Recent origin of low trabecular bone density in modern humans

Edited by Erik Trinkaus, Washington University, St. Louis, MO, and approved November 26, 2014 (received for review June 23, 2014)
December 22, 2014
112 (2) 366-371
Research Article
Gracility of the modern Homo sapiens skeleton is the result of decreased biomechanical loading
Timothy M. Ryan, Colin N. Shaw


The human skeleton is unique in having low trabecular density representing a lightly built human body form. However, it remains unknown when during human evolution this unique characteristic first appeared. To our knowledge, this study is the first to examine trabecular bone density throughout the skeleton of fossil hominins spanning several million years. The results show that trabecular density remained high throughout human evolution until it decreased significantly in recent modern humans, suggesting a possible link between changes in our skeleton and increased sedentism.


Humans are unique, compared with our closest living relatives (chimpanzees) and early fossil hominins, in having an enlarged body size and lower limb joint surfaces in combination with a relatively gracile skeleton (i.e., lower bone mass for our body size). Some analyses have observed that in at least a few anatomical regions modern humans today appear to have relatively low trabecular density, but little is known about how that density varies throughout the human skeleton and across species or how and when the present trabecular patterns emerged over the course of human evolution. Here, we test the hypotheses that (i) recent modern humans have low trabecular density throughout the upper and lower limbs compared with other primate taxa and (ii) the reduction in trabecular density first occurred in early Homo erectus, consistent with the shift toward a modern human locomotor anatomy, or more recently in concert with diaphyseal gracilization in Holocene humans. We used peripheral quantitative CT and microtomography to measure trabecular bone of limb epiphyses (long bone articular ends) in modern humans and chimpanzees and in fossil hominins attributed to Australopithecus africanus, Paranthropus robustus/early Homo from Swartkrans, Homo neanderthalensis, and early Homo sapiens. Results show that only recent modern humans have low trabecular density throughout the limb joints. Extinct hominins, including pre-Holocene Homo sapiens, retain the high levels seen in nonhuman primates. Thus, the low trabecular density of the recent modern human skeleton evolved late in our evolutionary history, potentially resulting from increased sedentism and reliance on technological and cultural innovations.
Obligate bipedalism—a defining feature of humans that distinguishes us from our closest living relatives, the African apes—has transformed the human skeleton. Among these unique features are long lower limbs with large joint surfaces. These large joint surfaces help distribute loads over a larger surface area and thus are better at resisting the high forces incurred during locomotion on two limbs instead of four (15). Early African Homo erectus at 1.8–1.5 Ma had enlarged lower limb joint surfaces (1, 3) and a larger stature (6) and body mass (7, 8) than many earlier hominins, and this pattern often is considered to reflect the emergence of a more modern human-like body plan (1, 3, 5, 6, 9; but also see ref. 7).
Recent modern human (Holocene Homo sapiens) skeletons also appear to be gracile as compared with earlier hominins (1014). Here, “gracilization” refers to the reduction in strength and bone mass relative to body mass inferred from osseous tissue and overall bone size and has been studied mainly using diaphyseal cortical bone cross-sections (1016). Although the relationship between mechanical loading during life and bone strength is likely to be complex (17), there is much evidence that increased mechanical loading leads to increases in relative bone strength (18). Thus, diaphyseal skeletal gracilization in recent modern humans relative to earlier hominins generally has been attributed to a decrease in daily physical activity via technological and cultural innovations (6, 10, 1315, 1922).
There also is evidence that increased activity level and mechanical loading increases trabecular bone mineral density within limb bones (ref. 23 and references therein). However, although there currently is extensive literature on the variation and evolution of long bone shaft strength in humans and fossil hominins (10, 15, 16, 2427), there has been comparatively less research on trabecular bone (2830) because of the technical challenges in quantifying its structure: limited access to high-resolution CT (microCT), problems with preservation and/or imaging of fine trabecular structures, particularly in fossils, and intensive processing time. A few studies examining individual limb elements have reported low trabecular density, as measured by volumetric density (the trabecular bone fraction, TBF, or bone volume relative to total volume), in recent modern human epiphyses. The recent modern human arm (humerus) and hand (metacarpals) have low TBF (31, 32) and mineral density (33) compared with chimpanzees and orangutans. This finding might be expected, because humans rarely use their upper limbs for locomotion and therefore do not habitually expose their upper limb bones to the high loads of body-weight support. Indeed, recent modern human upper limb bones have relatively low diaphyseal strengths compared with the lower limbs (34). However, recent modern humans also have low TBF in the calcaneus (35) and metatarsals (36) compared with great apes, despite the increased proportionate loading and full body-weight support incurred during bipedal locomotion.
To our knowledge, this study is the first to examine how trabecular density varies throughout the human appendicular skeleton, how that variation compares with other primates, and how trabecular density evolved in the hominin lineage. We test the hypotheses that (i) recent modern humans have lower TBF throughout the upper and lower limbs compared with that of other primates and (ii) the reduction in TBF first occurred in African Homo erectus, consistent with the shift toward a modern human locomotor anatomy, or more recently in concert with diaphyseal gracilization in recent modern humans. This study is the first, to our knowledge, to evaluate TBF in upper and lower limb joints in fossil hominins from late Pliocene Australopithecus to recent Homo.
To assess whether low trabecular density is a systemic phenomenon throughout the human skeleton, we examined trabecular density in seven epiphyseal elements throughout the upper limb (humeral head, proximal ulna, distal radius, metacarpal heads) and lower limb (femoral head, distal tibia, and metatarsal heads) (Table 1 and Tables S1 and S2). We measured trabecular density in a 2D image as the ratio of bone pixels/total pixels (i.e., the TBF) within a defined region of interest (ROI) for each epiphysis (Fig. S1). We first compared TBF across extant primate (baboon, orangutan, chimpanzee, and recent modern human) limb epiphyses (Table 1 and Table S2). We also compared TBF in late Pliocene and Pleistocene hominins (n = 42) within the context of changes in body form in early Homo at 1.8 Ma and throughout the Pleistocene (Table 1 and Table S2).
Table 1.
Sample size of taxa studied
TaxonProximal humerusProximal ulnaDistal radiusDistal metacarpalProximal femurDistal tibiaDistal metatarsal
Homo sapiens38383830383835
Pan troglodytes17171717171717
Pongo pygmaeus15161616161616
Papio anubis17171712181817
Australopithecus sp.1
Australopithecus africanus14533
Paranthropus robustus/early Homo122
Homo neanderthalensis32211
Early Homo sapiens112222
See Table S1 for a breakdown of each sex included in the extant samples and Table S2 for a complete list of the specimens included in each fossil taxon.


Recent Modern Human Trabecular Density.

Results show that the TBF is lower in recent modern humans than in the other primate taxa examined, including in lower limb joints that might be expected to have greater TBF in committed bipeds who bear comparatively more body weight through these limbs (Figs. 1 and 2). ANOVA with Tukey post hoc pairwise comparisons shows that recent modern human upper and lower limb joints have significantly (P < 0.001) lower TBF than all other primates (Fig. 2 and Table S3), albeit with some variation in values across elements (Fig. 3). Pooled data for all upper limb elements for each taxon indicate that baboons have the greatest TBF, and recent modern humans exhibit the least (Fig. 2A). Although there is some overlap in TBF between recent modern humans and chimpanzees, humans have systematically lower TBF in every anatomical element sampled (Fig. 3A and Table 2). Within lower limbs, the pooled results again reveal that the TBF is significantly (P < 0.001) lower in recent modern human epiphyses than in all other primate taxa (Fig. 2B). Relative to the upper limb, the TBF values of individual elements in the lower limb show an even greater separation between recent modern humans and chimpanzees (Fig. 3B).
Fig. 1.
Proximal femur and distal metacarpal 2D slices of trabecular structure and their associated trabecular fraction values in chimpanzees, recent modern humans, and a sample of fossil taxa demonstrating variation in trabecular fraction across taxa and time.
Fig. 2.
TBF in upper and lower limb epiphyses in extant taxa. The bar graphs show mean TBF with a 2-SD error bar for all upper limb (A) and lower limb (B) epiphyses. The asterisk indicates that the trabecular fraction is significantly (P = 0.001) lower in recent modern humans than in all other taxa.
Fig. 3.
TBF in each epiphysis across extant and fossil samples. In the box-and-whisker plots the horizontal line shows the median, the box defines the 25th and 75th percentiles, and the whiskers show the range of TBF variation in all upper limb (A) and lower limb (B) epiphyses. Asterisks indicate that recent modern human values are significantly lower than those of chimpanzees in each epiphyseal element. The Swartkrans sample represents either P. robustus or early Homo. MC, metacarpal; MT, metatarsal.
Table 2.
Mean and SD for each extant and fossil taxon and the number of SDs each fossil sample deviates from the recent human and chimpanzee means, which were treated as reference samples
Element and taxonMean trabecular fractionSDMean SDs from recent Homo sapiensMean SDs from Pan troglodytesRange
Proximal humerus
 Recent Homo sapiens0.200.020.16–0.26
Pan troglodytes0.300.030.25–0.37
Australopithecus africanus0.469.444.79 
Homo neanderthalensis0.272.48−0.950.24–0.29
 Early Homo sapiens0.262.10−1.26 
Proximal ulna
 Recent Homo sapiens0.250.050.16–0.39
Pan troglodytes0.330.050.16–0.29
Australopithecus africanus0.474.042.910.40–0.57
Homo neanderthalensis0.484.092.97 
 Early Homo sapiens0.331.37−0.11 
Distal radius
 Recent Homo sapiens0.210.030.16–0.27
Pan troglodytes0.310.030.23–0.39
Homo neanderthalensis0.385.301.78 
Proximal metacarpal
 Recent Homo sapiens0.230.030.16–0.33
Pan troglodytes0.320.030.25–0.38
Australopithecus africanus0.311.81−0.510.28–0.33
Paranthropus robustus/early Homo0.291.47−0.91 
Homo neanderthalensis0.250.29−2.27 
 Early Homo sapiens0.311.92−0.38 
Proximal femur
 Recent Homo sapiens0.230.020.19–0.29
Pan troglodytes0.410.020.35–0.46
Australopithecus sp.0.447.380.79 
Australopithecus africanus0.468.111.480.36–0.52
Homo neanderthalensis0.405.88−0.59 
 Early Homo sapiens0.375.10−1.31 
Distal tibia
 Recent Homo sapiens0.230.040.16–0.32
Pan troglodytes0.310.010.28–0.35
Australopithecus africanus0.445.036.330.37–0.52
 Early Homo sapiens0.291.51−1.10 
Distal metatarsal
 Recent Homo sapiens0.210.030.16–0.30
Pan troglodytes0.340.040.26–0.37
Paranthropus robustus/early Homo0.363.850.48 
 Early Homo sapiens0.466.683.76 
Positive and negative values refer to the number of SDs above and below the reference sample means, respectively. Ranges are provided only for those samples with n > 3.

Fossil Hominin TBF.

In the upper limb, the mean TBF of the humeral head in all pre-Holocene fossil hominins falls more than 2 SDs above the recent modern human mean (Fig. 3A and Table 2). In particular, the Australopithecus africanus humeral head has the highest TBF values of the taxa examined, having a substantially higher TBF than chimpanzee humeri and far greater (9 SDs) TBF values than the humeri of recent modern humans. Neanderthal and early modern human humeral TBF values are intermediate between those of chimpanzees and recent modern humans (Table 2).
The proximal ulna TBF is high in fossil hominins. A. africanus and Neanderthal mean TBF values are 4 SDs above the recent modern human mean. The early modern (i.e., pre-Holocene) human proximal ulna TBF is above the recent modern human mean (1.37 SDs) (Table 2), although it is within the overall range in recent modern humans (Fig. 3). Recent modern humans have a large range of variation, and overall the ulna has the highest TBF of all of the joints in the upper limb (Fig. 3A and Table 2). Interestingly, both A. africanus and Neanderthals have very a high ulnar TBF, several SDs above the chimpanzee mean (Table 2).
The mean trabecular density for all fossil hominin distal radii, including early H. sapiens, is above the recent modern human mean. Means of Neanderthal and Swartkrans specimens (possibly P. robustus or early Homo) are more than 5 and 2 SDs above the recent modern human mean, respectively, and are comparable to or higher than the mean of chimpanzees (Fig. 3A and Table 2). Apart from the Neanderthals, all fossil hominin TBF means for the metacarpal heads are 1 or more SDs above the recent modern human mean (Fig. 3A and Table 2). All fossil hominin metacarpal head means fall below the chimpanzee mean. Neanderthal metacarpals stand out in having a low TBF, comparable to that of recent modern humans.
In summary, in terms of TBF, the upper limbs of recent modern humans are lightly built compared with those of pre-Holocene hominins and extant nonhuman primates. A one-way ANOVA between pooled samples of Australopithecus and Swartkans upper limb elements vs. fossil Homo (including early modern humans) upper limb elements was not significant (P = 0.17), but the TBF of recent modern humans was significantly lower than that of fossil Homo considered separately (P < 0.01) and of all fossil hominins combined (P < 0.01) (Table S3).
In the lower limb, the femoral head TBF of all of the pre-Holocene fossil taxa falls above the observed range for recent modern humans, with fossil group means falling roughly 5 or more SDs above the recent modern human mean (Fig. 3B and Table 2). In particular, A. africanus femora and the fossil KNM-ER 738 (likely Paranthropus boisei, but attribution is uncertain) have TBF values lying more than 7 SDs above the recent modern human mean. In contrast, TBF values of pre-Holocene hominin femora overlap widely with those of chimpanzees, with all group means being within 1 SD of the chimpanzee mean.
Data on the distal tibia show that the mean TBF of A. africanus is much higher than that of chimpanzees and is 5 SDs above the mean in recent modern humans (Fig. 3B and Table 2). The tibial TBF in early modern humans is intermediate between that of chimpanzees and recent modern humans (Fig. 3B).
Results for the metatarsal head indicate that the mean TBF in the two fossil groups—Swartkrans and early modern humans—is more than 3 SDs above the recent modern human mean. Indeed, early modern humans exhibit the greatest mean value across the sample (Fig. 3B and Table 2), falling more than 3 SDs above the chimpanzee mean.
In summary, the lower limb pattern shows that recent modern humans have a lower mean TBF than all pre-Holocene hominins, including early modern humans, as well as chimpanzees. Compared with the results for mean upper limb TBF, the lower limbs of recent modern humans have consistently lower TBF values, exhibiting less overlap with any fossil or extant nonhuman sample. A one-way ANOVA between pooled samples of Australopithecus and Swartkrans lower limb elements vs. fossil Homo (including early modern humans) lower limb elements was not significantly (P = 0.18) different, but the mean TBF in the lower limb elements was significantly lower in recent modern humans than in fossil Homo (P < 0.01) and was lower than in all fossil hominins combined (P < 0.01) (Table S3).


Our results show that recent modern humans have low trabecular density, as assessed by TBF, throughout the limb skeleton; the TBF is about 50–75% of that of modern chimpanzees, and some fossil hominins have limb joints more than twice as dense as those of modern humans (Fig. 3). This decrease in trabecular density is a very recent evolutionary phenomenon. Hominins as late as the Late Pleistocene retain high trabecular density (i.e., TBF), suggesting that extant levels of human trabecular density likely emerged in the latest Pleistocene or Holocene. These results support recent studies showing little change in lower limb bone shaft robusticity among Homo throughout the Pleistocene, followed by a rapid decrease in recent modern humans in the Holocene (e.g., refs. 11 and 16). Across the sample, the mean trabecular density in all postcranial elements is lower in recent modern humans than in other extant and fossil taxa. In general, Australopithecus specimens displayed the greatest trabecular density in the lower limbs, exceeding that in chimpanzees. There is some overlap in the range of trabecular density between early modern humans and Neanderthals, especially in the upper limb. In recent modern humans the range of trabecular density also is closer to that of chimpanzees in the upper limb elements than in the lower limbs (Fig. 3B). Thus, although recent modern humans have significantly lower trabecular density in their upper limbs, both the general temporal trend across hominins and the distinctiveness of the low trabecular density in recent modern humans are more marked in the lower limb.
The hypothesis that lower limb trabecular density decreased in early Homo in conjunction with the shift toward a more modern human body plan is not supported. We show that both Neanderthals and early modern humans retain high trabecular density similar to that in earlier hominins, as compared with the low trabecular density found in recent modern humans. Thus, there is not a consistent, gradual shift toward lighter skeletons during the Pleistocene. On the contrary, our results suggest that there are generally high values, with some variation in skeletal robusticity, throughout the Pleistocene, as also has been observed in shaft strength of limb bones (11, 16, 27).
Similarly, the upper limb elements do not support the hypothesis of a temporal decrease in trabecular density throughout the Pleistocene. Instead, Early Pleistocene and Late Pleistocene taxa all have high trabecular density, supporting the idea that the shift from robust to gracile limbs, even in hands that are used for manipulation rather than for weight-bearing, occurred in the Holocene. Interestingly, the highest TBF values in the upper limb are observed in A. africanus, which had the highest lower limb values as well. Samples currently are too small to determine whether Australopithecus generally had denser trabecular bone than later hominins, chimpanzees, and other primates, but this observation deserves further attention.
The observation that low trabecular bone density (i.e., TBF) among recent modern humans is more marked in the lower limbs than in the upper limbs may argue for changes in mobility, i.e., increased sedentism, as the primary behavioral event driving skeletal gracilization in the Holocene (11, 15, 21, 37, 38). In addition, some behaviors characteristic of Holocene populations, including nonmechanized agriculture, may involve relatively heavy loading of the upper limbs (e.g., ref. 39) and thus serve to maintain relatively higher trabecular density in those skeletal elements.
Our hypothesis that trabecular density decreases at the same time as diaphyseal gracilization of the skeleton, in conjunction with increasing sedentism in the Holocene, is supported. In many ways, early anatomically modern (Pleistocene) humans were morphologically similar to more recent (Holocene) humans, with more linear bodies, including relatively narrow trunks and reduced mediolateral femoral buttressing (16, 40). Therefore, it is surprising that early modern humans do not have low trabecular density, as do recent humans, but instead have greater TBF than some of the Early Pleistocene taxa and chimpanzees. Again, these results support the hypothesis that trabecular reorganization occurred only later in human evolutionary history.
The results here imply that the higher trabecular density (i.e., TBF) throughout the skeleton of early modern humans compared with that of recent modern humans may be attributable to greater mechanical loading related to longer travel distances or running in the former. Arguably, the post-Pleistocene reduction in the trabecular density of the upper limb could result, at least in part, from a systemic effect (perhaps metabolic and/or pleiotropic) (41). Therefore it could have been driven indirectly by a decrease in mobility rather than by any specific reduction in upper limb bone loading, e.g., an increase in the use of more sophisticated tools.
A limitation of this study is that it did not assess other trabecular bone architectural parameters, such as trabecular strut anisotropy, thickness, and spacing, which can have significant effects on elastic modulus and strength (42, 43). Anisotropy in particular appears to be related to principal loading direction (42). Work to date suggests that in some anatomical regions, such as metatarsal heads, humans are characterized by higher degrees of anisotropy (36), but in others, such as metacarpal heads, humans and apes have comparable degrees of anisotropy (32). In both these cases, modern human joints are consistently low in TBF (32, 36). Therefore, within specific articulations, it is not clear to what extent parameters such as trabecular anisotropy, thickness, and spacing vary systematically among closely related taxa (32, 36, 44, 45). Incorporating additional structural properties across the same or different anatomical regions in future studies could help determine whether architectural changes (if any) paralleled the changes in trabecular density during hominin evolution and whether they were complementary or compensatory during the transition to recent modern humans.
Also, although mechanical loading can affect trabecular growth (23, 46, 47), direct developmental influences may not account entirely for the marked differences in trabecular density in recent modern humans and pre-Holocene hominins. It is plausible that, as sedentism became prevalent and food sources changed in the Holocene, selection pressure for robust skeletons diminished. In other words, more gracile bodies may have evolved rather than be a product of decreased mechanical stimulus during development. The lack of a significant difference in trabecular density between the (presumably) more active Puye and the industrial-era Terry samples also may argue for longer-term evolutionary as well as direct environmental effects. Moreover, other nonmechanical systemic factors that may influence skeletal morphology, including nutrition and disease prevalence (48, 49), also were affected by the transition to food production and increased sedentism in the Holocene (50, 51).
However, regardless of the proximate mechanisms involved, the reduction of trabecular density observed here among nonforaging Holocene humans relative to all earlier hominins examined and to nonhuman primates is consistent with a reduction in activity level and, in particular, mobility, among very recent human populations.

Materials and Methods

Trabecular density was examined using TBF measured from systematically imaged 2D cross-sectional slices of seven epiphyses throughout the appendicular skeleton: the femoral head, distal tibia, metatarsal head, humeral head, proximal ulna, distal radius, and metacarpal head (Table 1 and Tables S1 and S2).


Extant primates.

The comparative sample comprises recent modern humans (Homo sapiens), chimpanzees (Pan troglodytes), orangutans (Pongo pygmaeus), and olive baboons (Papio anubis) (Table 1 and Table S1). The samples were obtained from the National Museum of Natural History (NMNH), Smithsonian Institution. The nonhuman primates were wild-collected individuals. The recent human sample comprises individuals from the Terry and Puye collections (see SI Materials and Methods and Table S4 for more details). Only elements that did not exhibit any pathological signs were selected. Moreover, only adult individuals were selected, based on long bone epiphyseal fusion of all elements and the known ages of the Terry collection. Regarding the Terry collection, specific care was taken to sample individuals under the age of 45 y, mainly to avoid age-related effects (i.e., osteoporosis and osteopenia).

Fossil hominins.

The fossil sample included Australopithecus africanus, KNM-ER 738 (a femur likely representing Paranthropus boisei, but of an uncertain attribution), Swartkrans hominins attributed to either Paranthropus robustus or early Homo, Homo neanderthalensis, and early Homo sapiens, using all available postcranial elements from various geographic locations and time periods. However, not all postcranial elements were preserved for each fossil taxon (Table 1 and Table S2; see SI Materials and Methods for more details).

CT Data Acquisition.

Fossil and extant specimens were CT scanned using two methods, peripheral quantitative CT (pQCT) and microtomography (microCT), because specimen access prevented us from scanning the entire sample with one method. Consequently, to obtain a sufficiently large sample of fossil and extant individuals, trabecular bone was measured in two different ways, and a subsample was measured with both methods. The relationship between the two methods is discussed below.

pQCT data acquisition.

2D slices of extant H. sapiens, P. troglodytes, P. pygmaeus, and P. anubis (Table 1 and Table S1) were collected using a pQCT scanner (Stratec Research SA) at the NMNH, Smithsonian Institution. The pQCT scanner measures bone content within a given region; it reports bone mineral density based on the attenuation of X-rays. The raw data represent linear attenuation coefficients, which the pQCT calibrates into hydroxyapatite-equivalent densities, based on a calibration standard (52), and which are reported as density values in milligrams per cubic centimeter. All specimens were scanned at a resolution of 100 μm. Each anatomical element was positioned in a systematic orientation in the pQCT gantry to ensure homologous scanning of the same position and plane through the center of the joint in each specimen (Fig. S2 and see Fig. S4; also see SI Materials and Methods).

microCT data acquisition.

Like pQCT, microCT imaging relies on the attenuation of X-rays, but the images obtained are reconstructed in 3D. The 2D slices of fossil hominin specimens were derived from high-resolution microCT scans taken in the Department of Human Evolution at the Max Planck Institute for Evolutionary Anthropology (MPI-EVA), Leipzig, Germany, with permission from the respective institutions curating the fossil material (Table S2). All fossil epiphyses were scanned at ∼30 μm using a high-resolution, industrial BIR ACTIS 225/300 microCT scanner.
Four of the fossil specimens studied (Table S2) were affected by the presence of matrix, and therefore segmentation was used to eliminate the matrix. Segmentation identifies the boundary between bone and nonbone and eliminates foreign material that may be present because of the fossilization process. This study used the Ray-casting algorithm method (53) of segmentation because of its reliance on the gray value gradient rather than on absolute gray values. This method was applied to 3D microCT scans with matrix inclusions only (i.e., 4 of 42 samples), and thresholding values were image specific, based on the relative density of the included tissues, to ensure the best possible segmentation (Fig. S3).
To obtain TBFs from the fossil microCT data, 2D slices of each epiphysis were taken using Avizo 6.1 (Mercury Systems). To do so, the complete 3D TIF stack was reconstructed, and a single slice of trabecular bone (image) was taken from the same anatomical location used with the pQCT scanner protocol (see above and Figs. S2 and S4).

TBF Quantification.

pQCT trabecular quantification.

Upon scanning a single bone, an ROI initially was outlined around the entire image, including cortical bone as visualized in the scanner. Then, a Peel mode was used to eliminate concentric rows of voxels from the periosteal surface inward using a user-defined percentage of bone. The peel percentage was set between 60% and 70% in the femoral and humeral heads and between 40% and 50% in the other epiphyseal elements because the amount of trabecular bone vs. cortical bone varies depending on the size and shape of the bone. As noted above, the pQCT uses a calibration standard to convert the raw linear attenuation coefficients into hydroxyapatite-equivalent densities (expressed in milligrams per cubic centimeter).

microCT trabecular quantification.

The single 2D images of fossils were each converted into an eight-bit TIF file and were imported into Image J 1.48s (National Institutes of Health). Trabecular bone quantification was performed using the bone-specific plugin Bone J (54). This plugin quantifies TBF as a ratio of mineralized bone area to total area within an ROI. A circular ROI was centered on the center of the epiphysis (Fig. S1). The size of the ROI was standardized by the size of the epiphysis, and the ROI was positioned centrally within the epiphysis to avoid cortical bone. The ROI diameter was defined in the same manner used in the pQCT protocol: 50% of the mediolateral epiphyseal diameter in metacarpal and metatarsal heads, 40% of the mediolateral epiphyseal breadth in the distal tibia, distal radius, and ulnar head, and 60–70% of the epiphyseal diameter in femoral and humeral heads.

Comparison of Scanning Methods.

To compare the microCT fossil specimen images directly with those of the pQCT-scanned extant sample, a subset (n = 25) of the sample scanned with the pQCT (10 metacarpals, 5 humeri, 5 femora, and 5 metatarsals of chimpanzee skeletons housed at the NMNH) was scanned on the MPI-EVA BIR ACTIS 225/300 microCT scanner using a similar scanning resolution (∼30 μm). The size of the ROI was standardized, as was done for fossil hominin specimens, to test for any bias created by using different scanning techniques. Linear regression between the TBF values obtained using microCT and trabecular density using pQCT showed a highly significant correlation (TBF = 0.0006(density) + 0.1567, r = 0.81, P < 0.001; Fig. S5). This equation was used to predict the microCT-equivalent TBF of extant taxa from their pQCT values (see SI Materials and Methods for additional information). A comparison of the 25 chimpanzee specimens scanned with both methods demonstrated that they yielded comparable results (Fig. S6).

Statistical Analyses.

Because of the small within-taxon sample sizes for fossils, we used the extant human and chimpanzee means as reference samples to which TBF means, or the TBF value in the case of n = 1 for some fossil samples, were compared to show how many SDs a value or sample mean fell from the reference sample. One-way ANOVA with Tukey post-hoc correction was used to identify significant group differences among the extant primates and between pooled fossil Homo and Australopithecus samples (see SI Materials and Methods for additional statistical analyses).


We thank Z. Tsegai and D. Plotzki for help with microCT scanning and the curators at the following institutions for providing access to samples in their care: Tel Aviv University, Musee du Perigord Perigueux, Reihnisches Landesmuseum Bonn, University of the Witwatersrand, National Museums of Kenya, Distong National Museum of Natural History, and the National Museum of Natural History, Smithsonian Institution. MicroCT scans of Sterkfontein fossil material were produced through a collaborative project between the University of the Witwatersrand and the Department of Human Evolution, Max Planck Institute for Evolutionary Anthropology. This research was supported by the Wenner–Gren Foundation for Anthropological Research Wadsworth Fellowship (to H.C.), the Leakey Foundation Baldwin Fellowship (to H.C.), Smithsonian’s Peter Buck Postdoctoral Fellowship (to H.C.), and National Science Foundation Grants BCS-0521835 and DGE-0801634 (to B.G.R.).

Supporting Information

Supporting Information (PDF)
Supporting Information


CB Ruff, Hindlimb articular surface allometry in Hominoidea and Macaca, with comparisons to diaphyseal scaling. J Hum Evol 17, 687–714 (1988).
CB Ruff, Long bone articular and diaphyseal structure in old world monkeys and apes. I: Locomotor effects. Am J Phys Anthropol 119, 305–342 (2002).
WL Jungers, Relative joint size and hominoid locomotor adaptations with implications for the evolution of hominid bipedalism. J Hum Evol 17, 247–265 (1988).
WL Jungers, Scaling of postcranial joint size in hominoid primates. Gravity, Posture and Locomotion in Primates, eds FK Jouffroy, MH Stack, C Niemitz (Il Sedicesimo, Florence), pp. 87–95 (1990).
DM Bramble, DE Lieberman, Endurance running and the evolution of Homo. Nature 432, 345–352 (2004).
CB Ruff, A Walker, Body size and body shape. The Nariokotome Homo erectus Skeleton, eds A Walker, R Leakey (Harvard Univ Press, Cambridge, MA), pp. 234–265 (1993).
H Pontzer, Ecological energetics in early Homo. Curr Anthropol 53, S346–S358 (2012).
C Ruff, Body size and body shape in early hominins - implications of the Gona pelvis. J Hum Evol 58, 166–178 (2010).
BG Richmond, LC Aiello, BA Wood, Early hominin limb proportions. J Hum Evol 43, 529–548 (2002).
CB Ruff, E Trinkaus, A Walker, CS Larsen, Postcranial robusticity in Homo. I: Temporal trends and mechanical interpretation. Am J Phys Anthropol 91, 21–53 (1993).
BM Holt, et al., Temporal and geographic variation in robusticity. Skeletal Variation and Adaptation in Europeans: Upper Paleolithic to the Twentieth Century, ed CB Ruff (Wiley-Blackwell, New York, pp xx.
CB Ruff, Evolution of the hominid hip. Primate Locomotion: Recent Advances, eds E Strasser, A Rosenberger, H McHenry, J Fleagle (Plenum, Davis, CA), pp. 449–469 (1998).
D Marchi, V Sparacello, C Shaw, Mobility and lower limb robusticity of a pastoralist Neolithic population from north-western Italy. Human Bioarchaeology of the Transition to Agriculture, eds R Pinhasi, JT Stock (Wiley-Blackwell, New York), pp. 317–346 (2011).
CB Ruff, et al., Body size, body proportions, and mobility in the Tyrolean “Iceman”. J Hum Evol 51, 91–101 (2006).
BM Holt, Mobility in Upper Paleolithic and Mesolithic Europe: Evidence from the lower limb. Am J Phys Anthropol 122, 200–215 (2003).
E Trinkaus, CB Ruff, Femoral and tibial diaphyseal cross-sectional geometry in Pleistocene Homo. PaleoAnthropology 2012, 13–62 (2012).
LW Cowgill, The ontogeny of Holocene and Late Pleistocene human postcranial strength. Am J Phys Anthropol 141, 16–37 (2010).
C Ruff, B Holt, E Trinkaus, Who’s afraid of the big bad Wolff?: “Wolff’s law” and bone functional adaptation. Am J Phys Anthropol 129, 484–498 (2006).
CN Shaw, JT Stock, Extreme mobility in the Late Pleistocene? Comparing limb biomechanics among fossil Homo, varsity athletes and Holocene foragers. J Hum Evol 64, 242–249 (2013).
DW Frayer, Body size, weapon use and natural selection in the European Upper Paleolithic and Mesolithic. Am Anthropol 83, 57–73 (1981).
JT Stock, Hunter-gatherer postcranial robusticity relative to patterns of mobility, climatic adaptation, and selection for tissue economy. Am J Phys Anthropol 131, 194–204 (2006).
AA Macintosh, R Pinhasi, JT Stock, Lower limb skeletal biomechanics track long-term decline in mobility across ∼6150 years of agriculture in central Europe. J Arch Sci 52, 376–390 (2014).
U Pettersson, M Nilsson, V Sundh, D Mellström, M Lorentzon, Physical activity is the strongest predictor of calcaneal peak bone mass in young Swedish men. Osteoporos Int 21, 447–455 (2010).
JT Stock, SK Pfeiffer, Long bone robusticity and subsistence behaviour among Later Stone Age foragers of the forest and fynbos biomes of South Africa. J Arch Sci 31, 999–1013 (2004).
E Trinkaus, SE Churchill, CB Ruff, Postcranial robusticity in Homo. II: Humeral bilateral asymmetry and bone plasticity. Am J Phys Anthropol 93, 1–34 (1994).
CB Ruff, Mechanical determinants of bone form: Insights from skeletal remains. J Musculoskelet Neuronal Interact 5, 202–212 (2005).
LL Schackelford, Regional variation in the postcranial robusticity of late Upper Paleolithic humans. Am J Phys Anthropol 133, 655–68 (2007).
TM Ryan, RA Ketcham, Femoral head trabecular bone structure in two omomyid primates. J Hum Evol 43, 241–263 (2002b).
KA Robson Brown, EN Davies, DS McNally, The angular distribution of vertebral trabeculae in modern humans, chimpanzees and the Kebara 2 Neanderthal. J Hum Evol 43, 189–205 (2002).
NL Griffin, Bone architecture of the hominin second proximal pedal phalanx: A preliminary investigation. J Hum Evol 54, 162–168 (2008).
H Scherf, K Harvati, J-J Hublin, A comparison of proximal humeral cancellous bone of great apes and humans. J Hum Evol 65, 29–38 (2013).
ZJ Tsegai, et al., Trabecular bone structure correlates with hand posture and use in hominoids. PLoS ONE 8, e78781–e78781 (2013).
A Zeininger, BG Richmond, G Hartman, Metacarpal head biomechanics: A comparative backscattered electron image analysis of trabecular bone mineral density in Pan troglodytes, Pongo pygmaeus, and Homo sapiens. J Hum Evol 60, 703–710 (2011).
C Ruff, Femoral/humeral strength in early African Homo erectus. J Hum Evol 54, 383–390 (2008).
M Maga, J Kappelman, TM Ryan, RA Ketcham, Preliminary observations on the calcaneal trabecular microarchitecture of extant large-bodied hominoids. Am J Phys Anthropol 129, 410–417 (2006).
NL Griffin, et al., Comparative forefoot trabecular bone architecture in extant hominids. J Hum Evol 59, 202–213 (2010).
CB Ruff, CS Larsen, Postcranial biomechanical adaptations to subsistence changes on the Georgia Coast. Anthropol Pap Am Mus Nat Hist 68, 94–120 (1990).
CB Ruff, CS Larsen, Reconstructing behavior in Spanish Florida: The biomechanical evidence. Bioarchaeology of Spanish Florida: The Impact of Colonialism, ed CS Larsen (Univ Press of Florida, Gainesville, FL), pp. 113–145 (2001).
PS Bridges, Changes in activities with the shift to agriculture in the southeastern United States. Curr Anthropol 30, 385–394 (1989).
OM Pearson, Postcranial remains and the origin of modern humans. Evol Anthropol 9, 229–247 (2000).
DE Lieberman, How and why humans grow thin skulls: Experimental evidence for systemic cortical robusticity. Am J Phys Anthropol 101, 217–236 (1996).
TM Keaveny, EF Morgan, GL Niebur, OC Yeh, Biomechanics of trabecular bone. Annu Rev Biomed Eng 3, 307–333 (2001).
EF Morgan, HH Bayraktar, TM Keaveny, Trabecular bone modulus-density relationships depend on anatomic site. J Biomech 36, 897–904 (2003).
RJ Fajardo, R Müller, RA Ketcham, M Colbert, Nonhuman anthropoid primate femoral neck trabecular architecture and its relationship to locomotor mode. Anat Rec (Hoboken) 290, 422–436 (2007).
TM Ryan, A Walker, Trabecular bone structure in the humeral and femoral heads of anthropoid primates. Anat Rec (Hoboken) 293, 719–729 (2010).
H Pontzer, et al., Trabecular bone in the bird knee responds with high sensitivity to changes in load orientation. J Exp Biol 209, 57–65 (2006).
MM Barak, DE Lieberman, JJ Hublin, A Wolff in sheep’s clothing: Trabecular bone adaptation in response to changes in joint loading orientation. Bone 49, 1141–1151 (2011).
SM Garn, MA Guzmán, B Wagner, Subperiosteal gain and endosteal loss in protein-calorie malnutrition. Am J Phys Anthropol 30, 153–155 (1969).
SB Eaton, DA Nelson, Calcium in evolutionary perspective. Am J Clin Nutr 54, 281S–287S (1991).
MN Cohen, GJ Armelagos Paleopathology at the origins of agriculture (Academic Press, New York, 1984).
, eds MN Cohen, GMM Crane-Kramer (University Press of Florida, Gainesville Ancient Health: Skeletal indicators of agricultural and economic intensification, 2007).
P Augat, CL Gordon, TF Lang, H Iida, HK Genant, Accuracy of cortical and trabecular bone measurements with peripheral quantitative computed tomography (pQCT). Phys Med Biol 43, 2873–2883 (1998).
H Scherf, R Tilgner, A new high-resolution computed tomography (CT) segmentation method for trabecular bone architectural analysis. Am J Phys Anthropol 140, 39–51 (2009).
M Doube, et al., BoneJ: Free and extensible bone image analysis in ImageJ. Bone 47, 1076–1079 (2010).

Information & Authors


Published in

Go to Proceedings of the National Academy of Sciences
Go to Proceedings of the National Academy of Sciences
Proceedings of the National Academy of Sciences
Vol. 112 | No. 2
January 13, 2015
PubMed: 25535354


Submission history

Published online: December 22, 2014
Published in issue: January 13, 2015


  1. trabecular bone
  2. human evolution
  3. gracilization
  4. Homo sapiens
  5. sedentism


We thank Z. Tsegai and D. Plotzki for help with microCT scanning and the curators at the following institutions for providing access to samples in their care: Tel Aviv University, Musee du Perigord Perigueux, Reihnisches Landesmuseum Bonn, University of the Witwatersrand, National Museums of Kenya, Distong National Museum of Natural History, and the National Museum of Natural History, Smithsonian Institution. MicroCT scans of Sterkfontein fossil material were produced through a collaborative project between the University of the Witwatersrand and the Department of Human Evolution, Max Planck Institute for Evolutionary Anthropology. This research was supported by the Wenner–Gren Foundation for Anthropological Research Wadsworth Fellowship (to H.C.), the Leakey Foundation Baldwin Fellowship (to H.C.), Smithsonian’s Peter Buck Postdoctoral Fellowship (to H.C.), and National Science Foundation Grants BCS-0521835 and DGE-0801634 (to B.G.R.).


This article is a PNAS Direct Submission.



Habiba Chirchir1 [email protected]
Center for the Advanced Study of Hominid Paleobiology, Department of Anthropology, The George Washington University, Washington, DC 20052;
Human Origins Program, Department of Anthropology, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560;
Tracy L. Kivell
Animal Postcranial Evolution Laboratory, School of Anthropology and Conservation, University of Kent, Canterbury, Kent, CT2 7NR, United Kingdom;
Department of Human Evolution, Max Planck Institute for Evolutionary Anthropology, D-04103 Leipzig, Germany;
Christopher B. Ruff
Center for Functional Anatomy and Evolution, Johns Hopkins University School of Medicine, Baltimore, MD 21205;
Jean-Jacques Hublin
Department of Human Evolution, Max Planck Institute for Evolutionary Anthropology, D-04103 Leipzig, Germany;
Kristian J. Carlson
Evolutionary Studies Institute, The University of the Witwatersrand, Braamfontein 2000 Johannesburg, South Africa;
Department of Anthropology, Indiana University, Bloomington, IN 47405; and
Bernhard Zipfel
Evolutionary Studies Institute, The University of the Witwatersrand, Braamfontein 2000 Johannesburg, South Africa;
Brian G. Richmond1 [email protected]
Center for the Advanced Study of Hominid Paleobiology, Department of Anthropology, The George Washington University, Washington, DC 20052;
Human Origins Program, Department of Anthropology, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560;
Division of Anthropology, American Museum of Natural History, New York, NY 10024


To whom correspondence may be addressed. Email: [email protected] or [email protected].
Author contributions: H.C. and B.G.R. designed research; H.C. performed research; T.L.K., C.B.R., J.-J.H., K.J.C., and B.Z. contributed new reagents/analytic tools; H.C. analyzed data; H.C. wrote the paper; and T.L.K., C.B.R., and B.G.R. contributed to writing the paper.

Competing Interests

The authors declare no conflict of interest.

Metrics & Citations


Note: The article usage is presented with a three- to four-day delay and will update daily once available. Due to ths delay, usage data will not appear immediately following publication. Citation information is sourced from Crossref Cited-by service.

Citation statements



If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.

Cited by


    View Options

    View options

    PDF format

    Download this article as a PDF file


    Get Access

    Login options

    Check if you have access through your login credentials or your institution to get full access on this article.

    Personal login Institutional Login

    Recommend to a librarian

    Recommend PNAS to a Librarian

    Purchase options

    Purchase this article to get full access to it.

    Single Article Purchase

    Recent origin of low trabecular bone density in modern humans
    Proceedings of the National Academy of Sciences
    • Vol. 112
    • No. 2
    • pp. 297-E237







    Share article link

    Share on social media