Age differences

Most individuals over the age of 50 years across cultures and geographic regions will experience degenerative join changes associated with osteoarthritis (^{Felson, 2003}; ^{Molnar et al., 2011}). Some suggest that over the life course, articular cartilage will be subject to deterioration due to extended periods of mechanical use (^{Bettié et al., 2009}; ^{Loeser, 2009}, ²⁰¹⁰). During the aging process, progressive change in articular cartilage accumulate in the extracellular matrix (^{Loeser, 2010}). Degenerative changes may rapidly progress in older individuals because of the decreased ability of articular cartilage to withstand the pressures of mechanical loading (^{Molnar et al., 2011}; ^{Ortner, 2003}; ^{Waldron, 2009}). Outliers who suffered more severely than others, perhaps from a decreased resistance to stress from the accumulation of physical damage, chronic illness or genetic predisposition, increase in number as a function of advancing age (^{Marklein et al., 2016}).

Sex differences

There is much variation existing in the expression of degenerative joint changes in adult females and males. As a result, there are conflicting views on the relationship between sex differences and degeneration during the life course. According to ^{Larsen (2015}) when significant differences can be observed for degenerative joint changes associated with osteoarthritis, males tend to express a higher frequency than females (^{Molnar et al., 2011}). Similarly, ^{Woo & Pak (2014}) noted that males showed a greater severity than females. ^{Rogers & Dieppe (2003}) argue that while this is true for men before the age of 50 years, after that age women show greater prevalence of degenerative joint changes involved in the expression of osteoarthritis (^{Nevitt & Felson, 1996}; ^{Rogers & Dieppe, 2003}). This could possibly be related to declining levels of estrogen that women experience following menopause. ^{Quispe & Williams (2019}) found that males present a greater severity of degenerative markers than females for most indicators. However, older adult females, aged 50-55 years, exhibit significantly greater eburnation than their male counterparts, although this was not true of the elderly (^{Quispe & Williams, 2019}). Meanwhile, ^{Snodgrass (2004}) found both sexes exhibit the same progression of vertebral degeneration with respect to age, although females are more variable. ^{Srikanth et al. (2005}) analyzed 34 studies on sex differences and degeneration to determine whether there were an increased severity in males or females. Although only four studies on the spine were investigated, none yielded significant difference between males and females (^{Srikanth et al., 2005}).

Given these disparate results, it is possible that females and males may or may not differ in the degeneration of the thoracic skeleton with respect to age. The two sexes are compared at two post-reproductive age cohorts, separated by twenty years. The twenty-year interval between the two age cohorts conservatively estimates the amount of time needed for significant skeletal degeneration to occur based on the length of the human maturation sequence (^{Quispe & Williams, 2019}). The current study on the thoracic spine serves as a continuation of Quispe & Williams (²⁰¹⁹) who focused on the cervical vertebrae. Therefore, the same sample size and composition -both sexes at 50-55 and 70-75 years- and the use of the ^{Buikstra & Ubelaker (1994}), scoring system, is replicated in this study.

MATERIALS AND METHODS

To examine whether sex differences in degenerative markers exists with respect to age, thoracic vertebrae were examined from complete skeletons of the William M. Bass collection of the University of Tennessee, Knoxville, where an extensive collection of human remains with recorded sex, age and causes of death is curated (^{Shirley et al., 2011}). The socioeconomic status of 60% of donors can be categorized as low to middle income and 40% were laborers with a high school diploma (^{Wilson et. al., 2007}). The donor self-reported parameters included working-class, with a living stature between five-feet two inches and six-feet. Specific stature parameters decreased extrinsic factors that could potentially affect the expression of degenerative markers. Using these selection criteria, 99 individuals from the Bass collection were examined (Table 1). The sample included females and males between the ages of 50-55 years, and females and males between the ages of 70-75 years, replicating the research design of ^{Quispe & Williams (2019}). The total sample included 52 females and 47 males, of whom 53 were older adults and 46 were elderly individuals (Table 1). Each thoracic vertebra from T1 to T12 was examined. The superior and inferior surfaces of the body were scored for lipping, porosity, and eburnation.

There are multiple interpretations available for the diagnosis of osteoarthritis. Most acknowledge eburnation as its most severe manifestation and indeed pathognomonic of osteoarthritis, although many researchers also attribute this pathology to joint surfaces that exhibit at least two degenerative changes including porosity, osteophytosis or joint contour modifications (^{Austin, 2017}; ^{Becker & Goldstein, 2018}; ^{Gibbon et al., 2014}; ^{Molnar et al., 2011}; ^{Palmer et al., 2016}; ^{Plomp et al., 2015}; ^{Roberts et al., 2006}; ^{Suby & Giberto, 2019}; ^{Űstűndağ, 2009}). Yet there is variation in the way in which these diagnostic principles are applied. Whereas some collapse the categories of eburnation and other degenerative joint changes under osteoarthritis (^{Austin, 2017}; ^{Becker & Goldstein, 2018}; ^{Gibbon et al., 2014}; ^{Palmer et al., 2016}; ^{Űstűndağ, 2009}), others tabulate these markers separately (^{Henderson et al., 2013}; ^{Molnar et al., 2011}; ^{Plomp et al., 2015}; ^{Roberts et al., 2006}; ^{Suby & Giberto, 2019}; ^{Watkins, 2012}; ^{Woo & Sciulli, 2013}; ^{Yustos et al., 2021}; ^{Zoëga & Murphy, 2016}). Given these conflicting diagnostic criteria, the Standards (^{Buikstra & Ubelaker, 1994}) were chosen to rank degenerative joint changes independently as others have done (^{Henderson et al., 2013}; ^{Watkins, 2012}; ^{Williamson, 2015}; ^{Woo & Sciulli, 2013}; ^{Yustos et al., 2021}; ^{Zoëga & Murphy, 2016}).

Following ^{Buikstra & Ubelaker (1994}) and ^{Quispe & Williams (2019}), the degree of lipping was observed macroscopically and scored from 1- 4, 1) barely observable, 2) sharp ridge, often times curled and coupled with spicules, 3) pronounced spicule formation, and 4) ankylosis or fusion (Table 2; Fig. 1). Porosity was ascertained macroscopically using scores from 1-3 and included 1) observable pinpoint perforations, 2) presence of larger coalesced perforations, and 3) both pinpoint and coalesced expressed (Fig. 2) (^{Buikstra & Ubelaker, 1994}; ^{Quispe & Williams, 2019}). Eburnation was scored from 1-3 via macroscopic examination and included 1) unobservable or questionably present, 2) polish visible, and 3) polish present coupled with grooves (Fig. 3) (^{Buikstra & Ubelaker, 1994}; ^{Quispe & Williams, 2019}). In addition, the extent of lipping, porosity and eburnation were also recorded from 1-3, including 1) 1/3, 2) 1/3 to 2/3, and 3) >2/3 of the surface affected (Table 2) (^{Buikstra & Ubelaker, 1994}; ^{Quispe & Williams, 2019}). A scoring error study conducted at Georgia State University yielded no significant difference between scores for eight vertebrae conducted as two scoring attempts separated by two weeks.

FIGURE 1 Scoring system for lipping; (a) barely observable, (b) sharp ridge, often times curled and coupled with spicules, (c) pronounced spicule formation, and (d) ankylosis or fusion; scale: 1 cm. 

FIGURE 2 Scoring system for surface porosity; (a) observable pinpoint perforations, (b) presence of larger coalesced perforations, and (c) both pinpoint and coalesced expressed; scale: 1 cm. 

FIGURE 3 Scoring system for eburnation, identified with white arrows; (a) unobservable or questionably present, (b) polish visible, and (c) polish with groove(s); scale: 1 cm. 

Lipping, porosity and eburnation signal osteoarthritis but reflect the various responses of the body to the narrowing or obliteration of interior joint spaces (^{Robson-Brown et al., 2008}; ^{Zehra et al., 2015}). The scoring system used here preserves the independence of each degenerative marker. Although none of them specifically, such as eburnation, or in combination are used to diagnosis osteoarthritis, all of these degenerative markers are involved in the expression of this pathology.

The Mann-Whitney U statistic was included to test for differences within and across age cohorts. These tests included whether the sexes manifest thoracic degeneration differently by comparing older adults to elderly within each sex. Whether females and males differed as older adults and as elderly individuals was also explored. To compare thoracic degeneration across categories, 95% confidence intervals and mean values for all vertebrae scored for the four age/sex groupings are presented for each degenerative marker.

Ordinal regression was included to address the degree to which eburnation, widely considered to be diagnostic of osteoarthritis, compares to the other degenerative markers (^{Woo & Pak, 2014}). The resulting goodness-of-fit chi-square statistic was examined to evaluate the significance of the relationship between degenerative markers. Ordinal regression allowed for a comparison of all thoracic vertebrae elements pooled together to identify significant associations, or lack thereof, between degenerative markers. Marginal lipping, porosity and eburnation were compared with one another and each of these with their respective extent of degenerative joint changes on the superior and inferior surface.

RESULTS

Older adult males have higher or equal means than their female counterparts for all features excepting eburnation and extent of eburnation where females show greater or equal values (Table 3). Similarly, the means for elderly males are higher than those of elderly females with the exception of porosity, eburnation and extent of eburnation where the converse is present (Table 3). In fact, the means for elderly males surpass those of all other age and sex groups with the exception of porosity, eburnation and extent of eburnation where elderly females are higher (Fig. 4). The means for lipping and extent of lipping are higher than for porosity and eburnation (Fig. 4). The standard deviation is highest for extent of marginal lipping across age and sex categories (Table 3).

FIGURE 4 Means (circles) and 95% confidence intervals (lines) of all vertebrae scores for each age/sex category; older adult females = unfilled blue circles; older adult males = solid red circles; elderly females = green circles with blue shade; elderly males = solid dark blue circles. 

DISCUSSION

Walking erect places great strain on the thoracic vertebrae and has for millions of years (^{Cook et al., 1983}) as well as more recently in prehistory (^{Trinkaus, 1985}). Given this antiquity of vertebral degenerative markers it should be no surprise that deterioration of the spine is quite common in extant humans by 40 years (^{Bridges, 1994}; ^{Ortner, 2003}; ^{Quispe & Williams, 2019}), but its expression is minor compared to individuals of more advanced ages (^{López-Barrales et al., 2015}; ^{Mays & Dungworth, 2009}; ^{Snodgrass, 2004}; ^{Űstűndağ, 2009}; ^{Yustos et al., 2021}). Degeneration of the skeleton accelerates after 50 years and virtually everyone has some form of osteoarthritis by 65 years of age (^{Solomon, 2001}).

The thoracic vertebrae are believed to exhibit less extreme degenerative joint changes of the centrum than the lumbar and cervical regions because of the restricted movement potential of the thorax (^{Bridges, 1994}; ^{Larsen, 2015}). In contrast, the apophyseal joints across the vertebral column are similar in degree of degenerative changes associated with osteoarthritis (Bridges, 1994). In this study, the upper (T1-T3) and lower (T9-T12) thoracic regions appear to be affected by degenerative joint changes more severely than the middle (T4-T8) vertebrae (Tables 4 and 5). This is also the case for prehistoric Amerindians from Alabama where the greatest frequency of osteoarthritis occurred on T7-T10 and T12, with a peak at T9/T10 (^{Bridges, 1994}). Other studies have identified the lower thoracic as particularly vulnerable to degenerative joint changes (^{Űstűndağ, 2009}; ^{Yustos et al., 2021}); however, the upper thoracic is also cited as a location of extensive degenerative joint changes associated with osteoarthritis (^{Jurmain, 1990}). Larsen (²⁰¹⁵) suggests the lower thoracic vertebrae bear the greatest degree of mechanical loading. This observation is corroborated by the significant differences within age (Table 4) and sex (Table 5) categories in this region. The greater severity of degenerative joint changes observed in the lower thoracic vertebrae may be due to the amount of weight it must support and a reduced threshold for mechanical loads compared to the mid and upper thoracic regions.

Although there is no specific activity or labor linked to any unambiguous manifestation of degenerative joint changes, there is some degree of correspondence between repetitive strain, particularly involving mechanical loading and severity (^{Larsen, 2015}; ^{Ortner, 2003}; ^{Űstűndağ, 2009}), as well as genetic predisposition (^{Battié et al., 2009}). Specific occupational stressors nevertheless appear to operate on the vertebral skeleton over the life course, such as habitual horseback riding or 19^th century industrial labor in London (^{Larsen, 2015}), fishing in arctic waters (^{Zoëga & Murphy, 2016}) or labor as an enslaved person (^{Watkins, 2012}). The individuals in our sample from the University of Tennessee William M. Bass Osteology Collection where donated primarily from laborer occupations and from low and middle income service workers (^{Wilson et al., 2007}). Some of the degenerative features observed could have stemmed directly from manual labor, and in the case of many males, and perhaps some females, probably derived from the transmission of heavy loads from the arms to the lower appendicular skeleton via the thoracic spine. The significant differences between the sexes in both age categories (Table 5) provide evidence that the occupational stress from labor in low- and medium skilled work was at least in part gendered. In contrast, ^{López-Barrales et al. (2015}) found no statistically significant sex differences in vertebral osteoarthritis in prehistoric Andean peoples from Chile. However, prehistoric Tiwanaku females, in contrast to males, exhibit evidence of gendered repetitive movements in patterns of osteoarthritis (^{Becker & Goldstein, 2018}) whereas males of Iron Age Zambia present greater spinal osteophytosis than do females (^{Gibbon et al., 2014}). It appears the degeneration of the spine is particularly influenced by occupational stress associated with gendered differences in labor profiles.

However, males generally weigh more and have more massive skeletons than females and the greater body weight alone over decades would tend to usher in a mechanical response of the thoracic amphiarthrodial joints from the effects of gravity and daily life. In fact, most studies show a greater prevalence of degenerative joint changes in males compared to females (^{Molnar et al., 2011}), particularly when the sexual division of labor becomes intensified, as for example, after the adoption of agriculture (^{Larsen, 2015}). In this study, males overall presented a greater expression of degenerative markers than females. This was especially observable in margin lipping (Tables 3, 5 and 6; Fig. 4). It is therefore curious as to why older and elderly female adults, with presumably lower body weights, show significantly greater eburnation and extent of eburnation compared to males within each age category, at least for some thoracic vertebrae (Tables 4 and 5). Eburnation arises after the cartilaginous layer between mating pairs of a joint recedes or is dislodged, but the joint continues to operate producing bone-to-bone contact, which polishes and grooves the articulation (^{Molnar et al., 2011}; ^{Ortner, 2003}) and may be expressed independent of porosity or marginal lipping (Table 6). Hormonal disruptions associated with menopause may at least partly explain the greater severity of eburnation in females (^{Larsen, 2015}). In addition to the highly eburnated thoracic elements from females noted in this study, eburnation of the cervical vertebrae is also significantly elevated in older adult females compared to males (^{Quispe & Williams, 2019}).

Degenerative changes of the thoracic spine differ from those of the mobile joints of the skeletal system in the presence of ankylosis, which is more common in the vertebral column than in diarthrodial articulations (^{Larsen, 2015}). The smaller inferior surface bears a relatively greater mechanical load compared to the superior articular mate below it, which may increase the growth of osteophytic lipping and eventually ankylosis. The larger number of thoracic vertebrae affected on the inferior compared to the superior surface also could be related to the unique shapes of the vertebral bodies and the placement of the articular facets.

Spinal curvature also may influence degeneration, and in particular, the increasing kyphosis of the mid-thoracic region as a function of age (^{Toh et al., 2001}). Already by age 50 years, substantial kyphosis of the thoracic skeletal has occurred, particularly in males who generally must carry a heavier head, upper arms and torso than females. These differences are merely accentuated with age and lead to extensive lipping and eburnation of the upper and lower thoracic vertebrae. Although a higher degree of degenerative joint changes would be expected to occur where the spinal curvature is the greatest since these are areas of intense stress compared to the vertebrae aligned more directly in the line of gravity, this is not always the case (^{Bridges, 1994}). However, extensive lipping of the thoracic vertebrae is a risk factor for fractures in males older than 50 years, particularly within the mid-thoracic curvature (^{Pariente et al., 2017}), and may help to counteract thoracic kyphosis (^{Toh et al., 2001}). One might also expect a greater expression of degenerative markers as a function of placement, whereby larger inferior vertebrae withstand more weight than smaller superior ones. Although the lumbar skeleton generally exhibits a greater degree of degenerative change compared to the vertebrae positioned more superiorly, there is no direct order of increasing severity as a function of placement. However, the lower thoracic in this study do seem to exhibit a greater severity of degenerative change than superior elements confirming to some extent that position does influence the degree of expression, although other factors may be present. These include changes in spinal curvature anticipating the lordosis of the lumbar vertebrae that may place unique stresses on the lower thoracic skeleton.

CONCLUSION

One goal of this study was to examine whether older adults differed from the elderly within males and within females. Another was to directly compare the sexes within age categories, corresponding to older adults (50-55 years) and the elderly (70-75 years). Age and sex differences exist for degenerative markers of the thoracic region. When only females are considered, elderly individuals exhibit a higher severity than older adults with the exception of the extent of eburnation in T1 and the extent of lipping in T11 on the inferior surface. For males, a greater expression of osteoarthritis is observed for elderly individuals with the exception of porosity in inferior T9 and extent of porosity in inferior T6. Lipping is particular excessive in elderly males throughout the thoracic skeleton.

When comparing older adult of both sexes, males present a greater severity of lipping, extent of lipping, porosity and extent of porosity while females show a greater extent of eburnation. These significant values are restricted to the upper and lower thoracic region. Among elderly adults, males express a higher degree of lipping, extent of lipping and extent of porosity, whereas females exhibit a greater degree of eburnation. Contrary to ^{Rogers & Dieppe’s (2003}) predictions, elderly males exhibit a significantly higher severity of lipping throughout the thoracic region compared to elderly females.

Elderly adults of both sexes manifest a greater severity of degenerative markers than their older adult counterparts. Males also express a greater degree of severity than females, excepting eburnation and extent of eburnation for some vertebrae. These exceptions could be a result of differences in hormone profiles associated with reproductive cessation as well as gendered occupational stress.