Understanding Entheseal Changes: Definition and Life Course Changes

ABSTRACT

Over the past two decades, many articles have been published on entheseal changes (usually called ‘Musculoskeletal Stress Markers’) as activity markers in past societies. Over-simplified methods and over-interpretation of past activities have generated strong critiques of research results in this area of enquiry. While some significant improvements regarding the recording systems for entheseal changes have been applied more recently, many bioarchaeologists appear not yet to be fully aware of the multi-factorial aetiology of these alterations. In this article, we review the anatomical and clinical literature to discuss some of the difficulties associated with the recording of entheseal changes and the multiple factors leading to their appearance in the human skeleton. Thus far, fibrocartilaginous entheses appear to hold more promise for activity-related reconstruction than do fibrous ones, but these relationships remain an area of active research interest. Copyright © 2012 John Wiley & Sons, Ltd.

Introduction

La Cava (1959) appears to have been first to use the term ‘enthesis’ for creating the word ‘enthesitis’ to designate the inflammation of tendinous attachments. Subsequently, Ball (1971) and Niepel & Sit'Aj (1979) suggested using the words ‘enthesis’ to designate the area where a tendon, a capsule or a ligament attaches to bone and ‘enthesopathy’ to indicate pathological changes of this structure. ‘Enthesis’ and ‘enthesopathy’ are commonly used today in biomedical sciences, but other terms can also be found in the literature (Table 1).

Table 1. Common terminology used in biomedical sciences

Most common terms in biomedical sciences	‘Synonyms’
^a : There is no consensus in favour of either ‘entheseal’ or ‘enthesial’ and both are in standard use.
Enthesis (plural: entheses)	Insertion site / Attachment site
	Insertion area / Attachment area
	Tendon-to-bone insertion / Ligament-to-bone insertion
	Zone of insertion / Zone of attachment
Entheseal (adj.)	Enthesiala
	Enthesal
Enthesopathy (plural: enthesopathies)	Enthesiopathy
	Insertiopathy
	Insertional tendinopathy
	Enthesopathic change
Enthesitis (i.e. inflammation)	Insertitis
	Insertional tendinitis / Insertional tendonitis
	Insertional periostitis

An enthesopathy can be a radiological, clinical, histological, or osteological finding, and the notion of pain is not necessarily associated with its occurrence (e.g. Ball, 1971; Resnick & Niwayama, 1983; François et al., 2001). In biological anthropology, several terms are used to designate the osteological changes seen in entheses: enthesopathies (e.g. Dutour, 1986; Hawkey, 1988), muscle crests (Angel et al., 1987), musculoskeletal stress markers (Hawkey & Merbs, 1995) or muscle markings (Robb, 1998). Recently, Jurmain & Villotte (2010) proposed the generic and more neutral term ‘entheseal changes’ to designate all alterations of entheses seen in skeletal material.

Biological anthropologists have been interested in entheseal changes of the human skeleton for more than a century (e.g. Lane, 1887, 1888; Testut, 1889). Three main factors account for this interest. First, the areas of tendon or ligament attachment are usually easily visible on dry bones. Second, entheses exhibit a number of morphological variations, with more or less pronounced changes, such as irregularity and porosity, so changes can be scored. Finally, as entheses are regularly under heavy strain during physical activity, changes can, hypothetically, be used to reconstruct past physical activities. This type of study also has a long and contentious disciplinary history (for reviews, see Kennedy, 1989; Dutour, 2000; Jurmain et al., 2012), with some notable detractors (e.g. Jurmain, 1999; Zumwalt, 2006; Jurmain & Roberts, 2008; Alves Cardoso & Henderson, 2010) as well as supporters (e.g. Dutour, 1986; Hawkey & Merbs, 1995; Molnar, 2006; Villotte et al., 2010a, 2010b; Havelková et al., 2011).

Methodological research recently carried out by several scholars (Mariotti et al., 2004, 2007; Zumwalt, 2005; Villotte, 2006, 2009; Henderson & Gallant, 2007; Henderson et al., 2010; Villotte et al., 2010a) provides a renewed interest for the study of entheseal changes as potential markers of activity. This can be seen, for example, by the success of the Workshop in Musculoskeletal Stress Markers (MSM): limitations and achievements in the reconstruction of past activity patterns, held at the University of Coimbra, in Portugal (2^nd–3^rd July 2009) (Santos et al., 2011). However, this relative success could turn against itself, and one cannot entirely dismiss previous remarks on the limitations and pitfalls of this kind of study.

The purpose of this article is not to discuss the reliability of entheseal changes for the reconstruction of past activities, nor the strictly methodological aspects of this kind of study because those considerations have been largely discussed elsewhere (Dutour, 1992; Robb, 1998; Jurmain, 1999; Mariotti et al., 2004, 2007; Villotte, 2006, 2008a, 2009; Henderson & Gallant, 2007; Alves Cardoso & Henderson, 2010; Villotte et al., 2010a; Jurmain et al., 2012; Milella et al., 2012). The goal here is rather to attempt to clarify several points to anticipate continued use of entheseal changes as potential markers of activity in past populations. This perspective is based on the following premises: the most reliable data on entheses and entheseal changes are provided by their use in the biomedical sciences (even if some studies, as case reports, have apparently little broader application), and data from identified skeletal collections are far better than those provided by archaeological samples because of numerous inherent biases (e.g. the effect of age and the difficulty to assess the age-at-death in archaeological samples, their unknown genetic background).

Anatomical background

A brief overview of entheseal types

Two groups of entheses can be distinguished according to the tissue type present at the skeletal attachment site: fibrocartilaginous and fibrous (Benjamin & McGonagle, 2001; Benjamin et al., 2002). Recently, the importance of this anatomical distinction for recording entheseal changes in skeletal material has been highlighted, independently, by Henderson (Henderson & Gallant, 2007; Alves Cardoso & Henderson, 2010) and by Villotte (2006, 2008a, 2009; Villotte et al., 2010a).

Briefly, fibrocartilaginous entheses occur close to joints of the long bones, but also on short bones and some parts of vertebrae; fibrous entheses occur on the diaphysis of long bones and also on the vertebral column (see Tables A1 and A2 in Villotte et al., 2010a for a list of the main post-cranial fibrous and fibrocartilaginous entheses). Four histological zones are distinguished in fibrocartilaginous entheses (Cooper & Misol, 1970; Benjamin et al., 1986): (i) tendon or ligament, (ii) uncalcified fibrocartilage, (iii) calcified fibrocartilage, and (iv) subchondral bone. Zones 2 and 3 are separated by a regular calcification front called the ‘tidemark’. The tidemark, which is relatively rectilinear and not crossed by blood vessels, is the point at which soft tissues are removed during maceration (Benjamin et al., 1986). In contrast to fibrocartilaginous attachments, anatomical and anatomo-pathological descriptions for fibrous entheses are extremely rare. They attach soft tissues to bone directly or via a mediating layer of periosteum (Benjamin et al., 2002). The anchorage is achieved through collagen fibres from periosteum, tendon or ligament, which are embedded into bone (François et al., 2001; de Pinieu & Forest, 2003). At fibrous entheses, blood vessels from the tendon or the ligament may anastomose with those of the bone (Dörfl, 1969).

Defining a ‘normal’ and a ‘changed’ enthesis, is it simple?

The characteristic aspect of a healthy fibrocartilaginous enthesis is described by Benjamin et al. (2002: 939) in the following manner: ‘as the tidemark is relatively straight and the fibrocartilage zones avascular, the site of attachment in a healthy enthesis is smooth, well circumscribed and devoid of vascular foramina.’ This description fits well with the appearance of several attachment sites seen on the skeleton (Figure 1). Moreover, the accurate and numerous descriptions of changes that occur in enthesopathies in living people – erosion of the calcified fibrocartilage and subchondral bone, tidemark irregularity, vascularization of the fibrocartilage, calcification and ossification of soft tissues, cysts, and avulsions can be observed in skeletal material (Villotte, 2006, 2009; Villotte et al., 2010a and clinical references therein). Thus, it seems conceivable to define a ‘healthy’ (or ‘normal’) fibrocartilaginous enthesis in skeletal material as a smooth, well-defined imprint on the bone, without vascular foramina, and with a regular margin, and an enthesopathy in the other cases (Villotte, 2006, 2009; Villotte et al., 2010a). However, this definition can be applied for only some fibrocartilaginous attachments; it does not work as well, for instance, for the entheses of ligamenta flava on the vertebral column (Villotte, 2006) or for small attachment sites like the insertion of the M. brachialis (pers. obs., and Mariotti, pers. comm.). In the case of the M. brachialis, this may be related to the relatively thin layer of uncalcified fibrocartilage at the insertion, and the fact that this muscle is already attached to the bone at birth (cf. Benjamin et al., 1992).

Figure 1. Greater trochanter, insertion of the M. gluteus medius. Smooth imprint with regular margins. Scale: 1 cm.

Regarding the paucity of clinical and anatomical data for fibrous entheses, the definition of a ‘normal’ type is far more complex (Villotte, 2006, 2009; Alves Cardoso & Henderson, 2010). Osseous irregularity in the area of fibrous attachments is common in human skeletal remains, even in the first decades of adulthood (Villotte, 2009). Since the term ‘enthesopathy’ implies a pathological condition, it is not appropriate to designate all of these very common and probably asymptomatic changes as pathological (Jurmain & Villotte, 2010).

Possible pitfalls of anatomical over-simplification

Changes in the two types of enthesis do not indicate the same phenomena, and it is to be expected that biological anthropologists will no longer combine them in a single study. Although the distinction between fibrous and fibrocartilaginous entheses seems clear enough, one should not forget that some amendments to their apparent distinctiveness have been made. First, some attachments are ‘mixed.’ Hems & Tillmann (2000) showed that the majority of the entheses of the masticatory muscles are of this type. Thus, the insertion of M. masseter is partly periosteal, partly osseous and partly fibrocartilaginous. Second, in a fibrocartilaginous enthesis, the periphery has little or no fibrocartilage (Benjamin et al., 1986, 2002). Third, fibrocartilage may exist in a small quantity at a fibrous enthesis, particularly on the metaphysis, an example of which is the M. pectoralis major insertion on the humerus (Benjamin et al., 1986, 2002). Finally, one should not forget the morphological diversity in our species. For example, the most distal part of the insertion of the M. iliopsoas, at the junction between the lesser trochanter and the femoral shaft, may correspond to the inconstant fibrous insertion of the M. iliacus (Polster et al., 2008) and this region is highly variable as a consequence (Figure 2).

Figure 2. Lesser trochanter, insertion of the M. iliopsoas. White arrow: salient margin at the trochanter, taken into account in Villotte (2006). Black arrow: salient margin at the junction between the lesser trochanter and the femoral shaft, which can occur independently and was not taken into account in the same scoring system. Scale: 1 cm.

Several authors have discussed the boundaries of the concept of enthesis (for fibrocartilaginous attachments), and all agree that it cannot be reduced to the attachment of a tendon or ligament.1 In addition to the attachment zone, Niepel & Sit'Aj (1979) include under the term ‘enthesis’, all of the following: peritenon, associated bursae, fibrous tissues, fat-pads, and sesamoid bones. Benjamin et al. (2002, 2004) formalized this concept and suggest using ‘enthesis organ’ for the complex consisting of these anatomical structures around the fibrocartilaginous entheses sensu stricto because fibrocartilage is almost systematically observed in these structures. It is important that biological anthropologists consider this point before scoring entheseal changes. For instance, recent methodological papers (Mariotti et al., 2004, 2007; Villotte, 2006; Henderson & Gallant, 2007; Henderson et al., 2010; Villotte et al., 2010a) do not mention bursae and whether or not bursopathies are scored. Admittedly, this is a thorny problem. Villotte (2009), perhaps wrongly, did not score bone changes that can occur at the location of bursae (Figure 3) for several reasons: at present, data on bursopathies are still too scarce to accurately analyze the consequences on the skeleton, the presence of bursae is inconstant, and their exact location varies. However, these data may be of interest in the discussion of past activities - bursae, as a part of the ‘enthesis organ’, participate actively in the dissipation of mechanical stresses. Moreover, it should be noted that sometimes the distinction between a bony involvement of the tendon attachment site or of the bursa is difficult to establish.

Figure 3. Radial tuberosity. Changes occur at the medial part of the tuberosity – i.e. the attachment of the distal tendon of the M. biceps brachii (white arrow), and the lateral part that is the location of the bursa associated with this tendon (black arrow). Scale: 1 cm.

Entheseal changes through life: the effects of age, hormones and activity

Trauma and micro-trauma can produce entheseal changes, among numerous others factors (Resnick and Niwayama, 1983). Clinical literature also reports the presence of entheseal changes in many diseases. The goal is not to present these conditions in detail here (listed in Henderson, 2008; Villotte, 2009), but rather to focus on lesser-known aspects of entheseal changes related to age, hormones, and physical activities. However, it seems necessary to present briefly the two main causes of non-degenerative enthesopathies: seronegative spondyloarthropathies and diffuse idiopathic skeletal hyperostosis (DISH or hyperostotic disease). DISH is characterized by para-articular bridging osteophytes in the anterolateral aspect of the spine (Forestier & Rotes-Querol, 1950; Resnick et al., 1975). Exuberant bone production is seen at extra-spinal fibrocartilaginous and fibrous entheses (Resnick et al., 1975). Very early on, researchers recognized the enthesis as a primary target in ankylosing spondylitis and other seronegative spondyloarthropathies (Ball, 1971; Paolaggi et al., 1984a, 1984b). The inflammation occurs at the level of fibrocartilaginous entheses, leading to erosion of the fibrocartilage (Benjamin & McGonagle, 2001; Fournié, 2004). This erosive process is followed by deposition of reactive bone and the formation of an enthesophyte (Ball, 1971; Resnick & Niwayama, 1983). It is noteworthy that entheseal changes seen in cases of seronegative spondyloarthropathy are characterized, at least for some fibrocartilaginous entheses, by erosive lesions uncommonly seen in other individuals (Villotte & Kacki, 2009).

In studies of entheseal changes in skeletal samples with known age-at-death, age is the main aetiological factor identified (Shaibani et al., 1993; Cunha & Umbelino, 1995; Mariotti et al., 2004, 2007; Villotte, 2009; Alves Cardoso & Henderson, 2010; Villotte et al., 2010a; Niinimäki, 2011; Milella et al., 2012). However, the precise relation between age and entheseal changes remains poorly described (and in some aspects poorly understood). Properties of the entheses during skeletal immaturity are mainly described in studies of animal models and, to a lesser extent, in studies of human cadavers. For adulthood, data derive mainly from sports medicine or that associated with the aged, though the study of identified skeletal collections provides informative results.

Secondary ossification centres

In early development of humans and other mammals, the tendon or ligament attaches to the perichondrium (Hurov, 1986; Gao et al., 1996; Wei & Messner, 1996; Shaw et al., 2008). During growth, entheses seem to act as growth plates; the cartilage is resorbed at the inner side and produced at the outer side, possibly by metaplasia (Gao et al., 1996; Nawata et al., 2002). The classic appearance of a fibrocartilaginous enthesis (i.e. the four histological zones, see supra) appears in non-human mammals when growth slows or stops (Wei & Messner, 1996; Nawata et al., 2002; Wang et al., 2006). For instance, in the attachment zones in the rat anterior cruciate ligament, the boundary between uncalcified and calcified fibrocartilage is not clearly distinguishable before growth slows (Nawata et al., 2002). The process is not described for humans, but this progressive organization of the enthesis during growth and development could explain the lack of a clearly distinguishable area of attachment in juvenile human skeletons (Figure 4, compare with Figure 1). Indeed, in skeletal remains, the classic appearance of a fibrocartilaginous enthesis is seen when epiphyses of short and flat bones (also called apophyses) and long bones are partially or fully fused. The most common activity-related change occurring before the complete fusion of the epiphysis is a total or a partial bony avulsion (e.g. Resnick and Niwayama, 1983; Nakanishi et al., 1996; Stevens et al., 1999; Adirim & Cheng, 2003). Consequences of these avulsions (total or partial) can be observed on the adult skeleton (Villotte et al., 2010b; Knüsel, 2012).

Figure 4. Proximal humeral epiphysis of an immature individual (6–9 years old). The area of attachment of the M. supraspinatus and M.infraspinatus on the greater tubercle (white arrow) is not clearly distinguishable. Scale: 1 cm.

During adulthood, the degenerative process related to age and mechanical demands affects both the tendon and the fibrocartilaginous enthesis. Within the aging tendons, the amount of denatured collagen and proteolytic cleavage of matrix components increase (Riley, 2004). These changes lead to deterioration in the physical properties of the tendon (Riley, 2004), which, in turn, may favour the occurrence of mechanically induced alterations in the enthesis (Rodineau, 1991; Bard, 2003). From the sixth decade onwards, the fibrocartilaginous enthesis itself is the target of the degenerative process (Durigon & Paolaggi, 1991; Rodineau, 1991; Bard, 2003). Those degenerative changes are well described (Durigon & Paolaggi, 1991; Lagier, 1991; Kumagai et al., 1994; Jiang et al., 2002; Milz et al., 2004; Benjamin et al., 2007, 2009). They are:

• Microtears or microdamage of one of the four histological zones of the enthesis (tendon, uncalcified and calcified fibrocartilage, bone);
• Formation of enthesophytes (bony spurs at the enthesis), induced by the healing process, after microtears;
• Disturbance of collagen fibres and of the organization of cell columns;
• Calcific deposits;
• Increase of the thickness of the calcified fibrocartilage layer;
• Vascularization of the calcified and uncalcified fibrocartilage layers;
• Erosion of the surface and bone resorption beneath the enthesis.

Disturbance of enthesis organization and subsequent healing processes are visible on skeletal remains: vascularization, enthesophytes, calcific deposits, cysts, and irregularity of the surface are common in the skeletons of old individuals (Shaibani et al., 1993; Cunha & Umbelino, 1995; Mariotti et al., 2004, 2007; Villotte, 2009; Alves Cardoso & Henderson, 2010; Villotte et al., 2010a; Milella et al., 2012). Bony spur formation typically occurs in the most fibrous part of an enthesis (Villotte, 2006; Benjamin et al., 2009). It should be noted that, contrary to other fibrocartilaginous entheses, there is no correlation between frequency and size of enthesophytes at the ligamenta flava attachment sites and age-at-death (Cunha & Umbelino, 1995; Villotte, 2009).

Excessive mechanical stress in terms of frequency, speed, and/or intensity can cause a series of micro-traumatic insults that tend to disturb the tissue structure of the fibrocartilaginous enthesis (Husson et al., 1991; Khan et al., 1999; Benjamin et al., 2006). In young adults, these mechanical stresses are the main factor in the occurrence of an activity-related enthesopathy (Rodineau, 1991). In the older individual, on the contrary, it is the gradual depletion of tendon vascularity close to the insertion that favours the occurrence of lesions (Rodineau, 1991). Biomechanical parameters are, in this case, a secondary factor. Other factors may increase the risk of developing enthesopathy, for instance cold temperatures, the use of unsuitable equipment, very heavy muscular stresses endured without training or appropriate warm-up, and abnormal structures disrupting joint biomechanics (Commandré, 1977; Rodineau, 1991; Bard, 2003). The work of several authors, including Khan et al. (1999) and Milz et al. (2004), clearly indicate that overuse enthesopathies are similar to degenerative ones described in older individuals. Thus, micro- and macro-bony avulsions, tidemark irregularity (disorganization of the layer of calcified fibrocartilage), vascularization of the fibrocartilage, calcification, and ossification of soft tissues can be observed in cases of overuse enthesopathy (Dupont et al., 1983; Husson et al., 1991; Saillant et al., 1991; Potter et al., 1995; Selvanetti et al., 1997). It is noteworthy that in these sports medicine reports, skeletal alterations are, in most cases, inconspicuous.

Sports and occupational injuries at tendons and entheses seem, at present, more common in women (Punnett & Herbert, 2000; Bard, 2003). However, these differences are not observed for all anatomical sites and many sex-related parameters (e.g. muscle mass, fat mass, size, morphology), may interact (e.g. Punnett & Herbert, 2000; Bard, 2003). Among the factors involved, ovarian hormones, including estradiol and relaxin, could play an important role. The contribution of these hormones in reducing the amount of glycosaminoglycans and collagen has been demonstrated for fibrocartilaginous joints (Naqvi et al., 2005; Hashem et al., 2006). Moreover, these hormones promote hyper-laxity and increase the risk of intrinsic mechanical lesions (Punnett & Herbert, 2000; Bard, 2003). At the menopause, blood levels of estrogen drop significantly (Sowers, 2000). This decrease causes a change in the composition of collagen connective tissue, including ligaments, associated with loss of elasticity (Falconer et al., 1996; Ewies et al., 2003).

Diaphyses

As for fibrocartilaginous entheses, the following discussion on properties of fibrous entheses during skeletal immaturity is based on animal models. These models are of two kinds: those focusing on gross anatomy and those studying the histological properties of these entheses. The first type indicates that during growth, there is a relationship between muscle activity/properties and the morphology of attachment sites. For instance, Dysart et al. (1989) demonstrated that denervation of the rat forelimb is followed by an abnormally formed humerus, notably a smaller and less curved deltoid tuberosity. Based on these findings, they postulated ‘that muscle pull affects periosteal tension and consequently bone form and growth in length’ (Dysart et al., 1989: 158). In a study of mutant strains of mice, Montgomery et al. (2005: 819) reached a slightly different conclusion: ‘These findings suggest that muscle attachment sites expand during growth in order to accommodate increases in muscle size and mass, but that expansion of these bony regions is not necessarily dependent on increases in muscle contractile strength […]’.

If these experiments provide interesting insights into the effects of muscle properties and activity on bone morphology, the study of histological properties of attachment sites appears more informative for a better understanding of the ‘normal’ appearance of fibrous entheses in the immature human skeleton. In dog, rabbit and rat studies on diaphyseal entheses, tendons, and ligaments attach via periosteum during growth (Laros et al., 1971; Dörfl, 1980a, 1980b; Hurov, 1986; Matyas et al., 1990; Gao et al., 1996; Wei & Messner, 1996). Both osteoclastic and osteoblastic activity are seen at fibrous attachment sites during this period and appear to be mainly related to the migration of the attachments of tendons and ligaments during the growth in length of long bones (Hoyte & Enlow, 1966; Dörfl, 1980a, 1980b; Hurov, 1986). It is noteworthy that muscular traction plays no role in this migration (Dörfl, 1980a, 1980b; Grant et al., 1981). All these animal model studies included the tibial insertion of the medial collateral ligament. This ligament attaches to the tibia of growing individuals in an area called the ‘metaphyseal depression’, where growth-related osteoclastic resorption is more predominant than osteoblastic activity (Dörfl, 1980a; Matyas et al., 1990; Wei & Messner, 1996). Osteoclasts are most obvious at the periosteal side of the bone, but they are also seen at the endosteal or marrow side (Wei & Messner, 1996). It is noteworthy that the ‘metaphyseal depression’ disappears in mature rabbits (Matyas et al., 1990), but a shallow depression persists in rats up to 120 days of age, which was interpreted by Wei & Messner (1996) as a sign of continuing growth. Many biological anthropologists (e.g. Saunders, 1978; Castex, 1990; Stirland, 1996; Mariotti et al., 2004) reported high frequencies of a ‘fossa’ in juveniles and young adults for several metaphyseal attachment sites (e.g. humeral insertions of the Mm. pectoralis major and teres major and the femoral insertion of the M. gluteus maximus). Actually, these grooves or ‘fossae’ are very common in immature human skeletons, in frequency but also in their distribution in the body (Figure 5), and it may be tentatively suggested that they are related to a process similar to that described for the tibial ‘metaphyseal depression’ in rats and rabbits. These changes, especially for the humeral insertion of the M. pectoralis major, seem to be more dramatic in males during late adolescence and early adulthood, before the complete fusion of the epiphysis (Mariotti et al., 2004). Consequently, linking these erosions in young adult males solely to mechanical stresses (e.g. Hawkey & Merbs, 1995) appears, at least, highly hazardous (see Villotte, 2008b for a review of the possible causes of a ‘fossa’ at the humeral insertion of the M. pectoralis major in adults). Moreover, the bottom of these grooves is usually not smooth in juvenile human skeletons: marked porosity, short striae, and small asperities are often present (Villotte, 2006). One could speculate that those changes are related to the irregularity of the mineralization front (i.e. the superficial cortex) before skeletal maturity in other mammals (Matyas et al., 1990; Wei & Messner, 1996).

Figure 5. Proximal tibial shaft of an immature individual (6–9 years old). The area of attachment of the M. soleus displays the classic appearance of immature metaphyseal entheses, a ‘fossa’ with porosity, short striae, and small asperities/irregularities at its bottom. Scale: 1 cm.

In mature animals, the periosteal layer may or may not disappear, depending on the species and the enthesis. In adult humans, periosteal fibrous entheses are seen where muscles attach to a large area by short fibrous ends (Kenesi & Tallineau, 1991), and for some masticatory muscles (Hems & Tillmann, 2000). As the mediating layer of periosteum often disappears with age and leaves the soft tissue attaching directly to bone (Benjamin et al., 2002), it has been hypothesized that the physiological transition from a periosteal to a bony attachment in early adulthood may explain the high frequency of skeletal changes (i.e. irregularity) seen in young/middle-aged adults (Villotte 2009).

Benjamin et al. (2002: 934) note that ‘relatively little attention has been paid to fibrous entheses, even though they are associated with some of the largest and most powerful muscles in the body […]. This partly reflects a clinical bias toward fibrocartilaginous entheses – which are more vulnerable to overuse injuries, but also the attraction of working with a richer variety of tissues that such entheses can offer.’ In a study that focuses mainly on fibrocartilaginous entheses, Benjamin et al. (2007) briefly describe two modifications observed in elderly subjects at the fibrous insertion of M. pronator teres: a bony production and a vascular invasion of the fibrous tissue. Micro-trauma at fibrous entheses is described mainly for periosteal attachment sites (e.g. Condouret & Pujol, 1985); they lead to a periostitis. Only a few cases were reported for bony ones: enthesopathy at the M. deltoideus insertion on the humerus in golfers and ‘pala’ players (Commandré, 1977: 67) and small resorptive areas at the humeral insertion of the M. pectoralis major in gymnasts (Fulton et al., 1979).

To conclude this section, it seems important to report an interesting study on the effect of inactivity for several entheses in dogs (Laros et al., 1971). In active immature dogs, normal metaphyseal remodelling was seen with a marked bone resorption at the tibial insertion of the medial collateral ligament (i.e. a normal appearance, cf. supra). In inactive adolescent dogs, the reaction was more generalized. Contrary to the other entheses under study, simple caging for six weeks produced resorption at the tibial insertion of the medial collateral ligament in adult dogs. Moreover, after several weeks of immobilization in a plaster cast, resorptive changes at this enthesis were seen for the immobilized limb of adult dogs, but also in a lesser extent for the non-immobilized limb, free for activity, and weight-bearing! With continued caging (over a period of six months or more) and in dogs sacrificed 12 weeks after removal of plaster immobilization, bone resorption healed as fibrous tissue replaced resorbed bone and then became mineralized.

Conclusion

In the last 20 years, researchers have published important works on the limits and pitfalls of interpretations of entheseal changes as activity markers (e.g. Dutour, 1992; Jurmain, 1999; Jurmain et al., 2012), mainly related to the problem of false positives (in our case, an entheseal change not related to physical activity). Based on our experience, exuberant bone production is mainly seen in older individuals, individuals with systemic disease or, locally, in cases of trauma. In many cases, major entheseal changes are probably not directly related to physical activity, and certainly not only to micro-trauma at the enthesis.

While some biological anthropologists seem to have been completely unaware of these problems, others, notably Weiss (2003, 2004; Weiss et al., 2012), have attempted to identify the ‘confounding’ factors, but without adequately documented material. The archaeological record does not represent the best samples from which to identify the processes that produce alterations of an attachment site, for at least one good reason: the age-at-death assessment. This contribution illustrates the usefulness of clinical studies and studies based on identified skeletal collections for the understanding of entheseal changes. Based on the data obtained thus far, it seems that most of the changes seen for fibrous entheses cannot be directly associated with activity. In fact, some of the most common changes considered in biological anthropology – cortical defects at metaphyseal sites – may be related to growth and development, or even to inactivity. This hypothesis was formulated previously by Mafart (1996), though on archaeological criteria. If the study of enthesopathies for fibrocartilaginous sites appears more promising to attempt to reconstruct past activities, one cannot deny the numerous difficulties associated with their recording and the multiple factors leading to their appearance in the human skeleton.

Acknowledgements

The authors thank Charlotte Henderson and Francisca Alves-Cardoso for the invitation to participate in the Symposium and the two reviewers for their useful comments.

1. 1

   Opinions differ on the structures to be associated with the enthesis sensu stricto. For instance, Fournié proposed the extension of the concept of enthesis to include amphiarthroses and diarthro-amphiarthroses (Fournié and Fournié, 1991; Fournié, 2004). First, in histological and functional terms, these joints are closer to fibrocartilaginous entheses than to true diarthroses. Second, these joints and fibrocartilaginous entheses are predilected in seronegative spondyloarthropathies.