The authors have no conflict of interest.
Improved Bone Structure and Strength After Long-Term Mechanical Loading Is Greatest if Loading Is Separated Into Short Bouts†
Article first published online: 1 AUG 2002
Copyright © 2002 ASBMR
Journal of Bone and Mineral Research
Volume 17, Issue 8, pages 1545–1554, August 2002
How to Cite
Robling, A. G., Hinant, F. M., Burr, D. B. and Turner, C. H. (2002), Improved Bone Structure and Strength After Long-Term Mechanical Loading Is Greatest if Loading Is Separated Into Short Bouts. J Bone Miner Res, 17: 1545–1554. doi: 10.1359/jbmr.2002.17.8.1545
- Issue published online: 2 DEC 2009
- Article first published online: 1 AUG 2002
- Manuscript Accepted: 14 MAR 2002
- Manuscript Revised: 27 FEB 2002
- Manuscript Received: 28 DEC 2001
- mechanical loading;
- bone adaptation;
- mechanical properties;
Mechanical loading presents a potent osteogenic stimulus to bone cells, but bone cells desensitize rapidly to mechanical stimulation. Resensitization must occur before the cells can transduce future mechanical signals effectively. Previous experiments show that mechanical loading protocols are more osteogenic if the load cycles are divided into several discrete bouts, separated by several hours, than if the cycles are applied in a single uninterrupted bout. We investigated the effect of discrete mechanical loading bouts on structure and biomechanical properties of the rat ulna after 16 weeks of loading. The right ulnas of 26 adult female rats were subjected to 360 load cycles/day, delivered in a haversine waveform at 17 N peak force, 3 days/week for 16 weeks. One-half of the animals (n = 13) were administered all 360 daily cycles in a single uninterrupted bout (360 × 1); the other half were administered 90 cycles four times per day (90 × 4), with 3 h between bouts. A nonloaded baseline control (BLC) group and an age-matched control (AMC) group (n = 9/group) were included in the experiment. The following measurements were collected after death: in situ mechanical strain at the ulna midshaft; ulnar length; maximum and minimum second moments of area (IMAX and IMIN) along the entire length of the ulnas (1-mm increments); and ultimate force, energy to failure, and stiffness of whole ulnas. Qualitative observations of bone morphology were made from whole bone images reconstructed from microcomputed tomography (μCT) slices. Loading according to the 360 × 1 and 90 × 4 schedules improved ultimate force by 64% and 87%, energy to failure by 94% and 165%, IMAX by 13% and 26% (in the middistal diaphysis), IMIN by 69% and 96% (in the middistal diaphysis), and reduced peak mechanical strain by 40% and 36%, respectively. The large increases in biomechanical properties occurred despite very low 5–12% gains in areal bone mineral density (aBMD) and bone mineral content (BMC). Mechanical loading is more effective in enhancing bone biomechanical and structural properties if the loads are applied in discrete bouts, separated by recovery periods (90 × 4 schedule), than if the loads are applied in a single session (360 × 1). Modest increases in aBMD and BMC can improve biomechanical properties substantially if the new bone formation is localized to the most biomechanically relevant sites, as occurs during load-induced bone formation.
THE SIZE, shape, and strength of the adult mammalian skeleton is determined in large part by the mechanical forces endured during growth.(1) Consequently, mechanical loading of the skeleton during childhood and adolescence can significantly enhance the accumulation of bone mass and potentially delay age-associated bone loss and fracture susceptibility later in life. The collection of cell- and tissue-level mechanisms responsible for this adaptive process in bone is known as mechanotransduction. Although many pieces of the mechanotransduction pathways are becoming elucidated, the entire process—from cellular reception of the physical signal to effector cell response—is not well understood.(2)
The initial step in mechanotransduction concerns the detection of a physical stimulus or “mechanoreception.” Successful mechanoreception in bone cells requires that specific conditions are met in two domains, one on the physical signal side and one on the cellular side. On the signal side, the signal must be presented to the cell with specific physical characteristics. For example, the signal must be dynamic rather than static and the signal's frequency and magnitude must exceed a minimum value if it is to be detected by the cell.(3–6) On the cellular side of mechanoreception, the bone cell must be in a receptive state to detect the stimulus. Bone cells desensitize rapidly to mechanical stimuli, to the point where subsequent mechanical signals that would otherwise be osteogenic are largely ignored by the cell. Animal limb bones loaded in vivo exhibit a large gain in bone mass when administered relatively few (10-50) load cycles per day, but as that number is increased beyond ∼50 cycles/day, very little additional bone is formed.(7–9) These data suggest that the cells are receptive to the first 50 or so load cycles (first few minutes of loading), but as the loading bout is continued uninterrupted, the cellular response is greatly diminished.
Recently, we have conducted a series of in vivo experiments addressing desensitization and resensitization of bone cells, focusing on how these processes can be manipulated to enhance bone formation. The results from initial experiments show that rat tibias exposed to 360 daily load cycles increased osteogenesis nearly twice as much if the cycles were delivered in four discrete bouts of 90 cycles/bout, compared with tibias administered 360 cycles in a single uninterrupted bout.(10) Presumably, the bone cells in the tibias exposed to a single bout of 360 cycles (360 × 1) were sensitive to the first 50 or so cycles, but as desensitization developed, the cells were unable to transduce the remainder (majority) of the cycles. Tibias exposed to four bouts of 90 cycles (90 × 4) were allowed 3 h between each of the short bouts, which was sufficient to partially resensitize the bone cells before the second and subsequent bouts were administered. The 90 × 4 schedule was clearly more osteogenic that the 360 × 1 schedule despite identical mechanical inputs (same force magnitude, frequency, and number of cycles per day), but the loading protocol lasted only a week. We sought to determine whether the bout-scheduling effects would be maintained in the long run by evaluating the effect of the two loading protocols (90 × 4 and 360 × 1) on bone biomechanical and structural properties after an extended (16 weeks) loading period. We hypothesized that the 90 × 4 schedule, which produced bones with initially high bone formation rates, would result in stronger, more geometrically optimized bones than the 360 × 1 schedule after 16 weeks of loading.
Previously, we reported small but significant (5.4% and 8.6%) increases in areal bone mineral density (aBMD) in whole ulnas that had been loaded in vivo according to the 360 × 1 and 90 × 4 schedules, respectively, for 16 weeks.(11) We hypothesized that these modest gains in aBMD would translate into substantial improvement in bone strength, provided that bone is formed preferentially where the mechanical strains are greatest.
MATERIALS AND METHODS
Forty-four virgin female Sprague-Dawley rats were purchased from Harlan Sprague-Dawley (Indianapolis, IN, USA). The rats arrived at the age of 12 weeks, and were housed two per cage at Indiana University's Laboratory Animal Resource Center for 15 weeks before the start of the experiment. Water and standard rat chow were provided ad libitum during the acclimation and experimental periods. Body mass, collected throughout the acclimation and experimental periods, has been reported previously.(11) Under ether-induced anesthesia, the right distal forelimb of rats in the loading groups was subjected to axially applied compressive loads using a nonsurgical loading preparation that transmits mechanical force to the ulna through the olecranon and flexed carpus (Fig. 1A).(12) The natural curvature of the ulnar diaphysis results in a bending moment in the middiaphysis during end loading that produces tension in the lateral cortex and compression in the medial cortex. Load was applied as a haversine waveform at a frequency of 2 Hz and peak load magnitude of 17 N using an open loop stepper motor-driven spring linkage. All procedures performed in this experiment were in accordance with Indiana University Animal Care and Use Committee guidelines.
The rats were divided randomly into two loaded groups (n = 13/group) and two control groups (n = 9/group). The right ulnas of animals in the two loaded groups were subjected to 360 load cycles/day, 3 days/week, for 16 consecutive weeks. The two loaded groups differed from one another only in the timed delivery of the 360 load cycles received throughout each load day. One group was administered all 360 cycles in a single, uninterrupted session (360 × 1), which lasted 3 minutes. The other loaded group was administered the 360 cycles in four discrete bouts of 90 cycles/bout (90 × 4), with 3 h of recovery inserted between each of the brief (45 s long) loading bouts. All rats were allowed normal cage activity between bouts. The two control groups comprised a baseline control (BLC) group, which was killed on the first loading day, and an age-matched control (AMC) group, which was killed on the same day that the loaded groups were killed (16 weeks after baseline death). Neither control group was subjected to loading or anesthesia. In the loaded groups, the left ulnas were not loaded and served as internal controls for the loaded limb.
Ex vivo strain gauging
Immediately after (within 10 minutes of) death, the right distal forearm was minimally dissected to expose the medial surface of the midshaft ulna. The periosteum was removed and a single element strain gauge (EA-06-015DJ-120; Measurements Group, Inc., Raleigh, NC, USA) was fixed to the exposed medial ulnar surface with cyanoacrylate (M-Bond 200; Measurements Group, Inc.) at a point 6.8 mm distal to the brachialis insertion. Preliminary studies in our laboratory showed that this was the most reliable method for placing the gauge at midshaft in the fleshed forearm, where the total length cannot be measured without significant dissection/disruption of the surrounding tissues. Once fitted with a strain gauge, the forearm was loaded in cyclic axial compression using the same parameters as were used for in vivo loading (2 Hz, 17 N). The voltage signal was routed through a signal conditioning amplifier (model 2210; Measurements Group Inc.), and the peak-to-peak voltage was measured on a digital oscilloscope. The voltage measurements were later converted to strain using a calibration factor derived from measured (using strain gauges prepared identically to those used for collecting bone strains) and calculated (using beam theory) strains collected from an aluminum cantilever. The strain gauges were removed from the right ulnas by acetone dissolution of the cyanoacrylate, and the right and left ulnas were dissected free of the articulating bones, cleaned of soft tissues, and measured for total length to the nearest 0.1 mm with digital calipers. Subsequently, the bones were fixed in 10% neutral-buffered formalin for 48 h and then transferred to 70% ethanol for storage.
Peripheral quantitative computed tomography measurement of geometric properties
After tissue fixation, each ulna was fixed in a plastic tube filled with 70% ethanol and centered in the gantry of a Norland Medical Systems Stratec XCT Research SA + peripheral quantitative computed tomography (pQCT; Stratec Electronics, Pforzheim, Germany). The entire bone was scanned at 1-mm increments, beginning with the tip of the olecranon (proximal end) and ending with the distal ulnar epiphysis. The 33 slices per bone were scanned using a 0.46-mm slice thickness and a 0.08-mm voxel size. For each section, the X-ray source was rotated through 180° of projection (one block). The scans were imported into BonAlyse version 1.3 software (BonAlyse Ltd., Jyväskyla, Finland), in which the maximum and minimum second moments of area (IMAX and IMIN; mm4) were calculated for each slice. The second moment of area estimates a structure's resistance to bending by considering both cross-sectional area and the distribution of the material within the cross-section. Beams (or long bone shafts) with the material distributed farther from the neutral plane of bending will exhibit greater resistance to bending than beams with material distributed closer to the neutral plane of bending.
Qualitative changes in whole bone structure and morphology were evaluated from three-dimensional (3D) images of the whole bones reconstructed from individual micro-CT (μCT) slices. Each ulna was scanned on a desktop μCT (μCT 20; Scanco Medical AG, Bassersdorf, Switzerland). A microfocus X-ray tube with a focal spot of 10 μm was used as a source. For each slice, 600 projections were taken over 216° (180° plus half of the fan angle on either side). Approximately 1500 μCT slices were acquired per bone using a slice increment of 17 μm. A standard convolution-backprojection procedure with a Shepp and Logan filter was used to reconstruct the CT images in 1024 pixel × 1024 pixel matrices.
The ulnas were brought to room temperature slowly (∼2 h) in a saline bath. Subsequently, each bone was mounted between two opposing cup-shaped platens of a miniature materials testing machine (Vitrodyne V1000; Liveco, Inc., Burlington, VT, USA), which has a force resolution of 0.05 N (Fig. 1B). The bone was fixed in place using an ∼0.1 N static preload. The ulnas were loaded to failure in monotonic compression using a crosshead speed of 2 mm/s, during which force and displacement measurements were collected every 0.01 s. From the force versus displacement curves, ultimate force (FU; in N), stiffness (S; in N/mm), and energy to failure (U; in mJ) were calculated. After completion of the test, the two resulting broken pieces of each ulna were collected and photographed with a digital camera. The digital images were imported into Scion Image 4.0.2 for Windows (Scion Corp., Frederick, MD, USA), in which the position of the fracture was measured as a percent of the total bone length (expressed as percent of total length from the distal end).
Differences between right (loaded limb) and left (control limb) values were tested for significance using paired t-tests. Data from the right limb were compared with the left limb values by calculating percent differences between right (R) and left (L) limbs ([R − L]/L ∗ 100). One way analysis of variance (ANOVA) was used to test for significant differences among the four experimental groups. Significant ANOVAs were followed by Fisher's protected least significant difference (PLSD) post hoc tests to detect significant differences between individual experimental groups. Pearson's product moment correlation coefficient was used to assess the relation between mechanical and geometric properties. For the t-tests, correlations, ANOVAs, and post hoc tests, α = 0.05.
Three animals died during the 16-week experiment from anesthesia-related complications or unknown causes. Ulnar length in the left limb was not significantly different among the four groups (p = 0.51), but significant (p < 0.01) right versus left differences in length, of ∼0.5 mm, were detected for the 360 × 1 and 90 × 4 groups (Table 1).
Strain gauge measurements collected from the right ulna of the BLC group revealed that at the start of the experiment, 17 N of force produced ∼3300 με · of compression at the midshaft ulna (Table 1). Sixteen more weeks of normal cage activity (AMC group) failed to change strain significantly from that observed in the BLC group. However, 16 weeks of exogenous mechanical loading applied according to the 360 × 1 schedule resulted in a significant decrease in strain (∼40% reduction; p < 0.01) compared with the BLC and AMC groups. A similar, significant decrease in strain was measured in the rats loaded for 16 weeks according to the 90 × 4 schedule, but a significant difference was not found between strain measurements in the 360 × 1 and 90 × 4 groups.
Three-dimensional reconstructions of the whole ulnas from μCT slices showed that the right ulnas were substantially different from the left ulnas in the two loaded groups. The loaded bones exhibited substantial mediolateral thickening of the diaphysis near midshaft, and, more so, in the distal shaft, compared with the contralateral control bones (Fig. 2). The loaded bones also exhibited prominent osteophytic formation on and around the distal ulnar epiphysis (Fig. 2, arrow). The diaphyseal thickening and osteophytosis were evident in all of the right (loaded) ulnas from the loaded groups, regardless of loading schedule (90 × 4 or 360 × 1). Qualitatively, the degree of osteophytosis appeared slightly greater in the 90 × 4 ulnas than in the 360 × 1 ulnas, but no gross differences in mediolateral thickening were obvious between those two groups. No obvious differences in bone morphology were noted between right and left ulnas from either control group.
pQCT-derived IMAX and IMIN in the left (control) ulnas were significantly greater (13-31% and 19-23% greater, respectively; p < 0.05) in the three older groups compared with the younger BLC group, but only in the midshaft and distal regions (slices 13 to 26 for IMAX and slices 14 to 22 for IMIN). None of the three older groups was significantly different from another for either IMAX or IMIN at any of the slices, which suggests the presence of an aging effect on cross-sectional geometry, but no systemic effect of loading or anesthesia could be detected.
Percent difference (right vs. left) in IMAX was significantly (p < 0.05) different among groups for slices 16-25, 28, 29, and 31 (Fig. 3A). Post hoc comparisons revealed that these differences were mainly between the two loaded groups compared with the two control groups, with the loaded groups reaching up to a 23% increase in loaded versus control limb IMAX. However, in a subset of those slices, a scheduling effect was detected within the loaded groups. The 90 × 4 group exhibited significantly (p < 0.05) greater IMAX than the 360 × 1 group for slices 20-24.
Percent difference (right vs. left) in IMIN was significantly (p < 0.05) different among groups for slices 6 and 8-30 (Fig. 3B). Post hoc comparisons in all but five of those slices revealed significant (p < 0.05) differences among all groups (all pairwise comparisons significant) with the exception of the comparison between the two control groups, which was never significant in any of the slices. The 360 × 1 and 90 × 4 groups exhibited an ∼70% and ∼100% increase, respectively, in IMIN in the loaded arm compared with the control arm. Thus, a strong loading (360 × 1) effect on IMIN was detected, which was enhanced further by the load scheduling effect (90 × 4).
Mechanical testing of the left (control) ulnas resulted in significant (p < 0.05) differences between the BLC group and each of the three remaining (older) groups for both energy to failure and ultimate force, but stiffness was not significantly different among groups (Table 1). Compared with the three older groups, the BLC group exhibited significantly greater (∼55% greater; p < 0.001) energy to failure and significantly lower (∼20% lower; p < 0.001) ultimate force in the left ulnas. The greater energy to failure (despite lower ultimate force) can be explained by the large range of post-yield displacement exhibited by the younger, more ductile ulnas from the BLC group (Fig. 4). Thus, the control bones show the presence of an aging effect on energy to failure and ultimate force but no systemic effect of loading or anesthesia could be detected.
Percent difference (right vs. left) in ultimate force and energy to failure was significantly (p < 0.05) different among all groups (all pairwise comparisons significant), with the exception of the comparison between the two control groups, which was not significant for either energy to failure or ultimate force (Fig. 5). The loaded ulnas in the 360 × 1 group exhibited a 64% increase in ultimate force and a 94% increase in energy to failure as a result of exogenous loading once per day. Those gains in ultimate force and energy to failure were improved further by 35% and 75%, respectively, by partitioning the daily stimulus into discrete well-separated bouts (90 × 4). Percent difference in stiffness was not significantly different among groups (p = 0.12).
In vivo mechanical loading of the ulna altered the location of fracture resulting from the axial compression test (Table 1). The fracture location was not significantly (p = 0.10) different among groups for the left ulna, but the right ulna fracture location was significantly (p < 0.05) different among all groups (all pairwise comparisons significant), with the exception of the comparison between the two loaded groups. The BLC ulnas fractured at ∼40% of their total length, measured from the distal end (Table 1). Sixteen months later (AMC group), the fracture location shifted distally to ∼30%. Loading for 16 weeks according to both schedules reversed the age-associated distal migration of the fracture location to approximately midshaft (50%).
The objective of this study was to determine the effects of 16 weeks of mechanical loading, administered according to two different temporal schedules but with identical mechanical inputs, on bone structure and biomechanical properties in the rat ulna. We found that two of the three biomechanical properties measured (ultimate force and energy to failure) were improved considerably by loading. Ulnas loaded according to the 360 × 1 schedule absorbed 94% more energy before failing than the contralateral nonloaded control bones. Remarkably, loaded ulnas from the 90 × 4 rats were able to absorb 165% more energy before failing than the nonloaded control ulnas, which represents a 75% improvement in energy to failure over the 360 × 1 ulnas. Although less striking in magnitude, similar trends were observed for ultimate force; the right ulnas from the 360 × 1 group were able to support 64% more load before failing than the left ulnas, but the right ulnas from the 90 × 4 group supported 87% more load before failing than the left ulnas, representing a 35% improvement in ultimate force from the scheduling effect. These disparities in mechanical properties between the two loaded groups cannot be explained by differences in mechanical inputs, because the two loaded groups received the same total number of cycles per day, delivered at the same force magnitude and frequency. Rather, differences in mechanosensitivity of the resident bone cell populations are likely to have existed when the 360 daily cycles were applied. Presumably, most of the (later) cycles in each 360 × 1 bout were not processed because the cells had desensitized rapidly during the first 50 or so cycles. However, in the 90 × 4 animals, the first 50 or so cycles from each of the four shorter (90 cycles) daily bouts were processed, because the cells were allowed time to recover some of the lost mechanosensitivity during the 3-h rest periods between bouts.
Previously, we reported aBMD and bone mineral content (BMC) measurements from the same 90 × 4 and 360 × 1 ulnas reported here. The loaded ulnas exhibited 5.4% and 8.6% greater aBMD than the control ulnas in the 360 × 1 and 90 × 4 groups, respectively. BMC was increased by 6.9% and 11.7% in the loaded ulnas of the 360 × 1 and 90 × 4 groups, respectively. The present study shows that these small gains in aBMD and BMC imparted very large increases in ultimate force (64-87%) and energy to failure (64-165%) because the new bone formation was localized to the most biomechanically relevant sites. Thus, it might be possible to significantly enhance fracture resistance through mechanical loading (e.g., exercise), even if some of the noninvasive measurements of bone mass or density (e.g., DXA) reveal only slight changes.(13–15) For example, most exercise intervention studies yield differences in aBMD of only a few percent at most, but it is not known how such intervention affects fracture susceptibility.(16) Our data suggest that bone strength can be enhanced substantially from small changes in BMD or BMC if the bone is added to the mechanically appropriate (high strain) sites. This is perhaps an important difference between load-induced bone formation and pharmaceutically induced bone formation. Many of the osteogenic effects of anabolic agents are localized to biomechanically suboptimal bone surfaces. For example, intermittent parathyroid hormone (PTH) treatment adds bone primarily to the endocortical and trabecular surfaces, where it can contribute relatively little resistance to bending.(17,18) Sato et al.(19) showed that PTH (8 μg/kg per day for 6 months) increased midshaft femur BMC by 36% and ultimate force by 50% in the rat. Using the 360 × 1 schedule, we increased ulnar BMC by only 7% but improved ultimate force by 64%. In other words, comparison of the two experiments shows that mechanical loading can use less than one-fifth of the PTH-induced BMC to impart a 28% greater increase in strength than PTH, because the new bone is strategically (in mechanical terms) placed. Mechanical loading appears to optimize spatially the new bone formation for structural integrity.
Geometric properties varied substantially along the length of the ulnar diaphysis. IMAX was improved modestly by loading, reaching a peak increase over the control limb of 26% in the 90 × 4 group and 13% in the 360 × 1 group. However, IMIN was improved substantially by loading, reaching a peak increase over the control limb of 96% in the 90 × 4 group and 69% in the 360 × 1 group. The loading-induced percent increase in IMIN was roughly fourfold greater than that for IMAX. Reconstructed μCT images of whole ulnas confirm the geometric observations; substantial mediolateral thickening (IMIN plane) was clearly discernible in the loaded ulnas, particularly distally where IMIN peaked, but anteroposterior thickening was not observed on the μCT reconstructions, reflecting the modest increases found for IMAX. The large preferential increase in IMIN but not IMAX can be attributed to the preexisting geometry of the ulna and the strain distribution resulting from compressive axial loading. The IMAX (craniocaudal) plane at midshaft corresponds roughly to the neutral axis during applied loading, ∼90° from the plane of the ulna's curvature (and bending). Consequently, very little new bone was formed on the cranial and caudal periosteal surfaces because of loading, and IMAX increased only slightly. However, the medial and lateral surfaces of the rat ulnar diaphysis (1) endure the greatest strains during loading,(12,20) (2) are in the plane of the minimum second moment of area,(11) and (3) exhibit the greatest bone formation response to loading.(5,20–22) Consequently, the 4 months of in vivo loading in our experiment resulted in the addition of new bone primarily to the medial and lateral periosteal surfaces, which had the effect of substantially increasing IMIN and, ultimately, bone strength. The dominant role of IMIN in enhancing mechanical properties in the ulna is illustrated in Fig. 6, which shows that IMIN explains 92% of the variation in ultimate force.
The results from the BLC group indicate that at the start of the experiment (6 months of age), the ulnas were relatively tough and ductile, exhibiting a large degree of post-yield displacement (Fig. 4; Table 1). Four additional months of aging under normal mechanical usage (cage activity) resulted in a more brittle bone with reduced energy-absorbing capacity, as revealed by the measurements collected from the nonloaded control bones in the two loaded groups and from both right and left bones in the AMC group. The increase in brittleness with age is explained, at least in part, by a significant increase in volumetric BMD (vBMD) in the aged animals reported previously.(11) Loading for 4 months according to either schedule (90 × 4 or 360 × 1) reversed the loss of energy to fracture (area under the load-displacement curve) associated with aging by increasing ultimate force (Fig. 4), but the increase in brittleness (decrease in post-yield displacement) that normally occurs as a function of aging was not affected by loading.
The peak increase in diaphyseal IMAX and IMIN for both loaded groups occurred around slice 23-24, which was located approximately ¼ of the length from the distal end. This position corresponds to the site of maximal bone formation in the rat ulna observed by other investigators.(12,20) A substantial degree of variation was detected in IMAX and IMIN at the distal metaphysis/proximal epiphysis, beyond slice 29 or 30, indicated by the abrupt and large changes in percent difference for both loaded groups. The position along the length of the bone at which these anomalies was detected corresponds to the position of the osteophytic spicules detected from μCT reconstructions (Fig. 2). Therefore, the large but random differences at the metaphysis probably do not reflect a coordinated adaptive response like that observed in the diaphysis.
In situ mechanical strain measurements, collected from the medial surface of the midshaft ulna, showed a 40% reduction in peak mechanical strain resulting from 4 months of loading in vivo. It is interesting to note that the two different loading schedules, although significantly lower in strain than the two control groups, failed to exhibit significant differences from one another in mechanical strain. Likewise, the BLC and AMC groups failed to exhibit significant differences from one another in peak strain. These findings might be explained partly by the observation that stiffness was not significantly different among the four groups, although this would not account for the reduction in peak strain with loading.
A slight but significant difference in right versus left ulna length was detected in the two loaded groups. Longitudinal growth suppression is one of the inevitable side effects associated with axial loading of the ulna in growing rats,(20,22) although the degree of suppression depends on the force magnitude used.(5,20,23) Values for growth suppression in the growing male rat after only 2 weeks of loading at 17 N typically reach 1.2-1.3 mm (∼4% of total length).(5,20) We used adult animals (not rapidly growing) in this study to minimize the confounding effects of longitudinal growth suppression on the other endpoints measured, and detected only 0.5 mm of suppression (∼1.5% of total length) after 4 months of loading at 17 N.
The proximal shift in fracture location induced by loading is consistent with the site of maximum loading-induced bone formation along the diaphysis observed from μCT reconstructions and from the IMIN calculations. Initially (BLCs), the weakest point of the ulnar diaphysis was ∼3 mm distal to the midshaft. Sixteen weeks later (AMC group), the weakest point had migrated distally, to ∼6 mm distal to midshaft. In the ulnas loaded for 16 weeks according to either schedule, new bone formation was maximal 6 mm distal to midshaft, which had the effect of markedly strengthening the distal diaphysis and forcing the fracture site proximally to the midshaft.
The results of this study should be considered in light of several limitations of the experiment. To begin, the values for the mechanical properties listed in Table 1 could be greater than the values for mechanical testing of freshly dissected bone, because our bones were fixed in formaldehyde for 2 days.(24–26) However, Currey et al.(27) showed that formaldehyde fixation does not significantly affect mechanical parameters derived from quasi-static tests (only impact tests are significantly affected). Moreover, all of the ulnas from all groups were fixed for equal amounts of time (2 days) in the formaldehyde buffer; therefore, the comparisons among bones and groups within our experiment should not be affected. Second, we did not include sham-loading experiments. A potential shortcoming of the ulna loading model is the inability to conduct sham-loading control experiments, which can reveal the effects of soft tissue pressure/inflammation at the load contact points (flexed carpus and olecranon) on the osteogenic response in the diaphysis.(28) The geometric data illustrated in Fig. 3B address this issue: IMIN peaks in the middistal diaphysis and returns toward baseline values as the distal end is approached (excluding the large spikes at the distal tip resulting from the osteophytic formations). If inflammation at the distal load point-tissue interface was fueling the response seen in the diaphysis, the values observed in the middistal diaphysis should be maintained as the distal end is approached. However, the clear separation between the diaphyseal response and the distal metaphyseal/epiphyseal osteophytic response suggests that the diaphyseal results are not confounded by inflammation at the distal load point-tissue interface. Third, the force magnitude (17 N) used to load the rat forelimbs in our study induced peak strains that exceed those measured in humans during vigorous physical activity.(29) It is not known whether the bout-scheduling effects would yield similar benefits at substantially lower strains (e.g., those encountered during exercise).
In conclusion, we found that long-term mechanical loading protocols yield greater returns in biomechanical and structural properties if the daily regimen is divided into discrete bouts separated by recovery periods. Selectively stimulating bone cells when they are more receptive to mechanical signals has clear beneficial effects on bone mass, strength, and structure, even after an extended period of loading (16 weeks). In addition, exercise studies in which only small changes in aBMD or BMC are detected might be more beneficial to bone health and fracture resistance than has commonly been presumed. We found that very modest change in aBMD and BMC can translate into large changes in mechanical properties because mechanical loading tends to add bone to the most structurally relevant loci.
The authors thank Mary Hooser and Diana Jacob for assistance with tissue processing. This work was supported by National Institutes of Health (NIH) grants AR43730, T32, and AR07581.
- 12000 Does childhood and adolescence provide a unique opportunity for exercise to strengthen the skeleton? J Sci Med Sport 3:150–164., , , , , ,
- 21995 Mechanotransduction and the functional response of bone to mechanical strain. Calcif Tissue Int 57:344–358.,
- 31994 Mechanotransduction in bone: Do bone cells act as sensors of fluid flow? FASEB J 8:875–878., ,
- 41985 Regulation of bone mass by mechanical strain magnitude. Calcif Tissue Int 37:411–417.,
- 52001 Modulation of appositional and longitudinal bone growth in the rat ulna by applied static and dynamic force. Bone 29:105–113., , , ,
- 61999 Toward a mathematical description of bone biology: The principle of cellular accommodation. Calcif Tissue Int 65:466–471.
- 71984 Regulation of bone formation by applied dynamic loads. J Bone Joint Surg Am 66:397–402.,
- 81997 Five jumps per day increase bone mass and breaking force in rats. J Bone Miner Res 12:1480–1485., , , ,
- 91998 Three rules for bone adaptation to mechanical stimuli. Bone 23:399–407.
- 102000 Partitioning a daily mechanical stimulus into discrete loading bouts improves the osteogenic response to loading. J Bone Miner Res 15:1596–602., ,
- 112002 Shorter, more frequent mechanical loading sessions enhance bone mass Med Sci Sports Exer 34:196–202., , ,
- 121994 Noninvasive loading of the rat ulna in vivo induces a strain-related modeling response uncomplicated by trauma or periosteal pressure. Calcif Tissue Int 54:241–247., , ,
- 131998 Randomized controlled study of effects of sudden impact loading on rat femur. J Bone Miner Res 13:1475–1482., , , , ,
- 141999 Have the DXA-based exercise studies seriously underestimated the effects of mechanical loading on bone? J Bone Miner Res 14:1634–1635., ,
- 152001 Mechanical effects on the skeleton: Are there clinical implications? Osteoporos Int 12:77–83.
- 162000 Systematic review of randomized trials of the effect of exercise on bone mass in pre- and postmenopausal women. Calcif Tissue Int 67:10–18.,
- 172001 Repeated in vivo determinations of bone mineral density during parathyroid hormone treatment in ovariectomized mice. J Endocrinol 170:529–537., , , ,
- 182001 Intermittently administered human parathyroid hormone(1–34) treatment increases intracortical bone turnover and porosity without reducing bone strength in the humerus of ovariectomized cynomolgus monkeys. J Bone Miner Res 16:157–165., , , , ,
- 191997 Biosynthetic human parathyroid hormone (1–34) effects on bone quality in aged ovariectomized rats. Endocrinology 138:4330–4337., ,
- 201997 Strain magnitude related changes in whole bone architecture in growing rats. Bone 20:191–198., , ,
- 212001 Mechanical loading of diaphyseal bone in vivo: The strain threshold for an osteogenic response varies with location. J Bone Miner Res 16:2291–2297., , , ,
- 221998 Strain rate as a controlling influence on adaptive modeling in response to dynamic loading of the ulna in growing male rats. Bone 23:313–318.,
- 232002 The effects of dynamic axial loading on the rat growth plate. J Bone Miner Res 17:284–292., , ,
- 241973 Mechanical Properties of Bone. Charles C. Thomas, Springfield, IL, USA, pp. 1–322.
- 251965 A rheologic model for cortical bone. A study of the physical properties of human femoral samples. Acta Orthop Scand Suppl 83:1–77.
- 261964 Effect of embalming on the mechanical properties of beef bone. J Appl Physiol 19:1234–1236., , ,
- 271995 Effect of formaldehyde fixation on some mechanical properties of bovine bone. Biomaterials 16:1267–1271., , ,
- 282001 Skeletal loading in animals. J Musculoskel Neuron Interact 1:249–262., ,
- 291996 In vivo measurement of human tibial strains during vigorous activity. Bone 18:405–410., , , , , , , ,