SEARCH

SEARCH BY CITATION

Keywords:

  • bone;
  • hip structural analysis;
  • physical activity;
  • calcium intake;
  • aging

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

A population-based study on 1008 postmenopausal women identified that the 24% of women achieving high levels of PA and CI had 3.4–4.4% higher femoral bone strength in axial compression and 1.7–5.2% in bending than those achieving low levels, indicating that lifestyle factors influence bone strength in the proximal femur.

Introduction: Extensive research has shown that increased physical activity (PA) and calcium intake (CI) decrease the rate of bone loss; however, there is little research on how these lifestyle variables affect bone geometry. This study was designed to investigate the effects of modifiable lifestyle variables, habitual PA and dietary CI, on femoral geometry in older women.

Materials and Methods: Femoral geometry, habitual PA, and dietary CI were measured in a population-based sample of 1008 women (median age ± interquartile range, 75 ± 4years) enrolled in a randomized controlled trial (RCT) of calcium supplementation. Baseline PA and CI were assessed by validated questionnaires, and 1-year DXA scans (Hologic 4500A) were analyzed using the hip structural analysis technique. Section modulus (Z), an index of bending strength, cross-sectional area (CSA), an index of axial compression strength, subperiosteal width (SPW), and centroid position, the position of the center of mass, were measured at the femoral neck (NN), intertrochanter (IT), and femoral shaft (FS) sites. These data were divided into tertiles of PA and CI, and the results were compared using analysis of covariance (ANCOVA), with corrections for age, height, weight, and treatment (calcium/placebo).

Results and Conclusions: PA showed a significant dose–response effect on CSA all hip sites (p < 0.03) and Z at the narrow neck and intertrochanter sites (p < 0.02). For CI, there was a dose–response effect for centroid position at the intertrochanter (p = 0.03). These effects were additive, such that the women (n = 240) with PA in excess of 65.5 kcal/day and CI in excess of 1039 mg/day had significantly greater CSA (NN, 4.4%; IT, 4.3%; FS, 3.4%) and Z (NN, 3.9%; IT, 5.2%). These data show a favorable association between PA and aspects of bone structural geometry consistent with better bone strength. Association between CI and bone structure was only evident in 1 of 15 variables tested. However, there was evidence that there may be additive effects, whereby women with high levels of PA and CI in excess of 1039 mg/day had significantly greater CSA (NN, 0.4%; FS, 2.1%) and Z (IT, 3.0%) than women with high PA but low CI. These data show that current public health guidelines for PA and dietary CI are not inappropriate where bone structure is the health component of interest.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Hip fracture arising from bone fragility contributes to both morbidity and mortality.(1,2) BMD assessment has long been the standard measure for identifying those at risk for osteoporotic fracture; however, this measurement provides, at best, an indirect indication of the strength of the bone.(3) Bone strength declines in old age, especially in women.(3) The aging effects on bone are characterized by decreasing BMD, mean cortical thickness, and section modulus, with accompanying increases in subperiosteal and endocortical widths.(4,5) Statistically, BMD is a good predictor of fracture; however, different geometrical configurations of bone may have the same BMD but not necessarily the same strength.(6,7) Studies are now identifying an association between hip geometry and fracture.(8,9) Therefore, to prevent the development of bone fragility, an understanding of the variables affecting all aspects of bone strength is important.

There is extensive research on the relationship between BMD and lifestyle variables. Studies have reported positive associations between BMD and the lifestyle variables, physical activity (PA) and dietary calcium intake (CI).(6,10–18) An interaction between PA and dietary calcium that affects BMD has also been reported,(12,15,18) and randomized controlled trials have shown that both of these lifestyle variables have a positive affect on BMD.(19–22)

Less is known about the relationship between lifestyle variables and the underlying geometry, an important element in the determination of bone strength and subsequent fracture risk.(23) Available data suggest that CI and PA may also have a favorable effect on the geometry of the femur.(3,6,17,24) Aspects of hip geometry associated with bone strength of the proximal femur can be estimated from data derived from DXA scans.(1,25–31) Section modulus, an index of bending resistance, has been shown to be more strongly associated with PA level than BMD.(6) Moreover, postmenopausal women with a daily CI in excess of 1200 mg have been shown to have greater femoral shaft section modulus and shaft subperiosteal width (SPW) compared with those with a daily intake of <800 mg.(17)

The purpose of this study was to determine the association between the modifiable lifestyle variables, habitual PA and dietary CI, and geometric indices of bone strength in the proximal femur in postmenopausal women.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

This study was a cross-sectional analysis conducted within the Calcium Intake Fracture Outcome Study (CAIFOS), a 5-year prospective, randomized controlled trial of oral calcium supplementation in the prevention of osteoporotic fractures. The 1497 women <70 years of age included in the CAIFOS study were recruited using a population-based approach in which a random subset of women on the electoral role were sent a letter inviting them to participate in the study. More than 70% of women in this age group are on the electoral role, and 18% of the women approached responded.(18,26) Women were excluded if they had a significant current illness or if they were receiving any bone-active agent, including calcium supplementation. As previously reported, participants were similar in terms of disease burden and pharmaceutical consumption to whole populations of this age, but they were more likely to be from higher socioeconomic groups.(18,26) An earlier analysis of the data from this cohort has shown significant associations between BMD and the lifestyle variables, PA and CI.(18) The cross-sectional design used in this study enables examination of the relationships between bone structural variables and the self-selected lifestyle (PA and CI) of women >70 years. Randomized controlled studies of significant duration would be required to examine whether there is any potential for supplementation of PA and/or CI to favorably affect bone structural variables.

Reported PA

PA levels were ascertained by questionnaire at baseline.(18) Women were asked whether they participated in any sports, recreation, or regular PA. Those who answered “no” to this question scored zero, and those who answered “yes” were asked to list up to four sports, recreations, or forms of regular PA, including walking, that were undertaken in the last 3 months. Energy expenditure (kcal/day) for these activities was calculated using published energy costs.(27,28)

Dietary CI

The Anti-Cancer Council of Victoria Food Frequency Questionnaire (ACCVFFQ), a self-completed semi-quantitative questionnaire, was used to determine the average daily dietary calcium intake of the participants at baseline. The ACCVFFQ uses the NUTTAB95 database (Australian Government Publishing Service, Canberra, Australia) and has been validated against weighed diet records.(29) Comparison of these two methods revealed that mean intakes of calcium per day differed by only 30 mg (11% of the SD).(29)

DXA and hip structural analysis

Femoral geometric indices were determined at three narrow “cross-sectional” regions traversing the proximal femur, using the Hip Structural Analysis (HSA) Program Version 3 and incorporating correction for the array mode of the QDR 4500.(25) The regions corresponded to the narrow neck (NN), which traverses the femoral neck at its narrowest point, the intertrochanteric region (IT), which is positioned along the bisector of the angle between the neck and shaft axes, and the shaft region (FS), positioned at a distance of 1.5 times minimum neck width, distal to the axes intersection. The methods of computation and the derived variables have been described elsewhere.(1,6,30,31)

All DXA scans were acquired using the same Hologic QDR 4500A scanner (Bedford MA, USA). Because the primary outcome of the RCT was fracture, to reduce responder burden, BMD data were not collected until the 12-month visit. At this time, 1390 volunteers remained in the study, and these women were encouraged but not required to attend a second visit within 1 or 2 weeks of the primary follow-up visit for DXA scanning. A total of 1100 women had hip scans. Data from 43 of these women could not be used for HSA analysis. Scans were excluded when the program failed to complete a full analysis at all three regions (1.64%) or failed to correctly identify regions of interest (1.27%). Scans were also excluded where insufficient abduction during DXA scan acquisition resulted in the ischium being located too close to the femoral neck, so that overlap of the bones caused erroneous measurements (1.0%).

The variables included were BMD measured at each HSA site, cross-sectional area (CSA), section modulus (Z), SPW, and the centroid position. CSA is an index of the resistance of the bone to compression. It is equal to the amount of bone surface area in the cross-section after excluding all soft tissue space and is therefore proportional in principle to BMC.(31) Higher section modulus (Z) indicates greater resistance of the bone to mechanical failure under bending.(25,31) It is derived from cross-sectional moment of inertia (CSMI), which is an important index of structural rigidity. Because of the limitations of 2D DXA images, the CSMI and Z are only relevant to bending in the plane of the DXA image (the coronal plane) and cannot assess bending in any other direction.(1,6,25) The SPW is the outer diameter of the bone that is computed as the blur-corrected width of the mass profile.(25,31) The centroid position is the measured position of the center of mass of the cross-section with respect to the medial bone margin and is normalized by dividing this position by the SPW.(25) A lower centroid value in a bone subject to bending implies that there is an imbalance between bone mass in the superolateral zone compared with the inferomedial zone, in favor of the latter. This could occur, for example, as a result of thinning of superior cortex with aging, as has been observed in the upper femoral neck. This imbalance may reduce the resistance of the bone to local compressive or buckling forces and enhance fracture risk. The HSA program also generates estimates of mean cortical thickness and cortical stability (buckling ratio). However, because these require assumptions concerning cross-sectional shape and the ratio of trabecular to cortical bone mass at each measured site, they have not been included.(25,31,32)

All scans were analyzed by a single operator after extensive training. Before the start of this analysis, 30 randomly selected scans were analyzed in random order on two separate occasions 1 day apart to determine intraoperator reliability. Intraclass correlation coefficients (ICCs) varied between regions but were >0.96. No systematic difference occurred between the two analyses for any variable. This subset of scans were reanalyzed on four occasions during the course of the analysis to ensure that operator technique remained consistent. At each reanalysis, repeated-measures ANOVA models were constructed to test for systematic error. No significant difference was present between the serial measurements.

The short-term precisional errors of HSA variables, expressed in CV%, have been determined for the Hologic QDR4500 in a subject sample similar to this study, from replicate hip DXA scans.(31) For CSA, it ranges from 2.3% to 3%; for CSMI, it ranges from 3.2% to 4.6%; for Z, it ranges from 2.5% to 4.0%; for SPW, it ranges from 0.9% to 2.1%; and for centroid position, the range is 1.1–2.1%.

Data management and statistical analysis

Of the 1057 subjects with valid HSA data, 49 did not have complete data for PA and CI. Consequently, there were data from 1008 of the CAIFOS volunteers for this analysis. The subset of women with complete data for this study were slightly younger, lighter, taller, and participated in more activity but consumed less calcium than the women excluded from the analysis (Table 1). The statistical analysis was performed using SPSS for Windows Version 11.5 (SPSS, Chicago, IL, USA). Assessments of habitual PA and dietary CI were divided into tertiles, and two one-way analysis of covariance (ANCOVA) models were constructed to examine the independent dose responses of PA and CI. Age, height, weight, and treatment were added as covariates. Because one half of the women in the study had been receiving calcium supplementation for a relatively short time (12 months) before DXA measurements, there was a risk that this may confound the effects of their more prolonged, self-selected lifestyle. For this reason, treatment was included as a covariate in the models. The lowest PA tertile included those with <65.5 kcal/day expenditure, and the upper tertile included those with >176.5 kcal/day expenditure. The lowest tertile for dietary CI included those who consumed <780 mg of calcium/day, and the upper tertile included those who consumed >1039 mg of calcium/day.

Table Table 1.. Differences Between Included and Excluded Groups
Thumbnail image of

To examine whether their effects were additive and whether there were significant interactions between PA and CI, these variables were dichotomized with cut-points based on the tertile analyses. Daily PA was divided at 65.5 kcal and daily CI was divided at 1039 mg. Two-way ANCOVA models were constructed, with correction for the same potential covariates. Levine's test of homogeneity of variance was reviewed for all models. Statistical significance was inferred where p < 0.05. This was not adjusted for multiple comparisons because all hypotheses were defined a priori, and this would increase the chance of making type II errors, which may result in important differences being overlooked as insignificant.(33)

Ethics

This study was approved by the Human Resources Ethic Committees of the University of Western Australia and Curtin University of Technology.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Table 2 outlines the demographic characteristics of the sample.

Table Table 2.. Demographic Data of 1008 Subjects Included
Thumbnail image of

Habitual PA

The median habitual PA level in the 1008 postmenopausal women participating in this study was 119 kcal/day. However, 21.7% (219) of these women reported not participating in any regular PA. There was a dose–response effect of PA on femoral geometry and HSA-derived BMD, where higher levels of PA were associated with favorable changes in CSA at all three regions and Z and HSA-BMD at the NN and IT regions (Table 3). Of the covariates, treatment was not associated with CSA or Z at any of the sites measured (p > 0.19); height and weight were associated with CSA and Z at all of the sites (p < 0.001); and age was associated with Z at the IT and FS (p < 0.02) and was not associated with CSA at any of the sites (p > 0.34). The middle and upper tertiles of PA tended to have significantly greater values compared with the lower PA tertile (PA ≤ 65.5 kcal/day), for most of the variables.

Table Table 3.. HSA Variables in Subjects Undertaking Different Levels of PA
Thumbnail image of

Dietary CI

The median CI in the 1008 postmenopausal women in this study was 903 mg/day. There was little evidence of a dose–response effect of CI on HSA BMD or geometry variables. The only variable showing a significant difference was the centroid at the IT region, where the highest tertile was significantly greater than the lowest (Table 4). In this model, treatment was not a significant covariate (p = 0.63), whereas age, height, and weight were significant (p < 0.03).

Table Table 4.. HSA Variables in Subjects Consuming Different Amounts of Calcium
Thumbnail image of

PA and CI

Additive effects were identified with PA and CI for HSA-BMD and CSA at all three regions (NN, IT, FS) and for Z at the NN and IT regions (Table 5). In all these regions, women in the high CI/high PA group had a more favorable BMD and geometry than the women with low CI and PA. Women with combinations of high and low variables were intermediate (Figs. 1 and 2). There were no significant interactions between PA and CI for any of the variables (Table 5).

Table Table 5.. Results for Groups Dichotomized According to CI (≤1039 or >1039 mg/day) and PA (≤65.5 or >65.5 kcal/day)
Thumbnail image of
thumbnail image

Figure Figure 1. The effect of physical activity (A) and calcium intake (CI) on cross-sectional area in the dichotomised groups which include lower and higher CI (≤1039 or >1039 mg/day) and lower and higher PA (≤66 or >66 kcal/day). Data are estimated marginal means (95% CI). Bars with different letters are significantly different

Download figure to PowerPoint

thumbnail image

Figure Figure 2. Effect of physical activity (PA) and calcium intake (CI) on section modulus in the dichotomised groups which include lower and higher CI (≤1039 or >1039 mg/day) and lower and higher PA (≤65.5 or >65.5 kcal/day). Data are estimated marginal means (95% CI). Bars with different letters are significantly different.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

This study identified a positive association between femoral geometry and the modifiable lifestyle variables, PA and CI, in healthy postmenopausal women. The relationship between femoral geometry and modifiable lifestyle factors has not previously been studied for all three regions of the proximal femur: the NN, IT, and FS. This study also examined the additive effects and possible interactions between habitual PA and dietary CI.

Habitual PA

Animal studies suggest that mechanical loading generates an osteogenic response that causes adaptive changes in bone geometry.(24,34,35) This is consistent with studies of postmenopausal women reporting positive associations between femoral geometry and PA for CSA and section modulus at the NN, IT, and FS regions and SPW at the IT and FS regions.(6,17) The results of this study are consistent with previous studies, in that significantly greater levels of CSA and section modulus were found to be associated with increased levels of daily PA.

Additional energy expenditure in kilocalories per day is a common measure used to represent the magnitude of PA undertaken by an individual, although energy expenditure for the same activity varies between individuals in accordance with their body size. Moreover, bone geometry is associated with body size.(6,30) To account for this, height and weight were included as covariates in all ANOVA models used in this study. Therefore, the associations between PA and HSA variables, detected in this study, should be independent of any association between either of these variables and body size.

The significant differences in femoral geometry associated with greater levels of PA result in increased femoral strength with respect to axial and bending loads. Significantly greater femoral CSA with increased PA results in an increase in the axial strength of the bone.(30) PA levels in excess of 65.5 kcal/day were associated with CSAs that were 4.0%, 2.3%, and 2.4% greater at the NN, IT, and FS regions, respectively. A greater section modulus, as measured by HSA, represents an increase in resistance to femoral bending in the coronal plane.(25,30) Differences in bending resistance in other planes may also occur but cannot be evaluated using a single projection DXA scan. Section modulus of the group with PA in excess of 65.5 kcal/day was 3.8% and 3.5% greater at the NN and IT regions. These results support recommended daily PA of 150 minutes/week of moderate exercise (i.e., 116 kcal/day), which is suggested to achieve a range of health benefits.(36)

Associations between PA and section modulus have been previously explained by significantly greater SPWs in the population participating in greater levels of PA.(6) In this dataset, a positive association between PA and SPW was not apparent at any region. In this study, there were significantly greater values for section modulus and CSA, without accompanying increases in SPW at the NN and IT. This suggests that subjects who participate in greater levels of PA have a greater bone mass within cross-sections of similar outer diameters. This suggestion is supported by the increase in HSA-BMD with increasing PA levels at those sites with increased Z (Table 3).

Dietary CI

CI was associated with fewer HSA variables than PA. The effect of calcium on femoral geometry has only been studied previously at the FS region, where associations were reported between CI and section modulus and SPW.(17) Despite a considerable sample size and the wide range of CIs in this study of postmenopausal women, the only significant association identified was between CI and centroid position at the IT.

The observation that the centroid of the IT cross-section was more centrally located in those with higher CIs is worthy of comment. The position of the centroid is an indicator of the symmetry of the cross-section with respect to the image plane (a value of 0.5 would be centrally located, smaller values are shifted medially). The femoral neck biopsy studies of Bell et al.(37) showed that femoral neck fracture cases show preferential cortical thinning on the more lateral surfaces, which should shift the centroid medially. A medial shift in the femoral neck centroid with age was observed in U.S. whites from the NHANES III survey(38) and in the cadaver study of Mayhew et al.,(39) with a greater age shift in women in both studies.

Habitual PA and dietary CI

Both PA and CI have been identified as risk factors for osteoporosis.(2) The results of this study indicate that habitual PA and CI work independently; however, their effects are additive. Almost one quarter of the women (n = 240) in this study achieved levels of daily PA in excess of 65.5 kcal and daily CI in excess of 1039 mg. The femoral CSA of these women was 4.4%, 4.3%, and 3.4% greater at the NN, IT, and FS regions, respectively, compared with women with low levels (PA < 65.5 kcal/day and CI < 1039 mg/day) of both these lifestyle variables. The section modulus of these women was 3.9%, 5.2%, and 1.7% greater at the NN, IT, and FS regions, respectively, compared with women with low levels of these lifestyle variables. Based on the section modulus odds ratios reported for osteoporotic fractures in a population of postmenopausal white women,(9) the 24% of women in this study who have low levels of both PA and CI are 1.11–1.26 times more likely to suffer from a hip fracture compared with women achieving high levels of both PA and CI. Whereas both PA and calcium are beneficial to bone strength, PA seems to be a more important lifestyle variable with regard to optimizing femoral geometry.

Comparison of conventional BMD and HSA BMD

The regions of interest used for conventional BMD differ from those used in HSA.(31) Conventional BMD and HSA-BMD have approximately similarly located neck regions, although the latter is only one third as wide.(31) Because the IT HSA site traverses the conventional intertrochanteric and trochanteric BMD sites, it is not directly comparable.(31) No conventional BMD equivalent exists for the FS region. In addition, HSA regions are much smaller than conventional BMD regions so they are more prone to sampling error.(31)

A study on a similar subset of the CAIFOS cohort has already revealed associations between conventional BMD and modifiable lifestyle variables: habitual PA and dietary CI.(18) This allows a direct comparison between conventional BMD and HSA-BMD. The general response of conventional BMD and HSA-BMD is similar between the two studies for both PA and CI (Tables 2 and 3).(18) In addition, the additive effects of habitual PA and dietary CI, noted by Devine et al.,(18) for conventional BMD were also present for HSA-BMD at all comparable regions. An interaction was detected between habitual PA and dietary calcium for the femoral neck in conventional BMD, whereas no interaction was detected for any HSA-BMD site.(18) The general consistency between the results from conventional BMD and HSA-BMD suggests that these measures may be interpreted in a similar fashion. It should be noted, however, that, overall, there were fewer significant associations between the lifestyle variables and HSA variables than with the conventional BMD variables. These may reflect the lower precision of the HSA analysis compared with the conventional BMD analysis.(31)

Limitations

This is a cross-sectional study, and therefore, there is a need for prospective studies on the effects of modifiable lifestyle variables, PA and CI, on structural geometric indices to determine if these associations are causal. A 7-month exercise intervention has been shown to result in increased BMD, CSA, section modulus, and cortical thickness in girls 9–12 years of age.(40) Analogous prospective studies need to be conducted in an adult population at risk for osteoporotic fracture.

The data analyzed for this study were collected within a 5-year longitudinal randomized controlled trial so the women recruited were community dwelling and tended to be relatively healthy. In addition, the type of PA undertaken has not been specifically considered in this study. Further studies investigating mode of PA are required to determine what type of PA has the greatest effect on bone geometry. Some participants had also been receiving calcium supplementation for 1 year. HSA is dependent on the positioning of the subject's femur during acquisition of the DXA scan. Variation in this positioning alters the projected distribution of the mineral and increases the possibility of reproducibility error.(9,31) The fundamental 2D nature of a DXA image also limits HSA in that it can only estimate bending strength in the plane of projection of the DXA image.(31)

Conclusions

This study has identified positive associations between aspects of femoral geometry related to bone strength and the modifiable lifestyle variables, habitual PA and dietary CI. These results broadly support current public health guidelines for PA and dietary CI.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We thank Ritu Gupta for statistical advice and Ian Dick for involvement in the recruitment and acquisition of the data from the CAIFOS subjects. This study was supported by the Healthway Health Promotion Foundation of Western Australia and the Australasian Menopause Society.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  • 1
    Beck TJ, Ruff CB, Warden KE, Scott WW, Rao GU 1990 Predicting femoral neck strength from bone mineral data: A structural approach. Invest Radiol 25: 618.
  • 2
    Nguyen TV, Center JR, Eisman JA 2004 Osteoporosis: Underrated, underdiagnosed and undertreated. Med J Aust 180 (5 Suppl): S18S22.
  • 3
    Hayes WC, Bouxsein ML 1997 Biomechanics of cortical and trabecular bone: Implications for assessment of fracture risk. Mow VC, Hayes WC (eds.) Basic Orthopaedic Biomechanics, 2nd ed. Lippincott-Raven Publishers, Philadelphia, PA, USA, 69111.
  • 4
    Beck TJ, Looker AC, Ruff CB, Sievanen H, Wahner HW 2000 Structural trends in the aging femoral neck and proximal shaft: Analysis of the Third National Health and Nutrition Examination Survey Dual-Energy X-ray Absorptiometry Data. J Bone Miner Res 15: 22972304.
  • 5
    Duan Y, Beck TJ, Wang X, Seeman E 2003 Structural and biomechanical basis of sexual dimorphism in femoral neck fragility has its origins in growth and aging. J Bone Miner Res 18: 17661774.
  • 6
    Kaptoge S, Dalzell N, Jakes RW, Wareham N, Day NE, Khaw KT, Beck TJ, Loveridge N, Reeve J 2003 Hip section modulus, a measure of bending resistance, is more strongly related to reported physical activity than BMD. Osteoporos Int 14: 941949.
  • 7
    Miller PD, Bonnick SL 1999 Clinical application of bone densitometry. Favus MJ (ed.) Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism, 4th ed. Lippincott Williams and Wilkins, Philadelphia, PA, USA, 152159.
  • 8
    Beck TJ, Ruff CB, Mourtada FA, Shaffer RA, Maxwell-Williams K, Kao G, Sartoris DJ, Brodine S 1996 Dual-energy x-ray absorptiometry derived structural geometry for stress fracture prediction in male U.S. marine corps recruits. J Bone Miner Res 11: 645653.
  • 9
    Melton LJ III, Beck TJ, Amin S, Khosla S, Achenbach SJ, Oberg AL, Riggs BL 2005 Contributions of bone density and structure to fracture risk assessment in men and women. Osteoporos Int 16: 460467.
  • 10
    Fehling PC, Alekel L, Clasey J, Rector A, Stillman RJ 1995 A comparison of bone mineral densities among female athletes in impact loading and active loading sports. Bone 17: 205210.
  • 11
    Aptel I, Cance-Rouzaud A, Grandiean H 1999 Association between calcium ingested from drinking water and femoral bone density in elderly women: Evidence from the EPIDOS cohort. J Bone Miner Res 14: 829833.
  • 12
    Lunt M, Masaryk C, Scheidt-Nave C, Nijs J, Poor G, Pols H, Falch JA, Hammermeister G, Reid DM, Benevolenskaya L, Weber K, Cannata J, O'Neil TW, Felsenberg D, Silman AJ, Reeve J 2001 The effects of lifestyle, dietary dairy intake and diabetes on bone density and vertebral deformity prevalence: The EVOS study. Osteoporos Int 12: 688698.
  • 13
    Nguyen TV, Kelly PJ, Sambrook PN, Gilbert C, Pocock NA, Eisman JA 1994 Lifestyle factors and bone density in the elderly: Implications for osteoporosis prevention. J Bone Miner Res 9: 13391346.
  • 14
    Damilakis J, Perisinakis K, Kontakis G, Vagios E, Gourtsoyiannis N 1999 Effect of lifetime occupational physical activity on indices of bone mineral status in healthy postmenopausal women. Calcif Tissue Int 64: 112116.
  • 15
    Nguyen TV, Center JR, Eisman JA 2000 Osteoporosis in elderly men and women: Effects of dietary calcium, physical activity, and body mass index. J Bone Miner Res 15: 322331.
  • 16
    Heinonen A, Oja P, Kannus P, Sievanen H, Haapasalo H, Manttari A, Vuori I 1995 Bone mineral density in female athletes representing sports with different loading characteristics of the skeleton. Bone 17: 197203.
  • 17
    Uusi-Rasi K, Sievanen H, Vuori I, Pasanen M, Heinonen A, Oja P 1998 Associations of physical activity and calcium intake with bone mass and size in healthy women at different ages. J Bone Miner Res 13: 133142.
  • 18
    Devine A, Dhaliwal SS, Dick I, Bollerslev J, Prince R 2004 Physical activity and calcium consumption are important determinants of lower limb bone mass in older women. J Bone Miner Res 19: 16341639.
  • 19
    Nelson ME, Fiatarone MA, Morganti CM, Trice I, Greenberg RA, Evans WJ 1994 Effects of high-intensity strength training on multiple risk factors for osteoporotic fractures. JAMA 272: 19091914.
  • 20
    Heinonen A, Kannus P, Sievanen H, Oja P, Pasanen M, Rinne M, Uusi-Rasi K, Vuori I 1996 Randomised controlled trial of effect of high-impact exercise on selected risk factors for osteoporotic fractures. Lancet 348: 13431347.
  • 21
    Prince R, Devine A, Dick I, Criddle A, Kerr D, Kent N, Price R, Randell A 1995 The effects of calcium supplementation (milk powder or tablets) and exercise on bone density in postmenopausal women. J Bone Miner Res 10: 10681075.
  • 22
    Riggs BL, O'Fallon WM, Muhs J, O'Conner MK, Kumar R, Melton LJ III 1998 Long-term effects of calcium supplementation on serum parathyroid hormone level, bone turnover, and bone loss in elderly women. J Bone Miner Res 13: 168174.
  • 23
    Crabtree NJ, Kroger H, Martin A, Pols H, Lorenc R, Nijs J, Stepan JJ, Falch JA, Miazgowski T, Grazio S, Raptou P, Adams J, Collings A, Khaw KT, Rushton N, Lunt M, Dixon AK, Reeve J 2002 Improving risk assessment: Hip geometry, bone mineral distribution and bone strength in hip fracture cases and controls. The EPOS study. Osteoporos Int 13: 4854.
  • 24
    Noble BS, Peet N, Stevens HY, Brabbs A, Mosley JR, Reilly GC, Reeve J, Skerry TM, Lanyon LE 2003 Mechanical loading: Biphasic osteocyte survival and targeting of osteoclasts for bone destruction in rat cortical bone. Am J Physiol Cell Physiol 284: C934C943.
  • 25
    Beck TJ 2003 Hip Structural Analysis (HSA) Program (BMD and Structural Geometry Methodology). Department of Radiology, School of Medicine, Johns Hopkins Medical Institute, Baltimore, MD, USA.
  • 26
    Bruce DG, Devine A, Prince R 2002 Recreational physical activity levels in healthy older women: The importance of fear of falling. J Am Geriatr Soc 50: 8489.
  • 27
    McArdle W, Katch F, Katch V 1991 Energy, Nutrition and Human Performance. Lea & Febiger, Philidelphia, PA, USA.
  • 28
    Pollock M, Wilmore J, Fox S 1978 Health and Fitness Through Physical Activity. Wiley, New York, NY, USA.
  • 29
    Hodge A, Patterson AJ, Brown WJ, Ireland P, Giles G 2000 The Anti Cancer Council of Victoria FFQ: Relative validity of nutrient intakes compared with weighed food records in young to middle-aged women in a study of iron supplementation. Aust N Z J Public Health 24: 576583.
  • 30
    Kaptoge S, Dalzell N, Loveridge N, Beck TJ, Khaw KT, Reeve J 2003 Effects of gender, anthropometric variables, and aging on the evolution of hip strength in men and women aged over 65. Bone 32: 561570.
  • 31
    Khoo BCC, Beck TJ, Qiao Q, Parakh P, Semanick L, Prince R, Singer KP, Price R 2005 In vivo short-term precision of hip structural analysis variables in comparison with bone mineral density using paired dual-energy X-ray absorptiometry scans from multi-center trials. Bone 37: 112121.
  • 32
    Beck TJ, Stone KL, Oreskovic TL, Hochberg MC, Nevitt MC, Genant HK, Cummings SR 2001 Effects of current and discontinued estrogen replacement therapy on hip structural geometry: The study of osteoporotic fractures. J Bone Miner Res 16: 21032110.
  • 33
    Perneger TV 1998 What's wrong with Bonferroni adjustments. BMJ 316: 12361238.
  • 34
    Forwood MR, Turner CH 1995 Skeletal adaptation to mechanical usage. Bone 17 (4 Suppl): 197S205S.
  • 35
    Martin RB, Burr DB, Sharkey NA 1998 Skeletal Tissue Mechanics. Springer-Verlag, New York, NY, USA.
  • 36
    Technical Reference Group CAIT 2003 Physical Activity in Australia Population Health Monitoring and Surveillance: Question Development Background Paper. Australian Institute of Health and Welfare, Canberra, Australia.
  • 37
    Bell KL, Loveridge N, Power J, Rushton N, Reeve J 1999 Intracapsular hip fracture: Increased cortical remodeling in the thinned and porous anterior region of the femoral neck. Osteoporos Int 10: 248257.
  • 38
    Beck T, Looker A, Mourtada F, Dapthary M, Ruff C 2006 Age trends in femur stresses from a simulated fall on the hip among men and women: Evidence of homeostatic adaptation underlying the decline in hip BMD. J Bone Miner Res 21: 14251432.
  • 39
    Mayhew PM, Thomas CD, Clement JG, Loveridge N, Beck TJ, Bonfield W, Burgoyne CJ, Reeve J 2005 Relation between age, femoral neck cortical stability, and hip fracture risk. Lancet 366: 129135.
  • 40
    Petit MA, McKay HA, MacKelvie KJ, Heinonen A, Khan KM, Beck TJ 2002 A randomised school-based jumping intervention confers sit and maturity-specific benefits on bone structural properties in girls: A hip structural analysis study. J Bone Miner Res 17: 363372.