05). LSWV generated 10.7% greater PVGRFs than LS (p < 0.01) and 15.2% greater PVGRFs than FSWV (p < 0.01). FSWV did not produce significantly greater PVGRFs than FS. These findings suggest that lateral stepping exercises may be better at loading the musculoskeletal system than traditional forward stepping exercises. Further, they suggest that resistance doses of greater than 5% BW may be needed to increase musculoskeletal loading during forward stepping maneuvers in college-aged subjects. 3.1.4 Bench Stepping Kinetics in Older Adults W ang, M.-Y., Salem, G. J, Flanagan, S., and Greendale, G. A. Lower-extremity joint kinetics during forward stepping and lateral stepping in older adults. Journal o f Applied Biomechanics (In Review). We also investigated the lower-extremity joint kinematics and kinetics during forward and lateral stepping in older adults (116). Twenty-one healthy older adults (70-85 years) randomly performed 3 trials of forward stepping (FS) and lateral stepping (LS) at a self-selected pace (step height: 21 cm). Repeated measures analysis was used to test for differences in joint kinematics and kinetics between activities. Results indicated that LS produced significantly greater joint angular excursions than FS (P < 0.01) at the knee and ankle. Joint moments generated at the hip, knee, and ankle were similar between activities. At the hip joint, power and total work generated during FS were 32.4% and 26.1% greater than LS, respectively (p<0.001). Conversely, LS produced significantly greater R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 23 power (10.5%, p < 0.001)) and total work (8.1%, p < 0.01) than FS at the knee. At the ankle, LS produced 250.0% and 197.7% greater power and work (p < 0.001), respectively, than FS. The support impulse generated during LS was 13.8% greater than the FS (P < 0.01). FS induced a different pattern of musculoskeletal demand than LS in older adults. These findings may be used to prescribe exercise programs for improving performance of targeted lower-extremity muscle groups. 3.1.5 Squatting Exercises in older adults Flanagan, S., Salem, G. J., W ang, M.-Y., Sanker, S., and Greendale, G.A. Squatting exercises in older adults: kinematic, kinetic and electromyographic comparisons. Medicine and Science in Sport and Exercise (In Review). Squatting activities may be used, within exercise programs, to preserve physical function in older adults. In this study, we characterized the lower extremity joint excursions, peak moments, powers, work, impulse, and muscle recruitment patterns (EMG) associated with two types of squatting activities in elders (32). Twenty-two healthy, older adults (ages 70-85) performed three trials each of: 1) a squat to a self-selected depth (normal squat; SQ) and 2) a squat onto a chair with a standardized height of 43.8 cm (chair squat; CSQ). Descending and ascending phase joint kinematics and kinetics were obtained using a motion analysis system and inverse dynamics techniques. A 2 x 2 (activity x phase) ANOVA with repeated measures was used to examine the biomechanical differences among the two activities and phases. Results suggested that CSQ generated greater hip joint excursions, peak moments, power, and work; whereas SQ generated greater knee and ankle joint excursions, peak moments, power, and work. SQ generated a greater knee extensor impulse, a greater plantar flexor impulse, and a greater total support impulse. The R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 24 EMG data were consistent with the kinetic data. The findings suggest that, with older adults, CSQ places greater demand on the hip extensors, while SQ places greater demand on the knee extensors and ankle plantar flexors. Summary These previous studies have allowed us to quantify the mechanical loading associated with resistance exercise activities in the lower extremity in older and younger adults. They illustrate that vest loads imparted on the musculoskeletal system are region-specific, and that a minimum threshold (5%BW) of a vest load may be required to effect musculoskeletal loading specific joints. Exercises, such as aerobic stepping and squatting exercises, may invoke significantly different loading patterns to the musculoskeletal system depending upon the maneuver performed and the resistance used. This information may ultimately be used to develop an efficient exercise regimen, and prescribe exercise intervention for the preservation and/or enhancement of the lower-extremity musculoskeletal system. 3.2 QUANTIFYING EXERCISE-RELATED MUSCULOSKELETAL LOADING IN THE UPPER EXTREMITY Due to the novel nature of the upper extremity dynamic impact loading exercise program, introductory experiments were performed to determine the padding material (Pilot Study I) and the loading distance (Pilot Study II) used in the intervention study. Additionally, the repeatability o f the loading prescription, upper extremity kinematics (Pilot Study 111), and DEXA measurement of the radial bone mineral density (Pilot Study IV) was assessed. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 25 3.2.1 Pilot Study I: Damping Effects on Loading Forces 3.2.1.1 INTRODUCTION Purpose: The purpose of this pilot study was to select the padding material for the intervention study. Specific Aim s: a. To investigate the damping effects of 4 different padding materials. Mechanical characteristics of the loading forces including loading rate, peak load, and impulse (area under curve) were used for analysis. b. To assess the within-trial repeatability of the loading regimen. Subjects: Ten healthy adults were recruited from the USC Department of Biokinesiology and Physical Therapy. Subjects with musculoskeletal disorders of the upper extremities, such as arthritis, carpal tunnel syndrome and history of Colies’ fracture were excluded. 3.2.1.2 STUDY DESIGN AND M ETHODS This randomized-block pilot study investigated the damping effects of different padding materials on loading forces during DILE. Each subject performed DILE with their right arm using 5 different padding conditions. The sequencing of the padding conditions during the data collection was randomized. All experimental procedures were conducted at MBRL. Padding Conditions: There were a total of 5 padding conditions, N, TNE, TKE, TNP, and TKP in this study. a. N: Subject impacted against the force platform directly without any damping pad. b. TNE\ Subject impacted against the force platform covered with 1/4" thick ensolite close cell foam. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 26 C . TKE: Subject impacted against the force platform covered with 1/2" thick ensolite close cell foam. d. TNP\ Subject impacted against the force platform covered with 1/4" thick polyethylene close cell foam. e. TKP: Subject impacted against the force platform covered with 1/2" thick polyethylene close cell foam. Data Collection a. Anthropometries: Standing height in cm and body weight in N were measured. b. Loading distance: The loading distance was standardized. In order to obtain a standardized distance, we measured the distance between subject’s heel and the wall (Dh-w, a distance we believe to be more reliable measurement then the distance between wrist and the force plate surface). The following distances were obtained (Figure 1): (a) Dh-s: horizontal distance between heel and lateral styloid process, and (b) Dp-w: thickness of the force platform, i.e., platform surface to the wall. The loading distance Ds-p was set at 25 cm between styloid process and platform surface. Therefore, the distance Dh-w between heel and the wall was calculated as follow: Dh-w = Dh-s + Ds-p + Dp-w The Dh-w was then used as a standard loading distance for each subject. When measuring Dh-s, subjects were asked to stand against the wall with their heels, back, and scapulae contacting the wall. Subjects then elevated their right arms to a horizontal level. The horizontal distance between the heel and lateral styloid process was recorded. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 27 SiL Force Platform L J Dh-s Ds-p Dp-w (25 cm) Dh-w Figure 1: Loading distance c. Im pact Loading Forces: Loading forces associated with the exercise were measured using a strain gauge force platform (model OR6-6-1000, AMTI Inc., MA, USA) at 600 Hz for the first 5 subjects and 1200 Hz for the last 5 subjects. When analyzing these 2 sets of data (first 5 vs. last 5 subjects) separately, we found that the loading-rate measure was easier to calculate and more consistent when the data were collected at 1200 Hz. During data collection, subjects were asked to stand in front of the force platform with their right arm in the following position: shoulder 90° flexed, elbow fully extended, and wrist extended. The subjects were then asked to impact against the wall while maintaining an extended elbow. Each subject performed the loading regimen using the 5 padding conditions. Subjects performed this task 7 times (cycles) continuously with each padding condition. The force data were collected during the 2nd through 6th cycles. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 28 Data Analysis The analog data collected from force platform were converted to ASCII format using Vicon Reporter (Oxford Metrics Lt., Oxford, UK). The data were then read by Peak Fit (SPSS Inc., Chicago, IL, USA) which allowed us to offset the baseline, identify the start point of each loading cycle and generate the outcome variables. The start point of each loading cycle was defined as the first data point where the loading force was higher than 5 N. Outcome variables in this study were defined and calculated as follows: Outcome variables at each loading cycle a. Peak load (Pi): Peak load was defined as the maximum load generated during each cycle. In situations in which there were more than one peak, the peak with the greatest magnitude was used. b. Loading rate (Ri): The loading rate was calculated as peak load divided by the period of time from initial loading to peak load. In trials that had more than one peak within a loading cycle, the first peak was used. The formula can be described as follow, where Y is the loading magnitude and X is the time frame (in 1200 Hz increments). Loading rate Yfasl / (Xa t first p ea k ^ a t s ta rt) C. Im pulse (Ii): Linear impulse was the area under the force-time curve F(t), which can be calculated using the following formula where X is the time frame at the start and the end points of a loading cycle. Impulse = J ™ t F(t) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 29 Outcome variables following 5 loading cycles After obtaining the outcome variables for each loading cycle, additional (summary) variables were generated: d. Average peak load: Average peak load was the mean value of all 5 peak loads; i.e., (Pl+P2+P3+P4+P5)/5. e. Ultimate peak load: Ultimate peak load was the maximum value among the 5 peak loads; i.e., maximum (P1,P2,P3,P4,P5). f Average loading rate: Average loading rate was the mean value of all 5 loading rates; i.e., (Rl+R2+R3+R4+R5)/5. g. Ultimate loading rate: Ultimate loading rate was the maximum value among the 5 loading rates; i.e., maximum (R1,R2,R3,R4,R5) h. Average impulse: Average impulse was the mean value of all 5 impulses; i.e., (II+I2+I3+I4+I5)/5. i. Ultimate impulse: Ultimate impulse was the maximum value among the 5 impulses; i.e., maximum (11,12,13,14,15) j. Sum o f impulse: Sum of impulse was the sum of all 5 impulses; i.e., 11+I2+I3+I4+I5. 3.2.1.3 STATISTICAL ANALYSIS Within-trial reliability (i.e. among the 5 loading cycles) of peak load, loading rate, and impulse was assessed by generating intraclass correlation coefficients (ICCs) for each repeated measure. The effects of padding condition were analyzed using repeated-measures ANOVA. The null hypothesis was that there were no significant differences in the dependent variables among the 5 padding conditions. Dependent variables in this study included average peak load, ultimate peak load, average loading rate, ultimate loading rate, average impulse, ultimate impulse and sum of impulse. Post-hoc tests, including Tukey and Scheffe tests, were used to determine statistically significant R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 30 differences among padding conditions. All statistical procedures were conducted using SPSS 10.0 (SPSS Inc., Chicago, IL, USA). 3.2.1.4 RESULTS R epeatability: The ICCs for peak load, loading rate and impulse among the 5 loading cycles were excellent, ranging from .95 to .99 (Table 1). Table 1: Intraclass correlation coefficient for loading forces Variable IC C (n=10) Peak load .9817 Loading rate .9526 Impulse .9951 Padding effects: The repeated-measures ANOVA results indicated that there were statistically significant differences in average peak load (p<.05) and ultimate peak load (p<.05) among the 5 padding conditions (Table 2). When further analyzing the data of average peak load, Tukey’s post- hoc test revealed statistically significant differences between TNE and N, TKE and N, and TKP and N; whereas Scheffe’s post-hoc test indicated statistically significant differences between TNE and N, and TKP and N (Table 3). Similarly, when analyzing data of ultimate peak load, there were statistically significant differences between TNE and N, TKE and N, and TKP and N by Tukey’s test; while only TNE and N were statistically significantly different from each other using the Scheffe’s test (Table 4). Overall, only TNE and N were significant different from each other in average peak load and ultimate peak load for both Tukey and Scheffe’s post-hoc tests. There were no statistically significant differences in all outcome variables among all the padding groups. The results indicated that the padding groups were significantly different in peak values, including R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 31 average peak load and ultimate peak load, from the non-padding group, but there were no differences among the padding groups. Moreover, TNE (1/4" thick ensolite close cell foam) showed the most significantly different peak values from the non-padding group N. 3.2.1.5 CONCLUSIONS Within-trial reliability was excellent for peak load, loading rate and impulse. Data sampling rate at 1200 Hz tended to be more sensitive to detect differences in loading rate among padding conditions than that at 600 Hz. There were statistically significant differences in average peak load and ultimate peak load between padding and non-padding conditions. However, no statistically significant differences in all variables among the 4 padding conditions were observed. Based upon these results, the sampling rate at 1200 Hz was used in the Pilot Study II and will be used in the intervention study. Moreover, we selected 1/4" thick ensolite close cell foam (which showed the greatest mean difference and the highest significance level from non-padding condition) for subsequent investigations, including Pilot Study II, which determined the loading distance for the intervention study. Table 2: Repeated-measures ANOVA results for the outcome variables Variable ANO VA p value (n=10) Average peak .008* Average loading rate .071 Ultimate peak .008* Ultimate loading rate .171 Average AUC .268 Ultimate AUC .693 Sum AUC .268 *p<.05 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 32 Table 3: Post-hoc test results for average peak load Multiple Comparisons Dependent Variable: AVEM AXP Mean Difference 95% Confidence I nterval (I) PAD# (J) PAD# (i-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1 2 -5.33580 10.40603 .986 -35.21008 24.53848 3 -36.22166* 10.40603 .011 -66.09594 -6.34738 4 -15.69158 10.40603 .564 -45.56586 14.18270 5 -2.08038 10.40603 1.000 -31.95466 27.79390 2 1 5.33580 10.40603 .986 -24.53848 35.21008 3 -30.88586* 10.40603 .040 -60.76014 -1.01158 4 -10.35578 10.40603 .856 -40.23006 19.51850 5 3.25542 10.40603 .998 -26.61886 33.12970 3 1 36.22166* 10.40603 .011 6.34738 66.09594 2 30.88586* 10.40603 .040 1.01158 60.76014 4 20.53008 10.40603 .299 -9.34420 50.40436 5 34.14128* 10.40603 .018 4.26700 64.01556 4 1 15.69158 10.40603 .564 -14.18270 45.56586 2 10.35578 10.40603 .856 -19.51850 40.23006 3 -20.53008 10.40603 .299 -50.40436 9.34420 5 13.61120 10.40603 .688 -16.26308 43.48548 5 1 2.08038 10.40603 1.000 -27.79390 31.95466 2 -3.25542 10.40603 .998 -33.12970 26.61886 3 -34.14128* 10.40603 .018 -64.01556 -4.26700 4 -13.61120 10.40603 .688 -43.48548 16.26308 Scheffe 1 2 -5.33580 10.40603 .992 -39.10993 28.43833 3 -36.22166* 10.40603 .030 -69.99579 -2.44753 4 -15.69158 10.40603 .687 -49.46571 18.08255 5 -2.08038 10.40603 1.000 -35.85451 31.69375 2 t 5.33580 10.40603 .992 -28.43833 39.10993 3 -30.88586 10.40603 .088 -64.65999 2.88827 4 -10.35578 10.40603 .909 -44.12991 23.41835 5 3.25542 10.40603 .999 -30.51871 37.02955 3 1 36.22166* 10.40603 .030 2.44753 69.99579 2 30.88586 10.40603 .088 -2.88827 64.65999 4 20.53008 10.40603 .434 -13.24405 54.30421 5 34.14128* 10.40603 .046 .36715 67.91541 4 1 15.69158 10.40603 .687 -18.08255 49.46571 2 10.35578 10.40603 .909 -23.41835 44.12991 3 -20.53008 10.40603 .434 -54.30421 13.24405 5 13.61120 10.40603 .788 -20.16293 47.38533 5 1 2.08038 10.40603 1.000 -31.69375 35.85451 2 -3.25542 10.40603 .999 -37.02955 30.51871 3 -34.14128* 10.40603 .046 -67.91541 -.36715 4 -13.61120 10.40603 .788 -47.38533 20.16293 Based on observed m eans. * • The m ean difference is significant at the .05 level. Note: Pad #1 = TN E, Pad #2 = TKE, Pad #3 = N, Pad #4 = TN P, Pad #5 = TKP R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 33 Table 4: Post-hoc test results for ultimate peak load Multiple Comparisons D ependent Variable: UMAXP Mean Difference 95% Confidence Interval (I) PAD# (J) PAD# (l-J) Std. Error Sig. Lower Bound Upper Bound Tukey HSD 1 2 -4.0255 15.7470 .999 -49.2331 41.1821 3 -54.0859* 15.7470 .012 -99.2935 -8.8783 4 -19.5053 15.7470 .729 -64.7129 25.7023 5 -3.1305 15.7470 1.000 -48.3381 42.0771 2 1 4.0255 15.7470 .999 -41.1821 49.2331 3 -50.0604* 15.7470 .024 -95.2680 -4.8528 4 -15.4798 15.7470 .861 -60.6874 29.7278 5 .8950 15.7470 1.000 -44.3126 46.1026 3 1 54.0859* 15.7470 .012 8.8783 99.2935 2 50.0604* 15.7470 .024 4.8528 95.2680 4 34.5806 15.7470 .204 -10.6270 79.7882 5 50.9554* 15.7470 .021 5.7478 96.1630 4 1 19.5053 15.7470 .729 -25.7023 64.7129 2 15.4798 15.7470 .861 -29.7278 60.6874 3 -34.5806 15.7470 .204 -79.7882 10.6270 5 16.3748 15.7470 .835 -28.8328 61.5824 5 1 3.1305 15.7470 1.000 -42.0771 48.3381 2 -.8950 15.7470 1.000 -46.1026 44.3126 3 -50.9554* 15.7470 .021 -96.1630 -5.7478 4 -16.3748 15.7470 .835 -61.5824 28.8328 Scheffe 1 2 -4.0255 15.7470 .999 -55.1345 47.0835 3 -54.0859* 15.7470 .033 -105.1949 -2.9769 4 -19.5053 15.7470 .819 -70.6143 31.6037 5 -3.1305 15.7470 1.000 -54.2395 47.9785 2 1 4.0255 15.7470 .999 -47.0835 55.1345 3 -50.0604 15.7470 .058 -101.1694 1.0486 4 -15.4798 15.7470 .913 -66.5888 35.6292 5 .8950 15.7470 1.000 -50.2140 52.0040 3 1 54.0859* 15.7470 .033 2.9769 105.1949 2 50.0604 15.7470 .058 -1.0486 101.1694 4 34.5806 15.7470 .325 -16.5284 85.6896 5 50.9554 15.7470 .051 -.1536 102.0644 4 1 19.5053 15.7470 .819 -31.6037 70.6143 2 15.4798 15.7470 .913 -35.6292 66.5888 3 -34.5806 15.7470 .325 -85.6896 16.5284 5 16.3748 15.7470 .895 -34.7342 67.4838 5 1 3.1305 15.7470 1.000 -47.9785 54.2395 2 -.8950 15.7470 1.000 -52.0040 50.2140 3 -50.9554 15.7470 .051 -102.0644 .1536 4 -16.3748 15.7470 .895 -67.4838 34.7342 Based on observed m eans. * ■ The m ean difference is significant at the .05 level. Note: Pad #1 = TN E, Pad #2 = TKE, Pad #3 = N, Pad #4 = TN P, Pad #5 = TKP R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3.2.2 Pilot Study II: Distance Effects on Loading Forces 3.2.2.1 INTRODUCTION Purpose: The purpose of this investigation was to determine the loading distance for the dynamic impact loading exercise to be used during the intervention study. One-week test-retest reliability of the “novel” impact loading exercise was also assessed. Specific Aims: a. To investigate the effects of 3 different loading distances (defined as the distance between wrist joint and the wall; 20cm, 30cm and 40cm) on the mechanical characteristics (loading rate, peak load, and impulse) of the loading events. b. To assess one-week test-retest reliability of the peak loads, loading rates, and impulses associated with DILE. Subjects: Five healthy men (25-35 yrs) and five healthy women (26-34 yrs) were recruited from USC Department of Biokinesiology and Physical Therapy. Subjects with musculoskeletal disorders of the upper extremities, such as arthritis, carpal tunnel syndrome and history of Co lies’ fracture were excluded. 3.2.2.2 STUDY DESIGN AND METHODS This randomized-block pilot study investigated the relations between upper extremity loading distances and loading forces. Each subject performed DILE with their right arm at 3 different loading distances, 20, 30, and 40 cm, using padded and unpadded conditions. Thus, there were a total of 6 different loading conditions. The padding material was 1/4” thick ensolite close cell form selected based on the results of the Pilot Study I. The sequence of loading conditions was randomized. The same procedures were repeated at baseline and after one week. All experimental procedures were conducted in the Musculoskeletal Biomechanics Research Laboratory in the USC Department of Biokinesiology and Physical Therapy. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 35 Loading Conditions: There were a total of 6 loading conditions, P20, P30, P40, NP20, NP30, and NP40 in this study. a. P20: Subject impacted against the padded force platform at a loading distance of 20cm. b. P30: Subject impacted against the padded force platform at a loading distance of 30cm. c. P40: Subject impacted against the padded force platform at a loading distance of 40cm. d. NP20: Subject impacted against the force platform without padding at a loading distance of 20cm. e. NP30: Subject impacted against the force platform without padding at a loading distance of 30cm. f. NP40: Subject impacted against the force platform without padding at a loading distance of 40cm. D ata Collection Procedures of the data collection for anthropometries, loading distance, and impact loading forces were similar to those employed in Pilot Study I, except that the loading distance (Ds-p) varied in this study and all the loading forces were measured at 1200 Hz. a. Anthropometries: Standing height in cm and body weight in N were measured. b. Loading distance: In order to obtain a more reliable measurement of the loading distance, the distance between subject’s heel and the wall (Dh-w) was used (Figure 2). The procedures required measurements of the horizontal distance between heel and lateral styloid process (Dh-s) and the thickness of the force platform (Dp-w). The loading distances Ds-p between styloid process and platform surface were set at 20cm, 30cm, and 40cm. The distances Dh-w between heel and the wall were then calculated using the following formula, which served to standardize the loading distances during data collection. Dh-w = Dh-s + Ds-p + Dp-w R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 36 Force PlatformL J Dh-s Ds-p Dp-w (20, 30, 40cm) Dh-w Figure 2: Loading distance at 20cm, 30cm, and 40cm c. Im pact Loading Forces: Loading forces associated with the exercise were measured using a strain gauge force platform (model OR6-6-1000, AMTI Inc., MA, USA) at 1200 Hz. During data collection, subjects were asked to stand in front of the force platform with their right arm in the following position: shoulder 90° flexed, elbow fully extended, and wrist extended. The subjects were then asked to impact against the wall while maintaining an extended elbow position. Each subject performed the loading regimen 7 continuous times (cycles) for each of the 6 different loading conditions. The force data collected during the 2nd through 6th cycles were then used for analysis. Data Analysis The procedures of data analysis were the same as those used in Pilot Study I. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 37 3.2.2.3 STATISTICAL ANALYSIS Mechanical characteristics of loading forces, including peak load, loading rate, and impulse, were used for analysis. Intraclass correlation coefficients (ICCs) for each repeated measure were generated to assess (1) one-week test-retest reliability using average and ultimate peak load, loading rate and impulse, and (2) within-trial reliability (i.e., among the 5 loading cycles) of the baseline and the one-week retest measurements. Coefficients of Variation (CVs) were also calculated to compare the within-trial and between-trial variability of loading forces to the within-trial and between-trial variability of normalized loading forces. Normalized loading forces are the loading forces divided by the subject’s body weight. The effects of loading distance were analyzed using repeated- measures ANOVA. The null hypothesis was that there would be no significant differences in average peak load, average loading rate and average impulse among the 3 loading distances. Post- hoc tests, including Tukey and Scheffe tests, were used to determine the statistically significant differences among the 3 loading distances. All statistical procedures were conducted using SPSS 10.0 (SPSS Inc., Chicago, IL, USA). 3.2.2.4 RESULTS R epeatability W ithin-trial reliability The ICCs for peak loads, loading rates, and impulses among the 5 loading cycles were high both at baseline and one-week retest, ranging from .89 to .99 (Table 5). The CVs for peak loads, loading rates and impulses were the same as CVs for normalized peak loads, normalized loading rates and normalized impulses. This suggests that the variability of the raw data and normalized data was the same. TableS^ntraclas^orrelayoi^oefficienHryyoadnigJfcmce^r^^oadjn^Iistances Variables Loading Distance (n = 10) Baseline 20 cm 30 cm 40 cm Peak load .9757 .9446 .9731 Loading rate .9466 .9241 .9646 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 38 Table 5 (continued^ Impulse .9931 .9808 .9446 A fter 1 week Peak load .9775 .9644 .9630 Loading rate .9743 .8912 .8913 Impulse .9911 .9893 .9713 One-week test-retest reliability The ICCs for average measures of peak load, loading rate and impulse ranged from .71 to .93 and were higher than those associated with the ultimate measures of peak load, loading rate and impulse (ranged from .63 to .90) (Table 6). These data suggest that the average measures are more reliable than the ultimate measures. The ICCs for average peak loads, loading rates and impulses were slightly higher at loading distances of 20cm (.88-.93) and 40cm (.85-.91), and were lower for the 30cm distance (.71-.92). Similar results were also found in the ultimate measures. These findings indicate that performing the loading exercise at loading distances of 20cm and 40cm produced more reliable loading forces than that at 30cm. The CVs for average measures of peak loads, loading rates and impulses were the same as CVs for average measures o f normalized peak loads, normalized loading rates and normalized impulses. This suggests that the variability of the data was the same no matter the data were normalized or not. Table 6: Intraclass correlation coefficient for loading forces at 3 loading distances Loading Distance (n = 10) V ariables 20 cm 30 cm 40 cm Average peak load .9327 .7096 .9050 Average loading rate .8888 .7159 .8578 Average impulse .9078 .9172 .8758 Ultimate peak load .8998 .6612 .9041 Ultimate loading rate .7913 .7753 .6345 Ultimate impulse .8760 .8867 .8216 Distance effects: Repeated-measure analysis indicated that the results at baseline and afterl-week retest were similar. Therefore, only the 1-week retest data are reported here. The repeated-measures ANOVA results indicated that there were statistically significant differences in average peak loads, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 39 loading rates and impulses among the 3 different loading distances (p<.05) (Table 7-12). Distance- dependent responses in the loading forces were observed, i.e., the average peak loads, loading rates and impulses were the highest at the loading distance of 40cm and the lowest at the loading distance o f 20cm (p<.05) (Figure 3-5). No interactions between the padding conditions and the loading distances were seen. Similar results were also found when analyzing using ultimate peak load, loading rate, and impulse. Table 7: ANOVA table for average peak load Tests of Between-Subjects Effects Dependent Variable: BAVEMAXP Source Type III Sum of Squares df Mean Square F Sig. Intercept Hypothesis 1853589.800 1 1853589.800 198.274 .000 Error 84137.549 9 9348.617a PAD Hypothesis 779.082 1 779.082 2.403 .128 Error 14589.078 45 324.202b DISTANCE Hypothesis 74104.760 2 37052.380 114.288 .000 Error 14589.078 45 324.202b NAME Hypothesis 84137.549 9 9348.617 28.836 .000 Error 14589.078 45 324.202b PAD * DISTANCE Hypothesis 567.156 2 283.578 .875 .424 Error 14589.078 45 324.202b a. MS(NAME) b. MS(Error) Table 8: Post-hoc test results for the distance effect on average peak load BAVEMAXP S u b set DISTANCE N 1 2 3 Tukey HSD a.b 2 20 131.80076 3 20 177.67126 4 20 217.82151 Sig. 1.000 1.000 1.000 Scheffe a.b 2 20 131.80076 3 20 177.67126 4 20 217.82151 Sig. 1.000 1.000 1.000 M eans for groups in hom ogeneous su b se ts are displayed. B ased on Type III Sum of S q u a res T he error term is M ean Square(E rror) = 324.202. a- U ses Harm onic M ean Sam ple Size = 20.000. b. Alpha = .05. N ote: Distance 1 = 20cm , Distance 2 = 30cm, and Distance 2 = 30cm R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 40 Table 9: ANOVA table for average loading rate Tests of Between-Subjects Effects D ependent Variable: BAVERATE S ource Type III Sum of S q u ares df Mean S quare F Siq. Intercept Hypothesis 301440071 1 301440071.4 34.966 .000 Error 77589630.3 9 8621070.038a PAD Hypothesis 917300.780 1 917300.780 2.039 .160 Error 20247972.5 45 449954.944b DISTANCE Hypothesis 17895866.2 2 8947933.114 19.886 .000 Error 20247972.5 45 449954.944b NAME Hypothesis 77589630.3 9 8621070.038 19.160 .000 Error 20247972.5 45 449954.944b PAD * DISTANCE Hypothesis 18239.132 2 9119.566 .020 .980 Error 20247972.5 45 449954.944b a. MS(NAME) b. MS(Error) Table 10: Post-hoc test results for the distance effect on average loading rate BAVERATE Subset DISTANCE N 1 2 3 Tukey H S D ^ 2 20 1559.286 3 20 2268.799 4 20 2896.200 Sig. 1.000 1.000 1.000 Scheffea'b 2 20 1559.286 3 20 2268.799 4 20 2896.200 Sig. 1.000 1.000 1.000 Means for groups in homogeneous subsets are displayed. Based on Type III Sum of Squares The error term is Mean Square(Error) = 449954.944. a- Uses Harmonic Mean Sample Size = 20.000. b. Alpha = .05. Note: Distance 1 = 20cm, Distance 2 = 30cm, and Distance 2 = 30cm R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 41 Table 11: A N O V A table for average impulse Tests of Between-Subjects Effects D ependent Variable: BAVEAUC S ource Type III Sum of S q u ares df M ean S quare F Sig. Intercept Hypothesis 401670.715 1 401670.715 194.655 .000 Error 18571.501 9 2063.500a PAD Hypothesis 3.583E-02 1 3.583E-02 .000 .985 Error 4791.548 45 106.479b DISTANCE Hypothesis 23612.018 2 11806.009 110.877 .000 Error 4791.548 45 106.479b NAME Hypothesis 18571.501 9 2063.500 19.379 .000 Error 4791.548 45 106.479b PAD * DISTANCE Hypothesis 124.555 2 62.277 .585 .561 Error 4791.548 45 106.479b a- MS(NAME) b- MS(Error) Table 12: Post-hoc test results for the distance effect on average impulse BAVEAUC Subset DISTANCE N 1 2 3 Tukey H S D ^ 2 20 57.31687 3 20 82.23949 4 20 105.9036 Sig. 1.000 1.000 1.000 Scheffeab 2 20 57.31687 3 20 82.23949 4 20 105.9036 Sig. 1.000 1.000 1.000 Means for groups in homogeneous subsets are displayed. Based on Type III Sum of Squares The error term is Mean Square(Error) = 106.479. a- Uses Harmonic Mean Sample Size = 20.000. b. Alpha = .05. Note: Distance 1 = 20cm , Distance 2 = 30cm, and Distance 2 = 30cm R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 3: Average peak loads at 3 different loading distances Average Peak Load (N) 20cm 30cm Loading Distance * Significant different from 20cm, p < .05 ** Significant different from 30cm, p < .05 40cm Figure 4: Average loading rates at 3 different loading distances Average Loading Rate (N/sec) 5000 3000 2000 20cm 30cm Loading Distance * Significant different from 20cm, p < .05 ** Significant different from 30cm, p < .05 40cm R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 43 Figure 5: Average impulse at 3 different loading distances 140 120 100 80 | 60 40 20 0 * Significant different from 20cm, p < .05 ** Significant different from 30cm, p < .05 3.2.2.5 CONCLUSIONS Within-trial reliability was high for peak load, loading rate and impulse both at baseline and the one- week retest. The results are similar to those found in Pilot Study 1 . The variability of the loading forces was not affected by the normalization procedure. The one-week test-retest reliability was best for the loading distances of 20cm and 40cm, and the average measures of loading forces were more reliable than the ultimate measures. There were statistically significant differences in average peak loads, loading rates, and impulses among the 3 different loading distances, and dose-dependent responses were seen. These findings suggest that performing the loading exercise at loading distances o f both 20cm and 40cm will produce reliable loading forces, where 40cm will result in significantly higher peak load, loading rate, and impulse than 20cm. Because we expect to see significant changes in bone mass after exercise intervention, the loading distance which induces the Average Impulse * * 20cm 30cm 40cm Loading Distance R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 44 higher loading forces will be selected. Accordingly, a loading distance of 40cm will be used in the intervention study. 3.2.3 Pilot Study III: Reliability of Upper extremity Kinematics 3.2.3.1 INTRODUCTION Purpose: The purpose of this investigation was to determine the within-day and between-day reliability of the upper extremity kinematics associated with DILE. Subjects: Five healthy adults (25-29 yrs) were recruited from the USC Department of Biokinesiology and Physical Therapy. Subjects with musculoskeletal disorders of the upper extremities, such as arthritis, carpal tunnel syndrome and history of Colies’ fracture were excluded. 3.2.3.2 STUDY DESIGN AND M ETHODS This pilot study investigated the reliability of the upper extremity kinematics during DILE. Each subject performed DILE with their right arm at a loading distance of 40 cm. The same procedures were repeated at baseline and after 2-7 days. All experimental procedures were conducted in the Musculoskeletal Biomechanics Research Laboratory in the USC Department of Biokinesiology and Physical Therapy. D ata Collection Procedures of the data collection were similar to those employed in Pilot Studies 1 and II; however, reflective markers were also attached to subject’s bony landmarks in order to obtain the kinematic data. a. Anthropometries: Standing height in cm and body weight in N were measured. b. Loading distance: The loading distance was standardized at 40cm in this study. The measurement procedure of loading distance was the same as Pilot Study I. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. c. Im pact loading forces and kinematics: Both kinematic and kinetic data were collected while subjects performed DILE during the baseline and retest sessions. Loading forces associated with the exercise were measured using a strain gauge force platform (model OR6-6-1000, AMTI Inc., MA, USA) at 1200 Hz. A six-camera motion analysis system (Vicon 370, Oxford Metrics, Oxford, UK) was used to record three-dimensional coordinates o f the exercise arm, including the upper arm, forearm and hand during the impact loading exercise. A total of 11 reflective markers (2.5 cm diameter mounted balls) were firmly taped to the skin above the following positions: C l, T10, mid-point between the two clavicles, xiphoid process, dorsal surface of the third metacarpal head, radial and ulnar styloid processes, medial and lateral epicondyle of the humerus, olecranon, and acromial process. During data collection, subjects were asked to stand in front of the force platform with their right arm in the following position: shoulder 90° flexed, elbow fully extended, and wrist extended. The distance between the force platform and the subject’s wrist joint was 40 cm. The subjects were then asked to impact against the wall once while maintaining an extended elbow during each trial. A total of 3 loading trials were collected. D ata Analysis Marker-coordinate and force data were stored in a motion file generated by the Vicon 370 software. Data processing software, Polygon and BodyBuilder (Oxford Metrics, Oxford, UK) was used to calculate the wrist, elbow, and shoulder joint kinematics (i.e., peak joint angles and joint excursions) during the loading phase of the exercise. Peak joint angle was defined as the maximum angle of each upper-extremity joint (i.e., shoulder flexion, elbow flexion, and wrist extension) in the sagittal plane. Joint excursion was defined as the difference between the maximum and minimum joint angles during the loading event in the sagittal plane. Average peak joint angles were calculated by averaging the peak joint angles among the 3 loading trials collected on the same day. Similarly, average joint excursion was generated by averaging the joint excursions among the 3 loading trials R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 46 collected on the same day. The average measures were then used to assess between-day reliability of the joint kinematics. 3.2.3.3 STATISTICAL ANALYSIS Upper extremity kinematics during the loading event, including peak joint angle and joint excursion o f the shoulder, elbow, and wrist, were used for analysis. Intraclass correlation coefficients (ICCs) for each repeated measure were generated to assess (1) between-day reliability using average peak joint angle and average joint excursion, and (2) within-day reliability (i.e., among the 3 loading trials) of the baseline and the re-test measurements. All statistical procedures were conducted using SPSS 10.0 (SPSS Inc., Chicago, IL, USA). 3.2.3.4 RESULTS Within-day Reliability The ICCs for peak shoulder angle, shoulder joint excursion, peak elbow angle, elbow joint excursion, peak wrist angle, and wrist angle change among the 3 loading trials were high both at the baseline and the retest, ranging from .77 to .99 (Table 13). Table 13: Intraclass correlation coefficient for upper extremity kinematics at baseline and re- test Variables ICCs Baseline Peak shoulder angle .9837 Shoulder joint excursion .9244 Peak elbow angle .9197 Elbow joint excursion .8120 Peak wrist angle .9825 Wrist joint excursion .8336 Re-test Peak shoulder angle .9820 Shoulder joint excursion .9156 Peak elbow angle .9629 Elbow joint excursion .9269 Peak wrist angle .9898 Wrist joint excursion .7658 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Between-day Reliability The ICCs for average peak shoulder angle, average shoulder joint excursion, average peak elbow angle, average elbow joint excursion, average peak wrist angle, and average wrist joint excursion between the baseline and the re-test were high, ranging from .87 to .99 (Table 14). T abl^4^ntraclassa correlado^oefficien^fo^i£ge^x^em it^cinem atics______^^^^_ Variables___________________________________________________ ICCs Average peak shoulder angle .9865 Average shoulder joint excursion .9865 Average peak elbow angle .9086 Average elbow joint excursion .8235 Average peak wrist angle .8729 Average wrist joint excursion_________________________________.9635 3.2.3.5 CONCLUSIONS Within-trial and between-day reliability were high for peak joint angles and joint excursions at the shoulder, elbow, and wrist. These results suggest that DILE is reproducible. Moreover, the errors associated with the data collection procedures, including marker placement and data acquisition, were non-significant. 3.2.4 Pilot Study IV: Repeatability of DEXA Measurement on Forearm 3.2.4.1 INTRODUCTION Purpose: The purpose of this study was to determine the repeatability of DEXA measurements of the radius in healthy adults. Subjects: Five healthy adults (26-31 yrs) were recruited from the USC Department of Biokinesiology and Physical Therapy. Subjects with musculoskeletal disorders of the upper extremities, such as arthritis, carpal tunnel syndrome and history of Colies’ fracture were excluded. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 48 3.2.4.2 STUDY DESIGN AND METHODS This study examined the 7-day repeatability of bone mass measurements of the radius in healthy adults. Bone mass of the right radius was measured using dual-energy X-ray absorptiometry (DEXA, Hologic QDR1500, Waltham, MA, USA) at baseline and after 7 days. Standard procedures were performed by the same research associate. All experimental procedures were conducted in the Exercise Physiology Laboratory located in the USC Department of Biokinesiology and Physical Therapy. Data Collection The subject’s forearm length was measured and marked before being scanned. Subjects were asked to sit on the chair and place the forearm to be scanned on the table in the vertical position. Forearm length was measured from the ulnar styloid process to the table in centimeters. Half the forearm length down from the ulnar styloid process was then marked on the subject’s arm. This was the region being scanned. An x-ray transparent forearm block was placed on the table for forearm positioning. The entire length of the subject’s forearm and hand was then positioned to contact with the forearm block. The subject’s forearm was pronated and the hand was positioned in a loose fist. The elbow was bent at a 105-degree angle measured by a goniometer. The same height chair was used at each measurement. Two positioning aids were placed on the subject’s knuckles and medial surface of the elbow to provide additional fixation to the forearm. Once the subject was properly positioned, the forearm was scanned. The scanning procedure lasted approximately 6 minutes for each forearm. The standard position of arm is crucial since the position of the subject’s hand affects the relative position of the radius and ulna. DEXA measures “areal” bone density (g/cm2 ) which is the amount of bone mineral (grams) in a slice of bone divided by the projected area (cm2 ). Because the radius is not exactly a circle, rotating the arm can change the projected area. Consequently, changes in areal bone density can be resulted and measurement error will be introduced. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 49 Data Analysis Bone mass data at 3 different forearm regions, including 1/3 distal, mid-distal, and ultra distal regions, were generated by the DEXA. The 1/3 distal region is defined as a region 20mm wide, centered at a distance equal to 1/3 of the forearm length measured from the distal tip o f the ulna. This region is primarily cortical bone. The ultra distal is a region nominally 15mm in length positioned proximal to the end plate of the radius. This region excludes the end plate of the radius and is primarily trabecular bone. The mid distal is the region between the 1/3 and ultra distal regions and contains both trabecular and cortical bone. Radial bone mineral density reliability at these regions, measured at baseline and after 7 days, was then assessed using the within-subject coefficients of variation (CVs) between measurements. 3.2.4.3 RESULTS The CVs for 7-day test-retest bone mineral density measurement were 1.2%, 0.9%, and 1.1% at distal 1/3, mid-distal, and ultra distal regions, respectively. 3.2.4.4 CONCLUSION The 7-day reproducibility of the radial DEXA measurements is excellent. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 50 4. METHODS 4.1 OVERVIEW This is a randomized-controlled 6-month intervention study investigating the effects of dynamic impact loading exercise (DILE) on radial bone mineral density (BMD). The exercised arm was allocated randomly to either the dominant side or the none-dominant side with the contralateral side serving as the non-exercised control. In addition, subjects were randomly assigned into either damped or the non-damped treatment arms in order to examine the effects of different loading histories on forearm bone mass. Measurements including anthropometries, self-reported physical activity levels, hand grip strength, radial BMD, and impact reaction forces associated with the exercise were recorded at baseline and after 6 months. All experimental procedures were conducted in the Musculoskeletal Biomechanics Research Laboratory (www.usc.edu/go/mbrl) and the Clinical Exercise Research Center located in the USC Department of Biokinesiology and Physical Therapy. STUDY DESIGN Baseline Study Visit (0 month) Dynamic Impact loading Exercise Final Study Visit (6 month) Break-in Full Dose (3 Weeks) (21 Weeks) • Anthropometries • Self-reported physical activity • Hand grip strength measurement • Bone mineral density measurement • Upper extremity kinematics and reaction force measurement • Instruct dynamic impact loading exercise Exercised arm • All baseline measurement repeated (Damped, Non-Damped) Control arm Exercise Diary R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4.2 SUBJECTS A total of thirty women were recruited in this study. The number of subjects was determined from statistical power calculations (Lawrence Erlbaum Associates Inc., Hillsdale, NJ) based on two factors: 1) bone mass changes associate with exercise intervention; and 2) correlations between bone mass and it’s predictors. Due to the novel nature of the proposed study, data used for power analysis were obtained from the previous studies. To determine the sample size needed to detect exercise- induced bone changes, bone density data at baseline and after intervention in the exercise group reported by Ayalon (1987) were employed. In this study, the effects on distal radial bone density as a result of dynamic loading exercise designed specifically for loading the distal forearm was investigated. The power analysis revealed that a minimum of 20 subjects were required for intervention to detect a significant difference in main effects at the p < 0.05 level (1-p = 0.8). Correlations between reaction forces associated with exercise and bone changes are not available in the literature. However, several studies have reported significant predictors of forearm bone mass, including age, height, weight, lean mass and muscle strength (14, 29, 77,78, 102, 103, 118). The correlations between forearm bone mass and these predictors ranged from 0.2 to 0.5 in women. We believed that the mechanical characteristics of the reaction forces engendered by DILE correlated stronger with radial bone mass than the anthropometric measures do. Therefore, when expecting a correlation coefficient of greater than 0.5, the power analysis indicated that a sample size of 23 was needed for a statistically significant correlation r = 0.5 with an 80% power and alpha 1 tailed set at 0.05. Accordingly, 23 subjects were required in the study. Seven additional subjects were recruited to account for potential losses (drop out) from the study. Subjects were recruited from USC Campus. After hearing a description of the project, candidates completed a self-administered medical history form adapted from a USC IRB previously approved questionnaire (Proposal # 991003) (Appendix A) and a physical activity questionnaire (Appendix B) so that the information R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 52 relevant to inclusions and exclusions were obtained. Female candidates also underwent a pregnancy test immediately before the bone scans to ensure the safety of their participation. This information were reviewed for completeness and eligibility. Eligible and interested subjects were then contacted for an enrollment session at the study site. Inclusion Criteria Healthy women aged from 25 to 45 years were included. Female subjects must have normal menstrual cycles (28-32 days). Exclusion Criteria Individuals were excluded if the following conditions were present: Safety exclusions: cardiovascular disease, musculoskeletal disorders such as arthritis, carpal tunnel syndrome, and history of Colies’ fracture. Bone-related exclusions: women who regularly perform upper-extremity resistance exercises and/or play sports such as volleyball and tennis more than once a week in the previous 2 years; chronic diseases known to affect bone density such as insulin controlled diabetes, uncontrolled hypo- or hyperthyroidism and thyroid disease; current daily use of oral corticosteroids for more than 2 weeks; use of more than 1000 mg of calcium tablet supplementation per day for more than 3 months in the previous year; current pregnancy or have given birth; smoke more than five cigarettes per day; or take more than seven alcohol units per week. 4.3 EXERCISE INTERVENTION Subjects were instructed in the dynamic loading exercise activity on their first visit to the study site. The exercise arm was allocated randomly to either the dominant side or the non-dominant side with the contralateral side serving as the non-exercised control. In addition, subjects were randomly assigned into either damped or non-damped conditions. In the non-damped group, subjects R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 53 performed the impact loading exercise on the wall without any padding material. In the damped group, subjects performed the exercise on the wall covered with a 1/4" thick ensolite closed cell foam pad. Pilot investigations were used to select the foam material in order to vary the mechanical characteristics (i.e., loading rate and magnitude) of the loading prescription. To perform the dynamic loading exercise, subjects were asked to stand in front of a wall with the shoulder o f their exercise arm at 90° flexion, elbow fully extended, and wrist extended. The wrist joint was kept 40 cm from the wall. The pilot investigation indicated that the distance of 40 cm generated relatively reliable and high loading forces during the DILE. The subjects were then asked to impact against the wall while maintaining an extended elbow (Figure 6). The distance between each subject’s heel and the wall were recorded. This was the standard distance for each subject when performing DILE. A research associate instructed and supervised each subject until she performed the task correctly in the study site, i.e., USC Musculoskeletal Biomechanics Research Laboratory. All subjects were then required to perform their exercise on a load-bearing wall (i.e., a wall that structurally supports the roof) so that a stiff loading surface could be ensured. Subjects were suggested to perform the exercise at the same spot of the same wall throughout the study if possible. There was an initial break-in period during which the subject repeated the loading exercise consecutively 10 cycles per day, 2 days per week for week 1; 12 cycles per day, 3 days per week for week 2; and 24 cycles per day, 3 days per week for week 3. Beginning on week 4, the subjects performed the loading exercise 36 cycles per day, 3 days per week, for a total of 6 months. Supervision was not provided over the intervention period because the task was simple and easily reproducible. All subjects were required to maintain their lifestyle including physical activity level, exercise regimen, and diet throughout the length of the study. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 54 40cm Figure 6: Dynamic impact loading exercise 4.4 MEASUREMENTS Anthropometries Standing height was measured to the nearest centimeter with a stadiometer, and weight was measured to the nearest 0.1 kilogram using a balance scale with subjects wearing light clothing and without shoes. Height and weight were collected at the first and last laboratory visits. Additionally, forearm length, elbow width, wrist thickness, and hand thickness were measured to the nearest minimeter with a calipers and a tape measure. The forearm length is the distance between olecranon and ulnar styloid process, which was used to calculate regional bone mineral density. The elbow width is the distance between the medial and lateral epicondyles of the humerus. The wrist thickness is measured at the midpoint between the ulnar and radial styloid processes and hand thickness is measured at the midpoint of the 3rd metacarpal bone. These forearm measurements were used to generate upper extremity kinematics. They were taken at the first laboratory visit. Self-reported Physical Activity Level Self-reported physical activity levels were collected from each subject at the baseline, and after 6 months using a questionnaire adapted from previous studies (4, 56). This physical activity R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 55 questionnaire assessed self-reported upper-extremity activities at home, at work, and during leisure time (Appendix B). The assessment emphasized high-intensity activities which were performed specifically with the upper extremities. Subjects were asked to recall their physical activity level in the previous 6 months at the baseline and after 6 months of intervention. Exercise Diary Subjects were asked to keep a record of their training activities (Appendix C). Also, the research associate phoned and emailed the subjects every 2-4 weeks to facilitate exercise adherence. Hand Grip Strength Hand grip strength was measured on both hands with a JAMAR hydraulic hand dynamometer (Sammons Preston, Bolingbrook, IL) at each study visit. Subjects sat with their shoulder adducted, elbow flexed at 90°, and the forearm and wrist in neutral position (zero rotation) (63). Subjects were then asked to apply as much hand grip pressure as possible. The measurements were repeated three times at each side and the highest strength score was recorded. Radial Bone Mineral Density Bone mineral density at the distal third radius, ultradistal radius, and total distal 1/3 radius for both forearms were measured on each subject using dual-energy X-ray absorptiometry (DEXA, Hologic QDR1500, Waltham, MA, USA) at baseline and after 6 months. The distal third radius and ultradistal radius were chosen because: 1) they have been used repeatedly to assess bone mineral changes in the distal forearm (1, 7, 52), and 2) they allows for the characterization of both trabecular and cortical bone changes; the distal third radius is composed mainly o f cortical bone and the ultradistal radius is composed mainly of trabecular bone (38, 71). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 56 Upper Extremity Kinematics and Reaction Force Both kinematic and kinetic data were collected while subjects performed DILE during baseline and follow-up visits. Reaction forces associated with the exercise were measured on the exercised arm of each subject using a strain gauge force platform (model OR6-6-1000, AMTI Inc., MA, USA). This force platform was secured on the wall, and could be adjusted to fit subjects of various heights. The directions of the reaction forces were shown in Figure 7. Six raw voltage signals collected from the force platform were sampled at 1200 Hz, electronically processed, amplified by an A.M.T.I. Signal Conditioner (model MCA), and converted into digital signals with a sampling rate of 2500 Hz. A six-camera (60 frames/sec) motion analysis system (Vicon 370, Oxford Metrics, Oxford, UK) recorded three-dimensional coordinates of the exercised arm, including the upper arm, forearm and hand during the impact loading exercise. A total of 11 reflective markers (2.5 cm diameter mounted balls) were firmly taped to the skin above the following positions: C l, T10, mid-point between the two clavicles, xyphoid process, dorsal surface of the third metacarpal head, radial and ulnar styloid processes, medial and lateral epicondyles of the humerus, olecranon, and acromial process. Marker-coordinate and force data were stored in a motion file generated by the Vicon 370 software. Data processing software, BodyBuilder (Oxford Metrics, Oxford, UK) and Excel (Microsoft Corporation, USA) were used to calculate the wrist, elbow, and shoulder joint kinematics (i.e., peak joint angles and joint excursions) during the loading phase of DILE. The characteristics of the reaction forces including peak load, impact load, impulse, and peak loading rate were obtained using Peak Fit (SPSS Inc., Chicago, IL, USA). Joint kinematic data allowed the investigators to evaluate the repeatability of the loading regimen and to determine the possible injury mechanisms if any side effect occurred due to repetitive loading. Investigators could also assess the contribution of mechanical factors that cause osteogenesis by analyzing the kinetic data. During data collection, subjects were asked to stand in front of the force platform with their exercised arm in the following position: shoulder 90° flexed, elbow fully extended, and wrist R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 57 extended. The distance between the force platform and the subject’s wrist joint was 40 cm. The subject was then asked to impact against the wall while maintaining an extended elbow. This task was exactly the same as the impact loading exercise that the subjects performed during the 6-month intervention study. Subjects performed this task once during each trial. A total of 3 successful trials were collected. In order to assess the stability of this novel loading exercise, the reaction forces were measured at both baseline and after 6 months. X(+) Wall Force Platform >Y(+) Z(+) Figure 7: Loading force direction R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5. DATA AND STATISTICAL ANALYSIS 58 5.1 DATA ANALYSIS Reaction Force Outcome variables including peak load, impact load, loading rate, and impulse were obtained by Peak Fit (SPSS Inc., Chicago, IL, USA) which allowed us to offset the baseline, identify the start point of each loading cycle and generated the outcome variables. The start point of each loading cycle was defined as the first data point where the reaction force was higher than 5 N in Z direction (perpendicular to the wall). Peak load was calculated as the maximum load generated during each loading cycle. In situations in which there were more than one peak, the peak with the greatest magnitude was used. Impact load was defined as the earliest peak generated within the loading cycle. Loading rate was calculated as impact load divided by the period of time from initial loading to the impact load. Impulse was the area under the force-time curve F(t), which can be calculated using the following formula where X is the time frame at the start and the end points of a loading cycle. Impulse = / x a ! s t a dr t F ( t ) Average peak loads were calculated by averaging the peak loads among the 3 loading trials collected during the same session. The same algorithm was also applied to generate average impact load, average loading rate, and average impulse (please refer to Pilot Study I for detail). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 59 Upper Extremity Kinematics Joint kinematics including peak joint angles and joint excursions of the wrist, elbow, and shoulder (glenohumeral joint) during the loading phase of the exercise were calculated. Peak joint angle was defined as the maximum angle o f each upper-extremity joint (i.e., shoulder flexion, elbow flexion, and wrist extension) in the sagittal plane. Joint excursion was defined as the difference between the maximum and minimum joint angles during the loading event in the sagittal plane. Average peak joint angles were calculated by averaging the peak joint angles among the 3 loading trials collected on the same day. Similarly, average joint excursions were generated by averaging the joint excursions among the 3 loading trials collected on the same day. Radial Bone Mineral Density Bone mineral density (BMD) at the distal third radius (DR), ultra-distal radius (UD), and total distal 1/3 of radius (TOTAL) were measured. The DR is defined as a region 20mm wide centered at a distance equal to 1/3 of the forearm length measured from the distal tip of the ulna. DR contains mostly cortical bone. The UD is a region nominally 15mm in length positioned proximal to the end plate o f the radius. This region excludes the end plate of the radius and contains mostly trabecular bone. All o f the regional measures are default by the manufacture. Physical Activity Level Participant’s physical activity level was assessed in 3 domains, occupational activities, leisure activities, and home activities. A summary score at each domain was calculated as a summation of scores for all questions in that domain; thus, a total of 3 summary scores were generated. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 60 Hand Grip Strength The measurement of hand grip strength was repeated 3 times on each of subject’s arm. The highest score among the three tests was used for analysis. 5.2 STATISTICAL ANALYSIS All statistical procedures were employed using SPSS software (SPSS, Chicago, IL). Statistical significance was defined as an alpha level at or below 0.05. Repeatability of reaction force and upper extremity kinematics were assessed by Intraclass Correlation Coefficients (ICCs). Two sets o f ICCs were generated to test (1) within-session reliability (i.e., among the 3 loading trials) of the baseline and final measurements of impact loading characteristics including peak loads, impact loads, impulses, and loading rates, and (2) within-session reliability (i.e., among the 3 loading trials) o f the baseline and the final measurements of joint kinematics including peak joint angle and joint excursion at shoulder, elbow, and wrist. Paired t-tests were conducted to assess exercise effects, which compared 1) mean radial BMD between the exercised and non-exercised arms at the baseline and final, and 2) mean radial BMD between the baseline and final at both arms. The differences in radial bone adaptations, reaction force characteristics, and upper extremity kinematics between the damped and non-damped groups were tested by independent t-tests. Moreover, paired t-tests were used to detect any statistically significant changes in physical activity level and hand grip strength. One-tailed Pearson’s correlation analyses were used to investigate the relations between the reaction force characteristics including average peak load, average impact load, average loading rate, and average impulse and changes in radial bone measurements. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 61 6. RESULTS 6.1 SUBJECTS A total of 30 healthy females were recruited in this study. Two subjects were unable to keep the exercise schedule, one subject moved out of the state, one subject changed her life style, one subject had Colies’ fracture due to an accident not related to the study, and one subject experienced arm soreness at the first week o f the study and did not want to continue the study. Therefore, a total of twenty-four subjects completed the study. There were 14 subjects in the damped group and 10 subjects in the non-damped group. Table 15 shows the demographic characteristics of the subjects. Thirteen subjects performed the DILE with their dominant arm while eleven subjects performed the exercise with the non-dominant arm. The average exercise adherence rate was 93.3%, where adherence rate was calculated as the number of exercise sessions completed divided by the number of exercise sessions scheduled. Seventy-nine percent of the subjects performed the DILE without any side effects. Five subjects experienced mild to moderate arm soreness at the beginning of the study. The arm soreness disappeared after 3 weeks of performing DILE in 4 subjects who were in the damped group, whereas the soreness in the one subject who was in the non-damped group lasted until the 6th week. No other arm soreness was reported during the 6-month intervention. Seventy-five percent of the subjects performed the DILE at the same wall throughout the study. Four subjects performed the exercise on the same wall except 2-3 occasions. One subject performed 88% of the exercise on the same wall at home and 12% of the exercise on the same wall in her office. One subject performed the DILE in 2 different walls (31% and 69%; respectively) at home due to house remodeling. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 62 All Damped group Non-damped group _________________________________ (n=24)_____________ (n=14)________________(n=10) Age (yr) 29.5 (6.4) 27.5 (4.1) 32.3 (8.0) Height (cm) 163.2 (6.4) 163.4 (6.5) 162.8 (6.6) Weight (kg) 55.7(8.8) 56.1 (9.6) 55.3 (8.1) Exercised arm located at 54.2% 50% 60% dominant side Adherence rate 93.3% 95.5% 90.3% Mean (SD) 6.2 REPEATABILITY 6.2.1 Reaction Force The within-session ICCs for peak loads, impact loads, loading rates and impulses among the 3 loading trials were high, both at the baseline and final, ranging from .90 to .98 (Table 16). Table 16: Within-session intraclass correlation coefficient1 for reaction forces in Z direction at the baseline and final Variables Baseline Final Peak load .9398 .9525 Impulse .9670 .9157 Loading rate .9532 .8989 Impact load .9452 .9815 ICCs were calculated among the 3 loading trials within the same session. 6.2.2 Upper Extremity Kinematics The within-session ICCs for peak shoulder angle, shoulder joint excursion, peak elbow angle, elbow joint excursion, peak wrist angle, and wrist joint excursion among the 3 loading trials were high both at the baseline and the final sessions, ranging from .79 to .99 (Table 17). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 63 Table 17: W ithin-session and between-session intraclass correlation coefficient1 for upper V ariables Baseline Final Peak shoulder angle .9826 .9760 Shoulder join t excursion .8744 .9288 Peak elbow angle .9532 .9205 Elbow join t excursion .8798 .7859 Peak w rist angle .9899 .9873 W rist joint excursion .9102 .8886 ICCs were calculated among the 3 loading trials within the same session. Conclusions The reaction forces and joint kinematics collected at the final measurement session were chosen for further analysis in this study because we believed the reaction forces and joint kinematics measured after the 6-month intervention would be most representative o f the load ing during the intervention. Moreover, the high within-session ICCs observed at the final measurement indicated that the measurements of the reaction force and upper extremity kinematics were highly reliable. 6.3 Physical Activity Level and Hand Grip Strength 6.3.1 Physical Activity Level There were no statistically significant changes in physical activity levels at all three domains after 6 months of exercise intervention (Table 18). TablclSM ft^vsjcayictivitvJcveM nSdjfteren^omains^ Baseline Final P‘ W ork 9.0 (2.1) 9.1 (3.6) .93 Leisure 5.4 (4.2) 4.8 (5.3) .55 Home 10.8 (2.9) 10.0(3.8) .34 SD (mean) N = 23 1 Paired t-tests compared baseline and final R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 64 6.3.2 Hand Grip Strength Hand grip strength in the exercised arm increased statistically significantly by 7.9% (p < .001) after 6 months of exercise intervention (Table 19). Similarly, hand grip strength in the non-exercised arm increased statistically significantly by 4.8% (p < .05) after 6 months. No statistically significant differences in hand grip strength were found between the exercised arm and the non-exercised arm at both baseline and final measurements (p = .775 and p = .388, respectively). T a b l^ 9 ^ a n d i^rigjtrengtl^n^xercisedjand_nomexerase^iarms^tiftiea baseline^n^ina^ Exercised Arm Non-exercised Arm P2 Baseline (Kg) 30.2 (4.5) 30.4 (4.0) .78 Final (Kg) 32.6 (4.9) 31.8 (4.5) .39 P' p c .0 0 1 p < .05 Mean (SD) N = 24 1 Paired t-tests compared baseline and final 2 Paired t-tests compared exercised arm and non-exercised arm 6.4 BONE ADAPTATIONS AFTER 6 MONTHS OF DILE INTERVENTION Radial bone mineral density of the exercised and non-exercised arms at the baseline and final measurements were within normal range (Table 20). There were no statistically significant differences in mean radial bone measurements of the exercised arm after 6 months of exercise intervention. There were also no statistically significant differences in mean radial bone measurements of the non-exercised arm after 6 months. Similarly, when comparing radial bone measurements between exercised and non-exercised arms, no statistically significant differences were observed at both baseline and final (p ranged from .31 to .95). The differences in damping effects on radial bone adaptations after 6 months were further tested by independent t-tests in the exercised arm. The results showed no statistically significant differences R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 65 in radial bone changes after 6 months between the damped and non-damped groups at all regions (Table 21 and Figure 8-10). Following this initial analysis, we conducted post-hoc, independent t-tests to investigate the possible influences of oral contraceptives (OC) on the radial bone mineral density changes associated with DILE intervention. The results indicated that there were no statistically significant differences in mean radial BMD changes of the exercised arm after 6 months of exercise intervention between OC users and non-OC users (Table 22). There were also no statistically significant differences in mean radial BMD changes of the non-exercised arm after 6 months between the 2 OC groups (Table 23). Paired t-tests were further utilized to compare radial BMD between exercised and non-exercised arms at both baseline and final measurements. No statistically significant differences in radial BMD between exercised and non-exercised arms were observed in both OC group and non-OC group (p ranged from .26 to .84). When further examining the 18 subjects in the OC group, there were no statistically significant changes in radial BMD after 6 months of exercise intervention in both exercised arm and non-exercised arm (Table 24). In order to avoid possible statistical errors associated with examination of the numerous DEXA outcome variables, we limited our a priori analysis to the BMD data alone (a total of 36 BMD variables). Following this initial analysis and discussions with committee members, however, a post hoc analysis, that included an examination of bone mineral content (BMC) and area, was performed. This analysis is presented in Appendix D. Table 20: Radial BMD of exercised and non-exercised arm s at the baseline and final Exercised Non-exercised M easurem ents1 Baseline Final P2 Baseline Final P3 A ll (n=24) DR BMD .680 (.039) .676 (.042) .20 .678 (.041) .676 (.046) .47 UD BMD .453 (.059) .454 (.061) .63 .454 (.050) .452 (.049) .22 TOTAL BMD .570 (.043) .567 (.046) .18 .567 (.041) .565 (.044) .08 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 66 Mean (SD) 1 BMD = bone mineral density (g/cm2 ), DR = distal third radius, UD = ultra-distal radius, and TOTAL = total distal third of radius 2 Paired t-tests compared BMD at exercised arm between baseline and final 3 Paired t-tests compared BMD at non-exercised arm between baseline and final Measurements1 Damped (n =14) Non-damped (n = 10) P3 DR BMD Baseline .676 (0.038) .686 (0.043) Final .673 (0.034) .680 (0.053) Difference2 -.003 (.015) -.005 (.016) .87 UD BMD Baseline .448 (0.046) .460 (0.076) Final .447 (0.046) .463 (0.079) Difference -.001 (.009) .003 (.009) .91 Total BMD Baseline .565 (0.034) .577 (0.054) Final .561 (0.033) .575 (0.060) Difference l r , m 1 _____ ■ ___ 1 -.004 (.010) -.002 (.010) .97 TOTAL = total distal third of radius 2 Difference = Final - Baseline 3 Independent t-tests compared BMD changes after 6 months between the damped and non-damped groups Figure 8 : DR BM P 1 Percent change of exercised arm at damped and non-damped groups Exercised Arm DR 4.000 2.000 0.000 2.000 4.000 6.000 8.000 ♦ Damped ■ Non-damped 1 BMD = bone mineral density, DR = distal third radius R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 67 Figure 9: UD BM P 1 Percent change of exercised arm at damped and non-damped groups Exercised Arm UD 0 O ) c r e j= u 0 ) Q. C O - 4.000 3.000 2.000 1.000 0.000 1.000 2.000 3.000 4.000 5.000 t ♦ 4 1 1 ♦ Damped ■ Non-damped 1 BMD = bone mineral density, and UD = ultradistal radius Figure 10: TOTAL BMD Percent change of exercised arm at damped and non-damped groups Exercised Arm TOTAL o O ) c re m 4.000 3.000 2.000 1.000 0.000 - 1.000 - 2.000 -3.000 -4.000 ♦ t J L ■ * X ♦ Damped ■ Non-damped BMD = bone mineral density, and TOTAL = total distal third o f radius R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 68 Table22^RadialBM D^^xercisedarm[at< OC!and_non^OC^rou2s Measurements1 OC (n =6 ) Non-OC (n = 18) P3 DR BMD Baseline .683 (.035) .679 (.042) Final .677 (.030) .675 (.046) Difference2 -.006 (.013) -.004 (.016) .78 UD BMD Baseline .439 (.054) .458 (.061) Final .439 (.056) .459 (.063) Difference .000 (.007) .001 (.009) .78 Total BMD Baseline .566 (.024) .571 (.048) Final .561 (.025) .569 (.051) Difference -.005 (.009) - . 0 0 2 (.0 1 0 ) .47 | 1 1 1 1 1 K 1 ............. BMD = bone mineral density (g/cm ), DR = distal third radius, UD = ultra-distal radius, and TOTAL = total distal third of radius 2 Difference = Final - Baseline 3 Independent t-tests compared BMD changes after 6 months between the OC and non-OC groups Tabl^IS^Radial^BMD^ofnon^cerdsedjrm^^^^and^nomOC^grougs Measurements1 OC (n =6 ) Non-OC (n = 18) P3 DR BMD Baseline .675 (.018) .679 (.046) Final . 6 6 6 (.0 2 1 ) .680 (.052) Difference2 -.009 (.017) . 0 0 0 (.0 1 2 ) .14 UD BMD Baseline .440 (.040) .459 (.053) Final .436 (.036) .458 (.052) Difference -.004 (.009) - . 0 0 1 (.006) .37 Total BMD Baseline .559 (.015) .570 (.047) Final .555 (.015) .568 (.050) Difference i ,___ • -.005 (.008) , •. , . L .... -.002 (.007) .42 BMD = bone mineral density (g/cm ), DR = distal third radius, UD = ultra-distal radius, and TOTAL = total distal third of radius 2 Difference = Final - Baseline 3 Independent t-tests compared BMD changes after 6 months between the OC and non-OC groups R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 69 Tabje24^Ra{HaniMDj)fexercise(yuidnon^exercisecnM|m^^ Exercised Non-exercised M easurem ents1 Baseline Final P2 Baseline Final P3 DR BMD .679 (.042) .675 (.046) .36 .679 (.046) .680 (.052) .89 UD BMD .458 (.061) .459 (.063) .61 .459 (.053) .458 (.052) .51 TO TA L BMD .571 (.048) .569 (.051) .44 .570 (.047) .568 (.050) .23 Mean (SD) N=18 1 BMD = bone mineral density (g/cm2 ), DR = distal third radius, UD = ultra-distal radius, and TOTAL = total distal third of radius 2 Paired t-tests compared BMD at exercised arm between baseline and final 3 Paired t-tests compared BMD at non-exercised arm between baseline and final 6.5 REACTION FORCE CHARACTERISTICS Reaction force characteristics including peak loads, loading rates, impulses, and impact loads in X, Y, and Z directions were highly correlated. Average peak loads in the X and Y direction were 14.8% and 12.0% of the average peak loads at Z direction (i.e.; perpendicular to the wall); respectively. Shear forces in the X and Y directions were independent of the side of the exercised arm. Most of the peak loads occurred at positive X direction whereas only half of the peak loads occurred at positive Y direction. Table 25 shows the reaction forces perpendicular to the wall associated with DILE. The overall average peak load was 43.2 ± 21.7%BW and impact load was 38.9 ± 23.7%BW. T ablelS^ergendici^rjreacB o^fbrcesjdurin^th^loadin^has^^IL E All Damped group Non-damped group P‘ (n=24) (n=14) (n=10) Peak load (N) 242.8 (138.5) 212.2 (71.2) 285.7 (195.5) .21 Im pact load (N) 219.7(149.0) 186.6 (80.1) 266.1 (208.4) .21 Loading rate (N/sec) 6512.7 (10589.5) 4631.7 (4236.6) 9146.2 (15730.2) .31 Impulse (N*sec) 116.1 (29.2) 113.7 (25.9) 119.4 (34.4) .65 Mean (SD) 1 Independent t-tests compared damped and non-damped groups R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 70 6.5.1 The Influence of Damping Condition on Reaction Force Characteristics Independent t-tests indicated that the reaction force characteristics including peak loads, impact loads, loading rates, and impulses did not significantly differ between the two damping groups (Table 25). This was likely because of the large variation among subjects. These findings remained the same when the reaction force characteristics were further normalized by body weight. The average peak load was 37.6 ± 9.2%BW and average impact load was 33.1 ± 11.3%BW in the damped group. In the non-damped group, average peak load was 50.9 ± 31.0%BW and average impact load was 46.9 ± 33.6%BW. 6.6 ASSOCIATIONS AMONG REACTION FORCE CHARACTERISTICS AND BONE ADAPTATIONS Pearson’s one-tailed correlation analysis indicated statistically significantly positive correlations between changes in BMD (ABM D) and peak loads at all regions of the exercised arm, including distal third radius (DR), ultra-distal radius (UD), and total distal 1/3 of radius (TOTAL) (Table 26). Similarly, significantly positive correlations between ABM D and impact loads were also observed at all measurement regions. Furthermore, there was a statistically significantly positive correlation between UD ABM D and impulse (r = .401, p < .05). No statistically significant correlations between loading rates and changes in radial BMD were identified at any region; however, a marginally significantly positive correlation was identified at DR (r = .355, p = .045). Table 26: Correlations between reaction force characteristics associated with PILE and changes in radial bone density Peak Load Impulse Loading Rate Impact Load A BM D DR .425* .194 .355* .420* UD .437* .401* .191 .360* TOTAL .546** .315 .306 4 9 7 ** * p < .05; ** p < .01 ABM D = changes in bone mineral density R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 71 6.6.1 The Influence of Damping Condition on the Associations among Reaction Force Characteristics and Bone Adaptations Although the differences in reaction force characteristics were not statistically significant between the 2 damping groups, the relations between the reaction force characteristics and changes in radial bone mass differed with the damping condition. In the damped group, no statistically significant correlations were observed between the reaction force parameters and changes in bone measurements except a positive correlation between the peak loads and changes in UD BMD (r = 491, p < .05) (Table 27). In the non-damped group, ABM D significantly highly correlated with peak loads and impact loads at DR and TOTAL, r ranged from 0.733 to 0.818 (Table 28 and Figure 11-14). Moreover, changes in BMD also significantly highly correlated with impulse at TOTAL (r = .745; p < .01) and loading rate at DR (r = .615, p < .05). Following this initial analysis, we conducted a secondary analysis using a multiple linear regression model to further examine whether or not changes in body weight (A B W ) and/or changes in hand grips strength (A G rip) significantly contribute to the prediction of changes in BMD. The results indicated that neither A B W nor A Grip significantly correlated with changes in BMD at any region (p = .88 - .29). Consequently, the reaction force characteristics were the only significant predictors to the changes in BMD after considering A B W and AGrip. Hence, the results of the regression analysis were similar to those of the correlation analysis. The data o f this secondary regression analysis are presented in Appendix D for further references. Table 27: Correlations between reaction force characteristics associated with PILE and changes in radial bone density in the damped aroup Peak Load Impulse Loading Rate Impact Load A BM D DR .059 -.029 -.031 -.001 UD .491* .316 .242 .330 TOTAL .202 -.086 .079 .068 * p < .05; ** p < .01 ABM D = changes in bone mineral density R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 72 Table 28: Correlations between reaction force characteristics associated with PILE and changes in radial bone density in the non-damped group Peak Load Impulse Loading Rate Impact Load ABM D DR .733** .433 .615* .761** UD .448 .484 .171 .389 TOTAL .818** .745** .450 .807* * p < .05; ** p < .01 A BM D = changes in bone mineral density Figure 11: Correlations between peak load and changes in BMD at DR Changes in BMD at DR and Peak Load o o > a > O ) c ( 0 ■ E o Q 2 C D .030 .020 r = 733’ .010 r =.059 .000 700 N 400 500 600 -.010 ♦ ■ -.020 ♦ damped ■ on-damped — Linear (on-damped) Linear (damped) -.030 -.040 p < .0 1 Figure 12: Correlations between peak load and changes in BMP at TOTAL Changes in BMD at TOTAL and Peak Load E o s < / > <D O ) c .c C O o Q S C Q .030 .020 r =.818: .010 a 8 8 ♦ r =.202 .000 ♦400 7 (1 )0 N 100 300 500 600 -.010 ♦ damped ■ on-damped ■ — Linear (on-damped) Linear (damped) -.020 -.030 p c .O l R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 73 Figure 13: Correlations between impact load and changes in BMD at DR C h a n g es in BMD a t DR an d Im pact Load .030 .020 r =.761 g j .010 o O ) „ .000 0 o > 1 -.010 o Q 5 -.020 m -.001 600 400 500 ♦ d am p ed ■ o n -d am p ed L inear (on-dam ped) L inear (dam ped) -.030 -.040 ** p < .01 Figure 14: Correlations between impact load and changes in BMD at TOTAL C h a n g e s in BMD a t TOTAL an d Im pact Load .025 .020 .015 r =.807' E .010 _ o 3 .005 ( / > C D O ) c re ■ = -.005 Q s -.010 m -.015 f^=.068 .000 600 700 N 500 ♦ d am p ed ■ on-d am p ed — L inear (on-dam ped) — L inear (dam ped) -.020 -.025 * p < .05 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 74 6.7 JOINT KINEMATICS Table 29 contains the peak joint angle and joint excursion data of the shoulder, elbow, and wrist during D1LE. Subjects flexed their shoulder and elbow, and extended their wrist, apparently to absorb the reaction forces during DILE. There were no statistically significantly differences in joint kinematics between the 2 damping groups. TabI^9^jgge^x^emit^^inematicsi ^^in^theJoadingsg^s^o^DlLE Variables (degrees) All (n=24) Damped group (n=14) Non-dam ped group (n=10) P1 Peak shoulder angle 112.1 (7.6) 112.4 (8.1) 111.6 (7.3) .79 Shoulder join t excursion 9.2 (3.4) 9.3 (2.9) 9.0 (4.1) .83 Peak elbow angle 23.3 (7.8) 25.4 (8.2) 20.4 (6.5) .12 Elbow joint excursion 14.9 (4.7) 15.3 (4.3) 14.4 (5.4) .68 Peak w rist angle 63.1 (5.4) 62.8 (5.3) 63.6 (5.7) .73 W rist joint excursion 22.1 (6.3) 21.9 (6.6) 22.5 (6.2) .84 Mean (SD) 1 Independent t-tests compared damped and non-damped groups R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 75 7. DISCUSSION 7.1 SUMMARY This study investigated the effects of 6-month dynamic impact loading exercise on radial BMD in healthy women and determined the relations among the mechanical characteristics of the loading events and the bone changes. Participants demonstrated a high adherence rate to the prescription and no injuries associated with the exercise program were reported. Moreover, the loading profile and joint kinematics generated during DILE were highly reproducible. Although there were no statistically significant bone adaptations observed after 6 months of DILE intervention, the correlations between the radial BMD changes and mechanical characteristics of loading, particularly peak loads and impact loads, were statistically significant. Moreover, peak load and impact load were independent predictors of changes in BMD. Damping condition played a significant role in varying the loading history, which is evidenced by the differences in the relations between the reaction force characteristics and changes in radial BMD between the two different damping groups. More specifically, in the non-damped group, changes in BMD were strongly correlated with peak loads and impact loads at the DR and TOTAL measurement sites. Further, changes in BMD were also strongly correlated with impulses at TOTAL and loading rates at the DR sites. In contrast, only peak loads and UD BMD changes were modestly correlated in the damped group. DILE appears to be a valid and reliable upper extremity musculoskeletal loading model using humans, and results of correlation analyses parallel findings of previous bone adaptation models employing animals. Findings suggest that further research, investigating the associations between mechanical loading events and human bone health, is warranted. These studies should characterize the longer-term adaptations associated with DILE and the effects of DILE in other participant samples (e.g. older adults). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 76 7.2 PARTICIPATION AND ADHERENCE Our study participants demonstrated a 93.3% adherence rate in performing DILE over the 6-month period, suggesting that the unobtrusive nature of the DILE may prompt long-term adherence. Although this initial investigation used young adult women participants, we believe the greatest benefits of DILE may be experienced by older adults. Exercise adherence rate is a particularly important consideration when designing osteoporosis prevention programs for older adults, because poor adherence rates are commonly reported for exercise programs in older people (5). One can expect a dropout rate in exercise programs for adults of around 50% within 6 to 12 months (27). Furthermore, it has been suggested that the dropouts primarily comprise those participants who are most in need o f the exercise program to begin with (28). In order to facilitate exercise adherence, Barry and Eathorne (1994) have suggested that exercise should be low to moderate intensity, fun, interesting, convenient, effective and safe. Dynamic impact loading exercise appears to possess these characteristics because: 1) it is relatively easy to learn, 2) it can be performed in the home without specialized equipment, 3) it can be performed without supervision, 4) it takes approximately 3 min to complete each exercise session, 5) it can be performed relatively safely because no side effects due to DILE were reported except occasional arm soreness during the initial break-in period of the study, and 6) it can be performed using varying intensity (e.g. participants can perform the DILE with pad or without pad attached to the wall, and change the loading distance). Moreover, although DILE is novel in nature, we believe it is appropriate to introduce the program to older adults because it is time-efficient, safe, and is associated with a high adherence rate in young women. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 77 7.3 REPRODUCIBILITY OF REACTION FORCES AND JOINT KINEMATICS Our results indicated that the upper-extremity kinematics and reaction forces engendered during the DILE were reproducible even though the exercise regimen was novel in nature. The within-session ICCs ranged from .90 to .98 for reaction force characteristics, including peak loads, impact loads, loading rates, and impulse. Similar to the reaction forces, the within-session upper extremity kinematics were repeatable, with ICCs ranging from .79 to .99. The elbow joint excursion was least repeatable. This greater variability may be the result of skin movements between the reflective markers, and the bony humeral epicondyles and olecranon process. Target marker movement artifacts are a common source of errors associated with motion analysis. In the present study, subjects extended their elbow during the loading phase of DILE. In the extended position, the skin covering the dorsal elbow (olecranon process) is extremely loose; thus, marker movement artifacts are likely to be greatest in this joint position. 7.4 HAND GRIP STRENGTH AND PHYSICAL ACTIVITY LEVEL Our subjects demonstrated significant increases in hand grip strength by 7.9% in the exercised arm and 4.8% in the non-exercised arm. However, no statistically significant differences in hand grip strength were found between the exercised arm and the non-exercised arm at both baseline and final measurements. The hand grip strength was measured at a position that subjects sat with their shoulder adducted, elbow flexed at 90°, and the forearm and wrist in neutral position. Also, verbal encouragement was employed during testing. This standard position and procedure are reported to be highly reliable in the literature (43, 63). Moreover, our hand grip strength data were comparable to normative data (63). No physical activity level changes over the 6 months were reported. The physical activity level was evaluated in 3 domains, home, work, and leisure, and upper-extremity R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 78 specific activities were emphasized. These findings suggest that subjects did not alter their physical activity patterns during the study period. Moreover, the differences in grip strength performance between measurement periods were likely the result o f the learning effects as both the exercised arm and non-exercised arm experienced increases in grip strength, but no statistically significant differences were observed between the exercised arm and the non-exercised arms. 7.5 BONE ADAPTATIONS AFTER 6 MONTHS OF DILE INTERVENTION There were no statistically significant differences in mean radial bone measurements of exercised arm as well as non-exercised arm after 6 months o f exercise intervention. The lack of bone adaptation in the exercised arm may be due to the relatively short duration of the intervention, the large variation in reaction forces engendered during the DILE among the subjects, or a combination of factors. Bone undergoes remodeling through 4 major phases including resorption, reversal, resting, and formation. The pattern of cell appearance at a remodeling site is activation of osteoclasts followed by resorption of the existing bone. Osteoblasts then deposit osteoid (non-mineralized new bone) on the resorption site. All of these events occur within the same anatomical site and the linkage of resorption and formation is very tight, usually referred as “coupling”. The average normal duration of a resorption site in humans is about 4 weeks (83). Nevertheless, the functional lifespan of the team o f osteoblasts at a given bone remodeling site may ranged from 3-4 months to 1.5 years, with an average o f 5-6 months. It requires approximately 6 months to complete a remodeling cycle in trabecular bone and the time sequence for compact bone is longer. However, whether or not the time required to complete each remodeling cycle differs with different skeletal sites is still unknown. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 79 Previous studies, which included exercise programs with high magnitudes of load/strain and novel strain distributions, have demonstrated significant bone changes after 5-6 months of exercise intervention in both the upper extremity and lower extremity (1, 3, 6, 8, 100). Ayalon (1987) and Simkin (1987) reported a 3.8% increase in radial bone mass after 5 months of upper-extremity exercise intervention in 14 women aged 53-74 years. The exercise program was thought to generated varied loading patterns including tension, torsion, compression, and bending on the forearm, suggesting that radial bone responds rapidly to external loads that are diverse and novel in nature in postmenopausal women. Adami (1999) suggested that exercises may reshape the bone segment under mechanical stress although the changes in bone mass were not obvious after a 6- month exercise program. The exercise was designed to maximize stresses on the wrist. Peripheral quantitative CT (pQCT) was used to assess both trabecular bone and cortical bone at ultradistal and proximal radius in 250 postmenopausal women. No statistically significant changes in bone measurements were observed at the proximal radius. At the ultradistal radius, there were statistically significant increases in BMD (2.2%) and BMC (3.1%) of cortical bone and decreases in BMD (2.6%) and BMC (3.4%) in trabecular bone although the total BMC did not change significantly. Bassey (1998) reported significant increases in BMD within the lumbar spine, femoral neck and femoral trochanter after 5 months of vertical jumping exercise in premenopausal women. The exercise consisted of 5 bouts of 10 vertical jumps, 6 days a week (mean jumping height = 8.5 cm) and the ground reaction forces produced by the jumping exercise were 3 times body weight. Similar results were also reported by Bassey (1994). Here, 27 healthy premenopausal women were randomly assigned to either a control group that participated in low-impact exercise or a test group that participated in intermittent high-impact exercise. Bone mineral density at the femur and lumbar spine was assessed at 0, 6, and 12 months. At 6 months, the test group (n = 14) demonstrated a significant increase in trochanteric bone density (3.4%), a result significantly different from the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 80 control group. In the second 6 months, the control group was crossed over to high-impact exercise (n = 7) and showed a significant increase in trochanteric bone density (4.1%), while the original group maintained its improvement relative to the baseline. Our results were similar to Adami’s study (1), in that no changes in bone mass were observed after 6months of exercise intervention. Nevertheless, the radial BMD in this study was measured by DEXA, not pQCT. DEXA has been widely used in the clinic because of its numerous advantages over the traditional isotope-based absorptiometry (30, 38, 71). The x-ray tube of the DEXA is capable o f producing a higher radiation flux than the isotope source; thus, short- and long-term precision is greatly enhanced. Moreover, the increased output of the x-ray tube as compared to the isotope source has also reduced the scanning time and increased image resolution. The radiation dose associated with the DEXA measurement is low and there is no need for frequent changes in radiation source. Unfortunately, the three-dimensional distribution of the bone structure can not be detected by DEXA. PQCT can determine both bone density and distribution in three dimensions at any skeletal site. However, the pQCT has the disadvantage of relatively high radiation dose, high cost, and requirement of specific software and operation techniques, which were not appropriate for our experimental setting. The internal structure of the radius in our study might have been re arranged, but was not observed due to equipment limitation of the DEXA. We further analyzed BMC and area changes associated with DILE intervention, attempting to examine possible bone geometry changes. The results indicated no statistically significant changes in BMC and area of both the exercised and non-exercised arm after 6 months of exercise intervention except for a marginally significant 0.6% decrease in DR BMC (p = .04). This may suggest that the DR region experienced increased bone turnover and/or increased non-mineral ized bone component. However, a change of 0.6% was within the measurement error of the DEXA. Longer-term study with advanced image techniques is needed to further conclude the effects of DILE on bone geometry. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 81 Another possible reason for no significant changes in radial BMD after 6 months of exercise intervention is the large variation of reaction forces engendered during DILE. Participants performed the exercise with the peak loads ranging from 25% to 115% BW and the loading rates ranging from 3 to 91 times BW per second. It has been shown that the anabolic responses of bone occur only if the mechanical loads surpass a threshold (20, 90, 111). Rubin and Lanyon (1985) reported that applications o f mechanical strains greater than 1000 pstrain (100 cycles/day) were associated with periosteal and endosteal new bone formation. However, when the applied peak longitudinal strains were lower than 1000 pstrain, bone loss occurred by increased remodeling activity, endosteal resorption, and increased intra-cortical porosis. Similarly, Turner (1994) demonstrated that lamellar bone formation on the endocortical surface of rat tibiae increased linearly with increasing load for applied dynamic bending loads above 40N. No increase in bone formation was observed for applied bending loads less than 40N. If the threshold effect existed in our subjects, the effects of DILE on radial BMD may be underestimated due to the relatively large variation in loading forces. In fact, our results of regression analysis indicated that a minimum of 200, 250, and 400 N peak load at UD, TOTAL, and DR; respectively, was required for increases in radial BMD after 6 months of intervention in our study. This issue was also addressed by examining the relations between DILE loading events and BMD changes using correlational analysis. 7.6 ORAL CONTRACEPTIVES EFFECTS The use of oral contraceptives (OC) is associated with reduction of endogenous sex steroids (10) and suppression of bone remodeling activity (81). Consequently, OC use may influence the effects of exercise on bone mineral density. This hypothesis; however, is still debatable due to limited evidence. Hartard (1996) conducted a cross-sectional study to examine the effects of low-dosed oral contraceptives and physical activity on bone mineral density in 128 women aged 20 to 35 years old R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 82 (44). Subjects were assessed by questionnaire and interview and assigned to 4 groups according to their years of exercise and OC use. BMD for L2-4 and the femoral neck was evaluated by DEXA. The results indicated that subjects in the long-term exercise and short use of OC (group A) had the highest BMD values whereas the BMD values were statistically significant less than the group A in the rest of the 3 groups, long exercise and long OC use (group B), short exercise and long OC use (group C), and short exercise and short OC use (group D). Moreover, the BMD values for group B were not statistically different from values for group C and D. This suggests that after long-term intake of OC (>8 years), the physical active women in group B had similar BMD values to those less active women in group C and D. Therefore, long-term use of low-dose OC might counteract the beneficial influence of physical activity on bone formation. However, long-term physical activity was beneficial for bone health in women who took OC less than 2 years. Weaver et al (2000) investigated the effects of 2-year exercise on bone mass as modified by OC use in 141 women aged from 18 to 31 years (120). The exercise program consisted of resistance training at 70% 1RM for 3 sessions/wk and 60 min/wk of jumping rope. The results showed that total body BMC percent change from baseline was statistically significant greater in the exercise group than that in the non-exercise group at 6 and 24 months. When comparing subjects with OC use and without OC use, OC user had statistically significantly lower bone turnover at baseline and a decrease in total body BMC from baseline at 24 months. In addition, spine BMC and BMD in the exercise and OC use group decreased statistically significantly at 6 months and remained significantly below non-exercisers who used OC at 24 months. Further, femoral neck BMD also decreased in the exercise and OC use group at 6 months. These findings suggest that exercise had a beneficial effect and OC use had a negative effect on total body BMC. Moreover, combination of exercise and OC use compromised attainment of spine BMC and BMD observed in the non- exerciser plus OC group. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 83 Although our study was not designed to investigate whether or not the OC use could modify the effects o f exercise on bone density, the question could be examined by splitting the subjects into OC users and non-OC users. Among the 24 subjects in our study, 6 subjects were OC user and 18 of them were non-OC user. The independent t-tests revealed that there were no statistically significant differences in mean radial BMD change of exercised arm after 6 months of exercise intervention between the 2 OC groups. Similarly, when comparing radial BMD between exercised and non- exercised arms at both baseline and final, no statistically significant differences between OC users and non-OC users were observed. Moreover, when further examining the 18 subjects in the OC group, there were no statistically significant changes in radial BMD after 6 months o f exercise intervention in both exercised arm and non-exercised arm. These results suggest that our subjects’ responses to the 6-month exercise stimuli were similar whether or not they used OC. However, studies with longer duration and greater sample size are needed to conclusively understand the effects of OC on DILE intervention. 7.7 CORRELATIONAL ANALYSIS We found significantly positive correlations between changes in BMD and peak loads, as well as impact loads, at all measurement regions of the exercised arm. Also, there was a significant positive correlation between UD BMD changes and impulses. Moreover, peak loads and impact loads were independent predictors o f changes in BMD. These results were consistent with the literature where there was a dose-response relation between the peak strain magnitude and change in the mass of bone (80, 90). Qin (1998) reported a strong correlation between the change in cortical bone area and longitudinal normal strain (r = .91), as well as longitudinal normal stress (r = .92) and strain energy density (r = .92). The study evaluated the role o f daily stress stimulus on maintaining bone mass using R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 84 functionally isolated turkey-ulna model. Changes in bone mass were compared between an 8-week disuse group and a group of disuse plus daily exposure to mechanical loading. Animals in the loading group were subjected to a bending load of 9 N at 30 Hz for 108,000 cycles, approximately 60 min, per day. They found that mechanical loading resulted in a significant inhibition of bone loss observed in the disuse group. The predicted threshold for maintaining bone mass was 68 microstrain for this mechanical loading regimen. Similarly, Rubin and Lanyon (1985) demonstrated a high correlation, r = .83, between changes in cross-sectional area of ulna midshaft and longitudinal strains. The osteogenic responses to different magnitude of mechanical loading were examined over an 8-week period using turkey ulna model. Each bird was subjected to 100 consecutive 1 Hz load cycles of one of the following stain stimuli: 0, 500, 1000, 1500, 2000, 3000, or 4000 microstrain. The loading rate and magnitude of engendering strains within the bone tissue were within physiological range. The birds were allowed freedom of movement in large pens during the rest of the day. A graded dose-response relations between the peak strain magnitude and changes in bone tissue were evidenced in this study. There was a significant correlation between loading rates and changes in BMD at DR found in the present study. This result was consistent with the literature in that loading rates is an important factor for osteogenesis (69, 112). Turner (1995) suggested that the amount of new bone formation following 2 weeks of mechanical loading was directly proportional to the rate of strain experienced in the rat tibiae. In this study, the animals were divided into 4 groups with different strain rates, including 0, 0.013, 0.026, and 0.039 s'1 . The applied loads were 54 N, 2Hz, and 36 cycles/day for all groups. Bone formation and mineral apposition rates were measured at baseline and after 2 weeks. The results suggested that the relative bone formation rate in the 2 highest strain rate groups was significantly greater than that in the group with the lowest strain rate. Moreover, The group R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 85 with the highest strain rate had significantly higher relative mineral apposition rate than all other groups. O ’Conner (1982) also investigated the role of strain rate in new bone formation. Bending and compressive loads were applied intermittently at 0.5 Hz through implants permanently inserted into the radius and ulna of experimental sheep. The ratio between the maximum strain rate of the artificial loads and the maximum strain rate during walking, which had the greatest influence on every remodeling parameter, was used for analysis. The results showed that the variation in this ratio accounted for 68-81% of the variation in surface bone formation and 43% of the variation in intracortical remodeling. Only an additional 6-12% variation in surface bone formation could be explained by adding axial strain into the equation, suggesting that the effect of axial strain was less marked than that of strain rate. Also, the direction of bending and axial loading (tension or compression) appears to have had no effect on the course o f the remodeling observed. The lack of significant correlations between loading rates and changes in BMD at UD and TOTAL measurement sites may be associated with large variation of the loading rates in the present study. In addition, the UD contains mostly trabecular bone whereas the DR contains mostly cortical bone. The mechanical properties in trabecular and cortical bone are different (17). Cortical bone can withstand greater stress but less strain before failure than trabecular bone; thus, cortical bone is stiffer. In vitro, fracture does not occur until the strain exceeds 75% in trabecular bone, whereas cortical bone fractures when the strain exceeds only 2%. Because the porous structure of trabecular bone, it has larger capacity for energy storage than cortical bone. Due to the differences in mechanical properties, external loads transmitted to different regions of radius may cause different amounts of bone tissue deformation; thus, resulting in different bone responses. Moreover, bone is a viscoelastic material; its biomechanical behavior varies with the loading rate of the external force. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 86 Further research is needed to determine weather or not the osteogenic responses to various force characteristics are specific to the composition of bone; i.e., cortical bone vs. trabecular bone. The damping condition may have played an important role in varying the loading characteristics of participants in this study. After dividing the data into the 2 damping groups, the correlations between reaction force characteristics and bone density changes in the non-damped group were greater than the correlations calculated using the ‘collapsed’ data. Furthermore, the relations between the reaction force characteristics and changes in radial BMD were also greater in the non- damped group than in the damped group. More specifically, in the non-damped group, changes in BMD were strongly correlated with peak loads and impact loads at the DR and TOTAL measurement sites. Additionally, changes in BMD were strongly correlated with impulses at TOTAL and loading rates at the DR sites. In contrast, only peak loads and UD BMD changes were modestly correlated in the damped group. These results may be due to the differences in loading history between the different damping conditions. Although there were no statistically significant differences in reaction force characteristics between damped and non-damped conditions because of the large variation among subjects, the peak loads and impact loads in the non-damped group were 1.4 times greater than those in the damped group. Moreover, the loading rates in the non-damped group were 2 times greater than those in the damped group. Since it has been suggested that osteogenic response increases with increasing magnitude of load and loading rates (80, 90, 112), the greater correlations between reaction forces and bone adaptations after 6 months existed in the non- damped group than the damped group can be expected. The significant correlations between changes in UD BMD and loading parameters including peak loads (r = .437), impact loads (r = .360), and impulses (r = .401), calculated using all subjects, were no longer significant when they were generated in the non-damped groups. However, the correlations were actually higher (r = .448, .389, and .484 for peak loads, impact loads, and impulses; respectively) in the non-damped group R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 87 than the correlations for all subjects. This phenomenon is likely associated with decreased statistical power after splitting the samples. 7.8 CONCLUSIONS AND FURTHER DIRECTION This study is an important step in the development of a biomechanics-based research program in human subjects devoted to: 1) quantification of loading via exercise in the upper extremity, 2) investigation of the relation between bone adaptation and mechanical loading, and 3) assessment of the applicability o f the dynamic impact loading osteogenic exercise program. We demonstrated that DILE generated highly reproducible loading profiles and joint kinematics. In addition, the exercise was safe, easy to perform, time-efficient, and required no equipment, supervision or participant travel. The results of correlational analyses paralleled findings in the literature that peak loads and impact loads were statistically significant correlated with changes in radial bone mineral density at all measurement sites including distal third radius, ultradistal radius, and total distal 1/3 radius. The damping condition appeared to play an important role in varying the loading history, as evidenced by the differences in the relations between the reaction force characteristics and changes in radial BMD between the two different damping groups. The lack of statistically significant bone adaptations following 6 months of DILE intervention may be due to the relatively short duration of the intervention, the large variation in reaction forces engendered during the DILE among the subjects, equipment limitation of DEXA, or a combination of factors. DILE appears to be a valid and reliable upper extremity musculoskeletal loading model in humans; thus, using this model for further research is warranted. Further research directions should include: 1) investigating the longer-term effects o f DILE on bone mass on the forearm, 2) evaluating specific bone remodeling patterns at different regions using advanced image techniques, such as pQCT and MRI, 3) identifying effective mechanical-loading factors associated with DILE for osteogenesis as modified by bone-related factors such as age, gender, physical activity level, diet, and hormonal status, and 4) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 88 determining the effects of different loading regimens on loading characteristics associated with the DILE, such as various loading distances, loading conditions (type of padding), loading cycles, frequency, intensity, and duration. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 89 8. REFERENCES 1 Adami S; Gatti D; Braga V et al. Site-specific effects of strength training on bone structure and geometry of ultradistal radius in postmenopausal women [see comments]. Journal o f Bone & Mineral Research. 14:120-4, 1999. 2 Aloia JF; Cohn SH; Ostuni JA et al. Prevention of involutional bone loss by exercise. Annals o f Internal Medicine. 89:356-8, 1978. 3 Ayalon J; Simkin A; Leichter I et al. Dynamic bone loading exercises for postmenopausal women: effect on the density of the distal radius. Archives o f Physical Medicine & Rehabilitation. 68:280-3, 1987. 4 Baecke JA; Burema J; Frijters JE. A short questionnaire for the measurement of habitual physical activity in epidemiological studies. American Journal o f Clinical Nutrition. 36:936-42, 1982. 5 Barry HC; Eathorne SW. Exercise and aging. Sports Medicine. 78:357-75, 1994. 6 Bassey EJ; Ramsdale SJ. Increase in femoral bone density in young women following high-impact exercise. Osteoporosis International. 4:72-5, 1994. 7 Bassey EJ; Ramsdale SJ. Weight-bearing exercise and ground reaction forces: a 12-month randomized controlled trial of effects on bone mineral density in healthy postmenopausal women. Bone. 16:469-76, 1995. 8 Bassey EJ; Rothwell MC; Littlewood JJ et al. Pre- and postmenopausal women have different bone mineral density responses to the same high-impact exercise [see comments]. Journal o f Bone & Mineral Research. 13:1805-13, 1998. 9 Beaupre GS; Orr TE; Carter DR. An approach for time-dependent bone modeling and remodeling— theoretical development. Journal o f Orthopaedic Research. 8:651-61, 1990. 10 Bemben DA; Boileau RA; Bahr JM et al. Effects of oral contraceptives on hormonal and metabolic responses during exercise. Medicine & Science in Sports & Exercise. 24:434-41, 1992. 11 Bertram JE; Biewener AA. Bone curvature: sacrificing strength for load predictability? Journal o f Theoretical Biology. 131:75-92, 1988. 12 Bertram JE; Swartz SM. The 'law of bone transformation': a case of crying Wolff? Biological Reviews o f the Cambridge Philosophical Society. 66:245-73, 1991. 13 Beverly MC; Rider TA; Evans MJ et al. Local bone mineral response to brief exercise that stresses the skeleton [see comments], Bmj. 299:233-5, 1989. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 90 14 Bevier WC; Wiswell RA; Pyka G et al. Relationship of body composition, muscle strength, and aerobic capacity to bone mineral density in older men and women. Journal o f B om & Mineral Research. 4:421-32, 1989. 15 Black DM; Cummings SR; Karpf DB et al. Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Fracture Intervention Trial Research Group [see comments]. Lancet. 348:1535-41, 1996. 16 Bross R; Casaburi R; Storer TW et al. Androgen effects on body composition and muscle function: implications for the use of androgens as anabolic agents in sarcopenic states. Baillieres Clinical Endocrinology & Metabolism. 12:365-78, 1998. 17 Carter DR; Hayes WC. Bone compressive strength: The influence of density and strain rate. Science. 194:1174, 1976. 18 Cauley JA; Seeley DG; Ensrud K et al. Estrogen replacement therapy and fractures in older women. Study of Osteoporotic Fractures Research Group. Annals o f Internal Medicine. 122:9-16, 1995. 19 Cavanaugh DJ; Cann CE. Brisk walking does not stop bone loss in postmenopausal women. Bone. 9:201-4, 1988. 20 Chow JW; Jagger CJ; Chambers TJ. Characterization of osteogenic response to mechanical stimulation in cancellous bone of rat caudal vertebrae. American Journal o f Physiology. 265:E340-7, 1993. 21 Cooper C; Campion G; Melton LJd. Hip fractures in the elderly: a world-wide projection. Osteoporosis International. 2:285-9, 1992. 22 Cummings SR; Kelsey JL; Nevitt MC et al. Epidemiology of osteoporosis and osteoporotic fractures. Epidemiologic Reviews. 7:178-208, 1985. 23 Dalsky GP; Stocke KS; Ehsani AA et al. Weight-bearing exercise training and lumbar bone mineral content in postmenopausal women. Annals o f Internal Medicine. 108:824-8, 1988. 24 Dartsch PC; Betz E. Response of cultured endothelial cells to mechanical stimulation. Basic Research in Cardiology. 84:268-81, 1989. 25 Davidson RM; Tatakis DW; Auerbach AL. Multiple forms of mechanosensitive ion channels in osteoblast-like cells. Pflugers Archiv - European Journal o f Physiology. 416:646-51, 1990. 26 Dibbets JMH: One century of W olffs law. (Center for human growth and development, University of Michigan Press, Ann Arbor 1992). 27 Dishman RK: Exercise adherence: Its impact on public health (Human Kinetics Books, Illinois, Champaign 1988). 28 Dishman RK; Sallis JF; Orenstein DR. The determinants of physical activity and exercise. Public Health Reports. 100:158-71, 1985. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 91 29 Dook JE; Janies C; Henderson NIC et al. Exercise and bone mineral density in mature female athletes. Medicine & Science in Sports & Exercise. 29:291-6, 1997. 30 Faulkner KG; Gluer C-C; Majumdar S et al. Noninvasive measurements of bone mass, structure, and strength: current methods and experimental techniques. AJR. 157:1229-37, 1991. 31 Fitzsimmons RJ; Strong DD; Mohan S et al. Low-amplitude, low-frequency electric field- stimulated bone cell proliferation may in part be mediated by increased IGF-II release. Journal o f Cellular Physiology. 150:84-9, 1992. 32 Flanagan S; Salem GJ; Wang M-Y et al. Effect of two different squatting techniques on lower extremity kinetics in older adults. Medicine and Science in Sport and Exercise. 34, 2002 . 33 Frost HM. A determinant o f bone architecture. The minimum effective strain. Clinical Orthopaedics & RelatedResearch:2&6-92, 1983. 34 Frost HM. Bone "mass" and the "mechanostat": a proposal. Anatomical Record. 219:1-9, 1987. 35 Frost HM. The mechanostat: a proposed pathogenic mechanism of osteoporoses and the bone mass effects of mechanical and nonmechanical agents. Bone & Mineral. 2:73-85, 1987. 36 Frost HM. Vital biomechanics: proposed general concepts for skeletal adaptations to mechanical usage. Calcified Tissue International. 42:145-56, 1988. 37 Fryburg DA; Barrett EJ. Growth hormone acutely stimulates skeletal muscle but not whole-body protein synthesis in humans. Metabolism: Clinical & Experimental. 42:1223-7, 1993. 38 Genant H; Faulkner K; Gluer C-C. Measurement of bone mineral density: current status. The American Journal o f Medicine. 91:49S-53 S, 1991. 39 Gotshalk LA; Loebel CC; Nindl BC et al. Hormonal responses of multiset versus single-set heavy-resistance exercise protocols. Canadian Journal o f Applied Physiology. 22:244-55, 1997. 40 Greendale GA; Hirsch SH; Hahn TJ. The effect of a weighted vest on perceived health status and bone density in older persons. Quality o f Life Research. 2:141-52, 1993. 41 Greendale GA; Salem GJ; Young JT et al. A randomized trial of weighted vest use in ambulatory older adults: strength, performance, and quality of life outcomes. Journal o f the American Geriatrics Society. 48:305-11, 2000. 42 Grove KA; Londeree BR. Bone density in postmenopausal women: high impact vs low impact exercise. Medicine & Science in Sports & Exercise. 24:1190-4, 1992. 43 Hamilton A; Balnave R; Adams R. Grip strength testing reliability. Journal o f Hand Therapy. 7:163-70, 1994. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 92 44 Hartard M; Bottermann P; Bartenstein P et al. Effects on bone mineral density o f low- dosed oral contraceptives compared to and combined with physical activity. Contraception. 55:87-90, 1997. 45 Harter LV; Hruska KA; Duncan RL. Human osteoblast-like cells respond to mechanical strain with increased bone matrix protein production independent of hormonal regulation. Endocrinology. 136:528-35, 1995. 46 Heinonen A; Kannus P; Sievanen H et al. Randomised controlled trial of effect of high- impact exercise on selected risk factors for osteoporotic fractures [see comments]. Lancet. 348:1343-7, 1996. 47 Heinonen A; Sievanen H; Kannus P et al. Effects of unilateral strength training and detraining on bone mineral mass and estimated mechanical characteristics of the upper limb bones in young women. Journal o f Bone & Mineral Research. 11:490-501, 1996. 48 Ingber DE. Cellular tensegrity: defining new rules of biological design that govern the cytoskeleton. Journal o f Cell Science. 104:613-27, 1993. 49 Jones DB; Bingmann D. How do osteoblasts repond to mechanical stimulation? Cells Materials. 1:329-340, 1991. 50 Kanis JA; Pitt FA. Epidemiology of osteoporosis. Bone. 13 :S7-15, 1992. 51 Kassem M; Blum W; Ristelli J et al. Growth hormone stimulates proliferation and differentiation of normal human osteoblast-like cells in vitro. Calcified Tissue International. 52:222-6, 1993. 52 Kohrt WM; Ehsani AA; Birge SJ, Jr. Effects of exercise involving predominantly either joint-reaction or ground-reaction forces on bone mineral density in older women. Journal o f Bone & Mineral Research. 12:1253-61, 1997. 53 Kraemer RR; Heleniak RJ; Tryniecki JL et al. Follicular and luteal phase hormonal responses to low-volume resistive exercise. Medicine & Science in Sports & Exercise. 27:809-17, 1995. 54 Kraemer WJ; Fleck SJ; Dziados JE et al. Changes in hormonal concentrations after different heavy-resistance exercise protocols in women. Journal o f Applied Physiology. 75:594-604, 1993. 55 Kraemer WJ; Staron RS; Hagerman FC et al. The effects of short-term resistance training on endocrine function in men and women. European Journal o f Applied Physiology & Occupational Physiology. 78:69-76, 1998. 56 Kriska AM; Bennett PH. An epidemiological perspective o f the relationship between physical activity and NIDDM: from activity assessment to intervention. Diabetes- Metabolism Reviews. 8:355-72, 1992. 57 Kuchan MJ; Jo H; Frangos JA. Role of G proteins in shear stress-mediated nitric oxide production by endothelial cells. American Journal o f Physiology. 267:C753-8, 1994. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 93 58 Kudlacek S; Pietschmann F; Bernecker P et al. The impact o f a senior dancing program on spinal and peripheral bone mass. American Journal o f Physical Medicine & Rehabilitation, 76:477-81, 1997. 59 Lanyon LE; Rubin CT. Static vs dynamic loads as an influence on bone remodelling. Journal o f Biomechanics. 17:897-905, 1984. 60 Lean JM; Jagger CJ; Chambers TJ et al. Increased insulin-like growth factor I mRNA expression in rat osteocytes in response to mechanical stimulation. American Journal o f Physiology. 268:E318-27, 1995. 61 LeRoith D; Werner H; Burguera B et al. The insulin-like growth factor family of peptides, binding proteins and receptors: their potential role in tissue regeneration. Advances in Experimental Medicine & Biology. 321:21-8; discussion 29-30, 1992. 62 Lips P. Epidemiology and predictors of fractures associated with osteoporosis. American Journal o f Medicine. 103:3 S-8S; discussion 8S-11S, 1997. 63 Mathiowetz V; Weber K; Volland G et al. Reliability and validity of grip and pinch strength evaluations. Journal o f Hand Surgery - American Volume. 9:222-6, 1984. 64 Mazess RB. On aging bone loss. Clinical Orthopaedics & Related Researck.239-52, 1982. 65 Mundy GR. Peptides and growth regulatory factors in bone. Rheumatic Diseases Clinics o f North America. 20:577-88, 1994. 66 Murray DW; Rushton N. The effect of strain on bone cell prostaglandin E2 release: a new experimental method. Calcified Tissue International. 47:35-9, 1990. 67 Nelson ME; Fiatarone MA; Morganti CM et al. Effects of high-intensity strength training on multiple risk factors for osteoporotic fractures. A randomized controlled trial. Jama. 272:1909-14, 1994. 68 Nelson ME; Fisher EC; Dilmanian FA et al. A 1-y walking program and increased dietary calcium in postmenopausal women: effects on bone. American Journal o f Clinical Nutrition. 53:1304-1 1, 1991. 69 O'Connor JA; Lanyon LE; MacFie H. The influence of strain rate on adaptive bone remodelling. Journal o f Biomechanics. 15:767-81, 1982. 70 Olesen SP; Clapham DE; Davies PF. Haemodynamic shear stress activates a K+ current in vascular endothelial cells. Nature. 331:168-70, 1988. 71 Ott SM. Methods of determining bone mass. Journal o f Bone & Mineral Research. 6:S71- 6; discussion S83-4, 1991. 72 Otter M; Shoenung J; Williams WS. Evidence for different sources of stress-generated potentials in wet and dry bone. Journal o f Orthopaedic Research. 3:321-4, 1985. 73 Otter MW; Palmieri VR; Wu DD et al. A comparative analysis of streaming potentials in vivo and in vitro. Journal o f Orthopaedic Research. 10:710-9, 1992. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 94 74 Parifitt AM: The physiologic and clinical significance of bone histomorphometric data. (CRC Press, Boca Raton, FL 1983). 75 Pavalko FM; Otey CA; Simon KO et al. Alpha-actinin: a direct link between actin and integrins. Biochemical Society Transactions. 19:1065-9, 1991. 76 Phillips S; Fox N; Jacobs J et al. The direct medical costs of osteoporosis for American women aged 45 and older, 1986. Bone. 9:271-9, 1988. 77 Picard D; Ste-Marie LG; Coutu D et al. Premenopausal bone mineral content relates to height, weight and calcium intake during early adulthood. Bone & Mineral. 4:299-309, 1988. 78 Pocock N; Eisman J; Gwinn T et al. Muscle strength, physical fitness, and weight but not age predict femoral neck bone mass. Journal o f Bone & Mineral Research. 4:441-8, 1989. 79 Prince RL; Smith M; Dick 1M et al. Prevention of postmenopausal osteoporosis. A comparative study of exercise, calcium supplementation, and hormone-replacement therapy. New England Journal o f Medicine. 325:1189-95, 1991. 80 Qin Y-X; Rubin CT; McLeod KJ. Nonlinnear dependence of loading intensity and cycle number in the maintenance of bone mass and morphology. Journal o f Orthopaedic Research. 16:482-9, 1998. 81 Raisz LG; Wiita B; Artis A et al. Comparison of the effects of estrogen alone and estrogen plus androgen on biochemical markers of bone formation and resorption in postmenopausal women. Journal o f Clinical Endocrinology & Metabolism. 81:37-43, 1996. 82 Recker RR: Embryology, anatomy, and microstructure of bone. (Raven Press, New York, NY 1992). 83 Recker RR; Kimmel DB; Parfitt AM et al. Static and tetracycline-based bone histomorphometric data from 34 normal postmenopausal females. Journal o f BOne Mineral Reserve. 3:133-44, 1988. 84 Rockwell JC; Sorensen AM; Baker S et al. Weight training decreases vertebral bone density in premenopausal women: a prospective study. Journal o f Clinical Endocrinology & Metabolism. 71:988-93, 1990. 85 Rubin C; Gross T; Qin YX et al. Differentiation of the bone-tissue remodeling response to axial and torsional loading in the turkey ulna. Journal o f Bone & Joint Surgery - American Volume. 78:1523-33, 1996. 86 R ubin C T; B ain SD; M cL eod K J. S uppression o f the osteogenic response in the aging skeleton. Calcified Tissue International. 50:306-13, 1992. 87 Rubin CT; Gross TS; McLeod KJ et al. Morphologic stages in lamellar bone formation stimulated by a potent mechanical stimulus. Journal o f Bone & Mineral Research. 10:488- 95, 1995. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 95 88 Rubin CT; Lanyon LE. Limb mechanics as a function of speed and gait: a study of functional strains in the radius and tibia o f horse and dog. Journal o f Experimental Biology. 101:187-211, 1982. 89 Rubin CT; Lanyon LE. Regulation of bone formation by applied dynamic loads. Journal o f Bone & Joint Surgery - American Volume. 66:397-402, 1984. 90 Rubin CT; Lanyon LE. Regulation of bone mass by mechanical strain magnitude. Calcified Tissue International. 37:411-7, 1985. 91 Sachs F. Mechanical transduction in biological systems. Critical Reviews in Biomedical Engineering. 16:141-69, 1988. 92 Salem GJ; Wang M-Y; Azen S et al. Lower-extremity kinetic response to activity program dosing in older adults. Journal o f Applied Biomechanics. 17:103-12, 2001. 93 Salem GJ; Young J; Gregor G et al. Knee extensor moments and knee-extensor strength in older persons wearing a weighted vest. Medicine and Science in Sports and Exercise. 28:S110, 1996. 94 Salzstein RA; Pollack SR. Electromechanical potentials in cortical bone— II. Experimental analysis. Journal o f Biomechanics. 20:271-80, 1987. 95 Salzstein RA; Pollack SR; Mak AF et al. Electromechanical potentials in cortical bone— I. A continuum approach. Journal o f Biomechanics. 20:261-70, 1987. 96 Sandy JR; Farndale RW. Second messengers: regulators of mechanically-induced tissue remodelling. European Journal o f Orthodontics. 13:271-8, 1991. 97 Schwartz MA; Ingber DE. Integrating with integrins. Molecular Biology o f the Cell. 5:389- 93, 1994. 98 Scott GC; Korostoff E. Oscillatory and step response electromechanical phenomena in human and bovine bone. Journal o f Biomechanics. 23:127-43, 1990. 99 Shaw JM; Snow CM. Weighted vest exercise improves indices of fall risk in older women. Journals o f Gerontology. Series A, Biological Sciences & Medical Sciences. 53:M53-8, 1998. 100 Simkin A; Ayalon J; Leichter I. Increased trabecular bone density due to bone-loading exercises in postmenopausal osteoporotic women. Calcified Tissue International. 40:59-63, 1987. 101 Sims JR; Karp S; Ingber DE. Altering the cellular mechanical force balance results in integrated changes in cell, cytoskeletal and nuclear shape. Journal o f Cell Science. 103:1215-22, 1992. 102 Sinaki M; Fitzpatrick LA; Ritchie CK et al. Site-specificity of bone mineral density and muscle strength in women: job-related physical activity. American Journal o f Physical Medicine & Rehabilitation. 77:470-6, 1998. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 96 103 Sinaki M; Wahner HW; Offord KP. Relationship between grip strength and related regional bone mineral content. Archives o f Physical Medicine & Rehabilitation. 70:823-6, 1989. 104 Sinaki M; Wahner HW; Offord KP et al. Efficacy of nonloading exercises in prevention of vertebral bone loss in postmenopausal women: a controlled trial. Mayo Clinic Proceedings. 64:762-9, 1989. 105 Skerry TM; Bitensky L; Chayen J et al. Early strain-related changes in enzyme activity in osteocytes following bone loading in vivo. Journal o f Bone & Mineral Research. 4:783-8, 1989. 106 Smith EL, Jr.; Reddan W; Smith PE. Physical activity and calcium modalities for bone mineral increase in aged women. Medicine & Science in Sports & Exercise. 13:60-4, 1981. 107 Somjen D; Binderman 1 ; Berger E et al. Bone remodelling induced by physical stress is prostaglandin E2 mediated. Biochimica et Biophysica Acta. 627:91-100, 1980. 108 Turner CH; Akhter MP; Raab DM et al. A noninvasive, in vivo model for studying strain adaptive bone modeling. Bone. 12:73-9, 1991. 109 Turner CH; Forwood MR. What role does the osteocyte network play in bone adaptation? [editorial]. Bone. 16:283-5, 1995. 110 Turner CH; Forwood MR; Otter MW. Mechanotransduction in bone: do bone cells act as sensors of fluid flow? FASEB Journal. 8:875-8, 1994. 111 Turner CH; Forwood MR; Rho JY et al. Mechanical loading thresholds for lamellar and woven bone formation. Journal o f Bone & Mineral Research. 9:87-97, 1994. 112 Turner CH; Owan I; Takano Y. Mechanotransduction in bone: role of strain rate. American Journal o f Physiology. 269:E438-42, 1995. 113 Turner CH; Pavalko FM. Mechanotransduction and functional response of the skeleton to physical stress: the mechanisms and mechanics of bone adaptation. Journal o f Orthopaedic Science. 3:346-55, 1998. 114 Turner CH; Takano Y; Owan I. Aging changes mechanical loading thresholds for bone formation in rats. Journal o f Bone & Mineral Research. 10:1544-9, 1995. 115 Vandenburgh HH. Mechanical forces and their second messengers in stimulating cell growth in vitro. American Journal o f Physiology. 262:R350-5, 1992. 116 Wang M-Y; Salem GJ; Flanagan S et al. Lower-extreinity joint kinetics during forward stepping and lateral stepping in older adults. Medicine and Science in Sports and Exercise. 34, 2002. 117 Wang M-Y; Shuler CJ; Salem GJ et al. The effects of weighted and unweighted bench step aerobics on ground reaction forces. Proceedings o f the 42nd Annual Conference on Physical Therapy in Taiwan:8, 2000. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 97 118 Wapniarz M; Lehmann R; Reincke M et al. Determinants of radial bone density as measured by PQCT in pre- and postmenopausal women: the role of bone size. Journal o f Bone & Mineral Research. 12:248-54, 1997. 119 Watrous DA; Andrews BS. The metabolism and immunology of bone. Seminars in Arthritis & Rheumatism. 19:45-65, 1989. 120 Weaver CM; Teegarden D; Lyle RM et al. Impact of exercise on bone health and contraindication of oral contraceptive use in young women. Medicine & Science in Sports & Exercise. 33:873-80, 2001. 121 Weinbaum S; Cowin SC; Zeng Y. A model for the excitation of osteocytes by mechanical loading-induced bone fluid shear stresses. Journal o f Biomechanics. 27:339-60, 1994. 122 Weiss LW; Cureton KJ; Thompson FN. Comparison of serum testosterone and androstenedione responses to weight lifting in men and women. European Journal o f Applied Physiology & Occupational Physiology. 50:413-9, 1983. 123 Westerlind KC; Fluckey JD; Gordon SE et al. Effect of resistance exercise training on cortical and cancellous bone in mature male rats. Journal o f Applied Physiology. 84:459- 64, 1998. 124 W olff J; Maquet P; Furlong R et al.: The law of bone remodelling. (Springer-Verlag, Berlin, Germany 1986). 125 Wuster C. Growth hormone and bone metabolism. Acta Endocrinologica. 128:14-8, 1993. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 98 APPENDIX A Medical History Questionnaire R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 99 THE EFFECTS OF DYNAMIC IMPACT LOADING EXERCISE ON FOREARM BONE MASS IN HEALTHY ADULTS General Health and Medical History Questionnaire All information in this survey will remain strictly confidential. I agree to complete the following questionnaire with the understanding that all information is voluntary and completely confidential. I understand that my signature here does not mean that I have agreed to participate in the actual study. SIGNATURE_____________________________________________ N ame____________________ (last)_____________________(first)________(middle) Age_______ Height_________ Weight______ Gender_______ M / F Ethnicity______________________________ Email__________________________ Telephone Number (daytime)_____________________ (evening)________________ I. GENERAL HEALTH QUESTIONS The following questions are about activities you might do during a typical day. Does your health now limit you in these activities, if so, how much? 1. Does your health now limit you in sport activities, such as playing basketball, baseball, bowling, or golf? (1 )___________ Yes, limited a lot (2 )___________ Yes, limited a little (3 )___________No, not limited at all 2. Does your health now limit you in lifting or carrying groceries? (1 ) Yes, limited a lot (2 )___________ Yes, limited a little (3 )___________No, not limited at all 3. Does your health now limit you in house cleaning, such as mopping, sweeping, scrubbing or scraping? (1 ) Yes, limited a lot (2 )___________ Yes, limited a little (3 )___________No, not limited at all 4. Does your health now limit you in typing on the computer or typewriter for more than 1 hour? R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 100 (1 )___________ Yes, limited a lot (2 )___________ Yes, limited a little (3) _ _ N o , not limited at all 5. Do you have any current muscle soreness or injury? (1 )________ Yes (2 )___ No If yes, please state body region of soreness or injury. ______________________________ 6. Do you have any current joint soreness or injury? (1 ) Yes (2 )___ No If yes, please state body region of soreness or injury. 7. Do you have any current skeletal injury? (1 )_________ Yes (2 )___________No If yes, please state body region of injury. 8. During the past 6 weeks, have you had any muscle soreness or injury? (1 ) Yes (2) No If yes, please state body region of soreness or injury.______________ 9. During the past 6 weeks, have you had any joint soreness or injury? (1 ) Yes (2 )___ No If yes, please state body region of soreness or injury.___________ 10. During the past 6 weeks, have you had any skeletal injury? (1 )___________ Yes (2 )_________ No If yes, please state body region of injury.______ _______ 11. Have you engaged in upper-extremity training program during the previous 2 years, such as weight lifting? (1 ) Yes (2 )________ No If yes, please state the training p ro g ra m .____________________________________________ 12. Are you an athlete/sports player? (1 ) _Yes (2 )_________ No If yes, please state the sport you are playing. ______________________________________ R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 101 How often do you play the sport?_____________________ times/week 13. Have you taken more than 500mg of calcium tablet supplementation per day for more than 1 month in the previous year? (1 )_______ Yes (2 )_________ No 14. Do you have regular menstrual cycles within the last 3 years? (1 ) Yes (2 )_________ No How often are your menstrual periods?_______________________________________________ How many days do they last?________________________________________________________ Are you or have you been amenorrheic for reason other than pregnancy?__________________ 15. Are you pregnant now or do you intend to be pregnant in the following 1 year? (1 )_______ Yes (2 )_________ No 16. Do you use oral contraceptives currently or within the last 3 months? (1 ) Yes (2 )________ No R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 102 II. MEDICAL HISTORY Which medical conditions do you have or have you had in the past? A. EYE & EAR PROBLEMS (1 )___Cataracts (2 )___Glaucoma (3 )___Macular degeneration of the eye (4 )___Hearing loss/hearing aid B. HEART/CARDIOVASCULAR PROBLEMS (1 )___Angina (2 )___Heart attack:______ year (3 )___Heart failure (4 )___Coronary Artery Disease (5 )___High Blood Pressure C. LUNG PROBLEM S (1 ) _ _ Asthma (2 ) _ _ Bronchitis ("COPD") (3 ) _ _ Emphysema D. GASTROINTESTINAL PROBLEM (1 )__ Ulcers (2 )___Heartburn/Hiatal hernia (3 )___Diverticulosis (4 )___Liver Disease/Cirrhosis (5 )___Hepatitis (6 )_ _ Polyps (7 )_ _ Gallbladder disease E. NERVOUS SYSTEM PROBLEMS (1 )_ _ Stroke (2 )_ _ Dementia/Alzheimer's (3 )_ _ Parkinson's Disease (4 )_ _ Epilepsy/Seizures O ther problem s not listed above, please specify. F. BONE & JOINT PROBLEMS (1 )_ _ Osteo/arthritis, joint(s)_____________________________ (2 )_ _ Osteo/rheumatoid/arthritis, joint(s)_____________________________ (3 )_ _ Arthritis, type unsure, j oint(s)_____________________________ (4 )_ _ Osteoporosis (5 )_ _ Fracture, joint/bone(s)________________________ (6 )_ _ Low back pain, Specify (7 )_ _ Carpal tunnel syndrom (8 ) Bone marrow disorders G. GLAND PROBLEMS (1 )_ _ Ulcers (2 )_ _ Thyroid overactive (high) (3 )_ _ Thyroid underactive (low) (4 )_ _ Parathyroid disease (5 )_ _ Diabetes H. KIDNEY & URINARY TRACT (1 )_ _ Kidney disease (2 )_ _ Prostate disease (3 )_ _ Involuntary loss of urine I. OTHER HEALTH PROBLEMS (1 )_ _ Anemia (2 )_ _ Hernia (3 )_ _ Thrombosis (4 )__ Cancer, of what______________________________ (5 )__ Depression (6 )__ Long term immobilization, due to ________________________________ R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 103 Are you aware of any conditions that might limit your participation in physical activity? (1 ) Yes (2 )________ No If yes, what are the conditions?________________________________________________________ List all medicines that you use on a regular schedule at this time. Current medications used regularly What strength? How do you use it? (How many? How many times a day?) Example: Tylenol 500 mg. I pill 3 times a day List medicines that you used "as needed" at least twice in the last year. (Any medicines used daily or even weekly should be listed above.) Medications used "as needed" at least twice in the last year. How often? (weekly? monthly?) What strength? How do you use it? (How many? How many times a day?) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 104 APPENDIX B Physical Activity Questionnaire R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 105 THE EFFECTS OF DYNAMIC IMPACT LOADING EXERCISE ON FOREARM BONE MASS IN HEALTHY ADULTS Physical activity questionnaire All information in this survey will remain strictly confidential. I agree to complete the following questionnaire with the understanding that all information is voluntary and completely confidential. Name____________________ (last)_____________________(first)________(middle) ID #_________ Age________ Gender____________ Ethnicity________________________ Email__________________________ We would like to know about your physical activity during the past 6 months. Please read each question carefully and answer each as accurately and honestly as you can. Indicate only one answer for each question by circle the appropriate corresponding number (1-5). I. OCCUPATIONAL ACTIVITIES The following questions are about activities you did during your work in the past 6 months. 1. What is your main occupation?_______________________________________________________ 2. In comparison with others of your own age, you think your work is physically 1 2 3 4 5 much lighter lighter as heavy heavier much heavier 3. At work, you lift/carry heavy loads 1 2 3 4 5 never seldom sometimes often very often 4. At work, you do heavy work with your upper extremities such as construction or hard physical labor which requires digging or chopping with heavy tools 1 2 3 4 5 never seldom sometimes often very often 5. At work, the physical demands o f your upper extremities are 1 2 3 4 5 very light light moderate heavy very heavy R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 106 6. After working, your upper extremities are tired 1 2 3 4 5 never seldom sometimes often very often II. LEISURE ACTIVITIES The following questions are about activities you did during your leisure time in the past 6 months. 7. In comparison with others o f your own age, you think your physical activity during leisure time is 1 2 3 4 5 much less less the same more much more 8. Do you play sport or do exercise during your leisure time? 1 2 yes no If yes: Please list the sport or exercise starting with the one you performed most frequently in the past 6 months. Sport/Exercise You play this sport or do this exercise never seldom sometimes often very often 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 III. HOME ACTIVITIES The following questions are about activities you did at home in the past 6 months. 9. In comparison with others of your own age, you think your activities at home is physically 1 2 3 4 5 much lighter lighter as heavy heavier much heavier R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 107 10. At home, you do heavy cleaning such as scrubbing or scraping 1 2 3 4 never seldom sometimes often 11. At home, you do heavy gardening such as hoeing, digging, or shoveling 1 2 3 4 never seldom sometimes often 12. At home, you lift/carry heavy loads 1 2 3 4 never seldom sometimes often 13. At home, the physical demands of your upper extremities are 1 2 3 4 very light light moderate heavy 5 very often 5 very often 5 very often 5 very heavy R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. APPENDIX C Exercise Diary R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 109 2001 Dynamic Impact Loading Exercise Diary January_____ S M T W T F S 1 2 3 4 S 6 7 0 6 10 11 12 13 14 15 15 17 18 10 20 21 22 23 24 25 26 27 28 28 30 31 April July October February S M T W T F S 12 3 4 5 6 7 6 9 10 11 12 13 14 15 16 17 18 10 20 21 22 23 24 25 28 27 28 May August November S M T W T F 1 2 S 3 4 5 6 7 0 0 10 11 12 13 14 15 16 17 18 10 20 21 22 23 24 25 26 27 26 20 30 March S M T W T F S 1 2 3 4 5 6 7 0 0 10 11 12 13 14 15 16 17 10 19 20 21 22 23 24 25 26 27 28 20 30 31 June S M T W T F S 1 2 3 4 5 6 7 0 0 10 11 12 13 14 15 16 17 10 10 2 0 21 22 23 24 25 26 27 28 20 30 September S M T W T F 6 2 3 4 5 6 7 0 0 10 11 12 13 14 15 16 17 18 10 20 21 22 23 24 25 26 27 20 29 30 December * Please circle the day on which you perform the impact loading exercise. * Please put an "x" on the day that you should perfrome the exercise but you do not. * Please put an "v" on the day that you do perform the exercise, but is not completed. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 110 THE EFFECTS OF DYNAMIC IMPACT LOADING EXERCISE ON FOREARM BONE MASS IN HEALTHY ADULTS Exercise Diary Please record your training activities. Instructions: Date: Please write down the day you are scheduled to perform the impact loading exercise. C/I/N: C indicates the impact loading exercise is performed completely. I indicates the impact loading exercise is performed incompletely. N indicates the impact loading exercise is not performed. Please briefly describe the reason if I or N is indicated. Wall: Please indicate the location of the wall you perform the impact loading exercise. Side effects: Please simply describe any side effect associated with the impact loading exercise. Date C / I / N Wall Side effects R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. I l l Date C / I / N Wall Side effects R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 112 Date C / I / N Wall Side effects R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 113 APPENDIX D Secondary Statistical Analysis A. BMC and area changes after 6 months of DILE intervention B. Regression analysis: changes in BMD regressed on selected force characteristics, changes in body weight, and changes in hand grip strength R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 114 A. BMC and Area Changes After 6 Months of DILE Intervention Paired t-tests were conducted to detect any significant changes in BMC and area associated with the DILE intervention. There were no statistically significant differences in mean radial BMC of the exercised arm after 6 months of exercise intervention except for a marginally significant 0.6% decrease in DR BMC (p = .04) (Table 30). Moreover, there were no statistically significant differences in mean radial BMC of the non-exercised arm after 6 months. Further, no statistically significant changes in bone area after 6 months were observed in both exercised and non-exercised arm (Table 31). The differences in damping effects on radial bone adaptations after 6 months were further tested by independent t-tests in the exercised arm. The results showed no statistically significant differences in both BMC and area changes after 6 months between the damped and non-damped groups at all regions (Table 31 and 32). Table 30: Radial BMC of exercised and non-exercised arms at the baseline and final Exercised Non-exercised Measurements1 Baseline Final P2 Baseline Final PJ A ll 01=24) DR BMC 1.774 (.233) 1.763 (.231) .04 1.766 (.218) 1.758 (.223) .07 UD BMC 1.585 (.222) 1.595 (.216) .13 1.608 (.2 0 1 ) 1.590 (.228) .34 TOTAL BMC 6.986 (.997) 6.930(1.028) .15 7.017 (.939) 6.956 (.966) .36 Mean (SD) 1 BMC = bone mineral content (grams), DR = distal third radius, UD = ultra-distal radius, and TOTAL = total distal third of radius 2 Paired t-tests compared BMD at exercised arm between baseline and final 3 Paired t-tests compared BMD at non-exercised arm between baseline and final Table 31: Radial bone area of exercised and non-exercised arms at the baseline and final Exercised Non-exercised Measurements1 Baseline Final P2 Baseline Final PJ A ll (n=24) DR Area 2.604 (.259) 2.603 (.255) .93 2.600 (.254) 2.596 (.238) .69 UD Area 3.512 (.338) 3.537 (.296) . 2 2 3.549 (.316) 3.560 (.288) .60 TOTAL Area 12.266 12.213 .45 12.376 12.313 .52 (1.450) (1.401) (1.450) (1.309) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 115 Mean (SD) 1 Area = projected area (cm2 ), DR = distal third radius, UD = ultra-distal radius, and TOTAL = total distal third of radius 2 Paired t-tests compared BMD at exercised arm between baseline and final 3 Paired t-tests compared BMD at non-exercised arm between baseline and final Table 32: Radial BMC of exercised arm at damped and non-damped groups Measurements1 Damped (n =14) Non-damped (n = 10) P3 DR BMC Baseline Final Difference2 1.726 (.238) 1.719 (.232) -.006 (.021) 1.842 (.220) 1.825 (.228) -.017 (.027) .29 UD BMC Baseline Final D ifference 1.540 (.178) 1.561 (.185) .021 (.028) 1.648 (.270) 1.644 (.256) -.004 (.034) .06 Total BMC Baseline Final Difference 6.799 (.786) 6.778 (.857) -.021 (.212) 7.248 (1.233) 7.142(1.246) -.106 (.136) .28 1 BMC = bone mineral content (grams), DR = distal third radius, UD = ultra-distal radius, and TOTAL = total distal third o f radius 2 Difference = Final - Baseline 3 Independent t-tests compared BMC changes after 6 months between the damped and non-damped groups Table 33: Radial bone area of exercised arm at damped and non-damped groups Measurements1 Damped (n =14) Non-damped (n = 10) P3 DR Area Baseline Final Difference2 2.547 (.265) 2.549 (.268) .001 (.068) 2.684 (.239) 2.679 (.227) -.005 (.062) .82 UD Area Baseline Final D ifference 3.449 (.287) 3.497 (.288) .049 (.066) 3.601 (.398) 3.593 (.315) -.008 (.127) .17 Total Area Baseline Final Difference TT-" 1 ". • , , 12.041 (1.131) 12.076 (1.272) .035 (.362) i- , i • 12.581 (1.826) 12.405 (1.616) -.176 (.265) .13 Area = projected area (cm ), DR = distal third radius, UD = ultra-distal radius, and TOTAL = total distal third of radius 2 Difference = Final - Baseline 3 Independent t-tests compared area changes after 6 months between the damped and non-damped groups R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 116 B. Regression Analysis: Changes in BMD Regressed on Selected Force Characteristics, Changes in Body Weight, and Changes in Hand Grip Strength The relative contribution of reaction force characteristics, changes in body weight (A B W ), and changes in hand grips strength (AG rip) to the prediction of changes in BMD (A BM D ) were examined by multiple linear regression analysis. The results indicated that neither A G rip nor A B W significantly predicted ABM D at all regions. Peak load was an independent predictor of ABM D at all 3 regions (Table 34). With each 1 standard deviation (SD) increase of peak load, BMD increased by .43, .44, and .55 SD at DR, UD, and TOTAL; respectively. Moreover, each 100 N increase in peak load resulted in increases in BMD by .005, .004, and .003 g/cm2 at DR, UD, and TOTAL; respectively. The linear regression on peak load accounted for 18%, 19%, and 30% of variability in ABM D at DR, UD, and TOTAL. Impact load was also an independent predictor of ABM D at DR and TOTAL. Each 1 SD increase in impact load resulted in .42 and .50 SD increases in BMD at DR and TOTAL. Also, BMD at DR and TOTAL increased by .003 g/cm2 with every 100 N increase in impact load. The linear regression on impact load accounted for 18% and 25% of variability in ABM D at DR and TOTAL. Both impulse and loading rate did not significantly predict ABM D in any measurement region. The Influence of Damping Condition on the Predictability of Reaction Force Characteristics to Changes in Bone Mineral Density The roles of peak load and impact load as significant predictors of changes in BMD changed with the damping condition. None of the reaction force characteristics significantly predict changes in BMD in the damped group. Conversely, in the non-damped group, peak load and impact load were independent predictors o f ABM D at DR and TOTAL (Table 34). Each 1 SD increase in peak load R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 117 and impact load resulted in .73 and .76 SD increases in BMD at DR. Similarly, each 1 SD increase in peak load and impact load resulted in .82 and .81 SD increases in BMD at TOTAL. Moreover, BMD at DR and TOTAL increased by .006 and .004; respectively, with every 100 N increase in peak load and impact load. Impulse was also an independent predictor o f TOTAL ABM D. A .02 g/cm2 increase in TOTAL BMD was resulted from every 100 N*sec increase in impulse. The variability of ABM D explained by the linear regression on peak load and impact load was 54% and 58% at DR. The variability of ABM D explained by the linear regression on peak load, impulse, and impact load was 67%, 56%, and 65% at TOTAL; respectively. Table 34: Linear relations between selected force characteristics and changes in bone mineral density1 2 Coefficient4 M odel R 2 P value5 Model DR A B M D 3 Peak load .43 .18 .04 ABM D = -.02 + .00005 peak load Impulse .19 .37 Loading rate .36 .09 Impact load .42 .18 .04 ABM D = -.01 + .00004 impact load UD A B M D Peak load .44 .19 .03 ABM D = -.006 + .00003 peak load Impulse .40 .05 Loading rate .19 .37 Impact load .36 .08 T O T A L A B M D Peak load .55 .30 .01 ABM D = -.01 + .00004 peak load Impulse .32 .13 Loading rate .31 .15 Impact load .50 .25 .01 ABM D = -.01 + .00003 impact load N=24 1 Changes in bone mineral density regressed on selected force characteristics, changes in hand grip strength, and changes in body weight after 6 months 2 Neither changes in hand grip strength nor changes in body weight significantly predict changes in bone m ineral density in all m odels 3 ABM D = changes in bone mineral density, DR = distal third radius, UD = ultra-distal radius, and TOTAL = total distal third of radius 4 (3-coefficient for ABM D 5 P value for test that (3-coefficient for ABM D = 0 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 118 Table 35: Linear relations between selected force characteristics and changes in bone mineral density in the non-damned aroun1 2 Coefficient4 Model R2 P value5 Model DR A B M D 3 Peak load .73 .54 .02 ABM D = -.02 + .00006 peak load Impulse .44 .20 Loading rate .62 .06 Impact load .76 .58 .01 ABM D = -.02 + .00006 impact load UD A B M D Peak load .45 .19 Impulse .48 .16 Loading rate .17 .64 Impact load .39 .27 TOTAL A B M D Peak load .82 .67 .00 ABM D = -.01 + .00004 peak load Impulse .75 .56 .01 ABM D = -.03 + .0002 impulse Loading rate .45 .19 Impact load .81 .65 .01 ABM D = -.01 + .00004 impact load N=10 1 Changes in bone mineral density regressed on selected force characteristics, changes in hand grip strength, and changes in body weight after 6 months 2 Neither changes in hand grip strength nor changes in body weight significantly predict changes in bone mineral density in all models 3 ABM D = changes in bone mineral density, DR = distal third radius, UD = ultra-distal radius, and TOTAL = total distal third of radius 4 P-coefficient for ABM D 5 P value for test that P-coefficient for ABM D = 0 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The influence of patella alta on knee extensor mechanics and patellofemoral joint stress

PDF

Motor skill learning in individuals with Parkinson's disease: Consideration of cognitive and motor demands

PDF

The interaction between explicit knowledge and implicit motor -sequence learning following focal brain damage

PDF

Factors contributing to patellofemoral joint stress: a comparison of persons with and without patellofemoral pain

PDF

Biomechanical and neuromuscular aspects of non-contact ACL injuries: The influence of gender, experience and training

PDF

Satellite cell proliferation: two models and two anabolic stimuli

PDF

At what age are gait characteristics mature? Evaluation of gait kinematics, kinetics, and intersegmental dynamics in 7 year-old children

PDF

Health-related quality of life, occupation and prosthesis use in elderly people with lower extremity amputation in Taiwan

PDF

Quality of daily occupational experience and its relationship with adolescent tobacco smoking

PDF

Rehabilitative narratives of individuals with neurological illnesses and their caregivers in a rehabilitation unit in central Taiwan: An ethnographic study

PDF

The search for signals provoking hyperinsulinemic compensation for insulin resistance

PDF

Cardiovascular fitness, exercise, and metabolic disease risk in overweight Latino youth

PDF

Control and dynamics during angular impulse generation

PDF

The relationship between family routines and perceived stress in mothers with children with and without a disability

PDF

Role of intestinal hormone glucagon -like peptide-1 (GLP-1) in glycemic regulation

PDF

Bimanual coordination of goal-directed multi-joint rapid aiming movements following stroke

PDF

Function of the triceps surae during strength testing and gait

PDF

The Personal Projects Approach: A method to use with persons with multiple sclerosis

PDF

Narrative practices of intersubjectivity: An ethnography of children with autism in a sensory integration based occupational therapy clinic

PDF

The physiological effects of a single bout of eccentric versus concentric resistance exercise

Asset Metadata

Creator Wang, Man-Ying (author)

Core Title Quantifying musculoskeletal load and adaptation: Biomechanical consideration

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Biokinesiology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag biophysics, medical,health sciences, rehabilitation and therapy,OAI-PMH Harvest

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-546154

Unique identifier UC11339929

Identifier 3094380.pdf (filename),usctheses-c16-546154 (legacy record id)

Legacy Identifier 3094380.pdf

Dmrecord 546154

Document Type Dissertation

Rights Wang, Man-Ying

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA