- Email: [email protected]
Reliability testing of the patellofemoral joint reaction force (PFJRF) measurement during double-legged squatting in healthy subjects: A pilot study
Journal of Bodywork & Movement Therapies (2012) 16, 217e223
available at www.sciencedirect.com
journal homepage: www.elsevier.com/jbmt
RELIABILITY STUDY
Reliability testing of the patellofemoral joint reaction force (PFJRF) measurement during double-legged squatting in healthy subjects: A pilot study Javid Mostamand, MSc PT, PhD a,*, Dan L. Bader, DSc b, ¨ Hudson, PhD, MCSP c Zoe a Assistant Professor of Physiotherapy, Department of Physiotherapy, Faculty of Rehabilitation Sciences, Isfahan University of Medical Sciences, PO Box 164, Isfahan 8174673461, Iran b Professor of Medical Engineering, Department of Engineering, Queen Mary University of London, Mile End Road, London, E1 4NS, UK c MSc course leader and Senior Clinical Lecturer, Centre for Sports and Exercise Medicine, Barts and the London Queen Mary’s School of Medicine and Dentistry, Mann Ward, Mile End Hospital, Bancroft Road, London, E1 4DG, UK
Received 18 January 2011; received in revised form 10 March 2011; accepted 13 March 2011
KEYWORDS Patellofemoral pain syndrome; PFJRF reliability test; Double-legged squatting
Summary Introduction: Anterior knee pain or patellofemoral pain syndrome (PFPS) is supposed to be related to patellofemoral joint reaction forces (PFJRF). Measuring these forces may therefore provide reliable evidence for conservative treatments to correct probable malalignment in subjects with PFPS. The aim of the present study was to examine the reliability of PFJRF measurements during double-legged squatting in healthy subjects. Methods: Using a motion analysis system and one forceplate, PFJRF of 10 healthy subjects were assessed during double-legged squatting. Data were collected from superficial markers taped to selected landmarks. This procedure was performed on the right knees, at three different knee flexion angles of 30, 45 and 60 during three separate double-legged squats. Subjects were then requested to repeat this test procedure on two separate test sessions at different occasions. The PFJRF was calculated using a biomechanical model of the patellofemoral joint. Results: The data reveal an increase in PFJRF values (from mean, SD of 425.2, 35.5N to 1075.4, 70.1N)with an increase in the tibiofemoral joint angle during double-legged squatting. The CV (coefficient of variation) values during within and between session tests, revealed the high repeatability and reproducibility of PFJRF measurements, while the ICC (intra class correlation coefficient) values showed the low reliability of these measurements.
* Corresponding author. Tel.: þ98 (0) 311 792 2024; fax: þ98 (0)311 6687270. E-mail addresses: [email protected] (J. Mostamand), [email protected] (D.L. Bader), [email protected] (Z. Hudson). 1360-8592/$ - see front matter ª 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbmt.2011.03.003
218
J. Mostamand et al. Conclusion: The low reliability of PFJRF measurements suggests that the PFJRF measurement during double-legged squatting should be performed with caution with improving the method of kinetic measurement of the patellofemoral joint in healthy subjects. ª 2011 Elsevier Ltd. All rights reserved.
Introduction
Methods
Subjects with Patellofemoral pain syndrome (PFPS) suffer from pain, which is usually aggravated during activities associated with the flexed knee, such as squatting or kneeling. It is proposed that such activities result in higher magnitudes of patellofemoral joint reaction force (PFJRF) and associated patellofemoral joint stress (Reilly and Martens, 1972; Wallace et al., 2002), although the previous studies confirm abnormal knee joint proprioception in individuals with PFPS (Venessa et al., 2002; Akseki et al., 2008).Indeed, it was reported that increases in the knee flexion angle enhance the magnitude of resultant PFJRF (Fulkerson, 2004).It is also hypothesized that the values of PFJRF will be different in PFPS subjects compared to those associated with subjects with asymptomatic knees. Alterations in patellofemoral joint reaction force (PFJRF) may therefore explain the cause of pain during different functional activities in subjects with PFPS. Conservative treatment (physical therapy) of patellofemoral pain has focused on restoring normal patellar tracking by improving dynamic stability (Hanten and Schulthies, 1990; McConnell, 2002) through quadriceps (especially vastus medialis obliquus Z VMO) retraining or passive modalities (realignment procedures such as tape, brace, stretching) to reduce the symptoms of subjects. The traditional belief that the VMO is selectively affected compared with the other vasti is the origin of focusing on the VMO in the treatment of patellofemoral pain. However, the only support for this hypothesis was presented by Greber et al. (1985) who reported that in subjects with anterior-cruciate deficient knees, the vastus medialis was found to show greater atrophy than the other vasti. In contrast, the study performed by Spencer et al. (1984) has refuted the concept of selective VMO inhibition. Although it appeared that the vastus medialis was affected to a greater extent (the VMO was not isolated), there was no statistically significant difference to support VMO inhibition, selectively. Despite different views about treatment, it is hypothesized that reducing pain following application of both active and passive procedures of treatment is related to altering PFJRF in PFPS subjects. Measuring PFJRF may therefore reveal whether the pain relief is associated with altering this variable during functional activities. It was previously shown that the reliability of PFJRF measurement was high enough during single leg squatting (Mostamand et al., 2010). However, as far as the present authors know the reliability of PFJRF measurement during different activities has not yet been reported. Accordingly, the aim of the present study is to examine the reliability of PFJRF during double-legged squatting in healthy subjects. If the results show that the reliability of PFJRF measurement is high enough during various physical activities in healthy subjects, the future studies on the PFJRF values would be much more reliable in the symptomatics.
Subjects The present study was approved by the East London and City Research Ethics Committee before recruiting subjects. Written informed consent was provided by each subject. The study was designed to examine the PFJRF values and the reliability of outcome measurement (PFJRF) during double-legged squatting. For this, ten healthy volunteers were recruited into the pilot study. These volunteers had no any traumatic, inflammatory or infectious pathology in their lower extremity. Subjects with any history of surgery in their knees or dislocation or subluxation in their patellofemoral joints were also excluded from the study. Additionally, age more than 40 years was one of the exclusion criteria to ensure subjects had no signs of secondary osteoarthritis (Crossley et al., 2002). The subjects, selected from students studying at Queen Mary University of London, had no previous history of disorders in either the lower extremities or the spinal column. They were also required to perform a double-legged squat on repeated occasions.
Instrumentation Using a 2 camera (DCR-VX2000E, Sony, Japan) motion analysis system (SIMI Motion-2D&3D Motion Analysis, version 7.0, GmbH, Germany) and one forceplate (Kistler, 2812A-3, version 3.20, Switzerland), three dimensional movement and ground reaction force data of the subjects were recorded. Data were collected from superficial reflective markers taped to bony landmarks (Wallace et al., 2002). The five landmarks were the second metatarsal head, lateral malleolous, lateral aspect of the shank, lateral epicondyle of femur and lateral aspect of the thigh, as illustrated in Fig. 1.
Test procedure Before starting the main tests, all subjects were trained how to perform double-legged squat on their legs on the floor according to the required degrees of knee flexion, using verbal feedbacks (zero to approximately 75 of knee flexion). To control any trunk forward flexion or deviation, they were asked to keep their feet in full contact with the floor during double-legged squatting, while verbal feedback was used to encourage subjects to hold their trunks in a vertical position. After that, each subject was instructed to stand with their right leg on the force plate, as indicated in Fig. 1, with the other leg outside the force plate. They were then requested to execute three separate double-legged squats from a neutral position (0 of knee flexion) to a depth of approximately 75 of knee flexion (medium squatting),
Reliability testing of the patellofemoral joint reaction
219 Lq Z8:0e5 X 3 0:013X 2 þ 0:28X þ 0:046 where X is the tibiofemoral joint angle. PFJRF was calculated as the product of the quadriceps force (Fq) and a constant (k) as follow: PFJRFZFq k The constant k was estimated for knee joint angle (X ) using the following non-linear equation, based on the curve fitting to the data of van Eijden et al. (1986): 3:8e5 X 2 þ 1:5e3 X þ 0:462 kZ 7:0e7 X 3 þ 1:6e4 X 2 0:016X þ 1
Figure 1 Setup used to measure kinetics during doublelegged squatting.
while maintaining heel contact with the floor. Fig. 1 indicates an individual subject flexing to approximately 30 during the double-legged squatting protocol. Each squatting protocol was limited to about 3 s in duration which was monitored by a stop watch. Each subject was then requested to repeat this test procedure on two separate test sessions at different occasions. Clearly this protocol involved each subject removing these markers at the end of session 1 and having them reattached by the researcher on the subsequent two sessions. Thus each subject was assessed for 9 separate test procedures, incorporating 3 sets of 3 repetitions. When the subjects were recruited in the morning of the test day, the first and second set of tests was performed on the same day, separated by a minimum of at least 6 h. However, the second tests of subjects were performed on the next day, when they were recruited in the afternoon, for their first session of test. All third set of tests were also performed on the day after the first test session.
Data reduction Marker-coordinate and force data were processed by the SIMI motion analysis system. Using this system, the segmental kinematics for the foot, shank and thigh were computed. The inertial properties for the foot, shank and thigh were determined from the subject’s total body weight, segment geometry and anthropometric data (Winter, 1990). Sagittal-plane knee joint angles and net knee moments (Mk) were calculated from the inertial properties, segmental kinematics, and force platform data using inverse dynamics equations (Winter, 1990). The PFJRF was calculated using a biomechanical model of the PFJ (Salem and Powers, 2001). Based on the model, quadriceps muscle force (Fq) was calculated as the net knee moment (Mk) divided by the moment arm for the quadriceps (Lq). Fq ZMk =Lq The moment arm was estimated using the following nonlinear equation, based on the curve fitting to the data of van Eijden et al. (1987):
For each test, kinetic data (Mk, Fq, and PFJRF) were averaged through the 3 repetitions of double-legged squatting.
Data analysis All data were analyzed (SPSS-version 13) during the eccentric phase of squatting at 30, 45 and 60 of knee flexion. However, data for the reliability tests were confined to 30 of knee flexion. This angle was selected as it was thought to be most appropriate for comparative purposes with biomechanical variables associated with subjects with PFPS (Ernest et al., 1999). The ShapiroeWilk test was applied to all data sets (3 different sessions) of PFJRF measurements to test for normality. All data sets were found to be normally distributed and hence parametric statistics were used. Using the ANOVA test, the mean differences of PFJRF measurements during three different test sessions with 95% CI were calculated. From the mean and standard deviation of each data set, the coefficient of variation (CV) (Portney and Watkins, 2000) was calculated to describe the variability of the PFJRF measurements, both within and between each of the test sessions. Random two-way intra class correlation coefficients for a single measure (ICC type 2, 1) (Portney and Watkins, 2000) were also used to examine whether the corresponding values of PFJRF exhibited significant correlation, both within (R1, R2 & R3) and between sessions (S1, S2 & S3). This coefficient represented the reliability of the paired measurements (R1eR2, R1eR3, R2eR3 & S1eS2, S1eS3, S2eS3). In addition, two alternative statistics were used to establish the relevance and random error of each paired measurements namely, 1. the within and between session least significant difference (LSD) values of PFJRF for the three different test sessions. 2. the within and between session standard error of measurement (SEM) values of PFJRF for the three different test sessions(Portney and Watkins, 2000). Finally, using BlandeAltman plot (Bland and Altman, 1986), the difference scores of each paired between session measurements against the mean values was
220
J. Mostamand et al. in this measurement was considered adequate for normal subjects.
Table 1 Summarized data of three kinetic variables measured on a group of ten healthy subjects during the eccentric phase of double-legged squatting at different knee flexion angles. Data represents mean, SD. Knee extensor Quadriceps moment (N m) force (N) Knee 30 24.3, 2.0 flexion 45 29.3, 2.8 angle 60 39.8, 3.5
Intra class correlation coefficient (ICC) values
PFJRF (N)
540.6, 45.2 425.2, 35.5 750.3, 60.3 690.4, 50.6 1205.2, 85.2 1075.4, 70.1
calculated to show whether each one of these three measurements agrees sufficiently well with the other one.
Results Healthy volunteers (5 men and 5 women) with age (mean, SD) of 29, 6 y, weight of 75, 11 kg and height of 173, 9 cm were recruited into the pilot study. Summarized kinetic data of these subjects during eccentric phase of double-legged squatting at 30, 45 and 60 of the knee flexion are shown in Table 1.The data, relating to the right knees reveals an increase in all knee extensor moment, quadriceps force and PFJRF values with an increase in the tibiofemoral joint angle during doublelegged squatting. ANOVA test revealed that, there were significant differences between the PFJRF mean values for three different knee flexion angles of 30, 45 and 60 (p < 0.05).
Coefficient of variation (CV) values Values of the Coefficient of Variation (CV), both between and within sessions, which were obtained for the PFJRF measurements at 30 of knee flexion are indicated in Table 2. According to this table, the CV values for three between session repetitions varied between 0.7% and 10.9% in these ten healthy subjects denoted S110. It is also evident that for three within session repetitions, the variability varied between 0.6% and 16.9% for the same subjects. In 28 out of 30 cases and 26 out of 30 cases, the variability was less than 10% for the between session and within session, respectively. This value has previously been reported to indicate low scattered variability (Von Eisenhart- Rothe et al., 2004) and thus both reproducibility and repeatability
An analysis of the data using ICC, revealed very poor intratester reliability (0.17, 95% CI to0.34e0.62) for single measures of the PFJRF during first and second repetitions of double-legged squatting, while there were poor intratester reliability for single measures of PFJRF during repetitions 1, 3 (0.41, 95% CI to 0.10e0.75) and 2, 3 (0.47, 95% CI to 0.03e0.78) (Portney and Watkins, 2000). To find the reliability of paired measurements of PFJRF between sessions, the same statistic was used. An analysis of data revealed very poor intratester reliability for single measures of PFJRF during sessions 1, 2 (0.05, 95% CI to 0.53e0.45), 1, 3 (0.02, 95% CI to 0.47e0.51) and 2, 3 (0.27, 95% CI to 0.25e0.68).
Least significant difference (LSD) values The test-retest results of intratester variability of 3 sets of tests (within and between session results) are shown in Tables 3 and 4. Using t value at the 5% significance level, which is 2.77, the LSD values of within session repetitions (Cleophas et al., 2006), were between 1.6N and 4.9N (Table 3). According to the LSD results, there was no significant difference between the mean values for PFJRF measurements during different repetitions (p > 0.05), rejecting the effect of random chance in obtaining similar results during these repetitions. The LSD values for between session single measures were between 1.5 and 11.8 N (Table 4) rejecting the effect of random chance in obtaining similar results during different sessions.
Standard error of measurement (SEM) values The SEM values of PFJRF during different test repetitions were between 41.9 and 60.6N (Table 3). These values indicated that the repeated measures (within session repetitions) fall between 2 SEM of initial measurement. The SEM values of same variable during different test sessions were also between 63.3 and 80.0N (Table 4), revealing that the repeated measures (between session repetitions) lie between 2 SEM of initial measurement.
Table 2 Variabilityof the PFJRF measurements on a group of ten healthy subjects (S) during the eccentric phase of doublelegged squatting at 30 of the knee flexion.
CV of the PFJRF for BSR (%)
CV of the PFJRF for WSR (%)
1 2 3 1 2 3
S1
S2
S3
S4
S5
S6
S7
S8
S9
S10
4.4 6.1 2.9 7.7 7.2 6.4
5.8 4.5 9.1 7.6 6.1 8.4
5.8 9.5 10.9 13.7 13.0 16.9
8.4 6.8 5.3 10.2 0.6 9.8
2.7 1.1 10.3 12.6 4.5 7.6
3.2 4.9 4.7 7.4 4.0 3.8
6.1 6.5 5.2 7.4 8.4 6.6
2.9 4.3 2.8 5.5 8.2 6.1
5.7 1.3 1.5 6.2 7.1 5.3
4.1 5.0 0.7 5.9 9.1 5.5
PFJRF Z patellofemoral joint reaction force; BSR Z between session repetition; WSR Z within session repetition; CV Z coefficient of variation.
Reliability testing of the patellofemoral joint reaction Table 3
221
Variability of the PFJRF: Test- retest by 1 observer during three sets of three repetitions. PFJRF
Mean SEM LSD r
PFJRF
PFJRF
R1
R2
R1
R3
R2
R3
430.1
423.2 60.6 4.9 0.30
430.2
426.0 52.2 1.6 0.60
423.4
427.3 41.9 3.3 0.65
PFJRF Z patellofemoral joint reaction force; R Z repetitions; SEM Z standard error of paired measurement; LSD Z least significant difference within measures; r Z reliability coefficient of paired repetitions. All values are in Newtons; n Z 10, t Z 2.77 at 5% level of significance.
Indeed, these values demonstrated that the differences between repeated measures were not clinically relevant. According to the BlandeAltman plots (1986) for three paired between session measurements (Figs. 2e4), the high inconstancy of differences revealed that none of these three PFJRF measurements agree sufficiently well with the other one.
Discussion The results obtained from the present study indicated that PFJRF increased monotonically with increasing knee flexion angle (from 0 to 60 ). The monotonic increase in PFJRF can be attributed to the increased tibiofemoral joint angle. By increasing the knee flexion angle, the angle between the patellar tendon force and the quadriceps tendon force becomes more acute. This acute angle increases the magnitude of the PFJRF, which is the resultant of these forces (Reilly and Martens, 1972). However, the PFJRF also depends on the quadriceps muscle force (Reilly and Martens, 1972). With increasing knee flexion, the centre of gravity shifts farther away from the centre of rotation, with a consequent increase in the flexion moments, which needs to be counterbalanced by the increased quadriceps muscle force. Increasing the quadriceps muscle force may result in the increased PFJRF, (Reilly and Martens, 1972; Hungerford and Barry, 1979). As the reliability of PFJRF measurement during single leg squatting in healthy subjects had previously been reported (Mostamand et al., 2010), the current study was designed to compare the reliability of PFJRF measurements during double-legged squatting with those associated with single leg squatting.
Table 4
Variability of the PFJRF: Test- retest by 1 observer during three different test sessions. PFJRF
Mean SEM LSD r
In our previous study (Mostamand et al., 2010), using CV, ICC, LSD and SEM values, it was found that the reliability of PFJRF measurements was high during single leg squatting. The authors concluded that the future studies could rely on PFJRF measurements during this activity. However, the results in the present study are different from the previous one. The ICC values of within session measurements of PFJRF during double-legged squatting showed poor intratester reliability. The values of between session measurements also showed that there was very poor intratester reliability for these measurements during double-legged squatting. One probable reason of low reliability of both within and between session PFJRF measurements may relate to inequality of body weight distribution of subjects on both force plate and ground surface during double-legged squatting in the different repetitions and test times. Although the volunteers were asked to weight bear equally on both legs during the three different repetitions and sessions, the low ICC values of PFJRF measurements might be attributed to inability of subjects to distribute their weight equally on the both legs during double-legged squatting. The low reliability of PFJRF measurements during double-legged squatting reveals that these measurements are not necessarily the same during different functional activities. Clinically, different values of reliability during both single and double-legged squatting indicate that the researchers might need to improve the method of kinetic measurements of the patellofemoral joint. For example, they might require introducing a way to isolate the forces applied on both lower extremities during double-legged squatting. Indeed, it might be necessary to prevent translating the forces applied on one leg to another one during this activity. Recruiting such a method means that the
PFJRF
PFJRF
S1
S2
S1
S3
S2
S3
432.5
418.6 80 11.8 0.10
432.2
428.3 63.3 1.5 0.05
418.2
429.5 66.7 10.2 0.45
S Z sessions; SEM Z standard error of paired measurement; LSD Z least significant difference between session measurements; r Z reliability coefficient of paired repetitions. All values are in Newtons; n Z 10, t Z 2.77 at 5% level of significance.
222
Figure 2 BlandeAltman plot for paired PFJRF measurement of S1eS2 during double-legged squatting. S Z Session.
circumstances of reliability measurements would be similar to those associated with single leg squatting, as previously reported by the present authors (Mostamand et al., 2010). Although the ICC values demonstrated that the reliability of the PFJRF measurement during double-legged squatting was low, the CV values conversely revealed that the reliability of these measurements (both repeatability and reproducibility) was very high during this activity (majority of the values less than 10% during within and between session tests). One possible explanation for this contradiction may be related to different statistical calculation of reliability tests employed in the studies. The ICC tests use the single or average values of entire repetition of both within and between measurements of subjects which strongly depend on adequate number of subjects (Morrow and Jackson, 1993), while the CV values reveal the percentage of individual variability of within and between measurements in each subject. Indeed, using the ICC values the mean values of entire subjects with different variability undergo the reliability test, while the CV values individually explain the percentage of variability.
J. Mostamand et al.
Figure 4 BlandeAltman plot for paired PFJRF measurement of S2eS3 during double-legged squatting. S Z Session.
The multiple comparison tests of PFJRF measurements in the present study revealed that the measured differences of all paired mean values (both within and between test sessions) were below the level of LSD values. This indicated that there was insufficient evidence to conclude test-retest values are different, revealing accuracy of each paired measurements during double-legged squatting. The relatively low values of within and between session SEM during the present study, revealed that the random error of measurements were low, showing high precision of the PFJRF measurements during the similar activity. The precision of a measurement system, also called reproducibility or repeatability, is the degree to which repeated measurements under unchanged conditions show the same results (Taylor, 1999). Although the SEM is based on the characteristics of the normal curve and therefore a meaningful estimate can usually be made with a large sample of scores (Portney and Watkins, 2000), this statistic was used to express how far the standard deviation of errors reflects the reliability of the response, revealing precision of the measurements (Taylor, 1999). In the present study measurements were performed on small number of healthy subjects however, the similar judgment is true for patients, as these values demonstrated that the differences between repeated measures were not clinically relevant. Although the between session SEM values were low, indicating accuracy of the measurement method, the difference-vs- mean plots (Figs. 2e4) revealed that the magnitudes of the differences were inconstant throughout the range of measurement. It confirms that the PFJRF measurements were not reliable during double-legged squatting, as ICC values showed in the present study.
Conclusion
Figure 3 BlandeAltman plot for paired PFJRF measurement of S1eS3 during double-legged squatting. S Z Session.
The low reliability of PFJRF measurements suggests that these measurements during double-legged squatting should be performed with caution whilst improving the method of kinetic measurement of the patellofemoral joint in healthy subjects.
Reliability testing of the patellofemoral joint reaction
Conflict of interest statement We confirm that authors have no conflict of interests regarding this paper.
Acknowledgement This article was provided as a part of the study leading to the degree of PhD, which was financially supported by Isfahan University of Medical Sciences and Ministry of Health and Medical Education of Islamic Republic of Iran.
Appendix Abbreviations used in the present study: ANOVA test Analysis of variance test CV Coefficient of variation Fq Quadriceps muscle force ICC Intra class correlation coefficient Lq Quadriceps moment arm LSD Least significant difference Mk Knee moment PFJRF Patellofemoral Joint Reaction Force PFPS Patellofemoral pain syndrome SEM Standard error of measurement VMO Vastus medialis obliquus
References Akseki, D., Akkaya, G., Erduran, M., et al., 2008. Proprioception of the knee joint in patellofemoral pain syndrome. Acta Orthopaedica et Traumatologica Turcica 42 (5), 316e321. Bland, J.M., Altman, D.G., 1986. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet i, 307e310. Cleophas, T.J., Zwinderman, A.H., Cleophas, T.J., 2006. Statistics Applied to Clinical Trials, third ed. Springer, Netherlands. Crossley, K., Bennell, K., Green, S., Cowan, S., McConnell, J., 2002. Physical therapy for patellofemoral pain: a randomized, double blinded, placebo-controlled trial. The American Journal of Sports Medicine 30 (6), 857e865. Ernest, G.P., Kawaguchi, J., Saliba, E., 1999. Effect of patellar taping on knee kinetic of patients with patellofemoral pain syndrome. Journal of Orthopaedic and Sports Physical Therapy 29 (11), 661e667. Fulkerson, J.P., 2004. Disorders of the Patellofemoral Joint, fourth ed. Williams & Wilkins, Baltimore [MD]. Greber, C., Hoppeler, H., Claassen, H., et al., 1985. The lowerextremity musculature in chronic symptomatic instability of the
223 anterior cruciate ligament. The Journal of Bone and Joint Surgery 67A, 1034e1043. Hanten, W.P., Schulthies, S.S., 1990. Exercise effect on electromyographic activity of the vastus medialis oblique and the vastus lateralis muscles. Physical Therapy 70, 561e565. Hungerford, D.S., Barry, M., 1979. Biomechanics of the patellofemoral joint. Clinical Orthopaedics and Related Research 144, 9e15. McConnell, J., 2002. The physical therapist’s approach to patellofemoral disorders. Clinics in Sports Medicine 21, 363e387. Morrow Jr., J.R., Jackson, A.W., 1993. How “significant” is your reliability? Research Quarterly for Exercise and Sport 64 (3), 352e355. Mostamand, J., Bader, D.L., Hudson, Z., 2010. Reliability testing of the patellofemoral joint reaction force (PFJRF) measurement in taped and untaped patellofemoral conditions during single leg squatting: a pilot study. Journal of Bodywork & Movement Therapies 14 (4), 275e281. doi:10.1016/j.jbmt.2010.12.004. Portney, L.G., Watkins, M.P., 2000. Foundations of Clinical Research e Applications to Practice, second ed. Prentice Hall. Reilly, D.T., Martens, M., 1972. Experimental analysis of the quadriceps muscle force and patellofemoral joint reaction forces for various activities. Acta Orthopedia Scandinavia 43 (2), 126e137. Salem, G.J., Powers, C.M., 2001. Patellofemoral joint kinetics during squatting in collegiate women athletes. Clinical Biomechanics 16 (5), 424e430. Spencer, J.D., Hayes, K.C., Alexander, I.J., 1984. Knee joint effusion and quadriceps reflex inhibition in man. Archives of Physical Medicine and Rehabilitation 65, 171e177. Taylor, J.R., 1999. An Introduction to Error Analysis: The Study of Uncertainties in Physical Measurements. University Science Books, Sausalito, Calif. van Eijden, T.M.G.J., Kouwenhoven, E., Verburg, J., Weijs, W.S., 1986. A mathematical model of the patellofemoral joint. Journal of Biomechanics 19 (3), 219e229. van Eijden, T.M.G.J., Weijs, W.A., Kouwenhoven, E., et al., 1987. Forces acting on the patella during maximal voluntary contraction of the quadriceps femoris muscle at different knee flexion/ extension angles. Acta Anatomica 129, 310e314. Venessa, B., Bennell, K., Stillman, B., Cowan, S., et al., 2002. Abnormal knee joint position sense in individuals with patellofemoral pain syndrome. Journal of Orthopaedic Research 20, 208e214. Von Eisenhart- Rothe, R., Siebert, M., Bringmann, C., et al., 2004. A new in vivo technique for determination of 3D kinematics and contact areas of the patello-femoral and tibio-femoral joint. Journal of Biomechanics 37 (6), 927e934. Wallace, D.A., Salem, G.J., Salinas, R., Powers, C.M., 2002. Patellofemoral joint kinetics while squatting with and without an external load. Journal of Orthopedic Sports Physical Therapy 32 (4), 141e148. Winter, D.A., 1990. Biomechanics and Motor Control of Human Movement, second ed. A Wiley-Interscience Publication, New York NY.
available at www.sciencedirect.com
journal homepage: www.elsevier.com/jbmt
RELIABILITY STUDY
Reliability testing of the patellofemoral joint reaction force (PFJRF) measurement during double-legged squatting in healthy subjects: A pilot study Javid Mostamand, MSc PT, PhD a,*, Dan L. Bader, DSc b, ¨ Hudson, PhD, MCSP c Zoe a Assistant Professor of Physiotherapy, Department of Physiotherapy, Faculty of Rehabilitation Sciences, Isfahan University of Medical Sciences, PO Box 164, Isfahan 8174673461, Iran b Professor of Medical Engineering, Department of Engineering, Queen Mary University of London, Mile End Road, London, E1 4NS, UK c MSc course leader and Senior Clinical Lecturer, Centre for Sports and Exercise Medicine, Barts and the London Queen Mary’s School of Medicine and Dentistry, Mann Ward, Mile End Hospital, Bancroft Road, London, E1 4DG, UK
Received 18 January 2011; received in revised form 10 March 2011; accepted 13 March 2011
KEYWORDS Patellofemoral pain syndrome; PFJRF reliability test; Double-legged squatting
Summary Introduction: Anterior knee pain or patellofemoral pain syndrome (PFPS) is supposed to be related to patellofemoral joint reaction forces (PFJRF). Measuring these forces may therefore provide reliable evidence for conservative treatments to correct probable malalignment in subjects with PFPS. The aim of the present study was to examine the reliability of PFJRF measurements during double-legged squatting in healthy subjects. Methods: Using a motion analysis system and one forceplate, PFJRF of 10 healthy subjects were assessed during double-legged squatting. Data were collected from superficial markers taped to selected landmarks. This procedure was performed on the right knees, at three different knee flexion angles of 30, 45 and 60 during three separate double-legged squats. Subjects were then requested to repeat this test procedure on two separate test sessions at different occasions. The PFJRF was calculated using a biomechanical model of the patellofemoral joint. Results: The data reveal an increase in PFJRF values (from mean, SD of 425.2, 35.5N to 1075.4, 70.1N)with an increase in the tibiofemoral joint angle during double-legged squatting. The CV (coefficient of variation) values during within and between session tests, revealed the high repeatability and reproducibility of PFJRF measurements, while the ICC (intra class correlation coefficient) values showed the low reliability of these measurements.
* Corresponding author. Tel.: þ98 (0) 311 792 2024; fax: þ98 (0)311 6687270. E-mail addresses: [email protected] (J. Mostamand), [email protected] (D.L. Bader), [email protected] (Z. Hudson). 1360-8592/$ - see front matter ª 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbmt.2011.03.003
218
J. Mostamand et al. Conclusion: The low reliability of PFJRF measurements suggests that the PFJRF measurement during double-legged squatting should be performed with caution with improving the method of kinetic measurement of the patellofemoral joint in healthy subjects. ª 2011 Elsevier Ltd. All rights reserved.
Introduction
Methods
Subjects with Patellofemoral pain syndrome (PFPS) suffer from pain, which is usually aggravated during activities associated with the flexed knee, such as squatting or kneeling. It is proposed that such activities result in higher magnitudes of patellofemoral joint reaction force (PFJRF) and associated patellofemoral joint stress (Reilly and Martens, 1972; Wallace et al., 2002), although the previous studies confirm abnormal knee joint proprioception in individuals with PFPS (Venessa et al., 2002; Akseki et al., 2008).Indeed, it was reported that increases in the knee flexion angle enhance the magnitude of resultant PFJRF (Fulkerson, 2004).It is also hypothesized that the values of PFJRF will be different in PFPS subjects compared to those associated with subjects with asymptomatic knees. Alterations in patellofemoral joint reaction force (PFJRF) may therefore explain the cause of pain during different functional activities in subjects with PFPS. Conservative treatment (physical therapy) of patellofemoral pain has focused on restoring normal patellar tracking by improving dynamic stability (Hanten and Schulthies, 1990; McConnell, 2002) through quadriceps (especially vastus medialis obliquus Z VMO) retraining or passive modalities (realignment procedures such as tape, brace, stretching) to reduce the symptoms of subjects. The traditional belief that the VMO is selectively affected compared with the other vasti is the origin of focusing on the VMO in the treatment of patellofemoral pain. However, the only support for this hypothesis was presented by Greber et al. (1985) who reported that in subjects with anterior-cruciate deficient knees, the vastus medialis was found to show greater atrophy than the other vasti. In contrast, the study performed by Spencer et al. (1984) has refuted the concept of selective VMO inhibition. Although it appeared that the vastus medialis was affected to a greater extent (the VMO was not isolated), there was no statistically significant difference to support VMO inhibition, selectively. Despite different views about treatment, it is hypothesized that reducing pain following application of both active and passive procedures of treatment is related to altering PFJRF in PFPS subjects. Measuring PFJRF may therefore reveal whether the pain relief is associated with altering this variable during functional activities. It was previously shown that the reliability of PFJRF measurement was high enough during single leg squatting (Mostamand et al., 2010). However, as far as the present authors know the reliability of PFJRF measurement during different activities has not yet been reported. Accordingly, the aim of the present study is to examine the reliability of PFJRF during double-legged squatting in healthy subjects. If the results show that the reliability of PFJRF measurement is high enough during various physical activities in healthy subjects, the future studies on the PFJRF values would be much more reliable in the symptomatics.
Subjects The present study was approved by the East London and City Research Ethics Committee before recruiting subjects. Written informed consent was provided by each subject. The study was designed to examine the PFJRF values and the reliability of outcome measurement (PFJRF) during double-legged squatting. For this, ten healthy volunteers were recruited into the pilot study. These volunteers had no any traumatic, inflammatory or infectious pathology in their lower extremity. Subjects with any history of surgery in their knees or dislocation or subluxation in their patellofemoral joints were also excluded from the study. Additionally, age more than 40 years was one of the exclusion criteria to ensure subjects had no signs of secondary osteoarthritis (Crossley et al., 2002). The subjects, selected from students studying at Queen Mary University of London, had no previous history of disorders in either the lower extremities or the spinal column. They were also required to perform a double-legged squat on repeated occasions.
Instrumentation Using a 2 camera (DCR-VX2000E, Sony, Japan) motion analysis system (SIMI Motion-2D&3D Motion Analysis, version 7.0, GmbH, Germany) and one forceplate (Kistler, 2812A-3, version 3.20, Switzerland), three dimensional movement and ground reaction force data of the subjects were recorded. Data were collected from superficial reflective markers taped to bony landmarks (Wallace et al., 2002). The five landmarks were the second metatarsal head, lateral malleolous, lateral aspect of the shank, lateral epicondyle of femur and lateral aspect of the thigh, as illustrated in Fig. 1.
Test procedure Before starting the main tests, all subjects were trained how to perform double-legged squat on their legs on the floor according to the required degrees of knee flexion, using verbal feedbacks (zero to approximately 75 of knee flexion). To control any trunk forward flexion or deviation, they were asked to keep their feet in full contact with the floor during double-legged squatting, while verbal feedback was used to encourage subjects to hold their trunks in a vertical position. After that, each subject was instructed to stand with their right leg on the force plate, as indicated in Fig. 1, with the other leg outside the force plate. They were then requested to execute three separate double-legged squats from a neutral position (0 of knee flexion) to a depth of approximately 75 of knee flexion (medium squatting),
Reliability testing of the patellofemoral joint reaction
219 Lq Z8:0e5 X 3 0:013X 2 þ 0:28X þ 0:046 where X is the tibiofemoral joint angle. PFJRF was calculated as the product of the quadriceps force (Fq) and a constant (k) as follow: PFJRFZFq k The constant k was estimated for knee joint angle (X ) using the following non-linear equation, based on the curve fitting to the data of van Eijden et al. (1986): 3:8e5 X 2 þ 1:5e3 X þ 0:462 kZ 7:0e7 X 3 þ 1:6e4 X 2 0:016X þ 1
Figure 1 Setup used to measure kinetics during doublelegged squatting.
while maintaining heel contact with the floor. Fig. 1 indicates an individual subject flexing to approximately 30 during the double-legged squatting protocol. Each squatting protocol was limited to about 3 s in duration which was monitored by a stop watch. Each subject was then requested to repeat this test procedure on two separate test sessions at different occasions. Clearly this protocol involved each subject removing these markers at the end of session 1 and having them reattached by the researcher on the subsequent two sessions. Thus each subject was assessed for 9 separate test procedures, incorporating 3 sets of 3 repetitions. When the subjects were recruited in the morning of the test day, the first and second set of tests was performed on the same day, separated by a minimum of at least 6 h. However, the second tests of subjects were performed on the next day, when they were recruited in the afternoon, for their first session of test. All third set of tests were also performed on the day after the first test session.
Data reduction Marker-coordinate and force data were processed by the SIMI motion analysis system. Using this system, the segmental kinematics for the foot, shank and thigh were computed. The inertial properties for the foot, shank and thigh were determined from the subject’s total body weight, segment geometry and anthropometric data (Winter, 1990). Sagittal-plane knee joint angles and net knee moments (Mk) were calculated from the inertial properties, segmental kinematics, and force platform data using inverse dynamics equations (Winter, 1990). The PFJRF was calculated using a biomechanical model of the PFJ (Salem and Powers, 2001). Based on the model, quadriceps muscle force (Fq) was calculated as the net knee moment (Mk) divided by the moment arm for the quadriceps (Lq). Fq ZMk =Lq The moment arm was estimated using the following nonlinear equation, based on the curve fitting to the data of van Eijden et al. (1987):
For each test, kinetic data (Mk, Fq, and PFJRF) were averaged through the 3 repetitions of double-legged squatting.
Data analysis All data were analyzed (SPSS-version 13) during the eccentric phase of squatting at 30, 45 and 60 of knee flexion. However, data for the reliability tests were confined to 30 of knee flexion. This angle was selected as it was thought to be most appropriate for comparative purposes with biomechanical variables associated with subjects with PFPS (Ernest et al., 1999). The ShapiroeWilk test was applied to all data sets (3 different sessions) of PFJRF measurements to test for normality. All data sets were found to be normally distributed and hence parametric statistics were used. Using the ANOVA test, the mean differences of PFJRF measurements during three different test sessions with 95% CI were calculated. From the mean and standard deviation of each data set, the coefficient of variation (CV) (Portney and Watkins, 2000) was calculated to describe the variability of the PFJRF measurements, both within and between each of the test sessions. Random two-way intra class correlation coefficients for a single measure (ICC type 2, 1) (Portney and Watkins, 2000) were also used to examine whether the corresponding values of PFJRF exhibited significant correlation, both within (R1, R2 & R3) and between sessions (S1, S2 & S3). This coefficient represented the reliability of the paired measurements (R1eR2, R1eR3, R2eR3 & S1eS2, S1eS3, S2eS3). In addition, two alternative statistics were used to establish the relevance and random error of each paired measurements namely, 1. the within and between session least significant difference (LSD) values of PFJRF for the three different test sessions. 2. the within and between session standard error of measurement (SEM) values of PFJRF for the three different test sessions(Portney and Watkins, 2000). Finally, using BlandeAltman plot (Bland and Altman, 1986), the difference scores of each paired between session measurements against the mean values was
220
J. Mostamand et al. in this measurement was considered adequate for normal subjects.
Table 1 Summarized data of three kinetic variables measured on a group of ten healthy subjects during the eccentric phase of double-legged squatting at different knee flexion angles. Data represents mean, SD. Knee extensor Quadriceps moment (N m) force (N) Knee 30 24.3, 2.0 flexion 45 29.3, 2.8 angle 60 39.8, 3.5
Intra class correlation coefficient (ICC) values
PFJRF (N)
540.6, 45.2 425.2, 35.5 750.3, 60.3 690.4, 50.6 1205.2, 85.2 1075.4, 70.1
calculated to show whether each one of these three measurements agrees sufficiently well with the other one.
Results Healthy volunteers (5 men and 5 women) with age (mean, SD) of 29, 6 y, weight of 75, 11 kg and height of 173, 9 cm were recruited into the pilot study. Summarized kinetic data of these subjects during eccentric phase of double-legged squatting at 30, 45 and 60 of the knee flexion are shown in Table 1.The data, relating to the right knees reveals an increase in all knee extensor moment, quadriceps force and PFJRF values with an increase in the tibiofemoral joint angle during doublelegged squatting. ANOVA test revealed that, there were significant differences between the PFJRF mean values for three different knee flexion angles of 30, 45 and 60 (p < 0.05).
Coefficient of variation (CV) values Values of the Coefficient of Variation (CV), both between and within sessions, which were obtained for the PFJRF measurements at 30 of knee flexion are indicated in Table 2. According to this table, the CV values for three between session repetitions varied between 0.7% and 10.9% in these ten healthy subjects denoted S110. It is also evident that for three within session repetitions, the variability varied between 0.6% and 16.9% for the same subjects. In 28 out of 30 cases and 26 out of 30 cases, the variability was less than 10% for the between session and within session, respectively. This value has previously been reported to indicate low scattered variability (Von Eisenhart- Rothe et al., 2004) and thus both reproducibility and repeatability
An analysis of the data using ICC, revealed very poor intratester reliability (0.17, 95% CI to0.34e0.62) for single measures of the PFJRF during first and second repetitions of double-legged squatting, while there were poor intratester reliability for single measures of PFJRF during repetitions 1, 3 (0.41, 95% CI to 0.10e0.75) and 2, 3 (0.47, 95% CI to 0.03e0.78) (Portney and Watkins, 2000). To find the reliability of paired measurements of PFJRF between sessions, the same statistic was used. An analysis of data revealed very poor intratester reliability for single measures of PFJRF during sessions 1, 2 (0.05, 95% CI to 0.53e0.45), 1, 3 (0.02, 95% CI to 0.47e0.51) and 2, 3 (0.27, 95% CI to 0.25e0.68).
Least significant difference (LSD) values The test-retest results of intratester variability of 3 sets of tests (within and between session results) are shown in Tables 3 and 4. Using t value at the 5% significance level, which is 2.77, the LSD values of within session repetitions (Cleophas et al., 2006), were between 1.6N and 4.9N (Table 3). According to the LSD results, there was no significant difference between the mean values for PFJRF measurements during different repetitions (p > 0.05), rejecting the effect of random chance in obtaining similar results during these repetitions. The LSD values for between session single measures were between 1.5 and 11.8 N (Table 4) rejecting the effect of random chance in obtaining similar results during different sessions.
Standard error of measurement (SEM) values The SEM values of PFJRF during different test repetitions were between 41.9 and 60.6N (Table 3). These values indicated that the repeated measures (within session repetitions) fall between 2 SEM of initial measurement. The SEM values of same variable during different test sessions were also between 63.3 and 80.0N (Table 4), revealing that the repeated measures (between session repetitions) lie between 2 SEM of initial measurement.
Table 2 Variabilityof the PFJRF measurements on a group of ten healthy subjects (S) during the eccentric phase of doublelegged squatting at 30 of the knee flexion.
CV of the PFJRF for BSR (%)
CV of the PFJRF for WSR (%)
1 2 3 1 2 3
S1
S2
S3
S4
S5
S6
S7
S8
S9
S10
4.4 6.1 2.9 7.7 7.2 6.4
5.8 4.5 9.1 7.6 6.1 8.4
5.8 9.5 10.9 13.7 13.0 16.9
8.4 6.8 5.3 10.2 0.6 9.8
2.7 1.1 10.3 12.6 4.5 7.6
3.2 4.9 4.7 7.4 4.0 3.8
6.1 6.5 5.2 7.4 8.4 6.6
2.9 4.3 2.8 5.5 8.2 6.1
5.7 1.3 1.5 6.2 7.1 5.3
4.1 5.0 0.7 5.9 9.1 5.5
PFJRF Z patellofemoral joint reaction force; BSR Z between session repetition; WSR Z within session repetition; CV Z coefficient of variation.
Reliability testing of the patellofemoral joint reaction Table 3
221
Variability of the PFJRF: Test- retest by 1 observer during three sets of three repetitions. PFJRF
Mean SEM LSD r
PFJRF
PFJRF
R1
R2
R1
R3
R2
R3
430.1
423.2 60.6 4.9 0.30
430.2
426.0 52.2 1.6 0.60
423.4
427.3 41.9 3.3 0.65
PFJRF Z patellofemoral joint reaction force; R Z repetitions; SEM Z standard error of paired measurement; LSD Z least significant difference within measures; r Z reliability coefficient of paired repetitions. All values are in Newtons; n Z 10, t Z 2.77 at 5% level of significance.
Indeed, these values demonstrated that the differences between repeated measures were not clinically relevant. According to the BlandeAltman plots (1986) for three paired between session measurements (Figs. 2e4), the high inconstancy of differences revealed that none of these three PFJRF measurements agree sufficiently well with the other one.
Discussion The results obtained from the present study indicated that PFJRF increased monotonically with increasing knee flexion angle (from 0 to 60 ). The monotonic increase in PFJRF can be attributed to the increased tibiofemoral joint angle. By increasing the knee flexion angle, the angle between the patellar tendon force and the quadriceps tendon force becomes more acute. This acute angle increases the magnitude of the PFJRF, which is the resultant of these forces (Reilly and Martens, 1972). However, the PFJRF also depends on the quadriceps muscle force (Reilly and Martens, 1972). With increasing knee flexion, the centre of gravity shifts farther away from the centre of rotation, with a consequent increase in the flexion moments, which needs to be counterbalanced by the increased quadriceps muscle force. Increasing the quadriceps muscle force may result in the increased PFJRF, (Reilly and Martens, 1972; Hungerford and Barry, 1979). As the reliability of PFJRF measurement during single leg squatting in healthy subjects had previously been reported (Mostamand et al., 2010), the current study was designed to compare the reliability of PFJRF measurements during double-legged squatting with those associated with single leg squatting.
Table 4
Variability of the PFJRF: Test- retest by 1 observer during three different test sessions. PFJRF
Mean SEM LSD r
In our previous study (Mostamand et al., 2010), using CV, ICC, LSD and SEM values, it was found that the reliability of PFJRF measurements was high during single leg squatting. The authors concluded that the future studies could rely on PFJRF measurements during this activity. However, the results in the present study are different from the previous one. The ICC values of within session measurements of PFJRF during double-legged squatting showed poor intratester reliability. The values of between session measurements also showed that there was very poor intratester reliability for these measurements during double-legged squatting. One probable reason of low reliability of both within and between session PFJRF measurements may relate to inequality of body weight distribution of subjects on both force plate and ground surface during double-legged squatting in the different repetitions and test times. Although the volunteers were asked to weight bear equally on both legs during the three different repetitions and sessions, the low ICC values of PFJRF measurements might be attributed to inability of subjects to distribute their weight equally on the both legs during double-legged squatting. The low reliability of PFJRF measurements during double-legged squatting reveals that these measurements are not necessarily the same during different functional activities. Clinically, different values of reliability during both single and double-legged squatting indicate that the researchers might need to improve the method of kinetic measurements of the patellofemoral joint. For example, they might require introducing a way to isolate the forces applied on both lower extremities during double-legged squatting. Indeed, it might be necessary to prevent translating the forces applied on one leg to another one during this activity. Recruiting such a method means that the
PFJRF
PFJRF
S1
S2
S1
S3
S2
S3
432.5
418.6 80 11.8 0.10
432.2
428.3 63.3 1.5 0.05
418.2
429.5 66.7 10.2 0.45
S Z sessions; SEM Z standard error of paired measurement; LSD Z least significant difference between session measurements; r Z reliability coefficient of paired repetitions. All values are in Newtons; n Z 10, t Z 2.77 at 5% level of significance.
222
Figure 2 BlandeAltman plot for paired PFJRF measurement of S1eS2 during double-legged squatting. S Z Session.
circumstances of reliability measurements would be similar to those associated with single leg squatting, as previously reported by the present authors (Mostamand et al., 2010). Although the ICC values demonstrated that the reliability of the PFJRF measurement during double-legged squatting was low, the CV values conversely revealed that the reliability of these measurements (both repeatability and reproducibility) was very high during this activity (majority of the values less than 10% during within and between session tests). One possible explanation for this contradiction may be related to different statistical calculation of reliability tests employed in the studies. The ICC tests use the single or average values of entire repetition of both within and between measurements of subjects which strongly depend on adequate number of subjects (Morrow and Jackson, 1993), while the CV values reveal the percentage of individual variability of within and between measurements in each subject. Indeed, using the ICC values the mean values of entire subjects with different variability undergo the reliability test, while the CV values individually explain the percentage of variability.
J. Mostamand et al.
Figure 4 BlandeAltman plot for paired PFJRF measurement of S2eS3 during double-legged squatting. S Z Session.
The multiple comparison tests of PFJRF measurements in the present study revealed that the measured differences of all paired mean values (both within and between test sessions) were below the level of LSD values. This indicated that there was insufficient evidence to conclude test-retest values are different, revealing accuracy of each paired measurements during double-legged squatting. The relatively low values of within and between session SEM during the present study, revealed that the random error of measurements were low, showing high precision of the PFJRF measurements during the similar activity. The precision of a measurement system, also called reproducibility or repeatability, is the degree to which repeated measurements under unchanged conditions show the same results (Taylor, 1999). Although the SEM is based on the characteristics of the normal curve and therefore a meaningful estimate can usually be made with a large sample of scores (Portney and Watkins, 2000), this statistic was used to express how far the standard deviation of errors reflects the reliability of the response, revealing precision of the measurements (Taylor, 1999). In the present study measurements were performed on small number of healthy subjects however, the similar judgment is true for patients, as these values demonstrated that the differences between repeated measures were not clinically relevant. Although the between session SEM values were low, indicating accuracy of the measurement method, the difference-vs- mean plots (Figs. 2e4) revealed that the magnitudes of the differences were inconstant throughout the range of measurement. It confirms that the PFJRF measurements were not reliable during double-legged squatting, as ICC values showed in the present study.
Conclusion
Figure 3 BlandeAltman plot for paired PFJRF measurement of S1eS3 during double-legged squatting. S Z Session.
The low reliability of PFJRF measurements suggests that these measurements during double-legged squatting should be performed with caution whilst improving the method of kinetic measurement of the patellofemoral joint in healthy subjects.
Reliability testing of the patellofemoral joint reaction
Conflict of interest statement We confirm that authors have no conflict of interests regarding this paper.
Acknowledgement This article was provided as a part of the study leading to the degree of PhD, which was financially supported by Isfahan University of Medical Sciences and Ministry of Health and Medical Education of Islamic Republic of Iran.
Appendix Abbreviations used in the present study: ANOVA test Analysis of variance test CV Coefficient of variation Fq Quadriceps muscle force ICC Intra class correlation coefficient Lq Quadriceps moment arm LSD Least significant difference Mk Knee moment PFJRF Patellofemoral Joint Reaction Force PFPS Patellofemoral pain syndrome SEM Standard error of measurement VMO Vastus medialis obliquus
References Akseki, D., Akkaya, G., Erduran, M., et al., 2008. Proprioception of the knee joint in patellofemoral pain syndrome. Acta Orthopaedica et Traumatologica Turcica 42 (5), 316e321. Bland, J.M., Altman, D.G., 1986. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet i, 307e310. Cleophas, T.J., Zwinderman, A.H., Cleophas, T.J., 2006. Statistics Applied to Clinical Trials, third ed. Springer, Netherlands. Crossley, K., Bennell, K., Green, S., Cowan, S., McConnell, J., 2002. Physical therapy for patellofemoral pain: a randomized, double blinded, placebo-controlled trial. The American Journal of Sports Medicine 30 (6), 857e865. Ernest, G.P., Kawaguchi, J., Saliba, E., 1999. Effect of patellar taping on knee kinetic of patients with patellofemoral pain syndrome. Journal of Orthopaedic and Sports Physical Therapy 29 (11), 661e667. Fulkerson, J.P., 2004. Disorders of the Patellofemoral Joint, fourth ed. Williams & Wilkins, Baltimore [MD]. Greber, C., Hoppeler, H., Claassen, H., et al., 1985. The lowerextremity musculature in chronic symptomatic instability of the
223 anterior cruciate ligament. The Journal of Bone and Joint Surgery 67A, 1034e1043. Hanten, W.P., Schulthies, S.S., 1990. Exercise effect on electromyographic activity of the vastus medialis oblique and the vastus lateralis muscles. Physical Therapy 70, 561e565. Hungerford, D.S., Barry, M., 1979. Biomechanics of the patellofemoral joint. Clinical Orthopaedics and Related Research 144, 9e15. McConnell, J., 2002. The physical therapist’s approach to patellofemoral disorders. Clinics in Sports Medicine 21, 363e387. Morrow Jr., J.R., Jackson, A.W., 1993. How “significant” is your reliability? Research Quarterly for Exercise and Sport 64 (3), 352e355. Mostamand, J., Bader, D.L., Hudson, Z., 2010. Reliability testing of the patellofemoral joint reaction force (PFJRF) measurement in taped and untaped patellofemoral conditions during single leg squatting: a pilot study. Journal of Bodywork & Movement Therapies 14 (4), 275e281. doi:10.1016/j.jbmt.2010.12.004. Portney, L.G., Watkins, M.P., 2000. Foundations of Clinical Research e Applications to Practice, second ed. Prentice Hall. Reilly, D.T., Martens, M., 1972. Experimental analysis of the quadriceps muscle force and patellofemoral joint reaction forces for various activities. Acta Orthopedia Scandinavia 43 (2), 126e137. Salem, G.J., Powers, C.M., 2001. Patellofemoral joint kinetics during squatting in collegiate women athletes. Clinical Biomechanics 16 (5), 424e430. Spencer, J.D., Hayes, K.C., Alexander, I.J., 1984. Knee joint effusion and quadriceps reflex inhibition in man. Archives of Physical Medicine and Rehabilitation 65, 171e177. Taylor, J.R., 1999. An Introduction to Error Analysis: The Study of Uncertainties in Physical Measurements. University Science Books, Sausalito, Calif. van Eijden, T.M.G.J., Kouwenhoven, E., Verburg, J., Weijs, W.S., 1986. A mathematical model of the patellofemoral joint. Journal of Biomechanics 19 (3), 219e229. van Eijden, T.M.G.J., Weijs, W.A., Kouwenhoven, E., et al., 1987. Forces acting on the patella during maximal voluntary contraction of the quadriceps femoris muscle at different knee flexion/ extension angles. Acta Anatomica 129, 310e314. Venessa, B., Bennell, K., Stillman, B., Cowan, S., et al., 2002. Abnormal knee joint position sense in individuals with patellofemoral pain syndrome. Journal of Orthopaedic Research 20, 208e214. Von Eisenhart- Rothe, R., Siebert, M., Bringmann, C., et al., 2004. A new in vivo technique for determination of 3D kinematics and contact areas of the patello-femoral and tibio-femoral joint. Journal of Biomechanics 37 (6), 927e934. Wallace, D.A., Salem, G.J., Salinas, R., Powers, C.M., 2002. Patellofemoral joint kinetics while squatting with and without an external load. Journal of Orthopedic Sports Physical Therapy 32 (4), 141e148. Winter, D.A., 1990. Biomechanics and Motor Control of Human Movement, second ed. A Wiley-Interscience Publication, New York NY.