1. Top of page
  2. Abstract
  6. Acknowledgements


Patients with unilateral hip or knee replacements for end-stage osteoarthritis (OA) are at high risk for future progression of OA in other joints of the lower extremities, often requiring additional joint replacements. Although the risks of future surgery in the contralateral cognate joints (i.e., contralateral hip replacement after an initial hip replacement) have been evaluated, the evolution of end-stage hip OA to OA involving the knee joints, and vice versa (i.e., noncognate progression) has not been investigated. Because characterization of OA progression in noncognate joints may shed light on the pathogenesis of multijoint OA, we investigated the pattern of evolution of end-stage lower extremity OA in a large, clinical cohort.


Total joint replacement (TJR) was selected as a marker of end-stage OA, and a database comprising all lower extremity TJRs performed at a large referral center between 1981 and 2001 was accessed. Of the 5,894 patients identified, 486 patients with idiopathic OA who underwent hip replacement and 414 who underwent initial knee replacement were analyzed to determine the relative likelihood of subsequent TJRs. Patients with the systemic inflammatory arthropathy, rheumatoid arthritis (RA), were evaluated as a control population because RA progression is not considered to be a primarily mechanically mediated process.


The contralateral cognate joint was the most common second joint to undergo replacement in both the OA and the RA groups. However, in OA patients for whom the second TJR was in a noncognate joint, that joint was >2-fold more likely to be on the contralateral limb than on the ipsilateral limb (hip to knee P < 0.001; knee to hip P = 0.013). In contrast, among the RA cohort, the evolution was random and no laterality for noncognate TJR was observed at either the hip or the knee (P = 0.782).


This characterization of end-stage lower extremity OA demonstrates that the disease evolves nonrandomly; after 1 joint is replaced, the contralateral limb is significantly more likely to show progression of OA than is the ipsilateral limb. Thus, OA in 1 weight-bearing joint appears to influence the evolution of OA in other joints. The absence of such laterality in RA suggests that OA progression may be mediated by extrinsic factors such as altered joint loading.

The natural history and pattern of evolution of multiarticular osteoarthritis (OA) remain incompletely defined. Previous studies have shown a relationship between OA of the hands, knees, and hips, particularly in subjects with so-called generalized OA, which is widely regarded as genetically based (1–5). However, unlike the systemic inflammatory arthritides such as rheumatoid arthritis (RA), which tend to involve joints symmetrically, the arthropathy of OA is considered to be asymmetric (6, 7).

Nevertheless, there is abundant evidence for bilateral involvement in OA, such as in the relatively symmetric distribution of Heberden's and Bouchard's nodes in “nodal” hand OA (4) and in the high incidence of bilateral OA progression between paired (cognate) joints of the lower limbs (1, 8–10). In addition, patients undergoing unilateral total joint replacement (TJR) of the hip, a validated outcome measure of end-stage hip OA (11–13), have been reported to be more likely than the general population to have had a previous contralateral hip replacement (9, 14). Similarly, those undergoing unilateral knee replacement for idiopathic knee OA have a high frequency of previous contralateral knee replacement (9). The basis of this seemingly symmetric evolution is unknown. Does the presence of end-stage OA in 1 joint bring about OA in the same joint of the contralateral limb (the cognate joint)? Or, is the ultimate presence in both joints simply a reflection of the underlying pathophysiology of the OA process unique to the cognate joints?

No systematic study has examined the relationship of OA in 1 lower extremity joint to that in the other 3 large lower extremity joints. In particular, the evolution to end-stage OA from the initial involvement of 1 joint to the subsequent involvement of the ipsilateral or contralateral noncognate joint (e.g., initial end-stage hip OA followed by the development of end-stage knee OA) has never been investigated. Since OA of the lower extremities is at least partially related to excessive loading (15–18), we hypothesized that the sequence of progression of OA among the various joints of the lower extremities would be influenced by which joint developed end-stage OA first, and that this would be substantively different from the progression in nonmechanically mediated processes such as the inflammatory arthropathies. Using TJR as an indicator of end-stage OA (11–13), we sought to determine the relationship between end-stage OA in the hip or knee and the development of subsequent end-stage OA in the other large joints of the lower extremity.


  1. Top of page
  2. Abstract
  6. Acknowledgements

A database composed of all lower extremity TJRs performed at Rush–Presbyterian–St. Luke's Medical Center (Rush) in Chicago, Illinois between 1981 and 2001 was accessed. There were 5,894 patients in the database (excluding patients who underwent revision surgery only). The Department of Orthopedic Surgery at Rush compiled this database for research purposes. Information in the database included the type of TJR, whether the surgery was a revision surgery, date of surgery, and preoperative diagnosis.

Patients were included in this analysis if 1) their initial surgery was a primary TJR of the hip or knee, 2) they had at least 1 subsequent TJR, and 3) the preoperative diagnosis for the TJRs was idiopathic OA or RA. Patients were excluded if 1) they had multiple TJRs on the same date, or 2) the preoperative diagnosis was not OA or RA or was indeterminate. Patients with RA were included in the analysis as a control group because unlike OA, RA is a systemic inflammatory disease whose arthropathy is synovial based and is not regarded as primarily driven by mechanical loading (19).

The patients were divided into 2 groups for data analysis (Figure 1). The first group comprised those patients who had an initial hip replacement (group I) and the second group included those who had an initial knee replacement (group II). After distribution into group I or group II, patients were included if they had at least 1 subsequent TJR. Final analyses included only those patients with a diagnosis of OA or RA.

thumbnail image

Figure 1. Flow chart of patients in the study. The 5,894 patients in the database were divided into groups I and II based on the lower extremity site of the initial total joint replacement (TJR). Only those patients with a subsequent TJR and those with the diagnosis of osteoarthritis (OA) or rheumatoid arthritis (RA) were included in the final analysis (boxed areas). = total excludes patients with joint revision only. ∗∗ = total includes patients with other diagnoses and those who had multiple surgeries on the same date.

Download figure to PowerPoint

The Pearson chi-square test of hypothesized proportions was used to determine the relative likelihood of subsequent TJRs after the initial hip or knee replacement. Additional analysis using nonparametric Mann-Whitney test was performed to compare time from initial hip or knee replacement to subsequent TJRs. A significance level of P equal to 0.05 was determined a priori.


  1. Top of page
  2. Abstract
  6. Acknowledgements

In group I, 486 OA patients and 41 RA (control) patients fulfilled the inclusion criteria and had an initial hip replacement followed by a second TJR (Figure 1). In both the OA and the RA groups, the contralateral hip was the most common second joint to be replaced. In the OA group, 85% (411 patients) and in the RA group, 68% (28 patients) had their second TJR at the contralateral hip.

Of the 15% of OA patients (75 patients) in group I whose initial hip replacement was followed by a second TJR at the knee, a significantly higher proportion, 71% (53 patients), had TJR at the knee contralateral to the side of the initial hip replacement, while 29% (22 patients) had surgery at the knee ipsilateral to the initial hip replacement (P < 0.001). In contrast, of the 32% of RA patients (13 patients) in group I who had their second surgery at the knee, there was no significant laterality in the distribution of subsequent knee replacements. RA patients were equally likely to have a contralateral knee replacement (54% [7 patients]) as they were to have an ipsilateral knee replacement (46% [6 patients]) (P = 0.782). Table 1 summarizes the results on the OA and RA patients in group I. There was not a sufficient number of third TJRs among OA patients (7 hip, 15 knee) or RA patients (4 hip, 12 knee) to make statistically meaningful conclusions on the distribution of the third TJRs.

Table 1. Distribution of knee total joint replacements (TJRs) in patients with an initial hip TJR (group I)*
 Contralateral knee, no. (%)Ipsilateral knee, no. (%)
  • *

    OA = osteoarthritis; RA = rheumatoid arthritis.

  • P < 0.001 versus an ipsilateral knee replacement.

  • P = 0.782 versus an ipsilateral knee replacement.

OA (n = 75)53 (71)22 (29)
RA (n = 13)7 (54)6 (46)

In group II, 414 OA patients and 37 RA patients fulfilled the inclusion criteria and had an initial TJR of the knee followed by a second TJR (Figure 1). The contralateral knee was the most common second joint to be replaced in 92% of the OA patients (382 patients) and 86% of the RA patients (32 patients).

Of the 32 OA patients (8%) in group II who had their second surgery at the hip, a significantly higher proportion (72% [23 patients]) had a contralateral hip replacement as opposed to an ipsilateral hip replacement (28% [9 patients]) (P = 0.013) (Table 2). Only 5 patients in the RA group had their second surgery at the hip, of whom 4 had an ipsilateral hip TJR while only 1 had a contralateral hip TJR; there was therefore an insufficient number of patients to perform statistical analysis. Table 2 summarizes the results on the OA and RA patients in Group II. There was not a sufficient number of third TJRs in the OA group (7 knee, 10 hip) or RA group (1 knee, 6 hip) to make significant conclusions.

Table 2. Distribution of hip TJRs in patients with an initial knee TJR (group II)*
 Contralateral hip, no. (%)Ipsilateral hip, no. (%)
  • *

    See Table 1 for definitions.

  • P = 0.013 versus an ipsilateral hip replacement.

  • Insufficient number of patients to perform statistical analysis.

OA (n = 32)23 (72)9 (28)
RA (n = 5)1 (20)4 (80)

Analysis of the data using the Pearson chi-square test of association showed that a significantly larger proportion of subjects with an initial unilateral hip TJR (15%) required subsequent knee TJR compared with those with an initial knee TJR that required subsequent hip TJR (8%) (P < 0.0005). Results in the RA group revealed a similar trend, although the difference did not reach statistical significance, with 32% of the subjects with initial hip TJR undergoing subsequent knee TJR and only 14% of subjects with initial knee TJR receiving subsequent hip TJR (P = 0.057).

In order to further characterize the evolution of OA from 1 joint to the noncognate joint, the time intervals between the initial and subsequent TJRs were examined. In the case of initial TJR of the hip, there were no significant differences in the number of years to contralateral knee TJR versus number of years to ipsilateral knee TJR, in either the OA or RA group. The mean (±SD) time to second surgery for the OA patients was 4.6 ± 4.4 years for contralateral knee replacement and 3.5 ± 4.9 years for ipsilateral knee replacement (P = 0.144). For the RA group, the mean time until contralateral knee replacement was 2.0 ± 1.9 years, and until ipsilateral knee replacement was 2.8 ± 3.2 years (P = 0.568).

In the case of initial TJR of the knee (group II), there were similarly no significant interval differences between the contralateral and ipsilateral hip replacement. The time between initial knee TJR and subsequent TJR of the hip in the OA group was 2.3 ± 2.9 years for the contralateral hip replacement and 3.9 ± 3.6 years for the ipsilateral hip replacement (P = 0.082). In the RA group, the time interval between the contralateral hip replacement, 5.3 ± 0.0 years, and ipsilateral hip replacement, 4.9 ± 1.7 years, was not significantly different (P = 0.480).


  1. Top of page
  2. Abstract
  6. Acknowledgements

No systematic study has previously examined the influence of OA in 1 large joint of the lower extremity on the pattern of evolution of OA in the remaining lower extremity joints. Using TJRs as a validated surrogate marker for end-stage OA (11–13), this study demonstrates that the progression to end-stage OA in the lower extremities is not random. After a single unilateral TJR of the hip or knee, there are 3 remaining large joints potentially subject to replacement; of these, the cognate joint on the contralateral side is far more frequently replaced than either of the 2 noncognate joints. This is consistent with previous reports that hip or knee OA often evolves symmetrically to the contralateral cognate joint (8, 9, 13). Interestingly, evolution of noncognate end-stage OA appeared to occur significantly more often (P < 0.0005) from the hip to the knee (15% of subjects with an initial hip TJR) rather than from the knee to the hip (8% of subjects with an initial knee TJR). Moreover, in OA patients in whom the second TJR was in a noncognate joint, that joint was >2-fold more likely to be on the contralateral limb than on the ipsilateral limb (∼70% versus ∼30%, respectively). This preferential evolution to the contralateral noncognate joint was not found in the control group of RA patients, suggesting that a factor unique to the pathogenesis of OA may be responsible.

Chitnavis et al (9) performed a review of 402 consecutive patients undergoing hip or knee replacements, and reported that between one-fourth and one-third of patients had undergone bilateral TJRs at the time of survey; moreover, 70% of those having received unilateral hip replacement had radiographic evidence of OA at the contralateral hip, and nearly two-thirds of patients having received unilateral knee replacement had radiographic evidence of OA at the contralateral knee (9). The investigators did note that one-fourth of those with a TJR of the hip reported chronic knee pain, while one-eighth of those with a TJR of the knee reported chronic hip pain (9). Although the laterality of the pain was not reported, the observation of chronic noncognate joint pain is consistent with our findings in end-stage OA.

In primary idiopathic OA, joint degeneration has been related to abnormal mechanical loads and a significant “wear-and-tear” component (15–17). Investigators have suggested clinical correlates of the effect of loading on the evolution of OA, including a higher prevalence of radiographic knee OA observed in the “normal” limb of athletes who have undergone unilateral lower limb amputations (20, 21), and a higher prevalence of right-sided hip replacements over left-sided hip replacements for OA in a predominantly right-side dominant population (22, 23). We speculate that similar biomechanical factors may be responsible for the observed evolution of end-stage OA of the lower extremity as indicated by TJR in the present study. In this regard, neuromuscular adaptations (i.e., limping) performed to limit pain or improve function in patients with symptomatic unilateral OA may place relatively greater loads on the contralateral extremity. Whereas previous studies have primarily focused on the relationship and influence of loading on the progression of OA at a single joint (15–17), the results from this investigation suggest that loading alterations may have more extensive consequences, particularly on the multiarticular evolution of lower extremity OA.

A potential limitation of this study is that the order of TJR surgery does not necessarily reflect the order in which OA develops; for example, it is possible to have long-standing mild OA of the knees, and then subsequently to develop severely painful OA of the hip that requires TJR. However, the intention of the study was to characterize evolution to end-stage disease, rather than to identify the order of early disease onset. Although details concerning the symptoms and radiographic severity of OA among our study population are not available, investigators have previously validated the concept of using TJRs as an appropriate end point to assess severe OA (11–13, 24, 25). In these studies, pain and structural disease appear to be equally important contributors in the decision to perform TJR.

A second limitation involves the surgical decision of which joint to replace first in patients with multiple joints with end-stage disease; therefore, there may be surgical bias regarding whether it is intrinsically preferable to replace a hip or a knee first, or a left versus a right joint. However, the absence of significant interval differences between the times to ipsilateral versus contralateral TJRs and the absence of laterality among the RA control population suggest that there was no significant surgical bias in the order of joint replacements.

Finally, in light of the relative population prevalence of the 2 diseases, not unexpectedly, the number of RA subjects available for comparison with the OA group was small. Nevertheless, even with this small number of patients, the sequence of joint replacements appeared to be randomly distributed. Furthermore, the nonrandom distribution of joint replacements in OA is a novel and important clinical observation independent of comparison with the RA group.

Previous studies of the pattern of evolution of end-stage OA have dealt exclusively with the contralateral cognate joint. This is the first study to characterize that pattern among all 4 lower extremity joints. The evolution of end-stage lower extremity OA progresses in a nonrandom manner. The evolution of OA to the contralateral cognate and noncognate weight-bearing joint appears to be intrinsic to the nature of the disease. In particular, the presence of end-stage OA in 1 joint appears to increase the likelihood of developing OA in the contralateral noncognate joint. This may be related to biomechanical factors and is specifically not found in a nonmechanically mediated process, such as RA. However, the presence and extent of any alterations in loading that occur during OA evolution, which may contribute to the pathogenic process or processes that result in contralateral progression, remain to be characterized.


  1. Top of page
  2. Abstract
  6. Acknowledgements

The authors thank Laura Quigley, MS, RN (Department of Orthopedic Surgery) for assistance with access to the TJR database, and Dr. Debra E. Hurwitz and the members of the Rush Gait Laboratory for critical reading of the manuscript.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  • 1
    Cooper C, Egger P, Coggon D, Hart DJ, Masud T, Cicuttini F, et al. Generalized osteoarthritis in women: pattern of joint involvement and approaches to definition for epidemiological studies. J Rheumatol 1996; 23: 193842.
  • 2
    Cushnaghan J, Dieppe P. Study of 500 patients with limb joint osteoarthritis. I. Analysis by age, sex, and distribution of symptomatic joint sites. Ann Rheum Dis 1991; 50: 813.
  • 3
    Croft P, Cooper C, Wickham C, Coggon D. Is the hip involved in generalized osteoarthritis? Br J Rheumatol 1992; 31: 3258.
  • 4
    Hirsch R, Lethbridge-Cejku M, Scott WW Jr, Reichle R, Plato CC, Tobin J, et al. Association of hand and knee osteoarthritis: evidence for a polyarticular disease subset. Ann Rheum Dis 1996; 55: 259.
  • 5
    Holderbaum D, Haqqi TM, Moskowitz RW. Genetics and osteoarthritis: exposing the iceberg. Arthritis Rheum 1999; 42: 397405.
  • 6
    Solomon L. Clinical features of osteoarthritis. In: RuddyS, HarrisEDJ, SledgeCB, editors. Kelley's textbook of rheumatology. St. Louis, MO: W.B. Saunders; 2001. p. 140918.
  • 7
    Brandt KD. Osteoarthritis. In: BraunwaldE, FauciAS, KasperDL, HauserSL, LongoDL, JamesonAH, editors. Harrison's principles of internal medicine. New York: McGraw-Hill; 2001. p. 198794.
  • 8
    Spector TD, Hart DJ, Doyle DV. Incidence and progression of osteoarthritis in women with unilateral knee disease in the general population: the effect of obesity. Ann Rheum Dis 1994; 53: 5658.
  • 9
    Chitnavis J, Sinsheimer JS, Suchard MA, Clipsham K, Carr AJ. End-stage coxarthrosis and gonarthrosis: aetiology, clinical patterns and radiological features of idiopathic osteoarthritis. Rheumatology 2000; 39: 6129.
  • 10
    Gunther KP, Sturmer T, Sauerland S, Zeissig I, Sun Y, Kessler S, et al. Prevalence of generalised osteoarthritis in patients with advanced hip and knee osteoarthritis: the Ulm Osteoarthritis Study. Ann Rheum Dis 1998; 57: 71723.
  • 11
    Dougados M, Gueguen A, Nguyen M, Berdah L, Lequesne M, Mazieres B, et al. Requirement for total hip arthroplasty: an outcome measure of hip osteoarthritis? J Rheumatol 1999; 26: 85561.
  • 12
    Maillefert JF, Gueguen A, Nguyen M, Berdah L, Lequesne M, Mazieres B, et al. A composite index for total hip arthroplasty in patients with hip osteoarthritis. J Rheumatol 2002; 29: 34752.
  • 13
    Fox KM, Hochberg MC, Resnik CS, Kenzora JE, Hebel JR, Zimmerman SI, et al. Severity of radiographic findings in hip osteoarthritis associated with total hip arthroplasty. J Rheumatol 1996; 23: 6937.
  • 14
    Husted H, Overgaard S, Laursen JO, Hindso K, Hansen LN, Knudsen HM, et al. Need for bilateral arthroplasty for coxarthrosis: 1,477 replacements in 1,199 patients followed for 0–14 years. Acta Orthop Scand 1996; 67: 4213.
  • 15
    Hurwitz DE, Hong R, Case JP, Block JA. Progression of knee osteoarthritis is related to higher loading [abstract]. Trans Orthop Res Soc 2002; 27: 678.
  • 16
    Sharma L, Hurwitz DE, Thonar EJ, Sum JA, Lenz ME, Dunlop DD, et al. Knee adduction moment, serum hyaluronan level, and disease severity in medial tibiofemoral osteoarthritis. Arthritis Rheum 1998; 41: 123340.
  • 17
    Wada M, Tatsuo H, Sato M, Kawahara H, Shimada S, Sasaki S, et al. Dynamic load at baseline can predict radiographic disease progression of tibiofemoral osteoarthritis [abstract]. Osteoarthritis Cartilage 2000; 8: S8.
  • 18
    Prodromos CC, Andriacchi TP, Galante JO. A relationship between gait and clinical changes following high tibial osteotomy. J Bone Joint Surg Am 1985; 67: 118894.
  • 19
    Yamanishi Y, Firestein GS. Pathogenesis of rheumatoid arthritis: the role of synoviocytes. Rheum Dis Clin North Am 2001; 27: 35571.
  • 20
    Melzer I, Yekutiel M, Sukenik S. Comparative study of osteoarthritis of the contralateral knee joint of male amputees who do and do not play volleyball. J Rheumatol 2001; 28: 16972.
  • 21
    Burke MJ, Roman V, Wright V. Bone and joint changes in lower limb amputees. Ann Rheum Dis 1978; 37: 2524.
  • 22
    Newton JASV. Why is osteoarthritis of the hip more common on the right? [letter] Lancet 1993; 341: 179.
  • 23
    Seagroatt V, Tan HS, Goldacre M, Bulstrode C, Nugent I, Gill L. Elective total hip replacement: incidence, emergency readmission rate, and post-operative mortality. BMJ 1991; 303: 14315.
  • 24
    Mancuso CA, Ranawat CS, Esdaile JM, Johanson NA, Charlson ME. Indications for total hip and total knee arthroplasties: results of orthopaedic surveys. J Arthroplasty 1996; 11: 3446.
  • 25
    Dieppe P, Basler HD, Chard J, Croft P, Dixon J, Hurley M, et al. Knee replacement surgery for osteoarthritis: effectiveness, practice variations, indications and possible determinants of utilization. Rheumatology 1999; 38: 7383.