Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Determinants of bone mineral density in pre-and early-pubertal asian-and causcasian-Canadians : exercise… Petit, Moira Anne 2000

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2000-48694X.pdf [ 10.47MB ]
Metadata
JSON: 831-1.0077364.json
JSON-LD: 831-1.0077364-ld.json
RDF/XML (Pretty): 831-1.0077364-rdf.xml
RDF/JSON: 831-1.0077364-rdf.json
Turtle: 831-1.0077364-turtle.txt
N-Triples: 831-1.0077364-rdf-ntriples.txt
Original Record: 831-1.0077364-source.json
Full Text
831-1.0077364-fulltext.txt
Citation
831-1.0077364.ris

Full Text

D E T E R M I N A N T S O F B O N E M I N E R A L D E N S I T Y IN P R E - and E A R L Y - P U B E R T A L A S I A N - A N D CAUCASIAN-CANADIANS: E X E R C I S E INTERVENTION A N D FAMILY S T U D I E S  Moira Anne Petit B.A., St. Olaf College, 1992 M.S., St. Cloud State, 1994  A T H E S I S S U B M I T T E D IN PARTIAL F U L F I L L M E N T O F THE REQUIREMENTS FOR THE D E G R E E O F DOCTOR O F PHILOSOPHY in THE FACULTY O F G R A D U A T E STUDIES (Department of Education; School of Human Kinetics)  W e accept this thesis as conforrringjo the required standard  ' T H E UNIVERSITY O F BRITISH C O L U M B I A January, 2000 © Moira Anne Petit, 2000  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the university of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by her or his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  Department of  Human Kinetics  T h e University of British Columbia Vancouver, C a n a d a Date  January 30. 2000  ABSTRACT  Background: There is increasing evidence that the pre- and early-pubertal years may be a biologically optimal time for exercise to influence bone mass. T h e primary focus of this thesis was to examine the effects of a school-based exercise intervention in pre- and early- pubescent children (study 3). A secondary objective was to examine sex and ethnic differences in B M D (study 1), and familial resemblance of total body bone, lean and fat mass (study 2). Methods & Results: 1)  2)  3)  Sex and ethnic differences in BMD, physical activity and calcium intake were examined in 168 Asian (n=58) and Caucasian (n=110) boys (n=86) and girls (n=82). Physical activity, calcium intake, and femoral neck B M D were lower in Asian a s compared to Caucasian boys (p<0.05). Caucasian girls had lower physical activity, calcium intake and B M D values at the proximal femur and its regions compared with Caucasian boys. Familial resemblance was examined in 77 parents and their children. Significant correlations were observed between mothers and daughters for height, weight, lean and fat tissue, and total body bone mass (all ~r=0.50, p< 0.05). Resemblance for bone mass was also apparent for mother-son (r = 0.439, p<0.05) and father-son pairs (r=0.584, p<0.05). However, when adjusted for lean mass and height within each sex, ethnic and generation group, these correlations were no longer significant in any group. Exercise intervention: A total of 144 grade 3 and 4 children from 10 Elementary schools in Richmond B . C . were randomly assigned to exercise (n=63) or control (n=81) groups. Schools in the exercise group incorporated 10-20 minutes of moderate impact activities into their physical education classes. T h e exercise group showed significantly greater change in femoral trochanteric B M D (4.4 vs. 3.2%, p<0.05). There were no group differences at other sites either before or after controlling for potential confounding variables.  Conclusions: Results of these studies support the emerging notion that the immature skeleton is responsive to even moderate weight-bearing interventions and that other influences, including sex and ethnicity, are apparent in childhood. A school-based loading program incorporated into existing physical education curriculum is a practical way to influence bone mineral accretion during pre- and early-puberty. Longer-term intervention studies are needed to determine if benefits are maintained.  TABLE OF CONTENTS Abstract  11  List of Tables  vi  List of Figures  vii  Statement of contribution  viii  Acknowledgements  ix  Chapter 1 O V E R V I E W  1  Chapter 2 L I T E R A T U R E REVIEW  4  2.1  2.2  2.3  2.4  2.5  Bone Properties  5  2.1.1  Bone cells and cellular processes  6  2.1.2  Mechanical properties  8  Methods of Assessing Bone Strength  9  2.2.1  Issues related to D X A in pediatric populations  10  2.2.2  Summary of bone measurement  12  Bone Mineral Accrual  12  2.3.1  Rate and timing of mineral accrual  12  2.3.2  Cortical bone surfaces  14  2.3.3  Hormonal influences and assessment of maturation  16  2.3.4  Summary of bone mineral a c c r u a l . . :  18  Mechanical Loading  18  2.4.1  Mechanotransduction  18  2.4.2  Properties of osteogenic stimuli  19  2.4.3  Rationale for a greater osteogenic response in young vs. mature bone  21  2.4.4  Summary of mechanical adaptation theory  22  Heritability  .....  22  2.5.1  Twin models  23  2.5.2  Familial resemblance  24  2.5.3  Genotype  27  2.6  Prepubertal Gender Differences in a B M D  28  2.7  Ethnicity  31  2.7.1  Hip axis length  31  2.7.2  Bone mineral content/density  32  2.7.3  Lifestyle factors  33  iii  2.8  Body Weight and Lean Mass  34  2.8.1  35  Gender differences in body composition  2.9 Calcium 2.9.1  35  Interaction of calcium and exercise  36  2.10 Vitamin D and Seasonal Variation  38  2.11 Exercise Studies in Children and Adolescents  39  2.11.1 General physical activity  39  2.11.2 Studies of targeted/sport-specific loading in children and adolescents  42  2.11.3 Exercise intervention  44  2.11.4 Ground reaction forces from various activities  45  2.11.5 Limitations of exercise studies in children and adolescents  46  2.11.6 G a p s in the literature  48  Chapter 3 R A T I O N A L E , O J E C T I V E S & H Y P O T H E S E S  49  3.1 Rationale  50  3.1.1  Cross-sectional determinants of a B M D in prepubertal children  50  3.1.2  Familial resemblance  51  3.1.3  Mechanical loading intervention  51  3.2 Objectives  52  3.3 Hypotheses  53  Chapter 4 M E T H O D O L O G Y 4.1  54  Intervention S t u d y  56  4.1.1 Participants  56  4.1.2  Exercise intervention  57  4.1.3  Incentives  58  4.1.4  Measurements  59  4.1.5  Questionnaires  62  4.2  Parents  63  4.3  Statistical Methods  64  iv  Chapters R E S U L T S  67  5.1  S e x and Ethnic Differences at Baseline  68  5.2  Familial Resemblance  72  5.3  Mechanical Loading: 8-month School-based Intervention  75  Chapter 6 DISCUSSION  80  6.1  Sex and Ethnic Influences on a B M D in Prepubertal Children  81  6.2  Familial Resemblance in Total Body Bone, Lean and Fat M a s s  83  6.3  Mechanical Loading: Response to an 8-month School-based Intervention  87  Chapter 7 GENERAL SUMMARY AND CONCLUSIONS  92  7.1  Exercise as an Effective Intervention in Pediatric Groups  93  7.2  Ethnic Differences in Response to an Exercise Intervention  94  7.3  Other Determinants of the Bone Mineral Response to Exercise Intervention  95  7.4  Other Questions for Future Research  95  REFERENCES  97  Appendices  114  I.  Original manuscripts  I I 4  II.  Consent forms  I g, f  III.  Sample activities from the Healthy Bones Curriculum  I  IV.  Results for study participants  | q tj  V.  Questionnaires  i 9  v  a  "9-  LIST OF TABLES Page Chapter 2 LITERATURE REVIEW 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11  Bone cellular activity across the lifespan Cellular activity: Bone modeling vs. remodeling Studies of familial resemblance of bone mineral content (BMC) and areal bone mineral density (aBMD) measured by dual-energy X-ray absorptiometry Correlations and heritability estimates for prepubertal girls and their mothers Familial resemblance of bone mineral density (BMD), height, weight and body composition from Krall & Dawson-Hughes Studies of gender differences in areal bone mineral density (aBMD) measured by dual-energy X-ray absorptiometry in children and adolescents Descriptive data for Tanner 1 and 2 Asian and Caucasian children living in California Theoretical implication of calcium-exercise interaction Unadjusted values in peak B M C accrual velocity (Peak B M C V ) , 2 year accrual, and absolute B M C at 1 year post peak B M C V for inactive and active boys and girls Prospective or longitudinal studies of physical activity and bone in children and adolescents (< 18 years of age) Difference in percent change in a B M D between exercise and control groups in two intervention studies in pre- or early-pubescent children  7 7 25 26 27 30 33 37 41 43 45  Chapter 4 METHODS 4.1 4.2  Number and percentage of Asian children and place of birth Duration living in C a n a d a for Asian children  57 57  Chapters R E S U L T S 5.1 5.2 5.3  Descriptive variables for Asian and Caucasian boys and girls at baseline Percent of Asian and Caucasaian boys and girls participating in organized sport Areal bone mineral density regression model summaries including standardized and unstandardized Beta coefficients and adjusted r for the total body and femoral neck Number of Asian and Caucasian family groups participating in the study... Descriptive characteristics of Asian and Caucasian mothers and fathers Pearson product moment correlations comparing total body bone mineral content, fat and lean mass, height and weight between mothers, fathers, and their daughters and sons Number of participants in control and exercise groups by ethnicity and gender Anthropometric, body composition and unadjusted areal bone mineral density at baseline, and 8-month change (absolute and percent) values for exercise and control groups Beta coefficients for the final step of hierarchical regression models Pearson correlation coefficients for variables used in regression models 2  5.4 5.5 5.6 5.7 5.8 5.9 5.10  vi  69 68 72 72 73 74 75 77 79 79  LIST OF FIGURES Page Chapter 2 LITERATURE REVIEW 2.1 Stress-strain (load-deformation) curve 2.2 Velocity curve for bone mineral accrual in boys and girls. Data from the University of Saskatchewan Bone Mineral Accrual study 2.3 Cortical bone mineral accrual on the two cortical surfaces and effects on overall cortical width Chapter 4 METHODOLOGY 4.1 Overall study design showing the number of participants in each sex and ethnic group for the three measurement periods 4.2 Regions of the proximal femur in DXA scan analysis Chapters. RESULTS 5.1 Physical activity in Asian and Caucasian boys and girls 5.2 Femoral neck areal bone mineral density for Asian and Caucasian boys and girls at baseline 5.3 Change over 8-months in trochanteric areal bone mineral density for control and exercise groups  8 13 16  55 61  70 70 78  STATEMENT OF CONTRIBUTION  T h i s t h e s i s c o n s i s t s primarily of d a t a that a r e in p r e s s , o r s u b m i t t e d , f o r p u b l i c a t i o n . F o u r m a n u s c r i p t s (Table) r e s u l t i n g f r o m t h e s t u d i e s a r e a t t a c h e d in A p p e n d i x 1. In c o n j u n c t i o n with Dr. H e a t h e r M c K a y , t h e t h e s i s a u t h o r is a co-first author f o r e a c h m a n u s c r i p t . M o i r a Petit m a d e p r i m a r y c o n t r i b u t i o n s t o e v e r y a s p e c t of t h e s e m a n u s c r i p t s i n c l u d i n g : study d e s i g n ; subject recruitment; d a t a c o l l e c t i o n , m a n a g e m e n t , a n a l y s e s a n d interpretation; a n d m a n u s c r i p t writing a n d p r e p a r a t i o n . Dr. M c K a y is t h e senior author o n these papers.  oil- iu , Dr. H e a t h e r A . M c K a y  /  Date  f\fo{/, l b , M o i r a A . Petit  I  IV  Journal name, volume, year  Page limit  Augmented trochanteric bone mineral density after modified physical education classes: A randomized scholl-based exercise intervention study in pre- and early-pubescent children.  McKay, Petit,  Lifestyle determinants of bone mineral density: A comparison between prepubertal Asian- and Caucasian-Canadian boys and girls.  Calcified Tissue International (in Press), 1999  18  Petit, McKay  Familial resemblance of total body bone mineral, lean and fat mass in prepubertal Asianand Caucasian-Canadian children and their parents  Osteoporosis International (submitted), 1999  N/A  McKay, Petit,  Analysis of proximal femur D X A scans in growing children: Comparisons of different protocols for cross-sectional, 8-month and 7year longitudinal data  Journal of Bone and Mineral Research (in Press), 1999  N/A  Schutz, Khan  III  Article title  McKay, Petit,  Schutz, Prior, Barr, Khan  II  11*1  Date  T a b l e . M a n u s c r i p t s resulting f r o m this t h e s i s  Authorship  till  Bailey, Wallace, Schutz, Khan  viii  Journal of Pediatrics (in Press), 1999  20  ACKNOWLEDGEMENTS I have been fortunate to work with a group of exceptional researchers, role models and individuals. My appreciation and admiration go far beyond my ability to express them here. Special acknowledgment and appreciation to my supervisory committee for direction, guidance and support. Particularly to Dr. Heather McKay for her inspiration, creativity, hard work, trust and laughter. Heather has been (and remains) an extrodinary role model, mentor, teacher and friend. Also to Dr. Jerilynn Prior for providing endless opportunities, mentoring, friendship and passion. Dr. Susan Barrfor meticulous attention to detail and research design, and for teaching strong ethical principals. Dr. Robert Schutz for teaching me the importance of a sound study design and hypotheses prior to beginning a project and for making difficult concepts easy to understand (and for patience while teaching them!). My many thanks to honorary committee member, Dr. Karim Khan, for endless energy for your work, always teaching, and making time when you have none. I would also like to acknowledge Dr. Mike Houston, Dr. T e d Rhodes, and Dr. Alan Martin for serving as committee members at various points over the years. My appreciation to Dr. Brian Lentle, Dr. John Aldrich and Dr. Jack Taunton for serving as University examiners and to Dr. Cameron Blimkie for his through review of this document as the external examiner and for his many contributions to the field. T h e Healthy Bones Study was a collaborative effort with the support of the Richmond School Board. Laurie Nordin was particularly supportive of the project and assisted in recruiting teachers and principals. Without the teachers, principals, children and parents who participated in, and supported this work, the study would not have been possible. I a m particularly grateful to Lindsay Waddell who designed an exceptional Healthy Bones curriculum under an intense deadline, and to Gail Wilson and Judy Notte for their assistance and insightful comments on the curriculum. My appreciation and admiration to Colin Petit for being available anytime I called and for creativity and art work on Skully and posters in an exceptionally short time. I a m grateful to the many individuals who have provided me with friendship, balance, emotional (and financial!) support over the years and to the Hershey Wally's for endless humor, laughter and acceptance, regardless of my many extended absences. My love and appreciation to Libby and Pat for providing me with the best of the both worlds - the freedom to experience, and the comfort of security. Libby for your unconditional support, even when you weren't sure why/what I was doing (of course, neither was I at times!). Pat, for knowing me better than I know myself and for always having the right words to get me through. I am forever thankful for your good hearts, for teaching me what's important - and for teaching, by example, how to stick it out, regardless of the circumstances life throws you.  Chapter 1  OVERVIEW  Osteoporosis and related fractures represent an increasing health care problem which may reach crisis proportions based on current demographic trends [1, 2]. The number of fractures, their associated costs, and the resultant trauma on the lives of those affected, will increase dramatically in the next few decades, particularly in Asian and Caucasian men and women [3, 4]. Clearly, something must be done in the present to reduce the number and risk of fractures in the future. Until recently, prevention of osteoporosis and related fractures has been aimed at slowing bone loss associated with menopause through hormone replacement in postmenopausal women. However, low bone mass or bone mineral density (BMD) associated with osteoporosis is a function not only of age-associated bone loss but also of the proportion of the genetically predetermined peak bone mass attained during the growing years and preserved through mid-adulthood. Osteoporosis, therefore, may begin early in life when optimal bone mineral accretion is critical to the attainment of a healthy adult skeleton [5, 6]. Although osteoporosis is now acknowledged to have its antecedents in childhood, much less is known about the determinants of pediatric bone mineral accrual than bone in the postmenopausal years. Mechanical loading of bone through physical activity is a promising means of optimizing bone mineral accrual [7-9]. However, there has been relatively little research into the role of physical activity on BMD in children. The primary focus of this thesis was to address the broad question - Can a school-based physical activity intervention augment bone mineral in a pre- and early-pubescent population? In addition to mechanical loading, a number of other factors contribute to bone mass. Genetics, ethnicity, nutrition, body size and body composition, have all been shown to influence adult BMD and potentially interact with exercise in their effect on bone. Far less is known about the role these factors play in early childhood. Thus, a secondary focus of this thesis was to evaluate some of these determinants of bone mass, including sex, ethnicity, and heritability, in pre- and early-pubescent children.  2  This thesis consists of a literature review (Chapter 2) outlining the fundamentals of bone biology and detailing pediatric and exercise-related aspects of bone physiology. Chapter 3 lists the rationale, objectives, and specific hypotheses addressed in this research. The methodology, results and interpretation of the Healthy Bones exercise intervention study and the Family Bone Density study are presented in Chapters 4-6. Chapter 7 reviews the thesis and outlines areas for future research in mechanical loading and pediatric bone health.  3  Chapter 2  LITERATURE REVIEW  The first 4 sections (Part I) of this review provide a background for studies of mechanical loading effects on the pediatric skeleton including properties of bone, measurement issues, mineral accrual, and mechanical loading theory. Part II (sections 2.5-2.11) summarizes literature on determinants of bone mass and density in children and adolescents including genetics, sex, ethnicity, body composition, calcium intake, and exercise studies. These provide an overview of key issues related to pediatric exercise and bone mineral. Studies that are most relevant to this thesis are discussed in detail.  Part I. B a c k g r o u n d  2.1  B o n e Properties Bone is a unique form of connective tissue that provides mechanical support and has metabolic  functions. Bone matrix consists of collagen fibers and non-collagenous proteins (glycoproteins and proteoglycans). The mineral hydroxyapatite (primarily calcium and phosphate) gives bone matrix its strength and rigidity. The combination of elastic collagen and a mineralized matrix allows maintenance of some elastic properties while conferring marked rigidity and strength on the skeleton [10, Macroscopic  11]. structure.  Bone tissue is of two basic types; cortical and cancellous. Cortical or  compact bone is present primarily in the diaphyses (shafts) of long bones and the periphery of all bones, and has a relatively slow turnover rate [10].  Cancellous or trabecular bone is present  primarily in the vertebral bodies and the epiphyses of long bones and has a more rapid turnover rate [10]. Structural differences in cortical and cancellous bone are indicative of their varying functions. Collagen fibers in cortical bone are densely packed with small spaces forming a strong structure for mechanical support. Cancellous bone is made up of trabeculae organized in a more porous manner with a rich blood supply and it serves primarily metabolic functions [11]. Mechanical loading and hormonal milieu influence both types of bone, however cancellous bone (with its rich blood supply) appears to respond more readily to alterations in hormonal milieu, whereas cortical bone (with its muscular attachments) is more responsive to mechanical loading  [12, 13].  Microscopic structure. Adult bone, whether cortical or cancellous, is lamellated, meaning that bundles of collagen fibers are highly organized in alternating 'plywood-like' directions. In times of fracture or microfracture, woven bone may be formed as a temporary repair mechanism. Woven bone is less well organized than lamellar bone, but is formed and mineralized far more rapidly [14]. Eventually, woven bone will be resorbed and replaced by lamellar bone. New bone formed in response to mechanical loading is laid down initially as woven bone on the periosteal surface of cortical bone [15, 16].  2.1.1 Bone cells and cellular  processes  Bone cells are either quiescent, forming, or resorbing mineral at any given time. The activity of bone cells (osteoblasts, osteoclasts, osteocytes and bone-lining cells) determines whether, at any site, bone is lost, maintained, or formed. Osteoblasts are bone-forming cells which secrete collagen and ground substance. Osteoclasts synthesize and secrete lysosomal enzymes to digest the proteins linking hydroxyapatite crystals and collagen, resulting in resorption cavities within bone. Osteocytes (mature osteoblasts) are embedded in the bony matrix and form a communication network (via gap junctions) with osteoblasts and bone-lining cells. This network throughout the matrix allows cells to perceive and respond to mechanical stimuli [17]. Bone cells form and resorb bone to alter mass and architecture in response to varying hormonal and mechanical demands throughout life. This is accomplished by the processes of growth, modeling and remodeling. During growth, the skeleton develops into a genetically preprogrammed size and shape. Linear growth is primarily under control of the endocrine system, and ceases with epiphyseal closure. Growth itself requires the process of modeling, which alters bone shape and mass by resorbing or forming bone on diverse surfaces for an extended period of time. Osteoblasts are activated where bone is added (formation drifts) and osteoclasts resorb bone (resorption drifts). Overall, formation drifts dominate over resorption drifts and modeling's global effect is to increase bone mass and strength [18]. In contrast to modeling, in which resorption and formation are occurring on anatomically disparate surfaces, osteoblasts and osteoclasts are coupled during remodeling, the process of replacing old bone with new. Osteoblasts and osteoclasts function in organized units, or basic multicellular units (BMUs), during the bone remodeling process. The more BMUs that are 6  recruited, the greater the rate of bone turnover/remodeling. In remodeling, BMUs are active on one surface in a standardized sequence of events including: activation of quiescent surfaces, resorption, followed by formation and mineralization [19]. The entire remodeling process takes approximately 4-6 months to complete. The resorption phase occurs much more rapidly (over 3-4 weeks) than the formation and mineralization phases (4-5 months) [20], and there is a relative 1  deficit of bone over time (20 resorption cycles to 19 formation cycles) [19]. In adults, modeling drifts are basically inactive and remodeling is the dominant process. In contrast to modeling, remodeling can only maintain or decrease bone mass and strength but does not increase them [21]. Remodeling, or a form of modeling, also repairs bone microdamage that occurs in response to high mechanical loads. In young and growing animals most bone surfaces are active and primarily modeling, while in adult animals 80-90% of bone surfaces are quiescent or inactive [22]. Tables 2.1 and 2.2 summarize the activity of modeling and remodeling processes throughout life.  Table  2.1 Bone cellular activity across the life-span. Age  (y)  Primary s e q u e n c e / Process  Growth  Maintenance  Loss/Osteoporosis  0-20  20-50  50+  A-BF or A-BR Modeling  A-BR-BF Remodeling  A-BR-BF Remodeling  BF = BR BF > BR Uncoupled Coupled •adapted from Marks & Hermey [11] A=activation; BF=bone formation; BR=bone resorption Activity  Table  2.2 Cellular activity: Bone modeling vs. remodeling Modeling  BF < BR Uncoupled  Remodeling  Timing  Continuous  Cyclical  Surfaces of resorption and formation Extent  Different  Same  100%  20%  Activation*  not needed  Needed  Apposition rate  2-20 um/d  0.3- 1.0 um/d  Balance  Net gain  net loss or no change  Coupling  Systemic  local  *ln the sense of transformation of quiescent to active surface. Adapted from Parfitt 1984 [19] 1  formation and resorption time is in adult bone  2.1.2 Mechanical  properties  A few key biomechanical terms that are related to bone strength and stiffness will be defined in this section. Strength is defined as the load at which failure (or fracture in the case of bone) occurs [14] and is dependent on structural and material properties. To describe material behavior, which is independent of bone geometry, the variables stress and strain are used. When an external load is applied to bone, the bone will not only deform from its original shape, but an internal resistance, or stress, is generated (Figure 2.1). Stress is equal in magnitude and opposite in direction to the applied load. In other words, stress describes the "intensity" of the internal resistance and is expressed as the force per unit area (Pascal units or N/m ). Strain is the term used to describe the resulting deformation and is reported as the fraction 2  or percentage of the original bone dimension [1, 14, 23]. A 1% deviation in bone occurs at 10,000 microstrain.  Strain / Deformation  Figure 2.1 Stress-strain (load-deformation) curve  The structural properties of bone as a whole unit are influenced by both geometry and material properties. When a load is applied to bone, or any structure, deformation occurs in a linear fashion until the yieldpointls reached at which point the slope of the curve is reduced (Figure 2.1). In the elastic region before the yield point, a structure will return to its original shape if unloaded. When loaded within the plastic region the structure will undergo permanent deformation. The point of failure, or fracture in the case of bone, is eventually reached if the load continues to increase. Stiffness is defined as the load required to deform the structure a given amount and is equal to the slope of the curve in the elastic region [1, 14, 23]. The inorganic matrix, or mineral mass of bone,  8  determines up to 80% of a bone's stiffness and strength [23], while the organic component determines its elasticity, allowing temporary deformation of the bone under applied loading conditions. The direction of applied loads during physical activity are primarily described as either tensile (forces diverging to pull the material apart) or compressive (forces converging to push the material together) depending on the mode of loading. The load may also be described as changing the length (normal) or changing the angle (shear) of the structure. In reality, bone is rarely loaded in only one mode and most physical activities induce strains in a combination of directions. Mechanical behavior of bone is not easily measured in vivo. Estimates of the stress and strain can be calculated, however, using measures of external load such as ground-reaction forces and assumptions about related deformations. Measures of strain are generally obtained from animal models under controlled loading conditions. Although it is possible that bone responds specifically to stress, it appears more likely that bone cells respond to characteristics of strain, or some signal resulting from strain(s), engendered from mechanical loading during exercise. Osteogenic characteristics of strain are discussed in section 2.4.  2.2 Methods of Assessing Bone Strength In Vivo A number of non-invasive techniques have been used to estimate aspects of bone strength including bone geometry, and bone mass. Until recently, measures of bone geometry were limited to assessing cortical thickness or bone width from radiographs. The high radiation makes radiographs impractical for current use. Current measures of bone strength can be made from Magnetic Resonance Imaging (MRI) or estimated from Dual-energy X-ray Absorptiometry (DXA) measures of bone mineral content (BMC, g) and bone area. MRI is a precise method of assessing geometric properties of cortical bone, but it is expensive, time-consuming, and, at present, not easily accessible, making it impractical for studies of large populations. Rather than bone geometry, densitometric techniques provide a means of estimating B M C as a surrogate for bone mass, and thus provide some representation of bone strength. Bone mineral content is the amount of mineral in an anatomic region, and thus will vary simply due to size differences between individuals. In attempt to account for differences between individuals of 9  different size and growing at varying rates, BMC is typically expressed per unit area (cm ) d  and  termed bone mineral density, or areal BMD (aBMD, g/cm ). Areal bone mineral density (aBMD, 2  g/cm ) is most commonly measured by DXA or other absorptiometric techniques. A material's 2  Volumetric density implies measurement of the amount of mineral per volume (g/cm ) . 3  2  DXA is  therefore limited by its two dimensional representation of the three dimensional spatial properties of bone. As well, DXA is an integral measure of both cancellous and cortical bone. With quantitative computed tomography (QCT), it is possible to distinguish between cortical and cancellous bone in a 3-dimensional view. However, the high radiation makes it a poor choice for studies in pediatric groups. Thus, DXA remains the preferred method as it provides a precise, reasonably priced, and rapid tool for the assessment of bone mineral in vivo. DXA is widely used for both clinical and research purposes [24, 25] and aBMD measured from DXA is highly correlated with bone strength (r = 0.85 - 0.90) [26]. Because of the 2-dimensional measurement, it is necessary for researchers to account for bone size when using DXA to compare individuals who differ in size or are growing at different rates. Methods of adjusting for size and issues related to pediatric studies are discussed.  2.2.1  Issues related to DXA in pediatric  populations  A current issue in bone physiology is which value from DXA (BMC, aBMD, or bone area) is the best measure of bone strength and changes in bone quality in response to interventions such as exercise, calcium or drug therapies [27-29]. The "best" measurement of bone strength in a clinical context depends upon the ability of the procedure to predict fracture risk. The lumbar spine and proximal femur are the most common clinical sites for the measurement of aBMD and the potential prediction of fracture risk. Areal BMD identifies a large percentage of individuals at risk for hip and spine fractures [30-32] and thus is an appropriate measure as part of a risk profile. Outside of clinical assessment, the most appropriate means by which to report DXA estimates of bone strength depends on the question being asked. In studies of growing children, or when comparing individuals of different size, both B M C and aBMD are a function of, and only partially account for, differences in bone size or in rates of growth between individuals. In longitudinal studies of growth where the question is "how much total mineral is accrued?", B M C is typically reported. When the  All non-invasive methods of measuring BMD (volumetric or areal) are "apparent" densities because they include marrow space in trabecular bone and the medullary cavity in cortical bone. "True" density implies the amount of bone contained within a region, all of which is made up of its own material [Seeman]. True bone density seems to be constant. 2  10  question is looking at effects of an intervention or other determinants, it is important to account for individual differences in size and rates of growth. Areal density partially accounts for size and has the added advantage over BMC of being more precise and less prone to operator errors due to position or analysis as both BMC and bone area change in the same direction. The problem with aBMD is that it does not include bone depth and thus only partially accounts for size differences. Thus, it is necessary to use additional means of controlling for size differences between individuals [27, 33]. A popular approach to adjust for size differences is to assume a basic geometric shape for the bone region of interest and calculate an "apparent density" (BMAD) [27, 34]. Any equation estimating a volumetric density is dependent upon several assumptions about the shape of the bone and the geometric similarity between individuals within the cohort [35]. The validity of these assumptions, and the usefulness of estimated volumetric equations and other aspects of bone geometry measured from DXA (such as cortical cross-sectional area), has recently been challenged. For example, the difference in cortical cross-sectional area (CSA) estimated from DXA, and CSA measured from MRI techniques varies significantly between individuals of the same age. In two 18y-old female runners, the difference in CSA of the femoral shaft between the two techniques was 13 and 45% (Dr. Cameron Blimkie, personal communication). Also, relevant sites such as the femoral neck and vertebrae are more complicated in their geometry than is assumed by BMAD equations [35] and bone dimensions change during growth [36], and possibly with loading interventions: There may also be geometric dissimilarities in bone dimensions between individuals or groups (i.e. vertebral height and femoral neck differences). An alternate approach to using BMAD, in which fewer assumptions are made, is to account for size differences within multiple regression models [37] by entering size or growth indices in all regression models. For example, if BMC is the dependent variable, then bone area, height, body weight or composition, may be forced into the model prior to calcium or physical activity. To control for differences in rates of growth in longitudinal studies, and when change in BMC or BMD is the dependent variable, it would be appropriate to force baseline BMC or BMD, change in height and change in body weight, lean and/or fat mass, to control for differences in rates of growth in longitudinal studies. This approach allows for population specific models to be derived for both cross-sectional and longitudinal data sets. 11  2.2.2  Summary  of bone  measurement  Dual-energy x-ray absorptiometry estimates B M C and areal density of clinically relevant sites with a high degree of precision. The short scan time, and low effective radiation dose make DXA an ideal method of estimating B M C and aBMD in children. The primary limitation of DXA is that it only accounts for two, rather than three, dimensions of bone. Therefore, when comparing individuals who differ in size over a period of growth, it is important to account for size differences and differences in rates of growth. There are also a number of issues related to analysis of DXA proximal femur scans in pediatric populations. However, there are to date, no published studies outlining the effect of different scan analysis techniques in pediatric populations . This is a potential source of error as pediatric scan analysis is operator dependent.  2.3 Bone Mineral Accrual In a broad perspective, bone mineral is gained during the first three decades of life, and conserved from maturity until bone is lost in later adulthood. However, the view of skeletal growth simply as a gain in bone mineral, masks the differential rates of growth and heterogeneity at diverse bone sites and surfaces. Bone mass during childhood and adolescence increases via linear growth, increased cortical area through periosteal and endosteal apposition, and via an increase in trabecular thickness and number. The following section includes a summary of the pattern of bone mineral accrual during childhood and adolescence, sexual dimorphism in the rate and magnitude of mineral accrual, and accrual patterns on the two cortical surfaces. Studies of gender differences in absolute aBMD are reviewed in section 2.6.  2.3.1 Rate and timing of mineral  accrual  Most of what is known about the rate and timing of mineral accrual comes from cross-sectional, and a few longitudinal studies. The most comprehensive study of mineral accrual is the University of Saskatchewan Pediatric Bone Mineral Accrual Study, in which B M C was measured at the total body (TB), femoral neck (FN) and lumbar spine (LS) annually for 7 years in approximately 200 boys and girls. This study showed that 90% of adult bone mineral is acquired by the end of adolescence  12  at most sites [5, 7]. These data are in keeping with cross-sectional studies of BMC during growth suggesting there is minimal change in bone mass or density at lumbar spine and proximal femur sites in girls after 15-16y and in boys after age 18y [38-40]. Girls in the Saskatchewan study achieved -60% of their peak total body bone mineral (TBBMC) by 11.6 y, while boys reached 60% of TBBMC approximately two years later at 13.5 y [5]. The rate of mineral accretion accelerates dramatically during puberty (Figure 2.2) and is closely linked with sexual maturation and changes in hormonal milieu (section 2.3.3). Thus, in studies of skeletal development, it is important to control for the tremendous variability in maturation between children of the same chronological age [41]. An important aspect of the Saskatchewan study is that maturation is controlled for by aligning children at a common maturational landmark, peak height velocity (PHV), rather than by B M C T B Velocity Curve  chronological age. Of the 200 children  Cubic Spline  in the study, growth curves were fit for  450-, • Boys Age Ageeor Peak 14.05 Peak Value 409 Size Adjusted 394  400-  ro 350>» <u Q. 300O) c 250L_  iu  o  > ffi  Age P H V 13 44 yrs AgePHV / 11.77 yre*j  Age of Peak Peak Value  12.54 325  60 boys and 53 girls who had enough data points to identify PHV and peak BMC velocity. Peak total body BMC velocity was greater in magnitude for boys than girls (409 and 325 g/y) when  200-  controlling for height and weight [42].  1- 150O 2 m 100-  Boys also had greater peak BMC accrual rates than girls at the FN, but  50-  not at the LS.  09  10 11 12 13 14 15 16 17 18 19 Age in Years  Figure 2.2 Velocity curve for bone mineral accrual for boys (solid line) and girls (dotted line). Data from the University of Saskatchewan Bone Mineral Accrual study [42].  13  In addition to peak accrual rates, researchers identified the area under the curve (BMC accrued) in the 2 years around PHV. These 2 years are considered 'critical' years in skeletal development. Approximately 25% of TB BMC was laid down in the 2 years around PHV [42]. A 20% greater accrual rate for boys at the FN and TB translated to greater absolute BMC values 1 year post peak BMC velocity - which was the closest to adult values. In contrast, there were no gender differences at the spine in rates of accrual, or absolute values [42]. These accrual values are substantially higher than those reported in previous cross-sectional analyses [43, 44]. This study also clearly illustrated the, now well-established, dissociation between linear growth and bone mineral accrual. This study and others demonstrated a 7 month to 1 year lag time between PHV and peak TB BMC velocity in both males and females [7, 42, 45] (Figure 2.2). This lag time is theoretically necessary to cover the high calcium demand of both linear growth and mineral accrual. It may be that the calcium requirement for both increased skeletal size (length and width) and mineralization is too high to be covered from dietary sources alone. Parfitt theorizes that mineral is temporarily "borrowed" from the cortical shell during the period of most rapid linear growth (PHV) [46]. Theoretically, when linear growth ceases, the mineral "debt" is repaid and matrix mineralization can continue at a rapid rate [46]. Thus, the period between PHV and peak BMC velocity is a time of relative fragility for the skeleton. Interestingly, the prevalence of forearm fractures peaks during the time period between PHV and peak BMC accretion [47-49]. Further, recent data showed lower aBMD in young girls (aged 3-15 years) with forearm fractures compared with age-matched, fracture-free, controls [50].  2.3.2 Cortical bone  surfaces  Differences in bone mass between males and females are greater at predominantly cortical sites, likely due to larger bone size and cortical shell size in males [51, 52]. Cancellous bone mass (measured by QCT) at the lumbar spine is relatively stable and similar between boys and girls [53, 54]. Therefore, the overall gender differences appear to be primarily a reflection of cortical bone. The periosteal and endosteal cortical bone surfaces behave somewhat differently during growth in males and females.  14  The early work of Garn and colleagues [55-57] is invaluable in showing differential patterns of mineral accrual on the two cortical surfaces. The researchers monitored changes in cortical bone surfaces, reporting over 10 years of longitudinal and cross-sectional data. Apart from a short period just after birth, adolescence is the only time when cortical bone expansion occurs on both its inner (endosteal or endocortical) surface, as well as its outer (periosteal) surface. During other times of life periosteal apposition is matched by endocortical resorption [46, 56] (Figure 2.3). The pattern of periosteal growth parallels the pattern of linear growth, with a rapid spurt in early childhood, followed by a slower rate of growth until a second spurt begins in early adolescence. Apart from the brief growth spurt in early childhood, the greatest increase in periosteal diameter occurs in boys between 12-14 years (0.40 mm/yr) and in girls from 10-12 years (0.30 mm/yr). Periosteal diameter continues to expand throughout life adding about 2%/yr from age 30-80 [56] and is greater in males throughout life [55]. The cortical endosteal surface behaves much differently than the periosteal surface. A period of resorption in childhood (as indicated by medullary cavity expansion) is followed by a brief period of apposition during adolescence and a second resorptive phase in later life [55]. While boys gain more cortical bone periosteally, girls tend to gain bone on the endosteal surface. Apart from a brief period of apposition in the first half year of life, the first resorptive period lasts until the second decade in both genders. Boys have a larger medullary cavity than girls and a greater rate of resorption on the endosteal surface (6%) through the first 10 years of life. In girls, the greatest time for endosteal apposition is relatively late in puberty - after PHV, menarche and the periosteal growth spurt. While most pubertal events occur approximately 2 years earlier in girls than boys, endosteal apposition occurs around 4 years earlier in girls (14y vs. 18y) coinciding with menarche. The magnitude of apposition is larger in girls and the growth spurt occurs from age 14-16. Apposition rate slows, but continues in both genders until the second resorptive phase which begins sometime in the 4th decade [55]. The overall effect of increased periosteal expansion and medullary contraction is a larger cortex and stronger mechanical structure. The pattern of cortical growth follows the periosteal surface with a steady increase during childhood, an adolescent spurt, a slight gain during adulthood and decreased thickness in later adulthood [55].  15  a)  b)  2  II  I I I I I I I I I I I I I O  CN —  CO —  O •q-  i  r~vO  Age (years)  Figure 2.3 Cortical bone mineral accrual on the two cortical surfaces (a) and effects on overall cortical width (b). Graph is adapted using data from Garn [55] and Parfitt [46]. The upper lines show changes in periosteal diameter - increasing slope is representative of apposition, decreasing slope of resorption. The lower two lines show changes in the medullary cavity - positive slope represents endosteal resorption (i.e. a widening of the medullary cavity) and negative slope represents endosteal formation.  2.3.3 Hormonal  influences  and assessment  of maturation  The primary systemic hormones involved in skeletal development are growth hormone (GH), insulin-like growth factor 1, estrogens and androgenic hormones. GH is thought to be the most important regulator of longitudinal bone growth [58].  Although the mechanisms by which G H  stimulates long bone growth remain unclear, it is likely that effects are mediated through IGF-I. Both G H and IGFs stimulate osteoblast cell differentiation and thus bone formation. GH has also been associated with total body B M C in adolescent girls (14-18y) [59]. Estrogen appears to have a bimodal effect on bone. At low doses, that are inadequate to stimulate development of secondary sex characteristics, linear growth is stimulated, partially via its influence on G H [60], however children with G H insensitivity have a growth spurt at sexual maturity [58] suggesting an independent influence of estrogen. In contrast, higher doses of estrogen assist in decreasing the rate of linear growth and epiphyseal closure. A recent case-study illustrated the important role of estrogen for epiphyseal closure in males. A 28-year old male with an estrogen receptor gene mutation had continued long bone growth and lack of bone mineralization despite normal androgen levels [61].  16  For clinical purposes, estrogens are termed 'anti-resorptive', but because resorption and formation are linked, estrogen treatment slows both formation and resorption [62]. Progesterone stimulates bone formation [63], however levels of progesterone are low until normal ovulation is reached, and does not likely play a major role in stimulating bone modeling during growth. The exact role and mechanism of action for each of these hormones in bone mineral accrual remain unclear. However, the timing of mineral accrual is closely linked with indicators of endocrine maturation. In the 53 girls from the Saskatchewan Peak Bone Mineral Accrual Study that were monitored over 6-years, peak BMC velocity and menarche occurred at virtually the same age (12.5±0.9y, and 12.7±1 .Oy) [41]. Tanner breast and pubic hair stages, indicators of sexual maturation and hormonal milieu, are also closely linked with bone mineral accrual. Markersof bone turnover are maximal during Tanner breast stages 2-3 [64] corresponding to increases in pulse amplitude of gonadotrophins and high levels of GH and IGFs. Sex steroids are still generally low at this point but begin to rapidly increase around Tanner stage 3-4, coinciding with peak height velocity and peak BMC velocity [41]. It is important to note that there is tremendous variability in maturational stage and rate of maturation between same sex children of the same chronological age [41]. Clearly, bone mineral accretion is primarily dependent on stage and rate of maturation [39], and thus these need to be accounted for in studies of growing children. A practical means of assessing maturation in large studies is through self-assessment of Tanner stages. Self-assessment of maturity is strongly correlated with staging assigned by an endocrinologist [65, 66]. Matsudo [66] studied 352 human males and females 6-33 years. Maturity assessment was made both by medical doctors and by the particpants viewing line-drawings of Tanner stages [67]. Children and physicians agreed most often at Tanner stage I (93%) and V (100%), and least often at Tanner IV (23%). Among children 6-10 years, girls and physicians agreed on breast ratings 66% and pubic hair 78% of the time, while boys and physicians agreed on genital and pubic hair ratings 70% and 63% of the time respectively [66]. Duke et al. assessed selfratings in 43 females 9 to 17 years and 23 males 11 to 18 years [65]. Children and physicians agreed on 37/43, 40/43, and 21/23 of breast, female pubic hair, and male genitalia stages, respectively. None of the participants' ratings were different from the physician's rating by more than one stage [65].  17  2.3.4. Summary of bone mineral  accrual  The overall amount of bone mineral increases during linear growth when modeling is the predominant process and is generally believed to be maintained during adulthood when formation and resorption are coupled in remodeling. Childhood and adolescence are a unique time in which bone surfaces are active and rapidly forming and mineralizing bone matrix. Thus, any lifestyle factors influencing bone mineral accretion may have their greatest, and most lasting, effect if introduced during pre- or early puberty. Gender differences in the timing and rate of endosteal and periosteal mineral apposition may influence the bone response to loading during childhood and adolescence.  2.4 Mechanical Loading The advent of DXA, and a subsequent plethora of research on factors influencing bone mineral, has increased public awareness of the potential role of physical activity in achieving and maintaining a strong skeleton. Although only recently popularized, the concept that bone alters its shape and mass in response to mechanical loading is not a new idea. In the 17 century Galileo th  observed that bone may adapt its structure to its specific loading environment by noting that body weight and physical activity were related to bone size [68]. By the 19 century, a group of lh  researchers including Culmann, Meyer, Roux and the German anatomist Julius Wolff, had contributed considerable work to define what is now known as Wolff's law:  "The law of bone remodeling alterations  of the internal  alterations  of the external  is the law according to which  architecture...as form  well as secondary  of the bones...occur as a  consequence of primary changes in the shape...or in the stressing of the bones" [69].  2.4.1  Mechanotransduction  In recent years, researchers have focused on identifying mechanisms of cellular adaptation to the external loading environment. The transfer of mechanical loading-induced stimuli into chemical signals and subsequently to cell and tissue response is termed mechanotransduction.  18  Although  there are a number of theories, the prevalent view is that fluid flow mediates the mechanical signal either directly in response to strain (deformation), or by some electrical effect related to streaming potentials [70-72]. It has also been suggested that bone cells respond directly to deformation from mechanical loading [73]. A series of in vitro and in vivo loading studies in animals show that bone cells respond to strain (deformation), either directly or indirectly via fluid flow. Osteocytes are the most likely candidates to detect changes in extracellular fluid flow and transmit signals to the osteoblasts via gap junctions [17, 74, 75], however the exact biochemical pathways of mechanotransduction are unknown. It is clear that bone cells respond rapidly to mechanical loading both in vitro and in vivo. Cellular responses proportional to the degree of strain are observed in second messengers (i.e. c-AMP, IP , and DAG), growth factors (i.e. IGF-1), and these 3  are associated with subsequent changes in bone architecture and mass. Levels of osteopontin [76], IGF I mRNA [77], glucose-6-phosphate dehydrogenase activity, nitric oxide [78], and prostaglandin E have been shown to increase rapidly in response to loading bouts [79-81]. The 2  exact role of these gene products in altering the (re)modeling cycle are currently under investigation and not all models support current theories [82]. A detailed discussion of these theories is beyond the scope of this thesis, but there are several excellent reviews which discuss the evidence for cellular response to loading [17, 81, 83].  2.4.2 Properties  ol osteogenic  stimuli  While cellular mechanisms of bone adaptation are still unknown, biomechanical theories of bone adaptation are more clearly defined. Frost [84] introduced the term "mechanostat" to describe the mechanism(s) that control cellular activity of bone in response to strain induced by mechanical loading. Peak strains during a number of functional activities are remarkably similar ranging from 2000 to -3500IJE [85], suggesting that bone functions at an 'optimal' level of strain in all species. 3  The bone response to mechanical stimuli varies depending on the strain magnitude, distribution, rate and frequency. Strain magnitude: Strain magnitude is perhaps the most widely studied feature of mechanical stimuli. Frost [86] proposed three minimum effective strain thresholds (MES); the remodeling (MESr), modeling (MESm) and the microdamage (MESp) thresholds. The MESr occurs  3  strains are negative because primarily compressive forces activate (re)modeling  19  somewhere near 50-100 |iE. Strains below this range are perceived as "disuse", stimulating increased activation of bone resorption sites (and reduced formation by already activated sites) leading to an overall loss of bone mass. At strains above the MESr and below M E S m (800-4000 uE), remodeling is activated in the "conservation mode" so that the number of BMUs are maintained and formation and resorption are coupled [21]. Loads stimulating bone strains above the M E S m and below the MESp (center ~3000uE) stimulate modeling and deposition of lamellar bone, while above the MESp, microdamage occurs and woven bone is laid down. Animal data [87, 88] support Frost's threshold levels, but factors other than strain magnitude are important to optimize bone formation with loading. As well, threshold levels vary considerably depending on bone site, and can be altered by hormonal milieu [81, 89] or "error" of the strain. Strain distribution: Lanyon [90] added to the mechanostat theory by showing that even very high strains will not produce an adaptive response if they are normal in distribution. In other 4  words, bone cellular response is "error driven" and strains must be unusual in their distribution (the "error strain distribution" hypothesis) for adaptive (re)modeling to occur [90]. Only a few studies have examined this hypothesis to date, but results show that even strains within physiological levels can have significant osteogenic response when they are applied in an unusual direction [85, 88, 91]. In contrast, even very high strains are not osteogenic (although they may maintain bone mass) if presented in a normal distribution. Strain rate: Loads of high magnitude and unusual distribution can be less- or in-effective if they are applied statically or continuously. Rubin and Lanyon [85] showed that static/continuous application of a load was ineffective in protecting turkey ulnar bone, and indistinguishable from the effects of disuse. The same load applied at high rate of 100 cycles per day stimulated bone formation and increased ulnar area. Turner and colleagues [92] showed a marked linear response in both bone formation and mineral apposition with increased strain rate in rat tibias. A similar dose-response to strain rate was illustrated in a group of young, growing male rats. Loads between 1 and 20 N were applied to tibia and ulna at 3 strain rates. Across the range of load applied, the high-strain-rate group showed a 54% greater osteogenic response than the moderate- strain-rate group, who in turn had a 13% larger response than the low-rate group [93]. This study also  'normal' refers to the usual or customary load to which a specific bone has adapted. In a very sedentary individual, walking may cause a distribution that is unusual enough to stimulate an osteogenic response. However, walking is unlikely to lead to an osteogenic response in a normally active individual. 4  20  illustrated location-dependent increases and decreases in both formation and resorption - indicating a very complex architectural response of growing bone to mechanical loading [93]. In the context of mechanotransduction, high strain rates would have a greater potential to initiate cellular events because of a more rapid fluid flow, while low rate strains would be relatively slow and ineffective [94]. Strain frequency. If loading cycles are unusual in their distribution, a maximal response can be gained from only a few cycles of high impact loading. Cellular response is stimulated with one loading cycle and a maximal osteogenic response can occur after only a few strain cycles [85]. In a study of mature animals, 36 cycles of loading were as effective at stimulating bone formation as 360 and 1800 [85]. Optimal formation may occur with even fewer cycles in young animals. A recent study found that 5 jumps/day were as effective at increasing mass in young rats as 100 jumps/day [95]. Others have shown that running 3 or 18 km/day may benefit bone mass in rats [96] and that increasing the magnitude of the loads with weighted backpacks is more effective stimulus than increasing the number of strain cycles [97].  2.4.3 Rationale for a greater osteogenic  response  in young vs. mature bone  Animal studies are conclusive in demonstrating that mechanical loading is necessary to build and maintain bone mass and structure and that young bone responds more favorably to loading than old bone [81, 87]. Strain thresholds for bone formation from mechanical loading increase with age [98, 99]. In younger animals, mechanical loading induces net bone formation at loaded sites, but the same load only prevents bone loss in mature bone [99]. There are several plausible explanations for the greater response of immature bone to loading. First, young bone is incompletely mineralized, and thus any given load would induce a greater strain on young bone compared to more mature bone. During growth, there is already a rapid increase in muscle mass which would induce a high strain on undermineralized bone, and thus stimulate (re)modeling activity and global bone formation. Any additional stress due to exercise would increase bone strain, accelerating the formation process [21]. Another reason for the greater response of young bone to loading is that bone cells are highly active in the modeling process during growth. As previously discussed, unlike remodeling, which dominates in adult bone, modeling does not require resorption prior to bone formation. Formation 21  drifts during modeling occur in one place for several years, while in remodeling, osteoblasts function in one place for a short time. Also, more of the surface in adult bone is quiescent and must be activated prior to formation. Mechanical loading increases osteoblast recruitment and during childhood this increases the number of modeling drifts and thus, the potential for increased bone size/mass relative to adulthood.  2.4.4  Summary of mechanical  adaptation  theory  Bone cells respond to strain induced by mechanical loads by adapting their (re)modeling activity in favor of formation if strains are above the MESm. A number of variables increase or decrease the loading threshold including sex steroids, bone age, or other strain characteristics (strain distribution, rate, or frequency). Strains that are different from the normal loading patterns and high in magnitude, although not necessarily high in number, produce the greatest osteogenic response. Young bone has a greater potential to respond to exercise than older bone - partially because of the rapid modeling on a number of surfaces and the lower strain threshold. The challenge is to translate loading thresholds in animals into exercise prescriptions for humans.  Literature Review Part II. Determinants  2.5  of aBMD in Prepubertal  Children  Heritability The first step in the study of genetic influences on quantitative phenotypes, such as aBMD, is to  show that familial resemblance exists. A number of studies show resemblance for both B M C and aBMD using twin models (section 2.5.1) and correlations of (primarily) adult mother-daughter pairs (section 2.5.2). Simple correlations and/or heritability estimates are most commonly reported to describe the strength of familial resemblance. Heritability is defined as the proportion of total population variance in the trait that is attributable to genetic factors [100] and expressed as a percentage.  6  As discussed in a recent review [100], reporting heritability as a proportion gives a somewhat distorted representation of the change in genetic variance with age or in different populations. To  6  Heritability (%) = 100 x genetic variance/total variance  22  use the clear examples given by Seeman and Hopper [100] - if the total variance (which is a combination of environmental and genetic influences) increases with age due to both an increase in environment and genetic influences, but the environmental influence increases more than the genetic influence, then the heritability estimate would decrease. Similarly, if total variance increased due to an increase in environmental variance and genetic variance remained the same, the heritability estimate would decrease. Reporting heritability (a proportion) and genetic variance (an absolute) may give a very different picture of the 'strength' of the genetic influence [100]. This has limited our understanding of the relative contribution of genetic factors to aBMD at different stages of growth, and of the influence of environmental factors on a highly heritable trait. Despite the limitations of estimating heritability, a number of insights have been gained from twin models and family studies.  2.5.1 Twin models Studies of monozygotic (MZ) and dizygotic (DZ) twin pairs consistently indicate a strong genetic influence on aBMD [101-105]. In 56 MZ and 56 DZ adult female twin pairs, the genetic variance accounted for 65% of the total variance at the femoral neck [106]. The twins ranged in age from 24-67 y (mean 45 y). Other studies of adult and elderly twin pairs report heritability estimates ranging from 50-80% [101-105, 107]. A few assumptions may be violated with twin models, thus inflating heritability estimates [105]. Twin models assume environmental factors are similar between MZ and dizygotic DZ twin pairs, and that there are no gene-gene interactions. However, MZ twins may have more environmental commonalties than DZ twins [108] and gene-gene interactions are likely [109]. Thus, heritability estimates from twin models are considered to be maximum estimates. Further, the twin studies to date are primarily of adult or elderly twins [101104, 110]. There is some evidence that the absolute genetic variance changes depending on the age of twin pairs [111-113]. Depending on the age of the twins and the site of measurement, heritability estimates for aBMD may decrease after controlling for lean mass [106, 111]. Heritability estimates decreased by 16% at the femoral neck when lean mass was taken into account, while there was no change in heritability estimates at the lumbar spine [106]. The genetic factors accounted for 60-80% of the individual variances of both FN BMD and lean mass, and more than 50% of their covariance. The 23  association between greater muscle mass and greater BMD was attributed to genes regulating size [106]. In postmenopausal twin pairs, however, evidence was strong for an environmental influence on lean mass, and no evidence for a shared gene effect. These data suggest it is important to account for lean mass in studies of heritability or familial resemblance, yet few studies have. Most adjust for age alone. The lean mass-bone association is discussed further in section 2.8.  2.5.2 Familial  resemblance  In addition to twin studies, mother-child, father-child and sibling set studies also generally support a familial resemblance in BMC and aBMD [114-117]. There are, however, conflicting reports which found no relationship in mother-daughter pairs [118]. Early studies of familial resemblance utilized single- or dual- photon absorptiometry (SPA/DPA) and measured BMC/BMD at the radius [118-120]. More recent studies in which DXA measures of BMC/aBMD were utilized are summarized in Table 2.3. Most of the studies measured adult mother-daughter pairs. Correlations for mother-daughter pairs for BMC/aBMD averaged -0.40, and heritability estimates -50% across the studies. A limitation in several of the studies is the combining of pre- and postmenopausal daughters with their postmenopausal mothers [116, 121]. Heritability estimates are partially a function of the ages of each individual within the pair [111-112, 121] thus limiting studies with a wide variance in age. There are three studies of younger children and their parents. Lonzer et al. examined aBMD at the lumbar spine in 28 families [123]. Children ranged in age from 5 to 20 years and parents ranged from 30-55 years. Correlations were run on aBMD scores adjusted for age and sex for mother-child (n=16) and father-child (n=8) pairs [123]. The relatively high coefficients (0.58 to 0.85) are difficult to interpret in light of the small sample size, wide age range of parents and their children, and the combining of both sons and daughters. In another study, Francois et al. showed high heritability estimates in a large sample (n = 175 pairs) of mothers and their 16-20y old daughters [124]. Only the lumbar spine was reported and scores were adjusted for body mass index alone. Heritability estimates for BMC (53%) and BMD (63%) increased slightly, by 4 and 5%, when adjusted for body mass index. One study of growing boys (12.7y) and girls (11.8y) and their mothers (40.0 y) examined familial resemblance at a number of sites including the proximal femur, femoral neck and lumbar spine. Correlations were significant at all sites for mother-daughter pairs 24  when adjusted for maturity (daughters) and age (mothers), however resemblance was not apparent for sons and their mothers [117]. Overall, these studies suggest moderately high familial resemblance between adolescent children and their mothers. However, the wide age range, and lack of control for maturational differences in children, with the exception of one study [117], limit the conclusions that can be drawn from these studies.  Table 2.3  Studies of familial resemblance of bone mineral content (BMC)/areal bone mineral density (aBMD) measured by dual-energy X-ray absorptiometry (DXA). Factors adjusted sites r h'(%) n mean age (range) Pairs Reference for  Parent-Young Ferrari 1998  M-D  McKay 1994  M-D M-S  Lonzer 1996  M-C F-C Mid-C M-D  Francois 1999  M- 40.0 D - 8.1 (6.6-11.1)  138 pairs  M-40.0 D - 11.8 S-12.7 C - 5-20 P - 30-55  42 pairs  MD-16y+  M - 16 F-8 C-28 175 pairs  Child 18-36  rfor age h for calcium, ht. & wt.  LS FN MF  .25-.35*  PF LS FN LS  D-.37-.49* S-.22-.30  N/A  age (M), maturity (D/S)  .58-.85*  N/A  age,sex  LS  N/A  53-68  BMI  .16-.30*  34-63  age and weight  .30-.43  16  age, postmenopausal age (M), height, weight age, height, weight, and significant lifestyle factors weight, age or bone age (for D/S under 18y)  2  Parent-Adult Child Danielson 1999  M-D  M - 71.7 (65+) D - 48.5 (30-70)  207 pairs  Hansen 1992  M-D  M - 6 0 . 6 (45-72) D-38.5  144-D 101 - M  Krall 1993  Gueguen 1995  40 M - 6 0 (43-71y) families F - 63 (44-75) n = 160 D - 3 1 (21-45y) S - 32 (21-44y) M-129 models M-41.9 (34.2-49.9) using all F - 44.0 (35.5-53.4) F-129 D-18.1 (15.0-22.4) D-98 data S - 18.4 (14.2-25.8) S-85 25 pairs M-D M -72y D-41y M-D/S F-D/S mid-D/S  LS FN PF LS FN  TB LS FN TB  D - .22-.52 4 6 - 6 2 S - .27-.58  .32  64  age, body weight, ERT (M) and daughters' lifetime activity significant = p < 0.05; M=mother, F=father, D=daughter, S=son, mid=mid-parent value, C=child, h*=heritability estimate; r = Pearson Correlation Coefficient, TB - total body, PF - proximal femur, LS - lumbar spine, MF midfemoral diaphysis Ulrich 1996  25  TB LS Legs  .32 .15 .44*  N/A  Only one study, to my knowledge, has examined familial resemblance of BMC and aBMD in prepubertal girls [125] and there are no studies that investigate parents and their sons. In 138 prepubertal girls (age 6.6-11.1 y) and their premenopausal mothers (mean age 40.0y), associations for bone area, BMC and aBMD at the lumbar spine, femoral neck and midfemoral diaphysis were significant (Table 2.4). Correlation coefficients for age-adjusted bone scores ranged from 0.230.34. Heritability estimates derived from regression models adjusting for age, weight, height and calcium intake ranged from 18-36%. These estimates are -20% lower than in studies of adult women and their premenopausal mothers [115, 121]. Differences in heritability could be due to genetic or environmental resemblance, or to age-associated changes in total variance. Table 2.4 Correlations and heritability estimates for Variable Bone area LS  FN MF BMC LS  FN MF aBMD  Correlation (R)  Heritability estimate (%)  0.36 0.34 0.22  35 23 26  0.35 0.35 0.23  34 26 27  0.31 33 FN 0.34 36 MF 0.32 18* LS = lumbar spine, FN = femoral neck, MF = mid-femoral diaphysis. All correlations and heritability estimates were significant (p < 0.01); except *p = 0.03. LS  One key study of familial resemblance in bone at a number of sites has included parentoffspring associations with sons and fathers  [115].  Krall & Dawson-Hughes [115] reported one of  the few DXA studies to include mothers, fathers, daughters and sons from the same family. They studied a total of 40 families with one postmenopausal mother, father, premenopausal daughter and son from each family. Bone scores were adjusted for age, height, weight and significant lifestyle factors within each generation group. In a follow-up study, they also report resemblance for lean and fat mass  [126].  Specific correlations are listed (Table  2.5).  Overall, resemblance was  stronger for mother-daughter and mother-son pairs than for father-child pairs. It is not clear why these gender differences exist and studies identifying specific genotypes related to aBMD are necessary to further explore sex-related differences in heritability.  26  Table 2.5 Familial resemblance of bone mineral density (BMD), height, weight and body composition from Krall & Dawson-Hughes[115, 126]. F-D F-S mid-D mid-S M-S M-D .57* .11 .24 .46* .54* .54* TB BMD .22 .58* .31 .47* -.12 FN BMD .40 .34 .37 .24 .28* .16 LS BMD .30 .39 .72* .26 .58* .60* Height .38 .44* .57* .38 .38 Weight .49* .28 N/A N/A .36* .23 .05 Lean .35* N/A N/A .13 .07 .26 .07 Fat *M - mother, F - father, S - son, D - daughter, mid - midparent value TB = total body, FN = femoral neck, LS = lumbar spine 2.5.3 Genotype  Recent research has focused on identifying genotypes that are 1) related to aBMD and, 2) represent high risk for fracture or low aBMD. It has been suggested that aBMD is both polygenic and oligogenic meaning that a number of genes contribute small, additive effects (polygenic) and there are severai single genes with important independent effects (oligogenic) [109]. The specific genes responsible for regulating bone mineral accrual, peak bone density, and bone loss have not been identified clearly but a number of candidate genes are under investigation which may regulate aBMD either independently or in combination including the estrogen receptor, calcitonin, transforming growth factor B, interlueukin-6, and collagen type I (COLIA1 and COLIA2) [127-129]. A majority of the initial work centered around the relationship between aBMD and the vitamin D receptor gene. Morrison and colleagues [130] were the first to report variants in the vitamin D receptor (VDR) gene related to aBMD. The initial polymorphisms were identified as either the presence (b) or absence (B) of the Bsrrr\ restriction site. Twins with the bb genotype had up to 1.3 standard deviations higher aBMD than their siblings with BB genotype and those with Bb alleles had intermediate aBMD [130]. While the results of Morrison et al. have been supported by others [131, 132] there have been several studies showing no relationship between BMD and VDR genotype [133]. Controversy has ensued regarding the role of the VDR gene. Some researchers now suggest that this gene explains little, if any, of the genetic variance in aBMD, although others still support its effect [134, 135]. Part of the discrepancy maybe explained by gene-gene or geneenvironment interactions. It also appears that the genes influencing peak aBMD likely differ from those influencing bone loss [129].  27  A few studies have examined the relationship of VDR polymorphisms to bone mineral accretion. Results of two longitudinal studies in children and adolescents conflict. Ferrari et al. [136] reported a higher rate of aBMD accretion in bb compared'with BB 8 y old girls. In contrast, Gunnes et al. [137] found no influence of VDR genotype on 3 year aBMD gain or absolute aBMD in boys and girls aged 8-21.  Although the sample size in the Gunnes study was large (n = 494), the  large variation in age and maturity of the population limited its utility [137]. Although discrepancies exist in the literature at present, it is clear that aBMD is controlled by a number of genes and that they interact with environmental factors in their effect on bone. It would be useful if future research identified specific bone-related genotypes and the interaction with exercise response - particularly once specific genes controlling aBMD were defined.  2.6. Prepubertal Gender Differences in aBMD There are established differences in B M C and aBMD in adult men and women. It is generally believed that sex differences in mineral accrual and absolute values of bone mass are not apparent until puberty. A number of studies showed absolute values for BMC/aBMD are similar between boys and girls almost sites prior to puberty [138, 139]. Total body and femoral neck values in boys are greater, however, once puberty begins [138, 139]. However, there are conflicting reports and, until recently, most studies have included only small numbers of pre-pubescent children. Studies are summarized in Table 2.6. Most studies are cross-sectional and report aBMD at the lumbar spine only. Although in most cases, lumbar spine aBMD is similar between males and females from ages 1-20 y [140], there are a few reports showing greater LS aBMD or B M C in girls compared with boys of the same chronological age [141, 142]. It is likely that these differences are due to the earlier maturation of girls. In the few studies that matched for Tanner stage or size, there were no differences in LS aBMD [34, 143]. The few studies that report femoral neck aBMD show either no differences in absolute aBMD prior to puberty [34, 144] and no sex difference in prepubertal bone gain [40], or greater FN aBMD in prepubertal males [38, 145]. A majority of these studies have examined a wide age range of older children and adolescents and included relatively small numbers of prepubertal boys and girls. There are two recent cross-sectional studies of a large number of prepubescent boys and girls -  28  one from Tasmania in Australia [145], and one from Detroit, Michigan [146]. Jones and Dwyer [145] measured TB, LS, and FN aBMD in 330 children (115 girls, 215 boys) aged 7-9 y. In contrast to previous cross-sectional reports, aBMD was 9.6% higher in boys at the FN after adjusting for height, lean mass, fat mass, and exposure to sunlight. After adjusting for the same variables, L S aBMD was 3.2% higher in girls compared with boys. In a large sample of 773 grade 3 and 4 boys and girls, there were no significant gender differences in body size or unadjusted TB BMC/D, but girls' LS B M C values were slightly higher than boys' [146]. One report of a small number of prepubescent Asian and Caucasian children and adolescents living in the United States [147] showed higher LS aBMD in Tanner 1-2 Caucasian girls compared with Caucasian boys. Caucasian boys had higher femoral neck (+6%) and TB BMD (+4%) compared with the girls. Mean values for Asian boys were higher for LS (+2%), FN (+16%) and TB (+12%) aBMD compared with Asian girls (Table 2.7). However, the Asian sample size was small (13 males and 5 females). As well, statistical outcomes for gender comparisons were not reported so it is not clear if these differences are significant. The results of this study are reported in greater detail in section 2.7. Overall, there is a lack of studies that examine gender differences for bone mineral in prepubertal children. The two studies that included a large number of prepubertal children suggest gender differences may exist prior to puberty [145, 146].  29  Table 2.6  Studies of gender differences in areal bone mineral density (aBMD) measured by dualenerqy X-ray absorptiometry in children and adolescents Study Design Sample Age Sites Sunwnaty of Findings size Range -greatest increases: 11-13y for girls and 13-17y for 7-20y LS I 65 Kroger P F boys 37 F 1993 iy FN 28 M -boys had greater increases in F N B M C and FN width at all ages -no differences at the spine -no sig. change in B M D from T1-T2 1-15y LS 135 II Glastre -girls BMD > at 12y otherwise no sex differences 65 F 1990 70 M Southard 1991  I  Ogle 1995  II  Nelson 1997  II iy  Maynard 1998  i 1-7y  Boot 1997  I  Jones 1998  I  Lu 1994  r* & n  218 134F 84 M 265 128 F 137 M 773 @ time 1 561 @ iy 465* F 73 M75 500 295 F 205 M 330 115 F 215 M 266 130 F 136 M 778 433 F 345 M  1-19y  LS  4-26 y  TB Lean  Grades 3 & 4  TB LS Fat Lean TB LS  8-18 y  4-20 y  LS TB  7-9y  TB FN LS TB FN LS TB LS FN Troch LS FN FS  4-27y  2-20y  Zanchetta 1995  II  Bonjour 1991  n  207 98 F 109 M  9-18y  Theintz 1992  i 1y  198 98 F 100 M  9-19y  -no gender differences adjusting for Tanner stage - B M C / D increased with each Tanner stage group -no differences prepubertal (before age 10y) in B M C or lean - B M C increased until 17y in boys and 16y in girls -no sig. gender diffs in TB B M C / D -girls > LS B M C and fat -boys > lean mass -same trends for baseline and @ 1y -no diffs i n T B B M D before 15y, > in boys at 15-18y -LS BMD > in girls from 12-14y, > in boys 17-18y -LS B M C > in girls 11-13y, > in boys 16-18y -girls > LS BMD and BMAD at all ages -no diffs in TB BMD -males > F N B M D ; 9.6% greater after adjusting for size, body comp, and sunlight exposure -girls > LS BMD (+3.2%) after adjusting -males > TB B M D a t 1 7 y -peak LS BMD not different -TB B M C > in males at 17y -FN BMD > males at 16y -LS BMD > males at 17 y, > females at 13y -n.s. diffs at younger ages -no sex differences in 9-1 Oy olds -no LS BMD diffs by Tanner stage, LS B M C > males at T4 & T5 -FN BMD > males at T4 & T5 -no sex differences in gain prior to puberty -females greatest gains at 11-14y -males greatest gains at 13-17y  LS FN FS  : T = Tanner stage I - prospective design (all 12 months except Maynard et al. 1998 - 1-7 years) II - cross-sectional design *465 observations from 148 children measured 1-7 times **53 children had more than 1 mea  30  2.7  Ethnicity Rates of hip and spine fracture vary dramatically by region and within ethnic groups living in the  same country. As well, there is a trend for increased fracture rates in areas as they become industrialized. An example is the rapid increase in hip fracture rates in areas of Asia, such as Hong Kong, in the past 20 years. In 1960, fracture incidence in Hong Kong was lower than in Caucasian populations in North America or Europe [4, 148]. Further, the female:male rate of hip fracture was 1:1, while it was (and is) closer to 2 or 3:1 in Caucasians living in North America [4]. Since 1980, there has been a dramatic shift in fracture rates in Hong Kong, which are now closer to North American Caucasians. Accepting that vertebral fracture is defined differently in different populations, epidemiological data show higher vertebral fractures among Asian populations (from Hong Kong, Taiwan and China primarily) compared with Caucasians in North America [4, 148]. In contrast, hip fractures occur more commonly in Caucasian women living in North America or Europe. The incidence of fracture has doubled in Hong Kong in the last 20 years [148] and interestingly, the incidence of hip fracture in men in China is higher than that in women (men, 80.0/100,000 - women 67.2/100,000) [148, 149].  In North America and Europe, fracture rates in  older people at any site are higher in women than in men. In addition to differences in the site of fracture, cultural differences in lifestyle-related risk factors exist. Some of the factors that may explain differences in fracture rates between Asian and Caucasian populations are discussed in sections 2.7.1-2.7.3 including: hip axis length, aBMD, lifestyle and genetic factors. It is important to note that the term 'Asian' is used loosely in the literature and populations are not adequately defined in most studies. This limits the interpretation of results.  2.7.1. Hip axis length Hip axis length (HAL) is related to femoral bone strength [150]. A longer HAL significantly increases the risk of hip fracture, independent of bone density and body size [151, 152]. Hip axis length is shorter in Asian than Caucasian adults, which is consistent with the lower hip fracture rates reported in Asian populations [151]. Ethnic differences in HAL and evidence from twin studies suggest a strong genetic contribution to bone geometry [153]. However, Wang et al. [154] report a similar HAL between Asian and Caucasian children and adolescents after adjusting for  31  height. These researchers suggest that better nutrition among American-Asian vs. native Asian children could contribute to the lack of consistent differences in HAL [154].  6  2.7.2 Bone mineral  content/density  Morphological differences between individuals or populations influence BMC or aBMD outcomes (as discussed in section 2.2). Asians born in the United States are generally taller and heavier than their counterparts born in Japan or China [155]. North American Caucasians also tend to be taller, heavier, and have greater fat mass than American born Asians [147]. Size differences are reflected in traditional densitometric measures of BMC and aBMD. Studies in adult cohorts report 10-15% lower BMC or aBMD at the hip and spine in Asian compared with Caucasian reference populations [156-159]. These studies did not account for size differences. In other studies size has been controlled for in a number of ways; within multiple regression [160-162], using estimates of bone mineral apparent density (BMAD) [147], BMAD adjusted for age and BMI [163], BMD/height, or the square root of the bone's cross-sectional area [164]. In most cases, ethnic differences in aBMD could be accounted for by size. However, in a large (n=1529) multi-center trial (Hawaii, UK, Portland, and Denmark) even after accounting for height, lean and fat mass, postmenopausal Asian women had 4.4%, 2.2%, and 1.7% lower aBMD at vertebral, arm, and leg sites respectively, than their Caucasian counterparts [160]. On the other hand, femoral neck BMAD was higher in adolescent [147] and postmenopausal [163] Asian than Caucasian women. These results are consistent with the lower hip and higher vertebral fracture rates reported in Asian women. An overall limitation of studies that compare different ethnic groups is that most studies have not controlled for size. Only one study has compared young Asians and Caucasians living in North America (California) [147]. Of the 99 Asian children (aged 9-26y) approximately half were American-born, the other half had moved to the United States (U.S.) from China, Korea, and Vietnam and had been living in the U.S. for an average of 10 years. In this cohort there were 18 Asian (5 girls, 13 boys) and 34 Caucasian (14 girls, 20 boys) pre-/early-pubertal (Tanner stage 1 or 2) children (Table 2.7). Areal bone mineral density was not different between Asian and Caucasian pre-/early-pubertal  We were unable lo measure hip axis length in our studies. The Hologic software on the QDR-4500 does not have the option of measuring HAL. Other researchers have measured HAL off of the scans directly/manually, DXA pencil-beam technology was used. With fan beam technology (as is used on the 4500), the image is too distorted to accurately measure HAL directly from the scan. 6  32  boys, however, the Asian boys tended to be older at each Tanner stage. Prepubertal Caucasian girls averaged 20% higher aBMD at the femoral neck and whole body than prepubescent Asian girls, but there were no differences in BMAD [147]. The authors conclude that differences in aBMD can be explained by the greater height in the Caucasian groups. However, the validity of BMAD equations for size adjustments is questionable, particularly when comparing individuals of differing ethnicites (as discussed in section 2.2).  Descriptive data for Tanner stage 1 or 2 Asian and Caucasian boys and girls living in California. Values are Mean (SD) Girls Boys Asian Caucasian Asian Caucasian 5 14 13 20 N 11.1 (1.3) 10.1 (0.5) 10.5 (1.2) 10.8 (1.6) Age (years) 36.8 (6.8) 40.6 (8.5) 36.7 (8.1)" 28.9 (2.0) Weight (kg) 141 (10) 146 (8)* 134 (4) 144 (6)** Height (cm) 27.7 (4.6) 29.8 (4.8) 26.7 (4.5)** Lean mass (kg) 21.9 (1.8) 7.4 (3.4) 9.0 (3.8) 8.7 (3.7)** Fat mass (kg) 5.6 (1.3) 22 (5) 20 (6) 24 (4) Body fat (%) 20 (5) 6.9 (5.9) 11.7 (9.2) 1.5 (1.5) 5.5 (3.5)** Weight-bearing activity (h/wk) Calcium 400 (60) 760 (280)** 460 (180) 610 (160)** (mg/1000 kcal) LS aBMD 0.674 (0.048) 0.694 (0.069) 0.686 (0.087) 0.679 (0.059) 0.727 (0.077) 0.718 (0.071) 0.681 (0.060)* 0.600 (0.081) FN aBMD 0.816 (0.059) 0.836 (0.063) 0.804 (0.043) 0.772 (0.040) TB aBMD Adapted from Bhudhikanok et al. 1996 [147]. *p < 0.10, "p < 0.05  Table 2.7  2.7.3  Lifestyle  factors  Physical activity (section 2.11) and calcium intake (section 2.9) during childhood are two lifestyle factors associated with aBMD that may differ between Asians and Caucasians. It is likely that lifestyle changes will occur when individuals migrate from Asia to Canada although the variability is likely to diminish as the period of residency within Canada increases. This implies that the physical activity and nutrition patterns of Asian and Caucasian children (who share a common school environment) will be more similar compared with activity and nutrition patterns of their respective parents. There are, however, no studies that compare the influence of lifestyle or genetic factors on bone mass in Asian and Caucasian family groups. It appears that, in general, Asian adolescents living in the U.S. and Canada have higher calcium intakes than children in Hong Kong, but still lower than Caucasian children. Asian 33  (primarily of Chinese ancestry) high school students living in British Columbia reported lower calcium than their Caucasian peers (738 ± 32 vs. 1,127± 41 mg/day) [165]. Hoetal. [166, 167] suggest the calcium threshold in adult Asian women is around 600 mg/d, but optimal calcium intakes in Asians living in North America may be higher depending, in part, on body size or other factors. In a cross-sectional study, Boot [39] compared a multiethnic group of children including Caucasian and Asian boys and girls aged 4-20 years. Their findings were consistent with the other reports that found Asians as a group consumed significantly lower calcium than Caucasians (759 vs. 1180 mg/day). Caucasians living in California tended to consume a greater proportion of calcium in their diets than Asians in all age groups. The primary ethnic differences in the entire cohort were height, weight, exercise, calcium and vitamin D intake [147]. Few studies have compared physical activity levels in Asian and Caucasian children. In one study, Asian girls reported less time of general physical activity (4.9 vs. 7.5 h/wk), and had significantly lower total body aBMD than Caucasian girls [39]. In the Californian study of Asian and Caucasian children, the pre- and early-pubertal Caucasian girls engaged in more weight bearing physical activity [147]. Clearly there is a need for more descriptive data on this important lifestyle variable that influences bone health, and other important health issues (i.e. cardiovascular health).  2.8 Body Weight and Lean Mass Body weight is a well-established predictor of bone mass and areal density. Cross-sectional aBMD is strongly related with weight, and weight loss and gain are associated with loss and gain of aBMD, respectively [168-170]. DXA total body scans provide values for total body bone (g), lean (g), and fat mass (g). These tissues can be measured with a high degree of precision [171]. Thus, it is possible to elucidate the contribution of specific tissues to bone mineral. In adults, there is some controversy as to which component of weight best determines bone mass or density. Some studies showed effects of fat mass on bone, independent of lean mass [172, 173]. Others reported higher associations between lean mass and BMC/D [174, 175]. In growing children and adolescents, however, lean mass was a stronger predictor of absolute BMC/aBMD, and change in lean mass was strongly correlated with change in BMC/aBMD [111, 176, 177]. The relationship between lean mass and BMD may reflect muscle mass inducing tensile 34  strain during mechanical loading. Alternatively, the lean mass-bone association may be genetically determined. Common genes controlling overall growth may influence both lean mass and bone mass [106]. There is some evidence for a common genetic link between VDR genotype and IGF-1 (or other growth factors) which influence both bone and lean mass [178]. However data are preliminary and as yet unsubstantiated. The strong and consistent relationships of change in lean mass with change in aBMD in studies of children and adolescents, indicates that change in lean mass should be controlled for in longitudinal observational or intervention studies of bone mineral.  2.8.1 Gender differences  in body  composition  Fat mass increases gradually throughout childhood with a dissociation between boys and girls at about age 8 or 9 [179]. This dissociation increases dramatically at adolescence when girls average approximately twice the fat mass of boys [179]. Although lean mass has not been studied extensively in children, it is generally accepted that differences between boys and girls become apparent during puberty. These data are from studies which utilized skinfolds or underwater weighing to estimate fat and lean mass. Recent DXA studies challenge existing theories. Two studies with large samples of prepubertal children report greater fat mass in girls (+1.5kg), and greater lean mass in boys (+1.8kg), by ages7-9y (Grades 3 and 4) [145, 180]. Gender differences in unadjusted values of lean and fat mass have also been shown in children aged 3-8y [181]. Thus, utilizing DXA technology, it appears that significant gender differences in body composition are evident very early in life.  2.9 Calcium As a primary component of bone mineral, it is clear that calcium is necessary for adequate bone mineral accrual. What is unclear, is the amount necessary to maximize peak bone mass [8]. A recent increase in the recommended calcium intake (now called 'Adequate Intake') for children and adolescents reflects the apparently high need for calcium in the growing years [182]. Although some data support the argument that adequate calcium intakes during the growing years benefit bone mass [183-185], there is no direct evidence linking high calcium intakes to higher peak bone mass and lower fracture risk. 35  Evidence from retrospective and intervention studies generally support a positive effect of increased calcium intake'during childhood and adolescence. Retrospective, observation, and experimental studies of the mineral accrual-calcium relationship have been recently reviewed [8]. Retrospective studies in adults showed a relationship between milk consumption during childhood and adult aBMD [185-188]. Women who reported consuming a greater amount of milk during childhood or adolescence had higher aBMD as adults than those with low intakes early in life [186188]. In contrast, current (adult) calcium/milk consumption was not related to aBMD in crosssectional studies [187]. Nieves et al. estimated that increasing dietary calcium intake in teenagers from 800 mg/d to 1200 mg/d was associated with a 6% increase in proximal femur bone density [187]. Several randomized, placebo-controlled supplementation trials, showed increased BMC/aBMD with increased milk intake or calcium supplementation during childhood and adolescence [183, 184, 189-192]. Gadogen et al. [189] examined the effect of increased milk consumption in 82 (44 treatment and 38 controls) adolescent girls. Girls increasing milk intake had a 1.1% greater increase in BMC and 2.9% greater increase in aBMD over 18 months compared with controls [189]. In a younger group of girls (mean age 7 years), increased calcium intake (850 mg) over 12 months was associated with a 1.9% greater gain in aBMD of the femoral neck and 0.3% greater spinal aBMD [193]. Overall, supplemented children averaged a 2.5% greater increase in aBMD over 1-2 years compared with unsupplemented children [183, 184, 190, 191]. However, follow-up studies showed that the benefit decreased or disappeared after cessation of supplementation [194, 195]. Longitudinal studies are necessary to determine whether the benefit to bone would be maintained if calcium supplementation continued through adolescence.  2.9.11nteraction  of calcium and exercise  Asian children in North America have slightly lower calcium intakes than Caucasian children [147] and, when total energy intake is accounted for, girls tend to have lower calcium intakes than boys during adolescence [147]. As calcium and exercise both play a critical role in formation and mineralization of bone matrix, an interactive effect between the two has been postulated [196]. If a certain level of calcium is required for exercise to have a beneficial effect on the skeleton, then those children, or groups of children (i.e. race or gender) who have 'inadequate' 36  calcium intakes would not benefit from an intervention program. However, definitive studies needed to answer this question have not yet been carried out, but some evidence is available from animal studies and from review of adult data. Lanyon et al. suggested a possible interaction between calcium and exercise based on results from studies in which they monitored the bone response to loading in calcium-replete and calciumdeficient turkeys [197]. Turkeys in a calcium-deficient state lost bone despite sufficient loading to form bone in calcium-replete turkeys. However, the loss of bone was greater in those turkeys who were both calcium deficient and immobilized [197]. As well, high calcium intakes do not prevent bone loss associated with unloading, as in space flight [198]. These studies imply that only when 'adequate' calcium intakes and mechanical loading are present, will the bone response be positive (Table 2.8).  Loading (+)  Loading (-)  Calcium (+)  + bone  - bone  Caclcium (-)  - bone  - bone  +/- indicates presence (+) or absence (-) of 'normal'/'adequate' calcium or loading. A review of exercise/calcium trials in adult women supported the theoretical interaction between calcium and exercise [199]. Across the 17 studies reviewed, exercise resulted in increased  aBMD  only if calcium intake was above a 'threshold' level (>1000 mg/d) [199]. All but one of the studies reviewed was of peri- or postmenopausal women, yet estrogen/progesterone supplementation was not controlled for. Exercise was inadequately defined with either 'yes' or 'no', therefore the type of activity was not reported. The lowest average calcium intake in the studies reviewed [199], including calcium supplementation, was760mg/d, which is near the recommended level for this age group (60y). These factors limit conclusions from this review. Evidence for a calcium-exercise interaction effect on bone mass in children is also lacking. A one-year prospective study of boys and girls, 8-16 years, reported an interaction between calcium and "weight-bearing" physical activity on change in ultradistal (primarily cancellous bone)  aBMD  [200]. The researchers reported that "high" calcium intakes combined with "high" levels of exercise were positively related to change in aBMD. Average calcium intake was > 900mg/d across all age groups, however "high" was not defined for either calcium intake or physical activity level. 37  Furthermore, "weight-bearing exercise" was not detined, and the mechanism by which there would be a beneficial effect at the ultradistal radius is not clear. Slemenda et al. have also reported a positive effect of activity on change in radial aBMD and in contrast to Gunnes and Lehmann [200] they did not report an interaction between calcium and activity [44], However, as the calcium intakes were near or above the perceived threshold amount [44], it was unlikely that an interaction (if one existed) would have been detected. The absence of adequately designed calcium and physical activity intervention studies creates a noteable gap in the literature. An appropriate study would require a large number of subjects and thus be costly. Methodology is hampered by the definition of "adequate" calcium intake and physical activity. Calcium "adequacy" may vary between ethnic and age groups. The question remains as to whether calcium and exercise interact or if physiology adapts with increased retention to meet increasing demands for calcium during growth.  2.10 Vitamin D and Seasonal Variation  The active metabolite of vitamin D, 1,25-dihydroxy D3 (1,25(OH)2D3) enhances calcium absorption in the gastrointestinal tract, helps to regulate calcium handling in the kidney, and is an important regulator of parathyroid hormone (PTH), and thus bone resorption [201]. Any change in vitamin D formation, metabolism, or action profoundly influences bone metabolism. The primary source of vitamin D is sunlight. Ultraviolet B (UVB) rays stimulate the conversion of 7dehydrocholesterol to pre-vitamin D [201 ]. In the winter months at Northern latitudes, UVB rays are not strong enough to synthesize vitamin D [202, 203]. Winter sunlight exposure is associated with aBMD at the hip in prepubertal girls, but not boys [145]. In Maine (45 degrees latitude), vitamin D synthesis in the skin does not occur during late fall and winter [202, 203]. Vancouver is somewhat more northern (49 degrees) in latitude than Maine. Thus, it might be expected that during the winter months, vitamin D synthesis would not frequently occur. Not surprisingly, bone turnover slows during the winter months and longitudinal growth rate of bone is greatest in the summer. Summer velocities account for between 60-67% of total yearly growth [204]. Adequate intakes for vitamin D for children and adolescents (1-18y) is 200-400 IU [205].  38  2.11 Exercise Studies in Children and Adolescents  Evidence for a beneficial effect of physical activity on bone comes from epidemiological, laboratory, and exercise studies. Epidemiological studies of large populations show that women and men who are most physically active in adulthood have either higher aBMD or a lower risk of fracture [206-208]. These data are supported by animal studies that show increased bone density and/or changes in bone architecture in response to site-specific loads. It is also clear from animal studies that growing bone responds to loading more favorably than mature bone [87]. While the epidemiological and biological evidence clearly shows that physical activity is beneficial to bone, exercise studies are somewhat equivocal. The extent to which the bone response to exercise is affected by the genetic, nutritional and hormonal environment is unknown. In addition, the hormonal milieu in girls and boys is rapidly changing as they approach puberty, which creates additional challenges for researchers. Researchers have examined either 'general physical activity' or 'targeted/sport-specific' loading as they relate to BMC/aBMD in children and adolescents. In these age groups, general physical activity is typically assessed by questionnaire and reported as a total score of all daily activities. A number of physical activity questionnaires are administered throughout the year to account for seasonal variability and to adequately represent the potentially fluctuating physical activity levels of the child. Activities that put strain on (a) specific bone(s) are considered examples of targeted/sport specific loading and are generally examined by monitoring bone mineral of children in specific sports such as gymnastics or running. In the next section I review the primarily retrospective evidence for an effect of both general physical activity and sport-specific mechanical loading on bone mass and density in children and adolescents. The two currently published intervention studies in pre-pubescent children are also reviewed.  2.11.1 General  physical  activity  The association between aBMD and general physical activity in childhood has been examined using several study designs including: comparisons of dominant and non-dominant limbs; relating current physical activity to aBMD at the spine and femur; and by retrospective analysis of childhood physical activity as it relates to adult aBMD. Two recent reviews of the relevant literature suggest a generally beneficial effect of physical activity during childhood and adolescence on bone mineral 39  content/density [7, 8]. Normally active children had 3-6% greater bone mineral content or areal density in their dominant versus non-dominant limbs [9, 209]; children who were more physically active had higher hip and/or spine aBMD in cross-sectional studies [34, 210, 211]; and adults who were the most active during childhood had higher aBMD compared with their less active counterparts in most retrospective reports [212-215]. A few studies report no effect of childhood physical activity on aBMD. Boot et al [39] found that in boys, but not in girls (4-20y) self-reported physical activity was related to lumbar spine aBMD. One prospective study assessed aBMD change over 3 years in pre-, peri- and post-pubescent twin pairs. Self-reported physical activity was related to femoral neck aBMD in pre- (mean age 7.4 years) but not peri-or post-pubescent children [44]. Only one study of physical activity in North America has included children and adolescents of Asian descent. There was no relationship between daily physical activity and aBMD at any site in Asian and Caucasian males and females aged 9-26 years [147]. In 179 Chinese adolescents (12-13 years) living in Hong Kong, physical activity was related to LS aBMD at baseline, but not to gain in distal forearm BMC or LS aBMD over three years [216]. Pubertal stage was the single most important predictor of BMC/aBMD change at the LS and distal forearm [217]. The authors suggested that interventions should be implemented prior to puberty as they would have an insignificant effect compared to the process of growth at puberty [217]. The authors did not examine predictors of BMC/aBMD gain after pubertal stage was controlled for, nor did they report effects of physical activity at the proximal femur. The most convincing evidence for an effect of general physical activity on bone mineral accrual comes from the Saskatchewan Bone Mineral Accrual Study [42]. Physical activity was assessed on average 2-3 times per year by questionnaire over 6 years in boys and girls who ranged in age from 8-14 at baseline. Scores were averaged from all questionnaires up to the age of peak height velocity (PHV) and children divided into either low, average, or high, levels of physical activity. A unique aspect of this recently published study was the control of maturational differences by aligning children at a common maturational landmark, PHV. Researchers looked at the association between physical activity and three indicators of bone mineral accrual, and the difference between the 3 exercise groups, for the total body (TB), lumbar spine (LS), and femoral neck (FN); 1) peak BMC velocity (g/y), 2) mineral accrual during the 2 years around peak BMC velocity, and 3) absolute BMC values at 1 year post-peak (the closest to adult values). In the entire sample, 40  physical activity was correlated with indicators of BMC accrual at all sites (r = 0.39-0.47). Comparisons between the three exercise groups also showed a consistent association of physical activity on BMC accrual (Table 2.9). Within each sex, those children who were the most physically active had significantly (p < 0.01) greater gains at peak BMC velocity and in the two years around peak compared with the least active children. After adjusting for height and weight, accrual rates across all sites were on average 16% and 25% higher during the two years around peak BMC velocity in the most active boys and girls respectively compared to the children in the lowest quartile for physical activity [42]. Absolute values, after controlling for height and weight, were 9 and 16% greater for TB BMC, and 7 and 11 % greater for FN BMC in the most active boys and girls respectively at 1-year post peak [42]. These values are somewhat lower than the difference for unadjusted values (Table 2.9), but still highly significant. This study presents strong evidence for an association of every day physical activity on bone mineral accrual, however it is not yet known if greater accrual rates translate into higher absolute bone mass in adulthood. Longitudinal studies which follow children into adulthood are required to determine if benefits persist.  Table 2.9 Unadjusted values in peak BMC accrual velocity (Peak BMCV), 2 year accrual (PBMCV+1y), and absolute BMC at 1y post peak BMCV for inactive and active boys and girls. Inactive n=15  Boys Active n=15  Inactive n=13  %*  Girls Active n=13  %*  Total body 367 467 Peak BMCV 21% 280 367 24% 2y accrual 640 816 22% 503 618 19% BMC 1 y post 2104 2511 16% 1571 2003 22% Femoral neck Peak BMCV .73 .89 18% .50 .73 31% 2y accrual 1.2 1.4 14% .86 30% 1.12 BMC 1 y post 4.49 5.34 16% 3.49 23% 4.29 Lumbar spine Peak BMCV 11.1 13.2 16% 8.7 12.3 41% 2y accrual 19.1 21.9 13% 15.5 20.9 35% BMC 1 y post 52.4 57.4 9% 40.5 57.4 29% % = percent difference between the groups for unadjusted values [(active-inactive)/active]100 values were significantly different between inactive and active children within each sex (p < 0.01).  41  2.11.2 Studies of targeted/sport-specific  loading in children and  adolescents  Cross-sectional studies of young athletes assessed during childhood also suggest that activity is beneficial to bone. Bone mineral density at loaded sites is consistently higher in athletes than normally active controls. Consistent with biological theory [21,81], athletes in sports that induce the highest and most unusual loads also had the highest aBMD at loaded sites [218, 219]. Overall, in cross-sectional reports, early or late pubescent gymnasts had a 7-20% greater aBMD than agematched controls [218-220]. Prepubertal gymnasts also had 12-16% higher aBMD than controls and a substantially greater increase in bone density than controls over one year [220]. Gymnasts had 8-20% higher femoral neck and spine aBMD than swimmers or controls [218-220], and young figure skaters had higher aBMD than controls at lower body, but not upper body sites [221]. Swimmers and cyclists tend to have aBMD similar to controls and lower than athletes in weightbearing sports, despite high tensile loads from muscular activity [218]. In a recent study of prepubescent female artistic gymnasts, 1-year increase in aBMD was greater at a number of sites in gymnasts compared to normally active controls [222]. The difference was greater at the trochanter (+4.8%), than the lumbar spine (+1.0%), total body (+2.2%) or femoral neck (+2.2%). However, the sample size was small (n = 9/group), and none of these differences was statistically significant [222]. In one of the few studies of bone in active prepubescent boys, Daly et al. [223] found an increase in ultrasound attenuation of the calcaneus over 18-months in prepubescent male gymnasts (n = 31). There was no change in attenuation in 50 age-matched controls over the same time period [223]. Assessment of broadband ultrasound attenuation (BUA) is thought to measure aspects of bone quality, such as microarchetecture, bone elasticity, and/or bone density [224], but there remains debate as to the physiology underlying this technology. Athlete studies are limited by the problem of differentiating between the potential genetic predisposition of elite sports participants to high bone mineral density and the influence of the loads induced by the sport itself. Prospective studies of physical activity and mechanical loading are summarized (Table 2.10).  42  Table 2.10 Prospective or longitudinal studies of physical activity or mechanical loading and bone in children and adolescents (<18 yrs of age). Outcome Physical activity type, measureMean Author, year study duration age or (n in range exercise group) (y) elite gymnastics training, 20-30 hrs 30-85% greater increase in 10 Bass et al. aBMD at total body, spine and per week - 12 months 1998 [220] legs (45) weightbearing sporting activity normal sporting activities Slemenda et 6-14 correlated with increased monitored by questionnaire - 12 al. 1994 [44] proximal femur aBMD months (32) weightbearing physical activity weightbearing activity measured Gunnes et al. 8-17 had the greatest effect on over 12 months 1996 [200] vBMD (forearm trabecular) in (231) children below 11 yrs 3 groups: (I) little or no activity; (II) No relationship between PA Kroger et al. 7-20 3 hr/week; (III) regular athletes 1992 [34] (65) and aBMD at any site 5hr/week - 12 months 1.0-4.8% greater (but nonCompetitive artistic gymnastics 10 Nickolssignificant) increase in aBMD 1,6hrs/wk- 12 months Richardson et for gymnasts vs. controls al. 1999 [222] (9) General physical activity measured 26% greater increase in BMC Peak Bailey et al. over 2-yrs of greatest accrual over 7-years by questionnaire 1999 (26) [42] BMC in most active vs. least active velocity children Gymnastics training (males) 10-29 12.8% increase in BUA* at the 10.1 Daly et al. hr/wk (avg. 16.4 hr/wk) - 18 mos. calcaneus in gymnasts over 1999 [223] 18 mos. - no change in (31) controls. *BUA = broadband ultrasound attenuation  Retrospective studies: A few retrospective studies have shown gymnastics [220,225,226] and ballet [227] undertaken in childhood were related to adult bone status. In 99 retired ballet dancers (mean age 51 years), weekly hours of ballet performed between the ages of 10-12 years was related to femoral neck aBMD, while current physical activity and years of full-time ballet were not [227]. Adult aBMD at weight-bearing sites was also higher in retired gymnasts, who began training before menarche, than normally active controls [220, 225, 226]. Bass et al. found retired gymnasts to have 6-16% greater aBMD than controls regardless of the number of years since retirement. Kontulainen et al. described a similar phenomenon in a study of detraining in tennis players [228]. They measured side-to-side BMC differences in 13 tennis players and 13 controls in 1992 and 1996. At the first measurement, participants were 26 y old. Tennis players had been playing at the national level since age 11, but had retired an average of 2.3 y before follow-up measurements in 1996. Side-side BMC differences at the humeral shaft were 25% in 1992 and 26% in 1996, despite  43  a reduction in training (7 hrs/wkto 3 hrs/wk). Controls' side-side differences were less than 5% and did not change over time [228]. This study provides strong evidence of a benefit of childhood activity on adult bone mineral. However, it is important to note that while tennis players had retired, they still maintained an average of 3 hrs. of training per week. It is possible that this was enough to maintain their advantage in BMC. Optimal Age. A group of researchers from the UKK Bone Research Institute in Tampere, Finland have conducted a series of studies utilizing a unilateral training or loading model, with the contralateral limb serving as a control. They reported side-side bone mineral differences in 105 elite women squash and tennis players to examine if there was an optimal age during which exercise can effect bone. Bone mineral content (BMC) was 13% greater in the loaded (preferred) versus the non-loaded arm. When athletes were split according to when training began, differences for loaded versus non-loaded arm were 17-24% in those women who began training before menarche, and 8-14% different in those who began training after menarche [9]. In a followup study of 7-17 year old tennis players the same research group identified side-side differences only when girls reached the adolescent growth spurt (Tanner stage III) [229]. These data suggest a specific exercise effect which is dependent on and closely linked to maturational status.  2.11.3 Exercise  intervention  Two exercise intervention studies in pre- or early-pubescent children, one in pre-menarcheal girls [177] and one in prepubertal boys [230], have measured bone as a primary outcome. Morris et al. [177] monitored change in bone density over 10 months (January to October) in girls aged 910. Thirty-eight girls participated in a variety of activities for 30 minutes three times per week over 8 months (February to September), and 33 age-matched girls served as controls. The activities included "high-impact" aerobics, soccer, Australian football, step aerobics, bush dance, skipping, ball games, modern dance, and weight-training. When changes in height and weight were controlled for, girls in the intervention group had significantly greater increases in total body, lumbar spine, and proximal femur aBMD [177]. It is noteworthy that the 10% greater increase in femoral neck aBMD was not significant however after controlling for height (Table 2.11). This suggests that the groups may not have been matched for maturity [230], the most important determinant of aBMD change during the growing years. 44  A similar intervention program in prepubescent boys showed significant, but much smaller, effects [230]. The outcomes between studies were difficult to compare as in the Bradney et al. study, the measurement period ranged from 7-11 months and thus, changes in aBMD per month (rather than total change over the study period) were reported. Estimates for percentage difference over 8-months between the control and exercise groups are summarized in Table 2.11. Reasons for the differences between these studies are not clear as the interventions were similar, but could be due to a number of factors including 1) a greater maturational rate in the exercise vs. control girls in the Morris study [230], 2) differences in DXA measurement systems or analysis of DXA scans, 3) differences in stage of maturation between boys and girls, or 4) differences in implementation of the intervention. Thus, there is a need for an intervention study including both boys and girls controlling for stage of maturation and measured on the same densitometer to determine if response to intervention differs between boys and girls.  2.11 Difference in percent change in aBMD between exercise and control groups in two intervention studies in pre- or early-pubescent children. Morris et al. Bradney et al. Table  (girls)  (boys)  Total body  2.3%  1.0%  Lumbar spine  3.6%  2.3%  Proximal femur  3.2%  3.6%  Femoral neck  10.3%*  N/A  All values were significantly different between the groups (p < 0.05). *after controlling for height, this difference was not significant.  2.11.4 Ground reaction forces from various activities The optimal, or most osteogenic intervention program has yet to be defined in children. Animal studies indicate that the most osteogenic activities are high in magnitude and unusual in their distribution. Jumping or skipping activities fall into this category and are in line with Lanyon's error strain distribution hypothesis [81] and Frost's MES theory [84]. Strain magnitude from a number of activities has been estimated from measurement of ground reaction forces (GRF). Jumping and running induce ground reaction forces of 3-5 times body weight, which may equal up to 10 times body weight at the tissue level [231, 232]. Peak GRFs of 2.6-5.6 times body weight have been estimated for activities such as jumping, high impact and 45  step-aerobics in young adult groups [233, 234]. These studies support an important role of impact loading on the skeleton. Assuming that strain increases linearly with increasing load, activities that produce greater GRFs should have a larger osteogenic effect if they are unusual in their distribution.  2.11.5  Limitations  of exercise studies in children and  adolescents  A number of cross-sectional and retrospective studies show that both general physical activity and sports participation in childhood and adolescence are related to higher bone mineral density  (aBMD) during the years of growth.  The few prospective and intervention studies support these  findings. However, there are numerous limitations in existing studies including: 1) study design, 2) study duration, 3) maturational control, 4) interaction of major determinants, and 5) practicality/public health relevance.  Study design  While cross-sectional observation studies of athletes have been useful in demonstrating a potential influence of mechanical loading on aBMD, they can only be hypothesis generating, not hypothesis testing. Confounding factors such as genetic and nutritional influences are likely to be different in athletes than normally active individuals and may interact with exercise to positively affect  aBMD [196, 235, 236].  The studies that compare dominant and non-dominant limbs provide  a controlled model for examining the role of loading on bone. As both limbs share the same nutritional, genetic, and hormonal environment, differences in bone mineral can be attributed to loading. These studies [9, 237] and other retrospective studies show a beneficial effect of sportspecific mechanical loading on adult aBMD [227]. The two intervention studies in prepubertal children also indicate that exercise intervention increases bone mineral in prepubertal children [177, 230]. However, these studies are limited in their generalizability as only two schools were involved, and in one case [177] schools were not randomzied. Statistically, randomizing by school is considered a 'quasi-experimental' approach. A true experimental design means randomizing by student.  46  Study  duration.  A number of researchers have suggested that exercise started prior to- or in early- puberty is more beneficial to bone than exercise begun during or after puberty [9, 227, 229]. There may be a brief and unique period during the growing years when bone is most responsive to exercise [238, 229]. To determine if, and when, this opportunity exists, needs randomized intervention studies that follow girls and boys through puberty. Long-term follow-up studies are also needed to determine if benefits of childhood physical activity persist into adulthood. It is possible that aBMD in the controls will "catch-up" once intervention ceases or at some point over the growth process. This phenomena of "catch-up" occurred with cessation of calcium intervention in this age group [194, 195]. The long term (adult) benefits of childhood intervention have yet to be adequately studied.  Maturational  control  Observational and intervention studies that relate aBMD or change in aBMD to physical activity in children and adolescents may be confounded by pubertal status as children of the same chronological age can differ significantly in their maturational status. Also, children at the same maturity level at the start of an intervention trial may develop at different rates during the course of the study. Thus, growth-related changes in bone mineral may inappropriately be attributed to exercise. For example, in the one intervention study of premenarcheal girls, the greatest increase in BMD of exercisers compared with controls (10% at the femoral neck) was no longer significant when change in height was controlled for. This suggests a difference in pubertal status between the groups.  Interaction  of major  determinants  There are a number of determinants of bone mineral that may interact with the effect of mechanical loading on bone. Preliminary data suggest differences in calcium intake, gender, genetics, ethnicity, and hormones may interact with exercise response on bone. However, no studies have been adequately designed to test these interactions in children or adolescents.  47  Practicality/Public  health relevance  Programs need to be designed that are easily implemented within a community-based setting and in which a large proportion of the population can, and will, participate. In the two intervention studies in pre-pubescent boys and girls, children in the intervention group participated in 30 minutes of physical activity three times per week prior to school. Although dropout rates were low, the exercise schools self-selected, and thus are likely to have been more highly motivated to complete the intervention. If an exercise program could be implemented within existing physical education classes, it could easily reach a larger portion of the population [239].  2.11.6 Gaps in the literature  The existing gaps in the literature give rise to a number of important questions about the role of physical activity in bone mineral accrual [8] including: 1) Does increased activity during childhood lead to increased bone mineral accretion? 2) What is the 'optimal' osteogenic program? 3) Is there a "critical period" during childhood or adolescence when activity builds the most bone? 4) What other factors interact with exercise to influence bone? and 5) Are the benefits of childhood or adolescent exercise programs maintained into adulthood? The research necessary to conclusively answer these complex questions will require long term intervention and a large population of children across maturity ranges, and therefore may not be feasible. However, the studies in this thesis were designed to provide some insight into specific aspects of this complex puzzle.  48  Chapter 3  RATIONALE, OBJECTIVES & HYPOTHESES  49  The literature points towards the prepubertal years as a potentially optimal time for exercise intervention to have a lasting osteogenic effect. The primary objective of this study was to examine the effect of a school-based mechanical loading intervention in pre-/early-pubescent children. As relatively little is known about other determinants of  aBMD  in children, we also examined sex and  ethnic differences in aBMD, and familial resemblance of total body bone, lean and fat mass. A rationale for the studies (section 3.1), primary objectives (section 3.2), and hypotheses (section 3.3) are outlined for the three components of this thesis: 1) Cross-sectional determinants of aBMD  in prepubertal children, 2) Familial resemblance in total body bone, lean and fat mass, and  3) School-based, mechanical loading intervention.  3.1 Rationale  3.1.1 Cross-sectional  determinants of aBMD in prepubertal  children  General descriptive data of the determinants of aBMD in prepubertal children are scarce in the literature. As these years are deemed critical for bone mineral accrual, cross-sectional studies of determinants of bone in a large cohort of prepubertal children are necessary. It is also useful to compare ethnic and gender differences in aBMD and lifestyle factors related to aBMD as the occurrence of fracture differs among ethnic groups, and within genders of the same ethnic group. Gender differences in aBMD are apparent in adult groups and have been linked with differences in fracture prevalence. Much less is known about weather these differences exist in children and adolescents. Most studies have been limited to small numbers of prepubertal children and primarily to measurement of the lumbar spine. Thus, studies comparing aBMD, particularly at the proximal femur, and lifestyle factors related to  aBMD  in prepubertal boys and girls are  warranted. A study of Asian populations living in North America is needed for a number of reasons. First, Asians are at high risk for fracture and there are an increasing number of Asian residents in Canada. However, relatively few studies have examined lifestyle factors related to aBMD in Asians living in North America, especially in children. Further, lifestyle factors and morphologic parameters related to  aBMD  traditionally vary between Asian and Caucasians. It is not yet clear if differences 50  in adult aBMD between these groups can be attributed to differences in body or bone size. Nor is it known if differences are apparent early in life. Thus, studies of aBMD in prepubertal Asian and Caucasian children are important. Vancouver provides an opportune location in which to study these two ethnic groups.  3.1.2  Familial  resemblance  It is clear that genetic factors explain a large portion of the variance in peak bone mass [128], and likely that individual differences in response to exercise and/or calcium interventions are at least partially explained by genetic influences [236, 240, 241]. Despite the rapidly growing interest in the pre- and early-pubertal years as an optimal time for exercise or calcium intervention to influence bone mineral accretion, the contribution of genetic factors to pediatric bone mineral accrual remains unknown. Studies of familial resemblance in prepubertal children are needed to complement those in adult groups, and to explore the complex interactions between bone mass, heredity, and body composition at various stages of growth. Fracture risk appears to be determined equally, but at differential sites in sons and daughters, by paternal as well as maternal history of fracture [242]. However there are no studies of familial resemblance including prepubertal boys or fathers of prepubertal children. Further, only Caucasian family groups have been studied despite the high fracture risk in other ethnic groups [4, 148]. Studies of familial resemblance in Caucasian family groups and other ethnicities, including mothers, fathers, sons and daughters and which control for body composition factors are warranted to begin to explore these issues.  3.1.3 Mechanical  loading  intervention  As summarized in section 2.11.6, there are a number of important questions regarding the skeletal response to exercise intervention during childhood and adolescence. Overall, there is a need for randomized intervention trials in pre- and peri-pubescent children with adequate control for maturational status and growth. Most of the existing evidence of exercise benefits comes from retrospective or cross-sectional studies of highly trained athletes, but very few children are able to achieve this level of activity. Although the "ideal" exercise intervention has yet to be identified, for public health benefits to be realized a program should be designed to be easily implemented in the classroom [239]. As in other areas, there is also a need for exercise intervention studies involving 51  ethnic groups other than Caucasians, and especially Asians given the rising incidence in the prevalence of hip fractures in this group [243] and the increasing numbers of Asian residents in Canada.  3.2 Objectives  3.2.1 Cross-sectional  determinants  To examine ethnic and sex differences in absolute values of areal bone mineral density and to explore the contribution of calcium intake, physical activity, lean and fat mass to areal bone mineral density in pre- and early-pubescent children.  3.2.2 Familial  resemblance  To assess the strength of the relationship of total body BMC, lean and fat mass between mother-son, father-son, mother-daughter, and father-daughter pairs. To compare  these relationships  3.2.3 Mechanical  between  Asian and Caucasian  Family  Groups  loading  To determine the effect of an elementary school curriculum-based exercise intervention on change  in areal bone mineral density over 8 months after controlling for confounding variables  (initial aBMD, change in height, change in lean mass, sex, and ethnicity). To examine  sex and ethnic interactions  in the bone response  52  to  intervention  3.3 Hypotheses  3.3.1 Cross-sectional  >  determinants of aBMD  After controlling for size differences, there will be no sex or ethnic differences in absolute values of aBMD.  >  Dietary intake of calcium and physical activity will be less in Asian compared to Caucasian children living in a common geographical environment  >  Lean mass will explain a significant proportion of the variance in aBMD at all measured bone sites in prepubertal Asian and Caucasian boys and girls.  >  Physical activity and calcium intake will be secondary predictors of proximal femur, lumbar spine, and total body aBMD in prepubertal children.  3.3.2 Familial  >  resemblance  Resemblance in total body bone, lean and fat mass will be significant within Asian and Caucasian family groups.  > The strength of the relationship for TB BMC within family groups will decrease after adjusting for lean mass. >  There will be no significant difference in the strength of familial resemblance between Asian and Caucasian family groups.  >  Current and past physical activity and calcium intake will be lower in Asian compared to Caucasian families living in a common environment.  3.3.3 Mechanical  >  loading  intervention  A 10-20 minute.moderate impact loading intervention program incorporated into existing Physical Education classes will elicit increased proximal femur, femoral neck and trochanteric aBMD in the prepubescent skeleton.  >  A moderate loading intervention will have no significant effect on 8-month change in total body and lumbar spine aBMD.  >  Exercise intervention will produce a similar benefit to the growing skeleton regardless of ethnicity or gender. 53  Chapter 4  METHODOLOGY  An overview of the number of participants in each measurement period is presented in Figure 4.1. A total of 168 Asian and Caucasian children participated in the study at baseline. Baseline measurements were used to cross-section ally examine sex and ethnic differences in aBMD and lifestyle variables (objective 3.2.1). Of the children participating at baseline, 77 parents participated in the familial study (objective 3.2.2). A total of 144 children returned for follow-up measurement after the intervention period (objective 3.2.3). In addition to the primary studies, we addressed methodological issues in the analysis of pediatric DXA proximal femur scans. For this paper, we utilized scans from 40 of the Caucasian children who had 2 measurements, and scans from 10 children who participated in the University of Saskatchewan Bone Mineral Accrual Study. This original paper is included in its entirety in Appendix 1. Methodology, results and discussion of the intervention and family studies are included in chapters 4, 5, and 6. Three manuscripts resulting from these studies are also included in Appendix 1. Baseline (n = 168) Asian n = 58 (30 M. 28 F) Caucasian n = 110 (56 M, 54 F)  Objective 3.2.1 Sex and ethnic differences  /  Parents (n = 77) Mothers n = 49 (16 A. 33 C) Fathers n = 28 (10 A, 18 C)  Objective 3.2.2. Familial Resemblance  8-month followup (n = 144) Exercise group n = 63 (16 A, 47 C) Control group n = 81 (33 A, 48 C)  Objective 3.2.3. Intervention Study  4  Ancillary Study. Methodoloaical Issues  UBC Bone Study 80 Scans Irom 40 children (20F.20M)  Saskatchewan Growth Study (7-years) 70 scans from 10 children (5 girls, 5 boys)  Figure 4.1 Overall study design showing the number of participants in each sex and ethnic group for each of the measurement periods. M = male, F = female, A = Asian, C = Caucasian.  55  4.1 Intervention Study Prior to beginning the study, ethical approval was obtained from the Richmond School Board and UBC internal ethnical review board. Richmond was chosen for its high Asian population. Approximately 34% of the population in Richmond report either Mandarin or Cantonese as their first language. We had the support of a Richmond school board representative (Mr. Laurie Nordin) and he was invaluable for facilitating recruitment and generally promoting the study to principals and teachers. At the beginning of the 1997 school year, we presented the study to principals and teacher-development representatives from all elementary schools in Richmond. As well, letters were sent to all principals and teachers in the Richmond school district informing them of the study.  4.1.1 Participants Grade 4 or 3-4 combined classroom teachers who were interested in participating in the study contacted researchers who provided them with study details. A total of 18 teachers from 10 schools agreed to participate. We visited the classrooms in September to inform the children about the study and to answer questions from the teachers and children. At the time of the visit, consent forms (Appendix 2) were sent home with the children and collected by the teacher within 1 week. All forms were translated into Chinese for non-English speaking families and a Mandarin and Cantonese speaking translator was available to answer questions by phone. A total of 58 Asian and 110 Caucasian children returned signed consent forms at baseline. Repeat measurements were obtained on 144 children. Several of the children moved (n = 9) during the school year. Other reasons for not returning for final measurement included being absent from school on the day of measurement (n = 5) and failure to return the consent form (n = 7). Children were classified as Asian or Caucasian based on their parents' place of birth. Parents of Asian children were primarily from Hong Kong (63%), and also China, Taiwan, Japan, or Vietnam. The majority of Caucasian children and their parents were born in Canada (81%). Of the 58 Asian children in the study, 26% were born in Canada, while the others were born primarily in Hong Kong (41%). A majority (61%) of the Asian children had been living in Canada for less than 5 years. A breakdown of the Asian children's place of birth and duration living in Canada are presented (Tables 4.1 and 4.2). 56  Table 4.1 Number and percentage of Asian children born in areas listed. Birthplace N % of Asian children Canada  15  26%  Hong Kong  24  41%  Taiwan  10  17%  China  7  12%  Vietnam  1  2%  Japan  1  2%  Table 4.2 Duration living in Canada for Asian children n % of Asian Length of time in children  Canada <1 y  8  14%  1-3 y  16  28%  3-5 y  11  19%  5-7 y  3  5%  7-9 y  5  8%  Born in Canada  15  26%  4.1.2 Exercise  intervention  The 10 schools were stratified by student number per school and assigned to either large (n=4), medium (n=4), or small category (n=2). Schools within each tier were randomly assigned to either control or exercise groups. All classrooms at the intervention schools participating in the study completed the Healthy Bones intervention (described below), while teachers in control schools continued with their regular physical education program. The intervention program was designed to impose relatively high loads of unusual distribution on the skeleton. In a pilot study of a small number of children (n = 4) in the UBC biomechanics lab, various jumps included in the curriculum elicited ground reaction forces of 2-5 times body weight. The exercise intervention had two parts. 1) a curriculum to be implemented within the physical education classes which were held twice per week and 2) tuck jumps which were performed 3 times per week. Control schools continued with their regular physical education program.  57  Curriculum. The curriculum package consisted of: 1) games, dances, and other activities ranging from 10-30 minutes which included a minimum of 10 minutes of loading, 2) a comprehensive explanation of the exercises and the learning objectives, 3) a summary card for easy in-class instruction and 4) a wall chart depicting five basic skipping skills. The program was designed to be carried out by the classroom teacher as part of the usual PE class. Activities were consistent with the five "movement categories" mandated in the Instructional Resource Package (IRP) for physical education (dance, gymnastics, individual and dual activities, alternate environment activities, and games). A select sample of the activities used in the Healthy Bones curriculum are provided (Appendix 3). The program progressed based on level of ability and as children adapted they were able to spend more time jumping within each activity. The intervention began in the first week of November after baseline measurements and continued to the end of May with a two-week break at the end of December. New activities such as bench jumping and circuit training were added as program options at a training session in February.  Tuck Jumps. Children performed 10 tuck jumps 3 times per week to ensure a 'minimal' amount of loading. On 2 days jumps were performed as a warm-up to physical education, and on a third day in the classroom. Children were instructed to jump using both legs together and attempting to grab their knees, bringing them as close to the chest as possible. A 1 second rest between jumps was allowed to maintain quality of each jump.  4.1.3 Incentives and training  Teachers in the intervention schools received the curriculum, two in-service training sessions, musical tapes and supplies (jump ropes). The control schools received the curricula at the end of the study period. Small incentives such as skeleton stickers and pencils were given to the students. In addition, a printout of the children's skeleton from their DXA scan, along with a summary of the study results were sent home to the children and their parents at the end of the year (Appendix 4). To ensure appropriate implementation of the program, two in-service sessions were provided for teachers in the intervention schools (in October and February). These instructional workshops were directed by Lindsay Waddell who compiled the activity manual, and 58  Judy Notte, an elementary PE teacher who has a master's degree in curriculum studies. Teachers in intervention and control schools were also asked to log activities performed during PE class time. In addition, physical education classes were monitored 3 times during the year to ensure that appropriate and consistent implementation of the program was maintained. A measurement team, consisting of graduate students primarily, were trained ~1 week prior to the measurement start-date. Dr. Susan Barr trained the measurers to administer the food frequency questionnaire. Dr. McKay trained students in anthropometry, administration of the physical activity questionnaire, and vertical jump procedures. Although all measurers trained, measures of height and weight were taken by myself or Dr. McKay with the assistance of one other measurer to ensure consistency between measurement periods.  4.1.4  Measurements  Three groups of 4-6 children were shuttled between Richmond and the DiagnostiCare radiology clinic in the Fairmont Medical Building in Vancouver (~ 10 km) for measurement and supervised en route by a study staff person who also collected biographical data en route. Measurements were taken over a 3 week period in October, 1997 and again in June, 1998. The order in which schools were measured at baseline was randomized regardless of group status (exercise or control). This order was replicated at follow-up to standardize the duration between measurements. The pick-up and drop-off schedule was as follows:  Pickup from @ Clinic school Group 1 (n= 4-6) 9:00am 9:30-11:00 Group 2* (n = 4-6) 10:30 11:00-12:30 Group 3* (n = 4-5) 12:00 12:30-2:00 'Groups two and three ate their lunches at the clinic.  -Return to school 11:30 1:00 2:30  Children rotated through each of four stations at the clinic: 1. Anthropometry (height, weight, and calf girth) 2. Bone densitometry (total body, hip and spine) 3. Vertical jump 4. Questionnaires (physical, nutrition, and maturity) 59  In addition to questionnaires, other projects were available for the children to do once measurement was complete and all activities were supervised by a qualified study staff person.  Anthropometry. Body composition variables including fat mass and bone mineral free lean mass (lean mass) were estimated from DXA total-body scans [171, 244, 245]. Height (without shoes) was recorded to the nearest millimeter (mm) using a stadiometer. Stretch stature for standing height was measured by applying gentle upward traction on the base of the mastoid processes as per standard procedure [246]. Weight was measured using an electronic scale to the nearest 0.1 kg. Two measurements were taken unless values were more than 0.4 cm (for height) or 0.2 kg (for weight) apart, then a third measure was taken. The average of two values, or the median of three values was used as the final score.  Bone Measurement.  Bone densitometry scans were acquired and analyzed initially by the  same registered technologist. Total body, lumbar spine and proximal femur scans were acquired on a Hologic QDR-4500W bone densitometer (Hologic Inc., Waltham, MA). Proximal femur scans were analyzed for the total region and the femoral neck and trochanteric sub-regions (Figure 4.2). Participants wore plain, loose fitting shorts and a T-shirt. Shoes and all metal objects (jewelry, belts, glasses, etc.) were removed. The standardized positioning protocol for each site was used [247]. The procedure is painless and simply required that the child lay on the padded examination table for the duration of measurement, which includes scan time and positioning. In the array mode, approximate scan times for the regions of interest are: proximal femur 60 seconds, lumbar spine 60 seconds, and total body 6 minutes. The QDR 4500 has a mandatory quality control function and the spine and step (for body composition) phantoms were scanned daily. The coefficient of variation for the 4500 has been reported as less than 1 % at all sites [247, 248].  60  •^^I>^^^y  M'tW^  / \  Scan Analysis:  Figure 4.2 Regions of the proximal femur during scan analysis. FN femoral neck region; T = trochanteric region. The total proximal femur includes all shaded regions.  There are a number of methodological issues related to the analysis of DXA  scans in growing children. The influence of operator-dependent changes in proximal femur region and sub-regions of interest has not been previously addressed. Therefore, as an ancillary study to this project, we examined the effect of changing size, location and rotation of the proximal femur and femoral neck region of interest (ROI). The study results showed significant effects of small changes in the size and location of the ROIs. The original paper is presented in Appendix 1. As a result of that study, I re-analyzed proximal femur scans with a strict standardized protocol to ensure consistency between and within scans. To reduce the system error introduced by varying the size of the global and sub-regions of interest, the same global ROI was maintained for time 1 and time 2 within children. The femoral neck box size was standardized at the system default size of 1.5 cm (15 pixels) by 4.1 cm. The compare mode was used to ensure identical placement of the trochanteric line. Software version 8.20a was utilized for all analyses.  Vertical  Jump. A vertical jump test was performed during the first and final months of  intervention. Children followed the standard procedure for vertical jump testing [249]. A measurement of full reach was taken prior to jumping. A two-foot takeoff was performed from a standing position near a wall. Distance was measured from the highest reaching point with feet flat on the ground to the highest point touched during the jump. The highest of the two jumps is reported.  61  4.1.5 Questionnaires All questionnaires are attached (Appendix 5). Dietary Intake. A food frequency questionnaire (FFQ) was administered 3 times over the year (October, February, and June). During clinic visits (October and June), the FFQ was administered individually to children by trained graduate students. The questionnaire has been validated as a tool for assessing calcium intake in Asian and Caucasian high school students [250]. The FFQ was translated and sent home to parents in English or Chinese and parents' and childrens' questionnaires were compared. Estimated calcium intake (mg/d) from parents' questionnaire was correlated with childrens' (r = 0.58, p < 0.01). Thus, average calcium intake (mg/day) was calculated from children's FFQs (either 2 or 3 questionnaires). Although dietary records or food recalls are generally more accurate than FFQs for assessing total energy intake, the difficulties with administering and analyzing these records in young children and different ethnic groups made the FFQ the recommended (Dr. Susan Barr) option. In a recent review of reproducibility studies, Rockett & Colditz [251] concluded that FFQs are accurate assessments of dietary calcium in children and adolescents.  Activity. Physical activity questionnaires were also administered 3 times over the year. The activity record is a 7-day recall adapted from Slemenda et al. [211] and Sallis [252] (Appendix 5). The questionnaire has been validated against a number of physical activity measures including the Caltrac motion sensor [253]. Three activity scores were calculated to reflect the average activity level of the child: 1) An overall activity score was calculated by coding answers on a Likert scale from 1-5(1 = not active; 5 = always active) and totaling answers from questions 1 -8 and 11. Scores range from a minimum of 38 to a maximum of 190; 2) A 'load' score included the average times in the past week weightbearing activities were completed (from question 1) and; 3) A dichotomous (yes or no) variable indicated whether or not the child participated in extracurricular activities. As most children partook of primarily weight-bearing physical activity, there were no significant differences between the loading score and the overall score. Thus the overall activity score, averaged from 2-3 questionnaires, was utilized in all analyses. The extracurricular activity score was used as a descriptive variable. 62  Health History.  A health history questionnaire was sent  home  for parents to complete for their  children. It was designed as a screening tool to identify factors such as use of glucocorticoids and family history of osteoporosis or fracture, which may affect bone. All of the children were healthy and none were excluded based on the response to the questionnaire.  Maturity.  Pubertal development was evaluated by self-assessment of breast (girls) and pubic  hair (girls and boys), according to the method of Tanner [67]. The purpose of the rating procedure was explained to each child individually and children were allowed to privately select the picture that most accurately reflected their maturity level. The child returned their questionnaire in a sealed envelope. When there was a discrepancy between breast and pubic hair stage in girls, Tanner breast stage was used to represent the level of maturity.  4.2 Parents Letters were sent to the parents of children participating in the Healthy Bones Study inviting them to participate in the Family Study. Parents who expressed interest were contacted by phone, details of the study procedure were explained and an appointment was scheduled. Consent was obtained from interested parents when they first arrived at DiagnostiCare Radiology for measurement. Of the 168 Asian and Caucasian children enrolled in the intervention study, 88 parents (57 mothers and 29 fathers) agreed to participate in the study. Mothers were excluded if they had a history of ovariectomy or hysterectomy (n = 6), long term  Cortisol  use (n = 1), kidney disease (n = 1)  or scoliosis (n = 1). One father was excluded for treatment of hypothyroidism (n = 1). None of the women or men in the study had any other diseases, nor were they using medication known to influence bone metabolism.  Bone densitometry  and body composition.  Parents were scheduled for measurement in April,  1998. Lunch hour, evening and weekend appointments were made available. All bone densitometry procedures were completed as described for the children and scan acquisition and. analysis were performed by the same trained technician. Researchers checked scans following acquisition to identify errors in position or analysis. 63  Questionnaires. Parents completed the same calcium-specific FFQ as the children at the time of bone measurement. A milk history questionnaire developed for the NHANES III study [188] was also administered. The medical-history questionnaire utilized questions from Dequeker [254], McKay [117], and CaMOS (Canadian MuliCentre Osteoporosis Study, © CAMOS, 1995) questionnaires to assess demographic variables as well as past and present medical history and past and present physical activity. Questionnaires were traslated to Chinese and self-administered. 1 was available to clarify questions and translators were on-site to assist participants in interpretation of the questions if needed.  4.3 Statistical Methods  Data Checking. Scatterplots of initial values, final values and change for all variables were used for data checking. Outliers (values more than 1 SD above or below the next closest value) were checked to ensure that data were correctly entered and to determine if the value was reasonable for the individual. For example, one child weighed 1 SD more than any of the other children, but was also taller and more mature. One Caucasian boy reported a calcium intake of more than 2 SD above the other children, however, this was consistent across all 3 questionnaires and his parents questionnaire, so his data were retained. Dependent variables were checked for normal distribution by calculating skewness and kurtosis values. Values less than 2 were considered indicative of a normal population distribution. Regression assumptions of independence, linearity, normality and constant variance were checked by residual plots. All residual values were within a band of ±2 standard errors, and of a random pattern.  Cross-sectional Analysis: Descriptive statistics (mean±SD) were calculated for all variables. A 2 (male, female) by 2 (Asian, Caucasian) analysis of variance (ANOVA) was used to examine the main effects and interactions of sex and ethnicity on independent variables (aBMD, height, weight, lean mass, physical activity, dietary calcium). In a secondary analysis, height, and lean mass were used as covariates to determine if sex and ethnic differences in BMC/aBMD persisted when body size and composition were controlled for. Stepwise multiple regression models were fitted to estimate the contribution of the independent variables to absolute values of bone mineral. 64  Familial  Resemblance:  T o a s s e s s the familial r e s e m b l a n c e of B M C , P e a r s o n correlations w e r e  run for parent-child pairs on unadjusted s c o r e s and standardized s c o r e s . S t a n d a r d i z e d residual s c o r e s w e r e c a l c u l a t e d within e a c h generation, sex, a n d ethnic group adjusting for a g e , height, a n d l e a n m a s s . V a r i a b l e s were c h o s e n b a s e d on their theoretical a s s o c i a t i o n with T B B M C a n d w e r e significantly related to T B B M C in at least o n e of the groups. For c o n s i s t e n c y , the s a m e v a r i a b l e s w e r e e n t e r e d in regression e q u a t i o n s for all groups. Correlations were run for m o t h e r - d a u g h t e r ( M D), m o t h e r - s o n ( M - S ) , father-daughter (F-D), and father-son (F-S) pairs. To explore potential ethnic d i f f e r e n c e s in parent-child a s s o c i a t i o n s , mother-child a n d father-child correlations w e r e run within e a c h ethnicity. D u e to the small s a m p l e size, ethnic differences were not e x p l o r e d s e p a r a t e l y by s e x of the c h i l d . A s these data were exploratory, we did not adjust p-values for multiple comparisons.  Intervention  Study:  C h a n g e (final - initial value) a n d percent c h a n g e from b a s e l i n e w e r e  c a l c u l a t e d for all v a r i a b l e s a n d e x p r e s s e d a s m e a n ± S D . A 2x2x2 A n a l y s i s of V a r i a n c e ( A N O V A ) w a s utilized to test the main effects a n d interactions of group (exercise, control), s e x (M, F), a n d ethnicity ( A s i a n , C a u c a s i a n ) for both b a s e l i n e and c h a n g e variables. G r o u p d i f f e r e n c e s for c h a n g e in a B M D w e r e also tested with hierarchical regression to control for potential c o n f o u n d i n g factors. Independent v a r i a b l e s were entered in 8 steps including: b a s e l i n e a B M D , c h a n g e in height, c h a n g e in l e a n m a s s , a v e r a g e physical activity, average calcium intake, sex, ethnicity, a n d exercise group. D u m m y c o d i n g w a s u s e d to represent s e x (boys = 1, girls = 0), ethnicity ( C a u c a s i a n = 1, A s i a n = 0), a n d exercise group (exercise = 1, control = 0). T h e v a r i a b l e s c h o s e n for the regression m o d e l were b a s e d on e s t a b l i s h e d a n d theoretical relationships b e t w e e n these variables and c h a n g e in a B M D . Initial status w a s e n t e r e d first a s the relationship b e t w e e n initial status and c h a n g e in a B M D is c o n s i d e r e d important to all other relationships. C h a n g e in height a n d lean m a s s were entered next b a s e d on e s t a b l i s h e d relationships b e t w e e n these factors a n d c h a n g e in a B M D , e s p e c i a l l y in c h i l d r e n . M a i n effects of sex, ethnicity a n d group were entered prior to interaction terms. S e x w a s e n t e r e d prior to ethnicitythe q u e s t i o n of the effect of ethnicity controlling for s e x is a n e w o n e , while s e x d i f f e r e n c e s in b o n e  65  are well-established. Interaction terms are more meaningful with main effects partialled out, so these were entered in the last three steps. All data were analyzed using SPSS for Windows, Version 8.0 (SPSS Inc., Chicago). Results were considered significant if p < 0.05.  66  Chapter 5  RESULTS  Results are presented in three sections: Sex and ethnic differences in aBMD, physical activity and calcium intake are discussed in section 5.1. Section 5.2 describes familial resemblance in total body bone, lean and fat mass. Finally, results of the exercise intervention study are presented in section 5.3.  5.1 Sex and Ethnic Differences at Baseline  A total of 58 Asian and 110 Caucasian boys (n=86) and girls (n=82) participated in the study at baseline. Body composition, lifestyle characteristics and bone variables by ethnicity and sex are presented (Table 5.1). Children ranged in age from 6.9-10.2 years (mean 8.9±0.7 y). All of the boys and 89% of the girls were Tanner stage 1. The remaining girls were self-rated at Tanner breast stage 2.  Ethnicity.  As a group, the Asian children consumed less calcium (p<0.01) and reported participating in less physical activity (p<0.01) than their Caucasian counterparts. Differences were greatest between the Asian and Caucasian boys (Table 5.1). Estimated calcium intake was 506 mg/d lower for Asian as compared to Caucasian boys compared with a 338 mg/d difference for the girls. Similarly, the ethnic difference in physical activity estimated by questionnaire was greater between the boys (Figure 5.1). However, both Caucasian boys and Caucasian girls were more likely to participate in organized sport than their Asian counterparts (Table 5.2). While 70% of the Caucasian children participated in organized sport, only 14% of the Asian children reported similar involvement with sports teams. The Asian children (57%) were twice as likely to be attending academic lessons (Chinese school, mathematics, music, etc.) after school than the Caucasian children (28%).  Table 5.2 Percent of Asian and Caucasian boys and girls participating in organized sport Caucasian Total Asian Girls Boys Total  18% 11% 14%  65% 70% 67%  50% 49% 49%  68  as  u.  CM CO  tD CO  CO rsi •H  CO  ii o 0 CO  to to •H r-s is! CO  CO CO  m co"  m  •H  in  10  m  rs."  •tf  CM  0 CO  •tf  i  •tf  co  o to rs.  00  in  is. •tf  Cvj" CM  eg rs.  CO CO  CVJ  O) CO rs. •H CO o  •siCO cri  CJ) •tf  CO  i  CO rs:  to  CO  m  is. O  to o  CM CO  O CO  CO O  tf  •tf o  is. O tf CD  in  (3) rs.  to 0  in 0  to 0  CO d  in d  •tf  to 0  o  in  0  CM  in CO rs: •H  m  CO -H  rs. 0  rs. +1  to 0  tf tf tf  in  to o  co  rs. to d  in co d  co m d  Is. 0  rs. 0  CD 0  tf ii CM to  CM  CO d CO  c co 3 (0  O  CM  Si  CO CM  to  £ a> u_  to  s  to  8  to •H  m  CM CO  in  8  rs. +)  8  .2  co  CO  CO* CM CO  CO  p CO •H  CM  O  CO  •tf CM to •<t •tf"  CO  •H •tf to  O)  00  in  rs. OJ CM  in  at  c4 o4 " CO o>  o to to -tf  O)  0D CO CM  1--. •H rs.  o>  CO isi •H CO CM CM  co  CM •tf O  CO  is. •tf  0  •fl  CO CO CM  cu.  at  8 tf " 0 at  CO  in CO  •a tf  is. (si •H CO •tf CM  CO •H 00 0  in  rs.  0  tf tf tf tf tf •tf  •H  m to d  CO d  to o  CD O  CO in d  1— CO d  to in d  ai  in"  to •fl  o  CO  m co +1 o to  •tf CO •H O)  in  •tf  to 0  in  +1  to •H CM  M  CO  co" -H •tf at  CM CM CM CM  to  co"  •H o rs.  CM" CM  at  i  CO +l to  •H CO  CM  rs.  +i CM 0 is.  co  LO  in  rsi •H  co is.  tf  CD isi IS.  to CO •H CM  ai  ,_!  9 1 CO  d  d  to 0  to 0  m 0  m 0  to 0  ?H CO CO d  to d  in d  00 d  in  is. O  is. O  CD o  rs. o  3s  tf CM  0  CD  rs. CD CO •H  m  CO O  rs. 0  i  i  i  Is. o  rs.  0  m d  d  ? to  in d  in 0  00 0  i  i  00 rs. 0  CD O  m 0  CO 0  i  i  ? to  CM CD O  to o  CO to  •*  3 tf  8  o 00  d  Is. O CM CD O  CO  ? i ii ii to CM CO CO  CM to  00  CM 00  +i •tf rs: CD  r^ o  CM O) CO -H |srs. rs.  CO rs.  cb  tf tf i at  in  to d  CM" CM  co"  0  5 9  •tf  CM <tf CO  0  SO  CM  o 0  to  co  so  00 CM  -H is.  •fl  CO CM  i  ZZ  to > o  rs. •H  O O)  •H CD LO  to +1 o CM CO  CO  CM CM  in co"  4,563  ii  is. CO co" +1 rs. to O)  ,127  •5  CO  22,51  rs. •H 00  to  m d  CM 00 d  CD  in d  o LO" V O Q.  ^¥ CL S  CO (0 JO <5 co E O E E  1  2 2 c  c  CD CD TJ CJ) E,  ? 5 I  5  c sfe LL os  1 3  (0 cu  2 -a S?  rf  E  CO  O  TJ TJ  E  IQ E m eg  69  .>• >>J_ LL Q-  to  | | DJ g i  5) to  Figure 5.1 Physical activity in Asian (striped bar) and Caucasian (open bar) boys and girls. Significant sex*ethnicity interaction (p < 0.05).  Ethnicity ^ Asian H Caucasian  Error Bars show 95.0% Cl of Mean  o §  0.550  Asian Females Males  Caucasian Females Males  Figure 5.2 Femoral neck (FN) areal bone mineral density (aBMD, g/cm ) for Asian (grey markers) and Caucasian (black) boys (triangles) and girls (circles) at baseline. Significant sex by ethnicity interaction (p<0.05). Lines connect the mean values for boys (dotted line) and girls (solid line). 2  70  Ethnic differences were apparent for aBMD at the PF and FN (p<0.001) with Asian children having -3% lower values on average. At the FN, there was a significant sex by ethnicity interaction (p<0.05) both before and after adjusting for lean mass and height (Figure 5.2). Group means for FN aBMD were 7.6% lower in Asian than Caucasian boys prior to adjusting for covariates. This value decreased to a difference of 4.0% after controlling for size, but differences remained significant. Bone mineral density values were not different between Asian and Caucasian girls. TB, LS, PF and trochanter aBMD were not significantly different between the ethnic groups after adjusting for covariates.  Sex Differences. Differences were not significant between boys and girls for height, weight or dietary calcium intake (Table 5.1). However, as evidenced by the significant sex by ethnicity interaction (p<0.001) and the group means, the Asian girls were more active than the Asian boys whereas this trend was reversed amongst the Caucasian children where boys were more active than girls (Figure 5.1). The boys had significantly greater lean mass and % lean (p<0.05), and significantly less fat mass and % fat (p<0.001) as compared with the girls. These relationships were consistent within ethnicities as evidenced by the lack of a sex by ethnicity interaction (p>0.05). For aBMD there was a significant sex effect, the boys values being higher, at the femoral neck, proximal femur, trochanter and total body (p<0.01). However, the significant sex by ethnicity interaction (p<0.01) indicates that the differences in aBMD were within the Caucasian group where boys' values were approximately 10% higher than girls'. There was a negligible difference in PF and FN aBMD between the Asian girls and boys (Figure 5.2). Sex differences for aBMD remained significant after entering height and lean mass as covariates.  Regression. In multiple regression models, lean mass was the most significant predictor of aBMD at every site except the TB. For the TB, sex explained 9.4% of the variance and lean mass and fat mass an additional 16% (Table 5.3). For the PF and FN, lean mass, sex (both p<0.001) and physical activity (p<0.05) entered as significant predictors of aBMD and accounted for 29% (PF) and 37% (FN) of the  71  total variance respectively (Table 5.3). Lean mass was the only variable to enter for the lumbar spine regression model explaining 19% of the total variance in aBMD.  Areal bone mineral density (aBMD) regression model summaries including standardized and unstandardized Beta coefficients and adjusted/ for the total body and femoral neck sites.  Table 5.3  2  Total body aBMD  Variable Standardized Beta Sex .136 Lean .562 Fat -.436 'variables not in model: ethnicity, height, All predictors significant (p < 0.001)  Unstandardized Beta 1.33E-02 8.12E-06 -5.45E06 activity, calcium  Adjusted .094 .155 .252  Femoral neck aBMD  Variable Standardized B Unstandardized 13 Adjusted P/ Lean .234 9.50E-06 .295 Sex .470 3.18E-02 .352 Activity .154 5.73 E-04 .371 'variables not in model: ethnicity, height, fat, calcium All predictors significant. Lean & sex p < 0.001, physical activity p < 0.05.  5.2 Familial Resemblance  A total of 86 parents (56 mothers and 30 fathers) participated in the family study (Table 5.4). Asian families had moved to Canada 3 years previously on average. Children's height, weight, lean mass, fat mass and BMC, were expressed as a percentage of their parents' (average) values. On average, children (mean age 8.9 y) had reached -80% of parental height, but only -45% of parental weight, lean and fat mass. Children's total body BMC was -40% of parental values. Number of Asian and Caucasian family groups participating in the study. Asian Caucasian Total M-F-S 3 8 11 M-F-D 6 8 14 Total full families*: 9 16 25 M-S 9 20 29 M-D 8 19 27 Total M-C pairs 17 39 56 F-S 4 11 15 F-D 6 9 15 Total F-C pairs: 10 20 30 M = mother, F = father, S = son, D = daughter. *Total full family refers to families where both parents and one child (either a daughter or son) participated.  Table 5.4  72  Descriptive variables for parents are provided (Table 5.5). Caucasian parents were taller, weighed more and had greater fat mass and higher calcium intakes than Asian parents (p<0.05). Hours of weight-bearing physical activity (averaged from the past two weeks) did not differ between Asian and Caucasian parents. However, fewer Asian parents reported participating in regular physical activity both during childhood (Asian 40%, Caucasian 76%) and adulthood (Asian 17%, Caucasian 61%). Sex by ethnicity interactions were significant for BMC and lean mass which was explained primarily by ethnic differences within fathers. Unadjusted TB BMC was 30% lower in Asian compared with Caucasian fathers, and 10% lower in Asian mothers compared with their Caucasian counterparts. After adjusting for lean mass, height and age, BMC values were still -20% lower in Asian than Caucasian men, but only 1% different for the women.  Table 5.5 Descriptive characteristics for Asian and Caucasian mothers and fathers. Values are Mean±SD. Mothers  Fathers  Asian  Caucasian  Total  Asian  Caucasian  Total  n  16  33  49  10  18  28  Age (years)  40.3±4.3  40.9±4.1  40.9±4.0  Height (cm)  157.4±4.6  165.6±7.2  Weight (kg)  54.6±7.7  65.8±18.6  TB B M C (g)  1956±239  2192±229  TB B M C * *  2234  2268  Fat mass (kg)  16.1±3.9  22.3±12.8  Lean mass (kg)  37.8±4.5  %fat  28.6±3.9  %BMFL  +  t  42.0±4.0  41.0±5.4  162.9±6.8  X  167.8±6.0  179.8±4.0  62.1±16.7  X  68.6±10.6  94.6±16.4  2050±206  2924±238  2115±256  x  2251  2039  2549  20.3±10.8  15.2±2.6  23.1 ±9.1  46.7±6.6  43.8±7.3  X  52.0±9.0  69.5±8.1  29.6±8.0  29.3±6.9  X  22.3±4.1  23.4±5.9  67.8±3.7  67.2±7.5  67.4±6.4  X  74.8±4.2  73.4±5.6  Calcium (mg/d)  252±265  688±460  531±434  314±225  658±426  Current activity  1.9±2.1  2.2±1.5  2.1±1.7  1.8±1.5  2.4±3.1  +  t  41.4±4.9 1  175.5±7.5  t  85.3±19.2  n  2612±482 2294  1  t t  20.3±8.3 63.3±11.9 23.0±5.3 73.9±5.1  t  551±400 2.1±2.4  (hrs/wk) "Significantly different from males (p < 0.001), "Significantly different from Asians (p < 0.001), "Significant sex by ethnicity interaction; " B M C values are adjusted for lean mass, height and age; B M C = bone mineral content  73  Familial  resemblance  Parent-child correlations for TB BMC, height, weight, lean and fat mass are shown (Table 5.6). Daughters closely resembled their mothers for height, weight, lean and fat mass. Similarly, fatherdaughter correlations for weight, lean and fat mass (but not height) were also close to 0.50, however only weight was significant (p = 0.042). Unadjusted TB BMC correlations approached significance for mother-daughter (p = 0.054), but not father-daughter (p = 0.369) pairs. Mother-daughter and father-daughter correlations were not significant for adjusted TB BMC scores. There was a trend for sons to more closely resemble their fathers than their mothers for weight, lean and fat mass although none of the correlations reached significance. There was also a resemblance between sons and both parents for unadjusted TB BMC values. After adjusting TB BMC within each group, these associations decreased and became non-significant. When pairs were split by ethnicity, only the mother-child height association within Asians was statistically significant (r = 0.56, p = 0.019). There were trends for stronger mother-child correlations within Asian pairs for weight (r = 0.43, p = 0.096), lean mass (r = 0.476, p = 0.062), and unadjusted TB BMC (r = 0.42, p = 0.103). In contrast, father-child associations, although not significant, tended to be stronger for Caucasian than Asian pairs for weight, lean and fat mass (Caucasian, r = 0.42-0.44 vs. Asian, r=-0.12-0.03) but not BMC. Adjusted TB BMC values were not significantly related for any of the pairs in either ethnic group. There was no relationship between children's and parents' calcium intake (r=0.169, p=0.126), or between children's and parents' physical activity scores (r=0.042-0.173, p>0.05)..  Table 5.6 Pearson product moment correlations comparing total body bone mineral content (TB BMC), fat and lean mass, height and weight between mothers (M), fathers (F), and their daughters F-D F-S M-D M-S 21 28 14 14 N 0.584" 0.426 0.439" 0.354 TB BMC ' 0.365 0.277 -0.260 TB BMC residual* 0.156 -0.097 0.064 0.016 0.481" Height 0.455 0.538 0.025 0.550" Weight 0.454 0.512 0.537 0.310 Lean 0.390 0.204 0.508 0.478 Fat *residual scores adjusted for age, height, and lean mass within each sex, generation and ethnicity "p<0.05 x  x  x  74  5.3 Mechanical Loading: 8-month School-based Intervention A total of 144 children returned for measurements in June for a total of 81 children from control schools and 63 in the exercise group (Table 5.7). Several of the girls advanced from breast stage 1 to stage 2 during the study including 15 of the girls in the control group and 13 in the exercise group. Tanner stage was not related to change or percent change in body composition, anthropometric or aBMD variables and there were no significant differences in these variables between Tanner 2 and Tanner 1 girls. Teachers' logs indicated all intervention schools completed the minimum of 10 tuck jumps 3x/wk throughout the intervention period. A minimum of 10-20 minutes of the intervention games were reportedly completed within the PE classes 2x/wk with the exception of -3 wks during the intervention period for school breaks.  Table 5.7 Number of participants in control and exercise groups by ethnicity and gender. Exercise Control Total Female Asian 14 8 22 Caucasian 21 27 48 Total 35 35 70 Male Asian 19 8 27 Caucasian 27 20 47 Total 46 28 74 Total Asian 33 16 49 Caucasian 47 95 48 Total 81 63 144 Baseline  Characteristics.  Exercise and control groups were well matched at baseline (Table 5.8). Analysis of variance showed no significant (p > 0.05) baseline differences between the exercise and control groups in any of the anthropometric variables including calf girth, weight, height, percent fat or lean mass. Initial aBMD also did not differ between the groups at any site. Calcium intake was similar between the exercise groups (control = 988 ±611, exercise = 918 ± 404 mg/d; p = 0.456). Children from control schools reported significantly greater physical activity averaged from 3 questionnaires (control = 86.0 ± 15.4, exercise = 80.5 ± 12.0; p = 0.002) and performed better on the vertical jump at baseline (Table 5.8). Forty percent of the children in both groups reported participation in extracurricular  75  sports. There were no differences (p > 0.05) in baseline or change data based on participation or non-participation in organized sport.  8-month Change  There were no significant sex differences for any of the change variables (p > 0.05). Asian children had a greater increase in TB aBMD (Asian = 1.8%, Caucasian = 1.1%, p = 0.015). Other change variables were not significantly different between the two ethnic groups. Change in anthropometric and bone data for the two exercise groups are presented (Table 5.8). Absolute and percent change in weight, fat mass, and lean mass were similar between exercise and control groups. The control group had a significantly greater (+1.5 cm) increase in height. The exercise group increased their vertical jump significantly more than the control group (p = 0.003). With respect to aBMD, 8-month change was significantly greater than zero in both exercise and control groups and at all sites (p < 0.05). ANOVA showed no group differences for change in TB, LS, PF or FN aBMD. There was a significant effect of the intervention at the trochanter with the exercise group showing an average of 1.2% greater gain in aBMD than controls (p = 0.030). Group (exercise vs. control) differences appeared to be greater in the Asian children (2.3%) than the Caucasian children (0.9%), as shown in Figure 5.3. The group by ethnicity interaction effect was not significant (p = 0.146) however. Exercise group interactions with sex and/or ethnicity were also not significant at other sites.  76  c o  cvj  CM  ;_  CO CO  E  II CD  TJ C  Sc  CO  •H O  CO  CD £=  c co o xi "cc >,o t= CO g £ CD CO CD  CO  •tf  LO CO •H O)  +i  CO  d  u 4> X LU  "53  in  LO  CO +1  CD CO CO  d co  CO  CO +1  o  LO  CO  TJ  •5 CO CM CM  CO +1 CD •tf" IS-  •tf  LO  If) CO  « ^  CO  LO  CO  E c?  cb  CM  LO  CD D)  LO  OJ  o®  1  3  TJ  TJ C  * tf  c  CO  O C\i  O  -tf CO  LO  co  o  CM  CO •H CO co  •H CO  d  CO  CO O  >. TJ  o  CD C O  _g CD  CO •H  "55 CO CO  LO  d  CO  CO  o" "co  s > "55 c E  rfs^.  CO * CO  tf C\j  •tf  •tf"  •H  •tf •tf o  CM  CO CO  CM CO  00  00  i—  T-  T—  CM  o  d  d  CO CM O  CO  CM CD  CJ) CO  CO CD  d  d  d  •tf  q  d  LO  •H  •H rs. CJ) rs.  o  CM  CO  CO CM  O  d  •tf  CM .—  o  f—  •tf  o  .—  CD  CO (si  CJ)  d  O  •H 00 CD LO  o  o  •H  •tf  CO CO  d  CM  q  O  •H CO CM CO  d  tf  •tf  CD  o  d  -H r>-  i—  LO  d  d  CM  ^ CM  O  CM  CO  CO  CO  OJ  CD .—  00  o  o  CO CM  CD  o  O  O  tf tf CM  tf  O  o CM q  tf o  o  •tf"  $  cn co  $2  ^  J; •  CM  CD •H CO  cb Is.  CM CM CO  CJJ  Zi  ,-  rs.  O)  tf o o  d  o  d  q  CO CO  d  .—  d  CJ)  d  .—  |s-  d  ^  o  LO LO LO  co" co" •H •H o |sy—  n  0 0  3 tf  LO  CD  d  d  q  LO  •H  •H |s-  LO  CM CM  CM  q  •H  o  |s-  o  o  •H CO  •H CM  d  O  co  LO  o  o  O  CO  cb" co"  LO  CO  d  •tf  CO O  CO CO  q  d  •H CO CO LO  o  c  o <°  CD  CO LL  "3  t c CO _l  LL  ias  CD  CO CO  os CO  CD  CO CO  «  CD CO  CL  E  tica  "co O  s?  s?  ean rrias  CO  C5  O)  o v  ma  "I  _c  ^ ^ „^  ^  ma  4> o  o  ght (cm  CO 00 <D U) 3  E  _i  >  —>  CD  E o Q  s m Q 77  CD C  >,  CL  3  E  CD LL  o CD  CO "5 c CO CO E CO x SZ -.—• " co X I o o o o E h- 3 CL LL h—  po  <  E  ight  a. 2 O CD H CL|  ,5  cb  CO«  _d  !o i  CO CO •H  •tf  LO  CO  .£ co  § §* ^ &i W o T7 C co o c o  •tf is!  ±0.0  CO  CD  ±0.0  IS,  LO  ±0.0  cn c  ±0.0  CD  Error B a r s s h o w 95.0% Cl of Mean  OD -0.020  Asian Caucasain Controls  Asian Caucasian Exercise .  Figure 5.3 Change over 8-months in trochanteric areal bone mineral density for control (circles) and exercise (diamonds) groups. Lines connect mean values for Asian (grey line) and Caucasian (black line) children in the control and exercise groups.  After controlling for confounding variables within the regression analyses, the group effect remained significant at the trochanter. The final model (r = 0.37) explained -14% of the variance (adjusted r = 0.09) for change in trochanter aBMD. Of the variance explained, group (exercise, 2  control) contributed 3%, initial aBMD value 4%, and change in lean mass 5%. The remaining 2% is attributed to the non-significant variables within the model (change in height, sex, ethnicity, calcium and physical activity). Final regression models for change in TB, LS, PF and FN aBMD were not significant (Table 5.9). Pearson correlations of change in aBMD and independent variables entered in the regression models are also presented (Table 5.10).  78  in o  d v CO  o  r» co o  d  .—  n o  d  o  d  CD CD  o  d  CD E  ©  00 CO o  d  o  ^—  o  d  o Cvi  CO CD  CJ) o d  O c  CO x : o UJ c  CO  o  o ° CO CD  d  d  o  d  c  in o  d  CO  o  d  CO  Q  d  'a CD ^ O  CD  o co co c  00  CM  o  CO CO  o  CM  o  o o  CO o  o  c JS  CO CD ° CD x: -H- C D i_  O)  c  C ' D JS I  CN  M—  © 2  C CM O  ... —  CJ>  o  1—  d  o CO CM  CO ID o  CO  in  CO  d •  d  o  CO  d  CD  o  CM  3  CO  in  J3  c  C O o o  CO X) (D E .E 3 CL  —I co  CO  1-s  CD CD  CO E _  o  2 E  o  * 2  £ 3  c  o o  c :  i 5 £  O  CM m o  O  co  <  CO  m o  co"  CO O O  2  CO  CO CO CL  in o  3  CD O C CD  oo  CJ>  o  O  o  o  CD CO  CO  in  00  CO CD  co CJ)  o  d  d  d  CM d  co o o  CM O  T  d O  := "O  79  q d  o d  c  >>  q CO d cCD TJ  o  '3  c C CD Dc « . E | CCDO TJ CD CO  o CM d  CO T CM d  3  CD x ; o> CJ>  T—  CO CD c o o d  X)  o  CM  O CO O  CO  o  To CD CO c CD  in co  CJ) c  c  CD «S  CO c  CD  E Q o5 S co CD co co  CD  CM CM O  o  d  CJ) q d in  o E  CO CM CM d  00  CD CO co  a> Xl  o  >  -  CO CM  «  CO CJ CO ° CD TJ >. O CD Ec ® =o  CD  CO  CO  CD g E 0)0) ® CD TJ CD 2 CD £ "> VCO CO Q. 3 ID _CDCJ) CO C o => ° CD 2CD c•— c CD  d  CO CO  CD  O  "o co"  o o  CO O CO o TJ CO c Co CO CO > . _co O .1= CD TJ 55 > O E 0. <°  1-s i l l  ©  "5  CD £ Q  o  CM CM O  CM O  x:  « C y- O CD 2 <0 CD § CD 2 E CL E 2 CD O 2 co" o -5 x> -o ® <Cu o Si £ <  o  CO CO"  CO E CD  .2  3  CJ) CD  m  00  co o  ^ CD  Q. o in o  c  D) '  o -  CO  00 o  o  _co  CD TJ o E c o "to D. CO CD CD to "co O E E ?  TJ  CJ>  CO  d  TJ CD CD -o := CO  in  CM  c  o  X  o co o  'aO  O  ci o O) CD CO O 5)  CM  o o  CD  o  CD CL 3  CD  "C ooD TJ  CO  TJ o  ° 13 m  CJ)  oo  CD oo  O  .2=1 CO  c  TJ CD  "c  co CD CO U)  CD  J5  CO  5  "5  CD CO CO c o  If  d  o  O \  O) 'co  £ CD  in o  3  'E  ±= x:  d  c  CO  Cu CD CD - C  CO CO CM 00  CL  a.  co  O o  c  C O o o  C D C 'CL  CD O  E  CD Li_  co C O >  3  E _® "co  E 'x o  CD TT TJ o C XI C D Q_ o QCD 1  Chapter 6  DISCUSSION  This is the first comprehensive report of a school-based exercise intervention, and crosssectional determinants of aBMD in a large cohort of pre- and early-pubertal Asian and Caucasian children living in a common geographical region. Results of this study support the emerging notion that lifestyle and genetic influences on the skeleton are apparent early in life. The implications of the following aspects of this thesis are discussed as follows: Cross-sectional sex and ethnic differences in aBMD and lifestyle factors are discussed in section 6.1; Familial resemblance of total body bone, lean and fat mass are discussed in section 6.2; A discussion of the mechanical loading intervention is in the final section, 6.3.  6.1 Sex and Ethnic Influences on aBMD in Prepubertal Children  Little is known about ethnic or sex influences on bone mineral during the prepubertal years as most normative studies have been limited to Caucasian subjects [7, 255], and have had only a small number of prepubertal children. Ethnic differences in aBMD between Asian and Caucasian groups have been attributed primarily to variations in body (bone) size [147]. In this study of prepubertal Asian and Caucasian boys and girls living in geographical proximity, site-specific sex and ethnic differences in bone mineral were observed that were not fully accounted for by differences in body size or lean mass. Differences in lifestyle factors known to influence aBMD were also evident.  Ethnicity  Physical activity and dietary calcium, important determinants of aBMD, have rarely been directly compared in Asian and Caucasian children. In this study, Asian children were less physically active than their Caucasian counterparts. In the only published study that included both Asian and Caucasian children, Bhudhikanok et al. [147] reported similar results using an activity frequency questionnaire but with far fewer (18 Asian; 34 Caucasian) and slightly older (I0.5±1.1y) children than in the same maturity group we describe. The mean estimated dietary calcium intakes of Asian children in this study were greater than for Asian children living in Hong Kong [256] but lower than for the Caucasian children studied. In a study of five year old Chinese children in Hong Kong, those who consumed twice the dietary calcium intake of children in a mainland Chinese city had a 14%  81  greater radial BMC as measured by single photon absorptiometry [256]. The Caucasian-Canadian children in the present study had calcium intakes (estimated from a calcium food frequency questionnaire) comparable with those reported from 3-day food records in similarly aged Canadian children in the University of Saskatchewan Bone Mineral Accrual study [43]. The significant ethnicity effect for proximal femur and femoral neck aBMD within boys differs from the findings of Bhudhikanok et al. who reported no ethnic differences in boys until mid-puberty [147]. Bhudhikanok et al. attributed their outcomes to size differences. Although height and weight were not significantly different between Asian and Caucasian children in our cohort, ethnic differences in aBMD decreased after accounting for height and lean mass. However, FN aBMD was still 4% higher in Caucasian boys after controlling for height and lean mass. The lack of difference observed by Bhudhikanok et al. may reflect the small numbers of pre/early pubertal children in their sample [147]. A strength of the present study is the larger number of prepubertal (only) subjects than has previously been reported. For total proximal femur, trochanter, body and lumbar spine our data support the contention that, after accounting for size, there are no differences between Asian and Caucasian children.  Sex There was a trend toward greater physical activity in the Caucasian boys as compared to the Caucasian girls whereas the reverse trend was observed in the Asian children where the girls were more active. It is unlikely that this relatively small difference explains the observed sex difference in proximal femur and total body aBMD. It is accepted that fat mass increases gradually throughout childhood with a dissociation between boys and girls at about age 8 or 9 years [179]. This dissociation increases dramatically at adolescence when girls average approximately twice the fat mass of boys [179]. This sexual dimorphism was illustrated in the present study. Although lean mass has not been studied extensively in children, it is generally accepted that there is little difference between boys and girls prior to the adolescent growth spurt. In the present study, however, boys had significantly greater lean mass than girls by age 8.9 y. These data support results of recent studies of lean mass assessed by DXA in large groups of prepubertal children [145, 180].  82  Our finding of a male-female difference in aBMD at four sites (TB, PF, FN, and trochanter) in this age group is novel. There is one report of a difference at the femoral neck in a large cohort of prepubescent children from Australia [145]. Most studies of bone mineral in similarly aged and older girls and boys have examined the lumbar spine and report no difference between boys and girls. The few studies that examined the femoral neck included only a small number of prepubescent children in their samples [5, 38, 40]. Over 25 years ago, using radiogrammetry techniques, Garn reported a larger cortical area in boys at all ages [55]. These measurements were at different sites than those reported in the present study. Nonetheless, this structural difference might explain the significant sex difference in aBMD.  Summary of sex and ethnic differences  In summary, this study of a large number of Asian and Caucasian prepubertal children demonstrated differences in modifiable lifestyle factors between Asian and Caucasian boys and between Caucasian boys and girls. As ethnic and sex differences were also observed for femoral neck aBMD after adjusting for lean mass and body size, these data may have important long-term implications for fracture risk. It would, therefore, seem logical to suggest lifestyle interventions in this group. This argument is supported by the secular trend toward increasing incidence of osteoporotic femoral neck fracture among older Asian men [243].  6.2 Familial Resemblance in Total Body Bone, Lean and Fat Mass  This is the first report examining familial resemblance of total body bone mass in a multiethnic group of mothers and their prepubertal sons, and in fathers and their prepubertal sons and daughters. We found evidence of familial resemblance in TB BMC, as well as lean and fat mass, between these Asian- and Caucasian-Canadian children and their parents. Familial resemblance for total body bone mass was not apparent after accounting for lean mass and body size. These data support the emerging notion that heritability of bone mass is complex and interacts with variables such as age, sex, ethnicity, body size and composition [111, 178].  83  A number of previous studies have shown associations for regional (lumbar spine, femoral neck) bone mass or density between adolescent girls [117] or adult women [115, 121] and their mothers. There is only one report of familial resemblance for regional BMC and areal BMD (aBMD, g/cm ) in prepubescent girls and their mothers [125]. By way of comparing our results to these z  studies, we will address a number of factors which may influence the familial resemblance of bone including; 1) age of both parents and offspring, 2) the relationship between lean mass and bone, and 3) ethnicity of the population.  Age Relationships in bone mineral appear to be stronger for adult offspring and their parents than those for young or elderly offspring and their parents. Estimates of heritability are higher for motherdaughter pairs if both members are premenopausal as opposed to postmenopausal, and correlations decrease with increasing postmenopausal age of mothers [121, 257]. In younger age groups, our data indicate a weaker familial resemblance between prepubertal children and their parents. As compared to early adulthood, bone mineral content or density is -10-30% lower both prior to the attainment of peak bone mass (prepubertal) and during the postmenopausal years when bone mass is diminishing. It is plausible that familial comparisons would be more similar around the time of peak bone mass than either before or after peak when individuals are gaining or losing bone at variable rates. In a slightly older cohort (11.8±2.1y), McKay et al. [117] reported significant correlations (r = 0.31-0.49) for aBMD between daughters and their premenopausal mothers (mean age 40.0y). The girls' proximal femur and lumbar spine aBMD values were, on average, 83-93% of their mothers' values. The children in our study were -3 years younger and had reached 70% of their parents' values for total body aBMD, and 40% of their parents' TB BMC. Ferrari et al. [125] reported significant correlations for proximal femur, lumbar spine and midfemoral shaft BMC and aBMD in 138 prepubertal Swiss girls (mean age 8.1y) and their mothers (mean age 40.Oy). However, the reported heritability estimates of -18-36% for BMC and aBMD (adjusted for height, weight, age, and calcium intake) are -20% lower than those reported for adult daughters and their mothers [115]. Correlation coefficients for lumbar spine and femoral neck BMC and aBMD (standardized for age only) were 0.25-0.30 [125]. Caucasian mothers and daughters in  84  our study were similar in height and weight to the Swiss sample. It is possible that the differences in familial resemblance between the studies is due to population differences. The Swiss population may be more homogenous than our heterogeneous population from Europe, North America, and Asia. It is difficult to compare the two studies as variables adjusted for differed. However, the correlations for unadjusted TB BMC between mothers and daughters in our study were similar to the regional correlations adjusted for age [125]. Ferrari and colleagues adjusted for weight, but not lean mass, before predicting heritabiity estimates. Evidence for an interaction of heredity with lean and bone mass is discussed below. The present study suggests familial resemblance in bone mineral may be lower, or nonsignificant, before pubertal growth begins. In their study, Gueguen et al. [122] predicted heritability estimates for bone mass peak at age 26, and are lowest prior to age 8 and after age 44. Taken together, these data support a stronger resemblance in bone mass during the adult years when bone mass of both family members is at or near its peak, and a weaker association during prepubertal or postmenopausal years.  Association  between  bone and lean  mass  It is accepted that body weight and size influence DXA measures of BMC and aBMD. Relatively new DXA body composition software, has made it possible to accurately distinguish the independent contribution of lean and fat components of body weight to bone mineral. Although data in prepubertal children are limited, bone-lean mass associations in children and adolescents are higher than bone-fat mass associations [111, 180]. In a cohort of 215, 10-to 26-year old female twins, the genetic variance decreased by half in the early-adolescent years when lean mass was accounted for [111 ]. In the present study, univariate correlation showed a higher association between lean mass and total body bone mass (r = 0.80-0.94) than between fat mass and bone (r = 0.57-0.64). A strength of the present study is that we have controlled for differences in lean mass, whereas most studies adjust for age and/or weight alone. /  85  Ethnicity  To our knowledge, there are no other studies that compare familial resemblance in bone mass within Asian-Canadian family groups; or any ethnicity other than Caucasian. We found that Asian and Caucasian families had varying degrees of resemblance for bone, fat, and lean mass. These data must be interpreted cautiously given the small sample size. However, it is plausible that emigration from Asia to Europe or North America leads to alterations in lifestyle that, in turn, influence body size, bone mass or both. For example, estimated calcium intakes were lower for Asian than Caucasian children and parents in our study, but children's values were higher than the average of <500 mg/d reported for individuals living in Hong Kong [243]. Asian boys and parents also reported participating in less physical activity than Caucasians. Other studies have shown differences in adult bone mass with emigration. Premenopausal Chinese women who spent more than 12 years in Denmark had aBMD values similar to Danish women, while those Chinese women living in Denmark for less than 12 years had 4-7% lower aBMD at all sites [258]. Similarly, after adjusting for height and weight, U.S. born Japanese women had 45% higher calcaneal and radial BMD than women living in Japan [155]. In an adolescent cohort living in California, after body size was accounted for, there were virtually no differences in bone mass between Asian and Caucasian groups [147]. In our study, ethnic differences in bone mass were apparent for fathers even after adjusting for these variables. In contrast, TB BMC was similar between sons after adjusting for lean mass and height. The Asian families in our study had recently moved to Canada 3 years previously on average. These cultural and lifestyle factors together may serve to influence the genetic template for bone mass/density. A limitation of the present study is that bone mass in children and parents were measured during different seasons. Bone mineral accrual rates increase dramatically over the summer months [204]. In contrast, remodeling slows during winter when Vitamin D synthesis decreases. The seasonal difference in measurement periods between children (measured in October), and parents (measured in April), may have influenced familial resemblance. As this was a cross-sectional study,  86  the influence is likely to be minimal. Any seasonal effect on bone mass or density would be a particularly important consideration for longitudinal studies.  Summary of familial resemblance In summary, these data suggest familial resemblance for total body bone mass in prepubertal children is influenced by lean mass. Our data also suggest that ethnic differences in heritability of bone mass may be influenced by cultural differences in physical activity and calcium intake. Future prospective studies are warranted to further examine ethnic and sex differences in heritable influences on bone and lean mass and their interactions on the pediatric skeleton.  6.3 Mechanical Loading: Response to an 8-month School-Based Intervention Childhood appears to be a critical period when bone may be particularly responsive to weight bearing physical activity [42]. We intervened with a modified school-based physical education curriculum to moderately increase skeletal loading and found a significant increase in areal bone mineral density (aBMD) at the trochanteric region of the hip. The exercise intervention did not affect 8-month change aBMD at other skeletal sites prior to, or after, controlling for initial status and differences in growth. Although the difference between exercise and control groups for absolute trochanteric aBMD change were more than twice as high in the Asian (2.3%) compared to the Caucasian children (0.9%) the randomized nature of the design resulted in only a small number of Asian children in the intervention group (n=16) and this difference did not achieve significance (p=0.15). The aim of this research was to identify a practical method to optimize peak bone mass so as to minimize the risk of osteoporotic fracture in later life. To our knowledge, this is the first intervention study that has targeted a large number of pre- and early- pubertal children participating in existing physical education programs. Previous exercise intervention studies in children have had far fewer subjects and programs were offered outside of regular school time by professional instructors [177, 230, 259]. The additional expense and time such programs demand make them less practical, less economical and, thus, less likely to be implemented as a public health initiative. The program  87  designed for the present study was successfully implemented in 9 classes by teachers who were trained in the curriculum but, with the exception of one teacher, had no formal training in physical education. As the program conformed to the mandated Instructional Resource Package for elementary school PE, all of the children in the intervention classes (approximately 200 children in grades 3 and 4) participated in the program irrespective of whether or not they volunteered for bone mineral assessment. Further, although jump ropes were provided, activities could be completed without schools purchasing additional equipment. Therefore, our study suggests that a large-scale school-based intervention for bone health is eminently feasible. Costly and debilitating hip fractures commonly occur at trochanteric or cervical (femoral neck and intracapsular) sites. The etiology of hip fractures differs by site and trochanteric aBMD has been suggested to be a good predictor of fracture [260]. An increase in trochanteric aBMD with exercise has not been previously reported in children or adolescents. However, the osteogenic effect of jumping on the highly trabecular trochanter is biologically plausible and is consistent with current theories and research evidence concerning mechanical loading. Exercise has a site-specific effect on bone and bone responds to mechanical strains primarily from forces generated by muscle attachments and also ground reaction forces (GRFs) [261, 262]. The large muscles that are utilized during jumping (i.e. gluteus maximus, gluteus medius) [263, 264] attach at the greater trochanter. Also, the majority of forces placed on the proximal femur during loading are absorbed by the trochanter [265]. Our results differ somewhat from the two previous exercise intervention studies on bone in children of similar age, with respect to the magnitude of effect [177] and number of sites affected [177, 230], In a non-randomized study, Morris et al. reported greater increases in aBMD for all measured sites in premenarcheal girls. The largest difference was at the femoral neck with a 10% greater increase in aBMD for the exercise group [177]. A similar intervention program showed much smaller effects for the legs (3%), total body (1%) and lumbar spine (2.5%) in prepubescent boys [230]. Possible explanations for the observed differences are discussed below.  88  Maturational characteristics As rate of maturation explains most of the gain in bone mineral during growth [266, 267] differences in maturational development between groups represents a large potential confound in any study of growing children. In studies with small numbers of subjects in this age group, a difference in rate of maturation in even a few children between control or exercise groups can significantly influence results. A strength of the present study was the relatively large number of children at a similar stage of maturation at baseline. There were no differences in either the change, or absolute, values for bone mineral, height or weight between girls who remained in Tanner stage 1 and those girls who were Tanner 2 at final measurement. The large exercise intervention effect in the only other prospective exercise intervention study of premenarcheal girls [177] has been attributed to differences in maturity status between the groups [230]. As evidence of this, the 10% greater increase in femoral neck aBMD in the exercise group was no longer significant when the effects of change in weight and height were controlled for between groups [177]. In our study, there was a greater increase in height in the control, versus the exercise, group. This could potentially bias the outcome away from finding an effect of the intervention. By including initial status, change in height and change in lean mass in the regression models, we accounted statistically for any association between accelerated growth and change in bone mineral density. Pediatric exercise intervention studies may also differ in outcomes because of baseline maturational differences between cohorts. In a recent cross-sectional study of tennis players, Haapasalo et al. reported that side-to-side differences in humerus aBMD were not significant until Tanner stage 3 (12.6 y) [229]. Other cross-sectional and retrospective studies in both humans and animals, despite methodological limitations, suggest there may be an age-related effect of mechanical loading [9, 87, 99,220]. More than 50% of the Morris et al. subjects were Tanner breast stage 2 at entry as compared to our 89% Tanner stage 1 girls at baseline [177]. A greater osteogenic response may occur with an identical intervention in a slightly more mature (1-2y) age group [229].  89  Characteristics  of the  intervention  The exact load required to elicit an osteogenic response has not been clearly defined for young children. Animal studies suggest that the most osteogenic activities are of high magnitude and unusual distribution [81] but are not necessarily high in number [95]. To our knowledge, the loads imposed by jumping have not been published for this age group. In a pilot study of a small number (n=5) of prepubescent children in bur biomechanics laboratory, tuck jumps elicited ground reaction forces (GRFs) 2-5 times body weight. In older women, landing from an approximately 8 cm jump, induced GRFs 3-4 times body weight and joint reaction forces 3-5 times those forces at the hip [263, 268]. These moderate forces induced a 2-3% increase in trochanteric aBMD after 5 months of jumping in premenopausal women [269]. However, as children on average have relatively less muscle mass per unit body weight than adults, tensile forces at the trochanter could be lower than those experienced in adult studies. It is probable that the present intervention was of high enough impact to stimulate a (re)modeling effect at the direct site of muscle attachment (trochanter) but not at other regions of the proximal femur. As Bradney et al. [230] did not report bone outcomes for the regions of the proximal femur, comparisons at this site between studies for boys is difficult. However a more diverse program of sport intervention might explain their observed positive change for the total body and lumbar spine. Other aspects of the programs including progression, volume, duration and frequency, would also influence the physiological response to the intervention.  Differences  in physical  activity between  the groups at  baseline  Although the control and exercise groups in our study were well matched for anthropometric variables, the control group reported participating in more physical activity compared with the exercise group and had a higher vertical jump at baseline. A higher level of physical activity in the control group may have reduced the possibility of finding a significant effect of the intervention. However, the differences between the control and exercise groups in physical activity reported in the questionnaire were relatively small and unlikely to translate to significant effects on bone mineral accrual. In support of this, there was no association between physical activity score and change in aBMD at any site. As well, controlling for physical activity within regression models did not significantly alter the outcomes.  90  Limitations  of DXA  technology  A final issue that needs to be addressed in pediatric studies is the method of scan analysis. Software limitations for analyzing pediatric spine scans have been addressed [270] as have limitations of using areal estimates of BMD from DXA when comparing children growing at varying rates [27, 33]. For exercise studies, an important issues that is not yet widely recognized is the analysis of proximal femur scans. The largest effect of exercise interventions are reported at the proximal femur and its regions (femoral neck and trochanter). We have shown that, when using Hologic systems, small changes in the size of the global region of interest (ROI), and the size or location of the femoral neck ROI, can alter aBMD values by 3-5% even with small amounts of growth [271]. In our analysis, we compared all time 2 scans to time 1 utilizing the compare mode and thus reduced the system error by maintaining the same ROI between scans for the proximal femur and femoral neck. Others have not reported analysis techniques in depth so it is unclear if the ROIs were increased with growth or maintained. Analysis techniques should be standardized within and across studies and methods well-described in published papers.  Summary  of mechanical  loading  intervention  Our randomized prospective study demonstrated that an exercise program implemented within elementary school physical education, increased aBMD at the trochanteric region of the hip in preand early- pubescent children. Until such time that long term prospective studies address the question of whether benefits from childhood exercise persist into adulthood, pervasive public health care policies and programs that promote physical activity during the growing years seem welladvised.  91  Chapter 7  GENERAL SUMMARY AND CONCLUSIONS  92  It is increasing clear that prevention of osteoporosis should begin early in life when optimal bone mineral accretion is critical to the attainment of a healthy adult skeleton. Intervention with physical activity during the years of growth is a promising means of population-based osteoporosis prevention. Although awareness of this critical period has increased over the past few years, there are a number of gaps in the literature regarding the bone response to exercise intervention in children. The studies of this thesis begin to fill some of those gaps. The implications of the study results are summarized with respect to: 1) Exercise as an effective intervention in pediatric groups, 2) Ethnic differences in the response to an exercise intervention and 3) Other determinants of the bone mineral response to intervention. Questions for future research are also addressed.  7.1 Exercise as an effective intervention in pediatric groups  There is increasing evidence that the pre- and early-pubertal years are an important period during which lifestyle factors can influence aBMD. The results of our intervention study add to some of the key questions in the pediatric exercise bone literature.  >  Does increased  activity during childhood  lead to increased  bone mineral  accretion?  Our study supports existing literature showing benefits of childhood physical activity on bone mineral accrual. The 1% greater increase in trochanteric aBMD in the children adding jumping activities to their physical education classes is encouraging given the short duration of the study. As well, the intervention was easily implemented, with only minor modifications to elementary physical education classes. This has public health implications in that programs of this nature can be implemented on a large scale.  >  What is the 'optimal' osteogenic  program?  The increase at the trochanter in the children in our study, although small, was seen with a minimal intervention. Tuck jumps required less than 1 minute in the classroom or gym 3 times per week. The games children performed within physical education class on 2 days per week were performed for 10-20 minutes by modifying existing games and activities. Studies quantifying the ground reaction forces of the activities in this intervention are underway and will provide important  93  information. Nonetheless, the results of our intervention study support animal data that minimal loading can have site-specific effects if activities are high in magnitude and unusual. Future studies are required to determine if changes in loading characteristics, duration, or frequency of the intervention alter the osteogenic response. Until such studies are done, it seems prudent to recommend adding activities of moderate-high impact loading to physical education curriculum.  >  Is there a critical period during childhood  or adolescence  when physical  activity builds the most  bone?  Children in our study were primarily prepubescent. The 1% increase over 8-months was significantly less than that seen in two previous intervention studies in slightly older children, and less than in young gymnasts. It is unclear if the discrepancy is due to differences in the intervention, or if greater benefits would be seen with the same intervention program introduced in an older age group. Longer-term, randomized intervention studies utilizing the same intervention in pre- and peripubescent children are needed to answer these questions.  7.2 Ethnic differences in response to an exercise intervention The trend for a greater response to exercise intervention in the Asian children in our study population is intriguing. Given the lower bone mass in the Asian fathers in our study and the increasing incidence of hip fracture in Asians, an intervention aimed at increasing bone mass in Asian children is warranted. Asian boys in our study also had lower femoral neck aBMD, lower calcium intake, and reported participating in less physical activity than Caucasian boys. These results raise a number of interesting questions. >  Is there an ethnic difference  in the response  >  If so, what factors explain such a difference  to a longer-term in the  (2 or more years)  intervention?  response?  A greater response to intervention in Asian children, should one be found in other populations, may be due to a number of factors including lower level of physical activity. A greater response to a loading intervention in children with lower levels of physical activity would be expected based on theories of mechanical loading, given that calcium intake is 'adequate'. Another possibility is that the  94  modeling threshold is lower in Asian children based on genetic differences or a lower aBMD at baseline. Further long-term studies are warranted to examine these interactions.  7.3 Other determinants of the bone mineral response to exercise The results of our cross-sectional study show sex and ethnic differences in physical activity, calcium intake, and bone mineral in early life. Interestingly, sex differences were primarily among Caucasian children and ethnic differences primarily in boys. These differences are interesting, but need to be supported in studies which follow children through pubertal growth. Interactions between exercise and other factors, particularly hormonal milieu, genetics, and nutrition, are still poorly understood. Our cross-sectional and family data illustrate ethnicity, gender, lean mass, physical activity, calcium intake, and familial resemblance (implying genetic influences) influence aBMD very early in life. The exact role of these factors in the response to mechanical loading will require meticulously designed longitudinal studies.  7.4 Other questions for future research In addition to the questions raised above, there are a number of important questions rising for future research on the bone response to exercise intervention.  >  Are the benefits  of childhood  or adolescent  exercise programs  maintained  into  adulthood?  Without longitudinal follow-up, the long-term benefit of any intervention is unknown. It is possible the aBMD in control subjects will "catch-up" once intervention ceases, or at some point over the growth process, as has been shown to occur with cessation of calcium intervention in this age group [194, 195]. The long-term (adult) benefits of childhood physical activity intervention has yet to be adequately studied.  >  Does exercise  intervention  during growth influence  bone geometry  or  structure?  Bone mineral density is just one of the many factors determining bone strength. In addition to answering the above questions as they relate to aBMD, future research should also be directed towards examining the effects of mechanical loading on other aspects of bone strength, particularly  95  bone size and geometry. Childhood and adolescence are clearly crucial periods during which bone strength is determined. It seems intuitive that the greatest potential for altering the geometric properties of bone with a loading intervention would be during the time of rapid bone modeling and growth. As new technologies, such as MRI, become more accessible, such studies will provide important information.  Definitive studies to answer these questions will require long-term intervention, follow-up, and a large population of children across maturity ranges, and may, therefore, not be feasible. However, with evolving technology, meticulous attention to study design, and strong interdisciplinary research teams, pieces of the puzzle continue to be identified and put into place. Together these pieces contribute important information regarding the role of physical activity on the young skeleton.  96  REFERENCES 1.  Bouxsein ML, Myers ER, Hayes W. Biomechanics of age-related fractures, in: Osteoporosis.  R  Marcus, D Feldman, and J Kelsey (ed). Academic Press: 1996:373-393. 2.  Cooper C, Melton LJI. Magnitude and impact of osteoporosis and fracture, in: Osteoporosis.  R  Marcus, D Feldman, and J Kelsey (ed). Academic Press: 1996:419-434. 3.  Kanis J, Pitt F. Epidemiology of osteoporosis: Bone. 1992 13:S7-S15.  4.  Lau EM. Epidemiology of osteoporosis in urbanized Asian populations.  Osteoporosis  Int.  1997  7 Suppl:S91-S95. 5.  Bailey DA. The Saskatchewan Pediatric Bone Mineral Accrual Study: Bone mineral acquisition during the growing years.  6.  Int J Sports Med.  Seeman E. Reduced bone density in women with fractures: Contribution of low peak bone density and rapid bone loss.  7.  Osteoporosis  Int.  1994 Suppl 1:19-25.  Bailey DA, Faulkner RA, McKay HA. Growth, physical activity, and bone mineral acquisiton. in: Exercise  8.  1997 18:S191-S194.  and Sport Sciences  Reviews.  JO Holloszy (ed). Williams & Wilkins: 1996:233-266.  Barr SI, McKay HA. Nutrition, exercise and bone status in youth.  Int J Sport Nutr.  1998 8:124-  142. 9.  Kannus P, Haapasalo H, Sankelo M, Sievanen H, Pasanen M, Heinonen A, Oja P, Vuori I. Effect of starting age of physical activity on bone mass in the dominant arm of tennis and squash players.  10.  Ann Intern Med.  1995 123:27-31.  Jee SS. Introduction to skeletal function: Structural and metabolic aspects, in: A Basic Primer in Orthopaedics.  Science  F Bronner and RV Worrell (ed). Williams & Wilkins: Baltimore. 1991:3-  34. 11.  Marks SC, Hermey DC. The structure and development of bone, in: Principles  of Bone  Biology.  JP Bilezikian, LC Raisz, and GA Rodan (ed). Academic Press: San Diego. 1996:3-14. 12.  Rico H, Gonzalez-Riola J, Revilla M, Villa LF, Gomez-Castresana F, Escribano J. Cortical versus trabecular bone mass: Influence of activity on both bone components. Calcif  Tissue Int.  1994 37:325-330. 13.  Young N, Formica C, SzmuklerG, Seeman E. Bone density at weight-bearing and nonweightbearing sites in ballet dancers: The effects of exercise, hypogonadism, and body weight. J Endocrinol  14.  Clin  Metab. 1994 78:449-454.  Martin RB, Burr DB. Structure, function, and adaptation of compact bone. New York: Raven Press 1989.  15.  Burr DB, Schaffler MB, Yang KH, Wu DD, Lukoschek M, Kandzari D, Sivaneri N, Blaha JD, Radin EL. The effects of altered strain environments on bone tissue kinetics.  Bone.  1989  10:215-221. 16.  Lanyon LE. Strain-related bone modeling and re-modeling.  17.  Duncan RL, Turner CH. Mechanotransduction and the functional response of bone to mechanical strain. Calcif Tissue Int. 1995 57:344-358.  97  Top Geriatr Rehabil.  1989 4:13-24.  18.  Frost HM. Defining osteopenias and osteoporoses: Another view with insights from a new paradigm.  19.  Bone.  1997 20:385-391.  Parfitt AM. The cellular basis of bone remodeling: The quantum concept re-examined in light of recent advances in the cell biology of bone.  20.  Calcif Tissue Int.  1984 S6:S37-S45.  Eriksen EF, Gundersen HJG, Melsen F, Mosekilde L. Reconstruction of the formative site in iliac trabecular bone in 20 normal individuals employing a kinetic model for matrix and mineral apposition.  21.  Metab Bone Dis Rel Res.  Frost HM. Why do marathon runners have less bone than weight lifters? A vital-biomechanical view and explanation.  22.  1984 5:243-252.  Bone.  1997 20:183-189.  Parfitt AM. The physiologic and clinical significance of bone histomorphometric data, in: Histomorphometry:  Techniques  and Interpretations.  Bone  R Recker (ed). CRC Press: Boca Raton.  1983:143-152. 23.  Einhorn T. Bone strength: The bottom line.  24.  Blake GM, Fogelman I. Technical principles of dual energy X-ray absorptiometry.  Calcif Tissue Int.  1992 51:333-339. Sem Nucl  Med. 1997 27:210-228. 25.  Tothill P, Avenell A, Reid DM. Precision and accuracy of measurements of whole-body bone mineral: Comparisons between Hologic, Lunar and Norland dual-energy X-ray absorptiometers. Brit  26.  J Radiol.  1994 67:1210-1217.  Ashman RB. Experimental techniques, in:  Bone Mechanics.  SC Cowin (ed). CRC: Boca  Raton, FL. 1989:75-95. 27.  Carter DR.Bouxsein ML, Marcus R. New approaches for interpreting projected bone densitometry data.  28.  J Bone Miner Res.  1992 7:137-145.  Cowell CT, Lu PW, Lloyd-Jones SA, Briody JN, Allen JR, Humphries IRJ, Reed E, Knight J, Howman-Giles R, Gaskin K. Volumetric bone mineral density: A potential role in paediatrics. Acta Paediatr  29.  Suppl.  1995 411:12-16.  Seeman E. From density to structure: Gowing up and old on the surfaces of bone.  J Bone  Miner Res. 1997 12:509-521.  30.  Cummings SR, Black DM, Nevitt MC, Browner W, Cauley J, Ensrud K, Genant HK, Palermo L, Scott J, Vogt TM. Bone density at various sites for prediction of hip fractures.  Lancet.  1993  341:72-75. 31.  Kleerekoper M. Osteoporosis and the primary care physician: Time to bone up. Ann  Intern  Med. 1995 123:466-467. 32.  Wasnich RD, Ross PD, Heilbrun LK, Vogel JM. Prediction of postmenopausal fracture risk with use of bone mineral measurements. Am  33.  J Obstet Gynecol.  1985 153:745-751.  Katzman DK, Bachrach LK, Carter DR, Marcus R. Clinical and anthropometric correlates of bone mineral acquisition in healthy adolescent girls. J Clin  Endocrinol  Metab.  1991 73:1332-  1339. 34.  Kroger H, Kotameini A, Vainio P, Alhava E. Bone densitometry of the spine and femur in children by dual energy x-ray absorptiometry.  98  Bone Miner.  1992 17:75-85.  35.  Sievanen H, Kannus P, Nieminen V, Heinonen A, Oja P, Vuori I. Estimation of various mechanical characteristics of human bones using dual energy x-ray absorptiometry: Methodology and precision. Bone. 1996 18:17S-27S.  36.  Ott SM, O'Hanlan M, Lipkin EW, Newell-Morris L. Evaluation of vertebral volumetric vs. areal bone mineral density during growth.  37.  Bone.  1997 20:553-556.  Prentice A, Parsons TJ, Cole TJ. Uncritical use of bone mineral density in absorptiometry may lead to size-related artifacts in the identification of bone mineral determinants.  Am J Clin Nutr.  1994 60:837-842. 38.  Bonjour JP, Theintz G, Buchs B, Slosman D, Rizzoli R. Critical years and stages of puberty for spinal and femoral bone mass accumulation during adolescence. J  Clin Endocrinol  Metab.  1991 73:555-563. 39.  Boot AM, DeRidder MAJ, Pols HAP, Krenning EP, DeMuinck Keizer-Schrama SMPF. Bone mineral density in children and adolescents: Relation to puberty, calcium intake, and physical activity.  40.  J Clin Endocrinol  Metab.  1997 82:57-62.  Theintz G, Buchs B, Rizzoli R, Slosman D, Clavien H, Sizonenko PC, Bonjour JP. Longitudinal monitoring of bone mass accumulation in healthy adolescents: Evidence for a marked reduction after 16 years of age at the levels of lumbar spine and femoral neck in female subjects. J  41.  Clin Endocrinol  Metab.  1992 75:1060-1065.  McKay HA, Bailey DA, Mirwald RL, Davison KS, Faulkner RA. Peak bone mineral accrual and age of menarche in adolescent girls: A six-year longitudinal study.  J Pediatr.  1998 133:682-  687. 42.  Bailey DA, McKay HA, Mirwald RL, Crocker PRE, Faulkner RA. The University of Saskatchewan Bone Mineral Accrual Study: A six year longitudinal study of the relationship of physical activity to bone mineral accrual in growing children. J Bone  43.  1999 in Press:  Martin AD, Bailey DB, McKay HA, Whiting S. Bone mineral and calcium accretion during puberty. Am  44.  Miner Res.  J Clin Nutr.  1997 66:611-615.  Slemenda CW, Reister TK, Hui SL, Miller JZ, Christian JC, Johnston CC. Influences on skeletal mineralization in children and adolescents: Evidence for varying effects of sexual maturation and physical activity. J  45.  Pediatr.  1994 125:201-207.  Fournier PE, Rizzoli R, Slosman DO, Theintz G, Bonjour JP. Asynchrony between the rates of standing height gain and bone mass accumulation during puberty.  Osteoporosis  Int.  1997  7:525-532. 46.  Parfitt AM. The two faces of growth: Benefits and risks to bone integrity.  Osteoporosis  Int. 1994  4:382-398. 47.  Bailey DA, Wedge JH, McCulloch RG, Martin AD, Benhardson SC. Epidemiology of fractures of the distal end of the radius in children as associated with growth. J 71A:125-130.  99  Bone Joint Surg.  1989  48.  Blimkie CJR, Lefevre J, Beunen GP, Renson R, Dequeker J, VanDamme P. Fractures, physical activity, and growth velocity in adolescent Belgian boys. Med  Sci Sports Exerc.  1993  25:801-808. 49.  Landin LA. Fracture patterns in children. Acta  50.  Goulding A, Cannan R, Williams SM, Gold EJ, Taylor RW, Lewis-Barned NJ. Bone mineral density in girls with forearm fractures.  51.  Orthop Scand.  J Bone Miner Res.  1983 54 (Suppl):5S-9S.  1998 13:143-148.  Geusens P, Cantatore F, Nijs J, Proesmans W, Emma F, Dequeker J. Heterogeneity of growth of bone in children at the spine, radius and total skeleton. Growth  Develop Aging.  1991 55:249-  256. 52.  Lu PW, Cowell CT, Lloyd-Jones SA, Brody JN, Howman-Giles R. Volumetric bone mineral density in normal subjects aged 5-27 years. J Clin  53.  Metab.  1996 81:1586-1590.  Gilsanz V, Boechat Ml, Roe TF, Loro ML, Sayre I, Goodman WG. Gender differences in vertebral body size in children and adolescents.  54.  Endocrinol  Radiol.  1993 190:673-677.  Gilsanz V, Gibbens DT, Roe TF, Carlson M, Senac MO, Boechat Ml, Huang HK, Schulz EE, Libanati CR, Cann CE. Vertebral bone density in children: Effect of puberty.  Radiol.  1988  166:847-850. 55.  Garn SM. The Earlier Gain and Later Loss of Cortical Bone. Springfield, IL: CC Thomas 1970.  56.  Garn SM, Rohmann CG, Wagner B. Bone loss as a general phenomenon in man. Feder  Proc.  1967 26:1729-1736. 57.  Garn SM, Wagner B. The adolescent growth of the skeletal mass and its implications to mineral requirements, in:  Adolescent  Nutrition and Growth.  FP Heald (ed). Appleton-Century-  Crofts: New York. 1969:139-161. 58.  Ohlsson C, Isgaard J, Tornell J, Nilsson A, Isaksson O, Lindahl A. Endocrine regulation of longitudinal bone growth.  59.  Acta Paediatr  Suppl.  1993 391:33-40.  Rice S, Blimkie CJ, Webber CE, Vevy D, Martin J, Parker D, Gordon CL. Correlates and determinants of bone mineral content and density in healthy adolescent girls. Pharmacol.  60.  1993 71:923-930.  Ross J, Cassorla F, Skerda M, Valk I, Loriaux L, Culter G. A preliminary study of the effect of estrogen dose on growth in Turner's syndrome.  61.  Can J Physiol  New Engl J Med.  1993 309:1104-1108.  Smith EP, Boyd J, Frank GR, Takahashi H, Cohen RM, Specker B, Williams TC, Lubahn DB, Korach KS. Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N Engl J Med.  62.  1994 331:1056-1061.  Turner R, Riggs B, Spelsberg T. Skeletal effects of estrogen.  Endocrine  Reviews.  1994  15:275-309. 63.  Prior JC. Progesterone as a bone-trophic hormone.  64.  Blumsohn A, Hannon RA, Wrate R, Barton J, al-Dehaimi AW, Colwell A, Eastell R. Biochemical markers of bone turnover in girls during puberty.  65.  Endocrine  Clin Endocrinol.  Reviews.  1990 11:386-398.  1994 40:663-670.  Duke PM, Lift IF, Gross RT. Adolescents' self-assessment of sexual maturation. 66:918-920.  100  Pediatr.  1980  66.  Matsudo SM, Matsudo VR. Validity of self-evaluation on determination of sexual maturational level, in:  World-Wide  Variation  in Physical  Fitness.  AL Claessens, J Lefevre, and B Vanden  Eynde (ed). Institute of Physical Education: Leuven. 1993:106-110. 67.  Tanner JM. Growth at adolescence. Oxford: Blackwell Scientific Pub 1955.  68. • Galileo G. Discorsi e dimostrazioni matematiche, intorno a due nuove scienze attinente alia meccanica e i movimenti locali (Translated; University of Wisconsin Press). Madison Wl: 1638. 69.  Wolff J. The law of bone remodeling. Berlin: Springer-Verlag 1892.  70.  Aarden EM, Nigweide PJ, Vanderplas A, Alblas MJ, Mackie EJ, Horton MA, Helfrich MH. Adhesive properties of isolated chick osteocytes in vitro. Bone. 1996 18:305-313.  71.  Jones DB, Leivseth G, Tenbosch J. Mechano-reception in osteoblast-like cells.  Biochem  Biol  Cell. 1995 73:525-534.  72.  Reich KM, Gay CV, Frangos JA. Fluid shear stress as a mediator of osteobalst cyclic adenosine monophosphate production. J Cell  73.  Physiol.  1990 143:100-104.  Harter LV, Hruska KA, Duncan RL. Human osteoblast-like cells respond to mechanical strain with increased bone matrix protein production independent of hormonal control.  Endocrinol.  1995 136:528-535. 74.  McKee MD, Nanci A. A osteopontin at mineralized tissue interfaces in bone, teeth, and osteointegrated implants: Ultrastructural distribution and implications for mineralized tissue formation, turnover, and repair.  75.  Microsc Res Tech.  1996 33:141-164.  Owan I, Burr DB, Turner CH, Qiu J, Tu Y, Onyia JE, Duncan RL. Mechanotransduction in bone: Osteoblasts are more responsive to fluid forces than mechanical strain. Am  J Physiol.  1997 273:C810-C815. 76.  Terai K, Takano-Yamamoto T, Ohba Y, Hiura K, Sugimoto M, Sato M, Kawahata H, Inaguma n, Kitamura Y, Nomura S. Role of osteopontin in bone remodeling caused by mechanical stress. J Bone  77.  Miner Res.  1999 14:839-849.  Lean JM, Jagger CJ, Chambers TJ, Chow JWM. Increased insulin-like growth factor I mRNA expression in rat osteocytes in response to mechanical stimulation. Am  J Physiol.  1995  268:E318-E327. 78.  McAllister TN, Frangos JA. Steady and transient fluid shear stress stimulate NO release in osteoblasts through distinct biochemical pathways.  79.  J Bone Miner Res.  1999 14:930-936.  Lanyon L. Osteocytes, strain detection, bone modeling and remodeling. Calcif  Tissue Int.  1993  53:S102-S107. 80.  Lanyon LE. Control of bone architecture by functional load bearing. J Bone  Miner Res.  1992  7:S369-S375. 81.  Lanyon LE. Using functional loading to influence bone mass and architecture: Objectives, mechanisms, and relationship with estrogen of the mechanically adaptive process in bone. Bone. 1996 18:37S-43S.  82.  Mullender MG, Huiskes R. Osteocytes and bone lining cells: Which are the best candidates for mechano-sensors in cancellous bone? Bone. 1997 20:527-532.  101  83.  Smith EL, Gilligan C. Dose-response relationship between physical lodaing and mechanical competence of bone. Bone. 1996 18:45S-50S.  84.  Frost HM. Bone "mass" and the "mechanostat": A proposal. Anatomical  85.  Rubin CT, Lanyon LE. Regulation of bone formation by applied dynamic loads. Surg Am.  86.  J Bone  Joint  1984 66:397-402.  Frost HM. The role of changes in mechanical usage set points in the pathogenesis of osteoporosis. J Bone  87.  1987 219:1-9.  Record.  Miner Res.  1992 7:253-261.  Forwood MR, Burr DB. Physical activity and bone mass: Exercises in futility? Bone  1993  Miner.  21:89-112. 88.  Rubin CT, Lanyon LE. Regulation of bone mass by mechanical strain magnitude.  Calcif  Tissue  Int. 1985 37:411-417. 89.  Turner CH. Homeostatic control of bone structure: An application of feedback theory.  Bone.  1991 12:203-217. 90.  Lanyon LE. Functional strain in bone tissue as an objective and controlling stimulus for adaptive bone remodeling.  91.  J Biomech.  1987 20:1083-1093.  Rubin CT. Skeletal strain and the functional significance of bone architecture.  Calcif Tissue Int.  1984 36:S11-S18. 92.  Turner CH, Owan I, Takano Y. Mechanotransduction in bone: Role of strain rate. Am  J Physiol.  1995 269:E438-E442. 93.  Mosley JR, Lanyon LE. Strain rate as a controlling influence on adaptive modeling in response to dynamic loading of the ulna in growing male rats. Bone. 1998 23:313-318.  94.  Frost HM. Obesity, and bone strength and "mass": A tutorial based on insights from a new paradigm.  95.  Bone.  1997 21:211-214.  Umemura Y, Ishiko T, Yamauchi T, Kurono M, Mashiko S. Five jumps per day increase bone mass and breaking force in rats. J Bone  96.  Miner Res.  Newhall KM, Rodnick KJ, van der Meulen MC, Carter DR, Marcus R. Effects of voluntary exercise on bone mineral content in rats. J Bone  97.  1997 12:1480-1485.  Miner Res.  1991 6:289-296.  van der Wiel HE, Lips P, Graafmans WC, Danielsen CC, Nauta J, van Lingen A, Mosekilde L. Additional weight-bearing during exercise is more important than duration of exercise for anabolic stimulus of bone: A study of running exercise in female rats.  98.  Bone.  1995 16:73-80.  Rubin CT, Bain SD, McLeod KJ. Suppression of the osteogenic response in the aging skeleton. Calcif Tissue Int. 1992 50:306-313.  99.  Turner CH, Takano Y, Owan I. Aging changes mechanical loading thresholds for bone formation in rats. J Bone  Miner Res.  1995 10:1544-1549.  100. Seeman E, Hopper JL. Genetic and environmental components of the population variance in bone density.  Osteoporos  Int.  1997 7:S10-S16.  101. Arden NK, Baker J, Hogg C, Baan K, Spector TD. The heritability of bone mineral density, ultrasound of the calcaneous and hip axis length: A study of postmenopausal twins. J Miner Res. 1996 11:530-534.  102  Bone  102. Arden NK, Spector TD. Genetic influences on muscle strength, lean body mass, and bone mineral density: A twin study. J Bone  1997 12:2076-2081.  Miner Res.  103. Kelly PJ, Nguyen T, Hopper J, Pocock N, Sambrook P, Eisman J. Changes in axial bone density with age: A twin study.  1993 8:11-17.  J Bone Miner Res.  104. Pocock NA, Eisman JA, Hopper JL, Yeates MG, Sambrook PN, Eberl S. Genetic determinants of bone mass in adults: A twin study. J Clin  Invest.  1987 80:706-710.  105. Slemenda CW, Christian JC, Williams CJ, Norton JA, Johnston CCJ. Genetic determinants of bone mass in adult women: A re-evaluation of the twin model and the potential importance of gene interaction on heritability estimates. J Bone  Miner Res.  1991 6:561-567.  106. Seeman E, Hopper JL, Young NR, Formica C, Goss P, Tsalamandris C. Do genetic factors explain associations between muscle strength, lean mass, and bone density? A twin study. Am J Physiol.  1996 270:E320-E327.  107. Kelly PJ, Hopper JL, Macaskill GT, Pocock NA, Sambrook PN, Eisman JA. Genetic factors in bone turnover. J Clin  Endocrinol  Metab.  1991 72:808-813.  108. Slemenda CW, Turner CH, Peacock M, Christian JC, Sorbel J, Hui SL, Johnston CC. The genetics of proximal femur geometry, distribution of bone mass and bone mineral density. Osteoporosis  Int. 1996 6:178-182.  109. Rogers J, Mahaney MC, Beamer WG, Donahue LR, Rosen CJ. Beyond one gene-one disease: Alternative strategies for deciphering genetic determinants of osteoporosis.  Calcif Tissue Int.  1997 60:225-228. 110. Flicker L, Hopper JL, Rodgers L, Kaymakci B, Green RM, Wark JD. Bone density determinants in elderly women: A twin study. J Bone  Miner Res.  1995 10:1607-1613.  111. Hopper JL, Green RM, Nowson CA, Young D, Sherwin JA, Kaymakci B, Larkins RG, Wark JD. Genetic, common environment, and individual specific components of variance for bone mineral density in 10- to 26-year-old females: A twin study. Am J Epidemiol.  1998 147:17-29.  112. Loesch DZ, Hopper JL, Rogucka E, Huggins RM. Timing and genetic rapport between growth in skeletal maturity and height around puberty: Similarities and differences between girls and boys. Am J Hum Genet. 1995 56:753-759. 113. Smith DM, Nance WE, Kang KW, Jonnston CCJ. Genetic factors in determining bone mass. J Clin Invest.  1973 52:2800-2808.  114. Jouanny P, Guillemin F, Kuntz C, Jeandel C, Pourel J. Environmental and genetic factors affecting bone mass. Similarity of bone density among members of healthy families. Rheum.  Arth  1995 38:61-67.  115. Krall EA, Dawson-Hughes B. Heritable and life-style determinants of bone mineral density. J Bone Miner Res.  1993 8:1-9.  116. Lutz J, Tesar R. Mother-daughter pairs: Spinal and femoral bone densities and dietary intakes. Am J Clin Nutr.  1990 52:872-877.  117. McKay HA, Bailey DA, Wilkinson AA, Houston CS. Familial comparison of bone mineral density at the proximal femur and lumbar spine.  103  Bone Miner.  1994 24:95-107.  118. Sowers MR, Burns TL, Wallace RB. Familial resemblance in bone mass in adult women. Gen Epidemiol.  1986 3:85-93.  119. Lutz J. Bone mineral, serum calcium, and dietary intakes of mother/daughter pairs. Am J  Clin  Nutr. 1986 44:99-106.  120. Tylavsky FA, Bortz AD, Hancock RL, Anderson JJ. Familial resemblance of radial bone mass between premenopausal mothers and their college-age daughters.  Calcif Tissue Int.  1989  45:265-272. 121. Danielson ME, Cauley JA, Baker CE, Newman AB, Dorman JS, Towers JD, Kuller LH. Familial resemblance of bone mineral density (BMD) and calcaneal ultrasound attenuation: The BMD in mothers and daughters study. J  Bone Miner Res.  1999 14:102-110.  122. Gueguen R, Jouanny P, Guillemin F, Kuntz C, Pourel J, Siest G. Segregation analysis and variance components analysis of bone mineral density in healthy families.  J Bone Miner Res.  1995 10:2017-2022. 123. Lonzer MD, Imrie R, Rogers D, Worley D, Licata A, Secic M. Effects of heredity, age, weight, puberty, activity, and calcium intake on bone mineral density in children.  Clinical  Pediatrics.  1996 35:185-189. 124. Francois S, Benmalek A, Guaydier-Souquieres G, Sabatier JP, Marcelli C. Heritability of bone mineral density.  Rev Rhum Engl Ed.  1999 66:146-151.  125. Ferrari S, Rizzoli R, Slosman D, Bonjour JP. Familial resemblance for bone mineral mass is expressed before puberty. J  Clin Endocrinol  Metab.  1998 83:358-361.  126. Krall EA, Dawson-Hughes B. Soft tissue body composition: Familial resemblance and independent influences on bone mineral density.  J Bone Miner Res.  127. Freenfield EM, Goldberg VM. Genetic determination of bone density.  1995 10:1944-1950. Lancet.  1997 350:1263-  1264. 128. Kelly PJ, Harris M. Genetic regulation of peak bone mass.  Acta Paediatr  Suppl.  1995 411:24-  29. 129. Ralston SH. The genetics of osteoporosis. QJ Med. 1997 90:247-251. 130. Morrison NA, Qi JC, Tokita A, Kelly PJ, Crofts L, Nguyen TV, Sambrook PN, Eisman JA. Prediction of bone density from vitamin D receptor alleles.  Nature.  1994 367:284-287.  131. Fleet JC, Harris SS, Wood RJ, Dawson-Hughes B. The Bsml vitamin D receptor restriction fragment length polymorphism (BB) predicts low bone density in premenopausal black and white women. J  Bone Miner Res.  1995 10:985-990.  132. Spector TD, Keen RW, Arden NK, Morrison NA, Major PJ, Nguyen TV, Kelly PJ, Baker JR, Sambrook PN, Lanchbury JS, Eisman JA. Influence of vitamin D receptor genotype on bone mineral density in postmenopausal women: A twin study in Britian. Br Med J. 1995 310:13571360. 133. Kroger H, Mahonen A, Ryhanen S, Turunen AM, Alhava E, Maenpaa P. Vitamin D receptor genotypes and bone mineral density.  Lancet.  104  1995 8:11-17.  134. Coleman G, Eccleshall TR, Feldman D. Vitamin D receptor gene alleles and osteoporosis, in: Principles  of Bone Biology.  JP Biezikian, LG Raisz, and GA Rodan (ed). Academic Press: San  Diego. 1996:917-933. 135. Peacock M. Vitamin D receptor gene alleles and osteoporosis: A contrasting view. Miner Res.  J Bone  1995 10:1294-1297.  136. Ferrari S, Rizzoli R, Theintz G, Slosman D, Bonjour J. Vitamin D-receptor (VDR) gene polymorphisms and the effect of calcium supplementation on bone mass accumulation in prepubertal girls. J Bone  Miner Res.  1995 10:S187.  137. Gunnes M, Berg JP, Halse J, Lehmann EH. Lack of relationship between vitamin D receptor genotype and forearm bone gain in healthy children, adolescents, and young adults. Endocrinol  J Clin  Metab. 1997 82:851-855.  138. Rizzoli R, Bonjour JP. Determinants of peak bone mass and mechanisms of bone loss. Osteoporos  Int. 1999 Suppl 2:S17-S23.  139. Rubin K, Schirduan V, Gendreau P, Sarfarazi M, Mendola R, Dalsky G. Predictors of axial and peripheral bone mineral density in healthy children and adolescents, with special attention to the role of puberty. J Pediatr. 1993 123:863-870. 140. Kroger H, Kotaniemi A, Kroger L, Alhava E. Development of bone mass and bone density of the spine and femoral neck: A prospective study of 65 children and adolescents. Bone  Miner.  1993 23:171-182. 141. Glastre C, Braillon P, David L, Cochat P, Meunier PJ, Delmas PD. Measurement of bone mineral content of the lumbar spine by dual energy x-ray absorptiometry in normal children: Correlations with growth parameters.  J Clin Endocrinol  Metab.  1990 70:1330-1333.  142. McCormick DP, Ponder SW, Fawcett HD, Palmer JL. Spinal bone mineral density in 335 normal and obese children and adolescents: Evidence for ethnic and sex differences.  J Bone  Miner Res. 1991 6:507-513.  143. Southard RN, Morris JD, Mahan JD, Hayes JR, Torch MA, Sommer A, Zipf WB. Bone mass in healthy children: Measurement with quantitative DXA. Radiol. 1991 179:735-8. 144. Zanchetta JR, Plotkin H, Alvarez-Filgueria ML. Bone mass in children: Normative values for the 2-20 year-old population. Bone. 1995 16:393S-399S. 145. Jones G, Dwyer T. Bone mass in prepubertal children: Gender differences and the role of sunlight exposure.  J Clin Endocrinol  Metab.  1998 83:4274-4279.  146. Nelson DA, Simpson PM, Johnson CC, Barondess DA, Kleerekoper M. The accumulation of whole body skeletal mass in third- and fourth-grade children: Effects of age, gender, ethnicity, and body composition. Bone. 1997 20:73-78. 147. Bhudhikanok GS, Wang M, Eckert K, Matkin C, Marcus R, Bachrach LK. Differences in bone mineral in young Asian and Caucasian Americans may reflect differences in bone size. J MinerRes.  1996 11:1545-1556.  148. Liu Z, Zhao YL, Ding GZ, Zhou Y. Epidemiology of primary osteoporosis in China. Osteoporosis  Int. 1997 7:S84-S87.  105  Bone  149. Yan LY, Zhou B. An epidemiological study of hip fracture in Shenyang. Osteoporosis.  Chinese J  1996 2:69-71.  150. Geusens P. Geometric characteristics of the proximal femur and hip fracture risk.  Osteoporosis  Int. 1996 Suppl 3:S27-S30. 151. Cumimings SR, Cauley JA, Palermo L, Ross PD, Wasnich RD, Black D, Faulkner KG. Racial differences in hip axis lengths might explain racial differences in rates of hip fracture. Osteoporosis  Int. 1994 4:226-229.  152. Faulkner KG, Cummings SR, Gluer CC, Palermo L, Black D, Genant HK. Simple measurement of femoral geometry predicts hip fracture: The study of osteoporotic fractures.  J Bone  Miner  Res. 1993 8:1211-1217. 153. Flicker L, Faulkner KG, Hopper JL, Green RM, Kaymakci B, Nowson CA, Young D, Wark JD. Determinants of hip axis length in women aged 10-89 years: A twin study. Bone. 1996 18:4145. 154. Wang MC, Aguirre M, Bhudhikanok GS, Kendall CG, Kirsch S, Marcus R, Bachrach LK. Bone mass and hip axis length in healthy Asian, Black, Hispanic, and White American youths. J Bone Miner Res.  1997 12:1922-1935.  155. Nomura A, Wasnich RD, Heilbrum LK, Ross PD, Davis JW. Comparison of bone mineral content between Japan-born and US-born Japanese subjects in Hawaii. Bone  Miner.  1989  6:213-223. 156. Hagiwara S, Miki T, Nishizawa Y, Ochi H, Onoyama Y, Morii H. Quantification of bone mineral content using dual-photon absorptiometry in a normal Japanese population.  J Bone Miner  Res.  1989 4:217-222. 157. Kin K, Kushida K, Yamazaki K, Okamoto S, Inoue T. Bone mineral density of the spine in normal Japanese subjects using dual-energy x-ray absorptiometry: Effect of obesity and menopausal status. Calcif Tissue Int. 1991 49:101-106.  158. Tsai K, Huang K, Chieng P, Su C. Bone mineral density of normal Chinese women in Taiwan. Calcif Tissue  Int. 1991 48:161-166.  159. Tsai KS, Cheng WC, Sanchez TV, Chen CK, Chieng PU, Yang RS. Bone densitometry of proximal femur in Chinese subjects: Gender differences in bone mass and bone areas.  Bone.  1997 20:365-369. 160. Ross PD, He YF, Yates AJ, Coupland C, Ravn P, McClung M, Thompson D, Wasnich RD. Body size accounts for most differences in bone density between Asian and Caucasian women. Calcif Tissue Int. 1996 59:339-343.  161. Russell-Aulet M, Wang J, Thornton J, Colt EWD, Pierson RN. Bone mineral density and mass by total-body dual-photon absorptiometry in normal White and Asian men.  J Bone Miner Res.  1991 8:575-582. 162. Russell-Aulet M, Wang J, Thornton J, Colt EWD, Pierson RN. Bone mineral density and mass in a cross-sectional study of White and Asian women. J Bone  106  Miner Res.  1993 8:575-582.  163. Marcus R, Greendale G, Blunt BA, Bush TL, Sherman S, Sherwin R, Wahner H, Wells B. Correlates of bone mineral density in the postmenopausal estrogen/progestin interventions trial. J Bone  Miner Res.  1994 9:1467-1476.  164. Cundy T, Cornish J, Evans MC, Gamble G, Stapleton J, Reid IR. Sources of interracial variation in bone mineral density.  J Bone Miner Res.  1995 10:368-373.  165. Barr SI. Dieting attitudes and behavior in urban high school students: Implications for calcium intake. J Adolesc  Health. 1995 16:458-464.  166. Ho S, Leung P, Swaminathan R, Chan C, Chan S, Fan Y, Lindsay R. Determinants of bone mass in Chinese women aged 21-40 years II. Pattern of dietary calcium intake and association with bone mineral density.  Osteoporosis  Int.  1994 4:167-175.  167. Ho SC, Wong E, Chan SG, Lau J, Chan C, Leung PC. Determinants of peak bone mass in Chinese women aged 21-40 years. III. Physical activity and bone mineral density. Miner Res.  J Bone  1997 12:1262-1271.  168. Fehily AM, Coles RJ, Evans WD, Elwood PC. Factors affecting bone density in young adults. Am J Clin Nutr.  1992 56:579-586.  169. Henderson NK, Price Rl, Cole JH, Gutteridge DH, Bhagat CI. Bone density in young women is associated with body weight and muscle strength but not dietary intakes.  J Bone Miner Res.  1995 10:384-393. 170. Rico H, Revilla M. Bone mass, body weight, and seasonal bone changes.  Calcif Tissue Int.  1994 54:523-524. 171. Figueroa-Colon R, Mayo MS, Treuth MS, Aldridge RA, Weinsier RL. Reproducibility of dualenergy X-ray absorptiometry measurements in prepubertal girls. Obes  Res.  1998 6:262-267.  172. Reid IR, Evans MC, Ames RW. Volumetric bone density of the lumbar spine is related to fat mass but not lean mass in normal postmenopausal women. Osteoporosis  Int.  1994 4:362-367.  173. Reid IR, Plank LD, Evans MC. Fat mass is an important determinant of whole body bone density in premenopausal women but not in men. J Clin  Endocrinol  Metab.  1992 75:779-782.  174. Bevier WC, Wiswell RA, Pyka G, Kozak KC, Newhall KM, Marcus R. Relationship of body composition, muscle strength, and aerobic capacity to bone mineral density in older men and women.  J Bone Miner Res.  1989 4:421-432.  175. Sowers'MF, Kshirsager A, Crutchfield MM, Updike S. Joint influence of fat and lean body composition compartments on femoral bone mineral density in premenopausal women. Am J Epidemiol.  1992 136:257-265.  176. Courteix D, Lespessailles E, Loiseau-Peres S, Obert P, Ferry B, Benhamou CL. Lean tissue mass is a better predictor of bone mineral content and density than body weight in prepubertal girls.  RevRhum  Engl Ed.  1998 65:328-336.  177. Morris FL, Naughton GA, Gibbs JL, Carlson JS, Wark JD. Prospective 10-month exercise intervention in pre-menarcheal girls: Positive effects on bone and lean mass.  J Bone  Miner  Res. 1997 12:1453-1462. 178. Matkovic V. Nutrition, genetics and skeletal development.  107  J Am Coll Nutr.  1996 15:556-569.  179. Malina RM, Bouchard C. Growth Maturation and Physical Activity. Champaign, IL: Human Kinetics Publishers 1991. 180. Nelson DA, Barondess DA. Whole body bone, fat and lean mass in children: Comparison of three ethnic groups.  Am J Phys Anthropol.  1997 103:157-162.  181. Taylor RW, Gold E, Manning P, Goulding A. Gender differences in body fat content are present well before puberty.  Int J Obes Relat Metab Disord.  1997 21:1082-1084.  182. Standing Committee on the Scientific Evaluation of Dietary Reference Intakes. Dietary Reference Intakes: Calcium, Phosphorous, Magnesium, Vitamin D, and Fluoride. Washington: National Academy Press 1997. 183. Johnston C, Miller J, Slemenda C, Reister T, Hui S, Christian J, Peacock M. Calcium supplementation and increases in bone mineral density in children. New  Engl J Med.  1992  327:82-87. 184. Lee WTK, Leung SSF, Wang SH, Xu YC, Zeng WP, Lau J, Oppenheimer SJ, Cheng JCY Double-blind, controlled calcium supplementation and bone mineral accretion in children accustomed to a low-calcium diet. Am  J Clin Nutr.  1994 60:744-750.  185. Matkovic V, Kostial K, Simonovic I, Buzina R, Brodarec A, Nordin BE. Bone status and fracture rates in two regions of Yugoslavia.  Am J Clin Nutr.  1979 32:540-549.  186. Murphy S, Khaw KT, May H, Compston JE. Milk consumption and bone mineral density in middle aged and elderly women. Br J Med. 1994 308:939-941. 187. Nieves JW, Golden AL, Siris E, Kelsey JL, Lindsay R. Teenage and current calcium intake are related to bone mineral density of the hip and forearm in women aged 30-39 years. Am J Epidemiol.  1995 141:342-351.  188. Sandler RB, Slemenda CW, LaPorte RE, Cauley JA, Schramm MM, Barresi ML, Kriska AM. Postmenopausal bone density and milk consumption in childhood and adolescence.  Am J Clin  Nutr. 1985 42:270-274.  189. Cadogan J, Eastell R, Jones N, Barker ME. Milk intake and bone mineral acquisition in adolescent girls: Randomised, controlled intervention trial. BMJ. 1997 15:1255-1260. 190. Lee WTK, Leung SSF, Leung DMY, Tsang HSY, Lau J, Cheng JCY. A randomized doubleblind controlled calcium supplementation trial and bone and height acquisition in children. British J Nutr. 1995 74:125-139.  191. Lloyd T, Martel JK, Rollings N, Andon MB, Kulin H, Demers LM, Eggli DF, Kieselhorst K, Chinchilli VM. The effect of calcium supplementation and Tanner stage on bone density, content and area in teenage women. Osteoporosis  Int.  1996 6:276-283.  192. Nowson CA, Green RM, Hopper JL, Sherwin AJ, Young D, Kaymakci B, Guest CS, Smid M, Larkins RG, Wark JD. A co-twin study of the effect of calcium supplementation on bone density during adolescence. Osteoporosis  Int. 1997 7:219-225.  193. Bonjour JP, Carrie AL, Ferrari S, Clavien H, Slosman D, Theintz G, Rizzoli R. Calciumenriched foods adn bone mass growth in prepubertal girls: A randomized, double-blind, placebo-controlled trial. J Clin  Invest.  1997 99:1287-1294.  108  194. Lee WTK, Leung SSF, Leung DMY, Cheng JCY. A follow-up study on the effects of calciumsupplement withdrawal and puberty on bone acquisition of children.  Am J Clin Nutr.  1996  64:71-77. 195. Lloyd T, Rollings N, Andon MB, Eggli DF, Mauger E, Chinchilli V. Enhanced bone gain in early adolescence due to calcium supplementation does not persist in late adolescence. J  Bone  Miner Res. 1996 11 :S154 (abstract).  196. Kanders B, Dempster DW, Lindsay R. Interaction of calcium nutrition and physical activity on bone mass in young women.  J Bone Miner Res.  1988 3:145-149.  197. Lanyon LE, Rubin CT, Baust G. Modulation of bone loss during calcium insufficiency by controlled dynamic loading.  Calcif Tissue Int.  1986 38:209-216.  198. Bloomfield S. Changes in musculoskeletal structure and function with prolonged bed rest. Med Sci Sports Exerc. 1997 29:197-206.  199. Specker B. Evidence for an interaction between calcium intake and physical activity on changes in bone mineral density.  J Bone Miner Res.  1996 11:1539-1544.  200. Gunnes M, Lehman EH. Physical activity and dietary constituents as predictors of forearm cortical and trabecular bone gain in healthy children and adolescents: A prospective study. Acta Paediatr.  1996 85:19-25.  201. Clemens TL, O'Riordan JLH. Vitamin D. in: Metabolism  (2nd edition).  Principles  and Practice  of Endocrinology  and  KL Becker (ed). JB Lippincott: Philadelphia. 1995:483-490.  202. Rosen CJ, Morrison A, Zhou H, Storm D, Hunter SJ, Musgrave K, Chen T, Wei W, Holick MF. Elderly women in northern New England exhibit seasonal changes in bone mineral density and calcitropic hormones.  Bone Miner.  1994 25:83-92.  203. Webb AR, Kline L, Holick MF. Influence of season and latitude on the cutaneous synthesis of vitamin D3: Exposure to winter sunlight in Boston and Edmonton will not promote vitamin D3 synthesis in human skin. J Clin  Endocrinol  Metab.  1988 67:373-378.  204. Mirwald R, Bailey D, McKay H, Drinkwater D, Faulkner R. Seasonal variation in height, sitting height, and leg length in boys and girls age 9 to 18 (abstract).  Med Sci Sports Exerc.  1997  29:S98. 205. Holick MF. Vitamin D requirements for humans of all ages: New increased requirements for women and men 50 years and older. Osteoporosis  Int. 1998 8:S24-S29.  206. Elffors I, Allander E, Kanis JA, Gullberg B, Johnell O, Dequeker J, Dilsen G, Gennari C, Lopes Vaz AA, Lyritis G, Mazzuoli GF, Miravet L, Passeri M, Perez Cano R, Rapado A, Ribot C. The variable incidence of hip fracture in Southern Europe: The MEDOS Study.  Osteoporosis  Int.  1994 4:253-263. 207. Johnell O, Gullberg B, Allander E, Kanis JA, The MEDOS Study Group. The apparent incidence of hip fracture in Europe: A study of national register sources.  Osteop  Int. 1992  2:298-302. 208. Kanis JA. The incidence of hip fracture in Europe.  109  Osteoporosis  Int Suppl.  1993 1 :S10-S15.  209. Faulkner R, Houston C, Bailey D, Drinkwater D, McKay H, Wilkinson A. Comparison of bone mineral content and bone mineral density between dominant and non-dominant limbs in children 8-16 years of age. Am J Hum Biol. 1993 5:491-499. 210. Ruiz JC, Mandel C, Garabedian M. Influence of spontaneous calcium intake and physical exercise on the vertebral and femoral bone mineral density of children and adolescents. J Bone Miner Res.  1995 10:675-682.  211. Slemenda CW, Millder JZ, Hue SL, Reister TK, Johnston CC. Role of physical activity in the development of skeletal mass in children. J Bone Miner Res. 1991 6:1227-1233. 212. McCulloch RG, Bailey DA, Houston CS, Dodd BL. Effects of physical activity, dietary calcium intake and selected lifestyle factors on bone density in young women. Can Med Assoc J. 1990 142:221-227. 213. Talmage RU, Anderson JJB. Bone density loss in women: Effects of childhood activity, exercise, calcium intake and estrogen therapy. Calcif Tissue Int. 1984 36:S52. 214. Tylavsky FA, Anderson JJ, Talmage RV, Taft TN. Are calcium intakes and physical activity patterns during adolescence related to radial bone mass of white college-age females? Osteoporosis  Int.  1992 2:232-40.  215. Valimaki MJ, Karkkainen M, Lamberg-Allardt C, Laitinen K, Alhava E, Heikkenen J, Impivaara O, Makela P, Palmgren J, Seppanen R, Vuori I, Cardiovascular Risk in Young Finns Study Group. Exercise, smoking, and calcium intake during adolescence and early adulthood as determinants of peak bone mass. BMJ. 1994 309:230-235. 216. Cheng JC, Leung SS, Lee WT, Lau JT, Maffulli N, Cheung AY, Chan KM. Determinants of axial and peripheral bone mass in Chinese adolescents. Arch Dis Child. 1998 78:524-530. 217. Cheng JC, Maffulli N, Leung SS, Lee WT, Lau JT, Chan KM. Axial and peripheral bone mineral acquisition: A 3-year longitudinal study in Chinese adolescents. Eur J Pediatr. 1999 158:506512. 218. Dyson K, Blimkie CJR, Davison KS, Webber CE, Adachi JD. Gymnastics training and bone density in pre-adolescent females. Med Sci Sports Exerc. 1997 29:443-450. 219. Grimston SK, Wilows ND, Hanley DA. Mechanical loading regime and its relationship to bone mineral density in children. Med Sci Sports Exerc. 1993 25:1203-1210. 220. Bass S, Pearce G, Bradney M, Hendrich E, Delmas PD, Hardy A, Seeman E. Exercise before puberty may confer residual benefits in bone density in adulthood: Studies in active prepubertal and retired female gymnasts. J Bone Miner Res. 1998 13:500-507. 221. Slemenda CW, Johnston CC. High intensity activities in young women: Site specific bone mass effects among female figure skaters. Bone Miner. 1993 20:125-132. 222. Nickols-Richardson SM, O'Connor PJ, Shapses SA, Lewis RD. Longitudinal bone mineral density changes in female child artistic gymnasts. J Bone Miner Res. 1999 14:994-1002. 223. Daly RM, Rich PA, Klein R, Bass S. Effects of high-impact exercise on ultrasonic and biochemical indices of skeletal status: A prospective study in young male gymnasts. J Bone Miner Res.  1999 14:1222-1230.  110  224. Njeh CG, Boivin CM, Langton CM. The role of ultrasound in the assessment of osteoporosis: A review. Osteoporosis  Int. 1997 7:7-22.  225. Kirchner E, Lewis R, O'Connor P. Effects of past gymnastics participation on adult bone mass. JAppI  1996 80:226-232.  Physiol.  226. Lindholm C, Hagenfeldt K, Ringertz H. Bone mineral content of young female former gymnasts. Acta Paediatr.  1995 84:1109-1112.  227. Khan KM, Bennell KL, Hopper JL, Nowson C, Sherwin AJ, Flicker L, Crichton KJ, Harcourt P, Wark JD. Self-reported ballet classes undertaken between 10 and 12 years of age and hip bone mineral density.  Osteoporosis  Int.  1998 8:165-173.  228. Kontulainen S, Kannus P, Haapasalo H, Heinonen A, Sievanen H, Oja P, Vuori I. Changes in bone mineral content with decreased training in competitive young adult tennis players and controls: A prospective 4-year follow-up. Med Sci Sports Exerc. 1999 31:646-652. 229. Haapasalo H, Kannus P, Sievanen H, Pasanen M, Uusi-Rasi K, Heinonen A, Oja P, Vuori I. Effect of long-term unilateral activity on bone mineral density of female junior tennis players. J Bone Miner Res.  1998 13:310-319.  230. Bradney M, Pearce G, Naughton G, Sullivan C, Bass S, Beck T, Carlson J, Seeman E. Moderate exercise during growth in prepubertal boys: Changes in bone mass, size, volumetric density, and bone strength: A controlled prospective study. J  Bone Miner Res.  1998 13:1814-  1821. 231. Burdett RG. Forces predicted at the ankle during running. Med Sci Sports  Exerc.  1992 14:308-  318. 232. Payne AH. A comparison of the ground forces in race walking with those in normal walking and running, in:  Biomechanics  Vl-A.  E Asmussen and K Jorgensen (ed). University Park Press:  Baltimore. 1978:293. 233. McNair PJ, Prapavessis H. Normative data of vertical ground reaction forces during landing from a jump. J  Sci Med Sport.  1999 2:86-88.  234. Michaud TJ, Rodriguez-Zayas J, Armstrong C, Hartnig M. Ground reaction forces in high impact and low impact aerobic dance.  J Sports Med Phys Fitness.  1993 33:359-366.  235. Clark PA, Rogol AD. Growth hormones and sex steroid interactions at puberty. Metab Clin N Am.  Endocrinol  1996 25:665-681.  236. Eisman JA, Sambrook PN, Kelly PJ, Pocock NA. Exercise and its interaction with genetic influences in the determination of bone mineral density. Am J Med. 1991 91:5S-9S. 237. Haapasalo H, Kannus P, Sievanen H, Heinonen A, Oja P, Vuori I. Long-term unilateral loading and bone mineral density and content in female squash players.  Calcif Tissue Int.  1994  54:249-255. 238. Khan KM, McKay HA, Haapasalo H, Bennell KL, Fo.rwood MR, Kannus P, Wark JD. Does childhood and adolescence provide a unique opportunity for exercise to strengthen the skeleton? J Sci Med Sport. 1999 submitted: 239. Elders M. Schools and health: A natural partnership. J Scholastic  111  Health. 1993 63:312-315.  240. Ferrari SL, Rizzoli R, Slosman DO, Bonjour JP. Do dietary calcium and age explain the controversy surrounding the relationship between bone mineral density and vitamin D receptor gene polymorphisms? J Bone  Miner Res.  1998 13:363-370.  241. Salamone LM, Glynn NW, Black DM, Ferrell RE, Palermo L, Epstein RS, Kuller LH, CauleyJA. Determinants of premenopausal bone mineral density: The interplay of genetic and lifestyle factors.  J Bone Mineral  Res.  1996 11:1557-1565.  242. Soroko SB, Barett-Connor E, Edelstein SL, Kritz-Silverstein D. Family history of osteoporosis and bone mineral density at the axial skeleton: The Rancho Bernardo study.  J Bone  Miner  Res. 1994 9:761-769. 243. Lau EMC. The epidemiology of hip fracture in Asia: An update.  Osteoporosis  Int.  1996 3:S19-  S23. 244. Kohrt WM. Preliminary evidence that DEXA provides an accurate assessment of body composition.  J Appl Physiol.  1998 84:372-377.  245. Mazess RB, Barden HS, Bisk JP, Hansen J. Dual-energy x-ray absorptiometry for total-body and regional bone-mineral and soft-tissue composition.  Am J Clin Nutr.  1990 51:1102-1112.  246. Cameron N. The Measurement of Human Growth. London: Croon-Helm 1984. 247. Hologic. Model QDR-4500 User's Guide. Waltham, MA: Hologic, Inc. 1996. 248. Pouilles JM, Tremollieres F, Todorovsky N, Ribot C. Precision and sensitivity of dual-energy Xray absorptiometry in spinal osteoporosis. J Bone  Miner Res.  1991 6:997-1002.  249. ACSM. American College of Sports Medicine Position Stand: Exercise and physical activity for older adults. Med Sci Sports Exerc. 1998 30:992-1008.  250. Barr SI. Associations of social and demographic variables with calcium intakes of high school students. J Am  Diet Assoc.  1994 94:260-266,269.  251. Rockett RH, Colditz GA. Assessing diets of children and adolescents.  Am J Clin Nutr.  1997  65:1116S-1122S. 252. Sallis JF, Condon SA, Goggin KJ, Roby JJ, Kolody B, Alcaraz JE. The development of selfadministered physical activity surveys for 4th grade students.  Res Q Exerc Sport.  1993 64:25-  31. 253. Crocker PRE, Bailey DA, Faulkner RA, Kowalski KC, McGrath R. Measuring general levels of physical activity: Preliminary evidence for the Physical Activity Questionnaire for older children.  Med Sci Sports Exerc.  1997 29:1344-1349.  254. Dequeker J, Ranstam J, Valsson J, Sigurgevisson B, Allander E. The Mediterranean Osteopororsis Questionnaire.  Clin Rheum.  1991 10:54-72.  255. Bachrach LK. Bone mineralization in childhood and adolescence.  CurrOpin  Pediatr.  1993  5:467-473. 256. Lee WTK, Leung SSF, Ng MY, Wang SH, Xu YC, Zeng WP, Lau J. Bone mineral content of two populations of Chinese children with different calcium intakes. 206.  112  Bone Miner.  1993 23:195-  257. Hansen MA, Hassager C, Jensen SB, Christiansen C. Is heritability a risk factor for postmenopausal osteoporosis? J Bone  Miner Res.  1992 7:1037-1043.  258. Wang Q, Ravn P, Wang S, Overgaard K, Hassager C, Christiansen C. Bone mineral density in immigrants from southern China to Denmark. A cross-sectional study.  EurJ  1996  Endocrinol.  134:163-167. 259. Stephens MB, Wentz SW. Supplemental fitness activities and fitness in urban elementary school classrooms. Fam Med. 1998 30:220-223. 260. Mautalen CA, Vega EM, Einhorn TA. Are the etiologies of cervical and trochanteric hip fractures different?  Bone.  1996 18:133S-137S.  261. Kerr D, Morton A, Dick I, Prince R. Exercise effects on bone mass in postmenopausal women are site-specific and load dependent. J Bone  Miner Res.  1996 11:218-225.  262. Lanyon LE. Functional strain as a determinant for bone remodeling.  Calcif Tissue Int.  1984  36:S56-61. 263. Bassey EJ, Littlewood JJ, Taylor SJG. Relations between compressive axial forces in an instrumented massive femoral implant, ground reaction forces, and integrated electromyographs from vastus lateralis during various 'osteogenic' exercises. J  Biomechanics.  1997 30:212-223. 264. Bobbert MF, Huijing PA, van Ingen Scheanau GJ. Drop jumping I: The influence of jumping technique on the biomechanics of jumping. Med Sci  Sports Exerc.  1984 4:332-338.  265. Carter DR, van der Meulen MCH, Beaupre GS. Skeletal development: Mechanical consequences of growth, aging, and disease, in:  Osteoporosis.  R Marcus, D Feldman, and J  Kelsey (ed). Academic Press: San Diego, CA. 1996:333-350. 266. Gordon CL, Hamilton JM, Atkinson SA, Webber CE. The contributions of growth and puberty to peak bone mass.  Growth, Development  and Aging.  1991 55:257-262.  267. Krabbe S, Christiansen C, Rodbro P, Transbol I. Effect of puberty on rates of bone growth and mineralisation: with observations in male delayed puberty. Arch  Dis Child.  1979 54:950-953.  268. Bassey EJ, Ramsdale SJ. Weight bearing exercise and ground reaction forces: A 12-month randomized control trial of effects on bone mineral density in healthy postmenopausal women. Bone. 1995 16:469-476.  269. Bassey EJ, Rothwell MC, Littlewood JJ, Pye DW. Pre- and postmenopausal women have different bone mineral density responses to the same high-impact exercise. J Bone  Miner  Res.  1998 13:1805-1813. 270. Leonard M, Feldman H, Zemel B, Berlin J, Barden E, Stallings V. Evaluation of low density spine software for the assessment of bone mineral density in children. J Bone  Miner res.  1998  13:1687-1690. 271. McKay HA, Petit MA, Bailey DA, Wallace WM, Schutz RW, Khan KM. Analysis of proximal femur DXA scans in growing children: Comparisons of different protocols for cross-sectional, 8month and 7-year longitudinal data. J Bone  Miner Res.  113  1999 In Press.  APPENDIX 1  Original Papers  Journal of Pediatrics In Press  Augmented trochanteric bone mineral density after modified physical education classes: A randomized school-based exercise intervention study in pre- and earlypubescent children  Heather A McKay, PhD , Moira A Petit, MSc , Robert W Schutz, PhD , 1  1  1  Jerilynn C Prior, MD , Susan I Barr, PhD , Karim M Khan, MD, PhD 2  3  1  Departments of Human Kinetics , Medicine, Division of Endocrinology and Metabolism , and 1  2  Human Nutrition , University of British Columbia, Vancouver, British Columbia, Canada 3  KEY WORDS: pediatric, physical activity, DXA, bone mineral, intervention Correspondence to: Dr. Heather McKay School of Human Kinetics Room 210,6081 University Blvd. University of British Columbia Vancouver, B.C. V6T 1Z1 tele: 604-822-3120 fax: 604-822-6842 email: mckayh@interchange.ubc.ca  i»5  ABSTRACT Of the few exercise intervention studies focusing on pediatric populations, none have confined the intervention to the scheduled physical education curriculum. Objective: To examine the effect of an 8-month school-based jumping program on the change in areal bone mineral density (aBMD, g/cm) of healthy 3rd and 4th grade children. Study Design: Ten elementary schools were 2  randomized to exercise (n = 63) and control groups (n = 81). Exercise groups did 10 tuck jumps three times weekly and incorporated jumping, hopping and skipping into twice weekly physical education classes. Control groups did regular physical education classes. At baseline, and after 8 months of intervention, we measured aBMD, lean and fat mass by dual energy x-ray absorptiometry (Hologic QDR-4500). Calcium intake, physical activity and maturity were estimated by questionnaire. Results: The exercise group showed significantly greater change in femoral trochanteric aBMD (4.4 vs. 3.2%; p < 0.05). There were no group differences at other sites. Results were similar after controlling for covariates (baseline aBMD change in height, change in lean, calcium, physical activity, sex and ethnicity) in hierarchical regression. Conclusions: An easily implemented school-based jumping intervention augments aBMD at the trochanteric region in the pre- and early-pubertal skeleton.  INTRODUCTION Osteopenia and fragility fractures in older people are functions of failure to attain sufficiently high bone mineral during the years of growth and accelerated bone demineralization during aging. However, it is becoming increasingly apparent that the antecedents for osteoporosis begin in childhood (1,2). Establishing an optimum level of bone mineral in childhood and adolescence is essential to offset the inevitable loss of bone in later life. Recent work in pre- and early-pubertal children suggest that these years are biologically optimal for exercise intervention and are crucial if genetically pre-determined peak bone mass is to be achieved. Most of the evidence comes from retrospective or cross-sectional studies of highly trained athletes (3-6), but very few children are abletoachieve this level of activity and athlete studies may be confounded by self-selection. Exercise, to prevent osteoporosis, should target individuals who are not participating, or individuals who are sedentary, or those who are participating in moderate, recreational activities. A childhood exercise program that can be easily implemented within the existing physical education (PE) curriculum, and by elementary school teachers, is a practical means of population-based osteoporosis prevention (7). One exercise intervention study in premenarcheal girls (8) and one in prepubescent boys (9) have shown beneficial effects upon bone. In both cases, the interventions required additional time, resources and personnel. These studies, and those evaluating exercise intervention in older groups, were performed in Caucasian children and primarily in girls. Asians (10) and males (11) are increasingly at risk for osteoporosis so studies should include these populations. Therefore, we modified an existing PE curriculum to include activities known to be osteogenic (jumping, hopping) that could be implemented by teachers, using existing resources. This study aimedtoexamine the effect of this 8-month intervention on change in areal bone mineral density in a pre- and early-pubertal cohort of Asian and Caucasian boys and girls.  MATERIALS AND METHODS Subjects  11=*  Information about the study was first presented to principals and teacher representatives in the Richmond School District Richmond is a suburb of Vancouver, Canada in which 34% of the population report either Mandarin or Cantonese as theirfirstlanguage. Eighteen Grade 3 and 4 teachers from 10 schools agreed to participate. Schools were stratified by student number per school as either large, medium, or small. Within each tier, schools were randomized to either exercise or control group. Presentations about the study were made in each of the classrooms and consent forms were sent home to be signed by children and their parents. Approximately 45-50% of the children returned signed consent forms with a total of 210 volunteers at baseline of whom 165 were Asian or Caucasian. Repeat measurements were obtained on 144 healthy (Asian (n = 49) or Caucasian (n = 95)) children who are represented in the present analysis. Several of the children (n = 9) moved during the school year. Other reasons for not returning for final measurement included being absent from school on the day of measurement (n = 5) and failure to return the school consent to travel form (n = 7). Children were classified based on their parents' place of birth. Asian children had both parents born in Hong Kong primarily (63%), and also in China, Japan, Taiwan, or Vietnam. The majority of Caucasian children and their parents were born in Canada (81%).  Exercise intervention We designed a 'loading' curriculum that was easily implemented by classroom teachers into routine PE classes. The control schools continued their regular PE program. The program began in November of 1997 and continued through to June, 1998. The intervention was performed 3 times per week; twice within the PE class and once in the classroom. Teachers chose an activity from a variety of games, circuit training or dances which were described in detail in the curriculum manual. Activities required from 10-30 minutes, included a minimum of 10 minutes of loading and were consistent with the five movement categories mandated in the Canadian Integrated Resources Package for Physical Education (dance,  118  gymnastics, individual and dual activities, alternate environment activities, and games). For example, common games like "tag" which normally required running, were modified so that children became kangaroos or frogs and bounded in pursuit of classmates. To ensure a baseline amount of loading, children also performed 10 tuck jumps at the beginning of each PE class and once weekly in the classroom. Children were instructed to jump using both legs together and to grab their knees, bringing them as close to the chest as possible. Children rested for 1-second between jumps to maintain the quality of each jump. The program progressed based on level of ability, and as children adapted they were able to spend more time jumping within each activity. New activities such as bench jumping and circuit training were added as program options after 3 months of intervention. Teachers were instructed on the delivery of the curriculum at two training sessions offered twice during the year. To measure classroom compliance, teachers maintained logs in which they recorded the activities they selected and the approximate time spent on each activity. Research assistants observed PE classes three times during the year to assess the implementation of the program by teachers and the childrens' technique and ability.  Measurements The order in which schools were measured at baseline was randomized regardless of group status (exercise or control). This order was replicated at follow-up to standardize the duration between measurements. Bone mineral. Bone densitometry scans were acquired by the same registered technologist on a Hologic QDR-4500 bone densitometer (Hologic Inc., Waltham, MA), in array mode. Bone area (cm ), bone mineral content (BMC, g) and areal bone mineral density (aBMD, g/cm ) were 2  2  measured for the total body (TB), lumbar spine (LS, L1-L4), and total proximal femur (PF). Analysis of the PF included the regions of the femoral neck (FN) and trochanter (Troch). Standardized positioning protocol for each site (12) was used and the system phantom was  Mi  scanned daily to assure quality control. DXA measurement of BMC and aBMD is extremely precise with a coefficient of variation of less than 1% at all sites (12). Scan Analysis.  Hip and spine scans were analyzed by the same researcher to ensure  consistency. To reduce the system error introduced by varying the size of the global and subregions of interest (ROI), we maintained the same ROI for time 1 and time 2 within children. The femoral neck box size was standardized at the system default size of 1.5 cm (15 pixels) by 4.1 cm (12). Software version 8.20a was utilized for all analyses. Anthropometry. Absolute and percent fat and bone mineral free lean mass (BMFL) were estimated from DXA total body scans (13). Height (without shoes) and sitting height were recorded to the nearest 0.1 cm using a stadiometer. Weight was measured using an electronic scale to the nearest 0.1 kg and calf girth of the right leg was measured to the nearest mm. Measurements were taken twice and the average was used for analyses. A vertical jump test was performed using standard procedures (14) during thefirstand last months of the intervention. Calcium and Physical Activity. Dietary calcium intake and physical  activity scores were  averaged from 3 questionnaires administered at baseline andfinalmeasurement and during a midintervention school visit Fight of the children were absent from school during collection of the second questionnaire and thus had only two assessments. For calcium, a food frequency questionnaire (FFQ) was utilized which has previously been validated against food records as a tool for assessing calcium intake in Asian and Caucasian adolescents (15). The 7-day physical activity recall questionnaire asked the children the number of times they participated in a variety of primarily weight-bearing activities during recess, lunch and after school and specifically about their participation in extracurricular organized sports. The activity record is an adaptation of tools used by Slemenda et al. (16) and Sallis (17) and has recently been externally and internally validated as described elsewhere (18). Health History.  Parents were asked to complete a health history questionnaire for their  children. Children were included in the study if they were free of conditions, and were not using medications, known to affect bone mineral. All of the children met these inclusion criteria.  1X0  Maturity. Pubertal development was evaluated by self-assessment of breast (girls) and pubic hair (girls and boys), using a standard approach (19) that has been documented previously (1). Self-assessment of maturity is strongly correlated with staging assigned by an endocrinologist (20,21) and is practical for use in children in a school-based setting.  Statistical Methods Change and percent change from baseline were calculated for all outcome variables and expressed as mean ± SE. Data were checked for outliers. Dependent variables were checked for normal distribution and the regression assumptions checked by residual plots. A 2x2x2 Analysis of Variance (ANOVA) was utilized to test the main effects and interactions of group (exercise, control), sex (M, F), and ethnicity (Asian, Caucasian) for both baseline and change variables. Group differences for change in aBMD were also tested with hierarchical regression to control for potential confounding factors. Independent variables were entered in 8 steps including: baseline aBMD, change in height, change in BMFL mass, average physical activity, average calcium intake, sex, ethnicity, and exercise group. All data were analyzed using SPSS for Windows, Version 8.0 (SPSS Inc., Chicago). Results were considered significant if p < 0.05.  RESULTS Maturity Children ranged in age from 6.9-10.2 years at baseline with no between-group differences. All of the boys remained Tanner stage 1 throughout the study. Of the 74 girls, 66 were Tanner breast stage 1 at baseline and 8 girls were Tanner breast stage 2. Fifteen girls in the control group and 13 girls in the exercise group advanced from breast stage 1 to stage 2 during the study. Tanner stage was not related to change or percent change in body composition, anthropometric or aBMD variables and there were no significant differences in these variables between these early Tanner 2 and Tanner 1 girls.  JZl  Teachers' logs indicated all intervention schools completed a minimum of 10 tuck jumps 3 times per week throughout the intervention period. Teachers also reported including a minimum of 10 and a maximum of 30 minutes of activities from the study curriculum within the PE classes 2 times per week.  Baseline Characteristics Ethnic and sex differences at baseline have been reported elsewhere (22,23). There were no significant differences between the Asian and Caucasian children in height, weight, fat mass or BMFL mass. Asian children reported lower calcium intake and less physical activity, while Caucasian children had higher (8%) aBMD at the femoral neck site only (22,23). Analysis of variance showed no significant baseline differences between groups in any of the anthropometric variables including calf girth, weight, height, sitting height, leg length, percent fat or BMFL mass (Table 1). Similarly, initial aBMD did not differ between groups at any site (Table 1). Calcium intake was also similar between groups (control = 988 ± 68, exercise = 918 ± 51 mg/d; p = 0.456) and did not change over the course of the study for either group. Children from control schools reported significantly greater physical activity (control = 86.0 ± 1.7, exercise = 80.5 ± 1.5; p = 0.002) and performed better on the vertical jump at baseline. Forty percent of the children in both groups reported participation in extracurricular sports. There were no differences in baseline or change data based on participation or non-participation in organized sport.  8-Month Change  Asian children had a greater increase in TB aBMD (Asian = 1.8%, Caucasian = 1.1%, p = 0.015) than Caucasian children. Other change variables were not significantly different between ethnic groups. Children from both groups reported decreased activity at final measurement although this change was consistent between groups. The control group remained significantly more active than the exercise group (Control 81.7 ± 1.9; Exercise 74.7 ± 2.1) at final measurement. Anthropometry and bone data for the entire sample (Asian plus Caucasian) at  follow-up are provided (Table 1). Absolute and percent change in calf girth, weight, fat mass, and BMFL were similar between groups. The control group had a significantly greater (+1.5 cm) increase in height This difference was the result of a greater increase in sitting height (p = 0.028) with no difference in leg length change (p = 0.298). Vertical jump increased by 4.9±0.2 % in the control group and 6.5 ± 0.5% in the exercise group (p = 0.003). This difference was greater for the girls (control 5.0%, exercise 7.2%) than the boys (control 4.7%, exercise 5.5%) with a significant interaction effect (p = 0.043). For bone variables, 8 month change in aBMD was significantly greater than zero in both exercise and control groups and at all sites (p < 0.05). ANOVA showed no group differences in change in TB, LS, PF or FN aBMD. There was a significant effect of the intervention at the trochanter with the exercise group showing an average gain of 1.2% more aBMD than controls (p = 0.030). Figure 1 depicts absolute values for change in aBMD for the exercise and control groups by ethnicity. Although the difference for change between exercise and control groups tended to be much higher in the Asian (2.3%) compared to the Caucasian children (0.9%), the group by ethnicity interaction effect was not significant (p = 0.146). Exercise versus sex and/or ethnicity interactions were also not significant at other sites. After controlling for potential confounds within regression, the group effect remained significant at the trochanter. The final model (r = 0.37) explained approximately 14% of the variance (adjusted R = 0.09) for change in trochanter aBMD. Group (exercise, control) 2  contributed 3%, initial aBMD explained 4%, and change in BMFL mass explained 5% of the variance in the model. The remaining 2% is attributed to the non-significant variables within the model (change in height, sex, ethnicity, calcium and physical activity). Final regression models for change in TB, LS, PF and FN aBMD were not significant.  DISCUSSION Childhood is a critical period when bone may be particularly responsive to weight bearing physical activity (24). We intervened with a modified school-based physical education curriculum  to moderately increase skeletal loading and found a significant increase in areal bone mineral density (aBMD) at the trochanteric region of the hip. The exercise intervention did not affect 8month change in aBMD at other skeletal sites prior to, or after, controlling for initial status and differences in growth. Although the differences between exercise and control groups for absolute trochanteric aBMD change were more than twice as high in the Asian (2.3%) compared to the Caucasian children (0.9%) the randomized nature of the design resulted in only a small number of Asian children in the intervention group (n=16) and this difference did not achieve significance (p=0.15) (Figure 1). The aim of this research wastoidentify a practical method to optimize peak bone mass so as to minimize the risk of osteoporotic fracture in later life. To our knowledge, this is the first intervention study that has targeted a large number of pre- and early- pubertal children participating in existing physical education programs. Previous exercise intervention studies in children have had far fewer subjects and programs were offered outside of regular school time by professional instructors (8,9,25). The additional expense and time such programs demand make them less practical, less economical and, thus, less likelytobe implemented as a public health initiative. The program designed for the present study was successfully implemented in 9 classes by teachers who were trained in the curriculum but, with the exception of one teacher, had no formal training in physical education. As the program conformedtothe mandated Instructional Resource Package for elementary school PE, all of the children in the intervention classes (approximately 200 children in grades 3 and 4) participated in the program regardless of whether or not they volunteered for bone mineral assessment Further, although jump ropes were provided, activities could be completed without schools purchasing additional equipment. Therefore, our study suggests that a large-scale school-based intervention for bone health is eminently feasible. Costly and debilitating hip fractures commonly occur at trochanteric or cervical (femoral neck and intracapsular) sites. The etiology of hip fractures differs by site and trochanteric aBMD appears to be the best predictor of fracture (26). An increase in trochanteric aBMD with exercise has not been previously reported in children or adolescents. However, the osteogenic effect of  12.4  jumping on the highly trabecular trochanter is biologically plausible and is consistent with current theories and research evidence concerning mechanical loading. Exercise has a site-specific effect on bone and bone responds to mechanical strains primarily from forces generated by muscle attachments and also ground reaction forces (GRFs) (27,28). The large muscles that are utilized during jumping (i.e. gluteus maximus, gluteus medius) (29,30) attach at the greater trochanter. Also, the majority of forces placed on the proximal femur during loading are absorbed by the trochanter (31). Our results differ somewhat from the two previous exercise intervention studies on bone in children of similar age, with respecttothe magnitude of effect (8) and number of sites affected (8, 9). In a non-randomized study, Morris et al. reported greater increases in aBMD for all measured sites in premenarcheal girls. The largest difference was at the femoral neck with a 10% greater increase in aBMD for the exercise group (8). A similar intervention program showed much smaller effects for the legs (3%), total body (1%) and lumbar spine (2.5%) in prepubescent boys (9). Explanations for these observed differences are discussed below. Maturational Characteristics. Asrateof maturation explains most of the gain in bone mineral during growth (36,37) differences in maturational development between groups represents a large potential confound in any study of growing children. In studies with small numbers of subjects in this age group, a difference in rate of maturation in even a few children between control or exercise groups can significantly influence results. A strength of the present study was, therefore, the relatively large number of children at a similar stage of maturation at baseline. There were no differences in either the change, or absolute, values for bone mineral, height or weight between girls who remained in Tanner stage 1 and those girls who were Tanner 2 at final measurement. The large exercise intervention effect in the only other prospective exercise intervention study of premenarcheal girls (8) has been attributed to differences in maturity status between the groups (9). As evidence of this, the 10% greater increase in femoral neck aBMD in the exercise group was no longer significant when the effects of change in weight and height were controlled for between groups (8).  US  In our study, there was a greater increase in height in the control, versus the exercise, group. This could potentially bias the outcome away fromfindingan effect of the intervention. By including initial status, change in height and change in lean mass in the regression models, we statistically accounted for any association between accelerated growth and change in bone mineral density. Pediatric exercise intervention studies may also differ in outcomes because of baseline maturational differences between cohorts. In a recent cross-sectional study of tennis players, Haapasalo et al. reported that side-to-side differences in humerus aBMD were not significant until Tanner stage 3 (12.6 y) (32). Other cross-sectional and retrospective studies in both humans and animals, despite methodological limitations, suggest there may be an age-related effect of mechanical loading (3,5,6,33,34). More than 50% of the Morris et al. subjects were Tanner breast stage 2 at entry as compared to our 89% Tanner stage 1 girls at baseline (8). A greater osteogenic response may occur with an identical intervention in a slightly more mature (l-2y) age group as suggested in the Haapasalo study (32). Intervention Characteristics. The exact load required to elicit an osteogenic response has not been clearly defined for young children. Animal studies suggest that the most osteogenic activities are of high magnitude and unusual distribution (35) but are not necessarily high in number (36). To our knowledge, quantified loads imposed from jumping have not been published for this age group. In a pilot study of a small number of prepubescent children in our biomechanics lab, the various jumps utilized within the intervention elicited ground reaction forces (GRFs) 3-5 times body weight. In older women, landing from an approximately 8 cm jump, induced GRFs 34 times body weight and joint reaction forces 3-5 times those forces at the hip (29,37). These moderate forces induced a 2-3% increase in trochanteric aBMD after 5 months of jumping in premenopausal women (38). However, as children on average have relatively less muscle mass per unit body weight than adults, tensile forces at the trochanter could be lower than those experienced in adult studies. It is probable that the present intervention was of high enough impact to stimulate a (re)modeling effect at the direct site of muscle attachment (trochanter) but not at other regions of  \1\o  the proximal femur. As Bradney et al.(9) did not report bone outcomes for the regions of the proximal femur, comparisons at this site between studies for boys is difficult However a more diverse program of sport intervention might explain their observed positive change for the total body and lumbar spine. Other aspects of the intervention programs including progression, volume, duration and frequency would also influence the physiological response to the intervention. Conclusion. Our randomized prospective study demonstrated that an exercise program  implemented within elementary school physical education, increased aBMD at the trochanteric region of the hip in pre- and early- pubescent children. These outcomes have implications for fracture risk in later life. Cummings et al. (39) suggest that for every one standard deviation increase in BMD there is a seven times lower risk for hip fracture in older adults. Until such time that long term prospective studies address the question of whether benefits from childhood exercise persist into adulthood, pervasive public health care policies and programs that promote physical activity during the growing years seem well-advised.  I  21-  REFERENCES 1. Bailey DA. The Saskatchewan Pediatric Bone Mineral Accrual Study: bone mineral acquisition during the growing years. Int J Sports Med 1997; 18:S 191-4. 2. Seeman E. Reduced bone density in women with fractures: contribution of low peak bone density and rapid bone loss. Osteoporosis Int 1994; 1:19-25. 3.  Bass S, Pearce G, Bradney M, Hendrich E, De.mas PD, Hardy A, et al. Exercise before puberty may confer residual benefits in bone density in adulthood: studies in active prepubertal and retired female gymnasts. J Bone Miner Res 1998; 13:500-7.  4.  Dyson K, Blimkie CJR, Davison KS, Webber CE, Adachi JD. Gymnastics training and bone density in pre-adolescent females. Med Sci Sports Exerc 1997; 29:443-50.  5. Kannus P, Haapasalo H, Sankelo M, Sievanen H, Pasanen M, Heinonen A, et al. Effect of starting age of physical activity on bone mass in the dominant arm of tennis and squash players. Ann Intern Med 1995; 123:27-31. 6. Khan KM, Bennell KL, Hopper JL, Npwson C, Sherwin AJ, Flicker L, et al. Self-reported ballet classes undertaken between 10 and 12 years of age and hip bone mineral density. Osteoporosis Int 1998; 8:165-73. 7. Elders M. Schools and health: a natural partnership. J Scholastic Health 1993; 63:312-5. 8. Morris FL, Naughton GA, Gibbs JL, Carlson JS, Wark JD. Prospective 10-month exercise intervention in pre-menarcheal girls: positive effects on bone and lean mass. J Bone Miner Res 1997; 12:1453-62. 9. Bradney M, Pearce G, Naughton G, Sullivan C, Bass S, Beck T, et al. Moderate exercise during growth in prepubertal boys: changes in bone mass, size, volumetric density, and bone strength: a controlled prospective study. J Bone Miner Res 1998; 13:1814-21. 10. Lau EMC. The epidemiology of hip fracture in Asia: An update. Osteoporosis Int 1996; 3:S 19-23. 11. Grisso JA, Kelsey JL, O'Brien LA, Miles CG, Sidney S, Maislin G, et al. Risk factors for hip fracture in men. Hip Fracture Study Group. Am J Epidemiol 1997;145:786-793.  nh  12. Hologic. Model QDR-4500 User's Guide. Waltham: Hologic, Inc.; 1996. 13. Mazess RB, Barden HS, Bisk JP, Hansen J. Dual-energy x-ray absorptiometry for total-body and regional bone-mineral and soft-tissue composition. Am J Clin Nutr 1990;51:1102-1112. 14. The American College of Sports Medicine. ACSM Guidelines for Exercise Testing and Prescription. Baltimore: Williams & Wilkins; 1995. 15. Barr SI. Associations of social and demographic variables with calcium intakes of high school students. J Am Diet Assoc 1994; 94:260-6,269. 16. Slemenda CW, Millder JZ, Hue SL, Reister TK, Johnston CC. Role of physical activity in the development of skeletal mass in children. J Bone Miner Res 1991; 6:1227-33. 17. Sallis JF, Condon SA, Goggin KJ, Roby JJ, Kolody B, Alcaraz JE. The development of selfadministered physical activity surveys for 4th grade students. Res Q Exerc Sport 1993; 64:2531. 18. Crocker P, Bailey D, Faulkner R, Kowalski K, McGrath R. Measuring general levels of physical activity: preliminary evidence for the Physical Activity Questionnaire for older children. Med Sci Sports Exerc 1997; 29:1344-9. 19. Tanner JM. Growth at adolescence. Oxford: Blackwell Scientific Pub; 1955. 20. Duke PM, Litt IF, Gross RT. Adolescents' self-assessment of sexual maturation. Pediatrics 1980; 66:918-20. 21. Matsudo SM, Matsudo VR. Validity of self-evaluation on determination of sexual maturational level, in: World-Wide Variation in Physical Fitness. Claessens AL, Lefevre J, Vanden Eynde B, editors. Leuven: Institute of Physical Education; 1993. p. 106-110. 22. McKay HA, Petit MA, Khan KM, Schutz RW. Lifestyle determinants of bone mineral: A comparison between prepubertal Asian- and Caucasian-Canadian children (abstract). J Bone Miner Res 1998; 23:S493. 23. Petit MA, McKay HA, Khan KM, Schutz RW. Gender comparisons of bone mineral in prepubertal Asian and Caucasian children (abstract). J Bone Miner Res 1998; 23:S475.  24. Bailey DA, McKay HA, Mirwald RL, Crocker RPE, Faulkner RA. The University of Saskatchewan Bone Mineral Accrual Study: A six year longitudinal study of the relationship of physical activity to bone mineral accrual in growing children. J Bone Miner Res 1999; in Press. 25. Stephens MB, Wentz SW. Supplemental fitness activities andfitnessin urban elementary school classrooms. Fam Med 1998; 30:220-3. 26. Mautalen CA, Vega EM, Einhom TA. Are the etiologies of cervical and trochanteric hip fractures different? Bone 1996; 18:S 133-7. 27. Kerr D, Morton A, Dick I, Prince R. Exercise effects on bone mass in postmenopausal women are site-specific and load dependent J Bone Miner Res 1996; 11:218-25. 28. Lanyon LE. Functional strain as a determinant for bone remodeling. Calcif Tissue Int 1984; 36:S56-61. 29. Bassey EJ, Littlewood JJ, Taylor SJG. Relations between compressive axial forces in an instrumented massive femoral implant, ground reaction forces, and integrated electromyographs from vastus lateralis during various 'osteogenic' exercises. J Biomechanics 1997; 30:212-23. 30. Bobbert MF, Huijing PA, van Ingen Scheanau GJ. Drop jumping I: The influence of jumping technique on the biomechanics of jumping. Med Sci Sports Exerc 1984; 4:332-8. 31. Carter DR, van der Meulen MCH, Beaupre GS. Skeletal development: Mechanical consequences of growth, aging, and disease, in: Osteoporosis. Marcus R, Feldman D, Kelsey J, editors. San Diego: Academic Press; 1996. p. 333-50. 32. Haapasalo H, Kannus P, Sievanen H, Pasanen M, Uusi-Rasi K, Heinonen A, et al. Effect of long-term unilateral activity on bone mineral density of female junior tennis players. J Bone Miner Res 1998; 13:310-9. 33. Forwood MR, Burr DB. Physical activity and bone mass: exercises in futility? Bone Miner 1993;21:89-112.  130  34. Turner CH, Takano Y, Owan I. Aging changes mechanical loading thresholds for bone formation in rats. J Bone Miner Res 1995; 10:1544-9. 35. Lanyon LE. Using functional loading to influence bone mass and architecture: objectives, mechanisms, and relationship with estrogen of the mechanically adaptive process in bone. Bone 1996; 1&37S-43S. 36. Umemura Y, Ishiko T, Yamauchi T, Kurono M, Mashiko S. Five jumps per day increase bone mass and breaking force in rats. J Bone Miner Res 1997; 12:1480-5. 37. Bassey EJ, Ramsdale SJ. Weight bearing exercise and ground reaction forces: a 12-month randomized control trial of effects on bone mineral density in healthy postmenopausal women. Bone 1995; 16:469-76. 38. Bassey EJ, Rothwell MC, Littlewood JJ, Pye DW. Pre- and postmenopausal women have different bone mineral density responses to the same high-impact exercise. J Bone Miner Res 1998;13:1805-13. 39. Cummings SR, Black DM, Nevitt MC, Browner W, Cauley J, Ensrud K, et al. Bone density at various sites for prediction of hip fractures. Lancet 1993341:72-75.  1*1  OJ  1 g  5  1  <js  e i o  co © -H ON  00  CN  •*t ©  o 41  -H 00 VO  co  CN O 41 co O  CN © 41 co o  CN © 41  Tt  Tt  Tt  o 41  Tt  ON  © 41 vq co  o 41 CN  o  ON  VO  © 41 © 00 CN  H *«3 <U  [T.  w >cu  ON  41 ON  CN CO  41 VO VO CO  © 41 CN CN  o o o 4) oo o 00 ©  oo o 41 CN  W  VO  u  cn o  c  CN  ffl  CN  ON  41 00 o CO  41  © 41  co  CN CN  o o  00 © 41  ©  41  ON  l-s-  Tt  ON  t"-  o  1.  CN o 41  «o o 4)  1 & CN  VO  VO VO  41 vo CN  CO  00 © 41  00 © 41  o  00 II — - «  o  E  fl o  O -H p oo CN  «o  CN CO  ©  Tt  co  CN O 41 co  ©  1  CN O 41  41 vd  O -H  W co +1  IT)  CQ  CO CN  o co  ©  41 ON  co co ^  00 © 41 00 o CN  o  00 o o o 41 00  ON  O O 41  ON O  oo o o  Tt  O  VO ©  VO in  CO  CJ  O  c»-t  o  ©  41 co CN  i>  o  ©  41 ON  o 41  ©  ©  co  co  41 CN co  ON  00  00  vo  © ©'  o  © © ©  © © ©•  © ©'  41  41  41 00 «o  41 CN  ON  ©•  00 m  ©  C  •t-H  © ©  Tf  «o  00 o  ©  00  © © ©  41 VO VO  ©  00 o  VO  VO  o  ©'  o  H  132.  ©  o  ©  41 ©  m o cs-  © © ©  41 co co ©  oj  o  h-1  CO ©  o  CO  o  41  VO  41 CN  so  3 S 1  ©  41 00  ©  B  co  O  CO o 41  CD  C/3  ©  o 41 00  t—I  vd  Tt  © ©  oo o  o 41  Tt  ©  © © ©  VO  00 o 41  CO  O O O 41 o  ©  o  ©  ON  ©  CO  O  Tt Tt  VO  ©  41  o o 41 t-. «o  o o © 41  o 41  VO  CN  •vf  U  CO  ON  ©  VO  oo © 41 m o CN  00 o o © 41  CO  *  1 CJ  p  Figure 1. Fight-month change in trochanteric areal bone mineral density (aBMD, g/cm2) in Asian and Caucasian exercise (black bars) and control (white bars) groups. The difference between groups was significant for the collapsed (Asian plus Caucasian) population (p = 0.030). The ethnicity by exercise group interaction was not significant (p = 0.146).  133  13 if  ANALYSIS OF PROXIMAL FEMUR DXA SCANS IN CHILDREN  Analysis of proximal femur DXA scans in growing children: Comparisons of different protocols for crosssectional, 8-month and 7-yearlongitudinal data  1H.A. McKay, M.A. Petit, . D.A. Bailey, W.M. Wallace, R.W. Schutz and K.M.Khan 1  1  2 3  4  1  1  School of Human Kinetics, University of British Columbia, Vancouver, B.C. College of Kinesiology, University of Saskatchewan, Saskatoon  2  Human Movement Studies, University of Queensland, Australia  3  4  Computing Science, University of Saskatchewan, Saskatoon  Corresponding author: Dr. HA. McKay School of Human Kinetics 6081 University Blvd. 210 War Memorial Gym University of British Columbia V6T1Z1 Ph: (604) 822-3120 FAX: (604) 822-5884  Abstract Dual energy x-ray absorptiometry (DXA) is a widely used method for measuring bone mineral in the growing skeleton. As scan analysis in children offers a number of challenges we'compared DXA results using 6 analysis methods at the total proximal femur (PF) and 5 methods at the femoral neck (FN). In total, we assessed 50 scans (25 boys, 25 girls) from two separate studies for cross-sectional differences in area, BMC and BMD and for percent change over the short term (8-months) and long term (7-years). At the proximal femurforthe short-term longitudinal analysis, there was an approximate 3.5% greater change in bone area and BMC when the global ROI was allowed to increase in size between years as compared to when the global ROI was held constant. Trend analysis demonstrated a significant (p<0.05) difference between scan analysis methods for bone area and BMC across 7-years. At the femoral neck, crosssectional analysis using a narrower (from default) ROI, without change in location,resultedin a 12.9% and 12.6 % smaller bone area and BMC,respectively(both p<0.001). Changes in FN area and BMC over 8 months were significantly greater (2.3%, p<0.05)) using a narrower FNratherthan the default ROI. Similarly, the 7-yr longitudinal datarevealedthat differences between scan analysis methods were greatest when the narrower FN ROI was maintained across all years (p<0.001). For aBMD there were no significant differences in group means between analysis methods at either the PF or FN. Ourfindings demonstrate the need to standardize the analysis of proximal femur DXA scans in growing children.  This work is supported in part by the National Health and Research Development Program Grant (NHRDP) grantno.6608-1261, CANADA. Keywords: BMD, children, DXA, proximal femur, methodology, growth  INTRODUCTION DXA technology is a precise means to evaluate adult bone mineral, and because it is safe, fast and unimposing, it has also become the measurement system of choice for bone mineral assessment in children. Although analysis of DXA data is highly computerized, operators arerequiredto make key choices regarding the size and the placement of the total proximal femur(PF) global region of interest (ROI) and the femoral neck (FN) ROI. It was recently demonstrated that, in adults, misplacement of the FN ROI resulted in errors in 1% of 2300 proximal femur (PF) scans (1). The reported 2.7% precision was much poorer than the 1.3%reportedby the manufacturer. Serial scan analysis in children presents a number of challenges. Forexample, in prospective studies, should the proximal femurgtobal ROI remain constant between years regardless of increased subject size? What degree of error is introduced when the technologist decides to change the size of the region of interest (ROI) as the child grows? The femoral neck subregion is also oftenreportedand is sometimes adjusted for analysis of pediatric scans depending on the size of the child. If, in smallersubjects, the femoral neck ROI overlaps the greatertrochanter, laterally, or the ischium, inferomedially, the operator may manually alter the size (width and length) and location of the femoral neck ROI, or rotate the mid-line in order to avoid bony overlap. Few studies have addressed the methodological considerations presented by pediatric research with DXA (2). Thus, despite the increase in the number of pediatric DXA publications there are as yet no data evaluating thereliabilityof various methods commonly used for pediatric scan analysis. Therefore, todetermine whether a difference in DXA outcomes at the total proximal femur and the femoral neck is introduced when the regions of interest are adjusted, we addressed the following questions: 1. How does maintaining versus increasing the size of the global region of interest of the proximal femur affect area, bone mineral content and bone mineral density? 2. How does increasing or decreasing the size, or changing the location, orrotatingthe femoral neck subregion of interest, affect bone area, bone mineral content and bone mineral density? 3. How do these methodological differences affect percent change at the total proximal femur and the femoral neck a) over a short period (8 months) or b) over a long period (7 years) as a child grows.  MATERIALS AND METHODS Subjects and Scan Database  Two sets of pediatric DXA scans provided the material forthis study (Figure 1). The first set of scans (Dataset 1) were randomly selected from a longitudinal study of a multi-ethnic group of children in Vancouver (3,4) (Figure 1). All children were prepubescent (Tanner stage 1) ranging in age from 7-1 Oy (mean age 8.9±0.8y) at Timel. Right proximal femur scans were performed by the same certified technologist on a Hologic QDR 4500 (array mode) on two occasions, 8 months apart. Scans were analyzed with software version 8.20a:5 and standard positioning protocol for the PF was used (5). The effect of variations in analysis method on bone area, BMC and aBMD (DXA outcomes) of both the total PF and the FN, were determined from Timel scans. Absolute change and percent change from Time 1 to final (8 months) were also calculated. Precision values in vitro of less than 1 % have been reported for both total PF and FN on the QDR 4500 (6). The second set of scans {Dataset 2) were collected at the University of Saskatchewan as a part of a seven year longitudinal study of bone mineral accrual in healthy children (7) (Figure 1). Scans from 10 Caucasian children who had been measured annually over the 7 years were selected (Figure 1). These children were aged 9 years at first measurement (Time 1) and 16 years at final measurement (Time 7). The Saskatchewan data were collected by the same qualified researcher on a Hologic QDR-2000 in the array mode and analyzed using software version 4.55A:1. The coefficient of variation CV(%) in vivo for the total proximal femur and FN as measured on the QDR 2000 was 1.06 and 0.91, respectively (8). We selected a series of scan analysis methods based on personal experience and in discussion with groups involved in pediatricresearch.The analysis methods mostfrequentlyutilized to overcome problems that arise with pediatric scans were methods we evaluated in the present study. Cross-sectional and short-term (8-month) longitudinal analyses: Proximal femur  The cross-sectional and short-term longitudinal analyses of data set 1 (Figure 1) are described below and depicted in Figure 2. Standard/default: Time 1 and Final scans were analyzed by standard protocol as described by Hologic (5). The lower border of the ROI was placed 10 lines below the lesser trochanter if visible, or two times the length of the greatertrochanterif the lessertrochanter was not identifiable. Therightlateral edgeof the ROI wasfivelines outside the edge of the greater trochanterand the upper left corner was placed adjacent to the superioracetabularridge. Scans were analyzed manually, rather than with the system-defined 'Compare' function.  .„ „  Small ROI: 'Compare' was used to impose the global ROI from the Time 1 scan, as analyzed by standard method, on the Final scan. Large ROI: The Final scan was analyzed by standard protocol and 'Compare' was used to impose the larger Final ROI on the Time 1 scan. 7-year longitudinal analyses: Proximal femur To determine whether there was a difference in results obtained by various methods of DXA scan analysis (7 yeardata, Data set 2) we performed a longitudinal analysis utilizing protocols that varied the size of the global ROI. The procedures whereby the global ROI remained constant or increased overshorter or longerterm periods are described below. Standard/default: The system default analysis was utilized consistently across all years. Small ROI for 2 years: The global ROI was held constant across two years imposing the previous year's ROI on thefollowingyear's scan. That is, the Compare function was used to impose the Time 1 ROI on Time 2, the Time 3 ROI on Time 4 and the Time 5 ROI on Time 6. Time 1,3,5, and 7 scans were analyzed by standard protocol Large ROI for 2 years: The global ROI was held constant across two years imposing the subsequent year's ROI on the previous year's scan. That is, the Compare function was used to impose the Time 2 ROI upon Time 1, the Time 4 ROI on Time 3 and the Time 6 ROI on Time 5. Time 2,4,6 and 7 scans were analyzed by standard protocol Large ROI for 3 or 4 years: The global ROI was held constant as long as seemed biologically feasible based on growth, which was 4 years at the younger ages and 3 years at the older ages. Subsequent year's ROI was imposed on previous year's scans. That is, the Compare function was used to impose the Time 7 ROI upon Times 6,5 and 4 and the Time 3 ROI was imposed upon Times 2 and 1. Time 3 and 7 scans were analyzed by standard protocol. Cross-sectional and longitudinal analyses: Femoral neck The protocol as outlined below and depicted in Figure 3 was usedforthe cross-sectional and longitudinal evaluation of Data set 1. Standard/default: Time 1 scans were analyzed by standard protocol as described by Hologic (5). The default femoral neck ROI was 1.5 cm high and 4.1 cm wide and the system default location was perpendicular to the femoral midline with the lower edge aligned to the greatertrochanter which approximates the centerof the femoral neck (5). At the bone mapping stage, if the ischium was superimposed on the ROI, the overlapping portion was deleted.  ,  a  0  Decrease Width: The width of the femoral neck ROI was adjusted to a width relative to (1/3) the length of the long axis of the FN at each year for each subject. This width did not change over 8 months but was either i) maintained at the Time 1 width across the 7 years or ii) increased in size relative to growth from Time 2 to Time 7. Decrease Length: The default ROI was shortened by 2 lines within soft tissue (from 49 to 47 lines). The location and rotation of the ROI were unchanged.Forthe longitudinal analysis the decreased length remained constant across all years. Change Location: The FN ROI was positioned at the center of the femoral neck in which case the lower edge was not necessarily aligned to the edge of greatertrochanter. Forthe longitudinal analysis this location remained constant across all years. The size of the ROI was as described for C (above). Change Rotation: The default position of the femoral midline was rotated 2 lines superolaterally. The size and location of the ROI were unchanged. Statistical Analyses At the proximalfemur,we used paired student t-tests to compare the cross-sectional and short-term longitudinal difference in bone outcomes at final measurement between methods. At the FN, we used pained student t-tests to compare results obtained by the default method with results obtained when the ROI was decreased in width, length or rotated and to compare the narrower FN ROI with the narrower ROI at two locations. We also used t-tests to compare per cent change over 8 monthsforbone outcomes between methods. To adjustformultiple comparisons we applied the Bonferroni correction, post hoc. To assess the difference betweenregressionslopes at both the total PF and the FN across 7 years, we used repeated measures analysis of variance and trend analysis comparing cubic fits with the HunyhFeldt approach. Data were analyzed using SPSSforWindows, version 8.0 (SPSS Inc., Chicago). All results are presented as means± SD unless otherwise noted. Differences were considered significant when p < 0.05.  RESULTS Cross-sectional analyses: Proximal femur When the data were analyzed cross-sectionally, the various methods of scan analysis at the PF produced significantly different values for bone area and BMC (Table 1). A small decrease (2.5 pixels X 2.5 pixels) in the size of the ROI significantly decreased PF bone area and BMC (2.8% and 3% respectively, both p<0.001). A similar increase in the ROI significantly increased bone area and PF BMC (both 3.3%, p<0.001). Group means for aBMD did not differ significantly between analysis methods.  mo  8-month longitudinal analyses: Proximal femur On average, subjects grew 3.5 ± 0.9 cm and gained 2.2 ± 1.3 kg over 8 months. Eight-month change in bone outcomes differed significantly between analysis methods (Table I). We noted 8.3 and 13% (both p<0.001) increases in bone area and BMC, respectively, when the global ROI was increased with growth (Table I). This was an approximate 3.5% greater change than we observed with the size of the global ROI held constant between years. There were no significant differences between group means for aBMD regardless of analysis method used.  7-year bngitudinal analyses: Proximal femur The results of analyses of the 7-year data are shown (Figures 4A-4C). There was a significant (p<0.05) difference in 7-year change between analysis methods for mean bone area (4A). There was no significant difference in mean BMC (Figure 4B). There were no significant differences between group means for aBMD regardless of analysis method used (Figure 4C).  Cross-sectional analyses: Femoral Neck Group means (Cl) for FN bone area, BMC and aBMD are shown (Table II). A decrease in the  width of the FN ROI from the default width, without change in location, resulted in a 12.9% smaller area and a 12.6 % smaller BMC (both p<0.001). There was no difference in FN aBMD between methods. When the length of FN ROI was made smaller by 2 lines within soft tissue (that is, the amount of bone within the FN ROIremainedconstant), area and BMC were about 1.3% smaller (NS). Relocating the FN ROI from the default position (adjacent to the greater trochanter) to the centerof the long axis of the FN resulted in a significantly (p<0.001) smaller bone area (3.3%) and BMC (2.7%). Rotating the femoral neck midline 2 linesresultedin an approximately 1.6% smaller FN area and FN BMC (p<0.005).  8-month longitudinal analyses: Femoral neck Group means for change in FN ROI bone area, BMC and aBMD over 8 months using the five analyses are tabled (Table II). Eight-month changes in FN area were 2.3% (p<0.05) greater when analyzed by the narrow width ROI versus the default. There was nosignificant difference in change in FN aBMD between the two analysis methods. Reducing the length of the FN ROI by 2 lines in soft tissue, produced a 1.6% greater change in bone area (NS) and a 2.3% greater change in BMC (p=0.056). Differences for change in aBMD were not significant for any analysis method. Changing the location orthe rotation of the ROI did not significantly influence the 8-month change in FN ROI area, BMC or aBMD.  I4»  7-year longitudinal analyses: Femoral neck Results from all methods across 7 years are illustrated (Figures 5A-C). The pattern of bone area and BMC gain differed between analysis methods. The magnitude of this difference was the greatest when the narrower width FN ROI was maintained across all years (p<0.001, Figure 5A-B). The aBMD was consistent across all methods (Figure 5C) except when the FN ROI location was changed compared to standard positioning (significant differencefbrlineartrend (p<0.05).  DISCUSSION Our data show that varying the scan analysis method can produce different results. This could influence within-study comparisons, between-study comparisons and has the potential to significantly alter the interpretation of longitudinal studies. Therefore, it is vital that researchers recognize howdifferent methods - of analyzing pediatric DXAscans affect bone area and BMC in particular. Furthermore, the increasing use of bone densitometryforpediatric patients with chronic metabolicconditions, suggests that guidelines for analysis are necessaryforboth research and clinical practice. Implications of using various scan analysis methods Proximal femur: Observational longitudinal pediatric bonestudies provide valuable information regarding bone changes with growth and maturation (8). Wefoundthat several feasible methods of analyzing longitudinal data generated different results. When analyzing an 6-month longitudinal data set by increasing the global region of interest overtime wefounda 3.2% greater increase in bone area and a 3.7% greater increase in BMC compared with a longitudinal analysis where the ROI remained unchanged. This difference is substantial, approximating the magnitude of change in proximal femur BMC over an entire year's growth during prepuberty (3-4%) (7,9).This problem is not rectified by measuring change over longer time periods. When we applied the different analysis methods to 7-year longitudinal data, trend analysis revealed significant differences in regression slopesforarea and BMC depending on the method used to analyze the DXA scans. In intervention studies, errors are introduced ifthe rate of maturation differs between intervention and control groups. These errors are exacerbated by using customized, rather than standard, scan analysis protocols. Although researchers may make every effort to control studiesformaturational differences between groups, an imbalance of early orlate maturing subjects represents a potentially serious confound. If, as is likely, the technician were to increase the global ROIforthe more rapidly maturing (and thus, growing) 14^  children overtime, but did not increase the global ROI in the children who were not maturing as quickly, this would even further (but erroneously) magnify the difference between groups. As different analysis methods produced BMC and bone area results that were proportional, aBMD, the quotient of BMC and bone area, remained virtually constant irrespective of the analysis method. Thus, aBMD is a fairly consistent measure at this site. Using aBMD in studies of growing children, however, has several limitations. When measuring bone mineral by DXA, it is important to differentiate between bone outcomes that are strongly associated with increases in skeletal size (i.e., area and BMC) and those thatrelatespecifically to changes in BMD. Areal BMD (g/cm), although an attemptto control forsizedifferences, does not account 2  for growth of bone in three dimensions and, therefore,reflectsincreases in both bone size and bone density. This problem has been widelyrecognizedand numerous geometric andregressionapproaches have been proposed to estimate volumetric (g/cm) rather than areal density (10,11). In addition, as an option to 3  reporting BMD, various adjustments for BMC including corrections for height, weight or body mass index have been suggested (10-13). Femoralneck When analyzing the femoral neck the operator controls all aspects of ROI size and placement. As there are, to our knowledge, no published recommendations as to how pediatric hip scans should be analyzed, operators may narrow or shorten the FN ROI depending on the size of the child. This presents potential problems both within and between studies. If the femoral neck ROI size isreducedfor a smaller child as compared to a larger child within the same study the present study indicates that this will affect FN area and BMC. Analysis methods and, thus, outcomes can also vary considerably between studies. If, for example, a narrower femoral neck ROI were selected for one study,reportedvalues would be 13% less than for same aged children whose scans were analyzed using the system default. Similarly, as has been reported in studies with OCT (14) the location of the femoral neck ROI may also influence study outcomes. We found that when the FN ROI was moved to the center of the femoral neck and away from the medial border of the greater trochanter, bone area and BMC increased by approximately 3%. In longitudinal studies of growing children, the operator may select either the default ROI or a smaller ROI and maintain either of these over time. In the present study, changes over 8 months for FN BMC (8.6%) and BMD (4.2%) using the standard default across time were comparable to valuesrecordedfor the University of Saskatchewan 7 year Bone Mineral Accrual Study (7) in a similar aged cohort (unpublished data). The present study also demonstrated substantial differences in change in femoral neck BMC with very minimal decreases in width of the femoral neck ROI. If a pediatric bone mineral accrual studyreporteda 4%  annual increase in femoral neck BMC with the ROI kept constant at the default, we found that identical scans analyzed with a narrower ROI would generate a 6.3% annual increase. There are also implications at the femoral neck for long-term longitudinal studies. The 7-year longitudinal data suggest that the pattern of growth may appear to be different depending on the analysis method chosen for long term studies. As was the case at the total proximal femur, aBMD measurement at the femoral neck was consistent across analysis methods. Recommendations for standardizing analysis Proximalfemur The findings of the present study indicate the need to standardize analysis protocols for all studies of growing children. It is important to bear in mind, however, that the protocol of choice will depend on the research question being asked. For observational studies of normal growth we recommend that the systemderived global ROI be used overtime as the child grows. For intervention studies when the effect of the intervention (e.g. calcium or exercise) is being evaluated, we recommend that the global region of interest be held constant across time if biologically feasible. This may be impossible when children are growing very rapidly, as at the adolescent growth spurt. We contend, however, that in this instance the apparent effect of the intervention may be confounded based on a growth artifact and by a changed global ROI. We also recommend that researchers provide, andreviewersrequest,a description of how pediatric scans are analyzed if a protocol other than the system default is utilized. Femoral neck Given the significantly different outcomes of FN scan analysis by various methodsreportedin this paper we urge researchers to be aware of the implications of the choices they may make. If the outcome involves within or between subject comparisons werecommend,if at all possible, that the same ROI be maintained overtime and between subjects. The system default may not be appropriate forstudies of children younger than 8 years. If the FN ROI has been altered, researchers shouldreport,and reviewers should request, this information. Summary & Conclusions Although the present study demonstrates the need forstandardization between and within bone densitometry studies of growing children, theresultsare specific to Hologic systems. Protocols for, and outcomes from, the analysis of proximal femur scans may vary significantly based on hardware and software differences between manufacturers. Studies examining thereliabilityof pediatric scan analysis  m  using other DXA systems are needed. Agreement between research groups and institutions would permit between-study comparisons of cross-sectional and longitudinal outcomes. It is important that researchers differentiate between bone area and BMC outcomes due to subject differences, differences in the response to the intervention and results that reflect analysis method. We offerrecommendations forthe standardization of pediatric proximal femur scan analysis for Hologic systems. References 1. Staron R, Greenspan R, Miller T, Bilezikian J, Shane E, Haramati N1999 Computerized bone densitometricanalysis.-Operator-dependent errors. Radiology 211:467-70. 2. Leonard M, Feldman H, Zemel B, Berlin J, Barden E, Stallings V1998 Evaluation of low density spine software forthe assessment of bone mineral density in children. J Bone Miner Res 13:1687-1690. 3. McKay H, Petit M, Schutz R, Barr S, Prior J, Khan K1999 Augmented trochanteric bone mineral density after modified physical education classes: A randomized, school-based exercise intervention study in pre- and early-pubescent children. J Pediatr In press. 4. McKay H, Petit M, Khan K, Schutz R1999 Uestyle determinants of bone mineral: A comparison between prepubertal Asian- and Caucasian-Canadian boys and girls. Calcif Tissue Int In press. 5. Hologic 1996 QDR 4500 Users Guide. Hologic Inc. 6. Steiger P1998 Fan beam dual X-ray absorptiometry: an important advance in bone densitometry. Br J Radiol 71:993-994. 7. Bailey DA 1997 The Saskatchewan Pediatric Bone Mineral Accrual Study: Bone mineral acquisition during the growing years. Int J Sports Med 18 (Suppl3):S191-194. 8. Bailey D, Mirwald R, McKay H, Crocker P, Faulkner R1999 The University of Saskatchewan Bone Mineral Accrual Study: A seven year longitudinal study of the relationship of physical activity to bone mineral accrual in growing children. J Bone Miner Res 14. 9. Bonjour JP, Theintz G, Buchs B, Slosman D, Rizzoli R1991 Critical years and stages of puberty forspinal and femoral bone mass accumulation during adolescence. J Clin Endocrinol Metab 73:555-563. 10. Kroger H, Kotaniemi A, Kroger L, Alhava E1993 Development of bone mass and bone density of the spine and femoral neck- a prospective study of 65 children and adolescents. Bone Miner 23:171-182. 11. Katzman DK, Bachrach LK, Carter DR, Marcus R1991 Clinical and anthropometric correlates of bone mineral acquisition in healthy adolescent girls. Journal of Clinical Endocrinology and Metabolism 73:1332-1339. 12. Prentice A, Parsons T, Cole T1994 Uncritical use of bone mineral density in absorptiometry may lead to size-related artifacts in the identification of bone mineral determinants. Am J Clin Nutr 60. 13. Molgaard C, Thomsen B, Prentice A, Cole T, Michaelsen K1997 Whole body bone mineral content in healthy children and adolescents. Arch Dis Child 76:9-15. 14. Kuiper J, van Kuijk C, Grashuis J1997 Distribution of trabecular and cortical bonerelatedto geometry: A quantitative computed tomography study of the femoral neck. Investigative Radiology 32:83-89.  Address reprint requests to: Dr. HA. McKay School of Human Kinetics 6081 UniversityBlvd. 210 War Memorial Gym University of British Columbia V6T1Z1  14 5  i o IO  cn  t—  O CC or "co uu E  O z  .01tt 4; 1.1  CO  UJ  T—  <-;  co  C/3  CO O —:  CO  CD  o cx>  CD  cn  co  " " r  :  cvr  CM <P  NCM O CO  o> co  cEuP co  o  I—  o  UJ CO CO CO  o or o  <o  CM  O  CO CO o  or  CM  sCD  +cn  TJ-  CO  sCM  T  T  CO CO  CO  CD  co  CM  "55  ._  10  co  IO LO CO O  CM CO  co o  CO  e  CO o o  CO CM o  ST £5 co 0  o  T  v  CO  CL "O  o  I  CM CO  T—  ZJ  10 co  OJ  -o  CO «  I  CM  i  CO  TJ-  CM  T ? TT  CM CM  IO TT  CO* co LO  (13.64;  o  Tj"  (20.41;  <  CM  o  +-  O  Tj-  CO  oor  P  TT""  <  O  co  CM  o o o  ST 0  00 CO 0  v  o.  0  o o CD V Q.  E CD CO  "co c= OJ E E  (0.659;  2O  E?  *>!  24tt 5; 1.4  CO  r^.  TT  (0.02 2; 0.0  CU  I  co co  10  (0.6  oor  "ro  CM CD  CO OJ  J3  —'  c  1 i 0  |  LU  ct  •s =•» co I—  o -Q  8  o  f  P  •«-  o"  CO  CM  cn ^  uo co  CO  CO  10 p Tt  o  co  42  cn  10 10 <o  o  co  CO CO o.  M  "S  2  CO  <  3  2  E  .0  O  CQ  o  CO  cn  •e  . a "co «  14b  >  OJ  c  CO  o  _>»  "c= CO  CO  = J S>E «S OJ  Q OE E CD  T3  E  =3  eg  CO CJ  "e J2  o «£= "cz  "co  cu O)  CZ O  TO '16  UJ  O •z. <  6  3  X  o  o  a? s?  <D  LO  CO  oi di co  oo  £  iri  co"  lO  — f  oT od  * 2  ^  co  n  *  o  co.  oo  CM  to  •if  CO  CC LU  0-  cu c9>o T5S  8  Q  oo to CD  "E CD T3 CZ  CNJ  v  co  i  CO CN  t CD o^—  CO urj CM"  co cd  2^  co  LO  o _ fi  CO  _  t—  CNT  od  CM  £2-  CM •"3"  LO  CO  co  CO  cz o  i  ••  oo  o> cz  T3  co  CO  6T  co  CO  ..to  ^  CM  £2.  CM  tg «>  "*  O  S  o £r  S  £S  o  JZ  •CZ  to o o v Q.  o  O  o  CO  o> cz CD x: O  O tn  g o  CD  co  LU CO  5  if  CO  1 J  O)  ob CO  o o  o E  e  s  & § g  8  8  S  3  le  Q  o -Q ro h—  §> co j= o  m  .=  si  CT) CO  co  CM  5  IS 5  =S E  cz>  CO  •s  co CO CO  CO CZ CD  o  CO  CO CO CD CO  oi  oi  o o  CO  to CO o  co <z>  V  CO* CO  o  JZ  ro  CO  CM CM  CO CM CO . _ CO  o o o  V  co  CD  fi  CO  cr o>  CD  CO,  •It *  < i—  o>  CO CO  CO  5t  co cu  «  8t  CD  locatior  CU  CM" 1^ CMCO  oi  £  Oi  oo co co co  "E co  "S  CO  ro co  "8  CO  ro cz CO  cz  o  ro  cd cd  to oi  O  roi- -  1 0  co  oi  CO CO CO  Q  CO  u.  £  1^  S  CD CO  E .2 2 £  TZ> >% 33 C CO  o cz .!=?  CO  ro  CO CD CO  _»» CO  cz  ro *  03  XJ  co  8  a. t t ro IS ro  "E ro  o  a.  to  &  «o  CO  CO  "E ro -o cz  2  cz ca •E  CD T3 CO  I J M  CO  CO  •o  CD  ro 2 o  CD TZJ  E  e  "£ <D  I  £  C= 0 3  cz  co  co = CO  .>. co  ca cz ro "E co •o cz CO -i5'  c  CO CD CD  CD  -O T3  a> o cz  CD C3) CZ  ro JZ  E 2 cz 2 £ 1z CO cz o o>  £cz m CO  CO  • •  CO  II  o" cn CD ro CD  UBC  CQ E? Q. 13" CD  o  3  o o  20M,  o 13  >  na lyses  onth  3  Cro:ss-se ion  CO  CD T l  |ip!il|lll|li  CQ  *<  cB  >  —'  9L  CD »<  —< <n CO  o Z3 c z o II CO cn o  o'  — h  5F  CO  ro o CD CD  CA)  148  C  PROXIMAL FEMUR BONE MINERAL DENSITY  1.2&1  J  15  152  c.  FEMORAL NECK BONE MINERAL DENSITY 1.25n  Lifestyle determinants of bone mineral: A comparison between prepubertal Asian- and Caucasian-Canadian boys and girls H.A. McKay, M.A. Petit, K . M . Khan and R.W. Schutz School of Human Kinetics, University of British Columbia, Vancouver, B.C.  Send Correspondence to: H.A. McKay, Ph.D. School of Human Kinetics, University of British Columbia, Vancouver, B.C. V6J 1K7 TEL. (604) 822-3120 FAX (604)822-5884 mckayh@unixg.ubc.ca Number of text pages: 8 Number of reference pages: 3 Number of tables: 2 Key words: bone mineral-calcium-physical-activity-children-ethnicity  Calcified Tissue International; Accepted October, 1  Comparison of bone mineral between Asian and Caucasian-Canadian children  Summary The purpose of this study was to examine the difference in lifestyle and morphometric factors that affect bone mineral and the attainment of peak bone mass in 168 healthy Asian (n=58) and Caucasian (n=110) Canadian, prepubertal girls and boys (mean age 8.9±0.7) living in close geographical proximity. DXA (Hologic 4500) scans of the proximal femur (with regions), lumbar spine and total body (TB) were acquired. We report areal densities (aBMD g/cm ) at all sites and estimated volumetric density (vBMD, g/cm ) at the femoral neck. Dietary calcium, physical activity and maturity were estimated by questionnaire. Of these prepubertal children all of the boys and 89% of the girls were Tanner stage 1. A 2 X 2 ANOVA demonstrated no difference between ethnicities for height, weight, body fat or BMFL mass. Asian children consumed significantly less dietary calcium (35%) on average and were significantly less active (15%) than their Caucasian counterparts (p<0.001). There were significant ethnicity main effects for femoral neck BMC and rtBMD (both p<0.001) and significant sex by ethnicity interactions (p<0.01). The Asian boys had significantly lower femoral neck BMC (11%,), aBMD (8%) and vBMD (4.4%). At the femoral neck, BMFL, sex and physical activity explained 37% of the total variance in BMD (p<0.05). In summary, this study of a large number of Asian and Caucasian prepubertal children demonstrated differences in modifiable lifestyle factors and femoral neck bone mineral between Asian and Caucasian boys. The lower values observed in the Asian boys may be disadvantageous for achieving optimal peak bone mass. 2  3  Comparison of bone mineral between Asian and Caucasian-Canadian children INTRODUCTION There is increasing recognition that peak bone mass is an important determinant of adult bone mineral, and thus, risk of osteoporotic fracture [1-3]. Further, the prepubertal years are a particularly important time for bone mineral acquisition [1,46]. The major determinants of peak bone mass are genetic make-up, ethnicity, gender, weight-bearing physical activity, calcium intake, and soft-tissue composition [7, 8]. Within these determinants, ethnicity and sex are associated with differing rates of osteoporotic fracture and different amounts of bone mineral. Studies of ethnic differences in bone mineral are often confounded by the geographical separation of the ethnic groups being compared [9] and, in children, by variability in maturity and body size. Specifically, the only study of bone mineral in Asian and Caucasian children living in close proximity included subjects across a broad maturity range (9-26 y) and very few pre-pubertal children. To minimize age, maturity and size related variability, we studied bone mineral in Asian and Caucasian children aged 8 and 9 years who attend primary school classes together. The primary aims of this cross-sectional study were: (1) to report bone mineral content (BMC), bone mineral areal density (aBMD) and estimated bone mineral volumetric density (vBMD) across three skeletal sites, in Asian and Caucasian boys and girls; (2) to identify the determinants of bone mineral content and density and bone volume such as physical activity, dietary calcium intake and soft tissue composition in this population.  MATERIALS AND METHODS Subjects The study was presented to both the principals and the teacher-development representatives from a multiethnic school district which includes 34% of the population who report Chinese as their first language. Of the 15 schools that demonstrated initial interest, 10 schools agreed to take part in the study. Parents' place of birth was determined by questionnaire which was available in Chinese for the non-English speaking . Children were classified as Asian if both parents were born in China, Hong Kong (72%), Japan, Taiwan or Viet Nam and as Caucasian if parents were born in North America, Australia or Great Britain. Parents completed a health history questionnaire for their children. There were 168 healthy boys and girls (mean age 8.9± 0.7 y) enrolled in the study; 58 Asian (30 boys, 28 girls) and 110 Caucasian (56 boys, 54 girls) children. None of the participants had medical conditions or were taking medications known to influence bone mineral. Measurement DXA scans of the proximal femur, postero-anterior lumbar spine and total body (TB) were acquired and analyzed by the same qualified technologist using a Hologic QDR 4500 bone densitometer (Hologic Inc., Waltham, MA). Bone mineral content  Comparison of bone mineral between Asian and Caucasian-Canadian children (BMC, g) and areal density (aBMD, g/cm ) of these sites and the femoral neck (FN) region are reported. Also, in order to more adequately represent three dimensional size and to approximate true bone mineral density, bone volume and volumetric density were estimated at the femoral neck from areal density using appropriate formulae which have been used elsewhere for children [10]. Fat mass (g), bone mineral free lean mass (BMFL, g) and percentage of each was determined from the TB scan. Stretch statures (sitting and standing) were measured to the nearest millimeter using a wall stadiometer. Weight was measured on an electronic scale to the nearest 0.1 kg. Mean values were used for analysis. Questionnaires were administered at the time of bone densitometry measurement. Dietary calcium intake was estimated from a food frequency questionnaire that has previously been validated against food records (r=0.98) as a tool for assessing calcium intake in Asian and Caucasian high school students [11]. A physical activity questionnaire which has previously demonstrated adequate testretest reliability in children from grades 4 to 8 and has been validated against a seven day physical activity recall interview and a Caltrac motion sensor [12]. An activity score comprised of the amount of daily weight-bearing physical activity the children reported in the previous week was calculated and used to represent physical activity. The children were also asked if, and in every case indicated that, their previous week's activities represented their usual physical activity choices and patterns. Maturity was self-assessed by ratings of breast (girls) and pubic hair (girls and boys) development using a standard approach [13] that we have used previously [14]. Self-assessment has demonstrated strong associations with direct clinical observation [15] and as it is non-invasive is preferable for use with children. 2  Statistical analyses  A two by two, sex (M,F) by ethnicity (Caucasian, Asian), ANOVA was used to compare independent group means for body composition (leg length, sitting height, height, weight, fat and bone mineral free lean (and %)) variables. As there was no statistical difference between groups for height or weight, precluding the need to control for these factors, ANOVA was also used to determine between group differences in physical activity, dietary calcium, BMC and A B M D at the proximal femur (total), femoral neck (FN), lumbar spine and total body (TB) and for bone volume and estimated i?BMD at the femoral neck. Results are presented as means (SD) and differences were considered significant at p<0.05. Stepwise multiple regression models were fitted to estimate the contribution of the independent variables to absolute values of bone mineral. SPSS for Windows was used for all statistical analysis.  168  Comparison of bone mineral between Asian and Caucasian-Canadian children RESULTS All of the boys and 89% of the girls were at Tanner stage 1. Table 1 summarizes the body composition and lifestyle characteristics of the group by ethnicity. ANOVA demonstrated a significant main effect for ethnicity for both dietary calcium intake (p<0.001) and physical activity (p<0.001). In both cases, the Asian group reported lower values (Table 1). As a group, the Asian children consumed, on average, 35% less dietary calcium than their Caucasian counterparts. The differences were greatest between the Asian (mean=735 mg) and Caucasian (mean=1165 mg) boys where the Asians indicated a 41% lower calcium intake compared with a 29% difference for the girls. The ethnic difference in physical activity is a function of the much more pronounced level of activity in the Caucasian as compared to the Asian boys as noted by the significant sex by ethnicity interaction (p<0.001). The Asian children were, on average, 15% less active than their Caucasian counterparts. While 73% of the Caucasian boys participated in organized sport only 14% of the Asian boys reported similar involvement with sports teams outside of school. The Asian children (57%) were twice as likely to be attending academic lessons (Chinese, mathematics, music, etc.) after school than the Caucasian children (28 There was no difference in sitting height, leg length (p=0.06), soft tissue (fat mass or BMFL) between groups. Descriptive statistics for bone mineral variables are provided (Table 2). At the femoral neck there was a significant ethnicity effect for BMC and aBMD (both p<0.001) and for estimated bone volume (p<0.05), with Asian boys having consistently lower values for all of these measures (Table 2, Figure 2). There was no difference between groups for femoral neck vBMD. At the proximal femur (total) there was a significant sex by ethnicity interaction for aBMD (p<0.05) which is explained by the approximately 6.5% lower aBMD at this site in the Asian, as compared with the Caucasian, boys. There were no between group differences at either the TB or the lumbar spine. At both the total proximal femur and femoral neck, BMFL, sex (both p<0.001) and physical activity (p<0.05) entered as significant predictors of aBMD and accounted for 29% and 37% of the total variance, respectively. For TB aBMD sex, BMFL and fat explained 25% of the variance. For BMC at the femoral neck, BMFL (p<0.001), sex (p<0.001), calcium (p<0.05) and fat (p<0.05) were significant predictors, explaining 61% of the variance.  DISCUSSION  Comparison of bone mineral between Asian and Caucasian-Canadian children Little is known about ethnic influences on bone mineral during the prepubertal years as most normative studies have been limited to Caucasian subjects [1,16], and have had only a small number of prepubertal subjects. Ethnic differences in BMD have been attributed primarily to variations in body (bone) size [17]. In this study of prepubertal Asian and Caucasian children living in geographical proximity, we observed racial differences in bone mineral that are not accounted for by differences in body size. Physical activity and dietary calcium intake of Asian and Caucasian children have rarely been directly compared. We found that Asian children were less physically active than their Caucasian counterparts. In the only published study that included both Asian and Caucasian children, Bhudhikanok et al. [17] reported similar results using an activity frequency questionnaire but with far fewer (18 Asian; 34 Caucasian) and slightly older (10.5±1.05 y) subjects than in the maturity group we describe. Although there have been no studies evaluating the effect of physical activity on bone in boys of various pubertal stages, recent Finnish studies in girls showed that bone mineral appears to respond maximally to mechanical loading during Tanner stage 1 to 3 [6] and that physical activity was a major determinant of bone mineral [18]. The mean estimated dietary calcium intakes of Asian children in this study were greater than for Asian children living in Hong Kong [19] but substantially lower than for the Caucasian children studied. Five year old Chinese children in Hong Kong, who consumed twice the dietary calcium intake of children in a mainland Chinese city had a 14% greater radial BMC as measured by single photon absorptiometry [20]. A strength of the present study is that the same dietary calcium questionnaire was administered to children of both ethnicities allowing for direct comparisons. The significant ethnicity effect for femoral neck BMC within boys differs the findings of Bhudhikanok et al. who showed no ethnic differences in boys until midpuberty [17]. Unlike the present study, these authors also reported lower values between ethnicities for FN and total body BMC in Tanner stage 1 and 2 girls. This may reflect the small numbers of pre/early pubertal girls (5 Asian, 14 Caucasian), and earlier maturation or a greater proportion of girls in Tanner stage 2 in the Caucasian sample [17]. Bhudhikanok et al. attributed their outcomes to size differences which were not evident in our cohort who were of similar height and weight [17]. A strength of the present study is the larger number of prepubertal (only) subjects than has previously been reported. For T B and lumbar spine our data support the contention that after accounting for size there are no differences in BMC between Asian and Caucasian children. In our study when A B M D was the outcome measure, thus partially correcting for size, the lower values for the Asian boys at the femoral neck persisted. We report aBMD data as they remain the best predictor of fracture risk in adults and to compare our findings with those studies that reported aBMD only [21]. Our finding of no ethnic difference in uBMD is consistent with studies in adults [22, 23] and in older children [17] that suggest vBMD is similar in Asians and Caucasian and is independent of age. Recent work by Gilsanz in Caucasian- and AfricanAmerican children using QCT also supports this contention [24].  I IPO  Comparison of bone mineral between Asian and Caucasian-Canadian children In summary, this study of a large number of Asian and Caucasian prepubertal children demonstrated differences in modifiable lifestyle factors between Asian and Caucasian boys. As ethnic differences were observed for both femoral neck bone volume and «BMD and as both of these variables are independent risk factors for osteoporotic fracture in adults [25, 26] these data may have important long-term implications. It would, therefore, seem logical to suggest intervention via increased physical activity and increased dietary calcium intake in this group.  Comparison of bone mineral between Asian and Caucasian-Canadian children References 1. Bailey DA, Faulkner RA, McKay HA (1996) Growth, physical activity, and bone mineral acquisition. In: Holloszy JO (ed), Exercise and Sport Sciences Reviews, Williams & Wilkins, Baltimore, p 233-266 2. Kelly PJ, Eisman J, Sambrook PN (1990) Interaction of genetic and environmental influences on peak bone density. Osteoporos Int 1: 56-60 3. Ott SM (1990) Attainment of peak bone mass. J Clin Endocrinol Metab 71: 1082A-1082C 4. Khan K, Bennell K, Hopper J, Flicker L, Nowson C, Sherwin A, Crichton K, Harcourt P, Wark J (1998) Self-reported ballet classes undertaken at age 10-12 years and hip bone mineral density in later life. Osteoporos Int 8:165-173 5. Kannus P, Haaspasalo H, Sankelo M, Sievanen H, Pasanen M, Heinonen A, Oja P,Vuori I (1995) Effect of starting age of physical activity on bone mass in the dominant arm of tennis and squash players. Ann Int Med 123: 27-31 6. Haapasalo H, Kannus P, Sievanen H, Pasanen M, Uusi-Rasi K, Heinonen A, Oja P,Vuori I (1998) Effect of long-term unilateral activity on bone mineral density of female junior tennis players. J Bone Min Res 13: 310-319 7. Katzman DK, Bachrach LK, Carter DR, Marcus R (1991) Clinical and anthropometric correlates of bone mineral acquisition in healthy adolescent girls. J Clin Endocrinol Metab 73: 1332-1339 8. Chesnut CH (1989) Is osteoporosis a paediatric disease? Peak bone mass attainment in the adolescent female. Public Health Rep SI: 50-58 9. Villa ML,Nelson L (1996) Race, ethnicity, and osteoporosis. In: Marcus R, Feldman D, Kelsey J (ed), Osteoporosis, Academic Press, San Diego, CA, USA, p 435447 10. Kroger H, Kotaniemi A, Vainio P, Alhava E (1992) Bone densitometry of the spine and femur in children by dual-energy x-ray. Bone Miner 17: 75-85 11. Barr S (1994) Associations of social and demographic variables with calcium intakes of high school students. J Am Diet Assoc 94: 260-266 12. Crocker P, Bailey D, Faulkner R, Kowalski K, McGrath R (1997) Measuring general levels of physical activity: preliminary evidence for the Physical Activity Questionnaire for Older Children. Med Sci Sports Exerc 29: 1344-1349 13. Tanner J (1955) Growth at adolescence: Oxford: Blackwell Scientific Pub 14. Bailey DA (1997) The Saskatchewan bone mineral accrual study: Bone mineral acquisition during the growing years. Int J Sports Med 18 (Suppl3): S191-194 15. Matsudo S, Matsudo V (1993) Validity of self-evaluation on determination of sexual maturational level. In: Claessens A,Lefevre J,Eynde B (ed), World-Wide Variation in Physical Fitness., Institude of Physical Education, Leuven, p 106-110 16. Bachrach L (1994) Bone acquisition in childhood and adolescence. In: Marcus R (ed), Osteoporosis, Blackwell Scientific Publications, Boston, p 69-106 17. Bhudhikanok GS, Wang MC, Eckert K, Matkin C, Marcus R, Bachrach LK (1996) Differences in bone mineral in young Asian and Caucasian Americans may reflect differences in bone size. J Bone Min Res 11: 1545-1556  Comparison of bone mineral between Asian and Caucasian-Canadian children 18. Uusi-Rasi K, Nygard CH, Oja P, Pasanen M, Sievanen H, Vuori I (1997) Determinants of bone mineralization in 8-20 year old Finnish females. Eur J Clin Nutr 51: 54-59 19. Lee WTK, Leung SSF, Wang S, Xu Y, Zeng W, Lau J, Oppenheimer SJ, Cheng JCY (1994) Double-blind controlled calcium supplementation and bone mineral accretion in children accustomed to a low-calcium diet. Am J Clin Nutr 60: 744-750 20. Lee WTK, Leung SSF, Ng MY, Wang SF, Xu YC, Zeng WP, Lau J (1993) Bone mineral content of two populations of Chinese children with different calcium intakes. Bone Miner 23:195-206 21. Bonjour JP, Rizzoli R (1996) Bone acquisition in adolescence. In: Marcus R,Feldman D,Kelsey J (ed), Osteoporosis, Academic Press, San Diego, CA, USA, p 465476 22. Russell-Aulet M, Wang J, Thornton JC, Colt EWD,Pierson RNJ (1993) Bone mineral density and mass in a cross-sectional study of Asian and White women. J Bone Min Res 8: 575-582 23. Ross PD, He Y-F, Yates AJ, Coupland C, Ravn P, McClung M, Thompson D, Wasnich RD, Group ES (1996) Body size accounts for most differences in bone density between Asian and Caucasian Women. Calcif Tissue Int 59: 339-343 24. Gilsanz V, Boechat MI, Roe TF, Gilsanz R, Loro ML, Sayre J, Goodman WG (1994) Gender differences in vertebral body sizes in children and adolescents. Radiology 190: 673-677 25. Cummings SR, Black DM, Nevitt MC, Browner W, Cauley J, Ensrud K, Genant HK, Palermo L, Scott J, Vogt TM (1993) Bone density at various sites for prediction of hip fractures. Lancet 341: 72-75 26. Einhorn TA (1992) Bone strength: The bottom line. Calcif Tissue Int 51: 333-339  11*3  Comparison of bone mineral between Asian and Caucasian-Canadian children Table 1 Comparison of body composition and lifestyle factors between Asian and Caucasian males and females, mean (SD) Males n Sit Height (cm)  Activity score  67.4 (16.6)  28 74.5 (4.9) 135.2 (7.6) 60.8 (4.0) 30.8 (8.0) 8042 (4660) 24.8* (7.7) 22470 (3622) 73.1* (7.3) 73.3 (21.5)  Calcium (mg)  735" (367)  821 (418)  Height (cm) Leg Length (cm) Weight (kg) Fat (g) %Fat BMFL (lean) %BMFL  30 73.3 (3.8) 132.3 (6.6) 59.1 (3.8) 28.6 (5.6) 5894 (2947) 19.9 (6.0) 22516 (3127) 77.7 (5.7)  Asians Females  a  lower than Caucasians lower than Caucasian males  b  greater than Asian males *significance at (p < 0.05)  Caucasians Females  Total  Males  58 73.9 (4.4) 133.7 (7.2) 59.9 (3.9) 29.7 (6.9) 6931 (3984) 22.3 (7.3) 22494 (3345) 75.5 (6.8) 70.2 (19.2)  56 73.3 (3.3) 134.5 (6.6) 61.2 (3.9) 31.5 (8.0) 6645 (4624) 19.7 (7.7) 24563 (3842) 77.9 (7.3) 87.8 (17.1)  54 73.9 (4.5) 135.1 (7.6) 61.2 (4.5) 30.0 (5.2) 7302 (2902) 23.8* (5.9) 22430 (2853) 74.0* (5.6) 76.0* (13.5)  110 73.6 (3.9) 134.8 (7.1) 61.2 (4.2) 30.8 (6.8) 6967 (3872) 21.8 (7.2) 23516 (3543) 76.0 (6J) 82.0 (16.5)  777 (392)  1241 (786)  1159 (841)  1201 (811)  s  X  Total  Comparison of bone mineral between Asian and Caucasian-Canadian children Table  2 Comparison of B M C , A B M D , bone volume and vBMD between A s i a n and Caucasian males and females, mean ( S D ) Asians Females  Males  mem  2.20 (0.45)  2.40 (0.46)  12.52 (3.10)  13.75 (2.22)  12.33 (2.34)  13.05 (2.38)  20.40 (4.29) 843 (164)  20.42 (3.95) 843 (153)  21.44 (3.71) 907 (152)  20.22 (3.46) 833 (117)  20.85 (3.63) 871 (140)  0.61 (0.08)  0.67 (0.06)  0.60 (0.06)  0.64  (0.07)  0.63 (0.08)  0.61 (0.07) 0.66 (0.07)  0.68 (0.06)  0.61* (0.06)  0.65 (0.07)  0.56 (0.08) 0.78* (0.05)  0.56 (0.07) 0.80 (0.06)  0.56 (0.06) 0.82 (0.05)  0.56 (0.06) 0.79* (0.04)  0.56 (0.06) 0.81 (0.04)  6.82* (2.06) 0.34 (0.08)  7.85 (7.46)  7.19 (1.86)  7.53 (1.84)  0.34 (0.06)  0.32 (0.07)  0.33 (0.06)  2.20 (0.49)  Prox.Fem.Tot  12.12° (2.50) 20.43 3.68) 842 (145)  0.62" (0.06) 0.64  ttMIMjj/cm*) Femoral Neck Prox.Fem.Tot  (0.07) 0.56 (0.06) 0.82 (0.05)  a  PA Spine Total Body  Fern. Nk. volume (cm )  6.72 (2.29)  6.92 (1.82)  Fern. Nk. vBMD (g/cm )  0.35 (0.10)  0.32 (0.04)  3  3  x  2.58 (0.39)  2.18* (0.47)  Total Body  lower than Caucasians  " lower than Caucasian males greater than Asian males •significance at (p < 0.05)  Total  2.19 (0.48) 12.31 (2.78)  Femoral Neck  PA Spine  Total  Caucasians Males Females  X  x  I  \P\D  Familial resemblance of total body bone mineral, lean and fat mass in prepubertal Asian- and Caucasian-Canadian children and their parents M. A. Petit and H.A. McKay School of Human Kinetics, University of British Columbia, Vancouver Canada  Correspondence to: Heather McKay, Ph.D. School of Human Kinetics 210 War Memorial Gym University of British Columbia Vancouver, B.C. V6T1Z1 phone: 604-822-3120 fax: 604-822-6842  email: mckavh@interchange.ubc.ca  Osteoporosis International: Submitted, October, 1999  lot  Abstract There are few studies examining familial influences on bone and body composition in pediatric groups. We examined familial resemblance in total body bone, lean and fat mass in prepubertal Asian- and Caucasian- Canadian children (n=49) and their mothers (n=49) and fathers (n=28). Children ranged in age from 6.9-10.0 y (mean 8.9 y) and parents' mean age was 41.2y. Total body bone mineral content (TB BMC, g), lean and fat mass were measured using a Hologic QDR-4500 densitometer (DXA). Height and weight were also recorded. Calcium intake and physical activity were estimated by questionnaire. For both children and parent groups, Asians had lower calcium intakes and physical activity levels than Caucasians. Significant correlations were observed between mothers and daughters for height, weight, lean and fat tissue, and TB BMC (all - r= 0.50, p < 0.05). Resemblance for TB BMC was also apparent for mother-son (r = 0.439, p < 0.05) and father-son pairs (r = 0.584, p < 0.05). However, when adjusted for lean mass and height within each sex, ethnic and generation group, these correlations were no longer significant in any group (r = 0.156-0.365, p > 0.05). There were trends for ethnic differences in resemblance of BMC and body composition variables. These data suggest that familial resemblance of TB BMC is not apparent prior to puberty after adjusting for lean mass and height. The findings support the notion that heritability of bone mass is complex and interacts with a number of variables including age, sex, ethnicity, body size and lean mass. Keywords: Body composition; Bone mineral content; Dual-energy x-ray absorptiometry; Heritability; Asian; Pediatrics  :> lb 6  Introduction It is clear that genetic factors explain a large portion of the variance in peak bone mass, and it is likely that individual differences in response to exercise and/or calcium interventions are at least partially explained by genetic influences [1,2]. Current theories suggest that lifestyle interventions aimed at increasing peak bone mass will be most effective if introduced prior to puberty, and a number of intervention studies have been recently published [3-5] or are currently underway to test this theory. Despite the rapidly growing interest in the pre- and early-pubertal years as an optimal time for intervention, there is a lack of research examining the contribution of genetic factors to the pediatric skeleton. Prior to identifying specific genes, the first step in quantitative genetics research is to determine if a familial resemblance exists for the trait in question. A strong familial resemblance has been established for height throughout pubertal growth [6]. In contrast, there is only one study showing resemblance of bone mineral content (BMC) and areal density (aBMD) in prepubertal girls [7]. In addition to familial resemblance in height and peak bone mass, there is also a genetic influence on body weight, lean and fat mass. These factors, in turn, are significantly related to each other. The degree of association between these tissues, and the genetic effects on each, may vary according to the age, sex, and ethnicity of individuals [8], yet very few studies have examined these relationships. There is a need for studies of familial resemblance in prepubertal children to complement those in adult groups, and to explore the complex interactions between bone mass, heredity, and body composition at various stages of growth. There are no studies of familial resemblance including prepubertal boys or fathers of prepubertal children, or in ethnic groups other than Caucasians. Therefore, we examined familial resemblance in total body bone, lean and fat mass between prepubertal Asian- and Caucasian-Canadian boys and girls and their mothers and fathers. Methods Participants A total of 88 parents (57 mothers, 29 fathers) volunteered for the present study. Parents were biological mothers and fathers of Asian and Caucasian children participating in an exercise intervention study [4]. Children's (28 sons, 21 daughters) baseline measurements were utilized for the present analysis. Ethnicity was determined by parents' place of birth. Parents of Asian children were born in Hong Kong (56%), China (18%) or Taiwan (26%). The majority of Caucasian children and their parents were born in Canada (81%). Mothers were excluded if they had a history of ovariectomy or hysterectomy (n= 6), long term Cortisol use (n = 1), kidney disease (n=1) or scoliosis (n=1). One father was excluded for treatment of hypothyroidism (n =1). None of the women or men in the study had other diseases nor were they using medication known to effect bone metabolism. None of the women were had used oral contraceptives or other forms of estrogen or progesterone in previous 12 months. Bone and body composition Total body (TB) bone mineral content (BMC, g) was measured on a Hologic QDR 4500 densitometer (DXA). Children's measurements were taken during a 3 week period in October, 1997 and parents were measured in April, 1998. All scans were acquired and analyzed by the same certified  I  b°l  technologist. Analyses were completed using TB software version 8.20a. Body composition, including bone mineral free lean (lean, g) and fat (g) mass, was determined from DXA total body scans. The in vivo coefficient of variation (CV) is less than 1% for TB BMC. Anthropometry, maturity, and lifestyle characteristics  Height was measured to the nearest 0.1 cm using a stadiometer and headboard. Weight was measured to the nearest 0.1 kg with a digital scale. The average of two measures was utilized for analysis. If the two measurements differed more than 0.4 cm or 0.1 kg, a third measure was taken and the average of the two closest values was used. Pubertal development was evaluated by self-assessment of breast (girls) and pubic hair (girls and boys) using a standard approach that has been documented previously [9]. Calcium intake for both children and parents was estimated from a food frequency questionnaire (FFQ) which has been validated in Asian and Caucasian High School students [10]. Children's physical activity was estimated from a 7-day physical activity recall [11]. Parents' childhood, young adult and current physical activity were estimated by questionnaire. Parents were asked to rate overall physical activity on a scale of 1 to 5 (1 = seldom active, 5 = very active) as 1) a child/youth (prior to age 18y), and 2) as an adult. To confirm ratings of physical activity, parents were asked if they participated regularly in sport or physical activity during childhood and adulthood. Finally, an estimate of current hours of weight-bearing physical activity per week was averaged from a 2 week activity recall. Statistical methods Descriptive statistics (mean ± SD) were calculated for all variables. A 2 (male, female) by 2 (Asian, Caucasian) analysis of variance (ANOVA) was used to examine the main effects and interactions on BMC, height, weight, lean mass, fat mass, calcium and physical activity within parent and child groups. In a secondary analysis, height, lean mass and age were used as covariates to determine if differences in BMC persisted when body size and composition were controlled for. To assess the familial resemblance of BMC, lean mass, fat mass and lifestyle variables, Pearson correlations were run for all parent-child pairs (mother-daughter, mother-son, father-daughter, and father-son). Standardized residual scores for TB BMC were calculated within each generation, sex, and ethnic group adjusting for age, height, and lean mass. Variables were chosen based on their theoretical association with TB BMC and were significantly related to TB BMC in at least one of the groups. For consistency, the same variables were entered in regression equations for all groups. To explore potential ethnic differences in parent-child associations, mother-child and father-child correlations were run within each ethnicity. Due to the small sample size, ethnic differences were not explored separately by sex of the child.  Results Descriptive characteristics All of the boys and 86% of the girls were Tanner stage 1. Five of the 35 girls were self-reported Tanner breast stage 2. Mean (SD) values for all variables are listed in Table 1. Estimated calcium intake was significantly (p < 0.05) lower for Asian than Caucasian children (765±626 vs. 1021±454 mg/d). Asian boys reported participating in less physical activity compared to Caucasian boys, while there was no difference in activity between the girls. Asian children tended to be shorter, weigh less, and have lower total lean mass than Caucasian children. However, %fat and %lean values were not significantly different between groups. Total body BMC was significantly lower in girls than boys (p < 0.01), and lower in Asians than Caucasians (p = 0.004). Differences became non-significant after controlling for height and lean mass. The sex by ethnicity interaction was not significant for TB BMC (p = 0.868). Children's height, weight, lean mass, fat mass and BMC, were expressed as a percentage of their parents' values. On average, children (mean age 8.9 y) had reached -80% of parental height, but only -45% of parental weight, lean and fat mass. Children's total body BMC was -40% of parental values. Descriptive variables for parents are provided (Table 2). Caucasian parents tended to be taller, weigh more and have greater fat mass and higher calcium intakes than Asian parents. Hours of weight-bearing physical activity (averaged from the past two weeks) did not differ between Asian and Caucasian parents. However, fewer Asian parents reported participating in regular physical activity both during childhood (Asian 40%, Caucasian 76%) and adulthood (Asian 17%, Caucasian 61%). Sex by ethnicity interactions were significant for BMC and lean mass which was explained primarily by ethnic differences within fathers. Unadjusted TB BMC was 30% lower in Asian compared with Caucasian fathers, and 10% lower in Asian mothers compared with their Caucasian counterparts. After adjusting for lean mass, height and age, BMC values were still -20% lower in Asian than Caucasian men, but only 1% different for the women. Familial resemblance Parent-child correlations for TB BMC, height, weight, lean and fat mass are shown (Table 3). Daughters closely resembled their mothers for height, weight, lean and fat mass. Similarly, fatherdaughter correlations for weight, lean and fat mass (but not height) were also close to 0.50, however only weight was significant (p = 0.042). Unadjusted TB BMC correlations approached significance for mother-daughter (p = 0.054), but not father-daughter (p = 0.369) pairs. Mother-daughter and fatherdaughter correlations were not significant for adjusted TB BMC scores. There was a trend for sons to more closely resemble their fathers than their mothers for weight, lean and fat mass although none of the correlations reached significance. There was also a resemblance between sons and both parents for unadjusted TB BMC values. After adjusting TB BMC within each group, these associations decreased and became non-significant. When pairs were split by ethnicity, only the mother-child height association within Asians was statistically significant (r = 0.56, p = 0.019). There were trends for stronger mother-child correlations within Asian pairs for weight (r = 0.43, p = 0.096), lean mass (r = 0.476, p = 0.062), and unadjusted TB BMC (r = 0.42, p = 0.103). In contrast, father-child associations, although not significant, tended to be stronger for Caucasian than Asian pairs for weight, lean and fat mass (Caucasian, r = 0.42-0.44 vs.  Asian, r=-0.12-0.03) but not BMC. Adjusted BMC values were not significant for any of the pairs in either ethnic group. There was no relationship between children's and parents' calcium intake, or between children's and parents' physical activity scores. Discussion This is the first report examining familial resemblance of total body bone mass in a multiethnic group of mothers and their prepubertal sons, and in fathers and their prepubertal sons and daughters. We found evidence of familial resemblance in TB BMC, as well as lean and fat mass, between these Asian- and Caucasian-Canadian children and their parents. Familial resemblance for total body bone mass was not apparent after accounting for lean mass and body size. These data support the emerging notion that heritability of bone mass is complex and interacts with variables such as age, sex, ethnicity, body size and composition [8,12]. A number of previous studies have shown associations for regional (lumbar spine; femoral neck) bone mass or density between adolescent girls [13] or adult women [14,15] and their mothers. There is only one report of familial resemblance for regional BMC and areal BMD (aBMD, g/cm) in prepubescent girls and their mothers [7]. By way of comparing our results to these studies, we will address a number of factors which may influence the familial resemblance of bone including; 1) age of both parents and offspring, 2) the relationship between lean mass and bone, and 3) ethnicity of the population. 2  Age  Relationships in bone mineral appear to be stronger for adult offspring and their parents than those for young or elderly offspring and their parents. Estimates of heritability are higher for motherdaughter pairs if both members are premenopausal as opposed to postmenopausal, and correlations decrease with increasing postmenopausal age of mothers [14,16]. In younger age groups, our data indicate a weaker familial resemblance between prepubertal children and their parents. As compared to early adulthood, bone mineral content or density is -10-30% lower both prior to the attainment of peak bone mass (prepubertal) and during the postemenopausal years when bone mass is diminishing. It is plausible that familial comparisons would be more similar around peak bone mass than either before or after peak when individuals are gaining or losing bone at variable rates. In a slightly older cohort (11.8±2.1y), McKay et al. [13] reported significant correlations (r = 0.31-0.49) for aBMD between daughters and their premenopausal mothers (mean age 40.0y). The girls' proximal femur and lumbar spine aBMD values were, on average, 83-93% of their mothers' values. The children in our study were -3 years younger and had reached 70% of their parents' values for total body aBMD, and 40% of their parents' TB BMC. Ferrari et al. [7] reported significant correlations for proximal femur, lumbar spine and midfemoral shaft BMC and aBMD in 138 prepubertal Swiss girls (mean age 8.1 y) and their mothers (mean age 40.0y). However, the reported heritability estimates of -18-36% for BMC and aBMD (adjusted for height, weight, age, and calcium intake) are -20% lower than those reported for adult daughters and their mothers [15]. Correlation coefficients for lumbar spine and femoral neck BMC and aBMD (standardized for age only) were 0.25-0.30 [7]. Caucasian mothers and daughters in our study were similar in height and weight to the Swiss sample. It is possible that the differences in familial  resemblance between the studies is due to population differences. The Swiss population may be more homogenous than our heterogeneous population from Europe, North America, and Asia. It is difficult to compare the two studies as variables adjusted for differed. However, the correlations for unadjusted TB BMC between mothers and daughters in our study were similar to the regional correlations adjusted for age [7]. Ferrari and colleagues adjusted for weight, but not lean mass, before predicting heritabiity estimates. Evidence for an interaction of heredity with lean and bone mass is discussed below. The present study suggests familial resemblance in bone mineral may be lower, or nonsignificant, before pubertal growth begins. This is supported by a study from Gueguen et al. [17] that predicted heritability estimates for bone mass peak at age 26, and are low or non-significant prior to age 8 and after age 44. Taken together, these data support a stronger resemblance in bone mass during the adult years when bone mass of both family members is at or near its peak, and a weaker association during prepubertal or postmenopausal years. Association between bone and lean mass It is well accepted that body weight and size influence DXA measures of BMC and aBMD. Relatively new DXA body composition software, has made it possible to accurately distinguish the independent contribution of lean and fat components of body weight to bone mineral. Although data in prepubertal children are limited, bone-lean mass associations in children and adolescents are higher than bone-fat mass associations [8,18]. In a cohort of 215,10- to 26- year old female twins, the genetic variance decreased by half in the early-adolescent years when lean mass was accounted for [8]. In the present study, univariate correlation showed a higher association between lean mass and total body bone mass (r = 0.80-0.94) than between fat mass and bone (r = 0.57-0.64). A strength of the present study is that we have controlled for differences in lean mass, whereas most studies adjust for age and/or weight alone. Ethnicity To our knowledge, there are no other studies that compare familial resemblance in bone mass within Asian-Canadian family groups; or any ethnicity other than Caucasian. We found that Asian and Caucasian families had varying degrees of resemblance for bone, fat, and lean mass. These data must be interpreted cautiously given the small sample size. However, it is plausible that emigration from Asia to Europe or North America leads to alterations in lifestyle that, in turn, influence body size, bone mass or both. For example, estimated calcium intakes were lower for Asian than Caucasian children and parents in our study, but children's values were higher than the average of <500 mg/d reported for individuals living in Hong Kong [19]. Asian children and parents also reported participating in less physical activity than Caucasians. Other studies have shown differences in adult bone mass with emigration. Premenopausal Chinese women who spent more than 12 years in Denmark had aBMD values similar to Danish women, while those Chinese women living in Denmark for less than 12 years had 4-7% lower aBMD at all sites [20]. Similarly, after adjusting for height and weight, U.S. born Japanese women had 4-5% higher calcaneal and radial BMD than women living in Japan [21]. In an adolescent cohort living in California, after body size was accounted for, there were virtually no differences in bone mass between Asian and Caucasian groups [22]. In our study, ethnic differences in bone mass were apparent for fathers even after adjusting for these variables. In contrast, TB BMC was similar between sons after  adjusting for lean mass and height. The Asian families in our study had recently moved to Canada 3 years previously on average. These cultural and lifestyle factors together may serve to influence the genetic template for bone mass/density. In summary, these data show familial resemblance for total body bone mass in prepubertal children is influenced by lean mass. Our data also suggest that ethnic differences in heritability of bone mass may be influenced by cultural differences in physical activity and calcium intake. Future studies are warranted to further examine ethnic and sex differences in heritable influences on bone and lean mass and their interactions on the pediatric skeleton.  Acknowledgements This research was supported by a grant from British Columbia Health Research Foundation. We are grateful to Dr. Ari Heinonen and Dr. Karim Khan for their review of this manuscript, and to Dr. Robert Schutz for statistical advice.  References 1. Eisman JA, Sambrook PN, Kelly PJ, Pocock NA. Exercise and its interaction with genetic influences in the determination of bone mineral density. Am J Med 1991 ;91:5S-9S. 2. Kelly PJ, Harris M. Genetic regulation of peak bone mass. Acta Paediatr Suppl 1995;411:24-29. 3. Bradney M, Pearce G, Naughton G, et al. Moderate exercise during growth in prepubertal boys: Changes in bone mass, size, volumetric density, and bone strength: A controlled prospective study. J Bone Miner Res 1998;13:1814-1821. 4. McKay HA, Petit MA, Schutz RW, Prior JC, Barr SI, Khan KM. Augmented trochanteric bone mineral density after modified physical education classes: A randomized school-based exercise intervention study in pre- and early-pubescent children. J Pediatr 1999;accepted: 5. Morris FL, Naughton GA, Gibbs JL, Carlson JS, Wark JD. Prospective 10-month exercise intervention in pre-menarcheal girls: Positive effects on bone and lean mass. J Bone Miner Res 1997;12:1453-1462. 6. Malina RM, Mueller WH, Holman JD. Parent-child correlations and heritability of stature in Philadelphia Black and White children 6 to 12 years of age. Hum Biol 1976;48:475-86. 7. Ferrari S, Rizzoli R, Slosman D, Bonjour JP. Familial resemblance for bone mineral mass is expressed before puberty. J Clinc Endocrinol Metab 1998;83:358-361. 8. Hopper JL, Green RM, Nowson CA, et al. Genetic, common environment, and individual specific components of variance for bone mineral density in 10- to 26-year-old females: A twin study. Am J Epidemiol 1998;147:17-29. 9. Bailey DA. The Saskatchewan Pediatric Bone Mineral Accrual Study: Bone mineral acquisition during the growing years. Int J Sports Med 1997;18:S191-S194. 10. Barr SI. Associations of social and demographic variables with calcium intakes of high school students. J Am Diet Assoc 1994;94:260-266,269. 11. Crocker PRE, Bailey DA, Faulkner RA, Kowalski KC, McGrath R. Measuring general levels of physical activity: Preliminary evidence for the Physical Activity Questionnaire for older children. Med Sci Sports Exerc 1997;29:1344-1349. 12. Matkovic V. Nutrition, genentics and skeletal development. J Am College Nutr 1996;15:556-569. 13. McKay HA, Bailey DA, Wilkinson AA, Houston CS. Familial comparison of bone mineral density at the proximal femur and lumbar spine. Bone Miner 1994;24:95-107. 14. Danielson ME, Cauley JA, Baker CE, et al. Familial resemblance of bone mineral density (BMD) and calcaneal ultrasound attenuation: The BMD in mothers and daughters study. J Bone Miner Res 1999;14:102-110. 15. Krall E, Dawson-Hughes B. Heritable and life-style determinants of bone mineral density. J Bone Miner Res 1993;8:1-9. 16. Hansen MA, Hassager C, Jensen SB, Christiansen C. Is heritability a risk factor for postmenopausal osteoporosis? J Bone Miner Res 1992;7:1037-1043. 17. Gueguen R, Jouanny P, Guillemin F, Kuntz C, Pourel J, Siest G. Segregation analysis and variance components analysis of bone mineral density in healthy families. J Bone Miner Res 1995;10:2017-2022. 18. Nelson DA, Barondess DA. Whole body bone, fat and lean mass in children: comparison of three ethnic groups. Am J Phys Anthropol 1997;103:157-162.  19.  Lau EMC. The epidemiology of hip fracture in Asia: An update. Osteoporosis Int 1996;3:S19S23. 20. Wang Q, Ravn P, Wang S, Overgaard K, Hassager C, Christiansen C. Bone mineral density in immigrants from southern China to Denmark. A cross-sectional study. Eur J Endocrinol 1996;134:163-167. 21. Nomura A, Wasnich R, Heilbrum L, Ross P, Davis J. Comparison of bone mineral content between Japan-bom and US-bom Japanese subjects in Hawaii. Bone Miner 1989;6:213-223. 22. Bhudhikanok GS, Wang M, Eckert K, Matkin C, Marcus R, Bacharach LK. Differences in bone mineral in young Asian and Caucasian Americans may reflect differences in bone size. J Bone Miner Res 1996;11:1545-1556.  Table 1. Descriptive characteristics for Asian and Caucasian daughters and sons. Mean (SD) Sons Daughters Asian  Caucasian  Asian  Caucasian  N  7  Age (years)  9.0 (0.6)  14 8.8 (0.5)  9 8.9 (0.6)  19 8.7 (0.8)  129.4(4.7)  136.8 (8.2) *  134.1 (4.7)  136.0 (7.4)*  Height (cm) Weight (kg)  25.1 (1.9)  Fat (g)  5010(962)  8382(3980)  7456 (2725)  7650 (5550)  19833 (1357)  23969 (3188)«  22499 (2365)  25842 (5283)«  752 (119)t  835 (92)  942(190)  879  870  Lean (g) TB BMC (g)  32.6 (6.6) *  30.2 (4.8)  33.8 (10.7)  x  TB BMC*  862  848(85)* 851  %fat  19.5 (2.3)  24.0 (7.8)  23.5 (6.2)  20.2 (8.0)  73.3 (7.5)  73.7(5.9)  77.1 (7.7)  74.2 (13.2) 68.0 (8.1) 70.1 (25.8) Physical activity *TB BMC values are adjusted for height and lean mass "significantly greater than Asian children (p<0.05), ** (p< 0.01) tsignificantly lower than sons (p = 0.01) Significant ethnicity by sex interaction (p = 0.032)  89.6 (15.4)  %BMFL  r  77.6 (2.9)  xx  <  CO  o  CO  <  cr z> 3  , S.CO  co r o  f ? S  - CQ  CD  «>  ro —» CD  -Z2  1  CO QJ  ro co —*•  CD CO  S  OO  »  CQ  CQ  O  c/>  CD CO  =zi-  CQ  - co  7=r  v |  O)  bo CJ> bo —»• w co A co "-vi co cn co  O  < CD  » r2 Mw SP 10 01 K  ai  co  v  5> r o co  CO S  O ZT CO co o  Co'  »—  C D ^  CD  GO'  o'  ro ro J> ro CD CD  CD  CO  O 0) cz  ro N ro ro CD - v i , „ "oo "CD ro cn b cn bo  co bo £ P t O ^ O) CD co o co  > CO  c 2 3 J 2 - b  Q.  co' 3 CO 3  CO co'  O  CD r o ^ - ro r o -vi co co p J> co bo CO  vj  m  *1>  o  CO CO bo  ro ro  ro CD —1 ^ —' r o CD J> _1 ro o . , co co 0  1  CD  co cz o co CO  co"  j>  3  3  o Ex  J^-J _bo .CD "~x ~~><  CD  55 CO 3  - J ro cn —1. J> ro ro cn bo bo co 0 ro _. _ _ _ - CO ro L cn ro 0 CD  -g  CD ro co  jj, 9*  r°  co co  CO  01 01 j  w  M  J> J>  ^  00 co  CD CO  r o co co ro cn 3 -  ro S ro co . o o ^ co "cn Zl "oo co co co  ro CD  o 00 ro r to fo  co'  O  -P- r o - £ -r- -rz; co co CD 4^ cn  A  00  co cz o -n co. oV 0 5  §  ro a> CO —  g  00 01 -vj ^ A  •  CD ,ro j o .tn  CD CO  r o co —^  00 ^ b  3-  CO  a , J>  0  a.  >  —  ^  cf'  Table 3. Pearson product moment correlations comparing total body BMC, fat and lean mass, height M-D 21 .426 .156  M-S 28 .439* .277  F-D 14 .354 -.260  F-S 14 .584* .365  n TB BMC TB BMC residual* Height .481 * .064 .016 -.097 Weight .538" .025 .550* .455 Lean .537* .310 .454 .512 Fat .478 .204 .508 .390 *residual scores adjusted for age, height, and lean mass within each sex, generation and ethnic group *p<0.05 x  IgO  APPENDIX 2 Consent Forms  181  T H E  UNIVERSITY  OF BRITISH  C O L U M B I A  School of Human Kinetics 210, War Memorial Gym 6081 University Boulevard Vancouver, B.C. Canada V6T 1Z1 Tel: (604) 822-3838.  Appendix 1  Fax: (604) 822-6842  JUMP ROPE STUD Y INFORMATION TO FAMILIES  To The Family: A group of investigatorsfromthe University of British Columbia are beginning a study involving grade four and five children titled "The Effect of High Impact Loading on the Growing Skeleton". The Vancouver/Richmond Board of Education has granted approval for the study and [School names] Elementary Schools have been selected to take part in this investigation. The principals and staff at the schools have endorsed this project and are making available the time to allow children to participate. The primary purpose of the study is to investigate the effects of adding 10 minutes of jump rope activity to Physical Education classes on bone development in growing children. Also, to examine the role of nutritional patterns as they relate to bone development. The results of this study will establish whether increased physical activity and specific nutrient intakes should be recommended for growing children to optimize bone development and to reduce the risk offractureand osteoporosis in later life. This will be thefirstexercise intervention study in growing children. The results will provide important information regarding the role of physical activity and proper nutrition during the pre-teen years in the prevention of osteoporosis, a very prevalent and debilitating bone disease of older adults. Children who are involved in the study will have their bone status and growth and development evaluated two times over the year. Questionnaires concerned with physical activity and nutritional patterns of the children will be administered twice over the year. At each elementary school, one of the physical education classes will add 10 minutes of jump rope activity to their current curriculum, while the other classes will continue their regular program. The total time cornrnitment for your child outside of regular school time will be approximately 6 hours over the course of the year for bone measurements at the Vancouver Hospital and for completing questionnaires (see the attached "consent form" for details). Your child will be transported by minibus to Vancouver Hospital three times over the year and supervised en route by a study staff person in addition to the driver of the rriini-bus. If your child would like to participate in this study and undergo the measurements, we ask that you and your child sign the attached consent form and return it in the stamped, self-addressed envelope. Should you have any questions about this study please contact Dr. Heather McKay (822-3120) or Moira Petit (822-4281) at the University of British Columbia. Thank you for your interest in this study. We look forward to hearing from you. Sincerely, Heather McKay, Ph.D. & Moira Petit  THE  UNIVERSITY OF BRITISH  COLUMBIA  School of Human Kinetics 210, War Memorial Gym 6081 University Boulevard Vancouver, B.C. Canada V6T 1Z1 Tel: (604) 822-3838  Appendix 1 Pg 2  Fax: (604) 822-6842  JUMP ROPE STUD Y  CONSENT FORM FOR FAMILIES Procedures: Your child's participation in this project will involve two testing sessions (approximately one hour each) at Vancouver Hospital. Each session will include the following procedures: 1. Measures of height and weight will be taken. In addition your child will be asked to complete questionnaires that will provide physical activity, and dietary information. A parent will be given instructions on how to complete a three-day food diary that will provide information regarding your child's calcium and nutritional intake and a brief health history questionnaire. 2. Your child's whole body bone status will be evaluated by a bone densitometer and the bone density of the left hip and lumbar spine (lower back) will be measured. This procedure is painless and routinely used in the practice of modern medicine and there is a minimal radiation exposure. The total exposure per session will be less than 10 millirem which is similar to the background radiation one would receive making a one-wayflightfrom Vancouver to Halifax on a commercial airline. For comparative purposes, the average annual background radiation in Vancouver due to natural sources is around 150 millirem per year and the current permissible level for the general population is 500 millirem per year. These values can be used to compare the relative risk of the less than 10 millirem exposure from the bone density procedure. The typical exposure is less than the variability in annual background exposure in the Vancouver/Richmond area. All bone density measurements will be conducted in the Department of Nuclear Medicine at Vancouver Hospital and will be aclministered by qualified hospital technologists. About 20 minutes is required for all the bone measurement procedures. Rights and Welfare of the Individual: It is understood that you arefreeto withdrawfromany or all parts of the study at any time without penalty. Your child's identity will remain confidential and only those directly involved in the study (namely the investigators and Radiology staff) will have access to your child's records and results. All individual results will remain confidential. Please be assured that you may ask questions at any time. We will be glad to discuss your child's results with you and your child when they have become available and we welcome your comments and suggestions. Should you have any concerns about this study or wish further information please contact Dr. Heather McKay (822-3120) or Moira Petit (822-4281) at the University of British Columbia. If you have any concerns about your child's treatment, please contact Dr. R. D. Spratley at the office of Research Services and Aajrunistration at UBC (8228595).  193  THE  UNIVERSITY  OF BRITISH  C O L U M B I A  School of Human Kinetics 210, War Memorial Gym 6081 University Boulevard Vancouver, B . C . Canada V6T 1Z1 Tel: (604) 822-3838  Fax: (604) 822-6842  JUMP ROPE STUD Y  Appendix 1 Pg. 3  RESEARCH PROJECT CONSENT FORM Parent's Statement: I,  (please print the name of one or both parents) understand the purpose and procedures of this study as described and I voluntarily agree to allow my child to participate. I understand that at any time during the study we will be free to withdraw without jeopardizing any medical management, employment or educational opportunities. I understand the contents of the consent form, the proposed procedures and possible risks. I have had the opportunity to ask questions and have received satisfactory answers to all inquiries regarding this study.  Signature of Parent/Guardian  Date  Signature of Witness  Date  Signature of Investigator  Date  Child's Statement: I understand the purpose and procedures of this study as described and I voluntarily agree to participate. I understand that at any time during the study, I will be free to withdraw without jeopardizing any medical management, employment or educational opportunities. I understand the contents of the consent form, the proposed procedures and possible risks. I have had the opportunity to ask questions and have received satisfactory answers to all inquiries regarding this study. Signature of Child  Date  THE  UNIVERSITY  OF BRITISH  C O L U M B I A  School of H u m a n Kinetics 210, War Memorial G y m 6081 University Boulevard Vancouver, B . C . Canada V 6 T 1Z1 Tel: (604) 822-3838  Appendix 2  Fax: (604) 822-6842  JUMP ROPE STUD Y INFORMATION TO SCHOOLS (Principals and Teachers)  Dear Colleague: A group of investigators from the University of British Columbia are beginning a study involving grade four and five children titled "The Effect of High Impact Loading on the Growing Skeleton". The Vancouver/Richmond Board of Education has granted approval for the study and we are looking for three Elementary Schools to participate in the study. The primary purpose of the study is to investigate the effects of adding 10 minutes of jump rope activity to Physical Education classes on bone development in growing children. Also, to examine the role of nutritional patterns as they relate to bone development. The results of this study will establish whether increased physical activity and specific nutrient intakes should be recommended for growing children to optimize bone development and to reduce the risk of fracture and osteoporosis in later life. This will be thefirstexercise intervention study in growing children. The results will provide important information regarding the role of physical activity and proper nutrition during the pre-teen years in the prevention of osteoporosis, a very prevalent and debilitating bone disease of older adults. Children who are involved in the study will have their bone status and growth and development evaluated two times over the year. Questionnaires concerned with physical activity, and nutritional patterns of the children will be administered twice over the year. At each elementary school, we will be developing a program for physical education instructors to implement into the current curriculum. The program will consist of adding 10 minutes of jump rope activity as a warmup prior to, or in conjunction with, exisitng physical education programs. Two classes at each school will be involved with one class implementing the added jump rope activity, and the other class completing their normal physical education program: Children will be transportedfromthe schools by minibus to Vancouver Hospital three times over the year (in Oct. '97, & June '98) and supervised en route by a study staff person in addition to the driver of the mini-bus. This will require that children be releasedfromapproximately 2 hours of class time in October and June for hospital measurements. If your school would like to participate in this study, please sign the attached consent form. You will receive a copy of the consent form for your own records. A detailed proposal of this study is available to you and you will have access to a copy of the completed study. Should you have any questions please contact Dr. Heather McKay (8223120) or Moira Petit (822-4281) at the University of British Columbia. Thank you for your interest in this study. We look forward to hearingfromyou. Sincerely, Heather McKay, Ph.D. & Moira Petit  THE  UNIVERSITY OF BRITISH  C O L U M B I A  School of Human Kinetics 210, War Memorial Gym 6081 University Boulevard Vancouver, B.C. Canada V6T 1Z1 Tel: (604) 822-3838  Fax: (604) 822-6842  JUMP ROPE STUDY  Appendix 2 Pg. 2  Heather McKay PhD, Alan Martin PhD, Moira Luke PhD, Moira Petit MSc Department of Human Kinetics UBC  Consent Form for Principals/Teachers It is understood that you are free to withdrawfromany or all parts of the study at any time without penalty. Your schools identity, and that of the children involved in the study, will remain confidential. Please be assured that you may ask questions at any time. We will be glad to discuss any aspect of the study with you and your staff/colleagues. Should you have any concerns about this study or wish further information please contact Dr. Heather McKay (822-3120) or Moira Petit (822-4281) at the University of British Columbia. If you have any concerns about the childrenis treatment, please contact Dr. R. D. Spratley at the office of Research Services and Administration at UBC (822-8595). Teacher/Principal's Statement:  understand the purpose and procedures of this study as described and I voluntarily agree to allow my school/class to participate. I understand that at any time during the study we will befreeto withdraw without jeopardizing any employment or educational opportunities. I understand the contents of the consent form, the proposed procedures and possible risks. I have had the opportunity to ask questions and have received satisfactory answers to all inquiries regarding this study and wish to volunteer as a participant.  Signature of Teacher/Principal  Date  Signature of Witness  Date  Signature of Investigator  Date  T H E UNIVERSITY OF BRITISH  COLUMBIA  School of Human Kinetics 210, War M e m o r i a l G y m 6081 University Boulevard Vancouver, B . C . Canada V 6 T 1Z1 Tel: (604) 822-3838  Fax:(604) 822-6842  FAMILY BONES STUDY CONSENT F O R M  Purpose of the study: The primary purpose of this study is to examine lifestyle and genetic factors related to bone density within families of varying ethnicities. Research has suggested that exercise, calcium and genetics are the primary factors influencing bone health. To better understand the genetic-environment relationships, we will be examining the relationship of calcium and exercise to bone mass. There is some evidence that suggests that the prevalence of "osteoporosis prone" gene(s) may differ in various ethnic groups, thus we will be examining these genes as well. Study Procedures: Your participation in this project will involve one testing session (~lhr. long) at Diagnostic Care Radiology Clinic in Vancouver on West Broadway. The testing session will include the following procedures: 1. Measures of height and weight will be taken. In addition you will be asked to complete questionnaires that will provide health, physical activity, and dietary information. You will also be given instructions on how to complete a food diary that will provide information regarding your calcium and nutritional intake. 2. Your bone status will be measured by a bone densitometer. This procedure is painless and routinely used in hospitals and there is a minimal radiation exposure. The total exposure per session will be less than 10 millirem which is similar to the background radiation one would receive making a one-way flight from Vancouver to Halifax. The average annual background radiation in Vancouver due to natural sources is around 150 millirem per year. The total exposure currently allowed for the general population is 500 millirem per year . These values can be used to compare the relative risk of the less than 10 millirem exposure from the bone density procedure. All bone density measurements will be conducted in the Department of Nuclear Medicine at Vancouver Hospital and will be administered by qualified hospital technologists. About 20 minutes is required for all the bone measurement procedures. 3. As an optional, cheek smears will be collected that will allow us to identify a gene/genes that might be linked with bone density. This procedure involves the measurer passing a cotton (Q Tip-like) swab over the inside of your cheek. The DNA extractedfromthe tissue sample will be identified by a number and analyzed for the presence of bone density-related genes ONLY. Results will be available to Dr. McKay and Moira Petit by subject number only.  APPENDIX 3 Sample Activities from the Healthy Bones Curricul  A c t i v i t y : J U M P r u e SHOT  Aciivitg £Zme: 15 MiNUTeS . Five L O N G J U M P Ropes • five BeaN B S G S • aN iNPOOR OR GYMN3SHJM SPace •bjectfvas: •  •  TO pRovni^ STuDeNT5 wiTfi a N appROPRiaTe c a R D i o v a s c u L a R waRM-up F O R HiGHeR iMP3CT RUNNING 3NP JUMPING aCTiVTneS. . TO PROViDe STUPeNTS WITH Tfie OPPORTUNiTV TO eXPLORe 3ND PRSCTiCe Tfie SKiLLS O F TiMiNG 3ND aNTICiPaTiON.  • cReaTe Tfie 'snov B Y TYING a BeaN BaG TO ONe eND ' O F a cne TWO Ropes TOGerHeR TO app L C N G T H if rr is Neepep). ^"V .  LONG JUMP  • BReaK t H e . o a s 8 . - u p INTO F O U R O R Five G R O U P S O F S T U D € N T S , a •SflOT'TO WORK WiTfl 6 N P 3 N 3Rea OF Tfie G Y M TO P L 3 Y iN.  GIVING  rope,  eacn  GROUP  • seLecTONe STuPeNf IN e a c f i G R O U P T O Be'-Tfie S H O T tuRNeR. THIS INPIVIPUSL H O L P S Tfie eNP O F T f i e Rope IN H 3 N P 3 N P , B C N P I N G P O W N , T U R N S O N Tfie S P O T T O PR3G THe SflOT IN 3 CIRCLe 3L0NG Tfie GROUNP. T f i e ReMSiNiNG STUPeNTS iN eacfi G R O U P STep INTO THIS ciRCLe aNP sTreMpr TO J U M P oveR Tfie SflOT ss rr . COMeS T 0 W 3 R P TfieM.  •  wfieN a STupeNT Toucfies O R J U M P S O N Tfie Rope, TfieY acouiRe THe FIRST  i-erreR TO T f i e W O R P 'SKeLeTON'. s t u p e N t s GaiN sNOTHeR Le-rreR e a c f i nMe TfieY Miss O R L S N P O N Tfie Rope. wfieN TfieY mve coLLecrep S L L Tfie teTreRS TO spen 'SKeLeTON' TfieY BecoMe Tfie New ROPe/sfiOT T U R N C R .  Teacftbtg  fipsi  • seT THe RuLes •  cfiaNGe  FOR  we RuLes  Tfie G3Me BeFORe BResKiNG STUPeNTS INTO OF  GROUPS.  Tfie GaMe T O V S R Y Tfie acwiTY F O R S T U P C N T S :  G3Me 1: STUPeNTS MUST JUMP WITH BOTH FeeT TOGeTfieR (i.e. HOPPING). G3Me 2: STUPeNTS MUST HOP OVeR THe ROPe ON THeiR R FOOT ONLY (THeiR LeFT FOOT M3Y NOT TOUCH THe GROUNP PuRiNG THe G3Me). G3Me 3: S3Me aS G3Me NUMBeR TWO BUT STUPeNTS HOP O N THeiR L FOOT. Challenge iNCRease we PIFRCULTY O F we exercise B Y H S V . N G we Rope T U R N C R S speep U P THeiR R0T3T10N aNP/OR ST3RT TO TURN THe SHOT 3 COUPLe O F iNCHeS OFF THe GROUNP. CHaLLeNGe STUPeNTS FuRTHeR BY 3PPING 3 SeCONP SHOT IN escn G3Me (HeLP IN THe TURNeR'S OTHeR H3NP). NOW STUPeNTS MUST aNTOPaTe THe 3RR5V3L OF TWO MOVING OBJeCTS.  1^0  ju«vi*j£  FROG IN Tfie 5 e a  A c t i v i t y l i m e : 10-15 MiNUTeS  £qwipm«itt: •  •  ONe TUMBLING- M3T C3N B e PLayep IN S N INDOOR O R OUTDOOR  space  Objectives: ' •  TO PROViDe STUDeNTS WITH aN INTRODUCTION TO HIGHeR iMpacr weiGHT BeaR>Ne acnviTY BY eNCouRaG>NG THe D e v e L O P M e w O F L e s STRGNGTH.  A c t i v i t y ^ascription: •  ONe OR TWO STUDeNTS 3Re CflOSeN TO Be. FROGS 3ND S«T IN Tfie ceNTRe OF a C.RCLe FORMeD BY Tfie ReST OF THe CL3S5 (THe STUDeNTS FORMING Tfie CiRCLe aRe GRasswoppeRS).  •  puace a TUMBLING M3T a Few MeTeRS ouTsiDe Tfie eDGe OF m e ciRCLe. IT WILL Be a fiOMe-FRee OR saFe z o N e FOR Tfie GaMe.  •  FROGS ReMaiN IN Tfie ceNTeR O F m e ciRCLe wniLe we GRassnoppeRS H O P CLOCKWISe a ROUND TfieM OH3NT1NG * FROG iN Tfie Sea, CaN'T C3TCH M e !' UNTIL ONe OF Tfie FROGS CL3PS THeiR fl3NDS.  •  WfieN a FROG cLaP5 TfieiR fiaNDs, TfieY BeG'N TO cfiase Tfie GRsssfiOPpeRS WHO M3Y BReaK OUT OF Tfie GRCLe aND fieaD FOR Tfie MaT wfieRe TfieY' 3Re S3Fe.  .  3NY GRSSSflOPPeRS WHO aRe T3GGeD BeFORe ReaCHiNG THe M3T JOiN THe FROGS 3ND 3 New CiRCLe iS FORMeD WITH 3LL FROGS iN THe MiDDLe. THe CHaNTiNG 3ND cnasiNG ResuMes S N D IS RepeaTeD UNTIL ONLY ONe GRassHOppeR ReMa.NS.  .  3LL PaRTiCiPaNTS MUST MOVe ONLY BY HOPPiNG. FROGS 3ND GRSSSHOPPeRS HOP ON 3LL FOURS iN 3 CROUCHeD POSiTiON.  .  New FROGS C3N B e CHOSeN 3ND THe G3Me PL3YeD 3GaiN.  Teaching T i p s : •  EMPHaSiZe TO 3LL PaRTiCiPaNTS THe iMPORTaNCe OF NOT CHeSTiNG BY TRYING TO GeT UP 3ND SHuFFLe OR RUN. IT iS IMPORTaNT TH3T STUDeNTS HOP CROUCHeD DOWN FOR THe DuRSTiON OF THe G3Me.  "  THey'^FORM^a^ Ne^w^c'^cL^ ' T  H  a  V  e  S  T  u  D  e  N  T  S  S  T  0  P  F  0  R  a  Q o ; c K  STRetcH eacn  r^e  "30^  io MiNutes Ba^manfe N 0 N 6 ,  r- ,  sftdes)  -c .  «-.: >./ >'.*»•-•-•  '  • a- V • Keif,  •if  . THe^iRcurr ;OF ttiMpacT istaTioNS <s set UP aRouNP THe oursipe OF a GYMNasiuM (see statiON OPTIONS, eei-owx !  . pivipe Tfie cuass I N T O STanoN. •  e v e N GROUPS, ONe GROUP FOR  eacn ; "  THe aeRQftc Music -rape pRovipep FOR THe STUDY, seT UP a tape Pi-aveR IN a saFe aRea OF THe GYM. Keep THe MUSIC PLaYiNG FOR THe puRanoN OF THe CIRCUIT. USING  • stuPeNTS speNP 3 0 secoNPS ar eacn iMPacT sra-noN. O N cue FROM THe JNSTRUCTO^QTuPeNTS FiN.SH.THe aCTIVITY 3T THeiR " ~ • • ' sratiON aNP MaRcH INTO THe MiPPLe; OFkne GYM FOR RecoveRYr , . AGaiN ON Cue „FROM" THe INSTRUCTOR, STUPeNTS MaRCH OUT TO TH©* Nexr sranoN TO'BeGiN a New PaNce aenvrrY, ....... • STUPeNTS coNTjNue TO speNP 3 0 secoNPS aenve .ar each sianoN, MOVING aRouNP THe CIRCUIT IN a couNTeR cLocKwise piRecnoN. RecoveRY peRioPS SHOUUP LSST I - 1 j/2*MiNuTes eacH.  • ••Wfaf-'A-VgS •  IMPaCT  ST3T.ONS:  ; ^ : -  1) AeROBic KNees UP 2) HaMSTRING KICK BaCKS WITH POWeR BURST 3) JUMPING J3CKS  4) LUNGeS aLTeRNaTiNG LeGS  5) AeROBic KNees UP WITH CROSS IN FRONT 6) JUMPS 7) SKI JUMPING SKI JUMPING LaTeRaLLY Impact8)Emphasis:  neeL  => eNcouRaGe STUPeNTS TO FOCUS ON GemNG THeiR KNees upTO iNCRease THe iMpacT tHeY aRe Geir.NG. => eNSuRe T H a T STupeNTS Keep MaRcniNG FOR THe RecoveRY! peRiop - rr;jg jMpeRaTive THST ^rppeNTS Keep MOVING.  juttvftg: captu Re tfie FLae Activity Time: 15-30 MiNUTeS Equipment: » 30 PINNY'S IN JWO COLORS i • 2 BeaN-B3GS, 2 HOOPS 3ND 2 M3T5  Objectives: TO PRoviDe paRTiOpaNTs WITH aN INTRODUCTION TO iNvasive TeaM eaMes.  •  •  o DeveLOP STUDeNTS' SBILITY TO u s e S N D UNDCRSTSND Tfie coNceprs OF peRSONaL aNi> eeNeRaL space IN TeaM eaMes.  T  Activity 9escriptleit: •  BeFORe Tfie ST3RT OF CL3S5, PULL OUT Tfie eQUIPMeNT LISTeD aBOVe 3ND seT UP Tfie GYM (OR PL3YING 3Re3) 35 FOLLOWS: PLSCe Tfie TWO TUMBLING M3TS IN OPPOSITe CORNeRS OF Tfie GYM  1  a . P L a c e T f i e HOOPS iN T f i e R e M a i N i N S TWO coRNeRS WITH a B e a N Bas (RepReseNTiN© Tfie F L a G ) iNS«De e s c a Tfie TWO eNDs O F m e GYM SHOULD NOW MIRROR ONe s N o w e R . •  DIVIDe Tfie CL3SS INTO TWO eQU3L S Z e D TeSMS, 3ND 3SSIGN e a c n ONe OF Tfie TWO PINNY COLORS. eSCH PL3YeR TUCKS THeiR FiNNY INTO TfieiR W3i5TB3ND.  •  eaCH Te3M IS aSSiGNeD 3 N eND OF Tfie GYM TO DeFeND. Tfie OBJeCT O F THe G3Me.is TO  capruRe THe o m e R TeaMS FLSG (Be3N B3G) S N D RGTURN IT TO w e  HOOP iN THeiR OWN TeaM eND. RuLeS OF THe G3Me:. •  3 L L P3RT!C;P3NTS MUST JUMP WITH BOTH FeeT TOGeTHeR TO MOVe.3BOUT THe  PL3YING aRe3H THeY M3Y NOT MOVe 3 N Y OTHeR W3Y!  • THe M3TS RePReSeNT iM3GIN3RY PRISONS 3ND BeLONG TO THe Te3M DeFeNDiNG THe eND IN WHICH THeY 3Re LOC3TeD. •  PRiSONeRS M3Y B e TaKeN BY DISLODGING THe PiNNY TuCKeD INTO THeiR W3ISTB3ND. PRiSONeRS MUST THeN W3LK TO THeiR C3PTORS PRISON WHeRe THeY ST3Y UNTIL THeY 3Re TOUCHeD BY 3 MeMBeR O F THeiR OWN TeaM 3ND FReeD.  Teaching Tips: •  eNcouR3Ge eacH TeaM TO TaKe a Few MiNuTes BeFORe me G3Me TO DeveLOP 3. Te3M STR3TeGY  APPENDIX 4 Results for Participants  Healthy  Bone  Study1998Results  NICK's Skeleton June, "98  HICK's Skeleton October, '98 Table 1. NICK's Measurements October, *98  June,^ 147.5  142.5  Height (cm)  Just as different children grow at different rates, there i s a considerable difference i n the timing and rates of growth of the  different  body  systems.  For  example,  during  adolescence the lower extremities increase in length prior 68.2  63.8  Weight (kg)  to growth of the trunk.  In parental terms this suggests  that you will have to replace your child's jeans before  Calf Girth (cm)  35.7  37.2  Vertical Jump (cm)  17.0  17.1  your child's measurements over the school year to a large  1274.9  1362.4  population of healthy children of the same age and sex. The  replacing his or her s h i r t Table 2 compares the change in  T B B M C * (grams)  broken line shows the average range within which most  *TBBMC = total body bone mineral content  Table 2  children would be found. However, as there is a very wide  Average values for change over 8 months: LO  Height  HI  AVERAGE  variation i n the rate and timing of growth i n children at this age, a measurement falling outside of "average" is not unusual  For example, a child who begins to grow early  would fall at the HI end o f the range for height  ©  the LO end o f n o r m a l Weight  TBBMC  Both of these results represent a  healthy variation i n childhood growth.  Calf f i r t h  Vertical Jump  A child  who has not yet started their growth spurt, would fall at  If you have any questions about your child's results or the  ©  Bone Study, please call Motra at U B C ( 8 2 2 - 4 8 6 4 ) .  ©  Think You for participating m the Healthy Bone Study! We hope to seeyou again next year.  APPENDIX 5  Questionnaires  H E A L T H Y BONES STUDY: F O O D F R E Q U E N C Y Q U E S T I O N N A I R E  Child's Name:  :  .  Name of person completing the questionnaire:  Today's Date. '  Relationship to the child (i.e. mother, father, grandparent, guardian, etc.)  ;  We would like to know about some of the foods your child eats. For each food listed on the front and back of this sheet, please fill in how often your child usually eats a portion of the size stated. If s/he eats the food: -every day or more than once a day, fill in how many times they eat the food per day -less than once a day but more than once a week,fillin the times per week -less than once a week, but more than once a month,fillin the times per month -less often than once a month, or never eat it, put an "X" under "do not eat". Example: Janice has a glass of orange juice every morning, along with two slices of toast. She usually has two sandwiches at lunch, and eats french fries about three times a week. She almost never eats cauliflower. Orange juice, 1 cup French fries, regular serving Cauliflower, 1/2 cup (125 ml) Bread or toast, 1 slice  Per day 1  Per week  Per month  Don't eat  3 X 6  NUMBER OF TIMES M Y CHILD EATS T H E FOOD  Per day Bread or toast, 1 slice or 1 roll Muffin, 1 large Pizza, 1 medium slice Cheeseburger/Veggieburger with cheese. Cheese-1 slice processed OR 1 piece hard cheese (plain or in sandwich). Broccoli, 1/2 cup (125 ml) Gai-lan (Chinese broccoli), 1/2 cup.. Bok-choi (Chinese cabbage), 1/2 cup. Ice cream (large scoop)  Per week  Per month  Don't eat  NUMBER OF TIMES MY CHILD EATS THE FOOD Per day Per week Per month Don't eat Frozen yogurt (large scoop) Milkshake (including fast food shake). Yogurt, small (175 ml) carton or equivalent Canned salmon or sardines with bones 1/2 small can Soft drink 1 can or large glass Tofii, 2 oz (60 gm) Milk on cereal Orange juice, 1 cup Milk (any type including chocolate) 1 cup Macaroni & cheese, 1 cup (250 ml).. My Child usually drinks (choose one only): milk OR chocolate milk OR soy milk OR rice milk Is your child allergic to any foods? NO YES: (what foods?: Does your child use any vitamin and/or mineral supplements? Daily >3x/wk l-3x/wk Multivitamin • Multivitamin/mineral. Iron Vitamin C Calcium Other What is the brand/name of the supplement?  ) <l/wk  Name:  Healthy B o n e s Activity Questionnaire: Fall 1999 Aae:  Sex: M F Grade: We would like to know about the physical activity you have done in the last 7 days. This includes sports or dance that make you sweat or make your legs feel tired, or games that make you huff and puff, like tag, skipping, running, and climbing. Remember: A. There are norightor wrong answers - this is not a test. B. Please answer all questions as honestly and accurately as you can - this is very important. 1. PHYSICAL ACTIVITY IN YOUR SPARE TIME.  Have you done any of the following activities in the past 7 days? If yes, how many times and for how long? Tick only one circle per row* No 1-2 3-4 5-6 7 or more times time Skipping 0 0 0 0 0 Four Square  Creative Playground Tag Walking for exercise Bicycling Jogging or running  Aerobics  Swimming  Baseball, softball  Dance  Football Badminton Skateboarding  Soccer  Street Hockey Volleyball Floor Hockey Basketball Ice skating  Cross-country skiing loe hockey/ringette Other:  0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  200  0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  2. In the last 7 days, during you PHYSICAL EDUCATION (PE) CLASSES, how often were you very active (playing hard, running, jumping and throwing)? Check only one. 0 I don't do PE 0 Hardly ever 0 Sometimes  0 Quite often 0 Always  3. In the last 7 days, what did you do most of the time at RECESS? Check only one. 0 Sat down (talking, reading, doing school work) 0 Stood around or walked around. 0 Ran or played a little bit  0 Ran around and played quite a bit.  0 Ran and played hard most of the time. 4. In the last 7 days, what did you normally do AT LUNCH (besides eating lunch)? Check only one. 0 Sat down (talking, reading, doing school work) 0 Stood around or walked around. 0 Ran or played a little bit  0 Ran around and played quite a bit.  0 Ran and played hard most of the time. 5. In the last 7 days, on how many days RIGHT AFTER SCHOOL, did you do sports, dance, or play games in which you were very active? Check only one. 0 None. 0 1 time last week. 0 2 or 3 times. 0 4 times last week. 0 5 times last week. 6. In the last 7 days, on how many EVENINGS did you do sports, dance, or play games in which you were very active? Check only one. 0 None. 0 1 time last week. 0 2-3tJmes. 0 4 - 5 times last week. 0 6-7timeslastweek.  201  7. How many times did you do sports, dance, or play games in which you were very active LAST WEEKEND? Check only one. 0 None. 0 1 time. 0 2-3times. 0 4-5times. 0 6 or more times. /  8. Which ONE of the following five statements describes you best for the last 7 days? Read all 5 before deciding on the one answer that describes you. 0 All or most of my free time was spent doing things that involved little physical effort (e.g. watching TV, homework, playing computer games, Nintendo). 0 I sometimes (1-2 times last week) did physical things in my free time (e.g. played sports went running, swimming, bike riding, did aerobics). 0 I often (3-4 times last week) did physical things in my free time. 0 I quite often (5-6 times last week) did physical things in myfreetime. 0 I very often (7 or more times last week) did physical things in my free time. 9. How many hours per day did you watch television or play Nintendo last week? (each show is usually a half hour or 30 minutes). Check only one. 0 I watched less than 1 hour or have no TV. 0 I watched more than 1 hour but less than 2. 0 I watched more than 2 hours but less than 3. 0 I watched more than 3 hours but less than 4. 0 I watched more than 4 hours. 10. Were you sick last week, or did anything prevent you from doing your normal physical activities?  0 0  Yes  No  If yes, what prevented you?  10 1  11. Mark how often you did physical activity (like playing sports, games, doing dance or any other physical activity) for each day last week. None  Monday Tuesday Wednesday  Thursday Friday Saturday Sunday  0 0 0 0 0 0 0  Little Bit  0 0 0 0 0 0 0  Medium  0 0 0 0 0 0 0  12. Do you participate in organized sport or activity outside of school?  0 0  Yes No  If yes, what sport(s) or activities do you do?  How many days during the week do you do these activities?  01 02 03 04 05 06 07  i n  Often  0 0 0 0 0 0 0  Very Often  0 0 0 0 0 0 0  2.2 Is your child currently taking any medications? yes If Yes, What medication(s) is your child taking? __z  no  What are these medication(s) for? 3.0 Bone History  3.1 Has your child ever been hospitalized, confined to bed or had a limb immobilized (i.e. arm in cast)? yes no If Yes, list condition, approximate date and time involved (Example: wrist fracture summer, 1990 Reason  Date  lOwks) Time Involved  3.2 Is there a history of wrist, hip, or spinefracturesin your family? If Yes, indicate who was affected mother father maternal grandmother paternal grandmother maternal grandfather paternal grandfather 3.4. Is there a history of osteoporosis in your family? yes If Yes, indicate who was affected mother father maternal grandmother paternal grandmother maternal grandfather paternal grandfather  yes  no  no  3.5 Is there a history of any other bone disease in your family? yes no If Yes, please indicate the family member(s) affected 1.  2.  What is the name of the condition(s) affecting this family member? 1.  2.  PHYSICAL ACTIVITY  4. How would you rate the physical activity level of your child?, (physical activity is defined as vigorous activity that makes them sweat, huff & puff, or become out of breath). Inactive Sometimes Active Moderately Active Often Active Very Active THANK Y O U FOR Y O U R PARTICIPATION  Self-Assessment of Maturity Status  A s you keep growing over the next few years, you will see changes in your body. These changes happen at different ages for different children, and y o u may already be seeing some changes, others may have already gone through some changes. Sometimes it is important to know how a person is growing without having a doctor examine them. It can be hard for a person to describe themselves in words, so doctors have drawings of stages that all children go through. There are 5 drawings of pubic hair growth which are attached for you to look at.  W e want to know how well you can select your stage of growth from the attached set of drawings. A l l you need to do is pick the drawing that looks like you do now. Put a check mark above the drawing that is closest to your stage of development then put the sheet in the envelope and seal it so your answer w i l l be kept private.  10k  3  =  cr  .cc Sr „  3»  ro  £  a a  o'  -  ^ o tcou  3  o  CA  5  CJ  ro  Hi  c  3  CD  i C C CC 3 *  a  OJ  a CD  3 O Hi  OJ  a  OJ  «» Q)  (D  « 3  CD  ^  |  W  2 O  3  Q . CD  O .  -I  »  M = f  •  S  ID  OJ ~ t c n  I  "O  QJ  ro  co  -i  ST  =r w  =  3  -o  CD  £  g | g c c - ^ - _ tD « O 3  a a  3"  ^  —  MM  O H«  e -t  n.  N CD  CD  —• 3 * m OJ ~  OJ  CD  ST Hi CD  a.  -i OJ  CD  a  CD  a o  (A OJ  o  OJ  O  "I  3 Q,  C  CD OJ  CD  3*  3  ^ 3. i  OJ  CQ  — *  = ST Hi CO  CD  o  ex  CD 1  O  c OJ  c:  o  C  CD e g  C  3 CD OJ  •  CD  OJ OJ  Sj  OJ  cT  o  IQJ  H>  c  -T CD  TJ ro  CD  s/ce pj OJ  "0  O  CD  IQ3  c  a  HOi g £Q g - * i O J OJ 3  si  CD CA  OJCD CQ -I CD  ~* 5  CD •  OJ  3  a  —I  =T  -X g  a —.  H>  c  CD  Family Bone Study Questionnaire  The questions in this survey are directed towards those events in childhood, adolescence and your adult life which may have some influence on your bone mineral density. Read the questions carefully and mark the appropriate response with a check {/). Marie those questions which are not relevant or to which you are unable to respond, with N/A. All iruormation received remains strictly confidential. Thank you for participating in this part of the Family Bone Study.  ¥>2  1.1 Last Name  FirstName.  1.2 Addre ress 1.3 CitydrTown_  Po:  1.4 Telephone (Home) L.5 Date of Birtn: Day i.6 ! ! $  Month  Yet_  •  2.4 What language(s) do you speak at home? .5 How would you best describe your race or colour? (i.e. White, Chinese, East Indian, etc.) •8- ? 2.6 At what age did you finish schooling?  yearsfoY age  2.7 How many years of education did you complete?  years  zo°\  3.1 Have you ever smoked /  Yes'p U No % •  go to question 3.6 | | .£) ^ f j ^ t  3.2 Have you ever smokedror 6 months or more? Yes| • No-jj; • 3.3 Do you still smoke?  ^  go to question 3.6 % $jf- ^ #|  Yes, daily Yes, occasionally No, not at all  3.4  • •  3-6  ^> ^ , f^ f •  3.4 When you^ire/weresmoking, smoking,how howmany manycigarettes cigarettesdo/did do/didyoyou usually smoke per day? /hen you°are/were About KWteu  per day 9i r»l &  H - ^ ,  3.5 At about what age'did you start to smoke daily?  years"  ow often do you drink some kmdo^alco&olic kind or alcoholic beverage? t 3.6 How" yW'dnnk Daily or almost every day 3 or 4 times a week Once or twice a week Once or twice a month Less than once a month NEVER Don't know  n'#K%. 1 4fc&A_ • %_|ft 3 * • Q |_ |g I \ 2 -X • ^. | £ • jra_" i(24'^ CM£/s • 7^ y  3.7 How often did you drink (as a young adult) some kind of alcoholic beverage? Daily or almost every day 3 or 4timesa week Once or twice a week Once or twice a month Less than once a month NEVER Don't know  • *tf % % lfl 3 £ • f^fl • \\ \ % 2 ^ • *J'&J*\$K • ^'r • /r-£*il_ •  ft ffl^ft.fi 3.8 At about what age did you start to drink alcohol?  2J0  or Never •  3.9 H o ^ l U f s f ^ Childhood f^-lfi  Young t i | B Adulthood  Recent Past  Never Sometimes I to 2 cups per day j cups or more per day  3,0 H o C ^ p l M o ' ^ l / y l ^ & ^ t o ^ o a s Childhoodt^jtfl n  Young 4t|fl Adulthood' '  1  ^  Recent i ^ X ^ v ^ Past  NeveJ Sometimes 1 to 2 cups per day 3 cups or more per day  3.11 How'rntiy^ indicated?  flf-Jjft  Young Adulthood  Childhood  Recent Past  f%fiJb«X&  Never Somi  es  1 S2cans ler i i 3 cans or mc?e pef'day  3.12 Do you eat a special diet?  •  •  Yes/fl  —4m • —'?S§¥«r low cholesterol  ill  No nq .dairy (lactose (lacfcose mi intolerant) lo.uauy other: Please specr  •  YL  •  rt' „_  _ times/day  [QW many times a day do you take it? What is the name of the supplement?  X  How many milligrams of calcium does it contain?  3.14 Dqyou take a multivitamin supplement/ • Yes If Yes,  •  How many times a day do you take it?  . times/day  What is the name of the supplement? How many milhgrams of calcium does it contain?  3  '  fa • No • Yes ^ ? ° ****£ "^MS dafty basis? If Yes,' ho^i^ylu^s^aoay do you take it? _ tames/day"  l 5  y  U  o I a i d s o  r  o  na  3- 16JDp you take k bran orfibersupplement?  • Yes  i. day do you take it?  • No _times/day  % is mpnai^ofthe supplement? pro/serving  How rnSnygfams o ffiberSoes it contain?  . 1 Rate your overall level of physical activity as a child andyouth? youth?(circle' (circle one) 1 seldomfefc active #4f_  2 sometimesF^t active ?£<^j[  4 . _ moderatelyj^jl active  3 active  5  very^Sf active > t < ^ |  >w would you descnbe the games you played most often as a child/ (circle one) 4.2 How 1  2  3  games j £ | /fe games requiring mostly running, | such as board /j*)fc*-• fa(icjfo some running, ^ % £{fj ^ jumping, climbing, j^Jc. |^ ^ ^ games, drawing,)^, * ^ § ^ jumping, climbing Jfc | j , t h r o w i n g games. ^ Jfo puzzles, etc. ^ Iffofc^ throwing, etc. }fp. ^ * ' (  4.3 During^which years we^you physically active? (cirHe^l'?rii^ap^)ry^ ^ ^" ^ 5 2 3 4 30-40 10-15 15-20 20-30 age? 5-10 1  OST pnysically active? (circle one) mg which years were; you the MOS 4.4 During s you 3 4 2 15-20 20-30 age^ 5-10 10-15  11^  5 30-40  4.5 Did you participate in organized sport as a child or youth (to 18 years)U If Yes, list the sports you participated in and the approximate years of your participation: 7t ^rtrf^ if- $ 3 Example:^) jfc:  soccer-^M^ gymnastics ^ ACTrVTrY  5 years £ 4  ^  ^ ^ t ^ .  1  y  e  a  r  Nunffief of Years  ince the age or 18 years, did you regularly participate in sport (eg. hockey or tennis)? • Yesftn No#^ If Yes, list the sports you participated in and the approximate years of your participation: Example:Ufa:  hockey ft aerobic dance^  _  1 year 3 years  ACTTX/rTY  4  Numoef^f^ears  4.7 How would you rate your current level of physical activity (relative to others your age)? • very 414ft • lowfft • averagej£}ti • high fa • very ^ low high 4.8 How would you rate your current level of physicalfitness(relative to others your age)? °very^f<£ • low^ • average^ • high % Clvery^* low ^ high T  4.9 Approximately how many hours of television do you watch each day? hours on weekdays EJ ^ '1- ^r B  hours on Sat. and Sun.  111  f  K  Q^  4.  W  4-o  ^ . ± . * l f l l K j . 4 | | ^  4.10 During the last two weeks how many times did you do any of the activities listed below:  About how muchtimedid you spend on each occasion:  X  TWO WEEKS AGO LAST WEEK # of 14; -minutes #of ffiijr -minutes times #1 each timeife times ^ each timej&3 P lylL walking Home exercise &fli!^|6 Aerobics Weight training Bicycling  ——  Jogging or runnin^^ Bowling >liSti^ Social dancing^t^^-tjfe  Jazz, ballet or  §±% £%fy t  modem dancing^ Vf$fa %  Racquet sports^p Golf  t £  Swinuriing Gardening/ Yard work |J JL% House work^^  OR • I did nothing like this in the last two weeks. 4.11 Other comments concerning lifestyle/physical activity?  ^^^^^^^^-^^^^  5.0 ReproductiveHistory<femdes)j>j$4vr  77ze.ye questions will help us understand how women's hormones relate to bone structure.  ver gone J monuos or more WILUOUL a u i c u i u u a i • Yes^ • No '!-* go to 5.2  (iwiuu  7  What was the longest single period oftimewithout menstrual flow? months If you count all the penods you have missed throughouryour menstruating years, how many months would that be? (cumulative) 5.2 Have your menstrual penods stopped for more than one year? • YCSTL  ONo^  At whailifie? :'age? ive you 5.3 Have youhad your uterus removed (hysterectomy)? • Yesfc. DNo^C At what age? 5.4 Have you ever had •Yesrbh^ icvett • Yesfboth \ • Yes^o ™McIOW now many • No 5.5 Do you or did you ever take estrogen for menopause or for any other reason • Yes, currently & g&g Jftjfl • No ^ go to 5.6 &7fl |{j • Yes, but not now| ^ What type(s)?"?f-|f /?f *$#  N  frl^*  Piling Number of Name^ days/month  Age started  ftflflft *&3JHyg Total number of Age months taken stopped  Patch^ Number of Name days/month  Age started  Age stopped  ; %3 •rn Injection A  Ifl^itt*  How many times/year? How many years?  Total number of months taken  How frequently?  a/5  a months  5.6 Doyl/<3u or did you ever take Provera for menopause or for any other reason?  ft.I R f c f l  • Yes, currently|,g • Yes, but not now#  • No ^ go"o 5.7 * ^  ^  Whattype(s)?^F-H  Pillar Number of Name days/month  Age started  Age stopped  Total number of months taken  • Injection  5.7 Have you ever used birtfi control puis or oral contraceptives / $ • Yes^j D N c f ^ go to 5.8 4 ^ ^ - ^ ?(t^)At what age did you start (approximately)?  fl  5  ;  5  &  years^'  ^&fi$$$sLk%> For approximately how long did you use birth control pills?  years^  Are you stall using birth control pills? • No -> At what • Yes fa hat age did you stop?  4.  5.8 If you  removed:  months^  . years ^  i  ^ £10^  , >>3-  frA^llf  ft  ?.  o  Can you tell by the way you feelthat your period is coming? • Yes, every month ^ *>,}$_ • Yes, most monthsfc.£. S & #*»& • Yes, less than half the time # »S • Yes, once or twice a year r^fe'iL • Never tft£ 'I  If Yes to any of the above, /g ; What signs or symptoms indicate i • menstrual cramps or aching back or l e g s i ^ X % ^ ) ^ T % J f 9 ^ • bloating,fluidretention y%flf,4$ ^_ • increased appetite (in general orfor sweet, sagty orspicyfoods)^^^^ (^~fe^$,^*t • moodiness (frustration, irritability, sadness}^ ^ • breast tenderness in thefrontor the nipple j% fa  • • • •  - % A?» ^ l ^ i ) ' Jfjfj %  %\, £ |  breast tenderness up under the arm or on the outer sides of the I b ra ea se w ing orRftension) cA tfJL (fa £f he dsatch sel(migraine JL \ acne/pimples/blemishes 0^ jjg. ^ fyj yg,, 7 other (please specify): y  '  gft'T H Q%  5.9 How many times nave you been pregnant? 5.10 Hov/many ofthpSepregnancies resulted mat * leasf one live Birth (count twins and triplets as 1)1 . -> age at 1st birth?  '  5.11 Didyo  :9  for how many months total months (adding up the months with each child)(j\%v§ %fyjfc$-ti$ 5.12 How old were you when you had yourfirstmenstrual period? once they began?  • Yes^ go to 5.3  b>  .years^j •  No*"  . * t e * M * *  • No"^ go to 5.14;{j$% did they become regular?  M^illtSjtM^?  c) Have"yo'ur periods been made regular by medication?  • Yesj|_ years Atwhlafe? • Yesj|__  • N6% go to  Atwl  years  inyour20'sand30's? >. 14 On average, how often did you have menstrual penods when you were i • • • • • •  4.  20 days or less 2ok&yl~\-> 21-25 days 41—A££ 26-30 days 46 — £ o £ 31-36 days 31 — 3 6 ^ 37 days or more -j ^ jg_ wL do not know pr^ ^ 3  4  These questions will help us understand how men's hormones relate to bone structure. We as all men these questions. A^f^' \%M ^ % 41 # #fe£g * ft ##-W*ig)lfc, 6.1 Howiw^nlny^nt many < olfen^ave you fathered?;  1tea J ft  l  _ fertihty problem? 6.2 Have you ever oeefa diagnosed with a fertility prob • Don't Know • Yes 7^ rJNotf 6.3 Which of the following is your usual experience regarding spontaneous erections not related to sex? • one or more times a day (for example,firstthing when I wake upjjpx^rf^^j^ftfc)**  • most days tl^Zpfe S $r • some days 7jE| Q ^, • occasionally ^jft^ • rarely 4ft. + • never  11^  :f >• WJ  \  ^.0 developmental i r i s t o r y 7.1  '  ~ "  —  >  i > S ^ ^ 1 ^ ^ ^ M ^ ^ QtMtv *  **L  oihers your age in thetune intervals listed below somewhat^ £ average %fy somewhat very thin -£r overweight $Jf^ overweightj^ft  g Q t r e i a t I v e to  very 4^ &  n  6-10 yrs &*ojfc • 11-15  yrs I M S & D  16-20  yrs t i - Z o ^ O  *  21-25 yrs u*$ftp  • • • •  • • • •  7.2 In the last six months have you gained or lo^vreight  IffI HiLtiiU,  8.1 In the last 6 months have you  1  • •  • •  • •  • Yes ! •Noi&'fc 7  gat& _ M lbs -or-  kgs /  Io  kgs ^]  seeiVa dbctor?  It Yes, what was the reason for your visit?  If Yes, please specify  • •  lbs -or-  • Yes$- • No#^  \  41  T ^ I M ^ f t ^ - ^ 4 0 ^ ?  8.3 Have you ever been treated for any of the following conditions? fc tt— food allergies^'^ D^es • n$ ft iio' • yi asthma vS§ •• no k i d n e y d i s e a s e Qkfft • y e s • no other aUergies • yes • no back pain  Dyes • no  chrome liver problems  U yes  • no  scohosis^'^ll'jrilj • yes • no  gastrointestinal disease U yes • no  epilepsyy|| j$  muscular dystrophy  • yes • no  osteoporosis t j | ^ &kf$\ O yes • no  osteoarthritis  • yes • no  rheumatoid arthritis .  anemia j |  • yes • no  Dyes • no • yes • no  breast cancer %\jffa • uterine cancer^. V'yj;  v e s  •  prostate cancer j^&jijt^n yes • no  n o  hypertension (fjjaif|>  • yes • no  • yes • no  malabsorption^HlH-^L Dyes • no Insulin dependent  excess blood calcium  • yes • no  :xc5ssunnar\ excess unnary calcium • yes • no  c • yes • no No:nlnsulin dependent  ,4*  eaJIdife)^ • yes • no  *hyper = excess AijEJJ %,) **hypo = deficiency ;$j)(k_^  8.4  u  .  a  Dyes • no  **hypoparamyroidism • yes • no  ..^^erp^a^n^^d^^^ yes • no •hypo^yroidism  • yes • no  **hypomyrcudism^' Otherr conditions (please list)  e you you had any problems with your bones such as fracture? ive h ho'w ft^y^^tu^lh^av IT ive you had? Pleale^ist^e^ljypefl^n^^ Type of fracture ? p | | «f  • yes • no  tihe  date ofoccurrence  Date $  21?  • Yes^ • NOKS. ft  [i.e.:&. wrist fracturef Reason #g  flfcff^  summer, 1980 DateQ^ft  'e you 8.6 Have youhad any surgery in the past 2 years?  , l?*o  10 wks] "t Time Involved ^ / f l BtJ^ f  • Yes'fl • No^'jJ  ure and the approximate date [i.e. : gall bladder removed 3^#? f|. R e a s o n D a t  summer, 1990] j £ ' 1 ^ e  8.7 Is there a history of wrist, hip or spinefracturesm your family? • • • • •  • • • • •  • Yes/£j U Noj&'fy  maternal grandmother^ [>. t^maternal grandfather i j-^-jl^C. father paternal grandmother ^ J f L t t r paternal grandfather  8.8 Is mere^riistor^of osteoporosis m.your family? IfYes, inmcate wno* was"*affedfa^ "  0  • Yes'ft • Nora'p  • mother $ maternal grandmother %%J§~ maternal grandfather 5 L $ father / „ paternal grandmother yj~ paternal grandfather jr^  8.9 Is^here a history of any other bone disease in your family? • Yes']} • No$/(y %P IfVes,*  '  Please^dilatJthe farn^^einb%(fj affected 1.  What is me name of the condition(s) affecting this family member?  110  9.1 Are you currently taking any mearcations? • Yesft• .No'*f} /  ir Yes, what medicahon(s) are you taking?  What are these medications for?  9.2 Have you ever taken any of the following medications? Please specify at what age you began to use them and for how long you used them, fa [iS*" insuhn^ Medicafkm  • _  ft  Xes,^  s  15 yrs. old * ^fy  20 yrs.  ,'Ag&af s t a r t ] ,  Duration ofuse  calcium^preparations antacids*"^ . anabolic steroids fluoride^ vitamin D compounds calcitonin diuretics heparin '  ">  cortisone (oral) _ a.  • ,  coracosteroids (other) ann-inflarrrrnatones thyroid preparations other (please speafy):  0.2.1  m  ^ ,  ^  \  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0077364/manifest

Comment

Related Items