Association of Thigh Muscle Strength With Knee Symptoms and Radiographic Disease Stage of Osteoarthritis: Data From the Osteoarthritis Initiative

Transcription

1 Arthritis Care & Research Vol. 66, No. 9, September 2014, pp DOI /acr , American College of Rheumatology ORIGINAL ARTICLE Association of Thigh Muscle Strength With Knee Symptoms and Radiographic Disease Stage of Osteoarthritis: Data From the Osteoarthritis Initiative ANJA RUHDORFER, 1 WOLFGANG WIRTH, 2 WOLFGANG HITZL, 1 MICHAEL NEVITT, 3 AND FELIX ECKSTEIN, 2 FOR THE OSTEOARTHRITIS INITIATIVE INVESTIGATORS Objective. To determine whether thigh muscle strength differs between symptomatic and asymptomatic knees, and/or different radiographic strata of knee osteoarthritis (KOA). Methods. Isometric extensor and flexor strength were analyzed in 3,809 Osteoarthritis Initiative participants (2,201 women and 1,608 men) with central radiographic Kellgren/Lawrence (K/L) grade readings. Isometric strength measurements were stratified by radiographic disease status (K/L grades 0, 1, 2, and 3/4) and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain scores. Age-adjusted, separate-slopes analysis of covariance models was used to compare strength between symptomatic (WOMAC score range 5 20) and asymptomatic (WOMAC score 0) legs within and across K/L grade strata. Exploratory analyses focused on strength normalized to body weight and symptom frequency. Results. Isometric strength was significantly lower in symptomatic than in asymptomatic legs: 11 to 13% for extensor strength and 7 to 16% for flexor strength (P < for both) in men, and 9 to 17% (P 0.029) for extensor strength, and 10 to 21% (P 0.049) for flexor strength in women. Similar observations were made for pain frequency strata. Extensor and flexor strength were not significantly different across K/L grade strata in asymptomatic legs in either sex (P > 0.12). However, strength normalized to body weight was lower at higher K/L grades in both sexes (P < 0.02) because the body mass index was greater in participants with more advanced radiographic disease. Conclusion. Knee symptoms (i.e., pain) appear to be the relevant determinant of isometric knee extensor and flexor strength in KOA, whereas no direct association between strength and radiographic severity was observed. These findings suggest that the reduction in thigh muscle strength in KOA is related to pain but not to the structural (radiographic) disease status. INTRODUCTION Thigh muscle weakness has been commonly observed in patients with knee osteoarthritis (KOA) (1 5) or knee pain (6,7), and low muscle strength has been associated with Image and clinical data acquisition was supported by the Osteoarthritis Initiative (OAI), a public-private partnership comprising 5 contracts (N01-AR , N01-AR , N01-AR , N01-AR , and N01-AR ) funded by the NIH, a branch of the Department of Health and Human Services, and conducted by the OAI Study Investigators. Private funding partners of the OAI include Merck Research Laboratories, Novartis Pharmaceuticals Corporation, GlaxoSmithKline, and Pfizer Inc. Private sector funding for the OAI is managed by the Foundation for the NIH. The image analysis was supported by the Paracelsus Medical University Research Fund (PMU FFF E-11/14/ 073-WIR). 1 Anja Ruhdorfer, MD, Wolfgang Hitzl, DrMSc, MSc: Paracelsus Medical University, Salzburg, Austria; 2 Wolfgang incident symptomatic, but not with incident radiographic KOA (8). Muscle weakness is considered a potentially modifiable (9) risk factor of KOA and quadriceps muscle strengthening is recommended by the current Osteoarthritis Research Wirth, PhD, Felix Eckstein, MD: Paracelsus Medical Univer sity, Salzburg, Austria, and Chondrometrics GmbH, Ainring, Germany; 3 Michael Nevitt, PhD: University of California, San Francisco. Dr. Wirth has received consultant fees, speaking fees, and/or honoraria (less than $10,000 each) from Merck Sorono and Geneva, and owns stock and/or stock options in Chondrometrics GmbH. Dr. Eckstein has received consultant fees, speaking fees, and/or honoraria (less than $10,000 each) from Merck Sorono and AbbVie, and owns stock and/or stock options in Chondrometrics GmbH. Address correspondence to Anja Ruhdorfer, MD, Institute of Anatomy, Paracelsus Medical University, Strubergasse 21, A5020 Salzburg, Austria. anja.ruhdorfer@pmu.ac.at. Submitted for publication July 30, 2013; accepted in revised form February 18,

2 Symptoms and Disease Severity in KOA Thigh Muscle Strength 1345 Significance & Innovations Thigh muscle strength is lower in symptomatic than in asymptomatic legs; this applies to legs with mild, moderate, or severe, or without radiographic knee osteoarthritis (KOA). In contrast, muscle strength does not decline with higher grades of radiographic KOA, and symptomatic legs without radiographic KOA show lower strength compared with asymptomatic legs with severe radiographic KOA. These findings apply to extensor and flexor strength, and to both men and women. The findings stress the important role of pain, rather than structural (radiographic) disease stage, on muscle status and suggest that management of pain may be as important in maintaining adequate muscle strength as exercise. Society International KOA treatment guidelines (10). Previous studies reported a reduction of knee pain (11 13) following quadriceps strengthening exercise but no significant modification of structural (i.e., radiographic) disease progression (11) with training intervention. Although a discordance between structural (radiographic) changes and knee pain has been reported (14), pain and radiographic status are correlated to some extent as shown by a recent between-knee, within-person comparison (15). Previous studies reported a reduction in strength (5,16,17) and muscle quality (strength per muscle area unit) (18) in subjects with radiographic KOA, but it is controversial, i.e., whether muscle strength is directly associated with structural (radiographic) knee status, or whether the association is indirect, given the associations between radiographic change and knee pain (15). Previous studies on the association between muscle strength, pain, and radiographic change did not include the whole spectrum of radiographic status (from no to severe OA) (5,8,16,18,19) and did not fully dissect pain and radiographic strata in their analysis (7). For instance, Berger et al (7) reported a stronger correlation between isometric quadriceps muscle strength and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) total scores than with radiographic Kellgren/ Lawrence (K/L) grade status. Yet, pain and structural (radiographic) status are correlated with each other (15,20 22) and the WOMAC total score scale (0 96) encompasses other components than knee pain (i.e., stiffness and function). Also, analyses across men and women were combined, which increases the variability in strength values between groups and does not account for potential differences in pain/structure relationships between both sexes (23). In an attempt to disentangle the relationship between muscle strength, pain, and structural (radiographic) status, our current cross-sectional study uses a large sample of subjects with and at increased risk for KOA, covering the complete spectrum of radiographic status (from none to end-stage radiographic KOA [K/L grade 0 4]). Specifically, use of this large sample permitted us to study the relationship between muscle strength in symptomatic versus asymptomatic knees within each K/L grade and in asymptomatic knees across all radiographic strata (K/L grade 0 4). Further, previous findings of changes in gait patterns (3,24) and coactivation/elongation of the hamstrings (25) suggest that KOA may not only affect the quadriceps, but also other thigh muscles, specifically the hamstrings. Some studies also reported flexor strength (1) and/or the extensor:flexor strength ratio to be reduced in osteoarthritic knees (26). Therefore, our analyses not only included extensor strength, but also flexor strength and the extensor:flexor strength ratio. Given systematic differences of strength between men and women (27), and to account for a potentially different relationship between strength versus symptoms and strength versus radiographic status between sexes (28), current analyses were stratified for men and women. Specifically, we attempt to answer whether symptoms and/or radiographic status are more strongly associated with thigh muscle extensor and flexor strength in KOA. PATIENTS AND METHODS Participants. The Osteoarthritis Initiative (OAI) cohort comprises women and men between ages years from various ethnic and socioeconomic backgrounds ( (29). For the current study, the baseline data of the right knees of the incidence and progression cohorts (n 4,674) were included in the analysis. Participants for whom body height, weight, and/or body mass index (BMI) data were not available were excluded from the analysis (n 85). Participants that lacked information on the radiographic status of the right knee (i.e., K/L grading based on the central OAI radiographic readings) (30) (n 321), on symptom frequency (n 6), on the WOMAC knee pain score (31) (n 21), and/or on isometric extensor and/or flexor muscle strength (n 413) were also not included. The following reasons for missing (right extension/flexion) strength data are documented in the OAI database: equipment not yet available when baseline data was acquired (n 341), knee pain (n 2), equipment issues (n 66), refusal to participate (n 2), and other reasons (n 2). We also excluded the data from 36 participants who had a right knee arthroplasty at baseline. Ultimately, 1,610 men and 2,208 women were available for the analyses. Measurement of thigh muscle strength. Baseline measurements for maximum isometric extensor and flexor strength were drawn from the OAI database (clinical data set 0.2.2) (32). These measurements were obtained using the Good Strength Chair (Metitur Oy) (6,33,34), with the participant sitting upright with a fixed pelvis and thigh. The knee was placed in 60 of flexion for testing. In brief, the transducer/load cell was bolted to the lever arm and the lever arm was secured to the leg with a strap 2 cm

3 1346 Ruhdorfer et al proximal to the calcaneal bone. The participant first was familiarized with the testing procedure, performing 2 training trials with 50% effort. Then the participant was encouraged to pull (extension) and push (flexion) against the pad with maximal effort 3 times. The highest of the 3 measurements was used for the analysis (32). Because the strength measurements were taken at an anatomically consistent location, we used the isometric strength measurements directly (and not the moments) to estimate muscle strength, because both the lever arm between the load cell and joint center and that between the muscle tendons and joint center depend on body size and may be assumed to be roughly proportional. Nine participants with KOA (7 women, 2 men) had flexor strength values 10 newtons (N), whereas isometric extensor strength was in a normal range ( N), suggesting that the measurement of the isometric flexor strength was inaccurate. These 9 observations were excluded from the analyses. Evaluation of radiographic and knee symptom status. Radiographic status was assessed based on fixed-flexion radiographs using K/L grades (30) of the central readings provided by the OAI (imaging data set 0.E.2), where K/L grade 0 none, K/L grade 1 doubtful, K/L grade 2 definite, K/L grade 3 moderate, and K/L grade 4 severe radiographic KOA (29,30). Participants with K/L grades 3 and 4 were combined because of the low number of K/L grade 4 knees, but sensitivity analyses were performed to test whether results were consistent between these strata. WOMAC knee pain scores and pain frequency status were collected by the OAI (clinical data set 0.2.2). The WOMAC knee pain score comprises 5 questions scored from 0 (no pain) to 4 (extreme pain), resulting in a point scale with 20 being the worst pain. The 5 questions for the knee pain subscale covered the last 7 days and addressed pain during walking, climbing stairs, lying in bed, sitting or lying down, and standing. To efficiently contrast asymptomatic and symptomatic knees, symptomatic was defined as a WOMAC score 5, based on the Low Intensity Symptom State-Attainment Index (BLISS) cutoff reported by Bellamy et al (35), and asymptomatic as a WOMAC score of 0. Within the selected cohort, 19.7% had a WOMAC knee pain score 5. Among all knees with any baseline pain (WOMAC 0) this criterion (WOMAC 5) was fulfilled by 33.2% of all knees. Symptom frequency (OAI variable P01RKSX) was related to pain during the past 12 months. Frequent symptoms (Sx2) were defined as pain, aching, and stiffness on most days of at least 1 month (greater or equal to half the days of a month) in the past 12 months. Symptoms on less than half the days of a month within the last 12 months were considered infrequent (Sx1); no symptoms during the past year were classified as asymptomatic (Sx0). Statistical analysis. The primary focus of our analysis was the cross-sectional association of baseline extensor and flexor strength, in men and women separately, with the presence versus absence of knee pain (as defined above using WOMAC scores). After ensuring normality of the data, a separate-slopes analysis of covariance (ANCOVA) model, with age as a covariate, was used to apply global tests and planned contrasts (with 95% confidence intervals) to the difference between symptomatic versus asymptomatic knees across all K/L grade strata (global), and within each K/L grade stratum (contrasts). As a secondary analytic focus, the same model was used to compare the results across K/L grade strata within asymptomatic knees in an attempt to eliminate variability of pain between K/L grade strata from the analysis. For exploratory purposes, the above analyses were also performed for the extensor:flexor ratio, for quadriceps and flexor strength normalized to individual body weight, and with symptom frequency rather than with WOMAC pain scores to classify symptomatic (Sx2) versus asymptomatic (Sx0) knees. To achieve normal distributions, the extensor: flexor strength ratio was log-transformed. Further, BMI scores across K/L grades were compared using analysis of variance and Bonferroni correction across asymptomatic legs (WOMAC score 0). All analyses were performed using Statistica (data analysis software system; StatSoft), version 10 ( com), with the required significance level being set to P less than RESULTS Demographics. Of the 2,201 women and 1,608 men (demographics in Table 1), 35% and 38% had K/L grade 0, 18% and 19% had K/L grade 1, 31% and 23% had K/L grade 2, and 16% and 20% had K/L grade 3/4, respectively. Within these strata, the percentage of asymptomatic and symptomatic knees was 47% and 12% in men and 52% and 11% in women in K/L grade 0; 44% and 19%, and 49% and 12% in K/L grade 1; 32.5% and 26%, and 43.5% and 17% in K/L grade 2; and 23% and 40%, and 23% and 29.5% in K/L grade 3/4, respectively. Mean SD BMI was higher at higher K/L grades in both sexes (both P 0.001) as follows: kg/m 2 in K/L grade 0, kg/m 2 in K/L grade 1, kg/m 2 in K/L grade 2, and kg/m 2 in K/L grade 3/4. Muscle strength in symptomatic versus asymptomatic knees. Isometric extensor strength was significantly lower in symptomatic than in asymptomatic legs in men (P ) and women (P 0.029, global tests). The percent differences between the mean values of symptomatic and asymptomatic knees within the 4 K/L grade strata ranged from 10.8 to 13.4% (P ) in men and from 9.4 to 16.2% (P ) in women (Figure 1, contrasts). Isometric flexor strength was also lower in symptomatic versus asymptomatic legs in men (P ) and women (P 0.049, global tests). The percent differences within the K/L grade strata were 7.3 to 15.7% (P ) in men and 10.4 to 21.1% (P ) in women (Figure 1, contrasts). Sensitivity analyses showed that the differences between symptomatic and asymptomatic legs were consistent in K/L grade 3 and K/L grade 4 strata. Muscle strength in asymptomatic knees across different K/L grade strata. In asymptomatic knees, the isometric extensor strength did not significantly differ across K/L

4 Symptoms and Disease Severity in KOA Thigh Muscle Strength 1347 Table 1. Demographic data of women and men within each K/L grade (0 to 3/4) and WOMAC score (0 asymptomatic and >5 symptomatic) stratum* No. Age, years Weight, kg BMI, kg/m 2 WOMAC No. Age, years Weight, kg BMI, kg/m 2 WOMAC K/L grade 0 All WOMAC WOMAC K/L grade 1 All WOMAC WOMAC K/L grade All WOMAC WOMAC K/L grade 3/4 All WOMAC WOMAC * Values are the mean SD unless indicated otherwise. K/L Kellgren/Lawrence; WOMAC Western Ontario and McMaster Universities Osteoarthritis Index; BMI body mass index. grade strata (0, 1, 2, 3/4) in either sex (P 0.72 in men and 0.28 in women) (Figure 2). Similar results were observed for flexor strength (P 0.32 in men and 0.12 in women) (Figure 2). Exploratory analyses. Isometric extensor strength normalized to body weight was lower in symptomatic (WOMAC score 5) than in asymptomatic (WOMAC score 0) legs in both sexes (P in men and women) (Table 2). The percent differences between the mean values of symptomatic and asymptomatic knees within each K/L grade stratum ranged from 12.7 to 16.9% in men (P ) and from 14.8 to 22.9% in women (P ). Similar results were observed for flexor Figure 1. Isometric extensor (top) and flexor (bottom) strength measurements (newtons [N]) in women and men with asymptomatic legs (Western Ontario and McMaster Universities Osteoarthritis Index [WOMAC] score 0; lightly shaded bars) and symptomatic legs (WOMAC score 5; darkly shaded bars) in each Kellgren/Lawrence grade (KLG) stratum. Measurements are depicted as mean 95% confidence interval and with percent differences and P values for differences in strength between symptomatic and asymptomatic legs.

5 1348 Ruhdorfer et al Figure 2. Isometric extensor and flexor strength measurements (newtons [N]) in men and women in asymptomatic legs across Kellgren/Lawrence grade (KLG) strata (left) and extensor and flexor strength per body weight (N/kg) in men and women in asymptomatic legs across KLG strata (right). Measurements are depicted as mean 95% confidence intervals, with participant numbers (n) and significant P values between KLG strata. strength normalized to body weight in men (P 0.001) and women (P 0.051), with percent differences of 10.3 to 19.6% (P ) and 15.9 to 26.3% (P ) (Table 2). In asymptomatic legs, the separate-slope ANCOVA model showed that isometric extensor and flexor strength was lower with higher K/L grades in men (P and 0.02, respectively) and women (P and 0.02, respectively) when normalized to body weight (Figure 2 and Table 2). However, the extensor:flexor strength ratio was not significantly different between symptomatic and asymptomatic knees, and did not differ across K/L grades in either sex (P 0.13). Exploratory analyses with symptom frequency (Sx) confirmed the results for symptomatic versus asymptomatic (WOMAC) legs (Tables 3 and 4). Also, extensor strength was not significantly different in asymptomatic legs (Sx 0) across K/L grade strata in men (P 0.74) or women (P 0.67), and the same was observed for flexor strength in men (P 0.62). Only in women was the global P value 0.03, and flexor strength was lower in legs with K/L grade 1 than in all other K/L grade strata (P 0.03). DISCUSSION Quadriceps weakness in KOA has been described previously (1 5). However, these analyses have not taken into account knee pain (1,2,4,5) nor have they covered the entire spectrum of radiographic OA (5). Our current study attempts to disentangle the association of knee pain and radiographic disease stage with isometric knee extensor and flexor muscle strength in a large sample, with stratification for men and women, by contrasting symptomatic versus asymptomatic knees within K/L grade strata, and by contrasting K/L grade strata across asymptomatic knees. Our results demonstrate a strong association between knee pain and thigh muscle strength, as legs with symptoms (WOMAC score 5 or frequent symptoms) displayed significantly lower strength than those with no symptoms within each K/L grade stratum. In contrast, we did not find that radiographic disease stage (K/L grade) was related to thigh muscle weakness, as we did not find significant differences in (quadriceps) strength across K/L grade strata in asymptomatic legs. We did not pursue a formal statistical analysis of isometric strength across K/L grade strata in symptomatic knees because it was impossible to ensure a priori that the mean WOMAC pain score would be similar across these; therefore, we were unable to rule out a residual effect of pain on the analysis. However, when visually comparing isometric extensor and flexor strength in men and women across K/L grade strata in symptomatic knees post hoc (Figure 1), there is no indication that strength differs across symptomatic knees with different K/L grades. Of particular note is that asymptomatic knees with K/L grade 3/4 had greater muscle strength than symptomatic K/L grade 0 knees. The exploratory analyses with strength normalized to body weight confirmed the above findings of lower strength in symptomatic than in asymptomatic legs. Importantly, these data also indicated that, in asymptomatic knees, strength normalized to body weight decreased significantly with higher K/L grades, this being due to a higher BMI and body weight and similar isometric strength in knees with higher K/L grades. The WOMAC knee pain score (covering the last 7 days) has been described as a sensitive measure to study the

6 Symptoms and Disease Severity in KOA Thigh Muscle Strength 1349 Table 2. Extensor and flexor strength normalized to body weight in women and men within each K/L grade (0 to 3/4) and WOMAC knee pain score (0 asymptomatic and >5 symptomatic) stratum* WOMAC, no. participants Extensors/ weight, N/kg P Flexors/ weight, N/kg P K/L grade 0 Asymptomatic ( ) ( ) Symptomatic ( ) 1.5 ( ) Asymptomatic ( ) ( ) Symptomatic ( ) 1.9 ( ) K/L grade 1 Asymptomatic ( ) ( ) Symptomatic ( ) 1.2 ( ) Asymptomatic ( ) ( ) 0.07 Symptomatic ( ) 2.0 ( ) K/L grade 2 Asymptomatic ( ) ( ) Symptomatic ( ) 1.2 ( ) Asymptomatic ( ) ( ) 0.02 Symptomatic ( ) 1.8 ( ) K/L grade 3/4 Asymptomatic ( ) ( ) Symptomatic ( ) 1.2 ( ) Asymptomatic ( ) ( ) 0.21 Symptomatic ( ) 1.9 ( ) * Values are the mean (95% confidence interval) unless indicated otherwise. K/L Kellgren/Lawrence; WOMAC Western Ontario and McMaster Universities Osteoarthritis Index. Painful vs. painless. impact of strength on pain and knee function (13). Further, it has been shown that knee pain does not influence the measurements of strength per se (36). Therefore, the observed differences in strength between symptomatic and asymptomatic knees should not be associated with the measurement process of isometric strength, but should reflect true differences in muscle strength. Bellamy et al reported a BLISS cutoff of 25 normalized units (on a 100-point scale, equivalent to 5 units on the 20-point WOMAC knee pain scale), which defines a threshold for state attainment (at low and very low levels) in clinical trials (35). Hence, we chose a cutoff of at least 5 points on the 20-point WOMAC knee pain scale as being symptomatic. We felt this was adequate to contrast symptomatic versus asymptomatic legs (WOMAC score 0), as WOMAC score 5 represented the upper tertile of all participants with any pain in the cohort. Symptom frequency status, on the other hand, only encompasses a 3-grade scale but is related to a longer observation period (the previous year). Despite these conceptional differences, findings based on WOMAC and symptom frequency were highly convergent. The finding of lower isometric muscle strength in symptomatic versus asymptomatic legs is supported by previous studies reporting a protective effect of muscle strength on knee pain (12,37). O Reilly et al observed lower quadriceps strength in symptomatic versus asymptomatic legs, but did not match case and control knees for radiographic disease stage (38). Berger et al reported significant differences in knee extensor strength between WOMAC total score tertiles, but not across different K/L grade strata (K/L grades 2 versus 2 versus 2) (7). However, they did not stratify by sex and did not adjust for pain status within the above radiographic strata, although K/L grade and pain status have been shown to be associated with each other (15). Further, WOMAC total score is a broader assessment of knee symptoms (stiffness, function, pain) and therefore does not exclusively represent deficits in muscle strength related to knee pain. Previously, we performed a between-knee, within-person comparison of bilateral knees with frequent pain, with 1 knee displaying radiographic joint space narrowing (JSN) and the other a normal joint space width (i.e., persons concordant for pain status, but discordant for radiographic JSN status). In this study, we did not detect side differences in thigh muscle anatomic cross-sectional areas (MCSAs) and isometric strength (39). In contrast, a com-

7 1350 Ruhdorfer et al Table 3. Extensor and flexor strength in women and men for K/L grade (0 to 3/4) and symptom frequency strata (0 no pain and 2 frequent pain)* Symptom frequency, no. participants Extensor strength, N P Flexor strength, N P K/L grade 0 No pain ( ) ( ) Frequent pain ( ) 112 ( ) No pain ( ) ( ) Frequent pain ( ) 176 ( ) K/L grade 1 No pain ( ) ( ) 0.81 Frequent pain ( ) 109 ( ) No pain ( ) ( ) 0.12 Frequent pain ( ) 182 ( ) K/L grade 2 No pain ( ) ( ) Frequent pain ( ) 104 (99 110) No pain ( ) ( ) 0.01 Frequent pain ( ) 178 ( ) K/L grade 3/4 No pain ( ) ( ) 0.01 Frequent pain ( ) 105 (98 111) No pain ( ) ( ) 0.65 Frequent pain ( ) 174 ( ) * Values are the mean (95% confidence interval) unless indicated otherwise. K/L Kellgren/Lawrence grade. Painful vs. painless. parison of participants with unilateral frequent knee pain but concordant radiographic (K/L grade) knee status revealed lower thigh MCSAs and isometric strength in the symptomatic knee (6). However, both studies had lower numbers of participants ( 48) and there was concern that findings in control knees may be confounded by the contralateral case knees due to the between-knee, withinperson study design and lack of independence of observations in bilateral knees of the same person. For instance, strength in an asymptomatic knee may be reduced by contralateral pain status through a mechanism of central inhibition (40), or through reduced physical activity/function in general (41). However, our current data are in principal agreement with these findings and confirm, in a very large cohort, that pain is a more relevant factor in determining muscle strength than the radiographic disease stage in KOA in both men and women. In the current study, we did not adjust for pain status of the contralateral extremity. The rationale for this was that a potential inhibitory effect of a symptomatic contralateral (on an asymptomatic ipsilateral) extremity would supposedly negatively affect the latter and would therefore, if relevant at all, only result in an underestimation of the observed differences. We also did not adjust for leg dominance, because the OAI questionnaire (with which leg do you kick a ball?) only assesses coordinative dominance, whereas strength dominance (which leg to climb on a chair?) may be located in the other extremity. Further, an effect of leg dominance on thigh muscle strength was not supported by the findings of Ditroilo et al (42). It has been reported previously that thigh muscle weakness affects gait (43) and knee function (44) and is an important determinant of functional disability (45,46). The current results are suggestive of thigh muscle weakness being caused by pain but not by radiographic changes per se (47). This is in principal agreement with a previous observation that muscle strength is a stronger determinant of disability than radiographic disease severity (45). Alternatively, the observed thigh muscle weakness may be due to inhibitory effects of knee pain on muscle activation, which has been reported previously (25). When normalizing strength per weight, we did find significant differences across K/L grades in participants without pain. However, these differences were related to greater weight (and BMI) for the same isometric strength in subjects with a higher K/L grade. A previous study reported that muscle strength did not differ significantly between BMI strata (19) and it has been shown previously

8 Symptoms and Disease Severity in KOA Thigh Muscle Strength 1351 Table 4. Extensor and flexor strength normalized to body weight in women and men for K/L grade (0 to 3/4) and symptom frequency strata (0 no pain and 2 frequent pain)* Symptom frequency, no. participants Extensor strength, N/kg P Flexor strength, N/kg P K/L grade 0 No pain ( ) ( ) Frequent pain ( ) 1.7 ( ) No pain ( ) ( ) 0.01 Frequent pain ( ) 2.0 ( ) K/L grade 1 No pain ( ) ( ) 0.93 Frequent pain ( ) 1.5 ( ) No pain ( ) ( ) 0.16 Frequent pain ( ) 2.1 ( ) K/L grade 2 No pain ( ) ( ) Frequent pain ( ) 1.3 ( ) No pain ( ) ( ) 0.06 Frequent pain ( ) 2.0 ( ) K/L grade 3/4 No pain ( ) ( ) 0.01 Frequent pain ( ) 1.3 ( ) No pain ( ) ( ) 0.73 Frequent pain ( ) 1.9 ( ) * Values are the mean (95% confidence interval) unless indicated otherwise. K/L Kellgren/Lawrence grade. Painful vs. painless. that men and women with symptomatic (or doctor-diagnosed) KOA have a higher weight and BMI compared to those without symptomatic (or doctor-diagnosed) KOA (48). Hence, previously observed relationships between strength/weight and K/L grade status (17,49,50) may be confounded by an association of weight/bmi with radiographic disease status and may be reflective of people with higher body weight, but without greater strength, being at an increased risk of KOA. However, longitudinal data are needed to test this hypothesis. In conclusion, pain appears to be the relevant determinant of isometric knee extensor and flexor strength whereas no direct association between strength and radiographic severity was observed. Although a reduction in strength was observed at higher K/L grades when strength was normalized to body weight, this was due to a greater BMI (and weight) in subjects with more advanced radiographic disease, but not to a reduction in strength per se. These findings suggest that the reduction in thigh muscle strength in knee OA is related to pain, but not to radiographic (K/L grade) status, and that management of pain may be as important in maintaining adequate muscle strength as exercise. ACKNOWLEDGMENTS We thank the OAI participants, OAI investigators, and OAI Clinical Center staff for generating this publicly available image data set. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Ruhdorfer had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Ruhdorfer, Wirth, Hitzl, Eckstein. Acquisition of data. Ruhdorfer, Wirth, Nevitt. Analysis and interpretation of data. Ruhdorfer, Wirth, Nevitt, Eckstein. ROLE OF THE STUDY SPONSOR The Osteoarthritis Initiative private funding partners (Merck Research Laboratories, Novartis Pharmaceuticals Corporation, GlaxoSmithKline, and Pfizer Inc.) had no role in the study design, data collection, data analysis, or writing of this manuscript. Publication of this article was not contingent on the approval of these sponsors. The manuscript has received the approval of the Osteo-

9 1352 Ruhdorfer et al arthritis Initiative Publication Committee based on a review of its scientific content and data interpretation. ADDITIONAL DISCLOSURE Authors Wirth and Eckstein are employees of Chondrometrics. REFERENCES 1. Fransen M, Crosbie J, Edmonds J. Isometric muscle force measurement for clinicians treating patients with osteoarthritis of the knee. Arthritis Rheum 2003;49: Hurley MV, Scott DL, Rees J, Newham DJ. Sensorimotor changes and functional performance in patients with knee osteoarthritis. Ann Rheum Dis 1997;56: Diracoglu D, Baskent A, Yagci I, Ozcakar L, Aydin R. Isokinetic strength measurements in early knee osteoarthritis. Acta Reumatol Port 2009;34: Stefanik JJ, Guermazi A, Zhu Y, Zumwalt AC, Gross KD, Clancy M, et al. Quadriceps weakness, patella alta, and structural features of patellofemoral osteoarthritis. Arthritis Care Res (Hoboken) 2011;63: Petterson SC, Barrance P, Buchanan T, Binder-Macleod S, Snyder-Mackler L. Mechanisms underlying quadriceps weakness in knee osteoarthritis. Med Sci Sports Exerc 2008;40: Sattler M, Dannhauer T, Hudelmaier M, Wirth W, Sanger AM, Kwoh CK, et al. Side differences of thigh muscle cross-sectional areas and maximal isometric muscle force in bilateral knees with the same radiographic disease stage, but unilateral frequent pain: data from the Osteoarthritis Initiative. Osteoarthritis Cartilage 2012;20: Berger MJ, Kean CO, Goela A, Doherty TJ. Disease severity and knee extensor force in knee osteoarthritis: data from the Osteoarthritis Initiative. Arthritis Care Res (Hoboken) 2012;64: Segal NA, Torner JC, Felson D, Niu J, Sharma L, Lewis CE, et al. Effect of thigh strength on incident radiographic and symptomatic knee osteoarthritis in a longitudinal cohort. Arthritis Rheum 2009;61: Hudelmaier M, Wirth W, Himmer M, Ring-Dimitriou S, Sanger A, Eckstein F. Effect of exercise intervention on thigh muscle volume and anatomical cross-sectional areas: quantitative assessment using MRI. Magn Reson Med 2010;64: Zhang W, Moskowitz RW, Nuki G, Abramson S, Altman RD, Arden N, et al. OARSI recommendations for the management of hip and knee osteoarthritis, part II: OARSI evidence-based, expert consensus guidelines. Osteoarthritis Cartilage 2008;16: Roos EM, Herzog W, Block JA, Bennell KL. Muscle weakness, afferent sensory dysfunction and exercise in knee osteoarthritis. Nat Rev Rheumatol 2011;7: Amin S, Baker K, Niu J, Clancy M, Goggins J, Guermazi A, et al. Quadriceps strength and the risk of cartilage loss and symptom progression in knee osteoarthritis. Arthritis Rheum 2009;60: Maurer BT, Stern AG, Kinossian B, Cook KD, Schumacher HR Jr. Osteoarthritis of the knee: isokinetic quadriceps exercise versus an educational intervention. Arch Phys Med Rehabil 1999;80: Felson DT, Zhang Y, Hannan MT, Naimark A, Weissman BN, Aliabadi P, et al. The incidence and natural history of knee osteoarthritis in the elderly: the Framingham Osteoarthritis Study. Arthritis Rheum 1995;38: Neogi T, Felson D, Niu J, Nevitt M, Lewis CE, Aliabadi P, et al. Association between radiographic features of knee osteoarthritis and pain: results from two cohort studies. BMJ 2009; 339:b Ikeda S, Tsumura H, Torisu T. Age-related quadriceps-dominant muscle atrophy and incident radiographic knee osteoarthritis. J Orthop Sci 2005;10: Palmieri-Smith RM, Thomas AC, Karvonen-Gutierrez C, Sowers MF. Isometric quadriceps strength in women with mild, moderate, and severe knee osteoarthritis. Am J Phys Med Rehabil 2010;89: Conroy MB, Kwoh CK, Krishnan E, Nevitt MC, Boudreau R, Carbone LD, et al, for the Health ABC Study. Muscle strength, mass, and quality in older men and women with knee osteoarthritis. Arthritis Care Res (Hoboken) 2012;64: Segal NA, Zimmerman MB, Brubaker M, Torner JC. Obesity and knee osteoarthritis are not associated with impaired quadriceps specific strength in adults. PM R 2011;3: Felson DT, Naimark A, Anderson J, Kazis L, Castelli W, Meenan RF. The prevalence of knee osteoarthritis in the elderly: the Framingham Osteoarthritis Study. Arthritis Rheum 1987;30: Spector TD, Hart DJ, Byrne J, Harris PA, Dacre JE, Doyle DV. Definition of osteoarthritis of the knee for epidemiological studies. Ann Rheum Dis 1993;52: Sanghi D, Avasthi S, Mishra A, Singh A, Agarwal S, Srivastava RN. Is radiology a determinant of pain, stiffness, and functional disability in knee osteoarthritis? A cross-sectional study. J Orthop Sci 2011;16: Segal NA, Glass NA. Is quadriceps muscle weakness a risk factor for incident or progressive knee osteoarthritis? Phys Sportsmed 2011;39: Harding GT, Hubley-Kozey CL, Dunbar MJ, Stanish WD, Astephen Wilson JL. Body mass index affects knee joint mechanics during gait differently with and without moderate knee osteoarthritis. Osteoarthritis Cartilage 2012;20: Childs JD, Sparto PJ, Fitzgerald GK, Bizzini M, Irrgang JJ. Alterations in lower extremity movement and muscle activation patterns in individuals with knee osteoarthritis. Clin Biomech (Bristol, Avon) 2004;19: Hayes KW, Falconer J. Differential muscle strength decline in osteoarthritis of the knee: a developing hypothesis. Arthritis Care Res 1992;5: Berger MJ, McKenzie CA, Chess DG, Goela A, Doherty TJ. Sex differences in quadriceps strength in OA. Int J Sports Med 2012;33: Fragala MS, Clark MH, Walsh SJ, Kleppinger A, Judge JO, Kuchel GA, et al. Gender differences in anthropometric predictors of physical performance in older adults. Gend Med 2012;9: Eckstein F, Wirth W, Nevitt MC. Recent advances in osteoarthritis imaging: the Osteoarthritis Initiative. Nat Rev Rheumatol 2012;8: Kellgren JH, Lawrence JS. Radiological assessment of osteoarthrosis. Ann Rheum Dis 1957;16: Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol 1988;15: Osteoarthritis Initiative. Isometric Strength (Isometric Chair) Operations Manual. 1.2p, URL: oai.epi-ucsf.org/datarelease/operationsmanuals.asp. 33. Rantanen T, Era P, Heikkinen E. Physical activity and the changes in maximal isometric strength in men and women from the age of 75 to 80 years. J Am Geriatr Soc 1997;45: Rantanen T, Era P, Heikkinen E. Maximal isometric strength and mobility among 75-year-old men and women. Age Ageing 1994;23: Bellamy N, Bell MJ, Pericak D, Goldsmith CH, Torrance GW, Raynauld JP, et al. BLISS index for analyzing knee osteoarthritis trials data. J Clin Epidemiol 2007;60: Riddle DL, Stratford PW. Impact of pain reported during isometric quadriceps muscle strength testing in people with knee pain: data from the osteoarthritis initiative. Phys Ther 2011;91: Wang Y, Wluka AE, Berry PA, Siew T, Teichtahl AJ, Urquhart DM, et al. Increase in vastus medialis cross-sectional area is associated with reduced pain, cartilage loss, and joint replacement risk in knee osteoarthritis. Arthritis Rheum 2012;64:

10 Symptoms and Disease Severity in KOA Thigh Muscle Strength O Reilly SC, Jones A, Muir KR, Doherty M. Quadriceps weakness in knee osteoarthritis: the effect on pain and disability. Ann Rheum Dis 1998;57: Ruhdorfer A, Dannhauer T, Wirth W, Hitzl W, Kwoh CK, Guermazi A, et al. Thigh muscle cross-sectional areas and strength in advanced versus early painful osteoarthritis: an exploratory between-knee, within-person comparison in osteoarthritis initiative participants. Arthritis Care Res (Hoboken) 2013;65: Pietrosimone BG, Hertel J, Ingersoll CD, Hart JM, Saliba SA. Voluntary quadriceps activation deficits in patients with tibiofemoral osteoarthritis: a meta-analysis. PM R 2011;3: Van Dijk GM, Veenhof C, Spreeuwenberg P, Coene N, Burger BJ, van SD, et al. Prognosis of limitations in activities in osteoarthritis of the hip or knee: a 3-year cohort study. Arch Phys Med Rehabil 2010;91: Ditroilo M, Forte R, Benelli P, Gambarara D, De Vito G. Effects of age and limb dominance on upper and lower limb muscle function in healthy males and females aged years. J Sports Sci 2010;28: Segal NA, Yack HJ, Brubaker M, Torner JC, Wallace R. Association of dynamic joint power with functional limitations in older adults with symptomatic knee osteoarthritis. Arch Phys Med Rehabil 2009;90: Madsen OR, Bliddal H, Egsmose C, Sylvest J. Isometric and isokinetic quadriceps strength in gonarthrosis: inter-relations between quadriceps strength, walking ability, radiology, subchondral bone density and pain. Clin Rheumatol 1995;14: McAlindon TE, Cooper C, Kirwan JR, Dieppe PA. Determinants of disability in osteoarthritis of the knee. Ann Rheum Dis 1993;52: Chun SW, Kim KE, Jang SN, Kim KI, Paik NJ, Kim KW, et al. Muscle strength is the main associated factor of physical performance in older adults with knee osteoarthritis regardless of radiographic severity. Arch Gerontol Geriatr 2013;56: Fink B, Egl M, Singer J, Fuerst M, Bubenheim M, Neuen-Jacob E. Morphologic changes in the vastus medialis muscle in patients with osteoarthritis of the knee. Arthritis Rheum 2007; 56: Mehrotra C, Naimi TS, Serdula M, Bolen J, Pearson K. Arthritis, body mass index, and professional advice to lose weight: implications for clinical medicine and public health. Am J Prev Med 2004;27: Slemenda C, Brandt KD, Heilman DK, Mazzuca S, Braunstein EM, Katz BP, et al. Quadriceps weakness and osteoarthritis of the knee. Ann Intern Med 1997;127: Slemenda C, Heilman DK, Brandt KD, Katz BP, Mazzuca SA, Braunstein EM, et al. Reduced quadriceps strength relative to body weight: a risk factor for knee osteoarthritis in women? Arthritis Rheum 1998;41: