Coexisting lateral tibiofemoral osteoarthritis is associated with worse knee pain in patients with mild medial osteoarthritis

Osteoarthritis and Cartilage xxx (2017) 1e8

Coexisting lateral tibiofemoral osteoarthritis is associated with worse knee pain in patients with mild medial osteoarthritis H. Iijima y z, T. Aoyama y *, K. Nishitani x, H. Ito x, N. Fukutani y, T. Isho k, E. Kaneda ¶, H. Kuroki y, S. Matsuda x y Department of Physical Therapy, Human Health Sciences, Graduate School of Medicine, Kyoto University, Kyoto, Japan z Japan Society for the Promotion of Science, Tokyo, Japan x Department of Orthopedic Surgery, Graduate School of Medicine, Kyoto University, Kyoto, Japan k Rehabilitation Center, Fujioka General Hospital, Gunma, Japan ¶ Nozomi Orthopedic Clinic, Hiroshima, Japan

a r t i c l e i n f o

s u m m a r y

Article history: Received 14 September 2016 Accepted 21 February 2017

Objective: To examine the clinical impact of coexisting lateral osteoarthritis (OA) in knees with mild medial OA. Design: In patients with Kellgren/Lawrence (K/L) grade 2 OA in the medial compartment (n ¼ 100; age: 56e89 years; 80.0% female), anteroposterior knee radiography was used to assess the presence of lateral OA, using grading systems from the Osteoarthritis Research Society International (OARSI) atlas and the K/ L classification. The Japanese Knee Osteoarthritis Measure (JKOM), knee range of motion (ROM), and performance-based functional measures (10 m walk, timed up and go and five repetition chair stand maneuvers) were evaluated. The outcomes were compared between patients with and without lateral OA using an analysis of covariance (ANCOVA) or nonparametric rank ANCOVA. Furthermore, ordinal logistic regression analysis was performed, with responses on individual JKOM pain questionnaires as the outcomes and lateral OA as the predictor. Results: Knees with coexisting lateral OA had a significantly worse score of JKOM pain question compared with those without, after adjusting for covariates. The presence of lateral OA was significantly associated with knee pain while ascending/descending stairs and standing. These results were consistent between different definitions of the K/L and OARSI grading systems. The knee ROM and performancebased functional measures were not significantly different between patients with and without lateral OA. Conclusion: Knees with concomitant lateral and mild medial OA may be more symptomatic compared to those without lateral OA. These findings might help to define a clinically distinct subgroup based on a simple radiographic finding in mild knee OA. © 2017 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.

Keywords: Lateral tibiofemoral osteoarthritis Knee pain Radiography

Introduction Osteoarthritis (OA) of the knee, a leading cause of knee pain and chronic disability worldwide1, is the most common type of

*Address correspondence and reprint requests to: T. Aoyama, 53 Shogoin, Kawahara-cho, Sakyo-ku, Kyoto 606-8507, Japan. Tel: 81-75-751-3952; Fax: 8175-751-3909. E-mail addresses: [email protected] (H. Iijima), aoyama.tomoki.4e@ kyoto-u.ac.jp (T. Aoyama), [email protected] (K. Nishitani), hiromu@ kuhp.kyoto-u.ac.jp (H. Ito), [email protected] (N. Fukutani), isho.tak@gmail. com (T. Isho), [email protected] (E. Kaneda), [email protected] (H. Kuroki), [email protected] (S. Matsuda).

arthritis2 and is a heterogeneous disease3. The Kellgren/Lawrence (K/L) classification4 is a widely accepted scale that is used to grade disease severity and to define the presence of OA, in which grade 2 is usually used as the disease threshold5. Unlike the Osteoarthritis Research Society International (OARSI) atlas grading system defined by Altman and Gold6, the K/L classification grades medial or lateral compartment features without distinction. However, knee OA is a tricompartmental disease7, which indicates the importance of compartment-specific assessment and the need to investigate the pathogenesis and clinical impact of knee OA based on different radiographic patterns. Although knee OA commonly affects the medial tibiofemoral compartment of the knee joint8,9, possibly because of a higher

http://dx.doi.org/10.1016/j.joca.2017.02.801 1063-4584/© 2017 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Iijima H, et al., Coexisting lateral tibiofemoral osteoarthritis is associated with worse knee pain in patients with mild medial osteoarthritis, Osteoarthritis and Cartilage (2017), http://dx.doi.org/10.1016/j.joca.2017.02.801

2

H. Iijima et al. / Osteoarthritis and Cartilage xxx (2017) 1e8

percentage of loading in the medial compartment10, a coexisting marginal osteophyte in the lateral compartment in medial OA has been reported11,12. For example, Faschingbauer et al.11 showed that 170 (49.4%) of 344 knees with medial OA of K/L grade 2 or 3 had a radiographic-defined osteophyte in the lateral compartment. Coexisting medial and lateral OA has been suggested to be more of a synovial inflammatory disease than is isolated medial OA13. This finding indicates that bicompartmental (medial and lateral) and unicompartmental (medial alone) OA may belong to different subgroups, or may represent different stages of the same disease. However, no previous study has examined the clinical impact of coexisting lateral disease in medial knee OA, including knee pain and functional measures. Such knowledge would contribute to defining a subgroup of knee OA using simple radiographic features, and might help resolve the discordance between clinical symptoms and radiographic findings in heterogeneous OA disease14. This cross-sectional study sought to examine the clinical impact of coexisting lateral OA in knees with mild medial OA. Since mild knee OA shows a wide variability in clinical outcomes among patients, the current study focused on knees with K/L grade 2 in the medial compartment, thereby defining a potential subgroup in mild knee OA as suggested by Felson et al.15. We hypothesized that knees with coexisting lateral OA would have worse clinical symptoms, poorer physical function, and a more restricted knee range of motion (ROM) compared to knees without lateral OA.

Japanese Knee Osteoarthritis Measure [JKOM]). Measurement of ROM and functional performance were evaluated by trained physical therapists (HI and NF) who had >6 years of clinical experience with musculoskeletal disorders and a postgraduate master's (HI) or doctor's (NF) degree-level qualification.

The ethical committee of Kyoto University approved the study (approval number: E1923), and written informed consent was obtained from all participants before their enrollment. For the current cross-sectional study, we used data obtained from January 2014 to January 2015. Patients were identified through the medical record system, and were recruited from the community orthopedics clinic in Hiroshima which is located in a rural mountainous community. We distributed an advertisement requesting patients who were visiting the clinics for conservative treatment of knee OA. All recruited patients had a history of pain in one or both knees. Patients were included in this study if they met the following inclusion criteria: (1) age 50 years; (2) radiographic OA (K/L grade 2 according to the original version4) in one or both knees in the medial tibiofemoral compartment, evaluated using weightbearing anteroposterior radiographs of the tibiofemoral joint; and (3) ability to walk independently on a flat surface without any ambulatory assistive device. Patients with bilateral knee OA were not considered separately from unilateral cases. The exclusion criteria were (1) a history of knee surgery, (2) rheumatoid arthritis, (3) periarticular fracture, (4) present neurological problems, or (5) lateral knee OA (i.e., isolated lateral OA or coexisting medial and lateral OA but lateral OA is more severe than medial OA). Lateral knee OA was defined as a knee having a K/L grade 2, along with lateral joint space narrowing (JSN) > medial JSN, and lateral osteophytes > medial osteophytes using an OARSI atlas6 according to previously described methods16,17.

Radiographic evaluation Anteroposterior radiographs of the index knees in the fully extended weight-bearing and foot map positions were obtained within 3 months of enrollment. The index knee was defined as the more painful knee, as described in the Statistical analysis section. The beam was positioned parallel to the floor with no angle at the joint space. Fluoroscopic guidance with an anteroposterior X-ray beam was used to properly visualize the joint space. The radiographic severity of the lateral compartment in the tibiofemoral joint was assessed by trained examiner (HI) and experienced physician who had a >10 years of clinical experience as an orthopedic surgeon (KN) using the standard scale from the OARSI atlas6 and K/L classification system. Disagreement between two independent examiners was discussed and consensus was achieved. OARSI JSN (lateral tibiofemoral joint) and osteophytes (lateral femoral and tibial compartment) were graded on a four-point scale (0e3; 0: normal; 1: mild; 2: moderate: 3: severe) following the radiograph examples in the OARSI atlas6. Radiographic OA of the lateral compartment by the OARSI atlas grading system was considered to be present if: (1) JSN  grade 2, (2) the sum of the two marginal osteophyte grades from the lateral compartment 2, or (3) grade 1 JSN combined with a grade 1 osteophyte in the lateral compartment. Although this definition is supposed to be equivalent to grade 2 OA in the K/L classification18, a recent study showed that the OARSI atlas grading system differs from the K/L classification regarding the prevalence of radiographic knee OA19. Thus, we also defined radiographic OA according to the K/L grading system as the presence of a definite osteophyte (K/L grade ¼ 2) in the lateral compartment of the tibiofemoral joint in a compartment-specific manner20. Specifically, the K/L grade in the lateral compartment was scored as follows: 0 ¼ normal; 1 ¼ possible osteophyte; 2 ¼ definite osteophytes; 3 ¼ osteophytes and JSN; 4 ¼ large osteophytes, marked JSN, and definite deformity. We defined lateral OA if marginal osteophytes were confirmed in either femur or tibia. Although this definition of K/L grade 2 was different from the original version4, varus knee would be expected to lead to a lateral lift off, which affects JSN grading in the lateral compartment. Thus, radiographic lateral OA in the current study was defined only according to the presence of a definite marginal osteophyte. This is also an alternative criterion used in large studies, such as the Framingham study21. To assess intra-rater reliability, 100 randomly selected radiographs were rescored by the same examiner (HI) more than 1 week after the first assessment. The intra-rater reliabilities were excellent22 for JSN (k ¼ 0.84, 95% confidence interval [CI]: 0.73e0.96), osteophyte grade (femur, k ¼ 0.87, 95% CI: 0.81e0.93; tibia, k ¼ 0.90, 95% CI: 0.86e0.95), and K/L grade in the lateral compartment (k ¼ 0.88, 95% CI: 0.83e0.92). The inter-rater reliability between two examiners (HI and KN) was fair to good22 for JSN (k ¼ 0.66, 95% CI: 0.48e0.85), osteophyte grade (femur, k ¼ 0.73, 95% CI: 0.62e0.84; tibia, k ¼ 0.63, 95% CI: 0.50e0.75), and K/L grade in the lateral compartment (k ¼ 0.68, 95% CI: 0.56e0.80).

Outcome measurements

Statistical analysis

For all patients, the following outcome measurements were evaluated: (1) radiographic evaluation, (2) passive knee ROM measurement, (3) three functional performance measurements (the 10 m walk, timed up and go [TUG], and five repetition chair stand [5CS]), and (4) an OA-related health domain measure (the

To minimize any bias produced by similarities between the right and left knees of the same patients, only one knee per patient was analyzed (“index knee”; Supplementary methods). Data analyses were performed using JMP 11 (SAS Institute, Cary, NC, USA) or R (R Foundation for Statistical Computing, Vienna,

Method Patients

Please cite this article in press as: Iijima H, et al., Coexisting lateral tibiofemoral osteoarthritis is associated with worse knee pain in patients with mild medial osteoarthritis, Osteoarthritis and Cartilage (2017), http://dx.doi.org/10.1016/j.joca.2017.02.801

H. Iijima et al. / Osteoarthritis and Cartilage xxx (2017) 1e8

Austria). The normality of continuous variables was assessed with the ShapiroeWilk test. The homogeneity of the variances between groups for all parametric continuous variables was confirmed using the F test. Descriptive statistics were calculated as means and standard deviations (SD) for continuous variables, and as proportions for dichotomous/categorical variables. Characteristics between knees with and without lateral OA were compared using Student's t test for parametric continuous variables, the ManneWhitney U test for nonparametric continuous variables, and the chi-square/Fisher's exact tests for dichotomous/ categorical variables. The JKOM scores and functional performance measurements were also compared using an analysis of covariance (ANCOVA) or nonparametric rank ANCOVA, adjusted for age, sex, body mass index (BMI), and presence of patellofemoral (PF) OA. Parametric methods can result in inaccurate values when assumptions of normality and homogeneity of variance are not met23. Since the JKOM scores exhibited scattering distribution with narrow range of score points, a nonparametric rank ANCOVA24 was used to compare the JKOM scores in patients with and without lateral OA. Subsequently, ordinal logistic regression analysis was performed to examine the association of coexisting lateral OA with individual questions of the pain-related JKOM subcategory “pain and stiffness.” Ordinal logistic regression is a popular model for ordinal categorical outcome variables, which also works well for skewed continuous outcome variables using ranks of data25. Since few of the individual pain scores were high, individual scores of 2, 3, and 4 were collapsed into one level (moderate/severe pain), and included in the ordinal logistic regression model as a dependent variable (0 ¼ no pain; 1 ¼ mild pain; 2 ¼ moderate/severe pain). Associations between the presence of lateral OA and the individual items on the “pain and stiffness” scale were expressed as adjusted proportional odds ratios (ORs) with 95% CIs, adjusted for age (continuous), sex (0 ¼ male; 1 ¼ female), BMI (continuous), and presence of PFOA (0 ¼ absence; 1 ¼ presence). In the ANCOVA, rank ANCOVA, and ordinal logistic regression analysis, the abovementioned covariates were chosen a priori based on clinical judgment and because they might be associated with the radiographic lateral OA and clinical symptoms of OA and not on the causal pathway9,26. We checked the maximum number of independent variables included in the ordinal logistic regression model and confirmed that overfitting had not occurred (Supplementary methods). P-values <0.05 were considered statistically significant. Additional methods Additional methods for evaluation of the JKOM score, anatomical axis angle (AAA), presence of PFOA, passive knee ROM, spatiotemporal gait parameters, TUG, and 5CS are provided in Supplementary methods. Results In total, 214 patients (i.e., 214 index knees) were evaluated, of which 16 (7.5%) were excluded because of invalid (missing) data [Fig. 1(A)] and 98 were excluded because of pre-radiographic (K/L grade ¼ 1; n ¼ 41 knees) or severe radiographic OA (K/L grade ¼ 3, 4; n ¼ 51 knees) and lateral OA (i.e., isolated lateral OA or coexisting medial and lateral OA but lateral OA is more severe than medial OA; n ¼ 6 knees). Thus, a total of 100 patients (age, 56e89 years; 80.0% female) were included in the final analysis. According to the OARSI atlas criteria, 35 (35.0%) patients had radiographic OA, which was a slightly lower rate compared to that determined using the K/L classification (38; 38.0% patients). Figure 1(B) shows representative radiographs of knees with and

3

without lateral OA. Of the 35 and 38 patients with lateral OA according to the OARSI atlas or K/L classification, 29 (82.9% in the OARSI atlas; 76.3% in the K/L classification) met both criteria. Table I shows the outcome measures for patients with and without lateral OA. Importantly, patients with lateral OA had a significantly higher BMI when using the OARSI atlas and a higher proportion of radiographic PFOA when using both the OARSI atlas and K/L classification than did those without lateral OA. Knees with lateral OA had more severe knee pain during daily living regardless of the definition of radiographic OA The JKOM “pain and stiffness” score was significantly higher (i.e., worse pain) in patients with lateral OA than in those without lateral OA, after adjusting for age, sex, BMI, and the presence of PFOA (Table II), for both the OARSI atlas (9.89 ± 5.19 points vs 6.20 ± 4.40 points; P ¼ 0.011) and the K/L classification (10.0 ± 5.04 points vs 5.94 ± 4.30 points; P < 0.001). The other JKOM subcategories did not differ significantly between knees according to the OARSI atlas, whereas the “activities of daily living” (8.95 ± 6.76 points vs 4.77 ± 4.59 points; P ¼ 0.004), “general health conditions” (3.26 ± 1.64 points vs 2.55 ± 1.28 points; P ¼ 0.015), and “total score” (26.2 ± 14.0 points vs 16.1 ± 10.5 points; P ¼ 0.001) differed significantly between the two groups according to the K/L classification. Knee ROM, spatiotemporal gait parameter, TUG, and 5CS were compared between knees with and without lateral OA (Table III). However, ROM and the performance-based physical functions were not significantly different regardless of lateral OA, after adjusting for age, sex, BMI, and the presence of PFOA when using the OARSI atlas and the K/L classification. Coexisting lateral OA was associated with higher odds of knee pain while stairs climbing and standing regardless of definition of radiographic OA Supplementary Table 1 shows individual question scores of the JKOM “pain and stiffness” in patients with and without lateral OA. Ordinal logistic regression analysis (Table IV) showed that the presence of lateral OA was significantly associated with higher odds of worse knee pain when walking (proportional OR: 2.43; 95% CI: 1.05e5.63; P ¼ 0.038) and bending to the floor or standing up (proportional OR: 3.80; 95% CI: 1.60e9.04; P ¼ 0.003), after adjustment for age, sex, BMI and the presence of PFOA, when the K/ L classification was used. Notably, the presence of lateral OA was significantly associated with higher odds of worse pain when ascending stairs (proportional OR: 2.75; 95% CI: 1.09e6.92; P ¼ 0.031 in the OARSI atlas; proportional OR: 3.11; 95% CI: 1.27e7.60; P ¼ 0.013 in the K/L classification), descending stairs (proportional OR: 4.62; 95% CI: 1.87e11.4; P < 0.001 in the OARSI atlas; proportional OR: 4.76; 95% CI: 1.98e11.4; P < 0.001 in the K/L classification), and standing (proportional OR: 2.51; 95% CI: 1.04e6.05; P ¼ 0.040 in the OARSI atlas; proportional OR: 3.43; 95% CI: 1.45e8.14; P ¼ 0.005 in the K/L classification), after adjustment for age, sex, BMI and presence of PFOA, when using the OARSI atlas and the K/L classification. Discussion The current study revealed that patients with coexisting lateral OA had a worse JKOM “pain and stiffness” score compared to those without lateral OA, after adjusting for age, sex, BMI, and presence of PFOA for both the OARSI atlas and the K/L classification. Notably, the presence of lateral OA was significantly associated with higher odds of worse knee pain while ascending/descending stairs and standing which was consistent for two different definitions of

Please cite this article in press as: Iijima H, et al., Coexisting lateral tibiofemoral osteoarthritis is associated with worse knee pain in patients with mild medial osteoarthritis, Osteoarthritis and Cartilage (2017), http://dx.doi.org/10.1016/j.joca.2017.02.801

4

H. Iijima et al. / Osteoarthritis and Cartilage xxx (2017) 1e8

Fig. 1. Flow chart illustrating the distribution of study patients and radiographs of knees with and without lateral compartment OA. A. In total, 214 knees were recruited, and finally 100 knees with K/L grade 2 in the medial compartment. These knees were divided into groups based on the presence or absence of lateral OA according to two different definitions and subsequently analyzed. B. Representative radiographs of knees with (bicompartment OA: a) and without (unicompartment OA: b) lateral compartment OA. The lateral compartments in these radiographs were scored as follows: (a) OARSI atlas grading system: grade 1 lateral femoral osteophyte (upper white arrow), grade 1 lateral tibial osteophyte (lower white arrow), and grade 0 JSN; K/L classification: grade 2 in lateral compartment. (b) OARSI atlas grading system: grade 0 lateral femoral osteophyte, grade 0 lateral tibial osteophyte, and grade 0 JSN; K/L classification: grade 0 in lateral compartment.

Table I Characteristics of knees with and those without radiographic lateral OA in patients having knees with K/L grade 2 in the medial compartment (n ¼ 100)* Variables

Age, years Female, no. (%) Height, m Weight, kg BMI, kg/m2 Corrected AAA,  Alignment, no. (%) Neutral (corrected AAA 179 but <182 ) Valgus (corrected AAA 182 ) Varus (corrected AAA <179 ) Presence of radiographic PFOA, no. (%) Lateral joint space grade, no. (%) Grade 0 Grade 1 Grade 2 Grade 3 Lateral tibiofemoral joint K/L grade, no. (%) Grade 0 Grade 1 Grade 2 Grade 3 Grade 4

OARSI atlas

K/L

With lateral OA (n ¼ 35)

Without lateral OA (n ¼ 65)

P-valuey

With lateral OA (n ¼ 38)

Without lateral OA (n ¼ 62)

P-valuey

72.3 ± 8.53 28 (80.0) 1.55 ± 0.07 60.3 ± 11.4 25.2 ± 4.73 178.3 ± 2.78

71.9 ± 7.03 52 (80.0) 1.56 ± 0.08 56.5 ± 9.08 23.2 ± 2.83 178.6 ± 2.51

0.809 1.000 0.560 0.066 0.024 0.939 0.082

71.9 ± 8.25 31 (81.6) 1.55 ± 0.07 59.5 ± 10.6 24.9 ± 4.47 178.1 ± 2.18

72.1 ± 7.16 49 (79.0) 1.56 ± 0.07 56.9 ± 9.67 23.3 ± 3.05 178.7 ± 2.81

0.864 0.803 0.486 0.274 0.141 0.350 0.130

14 (40.0)

13 (20.0)

12 (31.6)

15 (24.2)

1 (2.9) 20 (57.1) 26 (74.3)

5 (7.7) 47 (72.3) 27 (41.5)

0 (0.0) 26 (68.4) 26 (68.4)

6 (9.7) 41 (66.1) 27 (43.5)

28 (80.0) 7 (20.0) 0 (0.0) 0 (0.0)

64 (98.5) 1 (1.5) 0 (0.0) 0 (0.0)

31 (81.6) 7 (18.4) 0 (0.0) 0 (0.0)

61 (98.4) 1 (1.6) 0 (0.0) 0 (0.0)

0 8 (22.9) 27 (77.1) 0 (0.0) 0 (0.0)

33 (50.8) 23 (35.4) 9 (13.8) 0 (0.0) 0 (0.0)

0 (0.0) 0 (0.0) 38 (100.0) 0 (0.0) 0 (0.0)

33 (53.2) 29 (46.8) 0 (0.0) 0 (0.0) 0 (0.0)

0.002

0.016

*

Except where otherwise indicated, values are mean ± SD. P-values are calculated using Student's t test (age), ManneWhitney U test (height, weight, BMI, and corrected AAA), chi-square test (female and presence of radiographic PFOA), and Fisher's exact test (alignment). In these analyses, JMP 11 (Student's t test, KruskaleWallis test, and chi-square test) and R (Fisher's exact test) software were used. Bold type represents a statistically significant result. y

Please cite this article in press as: Iijima H, et al., Coexisting lateral tibiofemoral osteoarthritis is associated with worse knee pain in patients with mild medial osteoarthritis, Osteoarthritis and Cartilage (2017), http://dx.doi.org/10.1016/j.joca.2017.02.801

H. Iijima et al. / Osteoarthritis and Cartilage xxx (2017) 1e8

5

Table II Comparisons of the JKOM scores between knees with and those without radiographic lateral OA used in patients having knees with K/L grade 2 in the medial compartment using rank ANCOVA (n ¼ 100)* JKOM scores

OARSI atlas

K/L

With lateral OA (n ¼ 35)

Pain and stiffness (0e32 points) Activities of daily living (0e40 points) Participation in social activities (0e20 points) General health conditions (0e8 points) Total score (0e100 points)

Without lateral OA (n ¼ 65)

Adjusted P-valuey

With lateral OA (n ¼ 38)

Without lateral OA (n ¼ 62)

Adjusted P-valuey

Mean ± SD

Range

Mean ± SD

Range

Mean ± SD

Range

Mean ± SD

Range

9.89 ± 5.19 8.37 ± 7.02 3.57 ± 3.41

0e21 0e28 0e15

6.20 ± 4.40 5.28 ± 4.83 3.12 ± 3.17

0e18 0e20 0e13

0.011 0.208 0.862

10.0 ± 5.04 8.95 ± 6.76 3.97 ± 3.74

1e21 1e28 0e15

5.94 ± 4.30 4.77 ± 4.59 2.85 ± 2.85

0e18 0e20 0e12

<0.001 0.004 0.248

3.03 ± 1.69 24.9 ± 14.5

0e6 4e66

2.71 ± 1.32 17.3 ± 11.1

0e6 1e49

1.000 0.093

3.26 ± 1.64 26.2 ± 14.0

0e6 4e66

2.55 ± 1.28 16.1 ± 10.5

0e6 1e49

0.015 0.001

*

Values are the mean ± SD and range. The lower and upper values of the ranges of JKOM scores were calculated because of the scattered distribution of the answered items. Adjusted P-values were calculated using a rank ANCOVA adjusted for age, sex, BMI, and presence of radiographic PFOA using R software. Bold type represents a statistically significant result. y

Table III Comparisons of the knee ROM, spatiotemporal gait parameters, and TUG and 5CS times between knees with and those without radiographic lateral OA in patients having knees with K/L grade 2 in the medial compartment using ANCOVA (n ¼ 100)* Variables

ROM,  Flexion Extensionz Spatiotemporal gait parameters Gait velocity, m/s Step length, %height Cadence, steps/min TUG, s 5CS, s * y z

OARSI atlas

K/L

With lateral OA (n ¼ 35)

Without lateral OA (n ¼ 65)

Adjusted P-valuey

With lateral OA (n ¼ 38)

Without lateral OA (n ¼ 62)

Adjusted P-valuey

140.7 ± 12.1 7.35 ± 7.94

147.0 ± 6.80 4.18 ± 5.97

0.083 0.052

141.4 ± 11.8 6.95 ± 7.65

147.0 ± 6.98 4.27 ± 6.15

0.067 0.092

1.16 ± 0.18 35.7 ± 4.80 126.3 ± 10.2 7.83 ± 1.44 9.35 ± 3.18

1.17 ± 0.17 36.7 ± 4.13 123.1 ± 12.9 8.10 ± 1.97 8.90 ± 2.27

0.430 0.907 0.258 0.059 0.828

1.15 ± 0.20 35.8 ± 5.00 124.9 ± 11.9 8.07 ± 1.92 9.66 ± 3.57

1.18 ± 0.15 36.7 ± 3.95 123.8 ± 12.3 7.97 ± 1.73 8.69 ± 1.73

0.922 0.749 0.785 0.756 0.200

Values are the mean ± SD. Adjusted P-values were calculated using an ANCOVA adjusted for age, sex, BMI, and presence of PFOA using JMP 11 software. A negative value for knee extension ROM means that knee is flexed.

Table IV Results of ordinal logistic regression analyses showing association between coexisting radiographic lateral OA and the task-specific knee pain intensity in patients having knees with K/L grade 2 in the medial compartment (n ¼ 100)* Individual item of JKOM pain and stiffness

Do you feel stiffness in your knees when you wake up in the morning? Do you feel pain in your knees when you wake up in the morning? How often do you wake up in the night because of pain in your knees? Do you have pain in your knees when you walk on a flat surface? Do you have pain in your knees when ascending stairs? Do you have pain in your knees when descending stairs? Do you have pain in your knees when bending to floor or standing up? Do you have pain in your knees when standing?

OARSI atlas

K/L

Proportional OR (95% CI)

P-value

Proportional OR (95% CI)

P-value

2.00 (0.86e4.66) 2.11 (0.88e5.08) 1.39 (0.59e3.25) 1.52 (0.64e3.58) 2.75 (1.09e6.92) 4.62 (1.87e11.4) 2.06 (0.87e4.88) 2.51 (1.04e6.05)

0.107 0.095 0.448 0.344 0.031 <0.001 0.099 0.040

2.07 (0.92e4.66) 2.20 (0.95e5.11) 1.52 (0.67e3.43) 2.43 (1.05e5.63) 3.11 (1.27e7.60) 4.76 (1.98e11.4) 3.80 (1.60e9.04) 3.43 (1.45e8.14)

0.077 0.066 0.313 0.038 0.013 <0.001 0.003 0.005

See Supplementary Table 1 for details of differences in individual item of the JKOM pain and stiffness between knees with and without radiographic lateral OA. Bold type represents a statistically significant result. * Proportional OR (95% CI) for a task-specific knee pain intensity (0: no pain, 1: mild pain, 2: moderate/severe pain) was calculated to indicate predictive ability of the presence of radiographic lateral OA while simultaneously including (one-step model) age, sex, BMI, and presence of radiographic PFOA in the ordinal regression model using JMP 11 software.

lateral OA. While spatiotemporal gait parameters, TUG and 5CS times, and knee ROM did not differ significantly between patients with and without lateral OA when the two different definitions were used, our findings indicate that knees with coexisting lateral OA in mild medial OA may represent a more symptomatic disease compared to medial OA alone. The OARSI atlas grading and K/L classification systems are two of the most widely used methods for radiographic definition of knee OA. While similar cutoffs have been suggested for both systems18, a recent study showed that the OARSI atlas yielded a higher

prevalence of radiographic OA compared with the K/L classification19. However, in the current study, we confirmed that 35 (35.0%) knees were defined by the OARSI atlas as having lateral compartment radiographic OA, compared to 38 (38.0%) knees for the K/L classification. Inconsistencies in the results of previous studies could be attributed to the classification used. We considered lateral OA to be present if marginal osteophytes were confirmed in either the femur or tibia, which is different from the original K/L classification. The lower cutoff for K/L grading would increase its sensitivity for detecting lateral disease.

Please cite this article in press as: Iijima H, et al., Coexisting lateral tibiofemoral osteoarthritis is associated with worse knee pain in patients with mild medial osteoarthritis, Osteoarthritis and Cartilage (2017), http://dx.doi.org/10.1016/j.joca.2017.02.801

6

H. Iijima et al. / Osteoarthritis and Cartilage xxx (2017) 1e8

The current study is the first to show that patients with coexisting lateral OA had worse knee pain. The mean group differences in JKOM score correspond to minimal yet clinically important differences (i.e., an absolute change of 10% for JKOM pain and stiffness [3.2 points] and activities of daily living [4.0 points]) according to the Outcome Measures in Rheumatology Clinical Trials and OARSI set of responder criteria27 when using the K/L classification. Although we cannot explain the mechanism driving the worse symptoms in patients with coexisting lateral OA, it should be noted that those knees had higher odds of reporting knee pain during weight-bearing related activities, such as stairs climbing and standing (Table IV). Weight-bearing pain is suggested to capture a different aspect of pain than is noneweight-bearing pain28, and is associated with magnetic resonance imaging (MRI) detected bone marrow lesion and joint effusion29. Although knees with coexisting lateral OA had a higher proportion of radiographic PFOA, the association between lateral OA and knee pain was significant, even after including radiographic PFOA as a covariate, indicating that the cause of worse knee pain during weight-bearing activities in knees with coexisting lateral OA is independent of PF joint disease. The proportional OR and its 95% CI were wide, indicating statistical instability. Thus, the effect estimate of weightbearing pain should be interpreted with caution. Pain in OA is generally characterized as pain that occurs during weight-bearing; thus, further studies with large sample sizes that investigate the relationship of lateral OA with knee pain would be of interest. Osteophyte in the lateral compartment and other knee joint compartments is suggested to be an indicator of lower knee flexion ROM30,31. However, knee ROM did not significantly differ between knees with and without lateral OA, after adjustment for age, sex, BMI, and presence of PFOA. In these analyses, the presence of PFOA was significantly associated with lower flexion ROM (data not shown), indicating an important role of PFOA for predicting lower knee flexion rather than that in the lateral OA in tibiofemoral joint. Knee ROM is generally considered to be restricted as the disease severity progressed; thus, further studies would be warranted to detect a risk factor of lower knee ROM in a mild stage of knee OA, which would be a basis for prevention of restriction of ROM. Marginal osteophytes were first described as an early sign of OA by Kellgren and Lawrence4. However, many authors have questioned the validity of a radiographic diagnosis and its ability to predict cartilage damage, particularly in the lateral compartment11,32,33. We could not evaluate actual cartilage damage in the lateral compartment; thus, osteophytes in the lateral compartment in the current study might be merely the signal feature of more severe tibiofemoral disease. Given that most of the knees with coexisting lateral OA had only marginal osteophytes without JSN, marginal osteophytes in the lateral compartment might be a sensitive marker of knee pain. Evaluation of lateral cartilage damage in knees with lateral osteophytes using MRI would be of particular interest to clarify differences between knees with and without lateral osteophyte. On a related note, an increase in varus and valgus alignments are associated with a reduced and increased risk of incidental damage in the lateral compartment34e36. Approximately 65% of patients in this study had a varus alignment knee. However, in this study, knees with and without lateral OA had a similar AAA, and most of the knees with valgus alignment lack radiographic lateral OA, indicating that frontal plane alignment had only a weak influence on the prevalence of lateral OA. We found that approximately 35e40% patients had lateral disease, which is within the previously reported prevalence range for lateral OA (10e55%)11,12,37e39. It is unknown whether these bicompartmental OA cases represent a different subgroup with a different pathophysiology than that of unicompartmental medial OA.

Demographic characteristics, particularly age, were similar in both groups; this may indicate that these groups are consecutive stages of the same disease. Bicompartmental OA is a significantly more inflammatory disease than unicompartmental OA13, and an inflammatory response in the synovial tissue is strongly associated with knee pain40, which might explain why knees with coexisting lateral OA had worse clinical symptoms. Furthermore, Nagaosa et al.39 suggested that an osteophyte at the lateral tibial plateau mainly appears as a traction spar extending upwards when the medial compartment is narrowed. Given that removal of osteophytes in the lateral compartment increases the varusevalgus motion41, the presence of lateral osteophytes in patients with coexisting lateral OA might be attributed to varusevalgus instability. Investigating the difference in varusevalgus instability between knees with and knees without lateral osteophytes in medial knee OA would be of interest in the verification of this hypothesis. However, knees with and knees without lateral OA could have different stages of the same disease, and knees without lateral OA can progress to having lateral OA. A prospective study with a large sample size for follow-up of the progression of the disease in patients without lateral OA would help address this question. The present study had some limitations. First, even as performed by a trained and experienced examiner, radiographic assessment can be a potential source of variability in diagnosing lateral OA. Despite this potential source, the inter-rater reliability in this study (k ¼ 0.63e0.79) was fair to good22, which is a similar value compared with previous studies19,42,43; thus, substantial reliability of radiographic assessment was achieved. Second, knee pain location was not evaluated in the current study. The PF joint is an important source of symptoms associated with knee OA44, and PFOA often coexists with bicompartmental medial and lateral OA37,38. The possibility that knees with coexisting lateral OA exhibited knee pain in their PF joint cannot be excluded, although the presence of PFOA was included as a covariate in the rank ANCOVA, ANCOVA, and ordinal logistic regression model. Third, we did not consider the effect of the contralateral (non-index) knee, particularly contralateral knee pain, on outcome variables. Bilateral knee OA is common in the Japanese population, accounting for 50% of knee OA with K/L grade 29; contralateral knee effects may contribute to the person-specific assessment, such as TUG and 5CS times and JKOM score. Knee pain in the contralateral knee is known to influence the score of functional measurements45,46. However, there are conflicting reports that contralateral knee pain and radiographic disease do not influence the score of functional measurements47,48. Further studies to investigate the effects of coexisting lateral OA on pain and functional outcomes with consideration of contralateral knee effects would be of interest. Fourth, the participants in this study were recruited from a rural mountainous community in Japan, and only those who approved a distributed advertisement were included in the analyses, which may attribute to higher proportion of female (80%) included in this study. Therefore, the participants may not be representative of a general population with knee OA. Finally, lack of patient information about pain medication and injury history may have restricted our analysis. Previous joint injury is a risk factor for progressing lateral JSN and osteophyte formation49. Injury history might attribute to the higher rates of lateral OA in this study. In conclusion, knees with coexisting lateral OA had more severe knee pain than those with medial OA alone, and the presence of lateral OA was significantly associated with a higher risk of knee pain while ascending/descending stairs and standing in patients with mild medial knee OA. These results were similar regardless of the use of OARSI atlas grading system or the K/L classification, indicating that coexisting lateral OA with mild medial OA in knees may represent a more symptomatic disease.

Please cite this article in press as: Iijima H, et al., Coexisting lateral tibiofemoral osteoarthritis is associated with worse knee pain in patients with mild medial osteoarthritis, Osteoarthritis and Cartilage (2017), http://dx.doi.org/10.1016/j.joca.2017.02.801

H. Iijima et al. / Osteoarthritis and Cartilage xxx (2017) 1e8

Authors' contributions All authors have made substantial contributions to (1) the conception and design of the study, acquisition of data, or analysis and interpretation of data; (2) drafting the article or revising it critically for important intellectual content; and (3) final approval of the version to be submitted. The specific contributions of the authors are as follows: (1) Conception and design of the study: HI, TA, KN, HI, NF, TI, and SM. (2) Analysis and interpretation of the data: HI, TA, KN, HI, HK, and TI. (3) Drafting of the article: HI, KN, NF, TI, EK, and SM. (4) Critical revision of the article for important intellectual content: HI, TA, KN, NF, and TI. (5) Final approval of the article: HI, TA, KN, HI, NF, TI, EK, HK, and SM. (6) Statistical expertise: HI, NF, and TI. (7) Obtaining of funding: TA and SM. (8) Collection and assembly of data: HI, TA, and NF.

6.

7.

8.

9.

10.

11. Conflicts of interest The authors have no financial support or other benefits from commercial sources for the work reported in the manuscript, nor any other financial interests that could create a potential conflict of interest or the appearance of a conflict of interest with regard to the work.

12.

13. Role of the funding source This work was supported by Grants-in-Aid from the Comprehensive Research on Aging and Health Science Research Grants for Dementia R&D (no. 16dl0110007h0003) and the Japan Society for the Promotion of Science (no. 16dk0110007h0003), from the Ministry of Education, Culture, Sports, Science, and Technology, and from the Ministry of Health, Labor, and Welfare.

14.

15.

Acknowledgments The authors thank Ms Yuko Yamamoto, Mr Masakazu Hiraoka, Mr Kazuyuki Miyanobu, and Mr Masashi Jinnouchi (Nozomi Orthopaedic Clinic, Hiroshima) for assistance and advice.

16.

Supplementary data

17.

Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.joca.2017.02.801. 18. References 1. Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, Ezzati M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380: 2163e96. 2. VanItallie TB. Gout: epitome of painful arthritis. Metabolism 2010;59(Suppl 1):S32e6. 3. Felson DT. Identifying different osteoarthritis phenotypes through epidemiology. Osteoarthr Cartil 2010;18:601e4. 4. Kellgren JH, Lawrence JS. Radiological assessment of osteoarthrosis. Ann Rheum Dis 1957;16:494e502. 5. Kerkhof HJ, Meulenbelt I, Akune T, Arden NK, Aromaa A, Bierma-Zeinstra SM, et al. Recommendations for standardization and phenotype definitions in genetic studies of

19.

20.

21.

7

osteoarthritis: the TREAT-OA consortium. Osteoarthr Cartil 2011;19:254e64. Altman RD, Gold GE. Atlas of individual radiographic features in osteoarthritis, revised. Osteoarthr Cartil 2007;15(Suppl A): A1eA56. Duncan RC, Hay EM, Saklatvala J, Croft PR. Prevalence of radiographic osteoarthritiseit all depends on your point of view. Rheumatology (Oxford) 2006;45:757e60. Wise BL, Niu J, Yang M, Lane NE, Harvey W, Felson DT, et al. Patterns of compartment involvement in tibiofemoral osteoarthritis in men and women and in whites and African Americans. Arthritis Care Res (Hoboken) 2012;64:847e52. Muraki S, Oka H, Akune T, Mabuchi A, En-yo Y, Yoshida M, et al. Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study. Osteoarthr Cartil 2009;17:1137e43. Johnson F, Leitl S, Waugh W. The distribution of load across the knee. A comparison of static and dynamic measurements. J Bone Jt Surg Br 1980;62:346e9. Faschingbauer M, Renner L, Waldstein W, Boettner F. Are lateral compartment osteophytes a predictor for lateral cartilage damage in varus osteoarthritic knees?: Data from the Osteoarthritis Initiative. Bone Jt J 2015;97-B:1634e9. Cotofana S, Buck R, Wirth W, Roemer F, Duryea J, Nevitt M, et al. Cartilage thickening in early radiographic knee osteoarthritis: a within-person, between-knee comparison. Arthritis Care Res (Hoboken) 2012;64:1681e90. Moradi B, Rosshirt N, Tripel E, Kirsch J, Barie A, Zeifang F, et al. Unicompartmental and bicompartmental knee osteoarthritis show different patterns of mononuclear cell infiltration and cytokine release in the affected joints. Clin Exp Immunol 2015;180:143e54. Bedson J, Croft PR. The discordance between clinical and radiographic knee osteoarthritis: a systematic search and summary of the literature. BMC Musculoskelet Disord 2008;9:116. Felson DT, Niu J, Guermazi A, Sack B, Aliabadi P. Defining radiographic incidence and progression of knee osteoarthritis: suggested modifications of the Kellgren and Lawrence scale. Ann Rheum Dis 2011;70:1884e6. Wise BL, Kritikos L, Lynch JA, Liu F, Parimi N, Tileston KL, et al. Proximal femur shape differs between subjects with lateral and medial knee osteoarthritis and controls: the Osteoarthritis Initiative. Osteoarthr Cartil 2014;22:2067e73. Van Ginckel A, Bennell KL, Campbell PK, Wrigley TV, Hunter DJ, Hinman RS. Location of knee pain in medial knee osteoarthritis: patterns and associations with self-reported clinical symptoms. Osteoarthr Cartil 2016;24:1135e42. Lohmander LS, Ostenberg A, Englund M, Roos H. High prevalence of knee osteoarthritis, pain, and functional limitations in female soccer players twelve years after anterior cruciate ligament injury. Arthritis Rheum 2004;50:3145e52. Culvenor AG, Engen CN, Oiestad BE, Engebretsen L, Risberg MA. Defining the presence of radiographic knee osteoarthritis: a comparison between the Kellgren and Lawrence system and OARSI atlas criteria. Knee Surg Sports Traumatol Arthrosc 2015;23:3532e9. Gudbergsen H, Lohmander LS, Jones G, Christensen R, Bartels EM, Danneskiold-Samsoe B, et al. Correlations between radiographic assessments and MRI features of knee osteoarthritisea crosssectional study. Osteoarthr Cartil 2013;21:535e43. Guccione AA, Felson DT, Anderson JJ. Defining arthritis and measuring functional status in elders: methodological issues in the study of disease and physical disability. Am J Public Health 1990;80:945e9.

Please cite this article in press as: Iijima H, et al., Coexisting lateral tibiofemoral osteoarthritis is associated with worse knee pain in patients with mild medial osteoarthritis, Osteoarthritis and Cartilage (2017), http://dx.doi.org/10.1016/j.joca.2017.02.801

8

H. Iijima et al. / Osteoarthritis and Cartilage xxx (2017) 1e8

22. Fleiss JL. Statistical Methods for Rates and Proportions. New York: Wiley; 1981. 23. Erceg-Hurn DM, Mirosevich VM. Modern robust statistical methods: an easy way to maximize the accuracy and power of your research. Am Psychol 2008;63:591e601. 24. Olejnik SF, Algina J. Parametric ANCOVA and the rank transform ANCOVA when the data are conditionally nonnormal and heteroscedastic. J Educ Behav Statistics 1984;9: 129e49. 25. McCullagh P. Regression models for ordinal data. J R Stat Soc Series B (Methodological) 1980:109e42. 26. Iijima H, Fukutani N, Aoyama T, Fukumoto T, Uritani D, Kaneda E, et al. Clinical impact of coexisting patellofemoral osteoarthritis in Japanese patients with medial knee osteoarthritis. Arthritis Care Res (Hoboken) 2016;68:493e501. 27. Pham T, van der Heijde D, Altman RD, Anderson JJ, Bellamy N, Hochberg M, et al. OMERACT-OARSI initiative: Osteoarthritis Research Society International set of responder criteria for osteoarthritis clinical trials revisited. Osteoarthr Cartil 2004;12: 389e99. 28. Stratford PW, Kennedy DM, Woodhouse LJ, Spadoni GF. Measurement properties of the WOMAC LK 3.1 pain scale. Osteoarthr Cartil 2007;15:266e72. 29. Lo GH, McAlindon TE, Niu J, Zhang Y, Beals C, Dabrowski C, et al. Bone marrow lesions and joint effusion are strongly and independently associated with weight-bearing pain in knee osteoarthritis: data from the osteoarthritis initiative. Osteoarthr Cartil 2009;17:1562e9. 30. Suzuki T, Motojima S, Saito S, Ishii T, Ryu K, Ryu J, et al. Osteoarthritis of the patella, lateral femoral condyle and posterior medial femoral condyle correlate with range of motion. Knee Surg Sports Traumatol Arthrosc 2013;21: 2584e9. 31. Ozdemir F, Tukenmez O, Kokino S, Turan FN. How do marginal osteophytes, joint space narrowing and range of motion affect each other in patients with knee osteoarthritis. Rheumatol Int 2006;26:516e22. 32. Kijowski R, Blankenbaker DG, Stanton PT, Fine JP, De Smet AA. Radiographic findings of osteoarthritis versus arthroscopic findings of articular cartilage degeneration in the tibiofemoral joint. Radiology 2006;239:818e24. 33. Koski JM, Kamel A, Waris P, Waris V, Tarkiainen I, Karvanen E, et al. Atlas-based knee osteophyte assessment with ultrasonography and radiography: relationship to arthroscopic degeneration of articular cartilage. Scand J Rheumatol 2016;45: 158e64. 34. Sharma L, Chmiel JS, Almagor O, Felson D, Guermazi A, Roemer F, et al. The role of varus and valgus alignment in the initial development of knee cartilage damage by MRI: the MOST study. Ann Rheum Dis 2013;72:235e40. 35. Khan FA, Koff MF, Noiseux NO, Bernhardt KA, O'Byrne MM, Larson DR, et al. Effect of local alignment on compartmental patterns of knee osteoarthritis. J Bone Jt Surg Am 2008;90: 1961e9.

36. Felson DT, Niu J, Gross KD, Englund M, Sharma L, Cooke TD, et al. Valgus malalignment is a risk factor for lateral knee osteoarthritis incidence and progression: findings from the Multicenter Osteoarthritis Study and the Osteoarthritis Initiative. Arthritis Rheum 2013;65:355e62. 37. Yamabe E, Ueno T, Miyagi R, Watanabe A, Guenzi C, Yoshioka H. Study of surgical indication for knee arthroplasty by cartilage analysis in three compartments using data from Osteoarthritis Initiative (OAI). BMC Musculoskelet Disord 2013;14:194. 38. Ledingham J, Regan M, Jones A, Doherty M. Radiographic patterns and associations of osteoarthritis of the knee in patients referred to hospital. Ann Rheum Dis 1993;52:520e6. 39. Nagaosa Y, Lanyon P, Doherty M. Characterisation of size and direction of osteophyte in knee osteoarthritis: a radiographic study. Ann Rheum Dis 2002;61:319e24. 40. Torres L, Dunlop DD, Peterfy C, Guermazi A, Prasad P, Hayes KW, et al. The relationship between specific tissue lesions and pain severity in persons with knee osteoarthritis. Osteoarthr Cartil 2006;14:1033e40. 41. Pottenger LA, Phillips FM, Draganich LF. The effect of marginal osteophytes on reduction of varus-valgus instability in osteoarthritic knees. Arthritis Rheum 1990;33:853e8. 42. Damen J, Schiphof D, Wolde ST, Cats HA, Bierma-Zeinstra SM, Oei EH. Inter-observer reliability for radiographic assessment of early osteoarthritis features: the CHECK (cohort hip and cohort knee) study. Osteoarthr Cartil 2014;22:969e74. 43. Klara K, Collins JE, Gurary E, Elman SA, Stenquist DS, Losina E, et al. Reliability and accuracy of cross-sectional radiographic assessment of severe knee osteoarthritis: role of training and experience. J Rheumatol 2016;43:1421e6. 44. Kornaat PR, Bloem JL, Ceulemans RY, Riyazi N, Rosendaal FR, Nelissen RG, et al. Osteoarthritis of the knee: association between clinical features and MR imaging findings. Radiology 2006;239:811e7. 45. Cotofana S, Wirth W, Pena Rossi C, Eckstein F, Gunther OH. Contralateral knee effect on self-reported knee-specific function and global functional assessment: data from the Osteoarthritis Initiative. Arthritis Care Res (Hoboken) 2015;67:374e81. 46. Riddle DL, Stratford PW. Unilateral vs bilateral symptomatic knee osteoarthritis: associations between pain intensity and function. Rheumatology (Oxford) 2013;52:2229e37. 47. Bindawas SM, Vennu V, Al Snih S. Differences in health-related quality of life among subjects with frequent bilateral or unilateral knee pain: data from the Osteoarthritis Initiative study. J Orthop Sports Phys Ther 2015;45:128e36. 48. Marmon AR, Zeni Jr JA, Snyder-Mackler L. Perception and presentation of function in patients with unilateral versus bilateral knee osteoarthritis. Arthritis Care Res (Hoboken) 2013;65:406e13. 49. Sward P, Kostogiannis I, Neuman P, Von Porat A, Boegard T, Roos H. Differences in the radiological characteristics between post-traumatic and non-traumatic knee osteoarthritis. Scand J Med Sci Sports 2010;20:731e9.

Please cite this article in press as: Iijima H, et al., Coexisting lateral tibiofemoral osteoarthritis is associated with worse knee pain in patients with mild medial osteoarthritis, Osteoarthritis and Cartilage (2017), http://dx.doi.org/10.1016/j.joca.2017.02.801