SEARCH

SEARCH BY CITATION

Keywords:

  • biomechanics;
  • detraining;
  • growth and development;
  • mechanical loading;
  • osteoporosis

Abstract

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

Short-term exercise in growing rodents provided lifelong benefits to bone structure, strength, and fatigue resistance. Consequently, exercise when young may reduce the risk for fractures later in life, and the old exercise adage of “use it or lose it” may not be entirely applicable to the skeleton.

Introduction: The growing skeleton is most responsive to exercise, but low-trauma fractures predominantly occur in adults. This disparity has raised the question of whether exercised-induced skeletal changes during growth persist into adulthood where they may have antifracture benefits. This study investigated whether brief exercise during growth results in lifelong changes in bone quantity, structure, quality, and mechanical properties.

Materials and Methods: Right forearms of 5-week-old Sprague-Dawley rats were exercised 3 days/week for 7 weeks using the forearm axial compression loading model. Left forearms were internal controls and not exercised. Bone quantity (mineral content and areal density) and structure (cortical area and minimum second moment of area [IMIN]) were assessed before and after exercise and during detraining (restriction to home cage activity). Ulnas were removed after 92 weeks of detraining (at 2 years of age) and assessed for bone quality (mineralization) and mechanical properties (ultimate force and fatigue life).

Results: Exercise induced consistent bone quantity and structural adaptation. The largest effect was on IMIN, which was 25.4% (95% CI, 15.6–35.3%) greater in exercised ulnas compared with nonexercised ulnas. Bone quantity differences did not persist with detraining, whereas all of the absolute difference in bone structure between exercised and nonexercised ulnas was maintained. After detraining, exercised ulnas had 23.7% (95% CI, 13.0–34.3%) greater ultimate force, indicating enhanced bone strength. However, exercised ulnas also had lower postyield displacement (−26.4%; 95% CI, −43.6% to −9.1%), indicating increased brittleness. This resulted from greater mineralization (0.56%; 95% CI, 0.12–1.00%), but did not influence fatigue life, which was 10-fold greater in exercised ulnas.

Conclusions: These data indicate that exercise when young can have lifelong benefits on bone structure and strength, and potentially, fracture risk. They suggest that the old exercise adage of “use it or lose it” may not be entirely applicable to the skeleton and that individuals undergoing skeletal growth should be encouraged to perform impact exercise.


INTRODUCTION

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

The ability of bone to respond to mechanical stimuli has been known for over a century, yet the clinical importance of this response continues to be debated.(1,2) Reduced bone strength is predominantly an age-related phenomenon,(3) whereas the ability of the skeleton to respond to mechanical loading decreases with age.(4) In fact, the skeletal benefit of a lifetime of exercise seems to occur mainly during the years of skeletal development.(5,6) The diminishing response of the skeleton to exercise during aging has raised the question of whether exercise-induced bone changes during growth persist into adulthood where they may have antifracture benefits. Initial clinical studies have shown that exercise-induced changes in the skeleton during growth may not be maintained long term after exercise cessation.(7–10) However, these studies primarily used DXA to assess bone status. DXA undoubtedly provides a reasonable picture of overall bone quantity; however, it has significant limitations in the assessment of the skeletal response to exercise.(11) Exercise generates large increases in bone strength without substantial increases in bone quantity because it causes new bone tissue to be placed where mechanical demands are greatest.(12,13) This site-specific adaptation is functionally important because its enables the skeleton to fulfil its mechanical requirements without overtly increasing its overall weight.(14) DXA does not provide an adequate measure of bone structure because of its inherently planar nature and low-spatial resolution. Bone strength, and the consequent risk for fracture, is dependent on not only how much bone is present (quantity), but also the distribution (structure) and composition (quality) of this bone. The aim of this study was to investigate whether a short period of skeletal exercise during growth results in lifelong changes in bone quantity, structure, and quality and whether these changes confer lifelong benefits on bone mechanical properties.

MATERIALS AND METHODS

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

Animals

Thirty-two female Sprague-Dawley rats (Harlan Sprague-Dawley) were acclimatized until 5 weeks of age before experimentation. All procedures were performed after approval of the Institutional Animal Care and Use Committee of Indiana University.

Forearm exercise

Forearm exercise was started at 5 weeks of age using an established axial compression loading model.(15) Initial peak strains were determined through a load-strain calibration experiment in four rats, as previously described.(13) Right forearms of remaining animals were exercised by axially loading across the olecranon and flexed carpus with the animal under isoflurane anesthesia (Abbott Laboratories). Exercise was introduced using an electromechanical actuator (Bose ElectroForce 3200 series; EnduraTEC), and applied using a 2-Hz haversine waveform for 360 cycles/day, 3 days/week for 7 weeks. The initial peak load was 8.4 N, which elicited a compressive strain (ε) of 3500 με on the medial surface of the ulna midshaft (as determined from the preliminary strain gauge experiment). This microstrain level is sufficient to induce bone adaptation in the rat ulna.(12,13,16,17) To counteract the increase in bone size associated with rapid growth in our animals, and consequent decrease in strain per given load, the peak load magnitude was increased incrementally in proportion to the increase in animal body weight. Left ulnas served as internal controls and were not exercised (nonexercised). Normal cage activity was allowed between exercise sessions and throughout detraining.

In vivo assessment of bone quantity and structure

Adaptation to exercise and bone status during detraining were determined in vivo by assessing both the right (exercised) and left (nonexercised) forearms using DXA and pQCT. Assessments were performed before (baseline) and at the completion of the exercise program (after exercise) to determine the extent of bone adaptation to exercise. During detraining, scans were performed at 2-week intervals for the first 6 weeks, 4-week intervals for the next 8 weeks, and 6-week intervals thereafter until a total detraining period of 92 weeks. Animals were anesthetized for assessments using ketamine (40–90 mg/kg) and xylazine (5–13 mg/kg) introduced intraperitoneally.

DXA was performed using a PIXImus II mouse densitometer (Lunar). Each forearm was scanned, and a region of interest box positioned over the forearm (ulna and radius) from the radial head to the radiocarpal joint. Ulna/radius BMC (mg) and areal BMD (aBMD; g/cm2) were collected. Short-term reliability for this procedure in four animals scanned four times with interim repositioning showed CVs of 2.9 ± 0.8% and 1.5 ± 0.6% (SD) for BMC and aBMD, respectively.

pQCT was performed using a Norland Stratec XCT Research SA+ pQCT (Stratec Electronics). Animals were positioned on a custom scanning platform and each forearm centered in the machine gantry. A cross-sectional scan was performed at the ulna midshaft using a 0.46-mm collimation and 70-μm voxel size. Analyses were restricted to the ulna cortical bone, with the bone edge detected using contour mode 1 at a threshold of 400 mg/cm3. Cortical area (Ct.Ar; mm2) was recorded for each bone, and the minimum (IMIN; mm4) and maximum (IMAX; mm4) second moments of area derived according to Gere and Timoshenko.(18) Short-term reliability for this procedure in four animals scanned four times with interim repositioning showed CVs of 1.8 ± 0.6%, 3.7 ± 2.0%, and 3.6 ± 1.6% for Ct.Ar, IMIN, and IMAX, respectively. In addition, forearm muscle cross-sectional area was determined by reassessing the ulna midshaft scans using contour mode 1 with an outer threshold of 30 mg/cm3 for soft tissue edge detection and peel mode 2 with an inner threshold of 400 mg/cm3 for bone edge detection.

Ex vivo assessment of bone structure

Animals were killed after 92 weeks of detraining (at 2 years of age), and the exercised and nonexercised ulnas dissected free and stored in 70% alcohol. pQCT was performed using the equipment and scanning parameters described above to assess bone structure (IMIN) at 1-mm increments along the entire bone length. In addition, humeral diaphyseal structure (Ct.Ar) in the exercised and nonexercised forearms was assessed by taking a cross-sectional scan at one third of the bone length from its distal end. Analyses were restricted to the cortical bone, and the bone edge detected using contour mode 1 with a threshold of 400 mg/cm3.

Ex vivo assessment of bone strength

Exercised and nonexercised ulna pairs from 10 animals were mechanically tested in axial compression, as previously described.(12) Ulnas were marked before testing at one-sixth intervals to permit site localization for poststrength testing measurements. Bones were rehydrated and slowly brought to room temperature overnight in a saline bath, photographed with a digital camera, and fixed using a −0.1 N static preload between two cup-shaped platens of a materials testing machine (Vitrodyne V1000; Liveco). Loading to failure was performed using monotonic compression at 2 mm/s during which force and displacement measurements were collected every 0.01 s. From the force versus displacement curves, ultimate force (N), yield force (N), stiffness (N/mm), energy to failure (mJ), postyield energy (mJ), and postyield displacement (mm) were derived. The yield point was defined using a 0.015-mm offset parallel to the stiffness.(19) The broken pieces of each ulna were rephotographed and replaced in 70% alcohol in preparedness for histological processing. The digital images of each bone were imported into Scion Image for Windows (Beta 4.0.2; Scion Corp.) wherein fracture position was measured as a percent of total bone length (expressed as percent from the distal end). In addition, bone curvature was determined from the pretest images, as previously described.(20)

Ex vivo assessment of bone fatigue

Exercised and nonexercised ulna pairs from the remaining 10 animals were tested until fatigue failure in axial compression, as previously described.(13) The ulna pairs were loaded at a constant peak axial load of 50 N/kg body weight until fatigue failure. This load was chosen to ensure that the specimens would break within a practical limit of 1–2 × 106 cycles. Fatigue failure was defined as occurring when the bone failed completely and could no longer withstand the applied load. For testing, ulnas were rehydrated and brought to room temperature overnight in a saline bath, fixed with 0.5 N of preload between two cup-shaped platens on an electromagnetic actuator (Bose ElectroForce 3200 series; EnduraTEC) and axially loaded until fatigue failure. Loading was performed in a room temperature (20°C) saline bath using a 5-Hz haversine waveform. The number of cycles until fatigue failure was recorded for each bone. After testing, the broken pieces of each ulna were replaced in 70% alcohol in preparedness for assessment of ash content.

Ex vivo assessment of bone gain during detraining

Intraperitoneal injections of calcein (7 mg/kg body mass; Sigma Chemical Co.) and alizarin (30 mg/kg body mass; Sigma Chemical Co.) were administered during the second week of exercise and at 6 months of age to allow histomorphometric assessments, respectively. Distal bone sections taken from ulna pairs that were strength tested (N = 10 pairs) were embedded in methylmethacrylate, as previously described.(16,17) Transverse bone sections were taken at 33% of the length of the ulna from its distal end, as indicated by an earlier placed mark. This site was within the region of maximal maintenance of exercise effects (see Results), and no ulnas fractured at this location during destructive mechanical testing. Six transverse thin (4 μm) sections were taken using a microtome (Reichert-Jung 2050) to assess bone quality through synchrotron infrared microspectroscopy (see Ex vivo assessment of bone quality). A transverse thick section was removed using a diamond-embedded wire saw (Histo-saw; Delaware Diamond Knives) to assess bone gain during detraining. This section was ground to a final thickness of 40–50 μm and viewed at ×200 magnification on a fluorescence microscope (Nikon Optiphot; Nikon). Total bone area at 6 months of age was determined by tracing the perimeter of the intracortical alizarin bone label, whereas total bone area after detraining was determined by tracing the periosteal and endocortical bone perimeters. Absolute bone gain during detraining (mm2) was determined as bone area after detraining minus bone area at 6 months of age.

Ex vivo assessment of bone quality

Bone quality was assessed by determining ash content of ulna pairs that were fatigue tested (N = 10 pairs) and performing synchrotron infrared microspectroscopy of bone sections taken from ulna pairs that were strength tested (N = 10 pairs). Ash content was determined by placing bone segments from exercised and nonexercised ulnas in a furnace oven (Thermolyne 1400). The oven was set at 100°C for 24 h to remove water, followed by 800°C for 24 h to remove organic material. Bone segments were weighed after the first and second 24-h periods to obtain dry and ash weight, respectively. Percent ash content was calculated as (ash weight/dry weight) × 100.

Synchrotron infrared microspectroscopy assessed the composition of ulna thin (4 μm) sections taken from exercised and nonexercised ulna pairs (N = 10 pairs). Sections were mounted between two aluminium plates containing a slit that permitted light to transmit through the section. The plates were placed in a compression cell on a motorized stage of an infrared (IR) microscope coupled to a Fournier transformed infrared (FTIR) spectrometer (Thermo Nicolet Instruments). Spectral mapping was accomplished using a pair of ×32 Schwarzchild objectives in a confocal arrangement, where a 10-μm aperture was used to illuminate the tissue with IR light. Synchrotron light from beamline U10B at the National Synchrotron Light Source (Brookhaven National Laboratory, Upton, NY, USA) replaced the conventional global source for IR spectral measurements.

Fifteen spectra were obtained in transmission mode from the middle of the medial cortex of each ulna, with each spectrum representing the accumulation of 256 co-added IR spectra. The final format of all spectral data were absorbance, where the background spectrum was collected through an empty sample holder. IR spectra were collected using a MCT-B detector (infrared region 4000–500 cm−1) cooled with liquid nitrogen. The chemical composition of bone tissue from the ulnas was determined by area integrations carried out on protein (Amide I: 1688–1623 cm−1; linear baseline at 1800 cm−1), carbonate (v2CO3−2: 905–825 cm−1; baseline at 905–825 cm−1), and phosphate (v4PO4−3: 650–500 cm−1; baseline 650–500 cm−1) IR bands,(21,22) and phosphate-to-protein and carbonate-to-protein ratios derived. Spectral bands from 1700 to 1760 cm−1 representing carbonyl (C = 0) stretching from plastic embedding material were identified in some sections and filtered.

Statistical analyses

Exercise effects on bone quantity, structure, quality, and strength were determined with mean percent differences and 95% CIs of the mean percent differences between exercised and nonexercised ulnas. 95% CI that did not cross 0% were considered statistically significant, as indicated by single sample t-tests on the mean percent differences with a population mean of 0%. Differences in bone structure between exercised and nonexercised ulnas at select time-points during the study were determined using paired t-tests. The Wilcoxon signed ranks test was used to assess differences in bone fatigue life between exercised and nonexercised ulnas, because values were not normally distributed. All statistical analyses were performed with the Statistical Package for Social Sciences software (SPSS), and all comparisons were two-tailed with a level of significance set at 0.05.

RESULTS

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

Animal characteristics

Eighty-two percent of animals (23 of 28) developed one or more benign mammary tumors during aging, consistent with previous reports.(23,24) These tumors occurred with equal frequency on the exercised and nonexercised sides of the body and were removed surgically by a veterinarian. Eight animals died before completion of the 92-week detraining period. Four animals died as a result of anesthetic complications, whereas four others died of natural causes. Necropsy by a veterinarian determined death in two of these latter animals to be caused by malignant cancer. No determinable cause of death was found in the other two animals. Data from these eight animals were excluded from analyses. Body weight in the remaining 20 animals at baseline, after exercise, and after detraining was 107.5 ± 8.6, 213.2 ± 13.8, and 354.4 ± 48.2 (SD) g, respectively.

Lifelong effects of exercise during growth on bone quantity and structure

In vivo bone status at baseline, after exercise, and after detraining is shown in Fig. 1. There were no side-to-side differences at baseline (all p = 0.20–0.91); however, bone quantity (aBMD and BMC) and structure (Ct.Ar and IMIN) were enhanced in exercised forelimbs after exercise (all p < 0.05). In terms of bone structure, exercise primarily induced a change in bone shape as opposed to bone size. The change in bone shape is indicated by the large (25.4%; 95% CI, 15.6–35.3%) after exercise difference in IMIN (p < 0.001), yet absence of an after exercise difference in IMAX (p = 0.13). This plane-selective distribution of exercise effects meant that there was only a small increase in actual bone size, as indicated by small (4.4%; 95% CI, 0.7–8.2%) postexercise differences in Ct.Ar (p < 0.05). Exercise-induced changes in bone quantity measures did not persist after detraining (all p = 0.17–0.34). In contrast, there was long-term maintenance of exercise-induced bone structural changes (Ct.Ar and IMIN; all p < 0.01). The latter is confirmed in Figs. 2A and 2B, which show that in vivo bone structural measures in exercised and nonexercised ulnas did not convergence over time. When assessed ex vivo after exercise, bone structural changes were found to have been predominantly induced and maintained in the distal portions of the bone (Fig. 2C).

thumbnail image

Figure Figure 1. Exercise during growth provides long-term benefits to bone structure but not quantity. Data show in vivo bone quantity and structure measures at baseline, after exercise, and after detraining (N = 10 animals). There were no side-to-side differences at baseline, improved bone quantity and structure in exercised forearms after exercise, and maintenance of the improved bone structure in exercised forearms after detraining. Data represent the mean percent difference between exercised and nonexercised forearms, with error bars indicating 95% CIs of the mean percent differences. CIs >0% indicate greater bone quantity and structure in exercised bones, whereas those <0% favor the nonexercised bones. Significance indicated by *p < 0.05; **p < 0.01; ***p < 0.001, as determined by single sample t-tests with a population mean of 0%.

Download figure to PowerPoint

thumbnail image

Figure Figure 2. Exercise-induced changes in bone structure are maintained long-term and are localized in the distal portions of the bone (N = 10 animals). Midshaft (A) Ct.Ar and (B) IMIN in exercised and nonexercised ulnas assessed in vivo at select time-points during the study. There was no evidence of convergence of values over time, indicating the absence of a “catch-up” phenomenon whereby nonexercised bones caught up the bone gain of exercised bones with age and concomitant growth. (C) Percent difference in IMIN between exercised and nonexercised ulnas at 1-mm increments along the bone length measured ex vivo after detraining. Exercise effects were predominantly induced and maintained in the distal portions of the bone. Data in A and B represent mean values, with error bars indicating SD. Data in C represent the mean percent difference between exercised and nonexercised forearms, with error bars indicating 95% CIs of the mean percent differences. CIs >0% indicate greater bone quantity and structure in exercised bones, whereas those <0% favor the nonexercised bones. Significance indicated by *p < 0.05; **p < 0.01; ***p < 0.001, as determined by (A and B) paired t-tests and (C) single sample t-tests with a population mean of 0%.

Download figure to PowerPoint

Bone sections showed the absolute gain in bone area during detraining to be equivalent between exercised and nonexercised ulnas (0.53 ± 0.10 versus 0.52 ± 0.10 mm2, respectively; p = 0.49; Fig. 3). Histological measures also confirmed that exercise effects were primarily mediated by bone apposition on the periosteal surface, as opposed to the endosteal surface, with exercised ulnas having elevated periosteal bone perimeter at 6 months of age (exercise = 5.16 ± 0.27 versus nonexercise = 4.95 ± 0.27 mm; p = 0.02). Exercised and nonexercised ulnas did not differ in curvature (4.51 ± 0.71° versus 4.45 ± 0.75°, respectively; p = 0.80). Similarly, forearm muscle cross-sectional area did not differ between exercised and nonexercised forelimbs at any time during the study (all p > 0.05, data not shown), and humeral diaphyseal Ct.Ar in exercised forelimbs was not enhanced after detraining (exercise = 4.65 ± 0.13 mm2 versus nonexercise = 4.70 ± 0.13 mm2; p = 0.25).

thumbnail image

Figure Figure 3. Representative histological images of the nonexercised and exercised ulna under (A) fluorescent, (B) white, and (C) polarized light. Sections were taken one third of the length of the ulna from its distal end. (A) Under fluorescent light, note the diffuse calcein (green) labeling in both ulnas. This labeling is indicative of woven bone formation and is consistent with the rapid skeletal growth that was taking place at the time of calcein administration (animal age = 6 weeks). In contrast, alizarin labeling formed a relatively uniform double label, indicating that rapid growth had declined and that bone formation was more uniform at the time of administration (animal age = 6 months). There were no differences in the absolute area of bone formed by nonexercised and exercised ulnas between 6 months (alizarin label) and 2 years of age (arrows, after detraining), indicating that there was no “catch-up” phenomenon whereby nonexercised bones caught up the bone gain of exercised bones. (B) Under white light, note the increased cortical area of the exercised ulna and, in particular, the significantly greater distribution of bone in the mediolateral plane. This plane approximates the plane of least bending resistance (IMIN) in the rat ulna. (C) Under polarized light, note the relatively disorganized and lamellar collagen alignments in the regions of calcein and alizarin labeling, respectively.

Download figure to PowerPoint

Bone strength after detraining

Exercised ulnas had greater strength (ultimate force) and stiffness after detraining than nonexercised ulnas (all p < 0.01; Figs. 4A and 4B). However, exercised ulnas exhibited reduced postyield displacement (p < 0.01). Fracture location was significantly more proximal in exercised ulnas compared with nonexercised ulnas (49.8 ± 2.5% versus 42.5 ± 6.3% of total bone length from the distal end, respectively; p = 0.02; Fig. 4C). Bone structure (IMIN) explained 76% of the variance in ultimate force (Fig. 5A), whereas bone quantity (DXA-derived BMC) explained 29% of this variance (Fig. 5B).

thumbnail image

Figure Figure 4. Exercise during growth provides long-term benefit to bone strength (N = 10 animals). (A) Representative force vs. displacement curves for a pair of exercised and nonexercised ulnas after detraining. The exercised ulna had greater ultimate force (peak of the curve on the y-axis) and stiffness (slope of the linear portion of the curve). The nonexercised ulna exhibited greater postyield displacement (displacement range between yield point and failure), which caused a larger amount of energy to be absorbed before failure (area under the curve). (B) Bone strength measures after detraining. Despite 92 weeks of detraining, ulnas exercised during growth had greater strength (ultimate force) and stiffness than nonexercised ulnas. In contrast, exercised ulnas had less postyield displacement, indicating greater brittleness than nonexercised ulnas. Data represent the mean percent difference between exercised and nonexercised forearms, with error bars indicating 95% CIs of the mean percent differences. CIs >0% indicate greater bone strength in exercised bones, whereas those <0% favor the nonexercised bones. Significance indicated by **p < 0.01, as determined by single sample t-tests with a population mean of 0%. (C) Fracture location during mechanical testing in ulna pairs. Ulnas broke significantly more proximal in exercised ulnas compared with contralateral nonexercised ulnas. Image on the right shows the mean (thickened horizontal line) and 95% CIs (box) for fracture location in exercised and nonexercised ulnas. Significance indicated by *p < 0.05, as determined by a paired t-test.

Download figure to PowerPoint

thumbnail image

Figure Figure 5. Bone strength (ultimate force) after detraining was highly related to (A) bone structure (midshaft minimum second moment of area; IMIN), but (B) poorly related to bone quantity (whole ulna BMC; N = 10 animals). This indicates that bone structure is a better predictor of bone strength changes in response to exercise and subsequent detraining.

Download figure to PowerPoint

Bone quality and fatigue life after detraining

Exercised ulnas had significantly greater whole bone ash content (p < 0.05; Fig. 6A), indicating greater tissue mineralization. This was corroborated by regional analyses using synchrotron infrared microspectroscopy that found exercised ulnas to have greater localized phosphate-to-protein and carbonate-to-protein ratios, indicators of bone mineralization (all p < 0.05; Fig. 6A). During fatigue testing, exercised ulnas had 10 times greater fatigue resistance than nonexercised ulnas (p < 0.05; Fig. 6B).

thumbnail image

Figure Figure 6. Exercise during growth impairs bone quality, but has long-term benefit to bone fatigue resistance (N = 10 animals). (A) Bone quality measures after detraining indicate greater whole bone (ash content) and localized (phosphate-to-protein) mineralization in exercised ulnas. Data represent the mean percent difference between exercised and nonexercised forearms, with error bars indicating 95% CIs of the mean percent differences. CIs >0% indicate greater bone mineralization in exercised bones. (B) The increased mineralization in exercised ulnas had no effect on bone function, as determined by fatigue life tests. Cycles to fatigue failure (fracture) is plotted on a log scale and was 10-fold greater in exercised ulnas. Data represent mean values, with error bars indicating SD. Significance indicated by *p < 0.05; **p < 0.01, as determined by (A) single sample t-tests with a population mean of 0% and (B) Wilcoxon signed ranks test.

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 found short-term exercise in rapidly growing rodents to provide lifelong benefits to bone structure, strength, and fatigue resistance. Skeletal exercise was achieved using an established axial compression loading model.(15) This model increases the stress distribution within the ulna and induces bone adaptation with similar locality and magnitude as observed in the rat ulna with functional activities.(25,26) After the completion of a 7-week exercise program, rats were restricted to home cage physical activities for a detraining period of 92 weeks (until 104 weeks of age). Thus, they were senescent by the end of the study because <60% of female Sprague-Dawley rats in captivity typically survive beyond 2 years of age.(27) The exercise-induced changes in bone quantity measures did not persist after detraining; however, there was long-term maintenance of exercise-induced bone structural changes. These structural changes endowed the exercised bones with enhanced functional properties, as indicated by increased strength and fatigue life.

The percent difference between the exercised and nonexercised ulnas in structural variables generally decreased with detraining; however, the absolute difference induced by exercise remained constant. For instance, 100% of the absolute difference in IMIN between ulnas after exercise was maintained throughout detraining, and there was no evidence of convergence of values over time. The latter indicates that there was no “catch-up” phenomenon whereby nonexercised bones caught up the bone gain of exercised bones with age and concomitant growth. This finding was confirmed histologically from labeled bone sections, which showed the absolute gain in bone area during detraining to be equivalent between exercised and nonexercised ulnas. These findings contrast those of Järvinen et al.(28,29) who failed to find long-term maintenance of exercise-induced bone changes within the rat femoral neck and suggested that nonexercised animals catch-up the bone advantage of exercised animals. Possible explanations for the current disparate findings include the site studied (ulna diaphysis versus femoral neck), sex (female versus male rats), mode of exercise (axial compression loading versus treadmill running), and study design. In terms of the latter, Järvinen et al.(28,29) used a between-animal study design. In contrast, this study used a within-animal study design wherein left forearms were not loaded and represented a powerful comparative group for internal control of genetic and environmental factors.

For the maintenance of structural differences after exercise to have functional relevance they need to be associated with the lifelong maintenance of exercise-induced changes in bone strength. Isolated ulnas were mechanically tested after detraining in the same direction that they were exercised and subsequently adapted during growth (axial compression). Ex vivo axial compression of the rat ulna generates the same strain pattern as engendered during in vivo axial loading of the rat forearm.(30) Exercised ulnas had greater strength than nonexercised ulnas, as indicated by greater ultimate force. As exercise effects were predominantly induced and maintained in the distal portion of the ulna, this region was endowed with greater fracture resistance. This was reflected during mechanical testing by fracture location being significantly more proximal in exercised ulnas compared with nonexercised ulnas. Confirming the importance of bone structure to bone strength, IMIN explained 76% of the variance in ultimate force. In comparison, DXA-derived BMC of the ulna explained only 29% of the variance in ultimate force. These findings indicate that exercise during growth could have long-term antifracture benefits through the structural changes it induces, independent of any long-term effects on bone mass.

A potential caveat to the observed lifelong benefit of exercise on bone strength was a seemingly negative effect of exercise on bone quality. Bone adaptation in response to exercise in the rat ulna is typically associated with enhanced postyield displacement(12); however, when exercise was coupled with detraining in this study exercised ulnas exhibited reduced postyield displacement. Postyield displacement indicates the ductility of a material before failure, with a material that exhibits less postyield displacement being more brittle.(31) The decrease in postyield displacement in exercised ulnas did not result from exercised-induced changes in ulna curvature; rather, it resulted from differences in bone tissue quality. Exercised ulnas had significantly greater whole bone ash content, indicating greater tissue mineralization. This was corroborated by regional analyses using synchrotron infrared microspectroscopy that found exercised ulnas to have greater phosphate-to-protein and carbonate-to-protein ratios. Increases in phosphate-to-protein and carbonate-to-protein ratios signify bone that is more mineralized and is typically found with bone that is more mature.(32)

The functional consequence of greater mineralization and increased brittleness in exercised ulnas may be that exercise during growth followed by detraining may create bones that are more susceptible to developing and propagating load-related (or exercise-related) damage.(33) We tested this theory by assessing the fatigue life of our ulnas in axial compression. Ulnas were loaded until fatigue failure in load control (50 N/kg body weight), as opposed to strain control, because the former better represents the requirements of the bone during function. Results showed exercised ulnas to have 10 times greater fatigue resistance than nonexercised ulnas. This indicates that the greater strength in exercised ulnas, and consequent decreased strain per given load, endowed these bones with improved fatigue properties.

The observed maintenance of structural and mechanical changes induced by skeletal exercise in our animal model is clinically relevant and translatable. Skeletal exercise was introduced using an axial compression loading model. While this model does not truly represent exercise as it does not involve physical exertion on behalf of the animal, the resultant bone stress and skeletal adaptation observed is consistent with that resulting from impact activities in rats.(26) Thus, the model provides a useful reproduction of the loading environment and subsequent skeletal effects of impact activities in humans, such as experienced during jumping and racquet sports. Skeletal exercise was for 3 minutes/day and 3 days/week, consistent with clinical studies showing skeletal benefits of impact (jumping) exercise in growing children.(34,35) Exercise during growth in humans positively alters bone structure by causing new bone to be preferentially laid down on the periosteal surface,(36–38) consistent with our animal model. Mechanisms exist for these exercise-induced structural benefits in humans to remain intact until senescence where they may have antifracture properties. Primarily, structural decay associated with age-related bone loss is principally mediated by endocortical and not periosteal surface changes.(39–43) Subsequently, exercise during growth has been advocated as a means of optimizing lifelong periosteal bone structure, with reduced periosteal bone apposition during growth being implicated in the pathogenesis of fragility fractures later in life.(43–46) The hypothesis of lifelong benefits of exercise on fracture risk has garnered initial clinical support, with former athletes >60 years of age having a lower risk of fragility fracture than matched controls.(47) Admittedly, the mechanisms for this observation were not explored, with maintenance of fracture modulating factors other than exercised-induced changes in bone structure being a possibility.

Several model-specific explanations for the observed maintenance of exercise effects warrant consideration. We considered the possibility that animals favored their exercised forelimbs during detraining, thereby facilitating the maintenance of their exercise-enhanced bone structure. Forearm muscle cross-sectional area did not differ between exercised and nonexercised forelimbs at any time during the study, and humeral diaphyseal structure in exercised forelimbs was not enhanced after detraining. These observations suggest that the rats did not favor one limb over the other. The maintenance of exercise effects may also be attributable to species selection. The rodent skeletal system lacks secondary remodeling of Haversian canals, which potentiates intracortical remodeling in human bone. Subsequently, rodents have limited ability to remodel and remove excess cortical bone after exercise. However, the rat ulna does remodel in response to various stimuli,(48–50) and we found evidence of intracortical remodeling in our bones in the form of elevated intracortical porosity (data not shown).

Overall, our findings suggest that exercise when young may provide lifelong benefits to bone structure and strength, and consequent fracture risk. As a result, the old exercise adage of use it or lose it may not be entirely true in relation to exercise effects on the skeleton, and individuals undergoing skeletal growth should be encouraged to perform impact exercise to potentiate their lifelong bone health.

Acknowledgements

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

The authors thank VR Barrett and A Jefcoat for vigilant maintenance of our animals; KW Condon for histological processing; and LM Miller for technical assistance with the U10B beamline. The National Synchrotron Light Source is funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract DE-AC01-98CH10886. This project was supported by the National Institutes of Health (R01 AR046530).

REFERENCES

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