pmc logo image
Logo of nihpaNIHPA bannerabout author manuscriptssubmit a manuscript
Journal List > NIHPA Author Manuscripts

Formats:
Zoology (Jena). Author manuscript; available in PMC 2009 January 1.
Published in final edited form as:
Zoology (Jena). 2008; 111(2): 135–147.
Published online 2008 January 24. doi: 10.1016/j.zool.2007.06.003.
PMCID: PMC2405767
NIHMSID: NIHMS42948
Copyright notice and Disclaimer
The relative importance of genetics and phenotypic plasticity in dictating bone morphology and mechanics in aged mice: evidence from an artificial selection experiment
Kevin M. Middleton,ad* Corinne E. Shubin,a Douglas C. Moore,b Patrick A. Carter,c Theodore Garland, Jr.,d and Sharon M. Swartza
aDepartment of Ecology and Evolutionary Biology, Brown University, Providence, RI 02912, USA
bBioengineering Laboratory, Rhode Island Hospital, Providence, RI 02903, USA
cSchool of Biological Sciences, Washington State University, Pullman, WA 99164, USA
dDepartment of Biology, University of California, Riverside, Riverside, CA 92521, USA
*Corresponding author. Present address: Department of Biology, California State University San Bernardino, 5500 University Parkway, San Bernardino, CA 92407, USA E-mail address: Email: kmm/at/csusb.edu (K.M. Middleton)
Small right arrow pointing to: The publisher's final edited version of this article is available at Zoology (Jena).
Publisher's Disclaimer
Abstract
Both genetic and environmental factors are known to influence the structure of bone, contributing to its mechanical behavior during, and adaptive response to, loading. We introduce a novel approach to simultaneously address the genetically mediated, exercise-related effects on bone morphometrics and strength, using mice that had been selectively bred for high levels of voluntary wheel running (16 generations). Female mice from high-running and control lines were either allowed (n = 12, 12; respectively) or denied (n = 11, 12; respectively) access to wheels for 20 months. Femoral shaft, neck, and head were measured with calipers and via micro-computed tomography. Fracture characteristics of the femoral head were assessed in cantilever bending. After adjusting for variation in body mass by two-way analysis of covariance, distal width of the femur increased as a result of selective breeding, and mediolateral femoral diameter was reduced by wheel access. Cross-sectional area of the femoral mid-shaft showed a significant linetype × activity effect, increasing with wheel access in high-running lines but decreasing in control lines. Body mass was significantly positively correlated with many of the morphometric traits studied. Fracture load of the femoral neck was strongly positively predicted by morphometric traits of the femoral neck (r2 > 0.30), but no significant effects of selective breeding or wheel access were found. The significant correlations of body mass with femoral morphometric traits underscore the importance of controlling for body size when analyzing the response of bone size and shape to experimental treatments. After controlling for body mass, measures of the femoral neck remain significant predictors of femoral neck strength.
Keywords: Artificial selection, Body mass, Bone morphometrics, Exercise, Mechanical loading
Genetics and exercise are foremost among many factors that act individually and in concert to determine the geometry and material properties of a bone. The shape and constituent materials of the bone, in turn, determine how it will respond to a given load. In mammals, genetic factors are thought to determine 50–80 % of the variation in mineral content of adult bone, with most of the balance dictated by amount, frequency, and/or intensity of exercise (Eisman, 1999; Ferrari et al., 1999). However, the interaction between genetic and environmental factors is complex, and the effects of loading may be modulated by the physiological response of bone, which itself is genetically determined to an important extent (Robling and Turner, 2002; Koller et al., 2003; Robling et al., 2003).
Although genetic factors are known to mediate exercise effects, the two are often addressed separately in experimental settings. For example, bone mineral content and activity effects have frequently been studied in rodents, without attention to possible interactions (Gordon et al., 1989; Umemura et al., 1995; Iwamoto et al., 1999; Klein et al., 2001; Turner and Burr, 2001; Bennell et al., 2002; Akhter et al., 2004a, b; Niehoff et al., 2004). Recently, however, more attention has been directed at variation among different strains of mice.
Several aspects of femoral geometry and strength have been shown to vary widely among different genetic strains of mice (Koller et al., 2003; Wergedal et al., 2005), as have responses to hind limb unloading (Amblard et al., 2003; Judex et al., 2004; Squire et al., 2004) and ovariectomy (Bouxsein et al., 2005). Expanded understanding of the role of mechanotransduction in bone has resulted from comparisons of different genetic strains of mice under artificial loading regimens (e.g., Robling and Turner, 2002; Robling et al., 2003; Kesavan et al., 2005; Li et al., 2005; Kesavan et al., 2006; Lau et al., 2006; Sawakami et al., 2006).
Although jump training (i.e., impact loading) has been used to study the effects of naturalistic loading on bone (Honda et al., 2001; Umemura et al., 2002; Welch et al., 2004), relatively rarely have the combined effects of genetic background and exercise been addressed under natural, musculoskeletal loading. One exception is a study by Kodama et al. (2000), who compared the response of bone to four weeks of jump training in C57BL/6J and C3H/HeJ mice. These authors found that both strains of mice responded to loading but that the relative response differed between strains and concluded that bone of the C3H/HeJ mice were less responsive to mechanical loading.
In addition to the complex interaction of genetics and loading history in forming bone, these two factors play an important role in the changes that bone exhibits as an animal ages. The effects of aging on bone properties have been studied extensively in rodents (Yamamoto et al., 1995; Mosley and Lanyon, 1998; Halloran et al., 2002; Mosley and Lanyon, 2002; Hamrick et al., 2006; McNamara et al., 2006) and in humans (Smith and Walker, 1964; Mosekilde, 1989; Russo et al., 2006). Similarly, exercise effects on age-related changes to bone have received a great deal of attention, both in humans (Weinstein and Hutson, 1987; Heaney et al., 1997; Kaptoge et al., 2006) and in rodent models (Raab et al., 1990; Søgaard et al., 1994; Silva and Gibson, 1997). One way to effectively probe relations between genetics and external influences, such as exercise, is to take advantage of artificial selection experiments. Recently, a few research programs have overcome the technical challenges of applying this labor- and time-intensive approach to mammals and have successfully used rodents in selection experiments targeting exercise-related phenotypes (Hussain et al., 2001; Koch and Britton, 2001; Henderson et al., 2002; Garland, 2003; Wisløff et al., 2005). Although diverse mammalian species vary considerably in bone structure and physiology (Bagi et al., 1997), and extrapolations from rodents to humans and other species must be made only with caution, it is now possible to add artificial selection to the range of techniques used to address questions in bone biology. Furthermore, selection experiments provide a valuable tool for assessing the relative magnitudes of evolutionary and phenotypic plasticity in skeletal biology (Garland and Kelly, 2006).
Here, we employ mice selectively bred for high levels of voluntary wheel running to explore the combined effects of genetic background and voluntary exercise on bone morphometrics and mechanics. We focus on whole-bone morphometrics and mechanical performance during loading, and do not consider variations in tissue-level properties such as bone mineral density and collagen fiber orientation that may affect mechanical behavior under load. For our study population we chose mice that were allowed or denied access to a running wheel for twenty months to explore whether presumptive changes in bone morphometrics and mechanical properties resulting from exercise would be retained as the animals aged.
We ask: 1) Does the size, shape, and mechanical performance of the femur differ between animals from lines that have been artificially selected for high voluntary exercise in comparison with unselected control lines (genetic effect)? 2) Do the same characteristics differ between individuals allowed free access to running wheels for twenty months in comparison with mice denied wheels (exercise effect)? 3) Are there interactions between these two factors such that the effect of exercise depends on genetic background?
Model system
We employed mice (Mus domesticus) in this study because they were part of a long-term artificial selection experiment on high levels of exercise (voluntary wheel-running behavior; for additional information, see Garland, 2003; Rhodes et al., 2005). Among the goals of this selection study are understanding heritable aspects of behavior and physiology in mammals, including neurobiological and physiological processes relevant to human health. The complete experimental design is described elsewhere (Swallow et al., 1998; Garland, 2003; Morgan et al., 2003), and we provide only a brief summary here.
From a base population of outbred Hsd:ICR mice, 8 closed lines were established in which the parents of each subsequent generation are those which exhibit the highest (4 selected lines) or typical (4 control lines) levels of voluntary wheel running. Mice from the ICR strain were chosen because they exhibit high levels of genetic variation and had previously been selected for large litter sizes and high weaning success (Swallow et al., 1998). Because genetic heterogeneity is a prerequisite for evolutionary change in response to selective breeding, the use of an inbred strain would have been impossible. Furthermore, we employed within-family selection to decrease the possible deleterious effects of inbreeding and potentially confounding maternal effects. Mean adjusted heritability of wheel running was 0.28 for the first 10 generations of selection (Swallow et al., 1998).
Running was scored as total number of exercise wheel revolutions run on day five plus day six of a six-day exposure to wheels. Total revolutions per day in the lines selected for high running increased rapidly for the first 16 generations and then reached an apparent plateau, at which point the selected lines ran about 170–200 % more revolutions per day than controls (Swallow et al., 1998; Garland, 2003). Summed revolutions on days five and six is the sole criterion for selection, but selection for high levels of voluntary exercise has the potential to lead to evolution in many diverse features because of pleiotropic effects of alleles that affect wheel running. The selected lines have diverged genetically and phenotypically from the control lines and are distinct morphologically, physiologically, and behaviorally (Swallow et al., 1999; Garland, 2003; Garland and Kelly, 2006; Kelly et al., 2006; Malisch et al., 2007).
Study population
Our sample consisted of femora of 47 female mice from generation 16 of the selection experiment. Mice were weaned at 21 days of age, housed individually, and either allowed or denied access to an exercise wheel (0.73 m circumference) for 80 weeks, beginning at 28±3 days of age (for additional details see Morgan et al., 2003; Bronikowski et al., 2006). Mean age at sacrifice was 596 days (range = 590–600). Approximately half of the mice were from the selected lines (n = 23) and half from the control lines (n = 24). Within each of these groups, half were raised with access to a wheel (n = 12 and 12, from selected and control lines, respectively) to allow for voluntary wheel running and half were sedentary (no wheel; n = 11 and 12, from selected and control lines, respectively). All mouse cages were standard size, and mice were permitted food and water ad libitum. All protocols were approved by and in compliance with guidelines set by the Washington State University IACUC.
Specimen preparation
After sacrifice, the mice were weighed and frozen until dissection. The mice were later thawed and both hind limbs dissected; the femora were separated from the distal limb elements and defleshed. Right femora were used for morphometric measurements and mechanical testing, and left femora were scanned using microcomputed tomography (µCT). The femur was chosen for detailed study because it is subject to the most direct loading from the axial skeleton, and the proximal femur is the site of most femoral fractures (Crawford and Fretz, 1985; Embertson et al., 1986; Zhang et al., 2000; Harasen, 2003).
Whole-bone morphometrics
Measuring with digital calipers to the nearest 0.01 mm, eight morphometric traits of the femur were quantified: length from the superior (articular) surface of the femoral head to the distal femoral condyles, proximal width (greater trochanter to medial surface of femoral head), distal width (mediolateral distance across distal femoral condyles), anteroposterior and mediolateral mid-diaphyseal diameter, proximodistal height and anteroposterior depth of the femoral neck, and height (proximodistal) and depth (anteroposterior) of the femoral head (Fig. 1Fig. 1). Many of these measurements (e.g., femur length, mid-shaft diameters, distal width, and head depth) overlap with morphometric data reported for exercise-selected mice from different generations (Garland and Freeman, 2005; Kelly et al., 2006) allowing comparisons among data sets. Note that our use of femoral head here equates to "femoral condyle" in prior studies (Garland and Freeman, 2005; Kelly et al., 2006). The angle of the femoral neck was measured from digital photographs as the angle formed between the femoral shaft and a line drawn parallel to the long axis of the femoral neck.
Fig. 1
Fig. 1
Fig. 1
Mouse proximal femur in anterior view. With the distal femur fixed, load is applied to the femoral head to cause cantilever bending at the femoral neck. The arrow indicates the point of load application. The most common fracture orientation is marked (more ...)
Microcomputed tomography
Left femora were scanned using a high-resolution, fan-beam microcomputed tomography scanner (µCT 20; Scanco Medical AG, Bassersdorf, Switzerland). For the mid-diaphysis, ten slices with a voxel size of 9 µm3 were acquired midway between the superior surface of the femoral head and the distal end of the femur. For the femoral neck, sixteen slices with a voxel size of 17 µm3 were acquired midway along the femoral neck. Mid-diaphyseal slices were acquired perpendicular to the long axis of the bone, and neck slices were acquired perpendicular to the long axis of the femoral neck. Cross-sectional areas, maximum and minimum second moments of inertia (Imax, Imin), and maximum section moduli (Imax/Cmin) were calculated for each slice using the scanner’s built-in software routines and averaged over all slices for the femoral diaphysis and neck for each individual.
Mechanical testing
To compare maximum load at failure among groups, we loaded the femoral neck to fracture in cantilever bending using an Instron 4222 materials testing machine (Instron Corporation; Norwood, MA). To approximate loading experienced by the femur during locomotion, a compressive load was applied to the femoral head with the distal end of the femur fixed in solid bismuth alloy (LMA-117; Small Parts, Inc., Miami Lakes, FL). This method of fixing the distal end of the bone facilitated later release of the entire bone specimen from the mounting medium by melting the potting material in a hot water bath (Fig. 1Fig. 1). Our methodology is similar to that used to test femoral neck strength in mice (Akhter et al., 2004b) and rats (Bagi et al., 1996). Tests were conducted under displacement control, with a constant loading velocity of 0.5 mm/min, and data were collected at a rate of 20 samples/sec. Load and displacement were recorded by computer and load at fracture was calculated.
Statistical analysis
Traits were analyzed by two-way analysis of variance (ANOVA) and analysis of covariance (ANCOVA) using SAS Procedure Mixed (version 9.1; SAS Institute; Cary, NC) with Type 3 tests of fixed effects with or without body mass as a covariate. Using ANOVA and ANCOVA, differences between control and selected lines (effects of line type), wheel and no wheel treatments (effects of activity), and the interaction between these two main effects were tested for statistical significance (Swallow and Garland, 2005; Kelly et al., 2006). In the analyses of predictors of fracture load, we included both body mass and femoral morphometric traits as covariates. Pearson correlations were calculated between body mass and the morphometric traits to check for potential multicollinearity. Correlations were always < 0.7, indicating that multicollinearity was not a concern in these data (Slinker and Glantz, 1985).
To address concerns related to performing multiple statistical tests on the same set of data, we carried out a false discovery rate (FDR) analysis using the qvalue software package (Storey, 2002) for R (version 2.4.1; R Development Core Team, 2007). The total number of hypotheses under test is 89: 60 in Table 1 (see Results), 19 in Table 2 (body mass covariate column only, as all others are redundant with Table 1), and 10 in Table 4 (only p values for the covariate correlation are included). Due to the relatively low number of p values, we used the "bootstrap" option of qvalue in estimating the proportion of true null hypotheses. Results of the FDR analysis indicate that, rather than the typical a = 0.05 for judging statistical significance, a more conservative level of a = 0.0374 (corresponding to a positive false discovery rate of 5 %; Storey, 2002) is appropriate given the number of hypotheses tested and the distribution of p values we obtained. In the tables, we present nominal p values for two-tailed tests, unless specifically noted for traits for which we had a priori predictions about the direction of main effect or covariate (e.g., smaller body mass in wheel-access and in selected-line mice).
Table 1
Table 1
Results (p values) of two-way analysis of variance models. p values < 0.05 are noted in bold. In the columns Line type and Activity, + indicates selected or wheel access is larger than control or sedentary, respectively; − indicates smaller. (more ...)
Table 2
Table 2
Results (p values) of two-way analysis of covariance models including body mass as a covariate. In the columns Line type and Activity, + indicates selected or wheel access is larger than control or sedentary, respectively; − indicates smaller. (more ...)
Table 4
Table 4
Results (p values) of two-way analysis of covariance models to determine best predictors of fracture load of the femoral neck including body mass as a covariate. In the columns Line type and Activity, + indicates selected or wheel access is larger than (more ...)
Morphometrics
Results of the two-way ANOVAs with activity level and line type (but not body mass as a covariate) show significant effects of activity on several traits (Table 1). Results of the two-way ANCOVAs show that activity level has a significant negative effect on body mass (p = 0.0374, one-tailed) and line type has a nearly significant effect (p = 0.0698, one-tailed), with both active and selected mice having lower body mass than sedentary or control mice, respectively (Table 2, Table 3). In ANCOVAs including body mass as a covariate, most femoral morphometric traits showed a significant positive correlation with body mass even after controlling statistically for main effects (Table 2). Mice bred for high levels of wheel running had wider distal femora than controls, regardless of wheel access (p = 0.0349), and wheel access led to mediolaterally narrower femoral shafts in both selected and control lines (p = 0.0166; Table 2, Table 3). Only one interactio was statistically significant: wheel access decreased midshaft cross-sectional area in control mice but increased it in selected mice (Table 2, Table 3).
Table 3
Table 3
Least squares means and standard errors for the four experimental groups (twoway analyses of covariance for which p values are presented in Table 2).
Mechanical properties
When the distal femur was fixed and loads were applied to the femoral head, the femur always fractured near the base of the neck, not distally along the shaft (Fig. 1Fig. 1). Table 4 summarizes results of analyses to determine the best morphometric predictors (i.e., correlates) of femoral neck fracture load, while controlling for all main effects and correlations with body mass. There was a significant positive correlation between load at fracture and both height and depth of the femoral neck, as well as between load at fracture and all femoral neck traits measured via µCT (r2 > 0.3; p < 0.01; Fig. 2Fig. 2; Table 4). In these analyses, the effects of line type and of activity were never statistically significant. None of the femoral shaft traits were significant predictors of fracture load (results not shown).
Fig. 2
Fig. 2
Fig. 2
Relationship between femoral neck fracture load, body mass, and morphometric traits of the femoral neck (height, Imax, cross-sectional (XS) area). A significant positive relationship was found with each trait. Points are coded according to ANCOVA grouping (more ...)
In this study, we explored the combined effects of genetic background and voluntary exercise on bone morphology and femoral neck strength using mice from lines that had been selectively bred for high levels of voluntary wheel running. We sought to determine whether the femora of mice selected for high levels of exercise would differ from controls (genetic effect), whether those with access to a running wheel would differ from those without (exercise effect), and whether the exercise effect, if any, was mediated by genetic background (interaction effect).
Wheel running and body mass
The mice analyzed in this study represent a subset of the female mice for which wheel running ontogeny was previously escribed (Morgan et al., 2003). Wheel running increased rapidly during the first 8 weeks of ontogeny, peaking at means of 76 km/week in mice selected for high levels of wheel running and 52 km/week in control mice (Fig. 3Fig. 3). From 8 weeks of age until the end of the experiment, wheel running declined roughly linearly, with means of 36 and 21 km/week at 84 weeks (selected and control lines, respectively). Both the position and the shape of the wheel running ontogeny were different between selected and control mice; selected mice ran significantly more across ontogeny and showed a steeper decline than control mice (Fig. 3Fig. 3; Morgan et al., 2003). In all mice, body mass was characterized by rapid early growth during weeks 4 through 20, followed by slower growth to week 84. Mice from selected lines were significantly smaller than controls, and those with wheel access were significantly smaller than those without (Morgan et al., 2003).
Fig. 3
Fig. 3
Fig. 3
Ontogeny of wheel running in lines of mice selectively bred for high levels of voluntary wheel running (closed symbols) and in unselected control lines (open symbols). Mean wheel running (km/week) is shown for each week of the 84 week duration. After (more ...)
Combined effects of genetic background and of exercise on skeletal traits
Of the 19 skeletal traits examined, only two (distal width of the femur, femoral head depth) showed significant effects of line type, one showed effects of exercise (mediolateral femoral diameter), and one (femoral shaft cross-sectional area) showed a line type-by-exercise interaction (Table 2, Table 4). These results suggest that neither selection over 16 generations for high levels of voluntary wheel running nor constant physical activity is associated with broad-scale modifications of the mouse femur at the whole-bone level, despite peak activity levels of 76 km/week in exercise-selected mice and 52 km/week in control mice (Fig. 3Fig. 3). However, these results differ substantially from those of previous training studies on mice from different generations of the same selection experiment (Garland and Freeman, 2005; Kelly et al., 2006) as well as twin and family studies in humans (Seeman et al., 1994; Cummings et al., 1995; Arden et al., 1996), which have demonstrated significant genetic underpinnings to the structure and function of bone.
In our study of musculoskeletal loading described here, the lack of significant differences between selected and control lines or between active and sedentary mice is potentially a consequence of the subjects’ age. At 84 weeks of age, these animals were almost certainly post-reproductive (Gosden et al., 1983) and had shown a gradual decrease in the level of exercise associated with aging in this study population (Fig. 3Fig. 3). Given the likelihood of post-reproductive status, they may have experienced a decrease in bone strength similar to that observed in ovariectomized mice (Peng et al., 1994; Mosekilde et al., 1998) and in post-menopausal women (e.g., Riggs et al., 1986; Garnero et al., 1996; Ahlborg et al., 2001, 2003).
Additionally, the beneficial effects to bone of repetitive loading are lost after loading ceases (i.e., detraining; Snow et al., 2001; Järvinen et al., 2003). As our study population aged, the number of wheel revolutions per day in the selected mice decreased to levels that may have been below the threshold for exercise-related benefits. Differences in bone morphometrics or strength that may have been present earlier in ontogeny could have been lost.
Musculoskeletal vs. artificial loading
Most studies of bone loading in rodents involve artificial loads exerted by mechanical bending apparatuses, either in axial or four-point bending (Robling and Turner, 2002; Robling et al., 2003; Kesavan et al., 2005; Li et al., 2005; Lau et al., 2006; Sawakami et al., 2006). While this technique does allow for controlled experimental conditions, loading patterns, and very high strains, the loads typically applied (often 10 N or more) are not biologically realistic. Jump-induced impact loading does provide a viable alternative to artificial loading (e.g., Honda et al., 2001; Umemura et al., 2002; Welch et al., 2004), but wheel running is perhaps the most realistic activity behaviorally. An additional drawback of three- or four-point bending is that, under natural conditions, bones are rarely loaded in pure bending but rather in a combination of axial compression and bending (Lovejoy et al., 1976; Biewener et al., 1983; Alexander, 2003; Lieberman et al., 2003, 2004).
The smaller body size of mice relative to other mammals may also be a factor in the apparent absence of large-scale effects of locomotion on bone morphometrics. Whereas a large body of literature exists regarding skeletal change due to loading in large mammals and birds (e.g., Rubin and Lanyon, 1984; Burr et al., 1998; Rubin et al., 2001; Lieberman et al., 2003, 2004), relatively fewer studies address skeletal mechanics of small mammals (exceptions include Biewener, 1983; Fischer 1994; Fischer et al., 2002; Witte et al., 2002). These studies have shown that small mammals generally employ more crouched limb postures than large mammals, which have relatively more erect limb postures that limit moment arms around joints. More crouched postures in small mammals may cause higher joint moments which may be resisted by relatively more massive limb bones, muscular forces, or a combination of both. Thus, in small mammals, such as mice, bone strains resulting from muscular forces may be relatively more important than in large mammals.
Morphometrics – importance of body mass
Previous studies of these mice in similar experimental designs have shown that body mass is lower in exercise-selected lines and is reduced in mice that are given constant wheel access (both selected and control lines; Swallow et al., 1999; Morgan et al., 2003; Swallow and Garland, 2005; Kelly et al., 2006). Our results here are consistent, although the line type effect did not reach statistical significance (Table 2, Table 3). This discrepancy may be attributable to the greater age of the mice in this study, and/or the relatively small sample sizes. In the previous study (Morgan et al., 2003), with a larger sample of mice, weighed at multiple points across ontogeny, the line type effect is statistically significant.
Body mass was the major determinant of variation in femoral size, shape, and strength in all of the mice in this study. All traits were positively correlated with body mass, even after accounting for effects of selective breeding and exercise, and 12 of 19 correlations were statistically significant (p < 0.05, not adjusted for multiple comparisons; Table 2). Of those traits that were not significantly correlated with body mass, two showed significant line type or activity effects (Table 2).
Although body mass effects were included in the mixed-model ANCOVA analysis (Table 2 and Table 4), morphometric traits that are related to load-bearing, such as cross-sectional diameters or areas, are often also normalized to femoral length (Ruff et al., 1993; Trinkaus, 1997). We performed an additional ANCOVA with both body mass and femur length as covariates (results not shown). Because of the high correlation between femur length and body mass, including both yielded little additional information at the expense of a more complicated statistical model, because femur length was rarely a significant covariate.
The widespread correlation of body mass with femoral morphometrics underscores the importance of accounting for body size in studies of the effects of exercise or other environmental factors on the skeleton. Without adequate control for the effects of body size, skeletal traits may appear to vary significantly among groups that experience different exercise regimens when the groups differ primarily in body size alone. Some studies on the effect of exercise or loading on bone do include body mass in statistical analyses (e.g., Gordon et al., 1989; Kannus et al., 1995; Järvinen et al., 2003; Binkley and Specker, 2004), but many do not, even when body mass differs significantly between control and treatment groups (e.g., Shaw et al., 1987; Niehoff et al., 2004; Wu et al., 2004).
Fracture strength and naturalistic loading
Three- or four-point bending tests are frequently employed to assess the mechanical strength of bones and can uncover differences in strength related to genetics, physiology, or experimental treatments (Kodama et al., 2000; Turner and Burr, 2001; Bennell et al., 2002; Akhter et al., 2004a; Silva et al., 2004). Although this methodology has the advantage of lending itself to calculations of the bone’s Young’s modulus, loading of the shafts of long bones in vivo is typically a combination of axial loading and a relatively small amount of bending, transmitted through the joints (Lovejoy et al., 1976; Biewener et al., 1983; Alexander, 2003; Biewener, 2003; Lieberman et al., 2003, 2004).
By loading the head in compression with the distal femur fixed, we mimicked natural loading, and thereby tested the strength of the femoral neck in cantilever bending (Fig. 1Fig. 1). Similar techniques have been used in mice (Akhter et al., 2004b), rabbits (Ohnishi et al., 2003), and humans (Augat et al., 1996). Because of its lower cross-sectional area and moments of inertia, coupled to large bending moments, the femoral neck could be expected to be the most likely site of fracture. This prediction is supported by studies in humans and other mammals (Crawford and Fretz, 1985; Embertson et al., 1986; Zhang et al., 2000; Harasen, 2003) and our experimental results: all femora fractured proximally at the neck, rather than in the shaft.
Some caution is warranted in the functional interpretation of differences in moments of inertia and section moduli. Loading in the femoral shafts and necks consists of a combination of axial and bending forces which shift the neutral axis away from the centroid of the bone (Lovejoy et al., 1976; Biewener et al., 1983; Lieberman et al., 2003, 2004). Without experimental measurement of the neural axis during loading (e.g., Judex et al., 1997; Demes et al., 1998, 2001; Lieberman et al., 2004) it is impossible to know the exact location of the neutral axis of the mouse femur at peak strain.
Although some aspects of gross bone geometry that can influence fracture strength, such as mineralization and collagen fiber orientation, were not quantified in this study, we were able to assess the net effect of variations in bone geometry on load bearing in a manner that is comparable to that experienced by intact organisms. Traits related to the structural geometry of the femoral neck are significantly correlated with fracture load even after accounting for the effect of body mass (Table 4). These traits include cross-sectional area of the femoral neck, maximum and minimum second moments of inertia, and maximum section modulus. In humans, femoral neck angle is correlated with fracture strength (Gnudi et al., 1999; Ciarelli et al., 2000; Gnudi et al., 2002; Center et al., 2004); however, in the mice studied here fracture load was not correlated with neck angle. We suggest that these differences in part reflect the differences in the anatomy of human and rodent femora; the human femoral neck is far longer than that of other mammals and is more highly angled with respect to the shaft (Bagi et al., 1997).
Conclusions
We introduce a unique model system to investigate the combined effects of exercise and genetics on femoral structure and performance of mice that have been selectively bred for high levels of voluntary wheel running. We found that body mass was a pervasive correlate of femoral structure, underscoring the importance of statistically controlling for body mass in studies of bone morphometrics. However, we did not find wide-spread effects of long-term wheel access. This general result differs from many previous studies and may result from differences in loading patterns (natural vs. artificial; postural differences) and allometric effects of body mass in small mammals vs. large ones.
Acknowledgements
The authors thank Scanco USA, Inc. for loan of the µCT system. We also thank E. Mathiowitz for allowing us access to equipment in her lab and E. Edwards for technical assistance. This manuscript benefited from the helpful suggestions of two anonymous reviewers. The project described was supported by NIH grant number 1F32AR053008-01 from the National Institute of Arthritis and Musculoskeletal and Skin Diseases to K.M. Middleton, National Science Foundation grant numbers DEB-0083638, DEB-0105079, and EF-0328594 to P.A. Carter, and IOB-0543429 to T. Garland, Jr. Additional support was provided by the Rhode Island Hospital Orthopaedic Foundation and University Orthopedics, Inc.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
  • Ahlborg HG, Johnell O, Nilsson BE, Jeppsson S, Rannevik G, Karlsson MK. Bone loss in relation to menopause: a prospective study during 16 years. Bone. 2001;28:327–331. [PubMed]
  • Ahlborg HG, Johnell O, Turner CH, Rannevik G, Karlsson MK. Bone loss and bone size after menopause. New Engl. J. Med. 2003;349:327–334. [PubMed]
  • Akhter MP, Fan Z, Rho JY. Bone intrinsic material properties in three inbred mouse strains. Cal. Tiss. Int. 2004a;75:416–420.
  • Akhter MP, Wells DJ, Short SJ, Cullen DM, Johnson ML, Haynatzki GR, Babij P, Allen KM, Yaworsky PJ, Bex F, Recker RR. Bone biomechanical properties in LRP5 mutant mice. Bone. 2004b;35:162–169. [PubMed]
  • Alexander RM. Principles of Animal Locomotion. Princeton: Princeton University Press; 2003.
  • Amblard D, Lafage-Proust M-H, Laib A, Thomas T, Rüegsegger P, Alexandre C, Vico L. Tail suspension induces bone loss in skeletally mature mice in the C57BL/6J strain but not in the C3H/HeJ strain. J. Bone Min. Res. 2003;18:561–569.
  • Arden NK, Baker J, Hogg C, Baan K, Spector TD. The heritability of bone mineral density, ultrasound of the calcaneus and hip axis length: a study of postmenopausal twins. J. Bone Min. Res. 1996;11:530–534.
  • Augat P, Reeb H, Claes LE. Prediction of fracture load at different skeletal sites by geometric properties of the cortical shell. J. Bone Min. Res. 1996;11:1356–1363.
  • Bagi CM, DeLeon E, Ammann P, Rizzoli R, Miller SC. Histo-anatomy of the proximal femur in rats: impact of ovariectomy on bone mass, structure, and stiffness. Anat. Rec. 1996;245:633–644. [PubMed]
  • Bagi CM, Wilkie D, Georgelos K, Williams D, Bertolini D. Morphological and structural characteristics of the proximal femur in human and rat. Bone. 1997;21:261–267. [PubMed]
  • Bennell KL, Khan KM, Warmington S, Forwood MR, Coleman BD, Bennett MB, Wark JD. Age does not influence the bone response to treadmill exercise in female rats. Med. Sci. Sports Exer. 2002;34:1958–1965.
  • Biewener AA. Locomotory stresses in the limb bones of two small mammals: the ground squirrel and chipmunk. J. Exp. Biol. 1983;103:131–154. [PubMed]
  • Biewener AA. Animal Locomotion. Oxford: Oxford University Press; 2003.
  • Biewener AA, Thomason JJ, Goodship AE, Lanyon LE. Bone stress in the horse forelimb during locomotion at different gaits: a comparison of two experimental methods. J. Biomech. 1983;16:565–576. [PubMed]
  • Binkley T, Specker B. Increased periosteal circumference remains present 12 months after an exercise intervention in preschool children. Bone. 2004;35:1383–1388. [PubMed]
  • Bouxsein ML, Myers KS, Shultz KL, Donahue LR, Rosen CJ, Beamer WG. Ovariectomy-induced bone loss varies among inbred strains of mice. J. Bone Min. Res. 2005;20:1085–1092.
  • Bronikowski AM, Morgan TJ, Garland T, Jr, Carter PA. The evolution of aging and age-related physical decline in mice selectively bred for high voluntary exercise. Evol. 2006;60:1494–1508.
  • Burr DB, Turner CH, Naick P, Forwood MR, Ambrosius W, Hasan MS, Pidaparti R. Does microdamage accumulation affect the mechanical properties of bone? J. Biomech. 1998;31:337–345. [PubMed]
  • Center JR, Nguyen TV, Pocock NA, Eisman JA. Volumetric bone density at the femoral neck as a common measure of hip fracture risk for men and women. J. Clin. Endocr. Metab. 2004;89:2776–2782. [PubMed]
  • Ciarelli TE, Fyhrie DP, Schaffler MB, Goldstein SA. Variations in three-dimensional cancellous bone architecture of the proximal femur in female hip fractures and in controls. J. Bone Min. Res. 2000;15:32–40.
  • Crawford WH, Fretz PB. Long-bone fracture in large animals – a retrospective study. Vet. Surg. 1985;14:295–302.
  • Cummings SR, Nevitt MC, Browner WS, Stone K, Fox KM, Ensrud KE, Cauley J, Black D, Vogt TM. Risk factors for hip fractures in white women. New Eng. J. Med. 1995;332:767–773. [PubMed]
  • Demes B, Stern JT, Jr, Hausman MR, Larson SG, McLeod KJ, Rubin CT. Patterns of strain in the Macaque ulna during functional activity. Am. J. Phys. Anth. 1998;106:87–100.
  • Demes B, Qin Y-X, Stern JT, Jr, Larson SG, Rubin CT. Patterns of strain in the Macaque tibia during functional activity. Am. J. Phys. Anth. 2001;116:257–265.
  • Eisman JA. Genetics of osteoporosis. Endocr. Rev. 1999;20:788–804. [PubMed]
  • Embertson RM, Bramlage LR, Herring DS, Gabel AA. Physeal fractures in the horse. 1. Classification and incidence. Vet. Surg. 1986;15:223–229.
  • Ferrari S, Rizzoli R, Bonjour J-P. Genetic aspects of osteoporosis. Curr. Opin. Rheum. 1999;11:294–300.
  • Fischer MS. Crouched posture and high fulcrum, a principle in the locomotion of small mammals: the example of the rock hyrax (Procavia capensis) (Mammalia: Hyracoidea). J. Hum. Evol. 1994;26:501–524.
  • Fischer MS, Schilling N, Schmidt M, Haarhaus D, Witte H. Basic limb kinematics of small therian mammals. J. Exp. Biol. 2002;205:1315–1338. [PubMed]
  • Garland T., Jr . Selection experiments: an under-utilized tool in biomechanics and organismal biology. In: Bels VL, Gasc J-P, Casinos A, editors. Vertebrate Biomechanics and Evolution. Oxford: BIOS Scientific Publishers; 2003. pp. 23–56.
  • Garland T, Jr, Freeman PW. Selective breeding for high endurance running increases hindlimb symmetry. Evol. 2005;59:1851–1854.
  • Garland T, Jr, Kelly SA. Phenotypic plasticity and experimental evolution. J. Exp. Biol. 2006;209:2344–2361. [PubMed]
  • Garnero P, Sornay-Rendu E, Chapuy M-C, Delmas PD. Increased bone turnover in late postmenopausal women is a major determinant of osteoporosis. J. Bone Min. Res. 1996;11:337–349.
  • Gnudi S, Ripamonti C, Gualtieri G, Malavolta N. Geometry of proximal femur in the prediction of hip fracture in osteoporotic women. Br. J. Radiol. 1999;72:729–733. [PubMed]
  • Gnudi S, Ripamonti C, Lisi L, Fini M, Giardino R, Giavaresi G. Proximal femur geometry to detect and distinguish femoral neck fractures from trochanteric fractures in postmenopausal women. Osteop. Int. 2002;13:69–73.
  • Gordon KR, Perl M, Levy C. Structural alterations and breaking strength of mouse femora exposed to three activity regimes. Bone. 1989;10:303–312. [PubMed]
  • Gosden RG, Laing SC, Felicio LS, Nelson JF, Finch CE. Imminent oocyte exhaustion and reduced follicular recruitment mark the transition to acyclicity in aging C57BL/6J mice. Biol. Reprod. 1983;28:255–260. [PubMed]
  • Halloran BP, Ferguson VL, Simske SJ, Burghardt A, Venton LL, Majumdar S. Changes in bone structure and mass with advancing age in the male C57BL/6J mouse. J. Bone Min. Res. 2002;17:1044–1050.
  • Hamrick MW, Ding K-H, Pennington C, Chao YJ, Wu Y-D, Howard B, Immel D, Borlongan C, McNeil PL, Bollag WB, et al. Age-related loss of muscle mass and bone strength in mice is associated with a decline in physical activity and serum leptin. Bone. 2006;39:845–853. [PubMed]
  • Harasen G. Common long bone fracture in small animal practice – part 2. Can. Vet. J. 2003;44:503–504. [PubMed]
  • Heaney RP, Barger-Lux MJ, Davies KM, Ryan RA, Johnson ML, Gong G. Bone dimensional change with age: interactions of genetic, hormonal, and body size variables. Osteop. Int. 1997;7:426–431.
  • Henderson KK, Wagner H, Favret F, Britton SL, Koch LG, Wagner PD, Gonzalez NC. Determinants of maximal O2 uptake in rats selective bred for endurance running capacity. J. Appl. Physiol. 2002;93:1265–1274. [PubMed]
  • Honda A, Umemura Y, Nagasawa S. Effect of high-impact and low-repetition training on bones in ovariectomized rats. J. Bone Min. Res. 2001;16:1688–1693.
  • Hussain SO, Barbato JC, Koch PL, Metting PJ, Britton SL. Cardiac function in rats selectively bred for low- and high-capacity running. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2001;281:R1787–R1791. [PubMed]
  • Iwamoto J, Yeh JK, Aloia JF. Differential effect of treadmill exercise on three cancellous bone sites in the young growing rat. Bone. 1999;24:163–169. [PubMed]
  • Järvinen TLN, Pajamäkai I, Sievänen H, Vuohelainen T, Tuukkanen J, Järvinen M, Kannus P. Femoral neck response to exercise and subsequent deconditioning in young and adult rats. J. Bone Min. Res. 2003;18:1292–1299.
  • Judex S, Gross TS, Zernicke RF. Strain gradients correlate with site of exercise-induced bone-forming surfaces in the adult skeleton. J. Bone Min. Res. 1997;12:1737–1745.
  • Judex S, Garman R, Squire M, Busa B, Donahue LR, Rubin C. Genetically linked site-specificity of disuse osteoporosis. J. Bone Min. Res. 2004;19:607–613.
  • Kannus P, Haapsalo H, Sankelo M, Sievänen H, Pasanen M, Heinonen A, Oja P, Vuori I. Effect of starting age of physical activity on bone mass in the dominant arm of tennis and squash players. Ann. Int. Med. 1995;123:27–31. [PubMed]
  • Kaptoge S, Jakes RW, Dalzell N, Wareham N, Khaw KT, Loveridge N, Beck TJ, Reeve J. Effects of physical activity on evolution of proximal femur structure in a younger elderly population. Bone. 2006;40:506–515. [PubMed]
  • Kelly SA, Czech PP, Wight JT, Blank KM, Garland T., Jr Experimental evolution and phenotypic plasticity of hindlimb bones in high-activity house mice. J. Morph. 2006;267:360–374. [PubMed]
  • Kesavan C, Mohan S, Oberholtzer S, Wergedal JE, Baylink DJ. Mechanical loading-induced gene expression and BMD changes are different in two inbred mouse strains. J. App. Physiol. 2005;99:1951–1957.
  • Kesavan C, Mohan S, Srivastava AK, Kapoor S, Wergedal JE, Yu HR, Baylink DJ. Identification of genetic loci that regulate bone adaptive response to mechanical loading in C57BL/6J and C3H/HeJ mice intercross. Bone. 2006;39:634–643. [PubMed]
  • Klein RF, Shea M, Gunness ME, Pelz GB, Belknap JK, Orwoll ES. Phenotypic characterization of mice bred for high and low peak bone mass. J. Bone Min. Res. 2001;16:63–71.
  • Koch LG, Britton SL. Artificial selection for intrinsic aerobic endurance running capacity in rats. Physiol. Genom. 2001;5:45–52.
  • Kodama Y, Umemura Y, Nagasawa S, Beamer WG, Donahue LR, Rosen CR, Baylink DJ, Farley JR. Exercise and mechanical loading increase periosteal bone formation and whole bone strength in C57BL/6J but not in C3H/HeJ mice. Cal. Tiss. Int. 2000;66:298–306.
  • Koller DL, Schreifer J, Sun Q, Shultz KL, Donahue LR, Rosen CJ, Foroud T, Beamer WG, Turner CH. Genetic effects for femoral biomechanics, structure, and density in C57BL/6J and C3H/HeJ inbred mouse strains. J. Bone Min. Res. 2003;18:1758–1765.
  • Lau KHW, Kapur S, Kesavan C, Baylink DJ. Up-regulation of the Wnt, estrogen receptor, insulin-like growth factor-I, and bone morphogenetic protein pathways in C57BL/6J osteoblasts as opposed to C3H/HeJ osteoblasts in part contributes to the differential anabolic response to fluid shear. J. Bio. Chem. 2006;281:9576–9588. [PubMed]
  • Li J, Liu D, Ke HZ, Duncan RL, Turner CH. The P2X7 nucleotide receptor mediates skeletal mechanotransduction. J. Bio. Chem. 2005;280:42952–42959. [PubMed]
  • Lieberman DE, Pearson OM, Polk JD, Demes B, Crompton AW. Optimization of bone growth and remodeling in response to loading in tapered mammalian limbs. J. Exp. Biol. 2003;206:3125–3138. [PubMed]
  • Lieberman DE, Polk JD, Demes B. Predicting long bone loading from cross-sectional geometry. Am. J. Phys. Anth. 2004;123:156–171.
  • Lovejoy CO, Burstein AH, Heiple KG. The biomechanical analysis of bone strength: a method and its application to platycnemia. Am. J. Phys. Anth. 1976;44:489–506.
  • Malisch JL, Saltzman W, Gomes FR, Rezende EL, Jeske DR, Garland T., Jr Baseline and stress-induced plasma corticosterone concentrations of mice selectively bred for high voluntary wheel running. Physiol. Biochem. Zool. 2007;80:146–156. [PubMed]
  • McNamara LM, Ederveen AGH, Lyons CG, Price C, Schaffler MB, Weinans H, Prendergast PJ. Strength of cancellous bone trabecular tissue from normal, ovariectomized and drug-treated rats over the course of aging. Bone. 2006;39:392–400. [PubMed]
  • Morgan TJ, Garland T, Jr, Carter PA. Ontogenies in mice selected for high voluntary wheel-running activity. I. Mean ontogenies. Evol. 2003;57:646–657.
  • Mosekilde L. Sex differences in age-related loss of vertebral trabecular bone mass and structure – biomechanical consequences. Bone. 1989;10:425–432. [PubMed]
  • Mosekilde L, Thomsen JS, Orhii PB, Kalu DN. Growth hormone increases vertebral and femoral bone strength in osteoporotic, ovariectomized, aged rats in a dose-dependent and site-specific manner. Bone. 1998;23:343–352. [PubMed]
  • Mosley JR, Lanyon LE. Strain rate as a controlling influence on adaptive modeling in response to dynamic loading of the ulna in growing male rats. Bone. 1998;23:313–318. [PubMed]
  • Mosley JR, Lanyon LE. Growth rate rather than gender determines the size of the adaptive response of the growing skeleton to mechanical stress. Bone. 2002;30:314–319. [PubMed]
  • Niehoff A, Kersting UG, Zaucke F, Morlock MM, Brüggemann G-P. Adaptation of mechanical, morphological, and biochemical properties of the rat growth plate to dosedependent voluntary exercise. Bone. 2004;35:899–908. [PubMed]
  • Ohnishi I, Oikawa K, Tsuji K, Ichikawa T, Kurokawa T. A femoral neck fracture model in rabbits. J. Biomech. 2003;36:431–442. [PubMed]
  • Peng Z, Tuukkanen J, Zhang H, Jämsä T, Väänänen HK. The mechanical strength of bone in different rat models of experimental osteoporosis. Bone. 1994;15:523–532. [PubMed]
  • R Development Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing; Vienna, Austria: 2007. http://r-project.org.
  • Raab DM, Smith EL, Crenshaw TD, Thomas DP. Bone mechanical properties after exercise training in young and old rats. J. App. Physiol. 1990;68:130–134.
  • Rhodes JS, Gammie SC, Garland T., Jr Neurobiology of mice selected for high voluntary wheel-running activity. Integr. Comp. Bio. 2005;45:438–455.
  • Riggs BL, Wahner HW, Melton LJ, III, Richelson LS, Judd HL, Offord KP. Rates of bone loss in the appendicular and axial skeletons of women. J. Clin. Inv. 1986;77:1487–1491.
  • Robling AG, Turner CH. Mechanotransduction in bone: genetic effects on mechanosensitivity in mice. Bone. 2002;31:562–569. [PubMed]
  • Robling AG, Li J, Shultz KL, Beamer WG, Turner CH. Evidence for a skeletal mechanosensitivity gene on mouse Chromosome 4. FASEB J. 2003;17:324–326. [PubMed]
  • Rubin CT, Lanyon LE. Regulation of bone formation by applied dynamic loads. J. Bone Joint Surg. 1984;66:397–402. [PubMed]
  • Rubin C, Turner AS, Bain S, Mallinckrodt C, McLeod K. Low mechanical signals strengthen long bones. Nature. 2001;412:603–604. [PubMed]
  • Ruff CB, Trinkaus E, Walker A, Larsen CS. Postcranial robusticity in Homo. I: Temporal trends and mechanical interpretation. Am. J. Phys. Anth. 1993;91:21–53.
  • Russo CR, Lauretani F, Seeman E, Bartali B, Bandinelli S, Di Iorio A, Guralnik J, Ferrucci L. Structural adaptations to bone loss in aging men and women. Bone. 2006;38:112–118. [PubMed]
  • Sawakami K, Robling AG, Ai M, Pitner ND, Liu D, Warden SJ, Li J, Maye P, Rowe DW, Duncan RL, Warman ML, Turner CH. The Wnt co-receptor LRP5 is essential for skeletal mechanotransduction but not for the anabolic bone response to parathyroid hormone treatment. J. Bio. Chem. 2006;281:23698–23711. [PubMed]
  • Seeman E, Tsalamandris C, Formica C, Hopper JL, McKay J. Reduced femoral neck bone density in the daughters of women with hip fractures: the role of peak bone density in the pathogenesis of osteoporosis. J. Bone Min. Res. 1994;9:739–743.
  • Shaw SR, Zernicke RF, Vailas AC, DeLuna D, Thomason DB, Baldwin KM. Mechanical, morphological and biochemical adaptations of bone and muscle to hindlimb suspension and exercise. J. Biomech. 1987;20:225–234. [PubMed]
  • Silva MJ, Gibson LJ. Modeling the mechanical behavior of vertebral trabecular bone: effects of age-related changes in microstructure. Bone. 1997;21:191–199. [PubMed]
  • Silva MJ, Brodt MD, Fan Z, Rho J-Y. Nanoindentation and whole-bone bending estimates of material properties in bones from the senescence accelerated mouse SAMP6. J. Biomech. 2004;37:1639–1646. [PubMed]
  • Slinker BK, Glantz SA. Multiple regression for physiological data analysis: the problem of multicollinearity. Am. J. Physio. Regul. Integr. Comp. Physio. 1985;248:R1–R12.
  • Smith RW, Jr, Walker RR. Femoral expansion in aging women: implications for osteoporosis and fractures. Science. 1964;145:156–157. [PubMed]
  • Snow CM, Williams DP, LaRiviere J, Fuchs RK, Robinson TL. Bone gains and losses follow seasonal training and detraining in gymnasts. Calc. Tissue Int. 2001;69:7–12.
  • Søgaard CH, Danielsen CC, Thorling EB, Mosekilde L. Long-term exercise of young and adult female rates: effects on femoral biomechanical competence and bone structure. J. Bone Min. Res. 1994;9:409–416.
  • Squire M, Donahue LR, Rubin C, Judex S. Genetic variations that regulate bone morphology in the male mouse skeleton do not define its susceptibility to mechanical unloading. Bone. 2004;35:1353–1360. [PubMed]
  • Storey JD. A direct approach to false discovery rates. J. Roy. Stat. Soc. B. 2002;64:479–498.
  • Swallow JG, Garland T., Jr Selection experiments as a tool in evolutionary and comparative physiology: insights into complex traits – an introduction to the symposium. Integr. Comp. Bio. 2005;45:387–390.
  • Swallow JG, Carter PA, Garland T., Jr Artificial selection for increased wheel-running behavior in house mice. Behav. Gen. 1998;28:227–237.
  • Swallow JG, Koteja P, Carter PA, Garland T., Jr Artificial selection for increased wheel-running activity in house mice results in decreased body mass at maturity. J. Exp. Biol. 1999;202:2513–2520. [PubMed]
  • Trinkaus E. Appendicular robusticity and the paleobiology of modern human emergence. PNAS. 1997;94:13367–13373. [PubMed]
  • Turner CH, Burr DB. Experimental techniques for bone mechanics. In: Cowin SC, editor. The Bone Mechanics Handbook. Boca Raton: CRC Press; 2001. pp. 7.1–7.35.
  • Umemura Y, Ishiko T, Tsujimoto H, Miura H, Mokushi N. Effects of jump training on bone hypertrophy in young and old rats. Int. J. Sports Med. 1995;16:364–367. [PubMed]
  • Umemura Y, Baylink DJ, Wergedal JE, Mohan S, Srivastava AK. A time course of bone response to jump exercise in C57BL/6J mice. J. Bone Min. Metab. 2002;20:209–215.
  • Weinstein RS, Hutson MS. Decreased trabecular width and increased trabecular spacing contribute to bone loss with aging. Bone. 1987;8:137–142. [PubMed]
  • Welch JM, Weaver CM, Turner CH. Adaptations to free-fall impact are different in the shafts and bone ends of rat forelimbs. J. App. Physiol. 2004;97:1859–1865.
  • Wergedal JE, Sheng MH-C, Ackert-Bicknell CL, Beamer WG, Baylink DJ. Genetic variation in femur extrinsic strength in 29 different inbred strains of mice is dependent on variations in femur cross-sectional geometry and bone density. Bone. 2005;36:111–122. [PubMed]
  • Wisløff U, Najjar SM, Ellingsen Ø, Haram PM, Swoap SJ, Al-Share Q, Fernström M, Rezaei K, Lee SJ, Koch LG, Britton SL. Cardiovascular risk factors emerge after artificial selection for low aerobic capacity. Science. 2005;307:418–420. [PubMed]
  • Witte H, Biltzinger J, Hackert R, Schilling N, Schmidt M, Reich C, Fischer MS. Torque patterns of the limbs of small therian mammals during locomotion on flat ground. J. Exp. Biol. 2002;205:1339–1353. [PubMed]
  • Wu J, Wang X, Chiba H, Higuchi M, Nakatani T, Ezaki O, Cui H, Yamada K, Ishimi Y. Combined intervention of soy isoflavone and moderate exercise prevents body fat elevation and bone loss in ovariectomized mice. Metab. 2004;53:942–948.
  • Yamamoto N, Jee WSS, Ma YF. Bone histomorphometric changes in the femoral neck of aging and ovariectomized rats. Anat. Rec. 1995;243:175–185. [PubMed]
  • Zhang L, Cheng A, Bai Z, Lu Y, Endo N, Dohmae Y, Takahashi HE. Epidemiology of cervical and trochanteric fractures of the proximal femur in 1994 in Tangshan, China. J. Bone Min. Metab. 2000;18:84–88.

PubMed articles by these authors
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph