Contribution of altered hip, knee and foot kinematics to dynamic postural impairments in females with patellofemoral pain during stair ascent

THEKNE-02224; No of Pages 6 The Knee xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

The Knee

Contribution of altered hip, knee and foot kinematics to dynamic postural impairments in females with patellofemoral pain during stair ascent☆ Danilo de Oliveira Silva a, Fernando Henrique Magalhães b, Marcella Ferraz Pazzinatto a, Ronaldo Valdir Briani a, Amanda Schenatto Ferreira a, Fernando Amâncio Aragão c, Fábio Mícolis de Azevedo a,⁎ a b c

Physical Therapy Department, School of Science and Technology, University of São Paulo State, Presidente Prudente, Brazil School of Arts, Sciences, and Humanities, University of Sao Paulo, Sao Paulo, Brazil Physical Therapy Department, State University of West Parana, Cascavel, Parana, Brazil

a r t i c l e

i n f o

Article history: Received 7 October 2015 Received in revised form 13 January 2016 Accepted 15 January 2016 Available online xxxx Keywords: Center of pressure Patellofemoral joint Knee injury Regression analysis Postural control

a b s t r a c t Background: Altered hip, knee and foot kinematics have been systematically observed in individuals with patellofemoral pain (PFP). However, less attention has been given to the altered dynamic postural control associated with PFP. Additionally, the relative contribution of kinematic impairments to the postural behavior of subjects with PFP remains an open question that warrants investigation. The aims of this study were: i) to investigate possible differences in hip adduction, rearfoot eversion, knee flexion and displacement area of the center of pressure (COP) in individuals with PFP in comparison to controls during stair ascent; and (ii) to determine which kinematic parameter is the best predictor of the displacement area of the COP measured during the stance phase of the stair ascent. Methods: Twenty-nine females with PFP and 25 asymptomatic pain-free females underwent three-dimensional kinematic and COP analyses during stair ascent. Between-group comparisons were made using independent ttests. Regression models were performed to identify the capability of each kinematic factor in predicting the displacement area of the COP. Results: Reduced knee flexion and displacement area of the COP as well as increased peak hip adduction and peak rearfoot eversion were observed in individuals with PFP as compared to controls. Peak hip adduction was the best predictor of the displacement area of the COP (r2 = 23.4%). Conclusions: The excessive hip adduction was the biggest predictor of the displacement area of the COP. Clinical relevance: Based on our findings, proximally targeted interventions may be of major importance for the functional reestablishment of females with PFP. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Patellofemoral pain (PFP) is a common and costly musculoskeletal disorder characterized by the presence of idiopathic anterior knee pain [1,2], which can severely affect quality of life by limiting the participation in functional activities [3]. PFP is usually observed in the physically active population [1] and accounts for 25 to 40% of all knee injuries observed in sports clinics [1]. Females are 2.23 times more likely to experience PFP than males [2], and the estimated prevalence of PFP among females aged 18 to 35 years is 13% [4]. A commonly cited hypothesis as to the cause of PFP is increased patellofemoral joint (PFJ) stress associated with abnormal lower extremity kinematics [5,6]. In this direction, studies have reported that

☆ This work was developed in the Laboratory of Biomechanics and Motor Control at the University of São Paulo State. ⁎ Corresponding author at: Rua Roberto Simonsen, 305, Presidente Prudente, SP 19060900, Brazil. E-mail address: [email protected] (F.M. de Azevedo).

stair ascent results in more challenging patellofemoral contact mechanics than walking [7,8], thereby being a useful experimental model to reproduce symptoms and abnormal movement patterns associated with PFP. Therefore, to investigate lower extremity mechanics during stair negotiation has been important to clarify the compensatory behavior shown by females with PFP [5]. It is generally agreed that the etiology of PFP is multifactorial and several factors have been proposed in an attempt to explain the pathomechanisms underlying PFP [9]. For instance, a large amount of biomechanical alterations have been observed in individuals with PFP [9], which have been grouped into three mechanistic categories: proximal factors, distal factors and local factors [1]. Studies approaching proximal factors have been focused on understanding how the hip, pelvis, and trunk may contribute to PFP. Local factors' studies have focused on the contribution of PFJ mechanics and surrounding tissues to PFP. Moreover, distal factors' studies are dedicated to the contribution of foot and ankle mechanics to PFP [1]. Currently, special attention has been given to the increased rearfoot eversion and hip adduction as well as reduced knee flexion that have

http://dx.doi.org/10.1016/j.knee.2016.01.014 0968-0160/© 2016 Elsevier B.V. All rights reserved.

Please cite this article as: de Oliveira Silva D, et al, Contribution of altered hip, knee and foot kinematics to dynamic postural impairments in females with patellofemoral pain during stair ascent, Knee (2016), http://dx.doi.org/10.1016/j.knee.2016.01.014

2

D. de Oliveira Silva et al. / The Knee xxx (2016) xxx–xxx

been observed in females with PFP as compared to controls [1,8,10–13]. However, to the best of our knowledge, there is no previous investigation on the specific contribution of such alterations to the dynamic postural impairments observed in subjects with PFP. More specifically, evidences from studies with traditional measurements based on center of pressure (COP) analysis (e.g. 95% ellipse area) have indicated impaired postural control in individuals with PFP during stair negotiation [14,15]. For instance, in a prospective study in which 43% of the participants developed PFP, the alteration of dynamic COP displacement was considered one of the three most important gait-related intrinsic risk factors for PFP development [16]. As proximal, distal and local kinematic alterations previously reported in subjects with PFP during stair ambulation [8,10,17] may contribute to changes in dynamic COP displacement, the relative influence of each of these alterations to PFPassociated postural impairments remains an open question that warrants investigation. Proximally, weakness or delayed onset of hip abductor and hip external rotator muscles are thought to contribute to excessive hip adduction during stair negotiation activities in individuals with PFP [17]. The importance of hip muscles to postural stability was demonstrated experimentally by Gribble and Hertel [18], who showed increased COP excursion velocity during single leg stance after fatigue of the hip muscles. Therefore, it is reasonable to suppose that biomechanical alterations at the hip level may account for the impaired postural performance observed in subjects with PFP. Locally, reduced knee flexion during stair negotiation is a common finding in individuals with PFP [10,19]. Recently, this mechanism was shown to alter vertical ground reaction forces [10], which is directly related to postural stability [20]. Therefore, altered knee kinematics may be associated with impaired postural stability in individuals with PFP. Distally, excessive rearfoot eversion has been reported during stair ascent and walking [8,21] in subjects with PFP, which is suggested to lead to greater PFJ stress due to the coupling between subtalar motion and tibia rotation [22]. A prospective study [16] showed that individuals who developed PFP (as compared to subjects that did not) had greater rearfoot pronation associated with greater pressure on the medial portions of the plantar surface during walking. A similar pattern was reported for subjects with PFP (as compared to healthy controls) during stair negotiation [23]. It is reasonable to speculate that individuals with PFP may demonstrate poorer postural control during dynamic activities due to altered foot motion. Assessment of postural control is of frequent interest to researchers and clinicians as postural steadiness is considered an important factor in functional reestablishment [24,25]. Identifying the kinematic alterations that most closely predict dynamic postural impairments would help clinicians to develop more specific and successful interventions. In this context, the aims of this study were (i) to investigate possible differences in hip adduction, rearfoot eversion, knee flexion and displacement area of the COP in individuals with PFP as compared to controls during stair ascent; and (ii) to determine which one of these kinematic parameters is the best predictor of the displacement area of the COP. Due to the previous literature mentioned above, it was hypothesized greater rearfoot eversion and hip adduction as well as decreased knee flexion in subjects with PFP in comparison to controls. The relative contribution of hip, knee and foot kinematics to dynamic postural behavior cannot be predicted beforehand due to contradictory results that have been reported with regard to the dynamic postural alterations observed in subjects with PFP. 2. Methods 2.1. Participants Fifty-four females aged 18 to 30 years were recruited and divided into two groups: PFP group (PFPG; n = 29) and control group (CG; n = 25). Only females were included due to high prevalence of PFP in

this population [2]. In addition, we assumed that including both sexes could be seen as a confounder because females are reported to exhibit different movement patterns than males [26]. Mean (SD) age, height, mass and physical activity level are presented in Table 1. Physical activity level was evaluated with the self-administered International Physical Activity Questionnaire long form [27]. Participants were recruited via advertisements in gyms, parks and Universities, between June and November 2014. The study was approved by the Local Ethics Committee and each participant gave written informed consent prior to participation. Diagnosis of PFP was completed following consensus from two experienced clinicians (N5 years of experience) and based on definitions used in previous studies [28–30]. The inclusion criteria were: (1) anterior knee pain during at least two of the following activities: prolonged sitting, squatting, kneeling, running, climbing stairs, and jumping; (2) pain during patellar palpation; (3) symptoms of insidious onset and duration of at least one month; and (4) worst pain level in the previous month at least three centimeters on a 10 cm visual analogue scale (VAS). Participants were required to fulfill all four requirements to be included in the PFPG. Subjects allocated in the CG could not present any signs or symptoms of PFP or other musculoskeletal impairments as well as no previous history of lower limb injuries. The presence of the following conditions were carefully screened: events of patellar subluxation, lower limb inflammatory process, patellar tendon tears, meniscus tears, bursitis, ligament tears or the presence of neurological diseases. Those who had undergone knee surgery, received oral steroids, acupuncture or physiotherapy during the preceding six months were excluded from this study. 2.2. Kinematic analysis Data collection included lower limb kinematic evaluation of each participant's symptomatic limb (unilateral symptoms) or most symptomatic limb (bilateral symptoms) during stair ascent. The dominant leg was evaluated in the CG. Motion analysis was collected using a three-dimensional motion analysis system (Vicon Motion Systems Inc.; Denver EUA) combined with four cameras (type Bonita®B10). Data were recorded with a sampling rate of 100 Hz and a resolution of one megapixel. Kinematic analysis was performed using the Oxford Foot Model (OFM) combined with plug-in gait (PIG-SACR), which was previously reported as a valid and reliable method [8,21,31]. Retroreflective markers (9.5 mm) were placed on specific anatomical landmarks by the same investigator, in accordance with the models specifications. The anatomical landmarks are described in detail in Appendix A. 2.3. Dynamic postural stability analysis Center of pressure data were obtained using a force plate (AMTI, OR6, Watertown, MA, USA). The COP signals were sampled at 2000 Hz. According to Rhea et al. [32], the inherent instability of upright posture requires anterior–posterior (AP) and medial-lateral (ML) sway

Table 1 Characteristics of the participants included in both groups. Variable

Age Mass (kg) Height (m) Cadence (steps/min) Physical activity (MET min·week−1) Worst pain level in the previous month (VAS) Pain level during stair ascent task (VAS)

Control group

PFP

p-Value

Mean (SD)

Mean (SD)

22.27 (3.52) 63.45 (6.31) 1.65 (0.04) 81.03(6.27) 4029.53 (595.32)

21.81 (2.69) 65.02 (9.06) 1.65 (0.06) 76.89(6.02) 4432.71 (437.02)

0.487 0.158 0.789 0.191 0.742

0.00 (0.00)

5.78 (1.99)

0.000⁎

0.00 (0.00)

2.02 (1.46)

0.000⁎

⁎ Statistically significant (p b 0.05) values.

Please cite this article as: de Oliveira Silva D, et al, Contribution of altered hip, knee and foot kinematics to dynamic postural impairments in females with patellofemoral pain during stair ascent, Knee (2016), http://dx.doi.org/10.1016/j.knee.2016.01.014

D. de Oliveira Silva et al. / The Knee xxx (2016) xxx–xxx

to be controlled simultaneously. Therefore, it is recommended not to separate the COP trajectory into two independent dimensions (i.e. AP and ML) for data analysis. While different patterns in the AP and ML dimensions may be observed, it is their combination that perhaps better provides a more complete picture of the dynamic postural control processes [33]. Therefore, we chose to quantify the 95% of the total area covered in the AP and ML directions using an ellipse fit to the data, which is a traditional metric that characterizes the magnitude of the resultant vector [32]. 2.4. Procedures Prior to data collection, an acceptable error of 0.08 mm was established for the motion-system calibration. A calibration trial was then carried out with the subjects in a relaxed standing posture, after which the participants performed practice stair ascent trials to allow familiarization with the instrumentation and environment. Subjects were not allowed to use handrails. A seven-step staircase, each step being 18 cm high and 28 cm deep, with a two meter walkway in front of and behind the staircase was used (Figure 1A). Data analyses were performed based on evaluation of the fourth step [8,10,29]. Each participant was asked to climb the staircase at their natural comfortable speed. As demonstrated by Jordan et al. [34], controlling the speed of walking can change the center of pressure signal, thus, the speed of stair ascent was not controlled in the present study. Five successful trials were collected for each participant. The mean value of these five trials was used for data analyses in order to attenuate the influence of speed, intra-subject variability, among other external factors [20,35]. To ensure a natural stair ascent pattern, participants were not aware of the presence of the force plate, which was hidden within the fourth step. The force plate was mechanically coupled to the ground (i.e. independent and uncoupled from the stair structure) (Figure 1B). The starting positions prior to each trial were the same to optimize the likelihood of a successful trial. The investigator was blind as to the group allocation of each participant. 2.5. Data analysis Signals from the motion analysis system were filtered with a fourthorder Butterworth low-pass filter with a cutoff frequency of six Hertz [20] and reconstructed within the Vicon Nexus® software. Heel strike and toe off were determined using vertical ground reaction force signals, which were used to identify the stance phase from which the measurements of interest were computed. The cadence was calculated based on the time of one entire gait cycle: single leg stance between toe-off of the opposite leg from the third step until foot contact on the fifth step. Each event was inspected manually by viewing the animated

3

visualization of the motion data [36]. The OFM and PIG-SACR models were applied to the data of each marker, which were then exported to a custom-developed Excel template for analysis. The variables of interest were cadence and peak angles during stance phase of hip adduction, knee flexion and rearfoot eversion. The 95% of the total area of the COP displacement (displacement area of the COP) were quantified using a customized Matlab program (Matlab 2009, The MathWorks Inc., Natick, MA, USA). The calculation of the dynamic postural stability parameter was previously described in detail [37], which was computed from COP signals recorded during the stance phase of the movement. 2.6. Statistical analysis Statistical analysis was performed using SPSS (version 18.0, SPSS inc., Chicago, Il). Prior to statistical analysis, all variables were assessed for normality and found to be normally distributed based on obtainment of p N 0.05 in the Shapiro–Wilk test. Mean age, height, mass, physical activity level and cadence were compared between groups using independent t-tests. Between group differences in kinematic and COP variables were also evaluated using independent sample t-tests. Effect sizes were calculated according to equations previous described [38], the guidelines for interpreting the Cohen's d are: 0 to 0.40 small effect, 0.41 to 0.70 moderate effect, 0.71 b large effect. A forced entry multiple regression model [39] analysis was carried out to determine which kinematic variable (independent variables) presented the best capability to predict the displacement area of the COP (dependent variable). The variables were inserted in the model separately to distinguish which was more predictive. Effects due to multicollinearity were limited by ensuring the Pearson's correlation coefficients between variables input in the regression model were less than 0.8 [39,40]. The assumption of homogeneity of variance and linearity was verified by qualitative inspection of the regression of standardized residual versus regression of standardized predicted value plot [39]. Overall performance of the final models was evaluated using Nagelkerke's r2 [41], which estimates explained variation of the model. 3. Results There were no significant differences between groups for age, height, mass, physical activity level or cadence (Table 1). As we expect PFP group presented pain level higher than three centimeters on a 10 cm VAS and the CG did not present any pain (Table 1). Peak rearfoot eversion and peak hip adduction were found to be significantly larger for the PFPG as compared to CG by 3.6 and 3.3 degrees, respectively. Peak knee flexion and displacement area of the COP were significantly decreased in the PFP group as compared to the CG by 3.8 degrees and 16.7 cm2, respectively (Table 2). Peak hip adduction was found to significantly predict the displacement area of the COP, explaining 23.4% of the COP area variance. In spite of not presenting significant results, peak knee flexion and rearfoot eversion explained 1.6% and 10.3% of the COP area variance, respectively. All kinematic parameters explained together 35.4% of the displacement area of the COP variance (Table 3).

Figure 1. A) The experimental set up used in data collection, participants were not aware of the presence of the force plate within the fourth step. B) The force plate was mechanically coupled to the ground (i.e. independent and uncoupled from the stair structure).

Please cite this article as: de Oliveira Silva D, et al, Contribution of altered hip, knee and foot kinematics to dynamic postural impairments in females with patellofemoral pain during stair ascent, Knee (2016), http://dx.doi.org/10.1016/j.knee.2016.01.014

4

D. de Oliveira Silva et al. / The Knee xxx (2016) xxx–xxx

Table 2 Between group comparisons for peak hip, knee and rearfoot angles and displacement area of the COP during stair ascent. Variable

Rearfoot eversion (°) Hip adduction (°) Knee flexion (°) Displacement area of the COP (cm2)

Control group

PFP

Mean (SD)

Mean (SD)°

3.1 (1.2) 9.2 (1.3) 42.9 (2.1) 43.1 (10.2)

6.7 (1.9) 12.5 (2.2) 39.1 (1.8) 26.4 (7.5)

t-value(df)

p-Value

Effect size

3.4(53) 5.1(53) 2.1(53) 6.7(53)

0.009⁎ 0.000⁎ 0.016⁎ 0.001⁎

0.38 0.55 0.33 0.62

Abbreviations: mean (SD) = mean and standard deviation. t-Value(df) = value of t (Student t-test) with degrees of freedom. ° = degrees. Effect size = effect size of Cohen's d. ⁎ Statistically significant (p b 0.05) values.

4. Discussion Biomechanical alterations in proximal, local and distal segments and joints are frequently discussed in the literature of PFP [1,8,10,17]. However, the association of such altered parameters with impaired postural control in individuals with PFP is poorly understood. Our findings indicated that reduced knee flexion and displacement area of the COP are presented in individuals with PFP during stair ascending as well as increased peak hip adduction and peak rearfoot eversion. Additionally, among the kinematic factors investigated, peak hip adduction was the best predictor of the displacement area of the COP (r2 = 23.4%), suggesting that altered hip motion may play a significant role on the dynamic postural impairments observed in this population. 4.1. Kinematic proximal, local and distal factors The present findings indicated that subjects with PFP perform stair ascent movements with greater peak hip adduction as compared to controls, which is in accordance with previous reports during stair ascent and descent [17]. Such an adducted hip position contributes to increased knee valgus and lateral tracking of the patella on the femur, which results in increased PFJ stress [5]. As compared to their matched controls, females with PFP showed reduced knee flexion during stair ascent, which is consistent with previous findings [10,19]. A previous study [19] reported a reduction of 6.8° in knee flexion, which is larger than the reduction of 3.8° found in our study. The 3.0° difference between the studies may be partly explained by differences in step number and height, as the stair used in our study had more steps (seven steps) and lower step height (18 cm) as compared to the study of Crossley et al. [19] (which used a four-step stair, each step being 20 cm height). Reduced peak knee flexion is thought to be a compensatory strategy of subjects with PFP in order to decrease the PFJ stress and perhaps reduce pain during stair ascent [10]. In addition, findings from the present study indicated larger peak rearfoot eversion during stair ascent in individuals with PFP, which is also in accordance with previous findings [8]. The relationship between

Table 3 Forced entry multiple regression model with kinematic parameters as independent variables and displacement area of the COP as dependent variable for PFPG. Regression model

Parameters (peak°)

r2

r2 change

F-change (sig F-change)

1 2†

Rearfoot eversion Hip adduction Rearfoot eversion Knee flexion Hip adduction Rearfoot eversion

0.103 0.338

0.103 0.234

2.307 (0.144) 2.725 (0.018)⁎

0.354

0.016

0.458 (0.507)

3

† Statistically significant (p b 0.05) regression model. ⁎ Statistically significant (p b 0.05) change when the variable was inserted.

excessive rearfoot eversion and PFP has been interpreted based on the assumption that, during the stance phase of gait, an exaggerated everted rearfoot induces excessive internal rotation of the tibia. Consequently, greater hip internal rotation and subsequent hip adduction may lead to increased PFJ stress [42]. 4.2. Dynamic postural control while ascending stairs In contrast to the frequently reported kinematic parameters mentioned above, less attention has been given to the dynamic postural control of individuals with PFP. The present findings indicated that, in comparison to controls, females with PFP had lower displacement area of the COP during stair ascent. Reduced dynamic COP-based measurements were also reported by Thijs et al. [16], who found decreased COP velocity (in the lateromedial direction) during the forefoot contact phase of the gait in subjects who developed PFP. On the other hand, Saad et al. [14] reported greater COP excursion during step up and down tasks performed by females with PFP as compared to matched controls. These seemingly contradictory results are probably related to the different tasks evaluated (i.e., gait [16] and step tasks [14]) and perhaps somewhat different measurements (i.e. COP velocity [16] and excursion [14]). Nevertheless, results from the present and previous studies consistently indicate an alteration in dynamic postural control (as measured by COP parameters) associated with PFP. Protection mechanisms of muscular, kinematic and kinetic [10,19, 43] origins have been reported in patients with PFP, which is thought to protect the knee joint from further injury by better distributing forces around the knee [44]. These protective mechanisms may lead to more constrained COP movements, which reduces the ability to adapt to postural fluctuations thereby limiting functional capacity [32]. Such a constrained pattern is frequently interpreted in the literature as maladaptive [45] as it limits the patient's ability to explore the boundaries of postural control, it could increase the probability of a given perturbation to threaten their dynamic stability [32]. In this line of reasoning, it could be argued that the more complex postural behavior (i.e., less constrained COP trajectories) observed in the pain-free females reflects their superior ability to perform postural corrections (rather than being “locked” into a particular pattern) [45]. In other words, in comparison to the subjects with PFP, pain-free participants showed a more flexible strategy that might be effective for adjusting to stance fluctuations. Constrained movement patterns seem to be a common finding in subjects with PFP, which may be harmful when repetitive actions are required. Non-complex movement patterns in the absence of neurological diseases led some authors to suggest that the pain diminished the variability of individuals with PFP by constraining movement patterns [46]. The same authors have performed a detailed analysis at the moment of heel strike in runners and have found a decrease in the variability of the thigh/leg coupling in individuals with PFP [47]. In this line of reasoning, individuals with PFP showed constrained lower limb kinematics during running, with greater hip adduction and internal rotation [48], which was associated with a reduced ability to change their movement pattern as the run progressed. Such constrained range of motion over the course of the run might have stressed repeatedly the same location and structures of the PFJ, which resulted in pain [48,49]. The present study showed that hip adduction explained significantly (23.4%) the alteration of the displacement area of the COP. Therefore, our and previous results suggest that individuals with PFP do not or are unable to adapt their mechanics during the course of a functional task but rather they remain restrained, thereby compromising dynamic postural control adjustments, eventually eliciting pain. Peak knee flexion was not able to significantly predict (1.6%) the displacement area of the COP. Reduced peak knee flexion during functional tasks has been often reported as a protection mechanism that females with PFP use to avoid pain [10,19]. A reasonable explanation to the lack of association between knee flexion and dynamic postural sway is

Please cite this article as: de Oliveira Silva D, et al, Contribution of altered hip, knee and foot kinematics to dynamic postural impairments in females with patellofemoral pain during stair ascent, Knee (2016), http://dx.doi.org/10.1016/j.knee.2016.01.014

D. de Oliveira Silva et al. / The Knee xxx (2016) xxx–xxx

that the PFP participants were not in their worst pain at the time of the data collection (Mean VAS at the time of data collection = 2.02 (SD = 1.46); worst pain level in the previous month = 5.78 (SD = 1.99)). Therefore, it is possible that the pain-related protection mechanism mentioned above was attenuated, which might not unequivocally depict whether reduced knee flexion is related to alterations in COP displacement. Intermittent and variable symptoms are typical characteristics observed in subjects with PFP [36], so that inter-subject variability in the amount of pain is common at the time of data collection. Therefore, further research that controls the level of pain during data acquisition (e.g., with PFJ loading protocols) is needed to clarify this question. Our findings indicated that peak rearfoot eversion during stair ascent was also not a good predictor of COP displacement (10.3%). Previous studies on the relative contributions of the proximal (hip) and distal (foot and ankle) mechanics in maintaining postural control showed that proximal mechanics caused greater impairment in balance stability than distal [50]. This could be attributed to the assumption that proximal mechanics of the lower extremity have a crucial role in maintaining balance under more challenging dynamic conditions [50]. As COP measure is a global status of all interactions occurring at the body, it is not surprising that peak rearfoot eversion explains only a minor amount of COP displacement during stair ascent.

5

5. Conclusion Our findings indicate that individuals with PFP climb stairs with greater rearfoot eversion, hip adduction as well as decreased knee flexion and displacement area of the COP. The excessive hip adduction while ascending stairs was the biggest predictor of the displacement area of the COP, suggesting the need to consider proximal factors in the management of PFP aimed at reestablishing dynamic postural control. Acknowledgment To São Paulo Research Foundation (FAPESP) for a grant (2014/ 24939-7) and a scholarship (2015/11534-1). The financial sponsors played no role in the design, execution, analysis and interpretation of data, or writing of the study. Appendix A. Anatomical landmarks of retroreflective markers Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.knee.2016.01.014. References

4.3. Clinical implications In a prospective study, Thijs et al. [16] addressed the importance of evaluating the dynamic postural control of individuals with PFP, as they reported alterations in some parameters of COP displacement as risk factors for the development of PFP. Similarly, other cross sectional studies have reported alterations in COP-based measurements in females with PFP [14,15]. However, to the best of our knowledge, the present study is the first to investigate the contribution of altered hip, knee and foot kinematics to the dynamic postural impairments associated with PFP. Based on the present findings, proximally targeted interventions (e.g. hip exercises and stretching) may be of major importance for the functional reestablishment of females with PFP, considering that peak hip adduction had the best capability in predicting dynamic postural impairments during stair ascent. Hip management has been considered an imperative factor in rehabilitation protocols of patients with PFP [11,51]. Therefore, the present findings suggest that further clinical trials may investigate dynamic postural control as an outcome to verify if such protocols are able to reestablish postural behavior. Future studies are required to investigate whether our results may be generalized to other functional tasks such as dynamic squatting, running and walking.

4.4. Limitations An obvious limitation of this study is that only females were included due to high prevalence of PFP in this population [2], so that further research using similar methodology in males with PFP is required. Another limitation of this study is that some participants were not in pain at the time of data collection, which may perhaps mask the real magnitude of biomechanical and postural alterations. Additionally, we have investigated only the three kinematic parameters that have received special attention in the recent literature (i.e. rearfoot eversion and hip adduction as well as decreased knee flexion). As patterns of muscle activation (e.g. hip and quadriceps muscle inhibition) and joint kinetics (e.g. knee flexion moment and hip adduction moment) may significantly influence the COP behavior during stair ascent, a complete analysis including such measurements is required to provide a more complete picture of which factors are contributing to the impaired dynamic postural control in subjects with PFP.

[1] Witvrouw E, Callaghan MJ, Stefanik JJ, Noehren B, Bazett-Jones DM, Willson JD, et al. Patellofemoral pain: consensus statement from the 3rd international patellofemoral pain research retreat. Br J Sports Med 2014;48:411–4. http://dx.doi.org/10.1136/ bjsports-2014-093450. [2] Boling M, Padua D, Marshall S, Guskiewicz K, Pyne S, Beutler A. Gender differences in the incidence and prevalence of patellofemoral pain syndrome. Scand J Med Sci Sports 2010;20:725–30. http://dx.doi.org/10.1111/j.1600-0838.2009.00996.x. Gender. [3] Cheung RTH, Zhang Z, Ngai SPC. Different relationships between the level of patellofemoral pain and quality of life in professional and amateur athletes. PMR 2013;5:568–72. http://dx.doi.org/10.1016/j.pmrj.2012.12.007. [4] Roush JR, Bay RC. Prevalence of anterior knee pain in 18–35 year-old females. Int J Sports Phys Ther 2012;7:396–401. [5] Salsich GB, Brechter JH, Powers CM. Lower extremity kinetics during stair ambulation in patients with and without patellofemoral pain. Clin Biomech 2001;16: 906–12. http://dx.doi.org/10.1016/S0268-0033(01)00085-7. [6] Liao T, Yang N, Ho K, Farrokhi S, Powers CM, Perry J, et al. Femur rotation increases patella cartilage stress in females with patellofemoral pain. Med Sci Sports Exerc 2015;47:1775–80. http://dx.doi.org/10.1249/MSS.0000000000000617. [7] Goudakos IG, König C, Schöttle PB, Taylor WR, Singh NB, Roberts I, et al. Stair climbing results in more challenging patellofemoral contact mechanics and kinematics than walking at early knee flexion under physiological-like quadriceps loading. J Biomech 2009;42:2590–6. http://dx.doi.org/10.1016/j.jbiomech.2009.07.007. [8] De Oliveira Silva D, Briani RV, Pazzinatto MF, Ferrari D, Aragão FA, Albuquerque CE, et al. Reliability and differentiation capability of dynamic and static kinematic measurements of rearfoot eversion in patellofemoral pain. Clin Biomech 2015;30:144–8. http://dx.doi.org/10.1016/j.clinbiomech.2014.12.009. [9] Lankhorst NE, Bierma-Zeinstra SM a, M. van Middelkoop. Factors associated with patellofemoral pain syndrome: a systematic review. Br J Sports Med 2013;47: 193–206. http://dx.doi.org/10.1136/bjsports-2011-090369. [10] De Oliveira Silva D, Briani RV, Pazzinatto MF, Ferrari D, Aragão FA, de Azevedo FM. Reduced knee flexion is a possible cause of increased loading rates in individuals with patellofemoral pain. Clin Biomech 2015;30:971–5. http://dx.doi.org/10.1016/ j.clinbiomech.2015.06.021. [11] Lack S, Barton C, Sohan O, Crossley K, Morrissey D. Proximal muscle rehabilitation is effective for patellofemoral pain: a systematic review with meta-analysis. Br J Sports Med 2015. http://dx.doi.org/10.1136/bjsports-2015-094723. [12] De Oliveira Silva D, Briani RV, Pazzinatto MF, Gonçalves AV, Ferrari D, Aragão FA, et al. Q-angle static or dynamic measurements, which is the best choice for patellofemoral pain? Clin Biomech 2015;30:1083–7. http://dx.doi.org/10.1016/j. clinbiomech.2015.09.002. [13] Santos MCA, Semeghuini TA, Azevedo FM, Colugnati DB, Negrão Filho RF, Alves N, et al. Analysis of localized muscular fatigue in athletes and sedentary subjects through frequency parameters of electromyographic signal. Braz J Sports Med 2008;14(6):509–12. http://dx.doi.org/10.1590/S1517-86922008000600007. [14] Saad MC, Felício LR, Masullo CDL, Liporaci RF, Bevilaqua-Grossi D. Analysis of the center of pressure displacement, ground reaction force and muscular activity during step exercises. J Electromyogr Kinesiol 2011;21:712–8. http://dx.doi.org/10.1016/j. jelekin.2011.07.014. [15] Lee S-P, Souza RB, Powers CM. The influence of hip abductor muscle performance on dynamic postural stability in females with patellofemoral pain. Gait Posture 2012; 36:425–9. http://dx.doi.org/10.1016/j.gaitpost.2012.03.024. [16] Thijs Y, Van Tiggelen D, Roosen P, De Clercq D, Witvrouw E. A prospective study on gait-related intrinsic risk factors for patellofemoral pain. Clin J Sport Med 2007;17: 437–45.

Please cite this article as: de Oliveira Silva D, et al, Contribution of altered hip, knee and foot kinematics to dynamic postural impairments in females with patellofemoral pain during stair ascent, Knee (2016), http://dx.doi.org/10.1016/j.knee.2016.01.014

6

D. de Oliveira Silva et al. / The Knee xxx (2016) xxx–xxx

[17] McKenzie K, Galea V, Wessel J, Pierrynowski M. Lower extremity kinematics of females with patellofemoral pain syndrome while stair stepping. J Orthop Sports Phys Ther 2010;40:625–32. http://dx.doi.org/10.2519/jospt.2010.3185. [18] Gribble P a, Hertel J. Effect of lower-extremity muscle fatigue on postural control. Arch Phys Med Rehabil 2004;85:589–92. http://dx.doi.org/10.1016/j.apmr.2003.06. 031. [19] Crossley KM, Cowan SM, Bennell KL, McConnell J. Knee flexion during stair ambulation is altered in individuals with patellofemoral pain. J Orthop Res 2004;22:267–74. http://dx.doi.org/10.1016/j.orthres.2003.08.014. [20] Winter DA. Biomechanics and motor control of human movement. Fourth Edi. New Jersey: John Wiley & Sons; 2009. [21] Barton CJ, Levinger P, Webster KE, Menz HB. Walking kinematics in individuals with patellofemoral pain syndrome: a case–control study. Gait Posture 2011;33:286–91. http://dx.doi.org/10.1016/j.gaitpost.2010.11.022. [22] Powers C. The influence of altered lower-extremity kinematics on patellofemoral joint dysfunction: a theoretical perspective. J Orthop Sports Phys Ther 2003;33: 639–46. [23] Aliberti S, Costa MSX, Passaro AC, Arnone AC, Sacco ICN. Medial contact and smaller plantar loads characterize individuals with patellofemoral pain syndrome during stair descent. Phys Ther Sport 2010;11:30–4. http://dx.doi.org/10.1016/j.ptsp.2009. 11.001. [24] Lin D, Seol H, Nussbaum MA, Madigan ML. Reliability of COP-based postural sway measures and age-related differences. Gait Posture 2008;28:337–42. http://dx.doi. org/10.1016/j.gaitpost.2008.01.005. [25] Peng H-T, Song C-Y. Predictors of treatment response to strengthening and stretching exercises for patellofemoral pain: an examination of patellar alignment. Knee 2015. http://dx.doi.org/10.1016/j.knee.2014.10.012 [in press]. [26] Willy RW, Manal KT, Witvrouw EE, Davis I. Are mechanics different between male and female runners with patellofemoral pain? Med Sci Sports Exerc 2013;44: 2165–71. http://dx.doi.org/10.1249/MSS.0b013e3182629215.Are. [27] Craig CL, Marshall AL, Sjöström M, Bauman AE, Booth ML, Ainsworth BE, et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 2003;35:1381–95. http://dx.doi.org/10.1249/01.MSS.0000078924. 61453.FB. [28] Ferrari D, Kuriki HU, Silva CR, Alves N, Mícolis de Azevedo F. Diagnostic accuracy of the electromyography parameters associated with anterior knee pain in the diagnosis of patellofemoral pain syndrome. Arch Phys Med Rehabil 2014;95:1521–6. http://dx.doi.org/10.1016/j.apmr.2014.03.028. [29] Briani RV, De Oliveira Silva D, Pazzinatto MF, De Albuquerque CE, Ferrari D, Aragão FA, et al. Comparison of frequency and time domain electromyography parameters in women with patellofemoral pain. Clin Biomech 2015;30:302–7. http://dx.doi. org/10.1016/j.clinbiomech.2014.12.014. [30] De Oliveira Silva D, Briani R, Pazzinatto M, Ferrari D, Aragão F, Azevedo F. Vertical ground reaction forces are associated with pain and self-reported functional status in recreational athletes with patellofemoral pain. J Appl Biomech 2015;31:409–14. http://dx.doi.org/10.1123/jab.2015-0048. [31] Stebbins J, Harrington M, Thompson N, Zavatsky a, Theologis T. Repeatability of a model for measuring multi-segment foot kinematics in children. Gait Posture 2006;23:401–10. http://dx.doi.org/10.1016/j.gaitpost.2005.03.002. [32] Rhea CK, Kiefer AW, Haran F, Glass S, Warren W. A new measure of the CoP trajectory in postural sway: dynamics of heading change. Med Eng Phys 2014;36:1473–9. http://dx.doi.org/10.1016/j.medengphy.2014.07.021. [33] Roerdink M, Hlavackova P, Vuillerme N. Center-of-pressure regularity as a marker for attentional investment in postural control: a comparison between sitting and

[34] [35]

[36]

[37]

[38] [39] [40] [41]

[42] [43] [44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

standing postures. Hum Mov Sci 2011;30:203–12. http://dx.doi.org/10.1016/j. humov.2010.04.005. Jordan K, Challis JH, Newell KM. Walking speed influences on gait cycle variability. Gait Posture 2007;26:128–34. http://dx.doi.org/10.1016/j.gaitpost.2006.08.010. Silva CR, Silva DO, Aragão FA, Ferrari D, Alves N, Azevedo FM. Influence of neuromuscular fatigue on co-contraction between vastus medialis and vastus lateralis during isometric contractions. Kinesiol 2014;46:179–85. Grenholm A, Stensdotter AK, Häger-Ross C. Kinematic analyses during stair descent in young women with patellofemoral pain. Clin Biomech 2009;24:88–94. http://dx. doi.org/10.1016/j.clinbiomech.2008.09.004. Prieto TE, Myklebust JB, Hoffmann RG, Lovett EG, Myklebust BM. Measures of postural steadiness differences between healthy young and elderly adults. Trans Biomed Eng 1996;43:965–6. Cohen J. Statistical power analysis for the behavioral sciences. New York: Routledge; 1988 2. Field A. Discovering statistics using IBM SPSS statistics. 4th ed. London: SAGE Publications Ltd.; 2013. Schmitz A, Noehren B. What predicts the first peak of the knee adduction moment ? Knee 2014;21:1077–83. http://dx.doi.org/10.1016/j.knee.2014.07.016. Harrell Jr FE, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996;15:361–87. Tiberio D. The effect of excessive subtalar joint pronation on patellofemoral mechanics: a theoretical model. J Orthop Sports Phys Ther 1987;9:160–5. Powers CM, Heino JG, Rao S, Perry J. The influence of patellofemoral pain on lower limb loading during gait. Clin Biomech (Bristol, Avon) 1999;14:722–8. Powers CM, Ho K-Y, Chen Y-J, Souza RB, Farrokhi S. Patellofemoral joint stress during weight bearing and non-weight bearing quadriceps exercises. J Orthop Sports Phys Ther 2014;44:320–7. http://dx.doi.org/10.2519/jospt.2014.4936. Rudolph KS, Axe MJ, Buchanan TS, Scholz JP, Snyder-Mackler L. Dynamic stability in the anterior cruciate ligament deficient knee. Knee Surgery, Sport Traumatol Arthrosc 2001;9:62–71. http://dx.doi.org/10.1007/s001670000166. Hamill J, Van Emmerik RE a, Heiderscheit BC, Li L. A dynamical systems approach to lower extremity running injuries. Clin Biomech 1999;14:297–308. http://dx.doi.org/ 10.1016/S0268-0033(98)90092-4. Heiderscheit BC, Hamill J, Van Emmerik RE a. Variability of stride characteristics and joint coordination among individuals with unilateral patellofemoral pain. J Appl Biomech 2002;18:110–21. Noehren B, Sanchez Z, Cunningham T, McKeon PO. The effect of pain on hip and knee kinematics during running in females with chronic patellofemoral pain. Gait Posture 2012;36:596–9. http://dx.doi.org/10.1016/j.gaitpost.2012.05.023. Noehren B, Pohl MB, Sanchez Z, Cunningham T, Lattermann C. Proximal and distal kinematics in female runners with patellofemoral pain. Clin Biomech (Bristol, Avon) 2012;27:366–71. http://dx.doi.org/10.1016/j.clinbiomech.2011.10.005. Gribble P a, Hertel J. Effect of hip and ankle muscle fatigue on unipedal postural control. J Electromyogr Kinesiol 2004;14:641–6. http://dx.doi.org/10.1016/j.jelekin. 2004.05.001. Barton CJ, Lack S, Hemmings S, Tufail S, Morrissey D. The “best practice guide to Conservative Management of patellofemoral pain”: incorporating level 1 evidence with expert clinical reasoning. Br J Sports Med 2015;49:923–34. http://dx.doi.org/10. 1136/bjsports-2014-093637.

Please cite this article as: de Oliveira Silva D, et al, Contribution of altered hip, knee and foot kinematics to dynamic postural impairments in females with patellofemoral pain during stair ascent, Knee (2016), http://dx.doi.org/10.1016/j.knee.2016.01.014