Over the last 90 years, numerous reports have concluded that acetabular dysplasia will cause osteoarthritis to develop in an otherwise normal individual, if the patient is followed for a sufficient period of time.1–8 Most of these studies are Level IV or V, using the JBJS rating system, with no Level I evidence available. However, if evidence is the thing that results in certainty, the evidence available is powerful, because most orthopaedic surgeons are certain that acetabular dysplasia precipitates osteoarthritis. We will analyze some well-known studies that illuminate the link between hip dysplasia and hip osteoarthritis to investigate the origin of this certainty.
EARLY EVIDENCE FOR A LINK BETWEEN HIP DYSPLASIA AND OSTEOARTHRITIS
Convincing reports of the connection between hip dysplasia and hip dysfunction and osteoarthritis begin with Le Scolan, in 1924.8 In her report, she described a large group of patients who appeared to have abnormal femoral-acetabular relationships and dysfunction. She described 5 different patterns of dysplasia, but did not quantify the dysplasia in any way. Fifteen patients were from 5 to 15 years of age. They tended to limp and have pain with exertion. Thirteen patients were over 16 years of age and 6 had osteoarthritis.
THE CLASSIC STUDY DEFINING ACETABULAR DYSPLASIA AND LINKING IT TO THE DEVELOPMENT OF OSTEOARTHRITIS
In 1939, Gunnar Wiberg published ‘Studies on Dysplastic Acetabula and Congenital Subluxation of the Hip Joint with Special Reference to the Complication of Osteoarthritis.”8 In this volume, he quantified acetabular dysplasia using a radiographic measurement he developed, the center-edge (CE) angle (Fig. 1). His report convinced the orthopaedic community that acetabular dysplasia caused osteoarthritis of the hip joint.
Wiberg established the CE angle as a reasonable way to describe hip dysplasia by first obtaining control data and then following patients that he defined as dysplastic.
He found 100 healthy, asymptomatic volunteers, 50 males and 50 females, between the age of 20 and 35 years and obtained hip x-rays. These included an AP and a frog lateral view (Lauenstein’s view) of the pelvis taken supine. He measured the CE angles. They ranged from 20 to 47 degrees. He postulated that a CE angle of under 20 degrees was abnormal and would cause hip osteoarthritis.
His hypothesis was that a smaller than normal acetabulum, as suggested by a small CE angle, would lead to increasing pressure between the femoral head and the acetabulum. This increasing pressure would precipitate osteoarthritis when it reached a critical level. He suggests that a hip with a CE angle of under 20 degrees had reached this critical level.
In an effort to test the validity of his hypothesis, he identified 44 arthritis-free adults with CE angles under 20 degrees. He was able to follow 18 patients (19 hips) from 4 to 29 years. All developed osteoarthritis. The lower the CE angle, the faster osteoarthritis supervened. The linear relationship between the CE angle and the rate of appearance of arthritis seemed reasonable, given the concept.
Other measurements have been proposed and other populations sampled in an effort to investigate the relationship between acetabular dysplasia and osteoarthritis. In the end, no single study has had the impact of Wiberg’s study in linking acetabular dysplasia and osteoarthritis. No single measurement has replaced the CE angle as the gold standard for describing acetabular dysplasia.
A STUDY DIFFERENTIATING BETWEEN STABLE AND UNSTABLE ACETABULAR DYSPLASIA AND THE IMPORTANCE OF THE DISTINCTION IN PREDICTING THE AGE OF ONSET OF OSTEOARTHRITIS
In 1983, Cooperman et al1 reexamined the Wiberg data. They noted that 7 of the 19 hips that Wiberg reported on were subluxated at presentation. Shenton’s line was broken by at least 5 mm in all of these cases, although no evidence of osteoarthritis was present at presentation. These patients tended to have low CE angles and all developed arthritis by the age of 42 years. Cooperman and colleagues concluded that hip subluxation was a very bad prognostic sign. The authors noted that after hips with subluxation were removed from the analysis, the magnitude of the CE angle was not a good predictor of the rate of appearance of osteoarthritis, that is, the patients with the lowest CE angles did not always develop osteoarthritis most rapidly.
To further evaluate prognosis in stable, dysplastic hips, Cooperman and colleagues identified 20 adults (average age, 43 y) with 32 stable, dysplastic hips. All had intact Shenton’s lines, CE angles under 20 degrees, and all were arthritis-free hips.
The patients were followed for an average of 22 years. In addition to the CE angle, the authors measured the acetabular angle of Sharp9 as well as 2 measurements suggested by Heyman and Herndon,9 acetabular depth, and the extrusion index (Fig. 1).
All of the patients developed osteoarthritis, but these 4 commonly used descriptors of hip dysplasia could not be used to predict the rate at which osteoarthritis would supervene. The authors emphasized that osteoarthritis may take decades to appear in the stable, dysplastic hip and that commonly used x-ray measurements are not reliable predictors of the age at which arthritis will be first noted. They confirmed that the CE angle is quite reliable in predicting the eventual development of osteoarthritis of the hip in patients with CE angles under 200 degrees.
STUDIES FURTHER LINKING STABLE AND UNSTABLE ACETABULAR DYSPLASIA TO OSTEOARTHRITIS
In 2005, Jacobsen and colleagues2,3 reviewed upright pelvis x-rays of 4151 adults living in Copenhagen. These individuals were randomly selected from a large cohort being followed in the third Copenhagen City Heart Study. Average age of the patients with hip x-rays was 61 years (range, 22 to 93 ). Shenton’s line was intact in all the hips.
They looked for factors that related to joint space narrowing, an early sign of osteoarthritis. They used the roentgenographic markers of hip dysplasia used by Cooperman and colleagues (Fig. 1). The CE angles ranged from 6 to 67 degrees. In addition, they recorded age, sex, BMI, occupation, and pack-years of smoking. They excluded individuals with previous hip surgery (except for total joint arthroplasty), previous hip fracture, treatment for a childhood hip disorder, or those with an inflammatory arthropathy at any joint. This resulted in a study group of 3859 individuals.
The significant factors for risk of joint space narrowing were a CE angle of under 20 degrees in men and women as well as age for women. Joint space narrowing was seen in women as a function of age alone, even when hip dysplasia was factored out.
At the end of the article, the authors included some random x-rays illustrating older arthritis-free participants with what the authors termed “moderate hip dysplasia.” These included a 65-year-old male with a CE angle of 16 degrees, a 65-year-old female with a CE of 17 degrees, a 52-year-old female with a CE angle of 15 degrees, and a 60-year-old male with a CE angle of 15 degrees (Fig. 2). This was Level II study.
In a separate article in 2005, Jacobsen et al10 reported on some dysplastic hips from this study. There were 247 hips (3.2%) with a CE angle of 20 degrees or less. They followed 81 stable, dysplastic hips (27 men and 54 women) for 10 years until the men reached an average age of 48 years and the women reached an average age of 53 years. They also followed 136 control subjects chosen from the same population. There was no significant difference in the rate of joint space narrowing between the dysplasia and the control groups. These patients with stable dysplasia were quite stable.
In 1995, Murphy et al,5 published a report on 286 patients who had received unilateral total hip arthroplasty for what the authors suspected was osteoarthritis secondary to hip dysplasia. They then characterized the opposite hip in an effort to describe factors that related to outcome in patients with hip dysplasia. This, of course, supposes that a reasonable number of the hips opposite a dysplastic hip are also dysplastic. The literature suggests that this is so.11–13
One hundred fifteen patients developed severe arthritis opposite their total hip arthroplasty. One hundred twenty-five patients had less than severe arthritis and were younger than 65 years of age.
Forty-six patients were over the age of 65 years with an opposite hip that did not have severe arthritis.
Forty-three patients had good records. None had subluxated hips. The average CE angle in this group was 340 degrees. The range was 16 to 49 degrees. The range is of great interest, as no patient who was free of severe arthritis and over 65 years of age, had a CE angle of under 16 degrees.
In 2010, Gosvig et al14 reviewed the same x-rays on 4151 citizens of Copenhagen that Jacobsen and colleagues reported on. They made a note of patients with a CE angle either under 20 or over 45 degrees and characterized these as either “dysplastic hips” or “deep hips.” They identified femora with pistol-grip deformities too.6 They analyzed 4 risk factors in an effort to correlate hip shape with joint space narrowing. They looked at: (1) acetabular dysplasia (CE angle <20 degrees); (2) a deep acetabulum (CE angle >45 degrees); (3) pistol-grip deformity; and (4) a deep acetabulum in concert with a pistol-grip deformity.
They excluded from consideration the patients who had had total hip arthroplasty, various childhood hip diseases, rheumatoid arthritis, unreadable x-rays, as well as those with excessive rotation. After exclusions, they analyzed 3620 individuals while using the same cohort as Jacobsen and colleagues, who reported on 3859 of these patients after their exclusions.
Gosvig and colleagues concluded that acetabular dysplasia was not a significant risk factor for joint space narrowing, nor was female sex a significant risk factor. The pistol-grip deformity and a deep socket were risk factors for the development of joint space narrowing. This was a Level II study.
A convincing Level I study assessing the relationship between acetabular dysplasia and osteoarthritis does not exist and is unlikely to appear, as this would require a huge cohort of individuals followed from skeletal maturity to senescence. Such a prospective study would require at least 50 years to complete. This time frame is prohibitive. In addition, at this time, the Orthopaedic community would not favor such a study, as we believe that a link between acetabular dysplasia and osteoarthritis is well established.
Jacobsen and colleagues and Gosvig and colleagues, give us large, Level II studies with different conclusions about the relationship between acetabular dysplasia and joint space narrowing. They used the same data set of 4151 pelvis x-rays. Each excluded certain individuals from analysis.
Gosvig and colleagues analyzed and reported upon 97 fewer men and 142 fewer women than Jacobsen and colleagues. Gosvig factored in femoral deformity in a way that Jacobsen did not, but did not evaluate the hips opposite total hip arthroplasty, that Jacobsen assessed.
These Level II studies present a cautionary tale, reminding us that the level of evidence does not judge the rational for the choices the authors make concerning exclusions or other aspects of their studies. Exclusions created different cohorts for these authors and resulted in different conclusions. This can be confusing. Level of evidence is a valuable tool for assessing methodology, but it does not assign power to conclusions. Although it may be the case that a potent study cannot flow from bad methodology, good methodology does not guarantee binding answers to important questions.
The Orthopaedic community is unlikely to change its mind about the relationship between acetabular dysplasia and osteoarthritis of the hip based on Gosvig’s interpretation of his sample for 4 reasons.
First, the P value linking joint space narrowing to acetabular dysplasia in Gosvig’s study was 0.053; missed it by 0.003, that is pretty close. Second, orthopaedic surgeons believe that the concept relating acetabular dysplasia to osteoarthritis is valid. Biomechanical forces can cause arthritis. A smaller than normal acetabulum will have a decreased area available for pressure transfer between the acetabulum and the femoral head, increasing the load on articular cartilage. A steeper than normal acetabulum will create more shear than normal at the femoral-acetabular interface. The mechanics of these situations have been well described by Pauwels and colleagues.15–20 At some critical point, greater than normal amounts of load/shear will result in significant cartilage damage. This damage will result in osteoarthritis. Individual biological factors and environmental factors may dictate that 1 person can tolerate more dysplasia than another, but at some level of dysplasia, all hips will fail.
Third, we have powerful clinical studies, a few of which we have reviewed, which suggest the linkage between acetabular dysplasia (as defined by the CE angle of Wiberg) and osteoarthritis. The CE angle is very attractive because it is easy to create, reproducible (both intraobserver and interobserver14), and the user can easily see that a small CE angle equates with a small or more vertical socket, which is the quality of the acetabulum that the angle was invented to represent.
Fourth, our casual experience suggests that quite dysplastic hips do not exist in the arthritis-free elderly. If we saw a few, our confidence might be shaken. At this time, they are unreported.
Now, there are some completely dislocated, arthritis-free hips at 65 years of age, but that is another subject.19–21 There is simply no case report or study reporting on a card-carrying Medicare patient who is arthritis-free with a subluxated hip, (Fig. 3) or one with a CE angle of 15 degrees or less. (Figs. 4A, B). We just do not see it.
At 16 degrees, we see a few individuals who are percolating through the filter that is life. We see at least 1 in the Murphy study, where the range of the individuals over the age of 65 years without severe osteoarthritis went down to 16 degrees. We see 1 in the Copenhagen study, when the authors illustrated a 65-year-old male with a CE angle of 16 degrees. These individuals are basically case studies. They tell us nothing about what is probable or improbable, they just tell us what is possible. It is undeniable that it is possible for an individual to reach 65 years of age, with a CE angle of 16 degrees and be arthritis-free.
Most orthopaedic surgeons believe that many factors could influence the development of hip osteoarthritis, including genetic and neuromuscular diseases, patient age, sex, obesity, ligamentous stability, occupation, recreational activities, femoral deformity, and the native toughness of the individual’s cartilage. Clearly, some patients do well with a CE angle of 16 degrees at 65 years of age, but I suspect that most do not.
It would be useful to delineate the factors that predict successful aging in the stable, dysplastic hip. The Orthopaedic community is still searching for something better than the CE angle because it does not reliably predict the rate of appearance of osteoarthritis in hips that range from 6 to 15 degrees, although it is adequate to determine the eventual fate of these hips.
At this time, groups are examining other ways to describe the femoral/acetabular relationship that may be better than measurements we can make on the plain AP x-ray of the pelvis. For instance, the false profile view is helpful in delineating anterior deficiency of the acetabulum although we do not have reams of longitudinal data that confirm the long-term usefulness of this view.22 Computed tomography scans are being used to better describe the volumetric relationships between the femur and pelvis.23 In addition, dynamic assessment of cartilage is beginning to come to the clinical realm.20 All of these have the potential to add value to our assessment of individual patients.
It is both peculiar and interesting that the CE angle works as well as it does. After all, it is a 2-dimensional representation of a 3-dimensional relationship. Logic suggests it would not be as useful as it has proved to be, but logic is not what rules our world, experience rules our world. And in our world, the CE angle works well to predict the eventual fate of a dysplastic hip.
1. Cooperman DR, Wallensten R, Stulberg SD.Acetabular dysplasia in the adult.Clin Orthop Relat Res.1983;175:79–85.
2. Jacobsen S, Sonne-Holm S.Hip dysplasia: a significant risk factor for the development of hip osteoarthritis. A cross-sectional survey.Rheumatology (Oxford).2005;44:211–218.
3. Jacobsen S, Sonne-Holm S, Soballe K, et al..Hip dysplasia and osteoarthritis A survey of 4151 subjects from the Osteoarthrosis Substudy of the Copenhagen City Heart Study.Acta Orthop.2005;76:149–158.
4. McWilliams DF, Doherty SA, Jenkins WD, et al..Mild acetabular dysplasia and risk of osteoarthritis of the hip: a case-control study.Ann Rheum Dis.2010;69:1774–1778.
5. Murphy SB, Ganz R, Muller ME.The prognosis in untreated dysplasia of the hip. A study of radiographic factors that predict the outcome.J Bone Joint Surg.1995;77-A:985–989.
6. Stulberg SD, Cordell ID, Harris WH, et al.Harris WH.Unrecognized childhood hip disease: a major cause of idiopathic osteoarthritis of the hip.The Hip. Proceedings of the Third Open Scientific Meeting of the Hip Society.1975.St Louis:CV Mosby.
7. Stulberg SD, Harris WHAmstutz HC.Acetabular dysplasia and the development of osteoarthritis of the hip.The Hip. Proceedings of the Second Open Scientific Meeting of the Hip Society.1974.St Louis:CV Mosby.
8. Wiberg G.Studies on dysplastic acetabula and congenital subluxation of the hip joint with special reference to the complication of osteoarthritis.Acta Chir Scand.1939;83suppl 587–135.
9. Sharp IK.Acetabular dysplasia. The acetabular angle.J Bone Joint Surg.1961;43B:268–272.
10. Jacobsen S, Sonne-Holm S, Geguhr P, et al..Joint space width in dysplasia of the hip.J Bone Joint Surg.2005;87B:473–477.
11. Heyman CH, Herndon CH.Legg-Perthes disease. A method for the measurement of the roentgenographic result.J Bone Joint Surg Am.1950;32:767–778.
12. Roberts JMFitzgerald RH, Kaufer H, Malkani AL.Developmental dysplasia of the hip: diagnosis and treatment of the non-ambulator.Orthopaedics.2002.St Louis, London, Philadelphia, Sydney, and Toronto:Mosby.
13. Kobayashi D, Satsuma S, Kuroda R, et al..Acetabular development in the contralateral hip in patients with unilateral developmental dysplasia of the hip.J Bone Joint Surg.2010;92-A:130–1397.
14. Gosvig KK, Jacobsen S, Sonne-Holm S, et al..Prevalence of malformations of the hip joint and their relationship to sex, groin pain, and risk of osteoarthritis. A population-based survey.J Bone Joint Surg Am.2010;92:1162–1169.
15. Bombelli R.Structure and Function in Normal and Abnormal Hips: How to Rescue Mechanically Jeopardized Hips.1993:3rd ed.New York:Springer.
16. Maquet P.Biomechanics of the Hip.1985.Berlin:Springer.
17. Jessel RH, Zurakowski D, Zilkens C, et al..Radiographic and patient factors associated with pre-radiographic osteoarthritis in hip dysplasia.J Bone Joint Surg.2009;91:1120–1129.
18. Millis MB, Kim Y-J.Rationale of osteotomy and related procedures for hip preservation: a review.Clin Orthop Relat Res.2002;405:108–121.
19. Pauwels F.Biomechanics of the Normal and Diseased Hip: Theoretical Foundations, Technique and Results of Treatment: An Atlas.1976.New York:SpringerTranslated by Furlong RJ, Maquet P.
20. Rhee PC, Woodcock JA, Clohisy JC, et al..The Shenton line in the diagnosis of acetabular dysplasia in the skeletally mature patient.J Bone Joint Surg Am.2011;93suppl 235–39.
21. Milgram JW.Morphology of untreated bilateral congenital dislocation of the hips in a seventy-four year old man.Clin Orthop Relat Res.1976;119:112–115.
22. Lequesne M, de Sezes S.Le faux profil du basin. Nouvelle incidence radiographique pour l'etude de la hanche. Son utilite dans les dysplasies et les differentes coxopathies. [The False Profile View. A New X-ray View of the Hip. For use in hip dysplasia and the other coxarthropathies].Rev Rhumat.1961;28:643–652.
23. Stubbs AJ, Anz AW, Frino J, et al..Classic measures of hip dysplasia do not correlate with three-dimensional computer tomographic measures and indices.Hip Int.2011549–558.