Secondary Logo

Journal Logo

Research Reports

Stiffness, Pain, and Hip Muscle Strength Are Factors Associated With Self-reported Physical Disability in Hip Osteoarthritis

Steinhilber, Benjamin PhD1; Haupt, Georg PT1; Miller, Regina; Grau, Stefan PhD2; Janssen, Pia MD1; Krauss, Inga PhD1

Author Information
Journal of Geriatric Physical Therapy: July/September 2014 - Volume 37 - Issue 3 - p 99-105
doi: 10.1519/JPT.0b013e3182abe7b5
  • Free



Osteoarthritis (OA) of the hip is a common and chronic disease that leads to a loss of articular cartilage.1 The prevalence of hip OA increases with age,2 and the major clinical symptoms are joint pain,3 stiffness of the affected joint,3 impaired hip muscle strength,4,5 and range-of-motion limitations of the affected hip joint.4 Hip OA is one of the main causes of physical disability (PD) among older adults,3,6 resulting in limitations in activities of daily living7 and reduced mobility.4,8 In addition, OA is associated with an extremely high economic burden due to the need for therapy, work-related losses, and home care costs.2,9 Therefore, it is necessary to determine the most effective treatments and cost-efficient therapies.10

Exercise therapy is a common and well-established treatment for hip OA.11 However, insufficient evidence suggests that exercise decreases PD.12,13 Detailed knowledge about the determinants of PD in hip OA could provide essential information for developing an effective exercise program to decrease PD in patients with hip OA3,14–18 and therefore alleviate the symptoms of this disease.

A number of factors are responsible for PD in patients with hip OA. These include age, body mass index (BMI), passive range of motion (PROM) of the hip joint,18,19 pain,19 hip muscle strength,8 the number of comorbidities,3,19 additional knee OA,20 prior level of physical fitness or conditioning,20 participation in a supervised exercise program,20 educational level,3 and patient self-efficacy.3,18,19,21 However, it remains unclear which of these factors are most important and whether the series of factors mentioned is complete or must be amended by additional yet unknown factors.

The aim of the present study was to identify factors that are related to the level of self-reported PD quantified according to the physical function subscale of the Western Ontario and McMaster (WOMAC) index in patients with hip OA. A better understanding of these factors will enable researchers and therapists to modify them and to investigate whether these modifications might help reduce PD or even prevent increased PD in patients with hip OA.


Data Collection

Data were collected during 3 studies in patients with hip OA. The first study evaluated isokinetic and isometric peak torque measures in subjects with hip OA.22 The second study evaluated a home-based strengthening exercise program in patients with hip OA.23 Finally, the third study was a randomized clinical trial in which an exercise program specially designed to increase hip muscle strength and postural control in patients with hip OA was evaluated.24 These 3 studies were conducted by the same research group,22–24 using identical measurement protocols and raters to evaluate isometric hip muscle strength, PROM of the hip joint, PD, pain, and stiffness, as well as anthropometric data. Data from the initial measurement day of each study for 160 patients (90 men and 70 women) were included in the present study. All of the study participants provided informed consent, and the studies were approved by the local ethics committee. The inclusion criteria for the studies were as follows: hip OA in at least one hip joint, a stable implantation of a hip endoprosthesis if there was a history of hip replacement, and the ability to walk safely without walking aids. Hip OA was assessed according to the clinical criteria of the American College of Rheumatology, which has a sensitivity of 86% and a specificity of 75%.25 The exclusion criteria were as follows: any operation involving the lower extremities during the last 3 months before the beginning of each study, any other pathology that would result in medical therapy, neurological disease leading to sensomotoric deficits, known endocrinological cause of hip OA, verified metabolic cause of hip OA, any acute disease, and drug or alcohol abuse.

Self-reported PD, Pain, and Stiffness

The level of self-reported PD was quantified using the physical function subscale of the WOMAC index. The WOMAC index is a reliable, valid, and responsive measure of self-reported disability in patients with hip OA.26,27 The WOMAC index includes 24 questions that can be subdivided into 3 categories: pain, stiffness, and difficulty with physical function. Response options for each question were 0 to 10 on numerical rating scales, with 0 representing no pain, stiffness, or difficulty with physical functions and 10 representing extreme pain, stiffness, or difficulty with physical functions. The aggregated scale ranges for the WOMAC pain, stiffness, and function subscales were 0 to 50, 0 to 20, and 0 to 170, respectively. These scores were normalized, and subsequently each subscale ranges from 0 to 10. Less than 1% of the values were missing. Pain and stiffness were quantified using the pain and stiffness subscales of the WOMAC index.

Hip Muscle Strength

Isometric peak torque for hip abduction (HAB), hip adduction (HAD), hip flexion (HF), and hip extension (HE) was assessed using an isokinetic dynamometer (Isomed 2000; D&R GmbH, Hernau, Germany). Both HAB and HAD were measured in a lateral position, while HF and HE were tested with the patients lying supine on the isokinetic device. All measurements were applied using defined angle positions, with 0° hip abduction for HAB, 20° hip abduction for HAD, 20° hip flexion for HF, and 40° hip flexion for HE. Patients were prohibited from participating in strenuous physical activity 24 hours before the measurements. Before a measurement was conducted, the patients warmed up on a bicycle ergometer for 5 minutes (self-chosen intensity from 20 to 60 W). The warm-up was followed by 5 minutes of stretching exercises for the lower extremities.28 A physiotherapist conducted manual facilitation to familiarize patients with the measurement procedure. In addition, the investigator gave a verbal explanation of the measurement procedure and emphasized the importance of maximal muscle contraction. The starting lower extremity was determined by drawing lots, and after 2 to 3 submaximal trials, the isometric peak torque of HAB and HAD was evaluated using the best of 3 maximal isometric muscle contractions. The same procedure was then done for the isometric measurements of HF and HE. Further details of the measurement procedure are available in Steinhilber et al.22 Strength variables were summarized to generate one total hip strength score by using the procedure presented by Chandler et al.29 The authors z-transformed each strength measure gender-specifically to account for differences in strength-generating capacities of different muscle groups.29 Strength is given as newton meter per body weight (Nm/kg).

PROM Measures

Hip flexion and internal rotation PROM were examined visually by a physiotherapist. The examiner identified whether impairments were present or absent, using the criteria established by the American College of Rheumatology. Impaired PROM in hip flexion was less than 115°, and in internal hip rotation, less than 10°.25 Patients were placed in a supine position on an examination table. The physiotherapist flexed the subject's thigh upward until reaching the end of the available range of motion without leaving the sagittal plane, while the contralateral lower extremity was extended and held down on the surface of the examination table. Internal hip rotation was also tested in a supine position, with the hip and knee flexed to 90°.

Additional Data

Age, gender, height, and bodyweight were assessed. Body mass index was calculated by dividing the patients' body weight (kg) by height (m2).

Statistical Analysis

Data were analyzed using JMP 9.0 (SAS Inc, Cary, North Carolina). Normal distribution was monitored graphically, using box plots and normal quartile plots. Extreme values were identified a priori to the further analysis, using the outlier box plot method, which detects extreme values that exceed the upper or lower quartile ± 1.5 × interquartile range.30 The Spearman rank correlations (P = .05) for nonparametric variables were conducted to assess multicollinearity among the independent variables. A stepwise multiple regression analysis was performed to evaluate the variables that contribute to self-reported PD, the independent variable. The forward technique was used by adding variables to the model until the minimum Bayesian information criterion31 was reached. The following independent variables were included in the stepwise regression model: pain, stiffness, total hip strength score, age, BMI, gender, internal hip rotation PROM, and hip flexion PROM. The variables that were identified by the stepwise procedure were entered into the multiple regression analysis. The adjusted R2 (to the number of factors of the whole model) was used to quantify the explanatory component within the variation of self-reported PD provided by the whole regression model. Standardized betas were determined to constitute the relative importance of the explanatory variables. The calculation of the standardized betas shows the parameter estimates that would have resulted from the regression, if all variables had been standardized to a mean of 0 and a variance of 1.32 All variables included into the statistical analysis were related only to the arthritic hip joint. In the case of 2 affected hip joints, the more severe side was included for further analysis. Before data analysis was done, the total data sample was randomly split (ratio: 70 to 30 percent) into an analysis sample (A-sample) and a validation sample (V-sample) for cross validation.33 The hypotheses of the present study were evaluated on the A-sample and validated using the V-sample. Furthermore, to validate the results of the A-sample, the formula of the predicted WOMAC physical function assessed by the multiple regression analysis of the A-sample was used to generate the predicted values of the WOMAC physical function scale in the V-sample. These predicted values have been correlated with the actual WOMAC physical function values of the V-sample and compared with the correlation in the A-sample.


Extreme Values

Significant extreme values were identified in 11 patients. Of the 11 patients with extreme values, 6 had BMIs higher than 35, and therefore associated with class II obesity. A high BMI may be associated with PD independent of the other outcome measures. Four extreme values were found for the physical function subscale of the WOMAC index, with 4 patients reporting severe impairments. One subject had very severe pain. We decided to exclude these 11 patients from further analysis because of the significant difference when compared with the rest of the study population.


The functional, clinical, and anthropometric characteristics of the study population are presented in Table 1. These data are provided for the total sample, the A-sample of 105 patients, which is equal to 70% of the total data sample, and the V-sample of 44 patients, which is equal to 30% of the total data sample (Table 1).

Table 1
Table 1:
Characterization of the Study Population

Normal Distribution

After excluding the extreme values, graphical examination suggested normal distribution for the variables stiffness, BMI, age, and the summary hip muscle strength score. Pain and physical function were not normally distributed. Furthermore, the residuals resulting from the multiple regression analysis were normally distributed, which is an application requirement of this analysis.34

Collinearity Among the Independent Variables

The Spearman rank correlation detected moderate to strong collinearity among the 4 strength measures: HAB, HAD, HF, and HE (r = 0.46−0.67; P < .0001; r2 = 0.21−0.45). This verifies the generation of a single summary strength score that is representative of hip muscle strength and that reduced collinearity among the factors that were included in the regression analysis. The correlation between the WOMAC pain and stiffness subscales was moderate (r = 0.5; P = .0001).

Stepwise Regression

The stepwise regression analysis revealed 3 of 8 variables to be significant. The consecutive steps are shown in Table 2. Consequently, only these 3 factors were entered into the multiple regression model, since the variables BMI, gender, age, and hip internal rotation and hip flexion did not contribute to explaining the variance of the dependent variable, the WOMAC physical function subscale.

Table 2
Table 2:
Factors That Are Related to Physical Disability in Patients With Hip OA

Multiple Regression Analysis

When the factors detected by the stepwise procedure were entered into the multiple regression model, r2 adjusted was 0.59. Stiffness and pain correlated positively with a decreased WOMAC physical function subscale, and a negative association between hip muscle strength and the WOMAC physical function subscale was found (Table 3). Stiffness and pain were the variables that explained most of the variation in the physical function scores, followed by hip muscle strength (Table 3).

Table 3
Table 3:
Factors of the Multiple Regression Analysis of the Analysis Sample and Validation Sample


The correlation of the predicted and actual values of the WOMAC physical function scores were 0.77 with P < .001 in the A-sample and 0.76 with P < .001 in the V-sample. The multiple regression analysis was repeated with the V-sample, including the same variables as applied in the A-sample. R2 adjusted of the V-sample was again 0.59. Stiffness was again the factor that explained most of the variation in the physical function scores. The variation explained by pain was less than in the A-sample. The explanatory component of hip muscle strength was higher than in the A-sample (Table 3). The characteristic of the V-sample and A-sample was well balanced (Table 1).


The objective of the present study was to investigate the relationship between several factors and self-reported PD assessed using the WOMAC physical function subscale in patients with hip OA. Stepwise regression analysis revealed stiffness, pain, and hip muscle strength to be significant factors related to the level of PD in hip OA. These factors explained 59% (r2 adjusted = 0.59) of the variance in the WOMAC physical function subscale. Body mass index, gender, age, and hip internal rotation and hip flexion PROM did not contribute to explaining significant parts of self-reported PD.


In the A-sample, the relationship of pain to PD was high with a standardized beta of .39. Only the relationship of stiffness was found to be slightly stronger (β = .42). In the V-sample, the relationship of pain and PD was to some extent lower, with a standardized beta of .33. Therefore, pain together with stiffness seems to be the most important factor related to PD. This result is consistent with the findings of previous studies. van Djik et al19 concluded that self-reported PD was largely related to pain. Juhakoski et al3 found a correlation (r = 0.853) between pain (WOMAC subscale pain) and physical function (WOMAC subscale physical function). The observed relationship between perceived pain and PD is obvious. The physical function subscale of the WOMAC index includes questions about functional activities that are likely to increase pain, such as getting in and out of a car or climbing up and down stairs. Patients with hip OA often complain about pain under weight-bearing conditions35 or maximal joint excursions.25,36 Both might appear during the tasks that are included in the WOMAC index.


Our findings indicate a considerable relationship between joint stiffness and self-reported PD in patients with hip OA. However, it seems that stiffness of the affected hip has not been of great interest in previous studies. Although Juhakoski et al3 and van Dijk et al19 used the WOMAC subscales for pain and physical function, they did not quantify the stiffness subscale of the WOMAC index and its relationship to PD. It is possible that they considered PROM of the hip that was evaluated in both studies would be sufficient to describe the flexibility and stiffness of the hip joint, as well. The stiffness subscale of the WOMAC index is assessed using only 2 items. The first item refers to the severity of stiffness that occurs after first awakening in the morning, and the second item refers to the severity of stiffness that occurs after periods of inactivity later in the day. It is not surprising that a high level of joint stiffness affects the physical performance of activities of daily living, which is associated with self-reported PD. However, a moderate collinearity exists between stiffness and pain. This might interfere with the results of the multiple regression analysis and overestimate one of these 2 factors. This is a known problem in regression analysis caused by multicollinearity.37 A subgroup analysis of patients with hip OA with low levels of pain probably might be an approach to investigate the influence of stiffness. Unfortunately, in our study, the sample was too small for an adequate subgroup analysis.

Hip Muscle Strength

The summarized total hip strength score of the affected hip joint was another significant factor related to PD in hip OA, though to a lesser degree than pain and stiffness. Our results showed a negative relationship between hip muscle strength and PD, which implies that increased hip muscle strength is associated with a lower degree of PD. Other authors have discussed this relationship between hip muscle strength and PD.3,8,15,19 Juhakoski et al3 found no relationship between hip extensor power and self-reported PD, and van Dijk et al19 found no association between hip abductor strength and self-reported PD. In their review article, Dekker et al8 concluded that reduced muscle strength was a risk factor for functional decline in patients with hip and knee OA. Pua et al15 reported a nonlinear relationship between decreased hip extensor strength and PD assessed using physical performance tests in patients with hip OA. They further pointed out that there might be a limit to the degree in which improvements in hip extensor strength would be beneficial. If hip extensor strength increases beyond a particular point, there may be no more additional benefits. Our results showed strong collinearity between HAB, HAD, HF, and HE strength. This collinearity leads us to question whether the level of PD should be attributed to the strength of one single muscle or group of muscles that provide one movement direction of the hip joint, for example, hip extension. The items of the WOMAC questionnaire that aim to quantify PD refer to different movement tasks of daily living, and these activities are determined by several muscles of the hip joint acting in all of the earlier-mentioned movement directions. The use of a summarized hip strength score to investigate the relationship between PD and muscle strength in hip OA might be a better approach, especially since a major part of the impairments in hip muscle strength is considered to be the result of disuse atrophy,38,39 which affects all muscles of the lower extremity.

Passive Range of Motion

Passive range of motion of hip flexion and hip internal rotation were not significant factors associated with PD. Passive range of motion at the hip was not examined as a continuous variable and identified only as being impaired or intact, which might have reduced the sensitivity of our investigation technique. Therefore, the relationship between PROM and PD might be underestimated in the present study. Van Djik et al,19 who measured PROM in degree, mentioned that hip flexion PROM is an important factor of PD. The main difference between our study and that conducted by van Djik et al,19 besides the 2 measurement approaches, was the inclusion of patients who were hospitalized. Our patients were recruited from the local population through newspaper advertisements. The degree of PD was higher in the sample of van Djik et al. They reported a WOMAC physical function subscale of 61.01, which corresponds to a normalized score of 3.4. In comparison, we assessed a normalized WOMAC physical function subscale of 2. It seems reasonable that the impairments in PROM were also more pronounced in the sample investigated by van Djik et al and would probably result in greater limitations in physical function. Unfortunately, this is speculative, since only the presence of impaired PROM was examined.

Consequences for Exercise Therapy

In the context of our study, it is important to state that there is scientific evidence that exercise reduces pain in patients with hip OA,40 especially if the exercise routine includes muscle strengthening.41 This emphasizes the role of strengthening exercise in the management of PD in patients with hip OA, since 2 main factors associated with PD (pain and hip muscle strength) are addressed by this kind of exercise therapy. However, it still remains unclear which exercise modalities are best to reduce pain and concurrently improve hip muscle strength. Furthermore, the present study implies that future research should investigate whether exercise therapy is also able to reduce stiffness, since stiffness is another major factor associated with PD in hip OA. Finally, we strongly recommend addressing exercises not only for single hip muscles but also for improvement in the strength of all muscles that are related to the hip joint.


A limitation of the present study may be the method used to determine PD. Physical disability could be quantified by self-reported questionnaires or physical performance tests.4 Juhakoski et al3 found good correlations between physical performance tests and the WOMAC subscale physical function. However, neither of these techniques, implemented independently or combined, would be able to completely describe the limitations in physical performance and activities of daily living caused by hip OA. Patients' lifestyles and activity levels are too different to allow this. We suggest that, although incomplete, self-reported PD is the aspect of PD most greatly perceived by patients, and therefore, it is an important measure. However, the World Health Organization has recommended that the WOMAC index is the most relevant instrument for assessing disability in OA of the lower limbs.42 As already mentioned in the “Discussion” section, the categorical assessment of PROM might be not as sensitive as the continuous assessment by using goniometry. Although studies report that the results of these 2 measurement approaches are quiet similar,43,44 the measurement technique of PROM must be considered as limitation, which might be the reason that PROM is not a related factor to PD in the present study. Other explanatory factors of PD have been proposed,3,8,19 including patients' level of education, self-efficacy, the amount of comorbidities. We did not measure these factors and therefore cannot include them in our regression model. Our statistical analysis revealed that about 40% of the variation in self-reported PD is still unexplained.


Although we did not include all potential factors, our study indicates that stiffness, pain, and decreased hip muscle strength are significant factors in the level of self-reported PD in individuals with hip OA. These 3 factors should be considered as potential targets to be addressed by exercise interventions to reduce PD in patients with hip OA.


This study was funded internally by the authors' department.


1. Felson DT, Lawrence RC, Dieppe PA, et al. Osteoarthritis: new insights. Part 1: the disease and its risk factors. Ann Intern Med. 2000;133(8):635–646.
2. Felson DT, Zhang Y. An update on the epidemiology of knee and hip osteoarthritis with a view to prevention. Arthritis Rheum. 1998;41(8):1343–1355.
3. 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(6):1066–1073.
4. Arokoski MH, Haara M, Helminen HJ, Arokoski JP. Physical function in men with and without hip osteoarthritis. Arch Phys Med Rehabil. 2004;85(4):574–581.
5. Rasch A, Dalen N, Berg HE. Muscle strength, gait, and balance in 20 patients with hip osteoarthritis followed for 2 years after THA. Acta Orthop. 2010;81(2):183–188.
6. Dagenais S, Garbedian S, Wai EK. Systematic review of the prevalence of radiographic primary hip osteoarthritis. Clin Orthop Relat Res. 2009;467(3):623–637.
7. Fautrel B, Hilliquin P, Rozenberg S, et al. Impact of osteoarthritis: results of a nationwide survey of 10,000 patients consulting for OA. Joint Bone Spine. 2005;72(3):235–240.
8. Dekker J, van Dijk GM, Veenhof C. Risk factors for functional decline in osteoarthritis of the hip or knee. Curr Opin Rheumatol. 2009;21(5):520–524.
9. Bitton R. The economic burden of osteoarthritis. Am J Manag Care. 2009;15(8) (suppl):230–235.
10. Pelland L, Brosseau L, Wells G, et al. Efficacy of strengthening exercises for osteoarthritis (part I): a meta-analysis. Phys Ther Rev. 2004;9(2):77–108.
11. Zhang W, Moskowitz RW, Nuki G, et al. OARSI recommendations for the management of hip and knee osteoarthritis, part II: OARSI evidence-based, expert consensus guidelines. Osteoarthritis Cartilage. 2008;16(2):137–162.
12. Fransen M, McConnell S, Hernandez-Molina G, Reichenbach S. Exercise for osteoarthritis of the hip. Cochrane Database Syst Rev. 2009;8(3):CD007912.
13. Zhang W, Doherty M, Arden N, et al. EULAR evidence based recommendations for the management of hip osteoarthritis: report of a task force of the EULAR Standing Committee for International Clinical Studies Including Therapeutics (ESCISIT). Ann Rheum Dis. 2005;64(5):669–681.
14. Pua YH, Wrigley TV, Cowan SM, Bennell KL. Hip flexion range of motion and physical function in hip osteoarthritis: mediating effects of hip extensor strength and pain. Arthritis Rheum. 2009;61(5):633–640.
15. 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(4):442–450.
16. Pua YH, Wrigley TV, Collins M, Cowan SM, Bennell KL. Self-report and physical performance measures of physical function in hip osteoarthritis: relationship to isometric quadriceps torque development. Arthritis Rheum. 2009;61(2):201–208.
17. Rosemann T, Kuehlein T, Laux G, Szecsenyi J. Factors associated with physical activity of patients with osteoarthritis of the lower limb. J Eval Clin Pract. 2008;14(2):288–293.
18. Steultjens MP, Dekker J, van Baar ME, Oostendorp RA, Bijlsma JW. Range of joint motion and disability in patients with osteoarthritis of the knee or hip. Rheumatology (Oxford). 2000;39(9):955–961.
19. van Dijk GM, Veenhof C, Lankhorst GJ, Dekker J. Limitations in activities in patients with osteoarthritis of the hip or knee: the relationship with body functions, comorbidity and cognitive functioning. Disabil Rehabil. 2009;31(20):1685–1691.
20. Juhakoski R, Malmivaara A, Lakka TA, Tenhonen S, Hannila ML, Arokoski JP. Determinants of pain and functioning in hip osteoarthritis—a two-year prospective study. Clin Rehabil. 2013;27(3):281–287.
21. Harrison AL. The influence of pathology, pain, balance, and self-efficacy on function in women with osteoarthritis of the knee. Phys Ther. 2004;84(9):822–831.
22. Steinhilber B, Haupt G, Boeer J, Grau S, Krauss I. Reproducibility of concentric isokinetic and isometric strength measurements at the hip in patients with hip osteoarthritis: a preliminary study. Isokinet Exerc Sci. 2011;19(1):39–46.
23. Steinhilber B, Haupt G, Miller R, et al. Feasibility and efficacy of an 8-week progressive home-based strengthening exercise program in patients with osteoarthritis of the hip and/or total hip joint replacement: a preliminary trial. Clin Rheumatol. 2011;31(3):511–519.
24. Krauss I, Steinhilber B, Haupt G, Miller R, Grau S, Janssen P. Efficacy of conservative treatment regimes for hip osteoarthritis—evaluation of the therapeutic exercise regime “Hip School”: a protocol for a randomised, controlled trial. BMC Musculoskelet Disord. 2011;12:270–282.
25. Altman R, Alarcon G, Appelrouth D, et al. The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hip. Arthritis Rheum. 1991;34(5):505–514.
26. Stucki G, Meier D, Stucki S, et al. [Evaluation of a German version of WOMAC (Western Ontario and McMaster Universities) Arthrosis Index]. Z Rheumatol. 1996;55(1):40–49.
27. 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(12):1833–1840.
28. Zakas A, Doganis G, Zakas N, Vergou A. Acute effects of active warm-up and stretching on the flexibility of elderly women. J Sports Med Phys Fitness. 2006;46(4):617–622.
29. Chandler JM, Duncan PW, Kochersberger G, Studenski S. Is lower extremity strength gain associated with improvement in physical performance and disability in frail, community-dwelling elders? Arch Phys Med Rehabil. 1998;79(1):24–30.
30. Williamson DF, Parker RA, Kendrick JS. The box plot: a simple visual method to interpret data. Ann Intern Med. 1989;110(11):916–921.
31. Schwarz G. Estimating the dimension of a model. Ann Stat. 1978;6(2):461–464.
32. SAS Institute Inc. JMP Statistics and Graphics Guide. Release 7 ed. Cary, NC: SAS Institue Inc; 2007.
33. Picard R, Cook D. Cross-validation of regression models. J Am Stat Assoc. 1984;79(387):575–583.
34. Brosius F. Lineare regression. SPSS 19. Heidelberg, Germany: Hüthig Jehle Rehm GmbH; 2011;574–580.
35. Netter F. Netters Orthopädie. 1st ed. Stuttgart, Germany: Georg Thieme Verlag; 2001.
36. Winkel D, van Paridon-Edauw D, Matthijs O. Diagnose- und Therapieschemata in der Nichtoperativen Orthopädie. Stuttgart, Germany: Gustav Fischer Verlag; 1995.
37. Tu YK, Kellett M, Clerehugh V, Gilthorpe MS. Problems of correlations between explanatory variables in multiple regression analyses in the dental literature. Br Dent J. 2005;199(7):457–461.
38. Arokoski MH, Arokoski JP, Haara M, et al. Hip muscle strength and muscle cross sectional area in men with and without hip osteoarthritis. J Rheumatol. 2002;29(10):2185–2195.
39. Rasch A, Dalen N, Berg HE. Test methods to detect hip and knee muscle weakness and gait disturbance in patients with hip osteoarthritis. Arch Phys Med Rehabil. 2005;86(12):2371–2376.
40. Zhang W, Nuki G, Moskowitz RW, 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(4):476–499.
41. Hernandez-Molina G, Reichenbach S, Zhang B, Lavalley M, Felson DT. Effect of therapeutic exercise for hip osteoarthritis pain: results of a meta-analysis. Arthritis Rheum. 2008;59(9):1221–1228.
42. WHO Scientific Group. The Burden of Musculoskeletal Conditions at the Start of the New Millennium. Geneva, Switzerland: World Health Organization; 2003. Report no.: 919.
43. Holm I, Bolstad B, Lutken T, Ervik A, Rokkum M, Steen H. Reliability of goniometric measurements and visual estimates of hip ROM in patients with osteoarthrosis. Physiother Res Int. 2000;5(4):241–248.
44. Rachkidi R, Ghanem I, Kalouche I, El HS, Dagher F, Kharrat K. Is visual estimation of passive range of motion in the pediatric lower limb valid and reliable? BMC Musculoskelet Disord. 2009;10:126–136.

hip muscle strength; hip osteoarthritis; physical function; stiffness; WOMAC

Copyright © 2014 the Section on Geriatrics of the American Physical Therapy Association