Reproducibility of concentric isokinetic and isometric ... - IOS Press

39

Isokinetics and Exercise Science 19 (2011) 39–46 DOI 10.3233/IES-2011-0395 IOS Press

Reproducibility of concentric isokinetic and isometric strength measurements at the hip in patients with hip osteoarthritis: A preliminary study Benjamin Steinhilber∗ , Georg Haupt, Johannes Boeer, Stefan Grau and Inga Krauss Medical Clinic, Department of Sports Medicine, University of Tübingen, Tübingen, Germany

Abstract. To determine the reproducibility of isokinetic and isometric hip flexion, extension, abduction and adduction strength scores in patients with hip osteoarthritis (N = 16) and healthy subjects, testing was conducted twice 7 days apart. The patient group (PG) consisted of 16 subjects with unilateral or bilateral osteoarthritis of the hip (11 women, 5 men; age 56–75y). The control group (CG) included 13 age-matched healthy subjects (9 women, 4 men; age 54–73y). The standard error of measurement (SEM) served as the reproducibility outcome parameter. The highest SEM values were obtained for hip extension measurements (15.4–16.7 Nm), followed by hip adduction (7.9–11.8 Nm) and hip flexion measurements (5.2–7.5 Nm). The smallest values were quantified for hip abduction measurements (4.7–7.7 Nm). PG showed larger values in all isometric measures except for hip extension and isokinetic hip abduction. When the 95% confidence interval is incorporated, these SEMs become significant. Thus clinicians should be aware of the large measurement error in patients with hip OA. Adequate preparations and firm fixation of the pelvis are key elements to achieve reproducible results. Keywords: Dynamometer, muscle strength, hip osteoarthritis, reproducibility

1. Introduction Osteoarthritis (OA) of the hip is a common, chronic disease among older people which leads to a loss of hyaline articular cartilage [24]. The prevalence of hip OA increases with age, and is expected to become more prevalent due to demographic changes in western societies [24]. Patients with hip OA have joint pain [16,20, 21], limitations in physical function [12,16,21,22] and impaired quality of life [16,20]. They are also typically impaired with weak hip muscles compared to healthy subjects [2,17,26,28], which is thought to be the re-

∗ Address

for correspondence: Benjamin Steinhilber, Silcherstrasse 5, 72076 Tübingen, Germany. Tel.: +49 7071 29 86488; Fax: +49 7071 29 4626; E-mail: [email protected].

sult of disuse atrophy caused by joint pain [2,28] or arthrogenous inhibition of muscle contraction [8]. Impaired hip muscle strength is a consequence of hip OA, yet is also associated with an accelerated progression of the disease [11,31]. Furthermore, hip muscle strength is an important correlate of physical function in people with hip OA [25,31] and is often used as an outcome measure in clinical trials [12–14,21]. Therefore, reliable measurements of hip muscle strength in patients with hip OA are crucial for diagnostic and rehabilitation processes. A common way to assess hip muscle strength in patients with hip OA is using an isokinetic device [21,27, 28,32], however only a few studies [2,25] have evaluated the reproducibility of such measurements in these patients. Moreover, neither of these studies was specifically designed to investigate whether hip OA influ-

ISSN 0959-3020/11/$27.50  2011 – IOS Press and the authors. All rights reserved

40

B. Steinhilber et al. / Reproducibility of concentric isokinetic and isometric strength measurements Table 1 Isokinetic and isometric measures Measurement condition Movement Body position HAB lateral HAD HF supine HE

AV (◦ /s) 30 30 40 40

Isokinetic Start-stop position (◦ ) 0–35 0–35 10–65 10–65

Total RoM (◦ ) 35 35 55 55

Isometric hip angle position (◦ ) 0 (hip abduction) 10 (hip abduction) 20 (hip flexion) 40 (hip flexion)

Abbreviations: HAB, hip abduction; HAD, hip adduction; HF, hip flexion; HE, hip extension; AV, angular velocity; RoM, range of motion.

enced reproducibility. Therefore, the measurement error (ME) of isokinetic and isometric strength measurements could be underestimated in these patients. Isokinetic devices are produced by several manufacturers using different software and hardware components. Additionally the applied measurement protocols differ substantially with respect to angular velocity, RoM and positioning of the tested subject. We found no study that assessed the reproducibility of strength measurements at the hip using the Isomed 2000 (Ferstl GmbH). Therefore the purpose of our study was to investigate the day-to-day reproducibility of isokinetic and isometric strength measurements at the hip using the Isomed 2000 in patients with hip OA and healthy subjects [10, 15,25,29].

2. Methods 2.1. Subjects Seventeen patients with unilateral or bilateral hip OA and sixteen healthy subjects volunteered to participate in the present study. The study protocol complied with the Helsinki Declaration and was approved by the Ethics Committee of the university. Subjects signed written informed consent forms. The patient group (PG) consisted of five males and twelve females aged 56 to 75 (mean and [SD] 64.6 [5.6] yrs). Hip OA was assessed according to the classification criteria of the American College of Rheumatology [1]. Subjects were admitted to the study if they had hip OA in at least one hip joint. Total hip replacement (THR) on the contralateral side was allowed. Neurological and advanced cardiopulmonary diseases, previous surgery of the lower extremities except THR, and insufficient mounting of the hip endoprosthesis led to exclusion. The control group (CG) included healthy subjects, five males and eleven females aged 54 to 73 (62.7 [6.1]).

2.2. Procedures The Isomed 2000 (D&R GmbH, Hernau, Germany) isokinetic dynamometer was used to measure concentric isokinetic and isometric peak torque for hip abduction (HAB), hip adduction (HAD), hip flexion (HF) and hip extension (HE). Subjects were tested twice over 7 days. All measurements were conducted by the same two raters, with the same measurement protocol and at the same time of the day to control for circadian variation in performance. In addition, subjects were urged to avoid any extra physical activity 24 h before each test occasion. The patient group was examined by an orthopedic surgeon on the first test occasion (TO1) before muscle strength was determined. Eleven hip patients were analyzed a third time one week after the second measurement to specifically examine whether familiarization with the test situation had increased the reproducibility of these measurements. The same protocol was used as on TO1 and TO2. Isokinetic peak torque was determined using a predefined range of motion (RoM) and angular velocities (AV) shown in Table 1. RoM and AV were chosen according to the impairments of subjects with hip OA and were determined by a series of pre-tests before the beginning of this study. Measurements in the frontal plane (HAB and HAD) were done first, followed by the sagittal plane (HF and HE). Subjects warmed up on a bicycle ergometer for 5min with a self chosen resistance between 20 and 60-W, followed by 5-min of stretching exercises for the lower extremities. Subsequently, the starting leg was allotted and subjects removed their shoes. Prior to the test, one rater familiarized the subjects with the measurement procedure and attached an orthosis to the tested leg. In the frontal plane, the orthosis was set at 90◦ at the knee (Fig. 1a). For the sagittal plane, the orthosis was set at 0◦ (Fig. 1b). The orthosis was necessary to decrease measurement artifacts related to oscillations and resulting moments of inertia of the shank and foot.

B. Steinhilber et al. / Reproducibility of concentric isokinetic and isometric strength measurements

41

Fig. 1. 1a and b. Measurement position and manual fixation during HAD and HAB (a) and HF and HE (b).

Fig. 2. Two raters conducting a hip extension strength measurement.

Subjects assumed the lateral and supine position for HAB/HAD and HF/HE testing, respectively. We ensured maximal axis-joint alignment before subjects were fixed to the surface and dynamometer. Subjects were stabilized using belts strapped around their pelvis and contralateral leg. Their arms were kept in a standardized position (Fig. 2) to ensure that they could not enhance contractions with their upper body muscles. Measurements were conducted by two raters (Fig. 2), decreasing the time necessary per measurement and ensuring a high level of standardization with respect to the test procedure. While rater one (R1) operated the isokinetic device and gave standardized instructions

to each subject, rater two (R2) was responsible for stabilization and measurement position by manually stabilizing the subject’s pelvis during all measurements of HAD and HE, where the pelvis tended to become instable. Additionally, R2 observed the subjects during the measurements for signs of pain or muscle cramps. Subjects completed a set of 15 low intensity dynamic repetitions to warm up and familiarize themselves with the motion characteristics. Subsequently, corrections were done to control for the effects of gravity on the leg and lever arm. Subjects were then instructed to push and pull their leg at maximum exertion as fast as possible over the entire RoM, breathe normally and keep

42


their hands in the standardized position. Subjects performed three maximum repetitions to familiarize themselves with the measurement procedure. They were again reminded to perform at maximum exertion and given a short rest. Subsequently, isokinetic peak torque of HAB and HAD were determined by 5 consecutive maximum repetitions at 30◦ /s over the entire RoM. R2 manually stabilized the measurement positions (Fig. 1a and b). Isokinetic peak torque was determined by the mean value of the best three of the last four repetitions. The first repetition was not considered for data analysis. Subjects were not verbally encouraged, but R1 accompanied the measurement verbally with “up” and “down” to ensure subjects remained active until the end of the measurement. After a 2-min rest, isometric peak torque of HAB and HAD was measured. Subjects were given a short practice period of two contractions followed by a break for regeneration. Subjects were instructed to completely relax their leg for 5-s at the beginning of the measurement to determine the baseline. They were then requested to contract slowly until their maximum strength was reached and to maintain this exertion for three seconds. Between contractions subjects were instructed to relax their leg until R1 gave the command for the next contraction. They were reminded to breathe normally and to perform at their maximum exertion. R2 stabilized the measurement position by manually fixating the subject’s pelvis during HAD (Fig. 1a). Isometric peak torque was determined by three maximal isometric contractions. The best of these three contractions was considered the maximum isometric peak torque. HF and HE were then measured analogously to the procedure described above in a supine position and at 40◦ /s. Pain was assessed using the subscale of the WOMAC questionnaire. 2.3. Statistical analyses

square, which describes within-subject variance [4]. This within-subject standard deviation, also called the standard error of measurement (SEM) is the most important type of reproducibility measure, since it affects the precision of estimates of change in the variable of an experimental study [15]. SEM also leads to the smallest real difference (SRD = 2.77SEM), which covers the individual differences on retest and thus serves as a cutoff for change at p = 0.05 [5]. 3. Results 3.1. Dropouts Four subjects did not finish the study. Two female subjects withdrew due to lower back pain. Two male subjects had to be excluded. One subject was not compliant with respect to not partaking in physical exercise 24-h before a test. The other subject had to be excluded because the isokinetic device failed to store his measurement data due to inadequate memory capacity. 3.2. Extreme values 3 extreme values were detected. One was found in the isometric HE measurement of CG from TO1 and TO2. As noted on the observation form, we excluded data from this subject due to gastroenterological problems during TO1, which could also have influenced TO2. Another two extreme values were found in PG for HE measures under isometric and isokinetic conditions from TO2 and TO3, both caused by the same subject. These plots indicate a bias between TO2 and TO3 that does not represent the data characteristics of all other observations. We therefore decided to exclude these measures. 3.3. Subjects

Findings were analyzed using JMP 7.0 (SAS Inc. Cary, NC, USA). A one way analysis of variance (ANOVA) with the factor group was conducted on the differences of the strength measures between TO1 and TO2 to control for significant differences between PG and CG. Bland-Altman plots [5] were used for visual inspection of heteroscedasticity and systematic bias (bias: mean difference between test occasions is not equal to zero). These plots show the difference in strength between TO1 and TO2 in relation to their mean strength value. Another ANOVA with subjects as a factor was conducted to calculate the residual mean

In all 16 patients with OA and 13 normal subjects completed the study and were included in the data analysis of test sequence one (TS1), consisting of the results from TO1 and TO2. Strength measurements were well tolerated in patients with hip OA. The two groups were similar in age, height and weight. Eleven subjects from PG participated in a third measurement one week after TO2. Their results are shown in test sequence two (TS2). The absolute values of strength of all of the three test occasions are presented in Table 2.


43

Table 2 Peak torque measures (in Nm) of TO1, TO2 (PG: n = 16; CG: n = 13) and TO3 (PG: n = 11) Measure Group Isokinetic HAB [SD] HAD [SD] HF [SD] HE [SD] Isometric HAB [SD] HAD [SD] HF [SD] HE [SD]

Peak torque TO1 PG CG

Peak torque TO2 PG CG

Peak torque TO3 PG (n = 11)

77.8 [21.7] 94.4 [32] 73.9 [21.5] 95.6 [27.8]

79.5 [23.1] 87.3 [17.4] 78.1 [17.8] 92.7 [32.3]

79.3 [24.6] 96 [30.4] 78.3 [21.2] 104.2 [29.2]

78.6 [23.8] 92.6 [21.6] 79.2 [17] 104.5 [29.4]

83.1 [31.3] 103.6 [32.3] 78.6 [232.6] 111.7 [43.2]

118.7 [38.1] 94.2 [38.3] 83.7 [24.2] 106.8 [42.1]

110.3 [24.3] 85.4 [23.8] 81.6 [17.5] 105.1 [44.3]

114.1 [40.4] 92.9 [34] 84.8 [26.3] 113.4 [47.5]

110.7 [28.2] 86.7 [25.2] 83.1 [15.4] 104.9 [42.3]

117.8 [46] 102.7 [40.2] 85.4 [29.8] 118.1 [59.8]

Abbreviations: TO, test occasion; HAB, hip abduction; HAD, hip adduction; HF, hip flexion; HE, hip extension; SD, standard deviation. Table 3 Measures of TS1 (PG: n = 16; CG: n = 13)

3.4. Test sequence one Bland-Altman plots indicated homoscedastic distribution of the test-retest differences in both the isokinetic and isometric hip strength measurements for either group. In addition, these plots indicated no relevant systematic bias except for HE, where larger mean values were generated for TO2 compared to TO1 (Table 2). The HE-related bias was 6–12% of the corresponding mean strength value with the largest values associated with the isokinetic test condition. The ANOVA revealed no significant influence of the factor group on differences between test days. 3.5. Measures of reproducibility The lowest measurement error (ME) was found in HAB, followed by HF and HAD. The greatest ME was quantified in HE. ME did not differ systematically between isokinetic and isometric measurements within each experimental group (Table 3). Day-to-day reproducibility among healthy and OA subjects under isokinetic conditions was similar, except for HF. ME for isokinetic HF was about one fourth less in healthy subjects. Under isometric conditions, reproducibility measures of isometric HAB, HAD and HF revealed smaller values for CG compared to PG. No differences were found between groups for the isometric HE measurement, which showed the lowest reproducibility.

Measure Group Isokinetic HAB HAD HF HE∗ Isometric HAB HAD HF HE

Bias (Nm) PG CG

SEM (Nm) PG CG

SRD (Nm) PG CG

−1.5 −1.6 −4.4 −8.6

1.0 −5.4 −1.1 −11.8

4.8 11.8 6.9 16.7

4.9 10.2 5.2 16.7

13.3 32.7 19.1 46.3

13.6 28.3 14.4 46.3

4.6 1.2 −1.1 −6.7

−0.4 −1.3 −1.5 −5.1

7.7 10.1 7.5 15.4

4.7 7.9 6.0 15.6

21.3 30.0 20.8 42.7

13.0 21.9 16.6 43.2

∗ One

healthy subject excluded (see extreme values). Abbreviations: HAB, hip abduction; HAD, hip adduction; HF, hip flexion; HE, hip extension; SRD, repeatability coefficient; SEM, standard error of measurement.

Table 4 Measures of TS1 and TS2 from eleven PG subjects Measure Test Sequence Isokinetic HAB HAD HF HE Isometric HAB HAD HF HE

Bias (Nm) TS1 TS2

SEM (Nm) TS1 TS2

−1.2 −1.2 −3.2 −2.3

0.1 −5.7 2.5 −1.6∗

5.4 10.3 6.7 12.0

7.7 13.3 6.7 8.0∗

3.9 −2.6 2.6 −1.3

1.2 −6.1 0.6 −1.5∗

7.3 10.3 7.1 12.3

4.9 7.4 8.0 10.6∗

∗ One

3.6. Test sequence two

subject excluded (see extreme values). Abbreviations: HAB, hip abduction; HAD, hip adduction; HF, hip flexion; HE, hip extension; SEM, standard error of measurement; TS1, test sequence 1; TS2, test sequence 2.

Bland-Altman plots have indicated homoscedasticity in both the isokinetic and isometric strength measurements. No systematic bias was found for either test sequence. Reproducibility did not differ system-

atically under isokinetic conditions between TS1 and TS2. A tendency of lower reproducibility in TS1 under isometric conditions has been found (Table 4).

44


4. Discussion Day-to-day reproducibility is quantified by relative and absolute measures of reproducibility. The information given by relative reproducibility measures is discussed in terms of clinical application [9]. It is unclear among experts [3,15] whether a specific index found to be statistically significant for a set of data also means that it is clinically meaningful. According to Atkinson and Nevill [3], the acceptance of a certain degree of ME is more important than a significant index. In our opinion, it is appropriate to quantify the day-today reproducibility using SEM and SRD, which are absolute measures of reproducibility. Complementary information is given by Bland-Altman plots. Studies investigating the effect of hip strengthening in patients with hip OA and total hip replacement (THR) suggest that strength improvements of about 15% are associated with improvements in physical functions [12,18]. These studies included a control group that received no treatment, thus differences between groups could be attributed to the interventional program. In clinical practice, the aim is to use strength measurements at the hip to monitor individual subjects and to assess therapy strategies. However, ME cannot be controlled in a clinical setting. According to the results of the present study (Table 3) a clinician would barely be able to distinguish whether changes were due to improvements in strength or to the repeated measurement alone. Without knowing the day-to-day reproducibility of strength measurements at the hip, misinterpretations are imminent. Our results indicate that the muscle group-direction of motion combination greatly influences reproducibility. The highest ME occurred in HE followed by HAD, independent of the group and unrelated to the testing condition. During these two measurements, the leg moved downward causing an upward movement of the pelvis. Although the pelvis was stabilized by a belt and an examiner, even small incongruence between the center of the lever arm and the hip joint center can induce errors by different moment-generating capacities [6]. In contrast, leg movements in HAB and HF were in an upward direction, resulting in a downward movement of the pelvis. The pelvis was firmly pressed against the contact surface and thus stabilized resulting in a smaller ME. Other authors have alluded to MEs caused by inadequate stabilization of the pelvis and the trunk during strength measurements at the hip [2,19, 33]. Nevertheless, ME is dependent on several factors like a subject’s motivation, physical condition on the

test day, pain and other factors. However, these factors were comparable among the different movements that were tested. The main factor that leads to different MEs between the four movement directions must be due to the above described different misalignment during the movements. In addition, we found that ME was smaller under isometric conditions, which was seen for both groups. We suggest that isokinetic measurements at the hip are more difficult to perform than their isometric counterparts, since adequate stabilization of the pelvis is more difficult under dynamic conditions, enabling subjects to perform the movement differently on separate test occasions. The main aim of this study was to determine whether ME is larger in OA patients. We found lower reproducibility in PG compared to CG under isometric conditions for HAB, HAD and HF. According to the force velocity relationship of isokinetic measurements [7], peak torque is larger in isometric measures, and dayto-day variability of physical performance of PG may have become more relevant in conjunction with maximum forces. No differences were found between PG and CG in isometric HE measurements, but the ME for both groups was large. The previously described motion characteristics during HE measurements point to inadequate fixation, which might lead to actual differences between PG and CG being overlooked. Further, hip OA did not seem to affect the reproducibility of isokinetic measurements. Reproducibility did not differ between groups, except for distinctly higher values for HF in PG, implying lower reproducibility in the PG. The lower reproducibility could be caused by pathological processes of the hip joint capsule [26,30], which is shortened and has lost flexibility and thus stressed by additional tension during hip flexion movements causing pain [23]. Although measurements were well tolerated by PG, we speculate that most subjects feared experiencing pain during this measurement, since the movement could be associated with painful daily situations. Another aim of our study was to determine whether subjects experience a practice-based improvement (PBI) in performance after repeated measurements. On the one hand, Bland-Altman plots of the subpopulation who participated in three test occasions indicate no systematic bias between TS1 and TS2. This indicates no PBI in one direction, which could for example be caused by habituation. On the other hand, the results from all subjects that participated in TS1 showed higher mean strength val-


ues in HE during TO2 independent of group or measurement condition. This indicates the existence of a PBI resulting in higher strength values at TO2. We hypothesize that this effect was probably associated with rater 2 who has been responsible for manual stabilization of the pelvis. As there were more possible pelvic movements in HE, it seems reasonable that R2 did not exert enough pressure to stabilize the pelvis appropriately at the beginning of our study. A further look into the raw data showed that the bias was caused by the same four subjects at the beginning of the study, which supports this hypothesis. Additionally, measures of reproducibility tended to be smaller for TS2 under isometric conditions. This indicates a PBI in terms of a more precise and stable performance during TS2, free from influencing factors like fear or uncertainness with the test situation. Thus, although it was a high priority to prepare subjects for the measurements, we could not completely avoid a PBI. Strength measurements at the hip could be used to compile an exercise regime for rehabilitation or therapy programs. However, differences in strength measures of individual test-retest settings in daily clinical routine should be interpreted carefully. Differences up to 46.3 Nm for hip extension reflect the measurement error albeit at its stringent envelope (95% CI). Clinicians should consider the larger measurement error in patients with hip OA. In general, isometric measures are more stable than isokinetic measures. Premeasurement preparation and firm stabilization, especially of the pelvis, are key elements to achieve reliable results in clinical studies and daily routine. A test measurement could improve these types of measurements.

[6]

[7] [8]

[9]

[10] [11]

[12]

[13]

[14]

[15] [16]

[17]

[18]

[19]

[20]

References [1]

[2]

[3]

[4] [5]

R. Altman, G. Alarcon, D. Appelrouth, D. Bloch, D. Borenstein, K. Brandt, C. Brown, T.D. Cooke, W. Daniel and D. Feldman, The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hip, Arthritis Rheum 34 (1991), 505–514. M.H. Arokoski, J.P. Arokoski, M. Haara, M. Kankaanpaa, M. Vesterinen, L.H. Niemitukia and H.J. Helminen, Hip muscle strength and muscle cross sectional area in men with and without hip osteoarthritis, J Rheumatol 29 (2002), 2185–2195. G. Atkinson and A.M. Nevill, Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine, Sports Med 26 (1998), 217–238. J.M. Bland and D.G. Altman, Measurement error, BMJ 313 (1996), 744. J.M. Bland and D.G. Altman, Measuring agreement in method comparison studies, Stat Methods Med Res 8 (1999), 135–160.

[21]

[22]

[23]

45

S.L. Delp and W. Maloney, Effects of hip center location on the moment-generating capacity of the muscles, J Biomech 26 (1993), 485–499. Z. Dvir, Isokinetics. Muscle Testing, Interpretation and Clinical Applications, (2 ed.), London: Elsevier, (2004). D.T. Felson and Y. Zhang, An update on the epidemiology of knee and hip osteoarthritis with a view to prevention, Arthritis Rheum 41 (1998), 1343–1355. M. Fransen, J. Crosbie and J. Edmonds, Isometric muscle force measurement for clinicians treating patients with osteoarthritis of the knee, Arthritis Rheum 49 (2003), 29–35. J.M. Gaines and L.A. Talbot, Isokinetic strength testing in research and practice, Biol Res Nurs 1 (1999), 57–64. E. Germanou, A. Beneka, P. Malliou, G. Tsiganos, St. Papalexandris, Ch. Bikos and G. Godolias, Reproducibility of concentric isokinetic strength of the knee extensors and flexors in individuals with mild and moderate osteoarthritis of the knee, Isokinetics and Exercise Science 15 (2007), 151–164. H.J. Gilbey, T.R. Ackland, A.W. Wang, A.R. Morton, T. Trouchet and J. Tapper, Exercise improves early functional recovery after total hip arthroplasty, Clin Orthop Relat Res (2003), 193–200. J. Green, F. McKenna, E.J. Redfern and M.A. Chamberlain, Home exercises are as effective as outpatient hydrotherapy for osteoarthritis of the hip, Br J Rheumatol 32 (1993), 812–815. R.S. Hinman, S.E. Heywood and A.R. Day, Aquatic physical therapy for hip and knee osteoarthritis: results of a single-blind randomized controlled trial, Phys Ther 87 (2007), 32–43. W.G. Hopkins, Measures of reliability in sports medicine and science, Sports Med 30 (2000), 1–15. D.J. Hunter, J.J. McDougall and F.J. Keefe, The symptoms of osteoarthritis and the genesis of pain, Rheum Dis Clin North Am 34 (2008), 623–643. M.V. Hurley, The role of muscle weakness in the pathogenesis of osteoarthritis, Rheum Dis Clin North Am 25 (1999), 283– 298. M.H. Jan, J.Y. Hung, J.C. Lin, S.F. Wang, T.K. Liu and P.F. Tang, Effects of a home program on strength, walking speed, and function after total hip replacement, Arch Phys Med Rehabil 85 (2004), 1943–1951. R.H. Jensen, G.L. Smidt and R.C. Johnston, A technique for obtaining measurements of force generated by hip muscles, Arch Phys Med Rehabil 52 (1971), 207–215. K.M. Jordan, N.K. Arden, M. Doherty, B. Bannwarth, J.W. Bijlsma, P. Dieppe, K. Gunther, H. Hauselmann, G. HerreroBeaumont, P. Kaklamanis, S. Lohmander, B. Leeb, M. Lequesne, B. Mazieres, E. Martin-Mola, K. Pavelka, A. Pendleton, L. Punzi, U. Serni, B. Swoboda, G. Verbruggen, I. Zimmerman-Gorska and M. Dougados, EULAR Recommendations 2003: an evidence based approach to the management of knee osteoarthritis: Report of a Task Force of the Standing Committee for International Clinical Studies Including Therapeutic Trials (ESCISIT), Ann Rheum Dis 62 (2003), 1145–1155. R. Juhakoski, S. Tenhonen, T. Anttonen, T. Kauppinen and J.P. Arokoski, Factors affecting self-reported pain and physical function in patients with hip osteoarthritis, Arch Phys Med Rehabil 89 (2008), 1066–1073. J.A. Kettunen and U.M. Kujala, Exercise therapy for people with rheumatoid arthritis and osteoarthritis, Scand J Med Sci Sports 14 (2004), 138–142. G.C. LLoyd-Roberts, The role of capsular changes in osteoarthritis of the hip joint, J Bone Joint Surg Br 35-B (1953), 627–642.

46 [24] [25]

[26]

[27]

[28]

[29]

B. Steinhilber et al. / Reproducibility of concentric isokinetic and isometric strength measurements NIH conference, NIH conference addresses conflict of interest concerns, J Investig Med 48 (2000), 389. Y.H. Pua, T.V. Wrigley, S.M. Cowan and K.L. Bennell, Intrarater test-retest reliability of hip range of motion and hip muscle strength measurements in persons with hip osteoarthritis, Arch Phys Med Rehabil 89 (2008), 1146–1154. Y.H. Pua, T.V. Wrigley, S.M. Cowan and K.L. Bennell, Hip flexion range of motion and physical function in hip osteoarthritis: mediating effects of hip extensor strength and pain, Arthritis Rheum 61 (2009), 633–640. A. Rasch, A.H. Bystrom, N. Dalen and H.E. Berg, Reduced muscle radiological density, cross-sectional area, and strength of major hip and knee muscles in 22 patients with hip osteoarthritis, Acta Orthop 78 (2007), 505–510. A. Rasch, N. Dalen and H.E. Berg, Test methods to detect hip and knee muscle weakness and gait disturbance in patients with hip osteoarthritis, Arch Phys Med Rehabil 86 (2005), 2371–2376. T.B. Symons, A.A. Vandervoort, C.L. Rice, T.J. Overend and G.D. Marsh, Reliability of a single-session isokinetic and iso-

[30]

[31]

[32]

[33]

metric strength measurement protocol in older men, J Gerontol A Biol Sci Med Sci 60 (2005), 114–119. S. Tarasevicius, U. Kesteris, A. Gelmanas, A. Smailys and H. Wingstrand, Intracapsular pressure and elasticity of the hip joint capsule in osteoarthritis, J Arthroplasty 22 (2007), 596–600. M.E. van Baar, J. Dekker, J.A. Lemmens, R.A. Oostendorp and J.W. Bijlsma, Pain and disability in patients with osteoarthritis of hip or knee: the relationship with articular, kinesiological, and psychological characteristics, J Rheumatol 25 (1998), 125–133. M.E. van Baar, J. Dekker, R.A. Oostendorp, D. Bijl, T.B. Voorn, J.A. Lemmens and J.W. Bijlsma, The effectiveness of exercise therapy in patients with osteoarthritis of the hip or knee: a randomized clinical trial, J Rheumatol 25 (1998), 2432–2439. K.S. Widler, J.F. Glatthorn, M. Bizzini, F.M. Impellizzeri, U. Munzinger, M. Leunig and N.A. Maffiuletti, Assessment of hip abductor muscle strength. A validity and reliability study, J Bone Joint Surg Am 91 (2009), 2666–2672.