Cross-sectional association between muscle strength and self-reported physical function in 195 hip osteoarthritis patients

Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]

Contents lists available at ScienceDirect

Seminars in Arthritis and Rheumatism journal homepage: www.elsevier.com/locate/semarthrit

Cross-sectional association between muscle strength and self-reported physical function in 195 hip osteoarthritis patients Michelle Hall, PhDa, Tim V. Wrigley, MSca, Jessica Kasza, PhDa,b, Fiona Dobson, PhDa, Yong Hao Pua, PhDc, Ben R. Metcalf, BSca, Kim L. Bennell, PhDa,n a Department of Physiotherapy, Centre for Health, Exercise and Sports Medicine, School of Health Sciences, Melbourne, The University of Melbourne, VIC, Australia b Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, VIC, Australia c Department of Physiotherapy, Singapore General Hospital, Singapore, Singapore

a r t i c l e in fo

Keywords: Hip osteoarthritis Muscle strength Physical function

a b s t r a c t Introduction: This study aimed to evaluate associations between strength of selected hip and knee muscles and self-reported physical function, and their clinical relevance, in men and women with hip osteoarthritis (OA). Methods: Cross-sectional data from 195 participants with symptomatic hip OA were used. Peak isometric torque of hip extensors, flexors, and abductors, and knee extensors were measured, along with physical function using the Western Ontario and McMaster Universities Osteoarthritis Index questionnaire. Separate linear regressions in men and women were used to determine the association between strength and physical function accounting for age, pain, and radiographic disease severity. Subsequently, magnitudes of strength associated with estimates of minimal clinically important improvement (MCII) in physical function were estimated according to severity of difficulty with physical function. Results: For men, greater strength of the hip extensors, hip flexors and knee extensors were each associated with better physical function. For women, greater muscle strength of all tested muscles were each associated with better physical function. For men and women, increases in muscle strength between 17–32%, 133–223%, and 151–284% may be associated with estimates of MCII in physical function for those with mild, moderate, and severe physical dysfunction, respectively. Conclusion: Greater isometric strength of specific hip and thigh muscle groups may be associated with better self-reported physical function in men and women. In people with mild physical dysfunction, an estimate of MCII in physical function may be associated with attainable increases in strength. However, in patients with more severe dysfunction, greater and perhaps unattainable strength increases may be associated with an estimate of MCII in physical function. Longitudinal studies are required to validate these observations. & 2016 Elsevier Inc. All rights reserved.

Introduction Hip osteoarthritis (OA) is a prevalent and costly chronic musculoskeletal condition [1,2]. After the knee, the hip is the most commonly affected lower-limb joint with 11% of older adults estimated to have hip OA [3]. Although pain is often considered the cardinal symptom of OA [4], physical function is also impaired in patients with hip OA [5] and is one of the drivers for costly joint replacement surgery [6]. Exercise, particularly strengthening, is among the core conservative treatments recommended by

n

Corresponding author. E-mail address: [email protected] (K.L. Bennell).

http://dx.doi.org/10.1016/j.semarthrit.2016.08.004 0049-0172/& 2016 Elsevier Inc. All rights reserved.

clinical guidelines [7–9] and can provide modest improvements in physical function for patients with hip OA [10]. However, a better understanding of the association between muscle strength and physical function in these patients may guide treatments, such that greater than modest clinical improvements can be achieved. There is evidence of lower-limb muscle weakness in people with hip OA [11]. Although muscle weakness has been associated with objective measures of physical impairments [12], little is known about the relationship between muscle strength and selfreported physical function in people with hip OA. This is important given that self-reported measures, such as the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) questionnaire [13], measure different constructs of physical functioning to performance-based measures [14] and are often used as

2

M. Hall et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]

end-points in OA clinical trials evaluating the efficacy of interventions [15–17]. Moreover, previous research has demonstrated that a minimal clinically important improvement (MCII) in physical function is dependent on the severity of difficulty with physical function in people with hip OA [13]. Thus, providing context for change in strength associated with a clinically relevant change in physical function according to severity of physical function can facilitate future longitudinal research. Therefore, the aim of this study was to test the hypothesis that greater strength of the major hip and thigh muscles would be associated with better self-reported physical function, assessed using the WOMAC, in men and women. If significant associations were found, we also aimed to explore the magnitude of strength gains associated with an estimate of MCII in physical function according to varying levels of initial physical dysfunction.

The most symptomatic hip was deemed the study hip in cases of bilaterally eligible cases. Ethical approval was obtained for the two studies from the University of Melbourne Human Ethics Committee and all participants provided their written informed consent.

Methods

Muscle strength was measured as maximal isometric muscle torque (Nm/kg) and normalized to body mass (kg) because a large proportion (75%) of the items on the WOMAC physical function subscale relate to weight-bearing activities [18]. The average of two maximal efforts was used to measure peak hip muscle torque (Nm); the peak of three trials was used for peak knee extensor torque. Differences in peak torque quantification (average versus peak) between hip and knee muscle strength assessments relate to differences in techniques (hand-held dynamometry versus isokinetic dynamometer). We have demonstrated excellent reliability for all strength measurement techniques used in this study [22,31].

Study design This was a cross-sectional study evaluating the relationship between muscle strength of selected hip and thigh muscle groups and self-reported physical function assessed using the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) [18]. The study conforms to the STROBE statement for reporting cross-sectional studies [19].

Dependent variable (outcome) Physical function was assessed using the WOMAC 17-item physical function subscale with hip-related questions on a scale from 0 (“none”) to 4 (“extreme”). The total score was normalized to a 0–100 score, where higher scores indicate extreme difficulty [18]. Physical function on the WOMAC is recommended for OA clinical trials [21]. Independent variables (predictors)

Participants Data from two studies [12,15] including 195 participants with symptomatic and radiographic hip OA [n ¼ 93 from a crosssectional study and n ¼ 102 from a randomised controlled trial (RCT) study utilizing baseline data] were used. Participants were recruited via community advertisements in Melbourne, Australia from November 2006 to May 2008 for the cross-sectional study and from May 2010 to April 2012 for the RCT. The two studies shared two inclusion criterion: (i) hip OA fulfilling America College of Rheumatology classification criteria of pain and radiographic changes [20] assessed on standing x-ray used standardised protocols and femoral or acetabular osteophytes and joint space narrowing greater than or equal to Grade 2 on a standing x-ray; and (ii) hip or groin pain on most days of the past month. The studies shared these exclusion criteria as follows: (i) presence of neurologic, cardiac, or other medical conditions that would compromise lower-limb function; (ii) back pain or other joint pain; (iii) lower extremity joint replacement; (iv) systemic arthritic conditions such as rheumatoid arthritis; (v) other previous pathology such as fracture; (vi) inability to walk unaided; and (vii) inadequate ability to understand English. The cross-sectional study had additional exclusion criterion including secondary hip OA due to trauma, inflammatory, or metabolic rheumatic diseases. The RCT had these additional inclusion criteria: (i) 50 years or older; (ii) average pain intensity in the past week of 40 or higher on a 100 mm visual analogue scale (VAS); (iii) at least moderate difficulty with daily activities. The RCT also had the following exclusion criteria: (i) hip joint surgery within past 6 months, (ii) planned lower-limb surgery; (iii) physiotherapy, chiropractic treatment, or prescribed exercises for hip, lumbar spine, or both in the past 6 months; (iv) walking continuously for more than 30 min daily and regular structured exercise more than once weekly; (v) uncontrolled hypertension, or morbid obesity (body mass index 4 40 kg/m2); (vi) unable to comply with study protocol; (vii) current or past (within 3 months) oral or intra-articular corticosteroid use; and (viii) inability to comply with study protocol.

Peak isometric hip muscle strength Peak isometric muscle strength was assessed on the study hip only. Before each hip muscle strength test, participants performed a submaximal and a maximal contraction warm-up effort for familiarization. Participants performed two maximal trials for approximately 3 s. Each participant received standardised, strong verbal encouragement to “push/pull” as hard as you can, respectively. Techniques used to assess muscle strength in this study have been previously described in detail along with excellent test– retest reliability in patients with hip OA (intraclass correlation coefficient 0.84–0.97; standard error of measurement 3–12 Nm) [22]. Therefore, we provide a brief description below. For hip extensor strength, participants were secured in a supine position with the study hip in 201 of hip flexion, aligned perpendicularly to a force transducer (FGV—10 force gauge; Shimpo Instruments, New York, USA) and digital inclinometer suspended from the ceiling. The lever arm was recorded as the distance from the most prominent aspect of the greater trochanter to the point of transducer attachment 5 cm proximal to the lateral malleolus. The transducer force was zeroed with the limb relaxed, to remove the gravitational weight of the limb from the recorded force. Hip abduction strength was measured with the Lafayette hand-held dynamometer (Lafayette Instrument Co, Lafayette, IN) in a stabilized, gravity-eliminated supine position, with both hips in neutral abduction–adduction. The lever arm was recorded as the distance from the most prominent aspect of the greater trochanter to the point of transducer attachment within 5 cm proximal to the femoral condyle. For hip flexion strength, participants were securely seated in an upright position with the hip and knee in 901 of flexion. Strength was assessed using a hand-held dynamometer placed 5 cm proximal to the superior aspect of the patella on the anterior thigh. The lever arm was recorded as the distance from the most prominent aspect of the greater trochanter to the point of transducer attachment 5 cm proximal to the femoral condyle. Correction for the gravitational weight of the flexed limb was deemed not possible for hip flexion strength in this position.

M. Hall et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]

Peak isometric knee extensor strength Isometric extensor torque at 601 of knee flexion was assessed on a KinCom 125-AP isokinetic dynamometer (Chattecx, Chattanooga, TN, USA). Participants were securely seated, with their trunk upright and hips flexed to approximately 901. Participants then performed a submaximal warm-up effort for familiarization. Participants performed three maximal trials for 3–5 s while receiving strong verbal encouragement to “push as hard as you can”. The lever arm length was recorded as the distance from the ankle cuff to the rotation axis of the dynamometer. Using the maximal force in Newtons (corrected for limb gravitational torque) multiplied by the lever arm length (m), peak isometric knee extensor torque (Nm) was obtained. We have demonstrated excellent knee extensor strength test–retest reliability in patients with hip OA (intraclass correlation coefficient 0.93; standard error of measurement 8 Nm) [31]. Covariates To optimize the pragmatic application of our findings age, pain, and radiographic disease severity were included as covariates given that each satisfied the criteria by Shrier and Platt [23]. Increasing age has been associated with reduced muscle strength [24], and difficulty with physical function in people with knee OA [25]. Pain was also considered a covariate given its association with worse difficulty with physical dysfunction [26]. Pain was assessed using the WOMAC pain-subscale [18], where participants reported the extent of pain while performing activities of daily living due to hip OA on a scale from 0 (“none”) to 4 (“extreme”). Five pain-related items were scored on a range from 0 to 20 and normalized to 100, where a higher score indicates greater pain severity. Lastly, radiographic disease severity was included as a covariate given its association with WOMAC physical function in hip OA patients [27]. In the current study, participants were graded as either “KL2” defined as definite narrowing of joint space inferiorly, definite osteophytes and slight sclerosis, “KL3” defined as marked narrowing of joint space, slight osteophytes, some sclerosis and cyst formation, and deformity of the femoral head and acetabulum or “KL4” defined as marked narrowing of joint space with sclerosis and cysts, marked deformity of the femoral head and acetabulum, and large osteophytes [28]. Statistical analysis To account for sex differences in lower-limb muscle strength [29], separate linear regression models for men and women were used to evaluate relationships between measures of peak isometric muscle strength (predictor variables) and WOMAC physical function (dependent variable). Analyses were performed both unadjusted and adjusted for age, WOMAC pain, and radiographic severity. To investigate if the relationship between the independent variables and WOMAC physical function differed across the two studies from which participants were drawn, a term for “study”, and an interaction between the study term and the strength variable were included. Interaction terms were not significantly different from zero and therefore these terms were excluded (see Supplementary material) Analysis were also performed stratified by study and overall results were consistent between the two studies. Residuals of each model were inspected and conformed to appropriate assumptions [30]. Stata version 13.1 Statacorp., College Station, TX, USA was used for all statistical analyses and significance was set at α o0.05. Data interpretation

meaningful change in physical function based on cross-sectional data. Specifically, estimated MCII values for WOMAC physical function previously reported according to tertiles of increasing difficulty with physical function in people with hip OA [13] ( 2.6,  14.8, and  15.1), were divided by coefficients in the final adjusted model (Table 2) to determine an increase in strength perhaps associated with an estimate of MCII in physical function for each sex and muscle group. Using cut-points associated with the estimated MCII [13] according to varying levels of difficulty with physical function, the average strength for each muscle group was calculated for participants with mild (WOMAC physical scores r 38.2), moderate (WOMAC physical scores between 38.3 and 52.9) and severe (WOMAC physical scores Z 52.9) physical dysfunction for each sex. The percentage difference in strength measure associated with MCII relative to the average strength value for each muscle group according to severity of difficulty with physical function was calculated. As an example of calculations presented in Table 3, we consider males with mild physical dysfunction and the association between hip extension strength and an estimate of MCII in physical function. Previous research estimates that a reduction of  2.6 units in WOMAC physical function is the equivalent to a MCII for hip OA patients with mild physical dysfunction (classified as having a WOMAC score r 38.2) [13]. Using the regression coefficient describing the association between strength and physical function from our regression model (Table 2), we divided  2.6 by  3.27 to determine the amount of strength associated with an estimated MCII in physical function for male with mild physical dysfunction (0.80 Nm/kg). Subsequently, to put this value (0.80 Nm/kg) in context, we expressed it as a percentage (32%) of the average hip extensor strength for males as described in Table 1 (0.80 Nm/kg divided by 2.48 Nm/kg).

Results Descriptive characteristics are presented in Table 1 for the cohort and according to sex. In general, the cohort was middleaged, overweight, and represented a female majority (60%).

Table 1 Participant characteristics Total (n ¼ 195)

Men (n ¼ 77)

Women (n ¼ 118)

63.0 7 8.9 1.66 7 0.1 79.1 7 15.2 28.6 7 4.8 64 (33%) 124:71 175:20

61.6 7 9.9 1.74 7 0.1 87.6 7 15.1 28.9 7 4.6 21 (27%) 51:26 73:4

64.0 7 8.1 1.61 7 0.1 73.6 7 12.6 28.3 7 4.9 43 (36%) 73:45 102:16

Radiographic severity KL grade 2, n (%) 110 (56%) KL grade 3, n (%) 49 (25%) KL grade 4, n (%) 36 (18%) WOMAC pain (0–100) 37.6 7 15.7 WOMAC physical function (0–100) 39.4 7 16.3

35 27 15 34.7 37.9

(46%) (35%) (19%) 7 13.4 7 14.5

75 22 21 39.4 40.4

(64%) (19%) (18%) 7 16.8 7 17.4

Isometric muscle torque (Nm/kg) Hip extension torque Hip flexion torque Hip abduction torque Knee extensor torque

2.23 1.42 1.34 1.89

7 7 7 7

1.49 1.11 0.99 1.33

7 7 7 7

Age, y Height, m Body mass, kg Body mass index, kg/m2 Bilateral symptoms, n (%) Affected hip, right:left, n Dominant leg, right:left, n

a

Similar to previous cross-sectional research [32], we provide context for a change in strength associated with a clinically

3

b c

1.78 1.23 1.13 1.54

Data available for 190 participants. Data available for 73 participants. Data available for 117 participants.

7 7 7 7

0.70 0.47 0.42 0.60a

0.68 0.45 0.42 0.60b

0.54 0.44 0.35 0.50c

M. Hall et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]

4

Table 2 Linear relationships between measure of peak isometric torque (independent variable) and WOMAC physical function 100 scale (dependent variable) Simple Regression

Slope

Multiple regressiona

Slope

Regression coefficient (95% CI)

p Value

Regression coefficient (95% CI)

p Value

Hip extension torque (Nm/kg) Men (n ¼ 77) Women (n ¼ 118)

 10.75 (  14.96 to  6.54)  12.11 (  17.61 to  6.60)

o0.001 o0.001

 3.27 (  6.51 to  0.04)  5.77 (  9.41 to  2.13)

Hip flexion torque (Nm/kg) Men (n ¼ 77) Women (n ¼ 118)

 17.16 (  23.36 to  10.97)  22.23 (  28.24 to  16.22)

o0.001 o0.001

 5.37 (  10.07 to  0.66)  10.88 (  15.20 to  6.55)

0.026 o0.001

Hip abduction torque (Nm/kg) Men (n ¼ 76) Women (n ¼ 118)

 11.85 (  19.01 to  4.69)  26.06 (  33.80 to  18.33)

o0.001 o0.001

 1.52 (  6.90 to 3.86)  10.94 (  16.64 to  5.23)

0.575 o0.001

Knee extension torque (Nm/kg) Men (n ¼ 73) Women (n ¼ 117)

 12.56 (  17.49 to  7.63)  20.03 (  25.33 to  14.73)

o0.001 o0.001

 4.02 (  7.88 to  0.16)  9.62 (  14.01 to  5.22)

0.041 o0.001

a

0.048 0.002

Adjusted for age, hip pain, radiographic disease severity.

Men

Women

In the unadjusted analyses, significant associations were observed between greater peak isometric strength of all tested muscles and better physical function, with 13–29% of the variance explained (Table 2, Fig. 1). In the adjusted analyses, significant associations were observed for the hip extensors, hip flexors, and knee extensors but not for the hip abductors (Table 2). For men with mild difficulty with physical function, moderate increases (31–32%) in muscle strength of the hip extensors, hip flexors, and knee extensors were associated with an estimate of MCII in physical function (Table 3). However, for men with more severe difficulty with physical function, considerably greater increases (197–284%) in strength of these muscles were associated with an estimate of MCII in physical function (Table 3).

In both the unadjusted and adjusted analyses, there were significant associations between greater peak isometric strength of all tested muscles and better physical function (Table 2; Fig. 2). In the unadjusted analyses, muscle strength explained 14–33% of the variation in physical function (Fig. 2). For women with mild difficulty with physical function, moderate increases (17–27%) in muscle strength of the hip flexors, hip extensors, hip abductors, and knee extensors were associated with an estimate of MCII in physical function (Table 3). However, for women with more severe difficulty with physical function, considerably greater increases (133–200%) in strength of these muscles were associated with an estimate of MCII in physical function (Table 3).

Fig. 1. Unadjusted association between peak isometric measures of strength (Nm/kg) and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) physical function score (range: 0–100, higher scores represent greater difficulty with physical function) for men.

M. Hall et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]

5

Table 3 Measures of peak isometric torque and associations with MCII for physical function stratified across increasing levels of difficulty with physical function Men

Peak hip extension torque (Nm/kg) Increase in hip extension (Nm/kg) associated with MCII % Increase in hip extension strength related to MCII Peak hip flexion torque (Nm/kg) Increased hip flexion (Nm/kg) associated with MCII % Increase in hip flexion strength related to MCII Peak hip abduction torque (Nm/kg) Increased hip abduction (Nm/kg) associated with MCII % Increase in hip abduction strength related to MCII Peak knee extension torque (Nm/kg) Increased knee extension (Nm/kg) associated with MCII % Increase in knee extension strength related to MCII

Women

Mild physical dysfunctiona (n ¼ 40)

Moderate physical dysfunctionb (n ¼ 23)

Severe physical dysfunctionc (n ¼ 14)

Mild physical dysfunctiona (n ¼ 55)

Moderate physical dysfunctionb (n ¼ 27)

Severe physical dysfunctionc (n ¼ 36)

2.48 7 0.72 0.80d

2.03 7 0.53 4.53

1.82 7 0.53 4.62

1.65 7 0.53 0.45

1.40 7 0.41 2.56

1.31 7 0.58 2.62

32%e

223%

254%

27%

183%

200%

1.58 7 0.46

1.40 7 0.38

0.99 7 0.26

1.32 7 0.40

1.02 7 0.38

0.84 7 0.37

0.48

2.76

2.81

10.24

1.36

1.39

31%

197%

284%

18%

133%

165%

1.30 7 0.33 NA

1.14 7 0.43 NA

1.14 7 0.34 0.24

0.93 7 0.35 1.35

0.79 7 0.25 1.38

21%

145%

175%

1.16 7 0.29 1.54

1.04 7 0.43 1.57

133%

151%

1.44 7 0.45f NA NA 2.12 7 0.62g 0.65 31%

NA

NA

1.78 7 0.40h 3.68

1.32 7 0.40i 3.76

1.61 7 0.48 0.27

207%

285%

17%

j

MCII, minimal clinically important improvement; NA, not assessed given the association between strength and physical function did not reach statistical significance (Table 1). a

Mild difficulty with physical function: defined as WOMAC physical scores between r 38.2. Moderate difficulty with physical function: defined as WOMAC physical scores between 38.3 and 52.9. c Severe difficulty with physical function: defined as WOMAC physical scores between Z52.9. d Sample calculation: MCII for mild difficulty with physical function coefficient from Table 2 males hip extension final adjusted model (e.g.,  2.6/  3.27 ¼ 0.80 Nm/kg). e Expressed as % ¼ (0.80/2.48)  100. f n ¼ 39. g n ¼ 39. h n ¼ 22. i n ¼ 12. j n ¼ 54. b

Fig. 2. Unadjusted association between peak isometric measures of strength (Nm/kg) and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) physical function score (range: 0–100, higher scores represent greater difficulty with physical function) for women.

6

M. Hall et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]

Discussion We found that in people with hip OA, peak isometric strength of major hip and knee muscle groups is associated with selfreported physical function. In women, greater strength of each muscle assessed was significantly associated with better physical function and in men significant associations were seen for all muscles, except the hip abductors. The amount of strength gain associated with a clinically relevant improvement in physical function depended upon the extent of difficulty with physical function. These data provide impetus for future longitudinal research to determine the validity of the cross-sectional associations between strength and physical function improvement observed in this study. Previous research has reported associations between hip and knee muscle strength and physical function assessed objectively in hip OA patients [31]. To our knowledge, this study is the first to evaluate relationships between muscle strength and self-reported physical function in hip OA. This is particularly relevant given the use of self-reported physical function as an endpoint in OA clinical trials [15–17]. Our results concur with observations in patients with knee OA, where greater knee extensor strength was significantly associated with better WOMAC physical function [32]. In the current study, stronger hip extensor, hip flexor, and knee extensor muscles were significantly associated with better physical function in both men and women with hip OA. These muscles are key to performing the majority of tasks assessed in the physical function component of the WOMAC questionnaire, such as walking, stair ambulation, and rising from a chair [33]. Stronger hip abductor muscles were also associated with better self-reported physical function in women. These muscles are also integral to weightbearing ambulation [33] and thus physical function. However, explanations for a lack of association between strength of hip abductors and physical function in men are not entirely clear. It is possible that demands on the hip abductor muscles are greater in women compared to men, given that women typically have a wider pelvis and would therefore require a greater internal hip abduction torque to ambulate. It is also possible that men have adequate strength in these muscles to perform tasks assessed in the WOMAC. We considered that men with strong hip abductors might have reached a ceiling in physical function, thereby attenuating a linear association between hip abductor strength and physical function. However, exploration of non-linear associations did not support our theory (data not shown). Our study extends existing knowledge by demonstrating that magnitudes of strength associated with estimates of clinically relevant improvements in physical function vary depending on initial levels of difficulty with physical function. Based on regression equations derived from cross-sectional data, increases between 17% and 32% in muscle strength may be associated with an estimate of MCII in physical function for those with mild difficulty with physical function. Such strength increases are likely attainable. For example, an uncontrolled study evaluating a 10-session resistance program improved hip and knee muscle strength by on average 30% (range: 26–33%) in people with hip OA [34]. Thus, patients with mild physical function difficulty may be a sub-group to target with muscle-strengthening interventions when aiming to improve self-reported physical function. However, our cross-sectional data suggest that for participants with moderate to severe difficulty with physical function (133–284%) larger and likely unattainable increases in muscle strength may be associated with an estimate of MCII in physical function. Overall, the magnitudes of strength associated with estimates of MCII together with observation that strength only explained 13–33% of variation in physical function, prompt consideration of other possible factors associated with physical function in hip OA

patients. Number of co-morbidities, back pain, fatigue, and hip flexion range of motion have previously been associated with WOMAC physical function in hip OA patients [27,35] and may in part further explain physical function. Physical activity may also be related to self-reported measures of physical function, as previously reported in patients with knee OA [37]. A better understanding of physical function determinants, according to severity of difficulty with physical function, could be used to further tailor clinical management of hip OA patients as appropriate. Although based on cross-sectional data, this study makes an important contribution to knowledge. Specifically, our findings suggest that severity of difficulty with physical function could mediate the effect of muscle-strengthening interventions on selfreported physical function in hip OA patients. This speculation may in part explain the overall modest effect of exercise interventions on physical function (standardised mean difference  0.38) in hip OA patients as reported in a systematic review of RCTs [10]. Future clinical trials are required to confirm if hip OA patients with mild difficulty with physical function have a better improvement in physical function compared to patients with greater difficulty with physical function following interventions that increase muscle strength. Muscle strength can be addressed in several ways. Resistance exercise that induces muscle hypertrophy will increase muscle strength. Weight-loss that reduces body mass will also increase muscle strength, given that by definition, normalized strength is in part dependent on body mass (Nm/kg). Pain and effusion may also influence maximal muscle strength [36] and thus if reduced will likely result in increased maximal muscle strength. Further to this, the extent of muscle weakness may vary among patients and thus scope to improve muscle strength should be considered in clinical management. Prospective research is required to determine the extent to which strength improvement, however achieved, is associated with an estimate of MCII in physical function in people with hip OA. Our findings should be interpreted with caution. Estimates from the regression equation are statistically independent of age, pain, and radiographic disease severity to ease translation of findings. Future research aiming to further evaluate the contribution of pain and radiographic disease severity to the association between strength and physical function would require groups with radiographic OA of the hip and both with and without hip pain. Our data are directional and represent an increase in strength associated with a MCII in physical function and not necessarily the minimal clinical important difference (MCID), which can reflect an improvement or worsening. As such, our data cannot be extrapolated in the opposite direction where a decrease in strength is associated with a reduction in physical function by a similar magnitude [13]. We elected to report a MCII rather than a MCID because we were interested in understanding how much change in strength is required to improve physical function in a meaningful way. Furthermore, it has been shown that the MCID is not necessarily consistent for improvement and worsening [38]. Also, the MCII estimates previously reported [13] are estimated from trials of non-steroidal anti-inflammatory drugs and thus may not necessarily reflect a MCII associated with strengthening interventions. Third, our results can only be generalized to the physical function subscale of the WOMAC and not to other self-reported measures of function nor to objective measures of physical performance. Similarly, caution should be used in extrapolating the results beyond our measurement of muscle strength. It is not known if magnitudes of normalized strength associated with an estimate of MCII MCII would be similar with other measures of muscle strength (e.g., isokinetic). Furthermore, maximal isometric muscle strength may not be the most valid or sensitive assessment of strength when investigating relationships between muscle

M. Hall et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]

strength and self-reported physical function. Individuals rarely maximally contract their lower-limb muscles during activities of daily living but use an unknown individual-specific proportion of their maximal strength. It is therefore possible that other submaximal muscle measures such as muscle endurance may be more closely related to physical function. Nevertheless, we assessed maximal muscle strength given its common use in clinical practice and our previous work demonstrating the reliability of the strength measures in people with hip OA [22]. Strengths of our study include a comprehensive assessment of hip and knee muscle strength using methods previously reported to be reliable in patients with hip OA [22]. Additionally, we applied relevant cut-points to determine the magnitude of strength associated with improved physical function to facilitate clinical interpretation. There are also limitations of this study. As a crosssectional study, cause-and-effect associations between muscle strength and physical function cannot be determined. We performed multiple statistical tests that increase the risk of Type 1 error (finding a statistical difference by chance). A group MCII rather than anchor-based MCII as used in the current study, would be more appropriate to determine magnitude of strength associated with an estimate of MCII in physical function improvement. However, in the absence of a distribution-based MCII in physical function available in the literature (to our knowledge), we applied an anchor-based MCII previously described according to physical dysfunction severity [13].

Conclusion Using cross-sectional data we found evidence to suggest that greater isometric strength of specific muscle groups was associated with better self-reported physical function in people with hip OA. Further to this, relatively modest increases in muscle strength of some muscle groups are likely to be associated with clinically relevant improvements in physical function in hip OA patients with mild difficulty with physical function. However, substantial increases in lower-limb strength may be associated with physical function improvements in patients with moderate and severe difficulty with physical function. Future longitudinal research is needed to determine the validity of the cross-sectional associations between strength and physical function observed in this study.

Funding The study was funded from a National Health and Medical Research Council (NHMRC), Australia project grant (628556). K.L.B. is supported by a NHMRC Principal Research Fellowship (1058440). M.H. is supported by a NHMRC program grant (1091302).

Appendix A. Supplementary material Supplementary data are available in the online version of this article at http://dx.doi.org/10.1016/j.semarthrit.2016.08.004

References [1] Lawrence RC, Felson DT, Helmick CG, Arnold LM, Choi H, Deyo RA, et al. Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part II. Arthritis Rheum 2008;58:26–35. [2] Kim C, Nevitt MC, Niu J, Clancy MM, Lane NE, Link TM, et al. Association of hip pain with radiographic evidence of hip osteoarthritis: diagnostic test study. Br Med J 2015:351–8.

7

[3] Pereira D, Peleteiro B, Araujo J, Branco J, Santos RA, Ramos E. The effect of osteoarthritis definition on prevalence and incidence estimates: a systematic review. Osteoarthritis Cartilage 2011;19:1270–85. [4] Hunter DJ, McDougall JJ, Keefe FJ. The symptoms of osteoarthritis and the genesis of pain. Med Clin North Am 2009;93:83–100 [xi]. [5] Bellamy N, Kirwan J, Boers M, Brooks P, Strand V, Tugwell P, et al. Recommendations for a core set of outcome measures for future phase III clinical trials in knee, hip, and hand osteoarthritis. Consensus development at OMERACT III. J Rheumatol 1997;24:799–802. [6] Dreinhofer KE, Dieppe P, Sturmer T, Grober-Gratz D, Floren M, Gunther KP, et al. Indications for total hip replacement: comparison of assessments of orthopaedic surgeons and referring physicians. Ann Rheum Dis 2006;65: 1346–50. [7] Zhang W, Nuki G, Moskowitz RW, Abramson S, Altman RD, Arden NK, et al. OARSI recommendations for the management of hip and knee osteoarthritis: part III: changes in evidence following systematic cumulative update of research published through January 2009. Osteoarthritis Cartilage 2010;18: 476–99. [8] Conaghan PG, Dickson J, Grant RL. Care and management of osteoarthritis in adults: summary of NICE guidance. Br Med J 2008;336:502–3. [9] Brosseau L, Wells GA, Pugh AG, Smith CA, Rahman P, Alvarez Gallardo IC, et al. Ottawa panel evidence-based clinical practice guidelines for therapeutic exercise in the management of hip osteoarthritis. Clin Rehabil 2015. [10] Fransen M, McConnell S, Hernandez-Molina G, Reichenbach S. Exercise for osteoarthritis of the hip. Cochrane Database Syst Rev 2014;4:1–54. [11] Loureiro A, Mills PM, Barrett RS. Muscle weakness in hip osteoarthritis: a systematic review. Arthritis Care Res 2013;65:340–52. [12] Pua Y-H, Cowan SM, Wrigley TV, Bennell KL. The lower extremity functional scale could be an alternative to the Western Ontario and McMaster Universities Osteoarthritis Index physical function scale. J Clin Epidemiol 2009;62:1103–11. [13] Tubach F, Ravaud P, Baron G, Falissard B, Logeart I, Bellamy N, et al. Evaluation of clinically relevant changes in patient reported outcomes in knee and hip osteoarthritis: the minimal clinically important improvement. Ann Rheum Dis 2005;64:29–33. [14] Stratford PW, Kennedy DM. Performance measures were necessary to obtain a complete picture of osteoarthritic patients. J Clin Epidemiol 2006;59:160–7. [15] Bennell KL, Egerton T, Martin J, Abbott JH, Metcalf B, McManus F, et al. Effect of physical therapy on pain and function in patients with hip osteoarthritis: a randomized clinical trial. J Am Med Assoc 2014;311:1987–97. [16] Chang RW, Semanik PA, Lee J, Feinglass J, Ehrlich-Jones L, Dunlop DD. Improving physical activity in arthritis clinical trial (IMPAACT): study design, rationale, recruitment, and baseline data. Contemp Clin Trials 2014;39: 224–35. [17] French HP, Cusack T, Brennan A, Caffrey A, Conroy R, Cuddy V, et al. Exercise and manual physiotherapy arthritis research trial (EMPART) for osteoarthritis of the hip: a multicenter randomized controlled trial. Arch Phys Med Rehabil 2013;94:302–14. [18] McConnell S, Kolopack P, Davis AM. The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC): a review of its utility and measurement properties. Arthritis Rheum 2001;45:453–61. [19] von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007;370:1453–7. [20] Altman R, Alarcon G, Appelrouth D, Bloch D, Borenstein D, Brandt K, et al. The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hip. Arthritis Rheum 1991;34:505–14. [21] Bellamy N. Osteoarthritis clinical trials: candidate variables and clinimetric properties. J Rheumatol 1997;24:768–78. [22] Pua YH, Wrigley TV, Cowan SM, Bennell KL. Intrarater test–retest reliability of hip range of motion and hip muscle strength measurements in persons with hip osteoarthritis. Arch Phys Med Rehabil 2008;89:1146–54. [23] Shrier I, Platt RW. Reducing bias through directed acyclic graphs. BMC Med Res Methodol 2008;8:70. [24] Doherty TJ. Invited review: aging and sarcopenia. J Appl Physiol 2003;95: 1717–27. [25] Elbaz A, Debbi EM, Segal G, Haim A, Halperin N, Agar G, et al. Sex and body mass index correlate with Western Ontario and McMaster Universities Osteoarthritis Index and quality of life scores in knee osteoarthritis. Arch Phys Med Rehabil 2011;92:1618–23. [26] White DK, Zhang Y, Niu J, Keysor JJ, Nevitt MC, Lewis CE, et al. Do worsening knee radiographs mean greater chance of severe functional limitation? The Multicenter Osteoarthritis Study. Arthritis Care Res 2010;62:1433–9. [27] Wolfe F. Determinants of WOMAC function, pain and stiffness scores: evidence for the role of low back pain, symptom counts, fatigue and depression in osteoarthritis, rheumatoid arthritis and fibromyalgia. Rheumatology 1999;38:355–61. [28] Kellgren JH. Atlas of standard radiographs of arthritis.The Epidemiology of Chronic Rheumatism. Oxford: Blackwell Scientific; 1963. [29] Frontera WR, Hughes VA, Lutz KJ, Evans WJ. A cross-sectional study of muscle strength and mass in 45- to 78-year-old men and women. J Appl Physiol 1991;71:644–50. [30] Kirkwood B, Sterne J. Essential Medical Statistics. 2nd Ed Malden (MA): Blackwell; 2003.

8

M. Hall et al. / Seminars in Arthritis and Rheumatism ] (2016) ]]]–]]]

[31] Pua YH, Wrigley TV, Collins M, Cowan SM, Bennell KL. Association of physical performance with muscle strength and hip range of motion in hip osteoarthritis. Arthritis Rheum 2009;61:442–50. [32] Ruhdorfer A, Wirth W, Eckstein F. Relationship between isometric thigh muscle strength and minimum clinically important differences in knee function in osteoarthritis: data from the osteoarthritis initiative. Arthritis Care Res 2015;67:509–18. [33] Neumann DA. Kinesiology of the hip: a focus on muscular actions. J Orthop Sports Phys Ther 2010;40:82–94. [34] Jigami H, Sato D, Tsubaki A, Tokunaga Y, Ishikawa T, Dohmae Y, et al. Effects of weekly and fortnightly therapeutic exercise on physical function and healthrelated quality of life in individuals with hip osteoarthritis. J Orthop Sci 2012;17:737–44.

[35] Juhakoski R, Tenhonen S, Anttonen T, Kauppinen T, Arokoski JP. Factors affecting self-reported pain and physical function in patients with hip osteoarthritis. Arch Phys Med Rehabil 2008;89:1066–73. [36] Bennell K, Hinman RS, Wrigley TV, Creaby MW, Hodges P. Exercise and osteoarthritis: cause and effects. Compr Physiol 2011;1:1943–2008. [37] Dunlop DD, Song J, Semanik PA, Sharma L, Chang RW. Physical activity levels and functional performance in the osteoarthritis initiative: a graded relationship. Arthritis Rheum 2011;63:127–36. [38] Schwartz AL, Meek PM, Nail LM, Fargo J, Lundquist M, Donofrio M, et al. Measurement of fatigue: determining minimally important clinical differences. J Clin Epidemiol 2002;55:239–44.