A hallmark of the human lineage is our bipedality. Among the many musculoskeletal anatomies that contribute to human bipedalism is an elongated Achilles tendon (Sellers et al., 2010). The Achilles tendon is comprised of the triceps surae muscle complex merging distally to form a single tendon, which inserts onto the tuber of the calcaneus (O'Brien, 2005; Harris and Peduto, 2006). The Achilles tendon has been found to increase locomotor efficiency by storing energy and returning it in elastic recoil (Alexander and Bennet-Clark, 1977; Alexander, 1984; Ker et al., 1987; Swartz, 1993; Sellers et al., 2010) and is subject to extremely high loads during various activities such as walking, running, and jumping (O'Brien, 2005). Because the Achilles tendon is particularly efficient at storing elastic energy during running, some (e.g., Bramble and Lieberman, 2004) have argued that an elongated tendon was a crucial adaptation that allowed for long distance endurance running in early members of the genus Homo.
Modern humans have the longest Achilles tendon relative to total muscle length of any living primate (Table 1), comprising approximately 65% of the total muscle length (Frey, 1913; Vereecke et al., 2005). In contrast, most other apes have a short tendon and their muscle mass is located more distally in the leg (Crompton et al., 2010). In chimpanzees, for example, the Achilles tendon constitutes a small percentage (∼7.5%) of the total muscle length (Table 1). This small tendon-to-muscle ratio permits a greater ankle joint excursion (Hanna and Schmidt, 2011; Myatt et al., 2011), which may permit the extreme dorsiflexion exhibited during vertical climbing in the apes (DeSilva, 2009). Interestingly, however, gibbons (Hylobates) have a relatively long Achilles tendon compared to the other nonhuman apes (∼35–45% of total muscle length) (Table 1). Although they are primarily arboreal brachiators, gibbons occasionally walk on their hind legs both terrestrially and arboreally (McCann, 1933; Sati and Alfred, 2002; Vereecke et al., 2006; Payne et al., 2006).
Table 1. Relative length of Achilles tendon across primates
Tendon length (as % of length of tendon-muscle unit)
All tendon lengths estimated from Figure 4 of original paper (reported as % of tibia length).
It is currently unclear when, during the course of human evolution, the Achilles tendon lengthened (Sellers et al., 2010). In fact, given the pattern of relative Achilles tendon lengths (Table 1) in primates, it remains possible that an elongated tendon is the ancestral condition and the great apes independently evolved larger muscle/tendon ratios as adaptations to vertical climbing (e.g., Lovejoy et al., 2009). Alternatively, some have suggested that an elongated Achilles may be an adaptation that developed in the genus Homo (Bramble and Lieberman, 2004; Lieberman et al., 2009). However, this hypothesis is difficult to test since a systematic comparison of the external anatomy of the calcaneus has yet to reveal an obvious skeletal correlation with Achilles tendon length. Though the Achilles tendon does leave evidence of its insertion on the calcaneal tuber, efforts to determine Achilles tendon length using such external markers have yielded mixed results, thereby limiting our ability to quantify its length based on fossilized bones (Bramble and Lieberman, 2004; Zipfel et al., 2011). For example, it is unclear whether Australopithecus had an elongated Achilles tendon based on external morphology (Bramble and Lieberman, 2004). Zipfel et al. (2011) found that the Au. sediba calcaneus (U. W. 88–99) possessed some similarities with modern humans related to the Achilles tendon insertion area, namely a smooth sloped surface for the retrocalcaneal bursa and well-developed ossified Sharpey's fibers at the Achilles enthesis. These anatomies are thought to indicate the presence of a long Achilles tendon, as they are present on human calcanei and absent on chimpanzee calcanei (Zipfel et al., 2011), but as the authors note, some of these features can be found in a small number of gorilla calcanei, which possess very short Achilles tendons (Table 1).
Complicating these efforts to reconstruct the natural history of the hominin Achilles is the near absence of fossil calcanei that preserve the calcaneal tuber (Table 2). The majority of fossil calcanei discovered have substantial damage to the area where the Achilles tendon inserts, thereby limiting the number of fossils that could potentially shed light on the evolution of this structure. Some researchers have therefore suggested that evidence for an elongated tendon may be found in the internal microstructure of the calcaneus (Crompton et al., 2010; Sellers et al., 2010). Maga et al. (2006) examined the trabecular architecture of the calcaneus in several hominoid species and found important, perhaps functionally relevant, differences. However, the sample sizes were too small to draw statistically significant conclusions, and the volume of interest was selected to study the overall stresses on the proximal calcaneus, rather than focusing specifically on the tensile pull of the Achilles tendon. In this study, we limited our volume of interest (VOI) to the trabecular bone directly underlying the tendon insertion on the calcaneal tuber since this bone should be most sensitive to the tensile strains produced during the contraction of the triceps surae.
Table 2. Complete record of calcanei currently known in the fossil record
Wolff's “Law” (Wolff, 1892) postulates that bone responds and adapts to its loading environment. Under this assumption, differences in strain (frequency or magnitude) are expected to yield differences in trabecular microarchitecture. There have been very few studies that test specifically whether the Achilles tendon has an impact on trabecular bone. In one such study, Biewener et al. (1996) investigated the effect that Achilles tendon removal had on the trabecular bone in potoroos (a small marsupial). They compared the trabecular architecture of a disuse group with an excised Achilles tendon to a control group of active adult potoroos. In accordance with the expectation of bony functional adaptation, Biewener et al. (1996) found that trabecular bone underlying the Achilles insertion differed substantially between the two groups, with Tb.N (/mm), Tb.Th (mm), and BV/TV (%) (see Table 3 for definitions) significantly lower in the disuse group. Interestingly, the degree of anisotropy (DA) did not differ between the groups, suggesting to Biewener et al. (1996) that the DA of trabecular bone in a given area may be established early in development and is not as responsive to disuse. To what degree the trabecular bone underlying the Achilles responds to increased use instead of disuse is unclear and is yet to be experimentally evaluated. These findings (Biewener et al., 1996) tentatively suggest that variation in Achilles tendon anatomy could result in different trabecular architecture in the calcaneus. If so, it is reasonable to hypothesize that primates with longer Achilles tendons should have quantifiable differences in the trabecular architecture underlying the tendon compared to primates with shorter tendons, since the longer tendons may exert more directed tensile forces on the bone. Specifically, the orientation of the trabeculae might be more unidirectional in long-tendoned primates than short-tendoned primates due to the pull of the tendon (Fig. 1). Conversely, a distal muscle belly attached to a short, broad tendon may exert a more multidirectional pull (Fig. 1), though this has never been experimentally demonstrated.
Table 3. Trabecular properties analyzed and definitions of each
Bone volume fraction
Ratio of bone volume to total volume
Average number of trabecular struts per unit length (/mm)
Average thickness of a trabecular strut (mm)
Degree of anisotropy
Maximum to minimum mean intercept length value (Whitehouse 1974), or preferential directionality
Principal trabecular orientation
Angulation of the principal trabeculae from the long axis of the calcaneus
Microcomputed tomography (micro-CT) allows researchers to quantify the complex three-dimensional geometry of trabecular bone and may shed light on this problem. Micro-CT is often used to assess microarchitecture of both cortical and trabecular bone (Bouxsein et al., 2010). We hypothesize that properties of trabecular bone, namely BV/TV, Tb.N, Tb.Th, and especially DA and PTO (Table 3) will differ in primates with different relative Achilles tendon lengths. Our hypotheses for this study are as follows:
H0: Variation in the Achilles tendon produces no discernible effect on the trabecular bone of the primate calcaneus.
H1: Variation in Achilles tendon anatomy among primates will correspond to differences in the morphology of the bone underlying the tendon insertion in the posterior calcaneus.
Seven cadaveric Homo sapiens calcanei were obtained from Boston University's medical campus. These calcanei were all from adults, but sex, exact age, and activity patterns are unknown. The calcanei were sectioned coronally at the superior most aspect of the talar articular surface so that they would fit into the micro-CT scanner.
Nonhuman primate calcanei were borrowed from the Harvard Museum of Comparative Zoology (MCZ). The sample was comprised of four baboons (Papio spp.), six gibbons (Hylobates lar), and four chimpanzees (Pan troglodytes). All were adult, wild-shot individuals (Table 4).
Table 4. Specimen list, separated by sex
Two scanners were used in this study due to size differences of the primate calcanei: a micro-CT 40 (SCANCO Medical, Brüttisellen, Switzerland) at the Center for Advanced Orthopaedic Studies (Beth Israel Deaconess Medical Center, Harvard Medical School), and a micro-CT 80 (SCANCO Medical, Brüttisellen, Switzerland) at the Department of Mechanical Engineering at Boston University. Both locations use SCANCO software to evaluate and quantify bony microarchitecture. Separating the four taxa into two groups (gibbon v. baboon; human v. chimpanzee) minimized complications related to size differences, but prevented comparisons between the two groups.
Baboon and gibbon calcanei were scanned in the micro-CT 40. Our goal in this part of the study was to compare the trabecular architecture of similarly sized primates with Achilles of similar lengths (Table 1) to test whether different locomotor patterns alone can explain trabecular differences. Gibbons are arboreal brachiators that also occasionally travel bipedally (Vereecke et al., 2006), whereas baboons travel quadrupedally on a semiplantigrade foot, and are primarily terrestrial (Rose, 1977). Small foam pieces were used to hold the calcanei in place during scanning. The outer circumference of each foam piece was cut to be the size of the scanning tube (36.9 mm). A hole was cut in the inside of the foam piece, forming a doughnut-like shape, into which the specimen was placed. The specimen was oriented so that a line connecting the superior most aspect of the talar articular surface to the superior most aspect of the calcaneal tuber was parallel with the vertical line of the scanning tube. The calcaneus was placed perpendicular to the scanning plane with the calcaneal tuber positioned superiorly. In scout view, half of the distance between the superior most talar articular surface and the most proximal part of the calcaneal tuber was selected so that homologous regions of the bone were represented in the different taxa (following Fajardo and Müller, 2001). The average number of slices scanned for baboons was 341 slices and 169 slices for gibbons. All baboons and gibbons were scanned at medium resolution, 70 kVp, 114 μA.
Chimpanzee and human calcanei were scanned in a micro-CT 80. These specimens were scanned in a 61.4 mm tube and oriented as described for the baboon and gibbon specimens. Again, a region comprising half of the distance between the posterior-most aspect of the calcaneal tuber and the superior-most aspect of the talar articular surface (or, in the case of the human specimens, the site of sectioning) was selected for scanning. The average number of slices scanned was 231 slices for chimpanzees and 282 slices for humans. All chimpanzee and human calcanei were scanned at medium resolution (60 μm), 70 kVp, 114 μA.
The scans were contoured to isolate the region of trabecular bone hypothesized to be most influenced by the pull of the Achilles tendon. Scans were oriented so that coronal plane slices could be contoured. Contouring of the trabecular VOI began at the first slice in which trabecular bone was clearly seen and a shell of cortical bone was restricted to the external perimeter of the calcaneus (cortical bone was not included in the contoured VOI). The maximum length and width of the slice were taken. A region 25% of the length away from the inferior border of the calcaneus was located and a circular contour was made (Fig. 2). Ideally, the circle would also be exactly 25% of the length away from the inferior and horizontal edges, but due to the irregular, vertically oblong shape of the calcaneus in some of the specimens (especially gibbons), the circle was closer to the mediolateral sides of the calcaneus. Contouring continued until half the number of slices scanned was reached ensuring that the bone quantified was comprised of the trabeculae nearest to the Achilles insertion. These methods resulted in a different number of contoured slices between specimens, but the same proportional region of each calcaneus by length (the posterior quarter of each specimen), as recommended elsewhere (Fajardo and Müller, 2001).
In addition to the 25% central location of the first contour, we did a second set of contours in chimpanzees and humans. These were made by simply moving the contours in an inferior direction (plantarly) so that the margin of the circular contour was 10% away from the inferior border of each slice of calcaneus (Fig. 2). Similar to the 25% contours, we continued until half the number of slices scanned was reached. This was done to add a second volume of interest (VOI) to the study, as X-rays and resin slices of humans (Milz et al., 2002) indicated that trabeculae located more inferior to the Achilles tendon insertion may be more oriented, and the tensile pull of the Achilles tendon may have a greater effect on this region.
Thresholds and Evaluation
Contours were evaluated using SCANCO software. An adaptive iterative thresholding (AIT) program was used to determine the ideal threshold for each specimen following the methodology of previous studies (Ridler and Calvard, 1978; Trussel, 1979; Ryan and Ketcham, 2002; Maga et al., 2006; Fajardo et al., 2007; DeSilva and Devlin, 2012). In the event that the two taxa being compared did not have statistically different thresholds, they were evaluated at one fixed threshold, which was the average of the two taxa. In the event that the two taxa being compared were different, each taxon was evaluated at its own ideal threshold. Statistical analysis was done using SPSS (IBM Corp., Armonk, NY). Because of small sample sizes, we used nonparametric statistics. Trabecular parameters were compared between species using a Mann–Whitney U-test while comparisons within calcanei between the 25 and 10% VOI were statistically evaluated using a paired Wilcoxon test.
The thresholds for the baboons and gibbons were not significantly different (P = 0.14); thus we used a fixed mean threshold for both taxa. The thresholds for the human and chimpanzee specimens were different from each other (P = 0.02). We therefore ran the analyses at each taxon's mean threshold, reported below. We also compared humans and chimpanzees at their fixed mean threshold and found similar results (data not shown).
There were no differences in the trabecular properties between baboons and gibbons (Table 5). In fact, the two taxa had nearly identical trabecular architecture in the calcaneal tuber despite having very different locomotive patterns. Surprisingly, no differences were found in the trabecular microarchitecture of chimpanzee and human calcanei either (Table 5; Fig. 3).
Table 5. Trabecular properties for all specimens at the 25% VOI
P-values for baboon/gibbon and chimpanzee/human comparisons are provided with significance determined at P<0.05.
In the 10% region of chimpanzee and human calcanei, there were also no differences in the trabecular properties of the two taxa (Table 6). However, there was a difference in DA between the two regions (25% VOI and 10% VOI) in humans (P = 0.03) and a strong trend towards a difference in chimpanzees (P = 0.06) (Fig. 3).
Table 6. Average BV/TV, Trabecular number, trabecular thickness, and degree of anisotropy for chimpanzee and human specimens at 10% VOI with P-values
0.20 ± .12
1.6 ± .44
0.25 ± .02
1.53 ± .12
0.30 ± .11
1.8 ± .33
0.24 ± .04
1.58 ± .29
We found post hoc that in gibbons, trabecular properties differed between males and females. Males had higher BV/TV values than females (P = 0.0001). Unfortunately, due to an insufficient number of female specimens in the other taxa (or, in the case of the humans, a lack of sex identification), we were unable to run similar tests on the remaining specimens.
The purpose of this study was to determine whether trabecular bone architecture in the calcaneus reflects Achilles tendon length variation in four different primates. Our results fail to reject the null hypothesis, suggesting that internal microarchitecture of the calcaneal trabeculae may not correlate with Achilles tendon length as we predicted.
Baboons and gibbons were compared due to their different forms of locomotion, but similar Achilles tendon lengths. A terrestrial, semiplantigrade quadruped (baboon) most likely has different loads on its calcaneus than an arboreal brachiator (gibbon), yet there were no differences in the trabecular properties of the bone underlying their Achilles tendon insertion sites. This similarity suggested to us that the similar Achilles tendon lengths of these two primates (Table 1) may be responsible for the similar trabecular architecture. If so, then humans and chimpanzees should have quite different trabecular properties given their different Achilles tendon lengths. Yet, the analysis of the trabecular bone in the calcanei of these taxa revealed no differences. The similarity in DA between humans and chimpanzees (P = 0.16 in the 25% region; P = 0.65 in the 10% region) was surprising since the Achilles tendon was predicted to have the greatest effect on this aspect of the trabecular architecture. Additionally, we calculated the angle of the principal trabecular orientation (PTO) to test whether the actual direction of the trabeculae differed between specimens, but found no significance between chimpanzees and humans (P = 0.23). We did find a difference in the DA of trabeculae between the 25% region vs. the 10% region of the same calcaneus. We found that there was a difference between the DA of the two regions in humans (P = 0.03) and a near significant difference in chimpanzees (P = 0.06), indicating that trabecular orientation does vary within the calcaneus; however, an identical pattern is found in both humans and chimpanzees.
Recently, both Sellers et al. (2010) and Crompton et al. (2010) suggested that evidence for a lengthy Achilles tendon may be found in the trabecular microarchitecture underlying the Achilles insertion. We agreed with these authors that studying the Achilles tendon's effect on trabeculae was an appropriate approach and attempted to test this hypothesis in a small sample of primates; however, our study failed to find any correlation between Achilles tendon morphology and parameters of trabecular microstructure in the region underlying the Achilles tendon. The evolution of the Achilles tendon therefore remains an open question. From the results of this study, we cannot discern the Achilles length or direction of pull based on the trabecular bone in the calcaneal tuber. Our failure to reject the null hypothesis could be due to the small sample sizes or the high degree of variation present in each taxon. For example, we found that there is a difference (P = 0.0001) between male and female gibbon BV/TV. We were only able to observe this in gibbons, as the other taxa either did not have enough specimens of both sexes, or the sex was unknown. We suggest that the difference in trabecular properties between the sexes may play a role in the substantial intraspecific variation (Hudelmaier et al., 2005; Eckstein et al., 2007). Though our volumes of interest were chosen to specifically isolate trabeculae that would be most impacted by the Achilles, there is a possibility that these VOIs were not in the correct locations or the correct size. Furthermore, we only tested four primate species and perhaps a broader interspecific study would yield different results. We would recommend that future researchers divide specimens by sex to decrease intraspecific variation, which may have been too extensive to see a clear pattern in this study.
There is also the possibility that the tensile strain produced at the junction of the Achilles tendon and the calcaneus has a stronger influence on the cortical bone than the trabecular bone. The focus of this study was trabecular bone, but we can preliminarily report that the cortical bone thickness directly overlying the trabecular VOI appears to be thicker in chimpanzees than in humans (Fig. 4). Although these data are consistent with other findings that apes tend to have thicker cortical bone than humans (Cotter et al., 2009), they are unexpected if tensile forces produced by the Achilles were influencing cortical thickness. Along these same lines, we do find one aspect of our negative results as potentially illuminating and worthy of further consideration. Comparative studies characterizing the microarchitecture of trabecular bone have consistently found that humans have a lower bone volume fraction (BV/TV) than chimpanzees. Whether in the humeral head (Shaw and Ryan, 2012; Scherf et al., 2013), first metacarpal head (Lazenby et al., 2011), thoracic vertebra (Cotter et al., 2011), femoral head (Shaw and Ryan, 2012), calcaneal body (Maga et al., 2006), talar body (DeSilva and Devlin, 2012), or medial metatarsal heads (Griffin et al., 2010), humans consistently possess lower BV/TV than chimpanzees. Under the Achilles tendon insertion, however, we found a trend towards slightly higher BV/TV in humans than in chimpanzees, though this trend was not statistically significant. The fact that humans have equal (and, if anything, greater) bone volume fraction than chimpanzees in the region underlying the Achilles is curious, given what appears to be a global pattern of reduced BV/TV in the human skeleton. We encourage future work investigating this observation in more detail.
Nevertheless, our study preliminarily suggests that varying Achilles tendon anatomy does not affect the underlying bone in primates as we suspected. One possible explanation for these results is that our model proposing that calcaneal loading direction differs in primates with long versus short Achilles tendons is incorrect or overly simplistic. Another possibility is that the forces at the Achilles tendon insertion have the same magnitude and direction regardless of tendon length, leading to an absence of differences in trabecular morphology at the enthesis. Experimental data on the Achilles/calcaneal junction are sorely needed to assess the validity of our model. Alternatively, if our model for how the Achilles tendon loads calcaneal trabecular bone is correct, then our findings contradict some predictions of Wolff's “Law.” Although many have found evidence for trabecular bone adaptation to changes in diet and hormones (Hodgkinson et al., 1978; Reeve et al., 1980; Devlin et al., 2010), and loading (Pontzer et al., 2006; Barak et al., 2011), others (Fajardo et al., 2007; Ryan and Walker, 2010; Shaw and Ryan, 2012; DeSilva and Devlin, 2012) have found the relationship between trabecular bone and inferred mechanical loading in primates to be more complicated. Although trabecular bone clearly is responsive to a variety of factors, Achilles tendon length may not affect trabecular properties as we had predicted.
In this study, we tested whether differing Achilles tendons in primates could be detected in the trabecular microarchitecture underlying the tendon insertion. Surprisingly, we failed to find differences in any trabecular bone parameter. These findings suggest that internal bone anatomy is not particularly useful in reconstructing tendon morphologies, thereby limiting its utility in reconstructing tendon length in extinct hominins. Thus, the evolutionary history of the Achilles tendon remains an open question.
The authors would like to thank Judy Chupasko at Harvard University's Museum of Comparative Zoology for loaning the non-human calcanei and Dr. Donald Siwek at Boston University's Forensic Anthropology Program for the use of the human calcanei. They are grateful to Beth Israel Deaconess Medical Center for the use of their micro-CT 40 scanner and Boston University Department of Mechanical Engineering for the use of their micro-CT 80 scanner. They are extremely grateful to Mary Bouxsein, Leeann Louis, Esther Zusstone, Cheyenne Lewis, and Matthew Reichert for their invaluable assistance and input. Emory Holland's artistic talent is on display in Fig. 1. They are grateful to Scott Miller and two anonymous reviewers for insightful comments that helped improve the manuscript. This project was funded by the Boston University Undergraduate Research Opportunities Program.