SEARCH

SEARCH BY CITATION

Keywords:

  • trabecular architecture;
  • anisotropy;
  • high-resolution computed tomography;
  • structural adaptation;
  • 3D morphometric analysis

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

The anisotropic arrangement of trabeculae in the proximal femur of humans and primates is seen as striking evidence for the functional adaptation of trabecular bone architecture. Quantitative evidence to demonstrate this adaptation for trabecular bone is still scarce, because experimental design of controlled load change is difficult. In this work, we use the natural variation of loading caused by a different main locomotor behavior of primates. Using high-resolution computed tomography and advanced image analysis techniques, we analyze the heterogeneity of the architecture in four proximal femora of four primate species. Although the small sample number does not allow an interspecies comparison, the very differently loaded bones are well suited to search for common structural features as a result of adaptation. A cubic volume of interest of size (5 mm)3 was moved through the proximal femur and a morphometric analysis including local anisotropy was performed on 209 positions on average. The correlation of bone volume fraction (BV/TV) with trabecular number (Tb.N) and trabecular thickness (Tb.Th) leads to the suggestion of two different mechanisms of trabecular bone adaptation. Higher values of BV/TV in highly loaded regions of the proximal femur are due to a thickening of the trabeculae, whereas Tb.N does not change. In less loaded regions, however, lower values of BV/TV are found, caused by a reduction of the number of the trabeculae, whereas Tb.Th remains constant. This reduction in Tb.N goes along with an increase in the degree of anisotropy, indicating an adaptive selection of trabeculae. Anat Rec, 2010. © 2010 Wiley-Liss, Inc.

The observation of a nonrandom and anisotropic trabecular architecture, for example in the proximal femur of humans and primates, led Julius Wolff and contemporaries to the conclusion that the bone structure follows a mechanical optimization principle (Huiskes,2000; Skedros and Baucom,2007). Much later the concept of a mechanostat controlling the process of bone (re)modeling was proposed (Frost,1987) to explain functional bone adaptation (Robling et al.,2006; Fratzl and Weinkamer,2007) during life. Further progress still needs to be made, however, both on the mechanical characterization of the loading conditions and the geometrical characterization of the bone. There are several difficulties in relating external mechanical loading to the resulting bone structure and the underlying remodeling process. The full information on how the bones were loaded over the life of the organism cannot be known. Even with complete knowledge about the loading history, it remains unclear to which mechanical signal bone reacts and how this signal is further processed (Burr,1993; Burger and Klein-Nulend,1998). In addition to a mechanical analysis, a detailed three-dimensional geometrical characterization of trabecular bone is essential to quantify the spatial heterogeneity of the trabecular bone architecture.

Adaptation via bone remodeling is the response of the organic system of an individual to a specific life style. This individual adaptation of the bone microarchitecture, however, is thought to be superimposed on genetic adaptation, which occurs on the time scale of many generations via random variation in the genetic code combined with natural selection. The result of genetic adaptation is bone architecture, which is “appropriate” for the habitual loading conditions of the whole species. Although a clear separation between the two adaptation processes seems unfeasible, investigations focus typically on one of the two processes. The study of differences in the bone architecture of many individuals of the same species with very different loading on the bone [e.g., by immobilization (Shen et al.,1997) or overloading (Goodship et al.,1979)] centers on individual adaptation. In our study, we used single representatives of different primate species with clearly differing locomotor repertoires to test a new method for detecting structural differences, which are expected to exist between these species. The nature of the adaptation, whether genetic and/or individual, is of minor relevance to this study.

Recent experimental and computational advances allow new ways to analyze trabecular bone architecture. High-resolution computed tomography (HrCT) provides detailed three-dimensional information of the spatial structure of trabecular bone in samples of several centimeters in size (Muller and Ruegsegger,1996; Muller,2009). Based on HrCT data, qualitative and quantitative assessments of the proximal femoral trabecular architecture of different primate species could demonstrate differences in the spatial arrangement of the trabecular bone (Fajardo and Muller,2001; MacLatchy and Muller,2002; Ryan and Ketcham,2002,2005; Fajardo et al.,2007; Scherf,2007,2008; Ryan and Walker2010). Interpretation of the results linked these differences to the preferred locomotor loading of the femora. Probably due to the absence of a standard analyzing protocol, for example, concerning the volume of interest (VOI) under investigation, these studies did not yield coherent relations connecting distinct locomotor habits and morphometric parameters such as the degree of anisotropy (DA) and BV/TV for the same species (MacLatchy and Muller,2002; Ryan and Ketcham,2002). Images obtained by HrCT can be directly used as input to microscopic mechanical analysis using microfinite element (μ-FE) models (Van Rietbergen et al.,2003; Homminga et al.,2004). Application of this method to the proximal femur of two different primates led to the conclusion that the differences in the femoral head trabecular architecture may not be fully explainable by differences of the hip joint loading (Ryan and van Rietbergen,2005).

The work presented here abstains from a micromechanical analysis but concentrates on a detailed morphometric analysis. A new approach is introduced, which combines a careful selection of bone samples and HrCT imaging with advanced position-resolved image analysis methods. As samples we chose the proximal femora of wild shot primates, which can be assigned by their locomotor preferences to four different locomotor groups: quadrupedal walker, springer, brachiator, and climber. Using HrCT, the trabecular architecture throughout the whole proximal femur was imaged. A position-resolved analysis of these images was performed by moving a cubic VOI of size (5 mm)3 through the proximal femur and analyzing the local trabecular architecture at 209 positions on average. Previous studies used volumes of interest (VOI) of rather large size with the disadvantage that the VOI included trabecular bone of different orientation (Fajardo and Muller,2001; MacLatchy and Muller,2002; Ryan and Ketcham,2002,2005; Ryan and van Rietbergen,2005). On the other hand, the VOI cannot be made arbitrarily small, as otherwise quantities like a principal orientation of trabeculae lose their significance (Harrigan et al.,1988). The size of our suggested VOI is therefore a compromise to obtain position-resolved and meaningful information about the trabecular architecture. In this way, position-resolved information not only about standard bone morphometric parameters such as bone volume fraction (BV/TV), trabecular number (Tb.N), and trabecular thickness (Tb.Th) can be obtained but also local measures of the anisotropy of trabecular bone. The potential of the method is demonstrated on single samples of each primate. The sample number is too small to permit a comparison between the different primates, but the structural analysis allows conclusions about common mechanisms of bone adaptation. The new technique of position-resolved analysis presented here will facilitate future studies on bone adaptation and allow a more thorough understanding of overall trabecular orientation in connection with locomotor behavior.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

The architecture of trabecular bone in the same regions of the proximal femora of the four primate species of Papio hamadryas, Semnopithecus entellus, Symphalangus syndactylus, and Alouatta seniculus was analyzed. These genera were chosen by virtue of their preferred mode of locomotion, with the main premise that each species exhibit a specialized mode of habitual locomotion, which differs from the other three species. The influence of body weight and body size was considered in the selection procedure of the species, so that differences in body weight and size were kept as small as possible (Table 1).

Table 1. Analyzed primates
SpeciesLocomotor groupBody weight rangea (kg)Head to trunk lengtha (cm)Length of entire femorac (cm)Voxel size of CT scan (μm)
  • a

    Male individuals; after Grzimek,1990.

  • b

    Direct measurement obtained from the analyzed animal.

  • c

    Average over five measurements obtained from each of the analyzed femora.

Papio hamadryas, [38] Sacred baboonQuadrupedal walker24b60–9424.961
Semnopithecus entellus, [39] Hanuman langurSpringer9–20.951–10822.748
Symphalangus syndactylus, [40] SiamangBrachiator∼1147–6020.643
Alouatta seniculus, [41] Red howler monkeyClimber6.5–8.149–7216.037

To exclude influences of unnatural locomotion, altered nutrition, as well as effects from growth, only femora of wild shot and adult individuals have been used. The bone specimens were provided by the collection “Dr. Senckenbergische Anatomie” (Johann Wolfgang Goethe University Frankfurt am Main, Germany) and by the Institute of Zoology, University of Hamburg (Germany). All specimens belonged to male individuals, except Semnopithecus e. whose sex is unknown. These four primate species are characterized by different locomotor behaviors resulting in different mechanical loading of the proximal femora. Table 1 lists important details of the study samples including the species and their preferred locomotor behavior.

Other locomotor abilities of these species are neglected by this classification, as the predominant locomotor mode is considered to have the greatest influence on load directed adaptation of bone. An outline of the range of the locomotor abilities of the chosen species is listed below:

Papio hamadryas—Sacred Baboon

Like all baboons, they usually walk quadrupedal and spend most of their time on the ground. Sometimes they gallop or trot. During tree climbing, they clutch the substrate only with their hands while placing the feet flat on the branches. Baboons rarely leap and their leaps are rather short and performed carefully (Hall,1962).

Semnopithecus entellus—Hanuman Langur

The individual used in this study originated from a subtropical forest region. Nikolei (2002) described that Hanuman langurs in wooded habitats are arboreal but frequently descending onto the ground. Their locomotor activities consist of quadrupedal walking, trotting, and galloping, whereas walking is the most frequent locomotor mode. Hanuman langurs leap quite often and leaps up to 10 m are reported. A further common kind of locomotion is climbing.

Symphalangus syndactylus—Siamang (Gibbon)

Siamangs live arboreally and like all gibbons, they predominantly brachiate and are able to cover distances up to 10 m by flight phases. Apart from brachiation, climbing is the most common mode of locomotion. Their locomotor repertoire includes bipedal walking which, unlike human walking, contains bouncing elements (Carpenter,1940; Fleagle,1976; Napier,1976; Preuschoft,1990; Vereecke,2006).

Alouatta seniculus—Red howler Monkey

Howler monkeys rarely come onto the ground and spend most of their time in the treetops where they rest up to 80% of their active daytime. They climb and walk very slowly and carefully and use their prehensile tail like an extra grasping limb to secure themselves during movement. Howler monkeys try to avoid jumps and prefer to climb from one tree to another. However, if necessary they can perform leaps over a distance of 3–4 m (Carpenter,1934; Welker and Schäfer-Witt,1990).

The assumption of different loading conditions due to different locomotor preferences in these four species is supported by the following studies. Hanna et al. (2006) noted an increase in the vertical peak force of the limbs, comparing walking and galloping in different primates, including Papio. Demes and Günther (1989) and Günther (1989) showed that during leaping, the experienced forces are even higher up to the fivefold of the body weight in bigger primates. Vice versa, locomotor modes, which involve less powerful locomotion of the hind limbs, such as climbing in Alouatta seniculus or brachiating in Symphalangus syndactylus, can be assumed to cause lower loading conditions on the hind limbs. In addition, the loading directions are supposed to vary considerably during climbing (Fleagel,1976). The difference between the “climber” Alouatta seniculus and the “brachiator” Symphalangus syndactylus was observed before (Scherf,2008). A possible reason for this difference could be the greater diversity in hind limb associated locomotion in Symphalangus syndactylus.

The collected proximal femora were imaged with the high-resolution computed tomography (HrCT) system (RayScan 200, Hans Wällischmiller GmbH, Germany) at the Foundry Technology centre of the Aalen University of Applied Sciences with a voxel size ranging from 37 μm to 60 μm as defined by the geometric size of the specimens (Table 1). The imaged part of the proximal femur included femoral head, femoral neck, greater trochanter, and part of the femoral shaft including lesser trochanter up to the region of the shaft containing no trabecular bone. Frontal central sections of the proximal femora reconstructed from HrCT data are shown in Fig. 1 for all four different primates.

thumbnail image

Figure 1. Frontal central sections of high-resolution computed tompgraphy (HrCT) images of the four analyzed proximal femora from primate species of (a) “walker” Papio h., (b) “springer” Semnopithecus e., (c) “brachiator” Symphalangus s., and (d) “climber” Alouatta s.

Download figure to PowerPoint

Our novel approach performs a local analysis of trabecular bone architecture within a cubic VOI, which was moved within the proximal femur of the primates (Fig. 2a). [For more details about the moving cube method see (Saparin et al.,2006; Marwan et al.,2009).] The size of the cubic moving window was 5 mm × 5 mm × 5 mm. To obtain consistent and comparable structural evaluation, the same window size was used to analyze all femora. The width of the cubic window was selected so that it included at least five trabeculae in each direction (Harrigan et al.,1988) to provide reliable estimation of morphometric parameters in all analyzed bone specimens. The shift between each position of the cubic window in x-, y-, and z-direction was one half of the window width or 2.5 mm. The cubic window was used to analyze exclusively trabecular bone. Positions where the parts of the cortical bone were included into the window were discarded. The number of the analyzed local VOIs per each proximal femur was defined by the geometric size and the shape of the bone. On average, 209 positions of the moving cubic VOI per bone sample were evaluated.

thumbnail image

Figure 2. (a) Cubic moving window of size 5 mm × 5 mm × 5 mm (red) shown within the 3D volume rendering of proximal femora of Papio h. (b) The schematic split of the proximal femur into four anatomic regions: femoral head (blue), femoral neck (red), greater trochanter (green), and femoral shaft (magenta). The boundaries between the regions were drawn based on anatomic features of the proximal femora and using several planes that were parallel to the main axes (see explanation in the text).

Download figure to PowerPoint

To compare the architecture of the trabecular bone in different parts of the proximal femur, it was divided into four regions: femoral head, femoral neck, greater trochanter, and femoral shaft (Fig. 2b). The boundaries between the regions were drawn based on anatomic features of the proximal femora and using several planes that were parallel to the main axes (Fig. 2b). The part of the femur below the onset of the smaller trochanter was assigned to the region of the femoral shaft. The parasagittal plane, which includes the symmetry axis of the shaft, divided the greater trochanter and femoral neck regions. For practical reasons, the oblique boundary between the femoral neck and femoral head was approximated by a stair-like set of segmented planes that were adjusted to the outer anatomic shape of the head-neck transition. These divisions match roughly biomechanical regions of the femur. The head and the greater trochanter are both regions, which are directly subjected to external loads. The body weight is the loading source of the femoral head, whereas the gluteal muscles transmit loads onto the greater trochanter. The stresses and strains arising on and in the femoral head are transmitted through the femoral neck. The femoral shaft is subjected to the combined stresses and strains coming from the femoral head via the femoral neck and directly from the region of the greater trochanter.

Three-dimensional data processing, analysis, and visualization were performed using the visualization system Amira 4.1 (http://www.amiravis.com/). For each analyzed femur, the optimal bone/soft tissue segmentation threshold was found by use of attenuation histograms. The resulting data sets were saved as a stack of 2D slices, which were imported for morphometric evaluation into CTAn software (SkyScan, Kontich, Belgium). The following local morphometric parameters were calculated for each position of the cubic window: bone volume fraction, BV/TV, trabecular thickness, Tb.Th, trabecular number, Tb.N, structure model index, SMI, DA, and local main direction of the trabeculae, MDT, based on minimum intercept lengths and eigenvalue analysis (Odgaard,1997).

The calculation of the BV/TV was based on the tetrahedron meshing of the bone volume and the tissue volume by the marching cube algorithm (Lorensen and Cline,1987). Model independent 3D estimation of the trabecular thickness Tb.Th was obtained by 3D sphere fitting (Hildebrand and Ruegsegger,1997a). An estimation of the trabecular number Tb.N was calculated using the relation,

  • equation image(1)

Calculation of the DA and the local main direction of trabeculae were based on the analysis of the mean intercept length (MIL) and eigenvector/eigenvalue analysis (Harrigan and Mann,1984; Simmons and Hipp,1997). The MIL analysis was performed in the maximal spherical region that fitted within each analyzed cubic VOI. Outcomes of the analysis were three eigenvectors and three corresponding eigenvalues. The direction given by the eigenvector, equation image, corresponding to the minimal eigenvalue, EVmin, defines the direction along which the smallest number of trabeculae was crossed, and therefore, indicates the local main direction of the trabeculae, MDT. The MDT is expressed by two angles in spherical coordinates, the polar angle, θ, 0 ≤ θ ≤ π, and azimuthal angle, φ, 0 ≤ φ ≤ 2π. The DA is calculated as DA = EVmax/EVmin and can vary from 1 (fully isotropic) to ∞ (fully anisotropic).

Having the local morphometric information in each cubic VOI, the data is evaluated with respect to the four anatomical regions. In case of scalar measures such as BV/TV, Tb.Th, Tb.N, and DA, this assessment can be displayed in form of frequency distributions (Fig. 3). The vectorial MDT polar plots give a good visualization of the orientation distribution of trabeculae within an anatomical region (Fig. 7). In addition, we propose a new parameter, called DC, which quantifies the collective alignment (or spread) of local MDTs (or minimal eigenvectors, respectively). The DC is calculated as

  • equation image(2)

where N denotes the total number of analyzed local VOIs (cubic windows) in the considered anatomical region. The ambiguity in the direction of the eigenvectors (either equation image or − equation image) is eliminated by choosing the eigenvectors out of a predefined halfspace (or hemisphere). We selected the low (southern) hemisphere and transformed all main trabecular directions to the interval 90° ≤ θ ≤ 180°, before calculating DC according to Eq. (2). The value of DC can change from 0% to 100%. For a set of randomly oriented vectors DC is 0%, for a set of parallel vectors DC is 100%, and for vectors whose orientations are uniformly distributed within one hemisphere the DC value is 63.3%.

thumbnail image

Figure 3. Stacked histograms of measured values of local trabecular thickness Tb.Th (a) and local trabecular number Tb.N (b) constructed from proximal femora of “walker” Papio h., “springer” Semnopithecus e., “brachiator” Symphalangus s., and “climber” Alouatta s. Color of the histogram bars represents the measurements obtained in different anatomical regions of the proximal femora.

Download figure to PowerPoint

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Relation between the Bone Volume Fraction, Trabecular Thickness, and Trabecular Number in Different Regions of Proximal Femora

The histograms of Fig. 3 show the frequency distributions of Tb.Th and Tb.N, and therefore, the variability of these quantities within a femur. The width of the distributions is remarkably different for the different femora and can deviate strongly from a normal distribution. The distributions for Tb.Th of Alouatta s. are narrower than for Papio h. in total and for each anatomical region separately. The frequency distribution of Tb.N for Alouatta s. displays a bimodal character, where the data from the femoral head is different to the data of all other anatomical regions.

The relation of the local bone volume, BV/TV, in each of the 836 cubic VOIs of all femora and the local trabecular number, Tb.N and Tb.Th is summarized in Fig. 4. Although the information content of the first plot in Fig. 4 can be transformed into the second plot by using Eq. (1), showing both plots is instructive. The dependence of local Tb.N and Tb.Th versus BV/TV exhibits two regions with either a linear or a constant relation (highlighted by yellow and orange underlying rectangles including some overlap). The first (yellow) region with the BV/TV from 0% to roughly 50%, is characterized by the linear relation between the BV/TV and Tb.N with a correlation coefficient rlin = 0.86 (Fig. 4a). This region in BV/TV comprises all data from femoral shaft, greater trochanter, and femoral neck of all primates, and additionally the data from the femoral head of the “climber” Alouatta s. In the plot BV/TV versus Tb.Th (Fig. 4b) this range of BV/TV values corresponds to a region of rather constant values for Tb.Th. The second (orange) region with BV/TV above about 40% is composed by the measurements from the femoral heads of Papio h., Semnopithecus e., and Symphalangus s. In Fig. 4a, this region is characterized by larger values of BV/TV but constant value of the Tb.N. Correspondingly, Fig. 4b demonstrates a strong linear correlation between the BV/TV and Tb.Th with rlin = 0.96.

thumbnail image

Figure 4. Relation between the local bone volume, BV/TV, and local trabecular number Tb.N (a) and local trabecular thickness Tb.Th (b), respectively. Data from all 836 evaluations within the moving cubic volume of interest of all femora are included. The position of the cubic VOI in different anatomical regions of the proximal femora is coded by color: femoral head (blue), femoral neck (red), greater trochanter (green), and femoral shaft (magenta). Different shape of the symbols represents different specimens: “walker” Papio h. (circles), “springer” Semnopithecus e. (diamonds), “brachiator” Symphalangus s. (triangles), and “climber” Alouatta s. (squares). The yellow and orange rectangles are guides to the eye to highlight the linear and constant relations between BV/TV, and Tb.Th and Tb.N, respectively.

Download figure to PowerPoint

Direction-Specific Optimization of Trabecular Number

The proximal femur of Symphalangus s. (brachiator) showed the lowest value for the DA in all four anatomical regions. Papio h. (walker) is characterized by large values of DA, which are roughly independent of the anatomical region. The proximal femur of Alouatta s. has very special characteristics. It combines a low value of BV/TV with a strong anisotropy with DA values ranging from 1.95 to 3.4. In all regions of the femur of Alouatta s., we found a strong linear relation (rlin = 0.98) between the amount of bone material, expressed by BV/TV, and the trabecular number Tb.N (Fig. 5a). For the relation between Tb.N and DA, we find a negative linear correlation (rlin = −0.69; Fig. 5b) demonstrating that lower values of Tb.N are accompanied by larger values of DA.

thumbnail image

Figure 5. Relation between the local trabecular number Tb.N and the local bone volume to total volume ratio BV/TV (a), and the relation between the local trabecular number Tb.N and the local degree of anisotropy DA (b) for the investigated proximal femur of the “climber” Alouatta s. Color of symbols denotes different anatomical regions of the proximal femur. The linear regressions (dashed line) resulted for (a) in a slope of 0.030 with r2 = 0.96 and for (b) in a slope of −1.71 with r2 = 0.38.

Download figure to PowerPoint

Orientation of Trabeculae

The analysis of the local main directions of trabeculae (MDT) is summarized in two figures. Although Fig. 6 gives the full three-dimensional information, the polar plots of Fig. 7 show the MDTs for the four anatomical regions separately. Among the four analyzed anatomical regions of the proximal femora, the region of the greater trochanter is characterized by the highest collective alignment of the trabeculae. This can be seen by the green straight-lines pointing downward in Fig. 6 and the accumulation of the data points around the center of the polar plot in Fig. 7. This observation is quantified in the DC, which takes for the great trochanter in all primates values from 96% to 97.8%. Differences between the four femora can be observed for the trabecular orientation in the femoral head. The femoral heads of Papio h., Semnopithecus e., and Alouatta s. display a DA that is 11% and 30% higher and a DC that is 20% to 48% higher (Table 1) when compared with the femoral head of Symphalangus s. (brachiator). The dominant direction of the trabeculae in the femoral head of Papio h., Alouatta s., and Semnopithecus e. is mainly limited to the intervals of −90° ≤ φ ≤ 90° and 120° ≤ θ ≤ 170° (Fig. 7) indicating a predominant trabecular orientation directed toward the femoral neck. In contrast, most trabeculae in the femoral head of the Symphalangus s. specimen are oriented along the radial directions, which results in a more uniform distribution of the elevation angle θ and the azimuth angle φ as shown in Fig. 7.

thumbnail image

Figure 6. 3D rendering of superimposed bone architecture and main local directions of the trabeculae obtained for different regions of differently loaded proximal femora of primates: “quadrupedal walker” Papio h. (a), “springer” Semnopithecus e. (b), “brachiator” Symphalangus s. (c), and “climber” Alouatta s. (d). Orientation of the lines represent local main direction of the trabeculae, their lengths are proportional to the local degree of anisotropy (DA), and the color of the lines corresponds to different anatomical regions of the proximal femur: femoral head (blue), femoral neck (red), greater trochanter (green), and femoral shaft (magenta).

Download figure to PowerPoint

thumbnail image

Figure 7. Main local directions of the trabeculae projected onto lower hemisphere obtained for different regions of differently loaded proximal femora of primates: “quadrupedal walker” Papio h., “springer” Semnopithecus e., “brachiator” Symphalangus s., and “climber” Alouatta s. Latitude and longitude represent correspondingly the elevation angle θ and the azimuth angle φ of the main direction of the trabeculae (MDT) is given by the local eigenvector with the minimal eigenvalue. An angle φ = 0 corresponds to the medial, φ = 180 to the lateral direction. Since all the bones were from the left side, φ = 90 corresponds to anterior, and φ = 270 to posterior direction.

Download figure to PowerPoint

Structure Model Index–Rod-Like Versus Plate-Like Trabeculae

The structure model index, SMI, characterizes the shape of the trabeculae (Hildebrand and Ruegsegger,1997b) with a value of 3 for an ideal rod-like, 0 for an ideal plate-like architecture, and negative values for “swiss-cheese” architectures of negative mean curvature (Stauber and Muller,2006). Strong correlation (rlin = −0.98) was found between the BV/TV and the SMI for the entire set of analyzed local VOIs in all four femora. If the proximal femora of different species are considered separately (Fig. 8), the Spearman's rank correlation coefficient between the SMI and the BV/TV for the entire proximal femora of Papio h., Semnopithecus e., Symphalangus s., and Alouatta s. was correspondingly −0.98, −0.99, −0.96, and −0.97. All the data lies reasonably well on a master curve defined by a polynomial of second order (with BV/TV given in %),

  • equation image(3)

with the only exception that the data points of Alouatta s. are slightly below. The strong correlation between SMI and BV/TV together with the correlation of BV/TV and Tb.Th (or Tb.N) shown in Fig. 4 translates naturally in a correlation between SMI and Tb.Th (or Tb.N).

thumbnail image

Figure 8. Relation between bone volume fraction, BV/TV, and the structure model index, SMI, for all four primates distinguishing with colors the different anatomical regions: (a) “walker” Papio h., (b) “springer” Semnopithecus e., (c) “climber” Alouatta s., and (d) “brachiator” Symphalangus s. With the exception of Alouatta s., all the data is well fitted with a quadratic function given by Eq. (3) (dashed line).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

A novel method was used to analyze the structure of trabecular bone position resolved in the proximal femur of primates with different locomotor behavior and therefore different loading conditions on their femur. Within a small moving cubic VOI, local morphometric parameters, and the local orientation of the trabeculae were analyzed. As demonstrated by Fig. 6, this method allows defining geometrical trajectories of the bone architecture, which is not possible with more global VOIs. Remarkable are the strong relationships, which are obtained, for example, between BV/TV and SMI in Fig. 8a over a very large range of BV/TV, which might be the result of the analysis performed within one bone of a single individual. The method allows quantifying the variability of architectural parameters (Fig. 3) and of the trabecular orientation (Fig. 7) within the proximal femur. However, for a statement that the observed differences are characteristic for the different primate species and to make a connection to their locomotor behavior, the number of analyzed individuals has to be larger. Concerning our approach it should be emphasized that the method cannot be further improved by decreasing the VOI (at least in the species sampled here), as a minimum volume is necessary for a reasonable calculation of the morphometric parameters (Harrigan et al.,1988).

The analysis showed striking evidence for a functional adaptation of the trabecular architecture. Taking the data from all 836 analyzed VOIs and relating bone volume fraction (BV/TV) with either trabecular number (Tb.N) or trabecular thickness (Tb.Th) revealed two different mechanisms of trabecular bone adaptation (Fig. 4).

  • A reduction of BV/TV is obtained by a decrease of Tb.N, while Tb.Th is kept constant.

  • An increase of BV/TV, however, occurs by an increase in Tb.Th by constant Tb.N.

This observation can be understood, based on the experimental finding that bone is formed with a roughly uniform Tb.N (Parfitt et al.,2000). In agreement with values found for children (Parfitt et al.,2000), we assume that this value is about 1.7 mm−1. Starting now from this value of Tb.N and intermediate BV/TV from about 45%, the application of Wolff's law and the idea of the mechanostat allow a simple interpretation of the observed distribution of data points. Where mechanically needed, new bone is deposited leading to an increase in BV/TV. As it is easier to thicken existing trabeculae than to create new ones, BV/TV and Tb.Th are linearly correlated as shown in Fig. 4b. The trabecular bone of the femoral head is known to be highly loaded (Van Rietbergen et al.,2003). Actually for all studied primates (with the exception of Alouatta s.), the trabeculae in the femoral head are reinforced by a thickening. This interpretation is in accordance to what is known from the development of human trabecular bone about 1 year after birth when infants start walking unassisted. In the proximal femur (Ryan and Krovitz,2006), in the proximal tibia (Gosman and Ketcham,2009) and in the iliac crest (Parfitt et al.,2000), it was found that during the juvenile period the trabecular number was almost unchanged, whereas the trabecular thickness continued to increase.

Inverting now the argument for the case of low mechanical load on the trabecular architecture, the rules of functional adaptation predict a reduction of BV/TV. Principally this can be achieved by either reducing Tb.Th or Tb.N. Reduction of the thickness of a trabecula runs into the problem that the formation of resorption cavity during bone remodeling can irreversibly perforate a trabecula, resulting in its loss. The “safer” solution evidently adopted by nature is to reduce the Tb.N (Fig. 4a). Anatomical regions where this adaptation mechanism is applied include the femoral neck, the greater trochanter, the femoral shaft, and the femoral head of Alouatta s. The femur of the “climber” Alouatta s. shows the lowest volume for BV/TV and consequently the lowest values for Tb.N. As another indication for functional adaptation, this strong reduction in Tb.N is not obtained by a random removal of trabeculae. As Fig. 5b demonstrates, a reduction in Tb.N is accompanied with a gain in the DA of the trabecular structure. This can be understood under the assumption that bone volume is reduced by preferentially removing mechanically unnecessary trabeculae. Investigations on the structural evolutions in human vertebrae showed that the alignment of the trabeculae along the main loading directions is not visible 5 weeks after birth but is well developed at an age of 8 years (Roschger et al.,2001). We hypothesize that also in primates the structural adaptation initiates early after birth (Ryan and Krovitz,2006), where it remains unclear how much of the structural adaptation is genetically determined and how much is the result of the actual mechanical loading after birth. Experiments on growing pigs demonstrated that adaptation of bone volume and bone architecture can occur with different speed. The bone volume fraction in the pigs increases rapidly, which can be ascribed to their enormous weight increase. The architectural anisotropy reached its maximum much later in the development than did the bone volume fraction (Tanck et al.,2001). The application of our analysis method to younger animals of the same species could clarify the time course of the structural adaptation.

Functional adaptation has been very well demonstrated for cortical bone. An increase in the external loading due to exercise or special devices resulted in an increase in the cross-sectional area of the cortex (Goodship et al.,1979). Much more difficult is the experimental design to demonstrate functional adaptation of trabecular bone architecture and therefore quantitative evidence is still rather scarce. This is the reason why computer simulations are used to understand the control of trabecular bone remodeling (Huiskes et al.,2000; Dunlop et al.,2009). A good experimental strategy is to use the natural variation of external loading because of the different locomotor preferences in diverse animals. The method presented in this study facilitates future studies on trabecular bone differences between species. Such studies clearly necessitate a larger sample set, ideally with animals of the same species but particular modes of locomotion. In addition, the local geometrical information about the bone architecture should be compared with local mechanical information, for example, from microfinite element analysis. Such knowledge can eventually explain intraspecies differences and further understanding of functional adaptation processes in bone.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

The authors thank Dr. Helmut Wicht for providing us with the femora of Papio hamadryas, Symphalangus syndactylus, and Alouatta seniculus; Prof. Dr. Harald Schliemann and Nelson Mascarenhas, Institute of Zoology, University of Hamburg, Germany, for provision of the femur of Semnopithecus entellus; Dr. Irmgard Pfeifer-Schäller for her patience and support in HrCT scanning. The authors thank John Dunlop for a careful reading of the manuscript and to Paul Zaslansky for help with the visualization.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED
  • Burger EH, Klein-Nulend J. 1998. Mechanotransduction in bone—role of the lacuno-canalicular network. In: Symposium of the European-Space-Agency/National-Aeronautics-and-Space-Administration Workshop on Cell and Molecular biology Research in Space. Leuven, Belgium: Federation of Amererican Societies for Experimental Biology. p S101S112.
  • Burr DB. 1993. Remodeling and the repair of farigue damage. Calcif Tissue Int 53: S75S81.
  • Carpenter CR. 1934. A field study of the behaviour and social relations of Howling monkeys (Alouatta palliata). Comparative psychology monographs X. Baltimore: The John Hopkins Press. p 1168.
  • Carpenter CR. 1940. A field study in Siam of the behavior and social relations of the gibbon (Hylobates lar). Reprinted in: Naturalistic behavior of non-human primates (Carpenter CR, 1964). University Park: The Pennsylvania State University Press. p 145271.
  • Demes B, Günther MM. 1989. Biomechanics and allometric scaling in primate locomotion and morphology. Fol Primatol 53: 125141.
  • Dunlop JWC, Hartmann MA, Brechet YJ, Fratzl P, Weinkamer R. 2009. New Suggestions for the Mechanical Control of Bone Remodeling. Calcif Tissue Int 85: 4554.
  • Fajardo RJ, Muller R. 2001. Three-dimensional analysis of nonhuman primate trabecular architecture using micro-computed tomography. Am J Phys Anthropol 115: 327336.
  • Fajardo RJ, Muller R, Ketcham RA, Colbert M. 2007. Nonhuman anthropoid primate femoral neck trabecular architecture and its relationship to locomotor mode. Anat Rec 290: 422436.
  • Fleagle JG. 1976. Locomotion and posture of the malayan Siamang and implications for hominoid evolution. Fol Primatol 26: 245269.
  • Fratzl P, Weinkamer R. 2007. Nature's hierarchical materials. Prog Mater Sci 52: 12631334.
  • Frost HM. 1987. Bone mass and the mechanostat—a proposal. Anat Rec 219: 19.
  • Goodship AE, Lanyon LE, McFie H. 1979. Functional adaptation of bone to increased stress. J Bone Joint Surg A 61: 539546.
  • Gosman JH, Ketcham RA. 2009. Patterns in ontogeny of human trabecular bone from SunWatch Village in the prehistoric Ohio Valley: general features of microarchitectural change. Am J Phys Anthropol 138: 318332.
  • Grzimek B. 1990. Grzimek's Encyclopedia of Mammals. New York: McGraw-Hill.
  • Günther MM. 1989. Funktionsmorphologische Untersuchungen zum Sprungverhalten an mehreren Halbaffenarten, Ph.D. thesis. Free University of Berlin. p 1183.
  • Hall KRL. 1962. Numerical data, maintenance activities and locomotion of the wild Chamca Baboon, Papio ursinus. Proc Zool Soc Lond 139: 181220.
  • Hanna JB, Polk JD, Schmitt D. 2006. Forelimb and hindlimb forces in walking and galloping primates. Am J Phys Anthropol 130: 529535.
  • Harrigan TP, Jasty M, Mann RW, Harris WH. 1988. Limitations of the continuum assumption in cancellous bone. J Biomech 21: 269275.
  • Harrigan TP, Mann RW. 1984. Characterization of microstructural anisotropy in orthotropic materials using a 2nd rank tensor. J Mater Sci 19: 761767.
  • Hildebrand T, Ruegsegger P. 1997a. A new method for the model-independent assessment of thickness in three-dimensional images. J Microsc 185: 6775.
  • Hildebrand T, Ruegsegger P. 1997b. Quantification of bone microarchitecture with the structure model index. Comput Meth Biomech Biomed Engin 1: 1523.
  • Homminga J, van Rietbergen B, Lochmuller EM, Weinans H, Eckstein F, Huiskes R. 2004. The osteoporotic vertebral structure is well adapted to the loads of daily life, but not to infrequent “error” loads. Bone 34: 510516.
  • Huiskes R. 2000. If bone is the answer, then what is the question? J Anat 197: 145156.
  • Huiskes R, Ruimerman R, van Lenthe GH, Janssen JD. 2000. Effects of mechanical forces on maintenance and adaptation of form in trabecular bone. Nature 405: 704706.
  • Lorensen WE, Cline HE. 1987. Marching cubes: a high resolution 3D surface construction algorithm. Proceedings of the 14th Annual Conference on Computer Graphics and Interactive Techniques. SIGGRAPH'87, Anaheim, July 27–31. p 163169.
  • MacLatchy L, Muller R. 2002. A comparison of the femoral head and neck trabecular architecture of Galago and Perodicticus using micro-computed tomography (μCT). J Hum Evol 43: 89105.
  • Marwan N, Kurths J, Thomsen JS, Felsenberg D, Saparin P. 2009. Three-dimensional quantification of structures in trabecular bone using measures of complexity. Phys Rev E 79: 021903-1021903-11.
  • Muller R. 2009. Hierarchical microimaging of bone structure and function. Nat Rev Rheumatol 5: 373381.
  • Muller R, Ruegsegger P. 1996. Micro-tomographic imaging for the nondestructive evaluation of trabecular bone architecture. In: LowetG, RuegseggerP, WeinansH, MeunierA, editors. 10th Conference of the European-Society-of-Biomechanics. Louvain, Belgium: IOS Press. p 6179.
  • Napier JR. 1976. Primate locomotion. Oxf Biol Readers 41: 316.
  • Nikolei J. 2002. Lokomotionsökologie adulter Hanuman Languren (Semnopithecus entellus) in einem saisonalen Waldhabitat in Ramnagar, Südnepal. PhD thesis, Free University of Berlin. p 1297.
  • Odgaard A. 1997. Three-dimensional methods for quantification of cancellous bone architecture. Bone 20: 315328.
  • Parfitt AM, Travers R, Rauch F, Glorieux FH. 2000. Structural and cellular changes during bone growth in healthy children. Bone 27: 487494.
  • Preuschoft H. 1990. Lesser Apes or Gibbons. In: GrzimekB, editor. Grzimek's encyclopedia of mammals. New York: McGraw-Hill. p 328356.
  • Robling AG, Castillo AB, Turner CH. 2006. Biomechanical and molecular regulation of bone remodeling. Annu Revi Biomed Eng 8: 455498.
  • Roschger P, Grabner BM, Rinnerthaler S, Tesch W, Kneissel M, Berzlanovich A, Klaushofer K, Fratzl P. 2001. Structural development of the mineralized tissue in the human L4 vertebral body. J Struct Biol 136: 126136.
  • Ryan TM, Ketcham RA. 2002. Femoral head trabecular bone structure in two omomyid primates. J Hum Evol 43: 241263.
  • Ryan TM, Ketcham RA. 2005. Angular orientation of trabecular bone in the femoral head and its relationship to hip joint loads in leaping primates. J Morphol 265: 249263.
  • Ryan TM, van Rietbergen B. 2005. Mechanical significance of femoral head trabecular bone structure in Loris and Galago evaluated using micromechanical finite element models. Am J Phys Anthropol 126: 8296.
  • Ryan TM, Krovitz GE. 2006. Trabecular bone ontogeny in the human proximal femur. J Hum Evol 51: 591602.
  • Ryan TM, Walker A. 2010. Trabecular bone structure in the humeral and femoral heads of anthropoid primates. Anat Rec 293: 719729.
  • Saparin P, Thomsen JS, Kurths J, Beller G, Gowin W. 2006. Segmentation of bone CT images and assessment of bone structure using measures of complexity. Med Phys 33: 38573873.
  • Scherf H. 2007. Locomotion-related femoral trabecular architectures in Primates. Ph.D. thesis, Darmstadt University of Technology. p 1203.
  • Scherf H. 2008. Locomotion-related femoral trabecular architectures in primates—high resolution computed tomographies and their implications for estimations of locomotor preferences of fossil primates. In: EndoH, FreyR, editors. Anatomical Imaging. Tokyo: Springer. p 3959.
  • Shen V, Liang XG, Birchman R, Wu DD, Healy D, Lindsay R, Dempster DW. 1997. Short term immobilization-induced cancellous bone loss is limited to regions undergoing high turnover and/or modeling in mature rats. Bone 21: 7178.
  • Simmons CA, Hipp JA. 1997. Method-based differences in the automated analysis of the three-dimensional morphology of trabecular bone. J Bone Miner Res 12: 942947.
  • Skedros JG, Baucom SL. 2007. Mathematical analysis of trabecular 'trajectories in apparent trajectorial structures: The unfortunate historical emphasis on the human proximal femur. J Theor Biol 244: 1545.
  • Stauber M, Muller R. 2006. Volumetric spatial decomposition of trabecular bone into rods and plates—a new method for local bone morphometry. Bone 38: 475484.
  • Tanck E, Homminga J, van Lenthe GH, Huiskes R. 2001. Increae in bone volume fraction preceedes architectural adaptation in growing bone. Bone 28: 650654.
  • van Hooff, JARAM. 1990. Macaques and Allies. In: ParkerSP, editor. Grzimek's Encyclopedia of Mammals. Vol. 2. New York: McGraw-Hill. p 208285.
  • Van Rietbergen B, Huiskes R, Eckstein F, Ruegsegger P. 2003. Trabecular bone tissue strains in the healthy and osteoporotic human femur. J Bone Miner Res 18: 17811788.
  • Vereecke E. 2006. The functional morphology and bipedal locomotion of Hylobates lar and its implications for the evolution of hominin bipedalism. PhD thesis, University Antwerpen. p 1308.
  • Welker C, Schäfer-Witt C. 1990. New World monkeys. In: GrzimekB, editor. Grzimek's Encyclopedia of Mammals. New York: McGraw-Hill. p 122177.