Secondary Logo

Journal Logo


Does Lumbar Facet Arthrosis Precede Disc Degeneration?

A Postmortem Study

Eubanks, Jason David MD*; Lee, Michael J MD; Cassinelli, Ezequiel MD*; Ahn, Nicholas U MD*

Author Information
Clinical Orthopaedics and Related Research: November 2007 - Volume 464 - Issue - p 184-189
doi: 10.1097/BLO.0b013e3181583d4e
  • Free


The functional spinal unit consists of a three-joint complex composed of the intervertebral disc and the facet joints. As such, degenerative changes in either the disc or the facet joints, whether in the form of disc desiccation, facet cartilage deterioration, vertebral rim or facet osteophytosis (VRO and FO, respectively), or spondylolisthesis, affect normal spinal mechanics and play an important role in progressive degenerative changes.4,11 It generally is believed the degenerative process begins in the intervertebral disc,2,5,14,20 where desiccation in the disc leads to loss of disc space height and subsequent increase in the pressure experienced by the facet joints leading to facet arthrosis, including FO.3 These conclusions, however, have been reached in the absence of definitive population studies, basing conclusions on small samples.2,5,6,18 For instance, the study by Butler et al2 includes only 58 individuals, whereas another, more recent study bases its conclusions on a sample size of 14 patients.5

In the process of studying skeletal lumbar spines for another purpose, it became apparent to us that some young individuals showed marked FO in the absence of bony evidence of VRO. Fujiwara et al5 independently highlighted this seemingly anomalous finding as well, noting, in 20% of degenerative spines, facet arthrosis appears to precede disc degeneration. This intriguing finding led the authors to question the commonly held belief that disc degeneration necessarily precedes facet arthrosis. In an era in which disc replacement surgery is being purported to save motion and avert further degeneration and, yet, is contraindicated in the presence of facet arthrosis,13,24 this issue has important implications.

Given our preliminary observations, we therefore hypothesized FO precedes VRO in the lumbar spine. Furthermore, given the known ubiquity of disc degeneration in the general population,15 we surmised, as the age of the specimens increased, the evidence of VRO would begin to outpace the early predominance of bony, arthritic changes in the facets.


From the more than 3000 treated and dried specimens contained in the Hamann-Todd Osteological Collection, Cleveland, OH, we chose 647 skeletal lumbar spines for examination. These specimens were chosen from the entire collection depending on their collective proximity within two rows of the collection and their ease of access (no ladder necessary to take them off the shelves). The selection was random insofar as the specimens in the collection are not arranged in any particular order. They represent individuals who died in Cleveland, OH, between 1893 and 1938. We included 475 men and 172 women ranging in age from 17 to 93 years, with a nearly bell-shaped distribution of ages for men and a larger proportion of younger ages in the female specimens (Table 1). Two hundred eighty-three specimens were of African-American ancestry; the remainder of the specimens had Caucasian ancestry. A power analysis was performed to determine the number of samples needed to arrive at a significant difference between the incidences of facet and disc arthrosis. Using a value of p < 0.05 and a power of 80%, we estimated, for a 20% difference between the incidences of facet and end plate arthrosis, a sample size of 91 would be needed.

Specimen Breakdown by Age and Gender

The gross specimens were examined subjectively by one examiner (JDE) for evidence of arthrosis. At the time of examination, specific attention was not paid to the coexisting presence or absence of spondylolysis nor was the examiner blinded to the age of the specimens. Some degree of VRO always accompanies disc degeneration.8,20 Using this definition, we evaluated inter-vertebral disc degeneration based on the presence of VRO, a commonly used marker for the presence of degenerative disease.1 Similarly, the evaluation and grading of FO we used were modeled after previously published work.22

We evaluated FO and VRO using a grading scale ranging from 0 to 4, as suggested in the guidelines presented by Kettler and Wilke10: Grade 0 = normal facet joints or vertebral end plates; Grade 1 = mild arthrosis with osteophytic reaction involving as much as 50% of the facet joint or vertebral end plates; Grade 2 = moderate arthrosis with osteophytic reaction involving 50% to 100% of the facet joint or vertebral end plates; Grade 3 = severe arthrosis with osteophytic reaction involving 100% of the facet joints or vertebral end plates with hypertrophic osteophytes bridging the joint space; and Grade 4 = complete ankylosis. This grading system models that described by Videman et al,22 which has no published interobserver or intraobserver reliability rating. Kettler and Wilke10 noted (in their specific review of grading systems of lumbar disc and facet degeneration) very few macroscopic grading systems have published reliability measurements. A notable exception is Thompson et al,19 who reported an intraobserver reliability of 87% to 91% and an interobserver reliability of 61% to 88%. We assume the intraobserver reliability in our study's macroscopic, morphologic evaluation of disc and facet degeneration approximates the intraobserver measurements noted by these authors.19 Finally, we evaluated the right and left facet joints at each lumbar level and graded them independently.

We analyzed the prevalence and the degree (Grade 0-4) of FO and VRO by decade at each lumbar level using analysis of variance. We then used Fisher's exact tests, in a post hoc fashion, to evaluate statistically significant differences in the prevalence and degree of FO versus VRO at each lumbar level in each age category. For every decade examined, the prevalence of FO was compared with the prevalence of VRO to determine the predominant, if any, site (facet versus disc) and level (L1-L2 to L5-S1) of degeneration for any given age category.


Specimens younger than 30 years had a greater prevalence (p < 0.05) of FO at the L1-L2 and L2-L3 levels than corresponding VRO. These level-specific findings stood in contrast to the other age categories in which we observed no differences between FO and VRO at these levels (Figs 1 and 2; Tables 2 and 3). In the specimens 30 to 39 years of age, an increased (p < 0.02) prevalence of FO as compared with VRO appeared at L4-L5. Again, this level-specific finding was particular to those specimens in their fourth decade of life when compared with all other age categories. Considerable FO, in the absence of VRO, was noted in specimens as young as 21 years (Fig 3).

Degree, Level, and Side-specific Facet Osteophytosis per Decade*
Degree and Level-specific Vertebral Rim Osteophytosis per Decade*
Fig 1
Fig 1:
This figure depicts the prevalence of lumbar FO in relation to age and lumbar level. There is an initial increase in prevalence of FO in the third and fourth decades followed by a more gradual increase in later decades. In comparison to Figure 2, this initial increase has a steeper slope.
Fig 2
Fig 2:
This figure shows the prevalence of lumbar VRO in relation to age and lumbar level. In comparison to Figure 1, there is a more steady and gradual increase of VRO in the early decades followed by an increased rate of degeneration beginning in the fifth decade, representing the age at which point VRO begins to outpace FO.
Fig 3A
Fig 3A:
B. (A) This is an example of early, multilevel FO in an African-American woman who died at the age of 21 years. (B) In contrast to the multilevel FO evident in illustration A, in this image, it is possible to observe the absence of corresponding VRO. In this young patient, bony evidence of arthrosis in the facets preceded bony evidence of degeneration in the discs.

By the fifth decade of life, however, this trend for the early predominance of FO changed. Whereas in the specimens younger than 40 years FO appeared to precede VRO, by the fifth decade, the appearance of VRO began to out-pace FO (Figs 1 and 2). Specimens aged 40 to 49, 50 to 59, 60 to 69, and older than 70 years showed more (p < 0.05) marginal osteophytosis than FO except at L4-L5 and L5-S1. The level-specific distributions of FO and VRO within age groups closely paralleled one another with the important exception of the findings previously noted at L1-L2 and L2-L3 for the younger than 30 years specimens and the L4-L5 findings in the younger than 40 years specimens (Figs 1 and 2). After the comparative early spike in FO, followed by the shift to predominant VRO in the fifth decade, by the sixth decade of life, evidence of arthrosis in the functional units of the lumbar spine was nearly ubiquitous, with overall greater prevalence in the degenerative discs (Figs 1 and 2).


Based on preliminary skeletal observations of early FO in young individuals, we challenge the commonly held belief that disc degeneration necessarily precedes facet arthrosis in the lumbar spine. Through an examination of 647 skeletal specimens, we found lumbar FO appears early in the degenerative cascade with greater prevalence in younger individuals than VRO. With continued aging, however, VRO appears to predominate after the fifth decade of life.

In exploring these questions, the limitations of our study are inherent in the study design. This is an observational, skeletal study in which arthrosis was evaluated solely on the bony evidence of arthritic change. We assumed bone degeneration in both locations equally reflected the progression of degeneration in the soft tissues even if the former appeared later in the process. With this assumption, we cannot conclusively answer the question of whether discs truly begin to degenerate before facets. This was not the intent of this analysis but, rather, to show evidence challenging the commonly held belief that facet arthrosis necessarily follows disc degeneration. Ideally, a prospective cohort study following a large group of patients with serial imaging studies and autopsy analysis after death would provide the most satisfactory answers to our study's questions. However, such a study would be logistically difficult and financially prohibitive. Contrary to our assumption, it could be argued the absence of MRI or fresh autopsy specimens in our study misses the early degenerative changes in the discs that precede the vertebral rim changes used as a reference in our study. Perhaps the early FO we found was accompanied by earlier, unmeasurable changes in the discs. Our findings allow for this possibility. In either case, the facets are, at a minimum, degenerating simultaneously and not merely after VRO. Additionally, the results of Vernon-Roberts and Pi rie20 remind us VRO appears with all degenerate discs regardless of the specific degree of disc degeneration. During examination of the specimens, special attention was not given to the concomitant presence of spondylolysis. It is possible the presence of this process might have influenced the resultant degenerative changes, specifically in the younger individuals. Finally, the evaluation of arthrosis was performed subjectively without measuring reliability. However, the analysis had the uniformity of one observer and used a grading system that was modeled on a previously published report22 and conformed to recently recommended guidelines10 for analysis.

With these considerations in mind, there is compelling evidence that degenerative changes in the intervertebral discs and facet joints parallel the aging process.2,12,14-17 What is less clear, however, is whether degeneration in the discs necessarily precedes arthrosis in the facet joints or whether the two processes proceed simultaneously or, furthermore, whether in some individuals facet arthrosis may precede degenerative changes in the discs. Clinical2,5 and experimental studies11 have suggested the degenerative changes in the lumbar spine begin in the intervertebral discs. The clinical studies2,5 have used MRI and/or computed to mography (CT) imaging in the examination of disc degeneration and facet arthrosis. Although MRI and CT imaging can be effective tools for assessing degenerative changes in the intervertebral disc and facet joints, the conclusions drawn from these studies sometimes have been based on small sample sizes and/or incomplete analysis.2,5,16 For example, in their classic article, Butler et al2 included scans from 68 patients and then chose not to examine L1-L2 and L2-L3 in their analysis of facet arthrosis. In our study and that of Kelsey and White,9 these were the levels in which a difference occurred in patients younger than 30 years.

In addition to MRI and CT analysis, gross morphologic examination of arthrosis in the lumbar spine has been used. Miller et al15 assessed the degree of disc degeneration in 600 autopsy specimens using a grading system described by Nachemson.17 Using a grading scale from 1 to 4, Miller et al15 macroscopically examined and categorized degenerative changes in lumbar intervertebral discs. They reported evidence of degenerative changes earlier in men than in women and most commonly at the L3-L4 and L4-L5 levels, with these changes first becoming apparent in the second decade of life (age 10-19 years). Only the Grade 4 discs in their study included evidence of VRO, and these changes were not recognized until the fourth decade.15 In contrast, we found evidence of marginal osteophytosis as early as the third decade of life.

Although the work of Miller et al15 helps us better understand the process of degeneration in the intervertebral disc, no mention was made in their study of facet arthrosis. In 20% of degenerative lumbar spines, facet arthrosis appears on MRI to precede disc degeneration.5 Some might question these results because they are based on MRI findings and numerous authors suggest CT is more accurate than MRI in assessing facet arthosis.7,16 Some who purport the genesis of degeneration in the intervertebral disc2,5 have done so through the additional implementation of CT scanning in the assessment of FO. However, Fujiwara et al5 and Weishaupt et al23 found MRI greater than 90% accurate in assessing facet arthrosis when compared with CT. Regardless, we provide gross morphologic evidence in a large sample suggesting FO precedes the bony vertebral rim changes associated with intervertebral VRO at the L1-L2 and L2-L3 levels during the third decade of life. Similarly, FO preceded VRO at the L4-L5 level during the fourth decade of life.

Videman et al,22 in a study of 86 male cadavers, reported evidence suggesting facet arthrosis is related to more than 3 years of heavy physical work before the age of 20 years. In our study, the specimens represented an urban population from an industrial city around the turn of the century when it was possible the nature of their work during early life loaded the posterior elements of the spine in such a way as to incite early arthrosis in facet joints. However, Videman et al21 noted osteophytes are associated with end plate degeneration and the degree of vertebral osteophytosis also was related to heavy work.22 One would expect, therefore, VRO would parallel the facet findings in our study, but it did not.

Additionally, Videman et al22 noted a J-shaped curve for facet arthrosis, indicating even a person who performed no physical work before age 20 was not without risk of having facet arthrosis develop. Although they did not examine the lumbar spines of women, this J curve might help explain our study's findings in which there was no difference between men and women in the temporal relationship of degeneration between the intervertebral discs and the facet joints.

Although we found bony degenerative changes beginning in the facet joints, we further noted the predominance of degenerative changes after the fifth decade appearing in the vertebral end plates. It seems the degenerative process appears early in the facet joints, but this process maintains a slow and progressive course. In the intervertebral discs, however, there was a marked increase in the degenerative changes noted beginning in the fifth decade, a process that outpaced the changes in the corresponding facet joints for the remainder of the individuals' lives (Figs 1 and 2). Again, it is unclear why there is this change in trend of the degenerative process. It could be, by the fifth decade, the biochemical, vascular, and biomechanical milieu of the intervertebral disc has reached a critical point beyond which subsequent degenerative changes occur at a heightened speed. The presence of existing arthrosis in the facet joints may accelerate this process because loss of normal anatomy can compromise disc mechanics,4,11 with possible further potentiation of disc degeneration.

We examined the natural history of degenerative changes in the functional unit of the lumbar spine. Through a postmortem, skeletal analysis, it seems bony evidence of FO appears early in the degenerative process, preceding the reactive VRO associated with degeneration of intervertebral discs. Once the facets start to deteriorate with age, the discs rapidly degenerate, with VRO developing more progressively than continued FO. These results challenge the belief that the process of degeneration necessarily begins in the disc; rather, it appears VRO progresses more rapidly in later years, but FO appears early with a predominance in younger individuals. This evidence of early facet degeneration questions the commonly held notion that facet arthrosis represents a mere sequela of disc degeneration.


We thank Lyman Jemella and the staff at the Cleveland Natural History Museum for assistance in facilitating our anatomic review of the Hamann-Todd specimens.


1. Battié MC, Videman T. Lumbar disc degeneration: epidemiology and genetics. J Bone Joint Surg Am. 2006;88 (suppl 2):3-9.
2. Butler D, Trafimow JH, Andersson GBJ, McNeill TW, Huckman MS. Discs degenerate before facets. Spine. 1990;15:111-113.
3. Dunlop RB, Adams MA, Hutton WC. Disc space narrowing and the lumbar facet joints. J Bone Joint Surg Br. 1984;66:706-710.
4. Fujiwara A, Lim TH, An HS, Tanaka N, Jeon CH, Andersson GB, Haughton VM. The effect of disc degeneration and facet joint osteoarthritis on the segmental flexibility of the lumbar spine. Spine. 2000;25:3036-3044.
5. Fujiwara A, Tamai K, Yamato M, An HS, Yoshida H, Saotome K, Kurihashi A. The relationship between facet joint osteoarthritis and disc degeneration of the lumbar spine: an MRI study. Eur Spine J. 1999;8:396-401.
6. Gotfried Y, Bradford DS, Oegema TR Jr. Facet joint changes after chemonucleolysis-induced disc space narrowing. Spine. 1986;11:944-950.
7. Grenier N, Kressel HY, Schiebler ML, Grossman RI, Dalinka MK. Normal and degenerative posterior spinal structures: MR imaging. Radiology. 1987;165:517-525.
8. Ingelmark BE, Moller-Christensen V, Brinch O. Spinal joint changes and dental infections. Acta Anat Suppl (Basel). 1959;38:1-60.
9. Kelsey JL, White AA3rd. Epidemiology and impact of low-back pain. Spine. 1980;5:133-142.
10. Kettler A, Wilke HJ. Review of existing grading systems for cervical or lumbar disc and facet joint degeneration. Eur Spine J. 2006;15:705-718.
11. Kirkaldy-Willis WH, Farfan FH. Instability of the lumbar spine. Clin Orthop Relat Res. 1982;165:110-123.
12. Lewin T. Osteoarthritis in lumbar synovial joints: a morphologic study. Acta Orthop Scand Suppl. 1964;73:1-112.
13. Lin EL, Wang JC. Total disk arthroplasty. J Am Acad Orthop Surg. 2006;14:705-714.
14. Lipson SJ, Muir H. Experimental intervertebral disc degeneration: morphologic and proteoglycan changes over time. Arthritis Rheum. 1981;24:12-21.
15. Miller JA, Schmatz C, Schultz AB. Lumbar disc degeneration: correlation with age, sex, and spine level in 600 autopsy specimens. Spine. 1988;13:173-178.
16. Modic MT, Pavlicek W, Weinstein MA, Boumphrey F, Ngo F, Hardy R, Duchesneau PM. Magnetic resonance imaging of intervertebral disk disease. Radiology. 1984;152:103-111.
17. Nachemson AL. Lumbar intervertebral pressure: experimental studies on post-mortem material. Acta Orthop Scand Suppl. 1960;43:1-104.
18. Oegema TR Jr, Bradford DS. The inter-relationship of facet joint osteoarthritis and degenerative disc disease. Br J Rheumatol. 1991;30 (suppl 1):16-20.
19. Thompson JP, Pearce RH, Schechter MT, Adams ME, Tsang IK, Bishop PB. Preliminary evaluation of a scheme for grading the gross morphology of the human intervertebral disc. Spine. 1990;15:411-415.
20. Vernon-Roberts B, Pirie CJ. Degenerative changes in the intervertebral discs of the lumbar spine and their sequelae. Rheumatol Rehabil. 1977;16:13-21.
21. Videman T, Battié MC, Gill K. Manninen H, Gibbons LE, Fisher LD. Magnetic resonance imaging findings and their relationship in the thoracic and lumbar spine: insights into the etiopathogenesis of spinal degeneration. Spine. 1995;20:928-935.
22. Videman T, Nurminen M, Troup JDG. 1990 Volvo Award in clinical sciences: lumbar spinal pathology in cadaveric material in relation to history of back pain, occupation and physical loading. Spine. 1990;15:728-740.
23. Weishaupt D, Zanetti M, Hodler J, Boos N. MR imaging of the lumbar spine: prevalence of intervertebral disk extrusion and sequestration, nerve root compression, end plate abnormalities, and osteoarthritis of the facet joints in asymptomatic volunteers. Radi ology. 1998;209:661-666.
24. Wong DA, Annesser B, Birney T, Lamond R, Kumar A, Johnson S, Jatana S, Ghiselli G. Incidence of contraindications to total disc arthroplasty: a retrospective review of 100 consecutive fusion patients with a specific analysis of facet arthrosis. Spine J. 2007;7:5-11.
© 2007 Lippincott Williams & Wilkins, Inc.