Degenerative spondylolisthesis in ancient and historic skeletons from New Mexico Pueblo sites.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 116:285–295 (2001) Degenerative Spondylolisthesis in Ancient and Historic Skeletons From New Mexico Pueblo Sites Charles F. Merbs* Department of Anthropology, Arizona State University, Tempe, Arizona 85287-2402 KEY WORDS olisthesis; osteoarthritis; lumbar; zygapophysial joints; disk degeneration; vertebral shift; age; sex ABSTRACT The objective of this study was to identify the presence of degenerative spondylolisthesis in adult Native American skeletons recovered from prehistoric and historic Pueblo sites in New Mexico. The vertebral columns of 491 individuals that were complete enough to allow the necessary observations produced 64 cases, with a total of 74 affected levels. The L5–S1 level was affected most frequently (78%), in contrast to clinical and anatomical studies where the L4 –L5 level was usually affected (80%). Age and sex were found to be important factors, with the condition limited to middle and old adults, and occurring in females over males at a ratio of approxi- mately 5:1. Olisthesis was shown to occur only in association with severe osteoarthritis of the zygapophysial joints, but severe osteoarthritis did not always result in olisthesis, particularly in males. Parallel or convergent inferior articular processes were found to increase in frequency in the presence of olisthesis, apparently resulting from degenerative remodeling that produced the olisthesis. The data also suggest that individuals with cranial border shifting in the column may be especially prone to olisthesis. Degenerative olisthesis was also observed in the cervical region. Am J Phys Anthropol 116:285–295, 2001. © 2001 Wiley-Liss, Inc. Spondylolisthesis, derived from the Greek roots spondylos (vertebra) and olisthesis (to slip), refers to the anterior slippage of a vertebra relative to the vertebra below (Merbs, 1996). Although known to medicine since the early 19th century, the cause of olisthesis was at first not determined. It was observed, however, that the superior surface of the sacrum and the bodies of the lower lumbar vertebrae were tilted forward and downward in humans when the trunk is erect, allowing gravity to affect them. Through careful dissection and removal of soft tissue, a Dr. Robert, working in Koblenz in the mid19th century, and known to history simply as Robert zu Koblenz (Newell, 1995), demonstrated that under normal circumstances it was impossible for a vertebra to slip forward as long as its neural arch was intact (Robert, 1855). However, if the body became separated from the arch, a condition now known as spondylolysis, it was free to slip and could do so with no greater force being applied than that equivalent to the usual weight-bearing load of the affected vertebra. The potential for anterior slippage is particularly great in the lower lumbosacral region, where the bodies are ventrally oriented downward and gravity can work upon them. The degree of olisthesis occurring varies greatly from one individual to another, ranging from none, or barely detectable, to complete, the rare instance where the body of the affected vertebra slips completely off that of the vertebra below. In general, the slippage is not gradual but episodic in nature. In time, spondylolisthesis as a sequel to complete bilateral spondylolysis be- came easy to recognize and was observed to be a relatively common phenomenon. Cases were observed, however, where olisthesis occurred in the absence of any arch defect. Junghanns (1931) referred to this as “pseudospondylolisthesis,” but the terms “spondylolisthesis with intact neural arch” (Macnab, 1950) and “spondylolisthesis without isthmus defect” (Potter and Norcross, 1954) were also introduced. Subluxation or disarticulation of the zygapophyses, or lengthening of the connection between the body and the inferior articular processes through dysplasia or repeated fracturing and healing, were seen as likely causes. A third, more common cause, was anterior migration of the zygapophyses due to severe osteoarthritic remodeling (Herron and Trippi, 1989). Newman (1955) referred to this as “degenerative spondylolisthesis,” a term that was descriptive and specific enough to gain wide popularity. Again it is the ventrally downward orientation of the vertebral bodies and the force of gravity that generate the slippage, but as opposed to olisthesis following lysis, the slippage here is very gradual and much more limited in extent. © 2001 WILEY-LISS, INC. *Correspondence to: Charles F. Merbs, Department of Anthropology, Arizona State University, Tempe, AZ 85287-2402. E-mail: Charles.Merbs@asu.edu Received 24 January 2001; accepted 30 July 2001. 286 C.F. MERBS Given its apparent cause, there is a temptation to see degenerative spondylolisthesis as merely a severe form of osteoarthritis affecting the lower back, particularly when considered entirely in its osteological context. Clinicians usually accord degenerative olisthesis a separate status, however, because zygapophysial osteoarthritis, even when severe, does not necessarily result in olisthesis, and because the condition is associated with a specific cluster of symptoms. PREVIOUS STUDIES Clinical studies demonstrated degenerative spondylolisthesis to be primarily a condition of older adults. It is usually not observed before age 40 (Rosenberg, 1975; Bolesta and Bohlman, 1989; Herron and Trippi, 1989; Takahashi et al., 1990), with most patients being in their sixties before they experience enough lower back pain to bring the condition to the attention of a physician (Rosenberg, 1975; Epstein et al., 1983; Brown and Lockwood, 1983). All studies agree that degenerative olisthesis occurs more frequently in women than in men (Rosenberg, 1975; Bolesta and Bohlman, 1989), but observed ratios vary from as low as 2:1 (Epstein et al., 1983) or 3:1 (Potter and Norcross, 1954) to as high as 6:1 (Herron and Trippi, 1989). The higher frequency in women suggests that pregnancy may play a role, and this appears to be the case. Sanderson and Fraser (1996) reviewed the records and radiographs of 949 women and 120 men aged 50 years and over who had received the attention of a spinal surgeon for low back pain over a 5-year period, and found that women who had borne children had a significantly higher incidence of degenerative olisthesis than nulliparous women (28% to 16.7%). The men they studied had a 7.5% incidence, significantly less than for nulliparous women. These results suggest that pregnancy is an important factor in the etiology of degenerative spondylolisthesis, but that females are still more prone to developing the condition, even when the pregnancy factor is eliminated. Because of the age factor, variable symptoms, and complicating conditions, overall frequency values are difficult to obtain. Potter and Norcross (1954) found only 19 affected individuals in a review of radiographs from 3,000 individuals, but their sample was skewed heavily in favor of the population segment least likely to show the condition, i.e., young males. Racial comparisons are rare, with the only agreement being that degenerative olisthesis is more common in “Blacks” than in “Whites” (Rosenberg, 1975; Bolesta and Bohlman, 1989). In a clinical study involving 200 patients, 80% showed olisthesis at the L4 –5 level, 8% at L3– 4, and 11% at L5–S1 (Rosenberg, 1975), and other studies confirmed that these frequencies are fairly typical (Junghanns, 1931; Potter and Norcross, 1954; Epstein et al., 1983; Bolesta and Bohlman, 1989, Herron and Trippi, 1989; Takahashi et al., 1990; Postacchini and Perugia, 1991). Two levels were involved in 5% of patients in two studies (Potter and Norcross, 1954; Rosenberg, 1975), and 8% in another (Takahashi et al., 1990). A study of lumbosacral spines (L1 through S2) collected at autopsy produced 19 cases of degenerative olisthesis: 18 at L4 –5, and 1 at L5–S1 (Farfan et al., 1976). Degenerative olisthesis was also found in 20 skeletons of the Hamann-Todd Collection at Case Western Reserve University, with L4 –5 being involved 18 times and L3– 4 twice (Rosenberg, 1975). Stewart (1935) reported three archaeological cases of olisthesis without lysis, one from Nunivak Island, Alaska, involving L4 –5, and one each from Hooper Bay, Alaska, and Hawikuh, New Mexico, involving L5–S1. Although degenerative olisthesis is relatively easy to detect in archaeological skeletons, no systematic studies of large series could be located. Because the neural arch remains intact in degenerative olisthesis, the maximum amount of slippage possible is considerably less than that following bilateral spondylolysis. Based on the percentage of anteroposterior vertebral body diameter exposed by the olisthesis (Laurent and Österman, 1976), a maximal slippage of approximately 30% is usually cited (Rosenberg, 1975; Herron and Trippi, 1989), although a maximal slippage of 43% has been reported (Postacchini and Perugia, 1991). At the minimum end of the scale, olisthesis of less that 5% is often excluded from studies because of the difficulty in making such small measurements (2 mm) on radiographs (Rosenberg, 1975). One study found a mean slippage of 14% among affected patients (Rosenberg, 1975). The presence of similar degenerative changes in the columns of monozygous twins suggests that some inherited aspects of basic vertebral anatomy may predispose an individual to eventual degenerative olisthesis. Although they do not comment on degenerative olisthesis per se, Palmer et al. (1984) found a close similarity in anatomy and general degenerative changes when viewing radiographs of the cervical vertebrae in 23 pairs of twins, particularly if the twins were monozygotic. All of their subjects were 50 – 60-year-old males. The shapes of vertebral bodies matched very well in twin pairs, but spinous processes less so. Ten of the 23 pairs had matching zygapophysial osteoarthritis, and 12 had matching anterior osteophytes indicative of disk degeneration. The authors suggested that the shapes of vertebrae are of considerable significance in the development of the changes associated with aging, and if this is true of the cervical region, it is likely true of the lumbosacral region as well. Ryan (1994), in fact, reported on twin brothers who were 73 years old when lumbar degenerative olisthesis was observed. Both individuals showed involvement at L4 – L5, along with symptomatic spinal canal stenosis. Lumbar degenerative spondylolisthesis has been associated with inferior articular processes that are more likely to be parallel to the midsagittal line, or even to converge caudally, rather than to diverge as DEGENERATIVE SPONDYLOLISTHESIS is usually the case (Sato et al., 1989). There is a question, however, as to whether parallel or converging processes lead to degenerative olisthesis, or whether they are a result of olisthesis. Another aspect of degenerative olisthesis involves long-standing cases of unilateral lumbar spondylolysis (Sherman et al., 1977; Aland et al., 1986; Gunzburg and Wagner, 1988; Waldron, 1992). Here the apophysial stress is reduced or eliminated on the side of the defect, but increased on the opposite side. If long-standing, this condition can also result in olisthesis due to the defect on one side and to degenerative changes of the zygapophysis on the opposite side. This results in little if any osteoarthritis on the side of the defect, but increasing osteoarthritis on the other side, leading to the condition often referred to clinically as “reactive sclerosis” (Albers and Yochum, 1980) and, potentially, the equivalent of degenerative olisthesis (Kornberg, 1988). The most obvious pathological features of lumbar degenerative olisthesis are the changes observable in the zygapophysial joints of the vertebrae involved. Rosenberg (1975) described examples in the Hamann-Todd Collection where “the articular processes had worn away, so that the articulation was formed by the anterior surface of the inferior process of the slipped vertebra and the entire posterior surface of the superior process of the subjacent vertebra. As the latter eroded posteriorly, hypertrophic new bone formed anteriorly, so that the superior process of the fifth lumbar vertebra encroached on the spinal canal and its lateral recesses.” Farfan et al. (1976) also observed at least moderately advanced osteoarthritis in the zygapophysial joints involved in degenerative olisthesis in the anatomical specimens they studied, noting that in 14 of 19 cases, the cartilaginous surfaces were almost completely destroyed or nonexistent. Other pathological conditions resulting from degenerative olisthesis have been identified, including slight stenosis (narrowing) of the spinal canal without compression of neural structures and stenosis of the nerve root canals (Postacchini and Perugia, 1991; Herkowitz and Kurz, 1991; Herkowitz, 1995). Joint involvement may not be uniform in all patients, and a rotary component, although small, has been reported (Bolesta and Bohlman, 1989). Intervertebral disk involvement was found with degenerative olisthesis to varying degrees (Elster and Jensen, 1985; Bolesta and Bohlman, 1989). Diskspace narrowing, end-plate sclerosis, or osteophyte formation, or a combination of these, were identified on radiographs of 139 of 200 patients showing degenerative olisthesis with a total of 197 levels involved (Rosenberg, 1975). Although disk degeneration and herniation have sometimes been suggested as factors contributing to degenerative olisthesis (Rosenberg, 1975; Epstein et al., 1976), poor correspondence was observed between the level of olisthesis and that of disk degeneration, the latter occurring most often at L5–S1 (100 cases), followed by 287 L4 –5 (91 cases) and L3– 4 (6 cases). Also, changes at L5–S1 were usually more profound, and occasionally fusion had occurred. It thus seemed likely to the clinicians involved that disk involvement was more an effect of olisthesis than a cause. In their study of 43 patients with degenerative olisthesis, Fitzgerald and Newman (1976) noted the virtual absence of congenital anomalies, but they did report 4 cases of transitional lumbosacral vertebrae. Epstein et al. (1973) also noted 3 examples of transitional lumbosacral vertebrae apparently associated with degenerative olisthesis. It is not clear in either study, however, whether the transitional vertebrae represented a cranial or caudal shift. Herron and Trippi (1989) reported that partial or complete sacralization of L5 (cranial shift) occurred four times more often in individuals with degenerative olisthesis than in the normal population. None of these studies appears to have investigated the possibility of shifting (as defined by Barnes, 1994) occurring at any other border, particularly the thoracolumbar border, in individuals with olisthesis. Although the picture is unclear, the possibility that vertebral border shift, particularly cranial shift, may predispose to degenerative olisthesis appears deserving of further investigation. Potter and Norcross (1954) noted what appeared to be a greater frequency of lumbar lordosis in individuals with degenerative olisthesis. More serious effects, however, are generally attributed to lateral stenosis that compromises the inferior nerve roots at the level of the olisthesis, but even superior nerve roots can be compressed in more advanced cases (Bolesta and Bohlman, 1989). Symptoms in the lower extremities had sometimes been present for 10 years, with varying degrees of back pain existing for even longer periods, before individuals presented themselves for treatment, according to Epstein et al. (1983). Five different clinical pictures were noted by Postacchini and Perugia (1991): no symptoms except occasional low back pain; chronic low back pain with no radicular symptoms; radicular symptoms and no signs of nerve root compression, with or without low back pain; radicular symptoms with neurologic deficit; and intermittent claudication. The marked disability caused by claudication contrasted sharply with the lesser neurological changes (Epstein et al., 1983). The extent of neural structure compression is quite variable (Laus et al., 1992), and appears to depend on three factors: the original size of the spinal canal, the extent of degenerative changes in the zygapophysial joints, and the extent of olisthesis (Postacchini and Perugia, 1991). Given the complexity of the clinical picture, particularly with respect to observable pathology, it seems unlikely that the study of an archaeological skeletal collection would be able to contribute much in relation to how individuals with lumbar degenerative olisthesis may have felt or how the condition may have affected their daily lives. 288 C.F. MERBS Degenerative olisthesis also occurs in the cervical region, but clinical reports are rare. Potter and Norcross (1954) reported seeing occasional cases of cervical involvement, most often between C7 and T1, which were usually associated with “exaggerated cervical lordosis in compensation for dorsal kyphos.” Only 2 of the 21 cases of degenerative olisthesis they found in an extensive radiographic study involved the cervical region; all others involved the lumbar region. Deburge et al. (1995) noted that cervical degenerative olisthesis usually occurred at C3–C4 or C4 –C5, immediately above a stiff lower cervical spine. They found the condition to be associated with two clinical patterns: neurological involvement causing cervicobrachial pain or myelopathy, or neck pain alone. Relatively little is known about the occurrence of spondylolisthesis in Native Americans. This study sought to identify degenerative spondylolisthesis in Native American adults, as shown in skeletons recovered from prehistoric and early historic Pueblo sites in New Mexico. MATERIALS AND METHODS Approximately 1,000 Native American skeletons from prehistoric or historic New Mexico Pueblo sites were examined at the National Museum of Natural History (Smithsonian Institution, Washington, DC), the Maxwell Museum of Anthropology (University of New Mexico, Albuquerque), the San Diego Museum of Man (San Diego, CA), and the Department of Anthropology at Arizona State University (Tempe). Some of the skeletons in Tempe were being curated at the time for the National Park Service. The minimal requirement of having the lower lumbosacral region (L3–S1) sufficiently intact to be examined for degenerative spondylolisthesis was met by 491 adult skeletons. These came from the following prehistoric and historic sites or areas in order of frequency: Puye, 92; Hawikuh (Zuni), 72; Gran Quivira, 54; Pottery Mound, 46; other Zuni sites (e.g., Heshatauthla, Helonawan), 43; Chaco Canyon (mostly Pueblo Bonito), 28; Paako, 27; Kuaua, 20; Sapawe, 18; Pajarito Plateau (Ottowi and Sankawi), 18; Navajo Reservoir, 15; Tijeras Canyon, 14; Giusewa, 12; Pecos Mission, 10; and Pindi, 9, with an additional 13 skeletons from nine different sites. Areas of analysis included age at death, sex, identification of degenerative olisthesis, measures of actual slippage, evidence of lumbosacral zygapophysial osteoarthritis, lumbosacral disk degeneration, and orientation of inferior articular processes. Also noted were cases of unilateral and bilateral spondylolysis, and, in the case of complete columns, information on border shifting. Where questions came up after initial analysis, it was possible, with one exception (National Park Service), to go back for restudy. Sex and age determinations are based on standard criteria (Buikstra and Ubelaker, 1994), with the condition of the vertebrae specifically eliminated Fig. 1. Left lateral view of lumbosacral region, showing simulated degenerative spondylolisthesis at L4 –L5 and L5–S1 levels (arrows). Dotted line indicates extension of anterior vertical plane of vertebral body, indicating approximate amount of slippage. as an age determinant. Sex in nearly all cases was based on the shape of the pelvis. Skeletons were placed in three approximate age categories: young adult, 18 –30 years; middle adult, 30 – 45 years; and old adult, over 45 years. Adjacent vertebrae were articulated and examined for anterior slippage of a body relative to the one below (Fig. 1). Measurements were taken using adjacent dorsal margins, rather than ventral margins as on clinical radiographs, because they are easier to take on bare bones (Fig. 2). The measure of olisthesis was based on the amount of the lower body exposed by the slippage when the affected vertebrae were placed in articulated position. Although this positioning does not take into account the cartilage that would have been covering the zygapophysial facets, measurements of slippage in columns lacking moderate to severe degenerative changes in the apophyseal joints were essentially zero, indicating that the contribution of cartilage is minimal. A millimeter scale was inserted in the midsagittal plane resting on the lower of the two bodies involved in the olisthesis, with zero corresponding to the anterosuperior margin of the lower vertebra. A reading was then taken corresponding to the length of the scale exposed by the positioning of the vertebra above. Extending a small rigid straightedge downward 289 DEGENERATIVE SPONDYLOLISTHESIS Fig. 2. Superior view of L5 and S1, showing how the typical olisthesis measurement (arrow) was taken in this study. against the dorsal surface of the upper body to intersect with the scale usually allowed for more precise measurement. All measurements were initially taken in the midsagittal plane, but additional right and left measurements were taken when the olisthesis was clearly asymmetrical. The asymmetry was noted, but the larger measurement was used in the final analysis. As in the clinical literature, the degree of olisthesis was interpreted as the amount of exposure divided by the anteroposterior diameter of the body of the lower vertebra. Body diameters were obtained using a standard sliding calliper, with osteophyte development being avoided or subtracted from the measurement. In actuality, osteophytes proved to be less of a problem than anticipated, because in this region of the column they tended to develop more to the right and left of the midline than in the midline itself. If the lower vertebra was too poorly preserved or too pathological to provide an adequate measure, the diameter of the upper vertebra was substituted. Where both body diameters were available (most cases), they were usually found to differ from each other by 1 mm or less. Since the a-p diameters of the vertebrae were relatively consistent, nearly all ranging between 30 –35 mm, the process was simplified for presentation purposes in this article by using actual measures of olisthesis rather than proportions. Zygapophysial osteoarthritis was identified primarily from roughening (e.g., pitting, grooving) and eburnation of articular surfaces, and enlargement of these surfaces through osteophyte formation. Osteoarthritis was recorded on a scale of 0 – 4 as follows: 0, none; 1, slight (usually limited to articular surface irregularities); 2, moderate (more advanced surface remodeling, slight marginal enlargement, and occasional traces of eburnation); 3, severe (larger areas of eburnation and abrasion wear through the cortical surface, and more advanced marginal enlargement); and 4, very severe (surface eburnated and deeply abraded into cancellous bone with parallel vertical grooving, and pronounced marginal enlargement). Osteophyte formation on body margins indicative of disk degeneration was scored on a 0 – 4 scale similar to that used for osteoarthritis. Scoring was based on how far osteophytes extended outward from the vertebral margin based on the following scale: 0, none; 1, 0.5–3.0 mm; 2, 3.0 – 6.5 mm; 3, 6.5–10.0 mm; and 4, more than 10.0 mm. Scoring was based on approximations of this scale as determined by visual estimation, not actual measurement. Six scores were obtained for each disk space: left, midline, and right on the ventral inferior margin of the body above, and the ventral superior margin of the body below. Marginal exostoses other than osteophytes indicative of disk degeneration, such as those associated with degenerative idiopathic skeletal hyperostosis (DISH), were rare and excluded from the scoring when encountered. Attempts to measure the angulation of the inferior articular processes and facets in the manner described by Sato et al. (1989), using radiographs, proved cumbersome and time-consuming when applied to bare bones. It was therefore abandoned in favor of simple visual assessment into three categories: diverging caudally from the midsagittal line, parallel to this line, or converging caudally toward this line. In a number of cases the processes were markedly asymmetrical, the two sides falling into different categories, and they were recorded as such. Evidence of cranial or caudal shift at vertebral borders, including partial and unilateral shift, was noted where columns were determined to be complete. Special attention was paid to lumbosacral shift, the border at or closest to lumbar degenerative olisthesis, and the one cited specifically in the clinical literature, but shift at any border was recorded. The presence of classic spondylolysis, specifically complete bilateral separation (separate neural arch), was also noted. Bilateral separation frequently results in olisthesis, but the process is very different from that producing degenerative olisthesis. A separate neural arch in fact appears to reduce stress at the zygapophysial joints, thus decreasing the likelihood of severe osteoarthritis or degenerative olisthesis at the affected level. Unilateral separation presents a very different situation. Complete separation on just one side appears to concentrate zygapophysial stress on the opposite (intact) side, thus increasing the likelihood of zygapophysial degeneration and degenerative olisthesis on this side. RESULTS The study produced 10 examples of separate neural arch (complete bilateral spondylolysis), which likely reduced stress in the area of potential degenerative olisthesis. These cases were thus eliminated from further consideration. Also noted were 8 examples of spondylolysis in which the separation was 290 C.F. MERBS Fig. 3. Cases of degenerative olisthesis in New Mexico Pueblo skeletons by level of involvement and amount (in millimeters) of slippage. complete on one side, with the other side remaining intact. In each case the affected unit was L5, and separation was through interarticularis. In only 2 cases, however, was the vertebra involved in degenerative olisthesis. Five of those unaffected were young adults (3 males and 2 females); the sixth was a middle adult female. In 5 of these individuals, the L5 laminae had not fused (spina bifida), making the situation similar to separate neural arch (actually hemiarch in this case). The 2 affected individuals were a middle adult male with 3 mm (0.09) of olisthesis, and an old adult female with 8 mm (0.28) of olisthesis. Although analogous to degenerative olisthesis without lysis, the primary subject of this report, these cases are not included in the analysis and will be examined more carefully in a separate study. The study identified 64 individuals with degenerative spondylolisthesis. A single level was involved in 55 cases, two levels in 8 cases, and three levels in 1 case, for a total of 74 levels affected. The distribution by level is as follows: L3–L4 ⫽ 2, L4 –L5 ⫽ 14, and L5–S1 ⫽ 58 (Fig. 3). Slippage ranged from the defined minimum of 2 mm (6 –7%) to a maximum of 8 mm (24 –27%). In 6 cases, all at the L5–S1 level, olisthesis was noticeably asymmetrical, with a difference of 2 mm or more between sides. The most extreme example of asymmetry, 8 mm on one side and 0 mm on the other, was observed in an old adult female (Fig. 4). The asymmetrical cases were equally divided with respect to side, 3 showing more olisthesis on the left side, and 3 on the right side. The degenerative olisthesis was well-distributed across archaeological sites and site groups, with no significant frequency differences observed. Degenerative olisthesis was observed in 41% (31/ 75) of old adults, 16% (33/271) of middle adults, and in none of the 145 young adults (Fig. 5). More females were affected than males (53:11). In fact, 60% (27/43) of all old females had the condition. Zygapophysial osteoarthritis scores rarely varied between superior and inferior articular facets, and in those few instances where they did vary, the maximum score was used (Fig. 6). The condition was Fig. 4. Dorsal view of sacrum from Gran Quivira, showing severe osteoarthritis in an individual with asymmetrical degenerative spondylolisthesis (8 mm right side, none on left side) at L5–S1. Note that the osteophytic margin of the right zygapophysial facet (arrow) is nearly in contact with the body of the sacrum, an example of severe degenerative remodeling. This specimen received an osteoarthritis score of 3 on the left side and 4 on the right. Fig. 5. Frequency of degenerative spondylolisthesis and separate neural arch in New Mexico Pueblo skeletons by age and sex. found to be absent in young adults and rare in middle adults (less than 5% of joints affected), but then to increase dramatically in old adults (23%). Frequencies were higher for males than females in both middle adults (6% vs. 3%) and old adults (27% vs. 19%). However, old adult females showed higher levels of involvement than males, with 11% of joints in females scoring higher than 1 on the scale of 1– 4 compared with just 5.5% in males. Values increased in frequency caudalward from the L3–L4 level, with the degree of involvement increasing dramatically at the L5–S1 level in old adult females. Especially striking is a comparison of old adult females with and without degenerative olisthesis: osteoarthritis was present in 43% of joints (nearly all at the 3– 4 level of severity in individuals with olisthesis, compared to only 5.5% of joints (all at the 1–2 level) in those without olisthesis). All individuals with degenerative olisthesis, regardless of sex, had level 3 or 4 DEGENERATIVE SPONDYLOLISTHESIS 291 Fig. 6. Distribution of zygapophysial osteoarthritis in vertebral columns of New Mexico Pueblo skeletons relative to age, sex, and level of involvement. osteoarthritis on at least one side at the affected level. In contrast to the osteoarthritis scores, those for disk degeneration often varied considerably between superior and inferior margins. The scores for inferior disk margins (superior vertebral margins) were usually higher, and they were the ones ultimately used in the analysis (Fig. 7). Observable changes associated with disk degeneration were found in young adults, with 6.3% of margins affected; this increased to 56% in middle adults and 77% in old adults. The amount of involvement by level proceeds in the opposite direction as that of osteoarthritis, with the highest frequencies and levels of intensity occurring at L3– 4, and the lowest at L5–S1. The amount of involvement by level proceeds in the opposite direction as that of osteoarthritis, with the highest frequencies and levels of intensity occurring at L3– 4 and the lowest at L5–S1 (Fig. 7). The vertebral level most likely to show degenerative olisthesis in this study, L5–S1, thus shows the least amount of disk degeneration. The only difference in pattern of disk degeneration between those with and those without degenerative olisthesis is a greater involvement (but with no greater intensity) at the L5–S1 level in those with olisthesis. Among young adults, more than 80% of L5 inferior articular processes fell in the divergent category, with the remainder being parallel (Fig. 8). One individual, a female, had processes that could be classified as convergent, but because they appear to have resulted from a developmental error rather than being a normal variant, they were eliminated from the analysis. Among middle adults without degenerative olisthesis, 25% of the processes were parallel or convergent, and this figure increased to 58% in old adults. A comparison of old adult females with olisthesis vs. those without produced a striking difference (Fig. 8): 87% of the processes in those with olisthesis were convergent or parallel, compared with just 44% in those without olisthesis. Border shifting was determined with reasonable certainty for 87 columns, but 4 were eliminated because separate neural arch was present. The distribution of the remaining 83 columns is shown in Table 1. This distribution produced a 2 value of 13.5 (df ⫽ 2, P ⬍ 0.01). Three of the 6 columns with olisthesis and cranial shift have only 11 thoracic vertebrae, and in one of these, the affected level involved a sacralized lumbar vertebra. A second example of transitional lumbosacral vertebra involvement was observed, but in this case the column was too incomplete to determine the direction of shift. All 3 cases of caudal shift associated with olisthesis involved the sacrocaudal border, a marginal kind of shift located well below the level of the olisthesis and probably irrelevant to the condition. Although the New Mexico collections were not examined systematically for degenerative olisthesis in the cervical region, the condition was observed in 5 individuals: 2 each from Tijeras Canyon and Navajo Reservoir, and 1 from Paako. Three of these 292 C.F. MERBS Fig. 7. Distribution of vertebral disc degeneration in New Mexico Pueblo skeletons relative to age, sex, and level of involvement. Fig. 8. Distribution of inferior articular process shape in New Mexico Pueblo vertebrae relative to degenerative spondylolisthesis. were females: 1 middle adult and 2 old adults, and 2 were old adult males. The C4 –C5 level was affected three times, and C5–C6 twice. All 5 cases were markedly asymmetrical, with the left side more severely affected three times and the right side more affected twice. In each case, olisthesis was accompa- DEGENERATIVE SPONDYLOLISTHESIS 293 TABLE 1. Distribution of New Mexico Pueblo vertebral columns, relative to border shift and degenerative spondylolisthesis With olisthesis Without olisthesis Totals Cranial shift No shift Caudal shift Totals 6 8 14 2 41 43 3 23 26 11 72 83 nied by severe osteoarthritis at the same level, with distinctly greater osteoarthritis on the side of greater olisthesis, but these cases were greatly outnumbered by others with severe cervical osteoarthritis and no olisthesis. The maximum slippage in the cases observed ranged from 2 mm (11%) to 4 mm (22%), with the greater slippage occurring in older individuals. The cervical olisthesis was accompanied by lumbosacral involvement in 3 of these individuals, i.e., the 3 females. One of the columns, of a female from Paako, also exhibited a retrolisthesis (posterior slippage) of 3 mm (0.17) at C1–C2. C2 in this individual presents an unusual anatomy, with the odontoid tilting back and the spinous process angling upward. DISCUSSION AND CONCLUSIONS New Mexico Pueblo skeletons closely resemble modern patients in terms of the age and sex of individuals most likely to develop degenerative olisthesis. They agree that this is a condition of older individuals, absent in young adults, sometimes occurring in middle age, but most likely to be seen in old adults, and females are more likely to be affected than males. The approximate 5:1 ratio observed for the New Mexico skeletons falls between the extremes of 2:1 and 6:1 reported in the clinical literature. Also, the maximum degree of slippage observed here (27%) is similar to that observed clinically. Little information on the overall frequency of degenerative olisthesis is provided by clinical studies. The impression given, however, is that it is a relatively uncommon condition, certainly less frequent than the 41% reported for old adults in this study. Before much significance is read into this, however, the nature of the two samples must be considered. In this study, all skeletons that met the lumbosacral completeness criterion were carefully examined, with olisthesis as little as 2 mm recorded. The same would be true, presumably, for an anatomical collection study or one of autopsy specimens. Subjects ending up in the clinical studies, on the other hand, usually present themselves for medical care with pain threshold as a key factor. Small degrees of olisthesis may not always be associated with pain and thus may not come to the attention of a clinician. It is thus reasonable to assume that for degenerative olisthesis, clinical frequencies are underreported, but by how much it is difficult to say. It is thus difficult to say if the high frequency observed in Fig. 9. Percent of total degenerative spondylolisthesis according to level affected in New Mexico Pueblo skeletons and clinical patients. the New Mexico skeletons has any significance, especially since age and sex are key variables. The results of this study are very different from those obtained from clinical patients in one respect: level of involvement (Fig. 9). Using the study of Rosenberg (1975) as typical, 80% of the olisthesis is expected to occur at level L4 –L5. This stands in sharp contrast to our New Mexico study, where 79% occurred at L5–S1. It should be noted, however, that in both studies, “level” was identified by morphological appearance, and not by number of vertebrae down from the cranium. The relationship between degenerative olisthesis and zygapophysial osteoarthritis appears fairly straightforward: severe osteoarthritis must be present for olisthesis to occur. The two conditions follow the same pattern, increasing in frequency from L3–L4 down to L5–S1 (Fig. 6). Even more convincing is the fact that every case of degenerative olisthesis observed was accompanied by severe zygapophysial osteoarthritis at the affected level, and all examples of grade 4 zygapophyseal osteoarthritis were associated with degenerative olisthesis. In addition, asymmetrical degenerative olisthesis was always accompanied by asymmetrical osteoarthritis, with the side of greater involvement always in agreement. However, the results also show that severe osteoarthritis need not result in degenerative olisthesis, especially in males. Occasionally, males showed high osteoarthritis scores in the L3–S1 level range with no olisthesis; such cases were rare in females. In contrast, there appears to be little causal relationship between degenerative olisthesis and disk degeneration in the lumbosacral region. It was common, for example, to encounter individuals with severe disk degeneration and no degenerative olisthesis. Also, the pattern of disk degeneration is the opposite of that of osteoarthritis and degenerative olisthesis, occurring with greatest frequency and intensity at L3–L4 and decreasing down to L5–S1 (Fig. 7). The one exception to this pattern can be seen in old adult females with degenerative olisthe- 294 C.F. MERBS sis, where the L5–S1 level shows more disk degeneration than expected. All results are consistent with osteoarthritis predisposing to olisthesis, whereas osteophyte formation is an adaptive response to instability in the vertebral column, with olisthesis being one source of instability. This study substantiated what has been observed in the clinical context, that vertebrae involved in degenerative olisthesis are more likely to have inferior articular processes that are parallel to the midline or actually converge toward it, rather than diverge from it, which is generally thought to be the normal condition. The presence of parallel processes, even in young adults with no signs of osteoarthritis, degenerative olisthesis, or any other abnormal condition, suggests that they are part of the normal range of variability of this feature. Although it cannot be ruled out that parallel or convergent processes may predispose an individual to degenerative olisthesis, their increase with age, and particularly their extremely high occurrence in old adult females with degenerative olisthesis (Fig. 8), indicate that they most often develop as part of the degenerative remodeling process that produces olisthesis. Since identifying vertebral border shift requires greater completeness of the column than identifying degenerative olisthesis, and since data on shift were collected only for individuals with olisthesis at the beginning of the study, the sample size of individuals for which both olisthesis and shift is available is relatively small (83). Nevertheless, the results are quite interesting and may provide direction for future work, especially as border shift is thought to have some genetic basis (Merbs, 1974). A 2 value significant at a probability level of greater than 0.01 suggests that a relationship may indeed exist between degenerative olisthesis and cranial shift, but the nature of this relationship is difficult to ascertain. However, if the affected vertebrae were identified based on their distance from the cranium (e.g., C1 ⫽ V1, T1 ⫽ V8, L1 ⫽ V20) rather than their morphological appearance, the New Mexico results would more closely resemble those of clinical studies. In 3 of the 6 cases for which this information is available, the affected levels would be V23–V24 (ordinarily L4 –L5), and not L5–S1 as based on their morphological appearance. Approached in this way, the level of olisthesis in these 6 columns (one with two levels affected) would change from L4/L5 ⫽ 3 and L5/S1 ⫽ 4, to L3/L4 ⫽ 1, L4/L5 ⫽ 5, and L5/S1 ⫽ 1. Add to this the two columns showing no shift, and the distribution becomes equivalent to L3/L4 ⫽ 1, L4/5 ⫽ 7, and L5/S1 ⫽ 1. Even at that, however, thoracolumbar cranial shift cannot account for all of the difference in level of involvement seen in the two kinds of studies. One possible explanation for the difference in level of involvement takes into consideration a contrast between studies involving archaeological skeletons and studies involving living patients. Since degen- erative olisthesis produces stenosis of the neural canal and impingement of the spinal nerves that could result in enough pain to send the victim for medical assistance, the pain factor must be taken into consideration when evaluating clinical studies. Generally, the higher the nerve impingement takes place in the lumbar column, the greater the pain produced. This might explain the higher level of involvement in the column recorded in the clinical literature (L4 –L5) than that observed in this archaeological study (L5–S1). It is possible that individuals with L4 –L5 involvement experienced greater pain than those with L5–S1 involvement, making them more likely to seek medical assistance and thus to end up being cited in the clinical literature. Studies of archaeological specimens, on the other hand, are based on pathological appearance; pain cannot be a factor. Archaeological studies would thus lack any bias for L4 –L5 involvement. Although this suggestion makes some sense, it appears to be contradicted by the study by Farfan et al. (1976) of lumbosacral spines collected at autopsy, and by Rosenberg (1975) of anatomical skeletons. Both studies presumably involved complete ascertainment based on pathological appearance with pain not a factor, but both recorded degenerative olisthesis primarily at the L4 –L5 level. The difference in level of involvement might also be attributed to differences between the Native Americans who were the subjects of this study and the Europeans or European-Americans who made up the majority of subjects of the clinical and anatomical studies. Twin studies suggest that vertebral anatomy, some of which may predispose to degenerative changes, may indeed be inherited. This suggests that the level differences might have some genetic basis. Inherited morphological features may produce subtle differences in the lumbar curve, for example, that may have a significant effect on stress patterns in the lower column which would be difficult if not impossible to determine from bare skeletons. Another possibility is that the differences are essentially behavioral in nature, resulting from activity patterns that are specific to these prehistoric and early historic Native Americans. Determining what these might be, however, is beyond the scope of the present study. Although degenerative olisthesis in the cervical region is represented by just 5 examples in this study, some interesting observations can be made. As with its lumbosacral equivalent, old adult females appear to be at greatest risk. The middle cervical region, specifically C4, C5, and C6, appears at greatest risk, and all the cases observed were markedly asymmetrical, but with no consistency in the side involved. As with its lumbosacral equivalent, the condition appears closely related to osteoarthritis, but not disk degeneration. Lumbosacral olisthesis was also present in the 3 females affected, but not in the 2 males. In one case, cervical olisthesis was accompanied by posterior slippage at C1–C2. DEGENERATIVE SPONDYLOLISTHESIS ACKNOWLEDGMENTS I acknowledge the valuable assistance provided by my wife, Barbara Merbs, who served as my recorder for the data collected in this study. I also thank the people who provided valuable assistance at the various institutions that curated the New Mexico collections, particularly Diane Hawkey and Chelsea Klein at Arizona State University, Rose Tyson at the San Diego Museum of Man, Joseph Powell at the Maxwell Museum of Anthropology, and Donald Ortner and David Hunt at the Smithsonian Institution. LITERATURE CITED Aland C, Rineberg BA, Malberg M, Fried SH. 1986. Fracture of the pedicle of the fourth lumbar vertebra associated with contralateral spondylolysis. J Bone Joint Surg [Am] 68:1454 –1455. Albers VL, Yochum TR. 1980. Reactive sclerosis of a pedicle due to unilateral spondylolysis. ACA J Chiroprac 14:56 –59. Barnes E. 1994. Developmental defects of the axial skeleton in paleopathology. Niwot: University Press of Colorado. Bolesta MJ, Bohlman HH. 1989. Degenerative spondylolisthesis. Instr Course Lect 38:157–165. Brown MD, Lockwood JM. 1983. Degenerative spondylolisthesis. Instr Course Lect 32:162–169. Buikstra JE, Ubelaker DH. 1994. Standards for data collection from human skeletal remains. Arkansas Archeological Survey Research Series No. 44. Fayetteville, AR. Deburge A, Mazda K, Guigui P. 1995. Unstable degenerative spondylolisthesis of the cervical spine. J Bone Joint Surg [Br] 77:122–125. Elster AD, Jensen KM . 1985. Computed tomography of spondylolisthesis: patterns of associated pathology. J Comput Assist Tomogr 9:867– 874. Epstein JA, Epstein BS, Lavine LS, Carras R, Rosenthal AD, Sumner P. 1973. Lumbar nerve root compression at the intervertebral foramina caused by arthritis of the posterior facets. J Neurosurg 39:362–369. Epstein JA, Epstein BS, Lavine LS, Carras R, Rosenthal AD. 1976. Degenerative lumbar spondylolisthesis with an intact neural arch (pseudospondylolisthesis). J Neurosurg 44:139 – 147. Epstein NE, Epstein JA, Carras R, Lavine LS. 1983. Degenerative spondylolisthesis with an intact neural arch: a review of 60 cases with an analysis of clinical findings and the development of surgical management. J Neurosurg 13:555–561. Farfan HF, Osteria V, Lamy C. 1976. The mechanical etiology of spondylolysis and spondylolisthesis. Clin Orthop 117:40 –55. Fitzgerald JAW, Newman PH. 1976. Degenerative spondylolisthesis. J Bone Joint Surg [Br] 58:184 –192. Gunzburg R, Wagner J. 1988. Degenerative spondylolisthesis with unilateral spondylolysis. A case report. Int Orthop 12: 139 –141. Herkowitz HN. 1995. Degenerative lumbar spondylolisthesis. Spine 20:1084 –1090. Herkowitz HN, Kurz LT. 1991. Degenerative lumbar spondylolisthesis with spinal stenosis. J Bone Joint Surg [Am] 73:802– 808. 295 Herron LD, Trippi AC. 1989. L4 –L5 degenerative spondylolisthesis: the results of treatment by decompressive laminectomy without fusion. Spine 14:534 –538. Junghanns H. 1931. Spondylolisthesen ohne Spalt im Zwischengelenkstück (“Pseudospondylolisthesen”). Arch Orthop Unfall Chir 29:118 –127. Kornberg M. 1988. Spondylolisthesis with unilateral pars interarticularis defect and contralateral facet joint degeneration. Spine 13:712–713. Laurent LE, Österman K. 1976. Operative treatment of spondylolisthesis in young patients. Clin Orthop 117:85–91. Laus M, Tigani D, Alfonso C, Giunti A. 1992. Degenerative spondylolisthesis: lumbar stenosis and instability. Chir Organi Mov 77:39 – 49. Macnab I. 1950. Spondylolisthesis with an intact neural arch: the so-called pseudo-spondylolisthesis. J Bone Joint Surg [Br] 32: 325–333. Merbs CF. 1974. The Effects of cranial and caudal shift in the vertebral columns of Northern populations. Arctic Anthropol [Suppl] 6:12–19. Merbs CF. 1996. Spondylolysis and spondylolisthesis: a cost of being an erect biped or a clever adaptation? Yrbk Phys Anthropol 39:201–228. Newell RLM. 1995. Spondylolysis: an historical review. Spine 20:1950 –1956. Newman PH. 1955. Spondylolisthesis, its cause and effect. Ann R Coll Surg 16:305–323. Palmer PES, Stadalnick R, Arnon S. 1984. The genetic factor in cervical spondylosis. Skeletal Radiol 11:178 –182. Postacchini F, Perugia D. 1991. Degenerative lumbar spondylolisthesis. Part I: etiology, pathogenesis, pathomorphology, and clinical features. Ital J Orthop Traumatol 17:165–173. Potter RM, Norcross JR. 1954. Spondylolisthesis without isthmus defect. Radiology 63:678 – 684. Robert. 1855. Eine eigenthümliche angeborene Lordose, wahrscheinlich bedingt durch eine Verschiebung des Körpers des letzten Lendenwirbels auf die vordere Fläche des ersten Kreuzbeinwirbels (Spondylolisthesis Kilian), nebst Bemerkungen über die Mechanik dieser Beckenformation. Monatsschr Geburtsk Frauenkr (Berlin) 5:81–94. Rosenberg NJ. 1975. Degenerative spondylolisthesis: predisposing factors. J Bone Joint Surg [Am] 57:467– 474. Ryan MD. 1994. L4 –5 degenerative spondylolisthesis in monozygous twins. Spine 19:985–986. Sanderson PL, Fraser RD. 1996. The influence of pregnancy on the development of degenerative spondylolisthesis. J Bone Joint Surg [Br] 78:951–954. Sato K, Wakamatsu E, Yoshizumi A, Watanabe N, Irei O. 1989. The configuration of the laminas and facet joints in degenerative spondylolisthesis. A clinicoradiologic study. Spine 14:1265–1271. Sherman FC, Wilkinson RH, Hall JE. 1977. Reactive sclerosis of a pedicle and spondylolysis in the lumbar spine. J Bone Joint Surg [Am] 59:49 –54. Stewart TD. 1935. Spondylolisthesis without separate neural arch (pseudospondylolisthesis of Junghanns). J Bone Joint Surg 17:640 – 648. Takahashi K, Kitahara H, Yamagata M, Murakami M, Takata K, Miyamoto K, Mimura M, Akahashi Y, Moriya H. 1990. Longterm results of anterior interbody fusion for treatment of degenerative spondylolisthesis. Spine 15:1211–1215. Waldron T. 1992. Unilateral spondylolysis. Int J Osteoarch 2:177–181.