Braking and propulsive impulses in individuals with patellofemoral pain syndrome when walking up and down stairs

Patellofemoral pain syndrome (PFPS) is a prevalent clinical condition and it affects gait behavior. Braking and propulsive impulses are important biomechanical parameters obtained from ground reaction forces (GRF), which combine the amount of force applied over a period of time. The aim of this study was to evaluate these impulses while walking up and down stairs in healthy controls and PFPS individuals. The results did not reveal significant differences in braking and propulsive impulses between groups during these activities. Thus, the painful condition on a simple functional activity was insufficient to change the motor strategy to walking up or down the stairs.


Introduction
Patellofemoral pain syndrome (PFPS) is a prevalent clinical condition (Boling et al., 2010;Da Cunha et al., 2013;Davis & Powers, 2010).Several researchers investigated different motor tasks such as running, walking, jumping, and walking up and down stairs in patients with PFPS (Magalhães et al., 2010;Nakagawa, Moriya, Maciel, & Serrão, 2012).Among these motor tasks, walking up and down stairs are the most painful and challenging activities for individuals with PFPS because they can exacerbate the pain symptoms (Bretcher & Powers, 2002).
Different parameters such as ground reaction forces (GRF), center of pressure displacement, electromyography activity of muscle groups mainly involved in these tasks, and movement analysis of hip, knee and trunk segments are studied to attempt to identify changes in movement control in these individuals (Boling et al., 2009;Nakagawa et al., 2012;Saad, Felício, Masullo, Liporaci, & Bevilaqua-Grossi, 2011).Overall, these studies showed that PFPS individuals presented lower values of vertical GRF for the affected limb in a jump-landing task (Boling et al., 2009) and when walking up and down stairs (Saad et al., 2011).They also presented reduced levels of EMG activity (Saad et al., 2011) and increased hip adduction, knee abduction, and trunk lean in the frontal plane (Nakagawa et al., 2012) when walking up and down stairs.
Although previous studies showed that PFPS individuals reduced their vertical GRF, a more comprehensive view of the changes in movement control can be achieved by looking at the force applied on the ground over a period of time, that is, the impulse, which measures how the momentum of the body changes (Enoka, 2008).Walking up or down stairs involves changing body momentum appropriately to successfully perform these tasks.Gait studies have split braking impulse from propulsive impulse (Moraes, Allard, & Patla, 2007;Peterson, Kautz, & Neptune, 2011).Braking impulse corresponds to the deceleration period of the support phase, whereas the propulsive impulse corresponds to the acceleration phase of the support phase (Moraes et al., 2007).Both braking and propulsive impulses are computed for vertical and horizontal directions.Both vertical and horizontal components of the impulses are regulated to properly control step length and width (Begg, Sparrow, & Lythgo, 1998;Moraes et al., 2007, Patla, Robinson, Samways, & Armstrong, 1989;Warren, Young, & Lee, 1986).For instance, step lengthening is associated with an increase in propulsive impulse in both vertical and anterior-posterior (AP) directions, whereas increase in step width is associated with greater medial-lateral (ML) propulsive impulse.Mochida et al. (2009) found that individuals with chronic knee pain showed increased GRF braking and propulsive peaks while stepping, indicating knee overload.Similarly, Zeni, and Higginson (2009) also found changes in these parameters in individuals with moderate to severe pain.Saad et al. (2011) assessed only the vertical component of the GRF while walking up and down stairs.They showed a reduction in both active and passive peaks for individuals with PFPS when walking down a staircase.This variable represents essentially the amount of overload applied on to the musculoskeletal system.To fully understand how this task is controlled by individuals with PFPS, a more comprehensive analysis of the GRF is required.Therefore, the present study re-analyzed this previous data set in order to compute braking and propulsive impulses in both AP and vertical directions.Thus, the purpose of this study was to evaluate the braking and propulsive impulses derived from the GRF while healthy individuals and patients with PFPS walked up and down stairs.Based on the data from Saad et al. (2011), we hypothesized that PFPS individuals would reduce braking and propulsive impulses when walking down stairs, but would not change them when walking up stairs.

Participants
Thirty sedentary female volunteers participated in this study, divided into two groups: control group (CG) (n=15, age: 23.3 ±2.1 years, height: 160.4 ±3.4 cm, body mass: 53.5 ±2.2 kg) and patellofemoral pain syndrome group (PFPSG) (n=15, age: 23.2 ±2.3 years, height: 159.7 ±2.8 cm, body mass: 58.7 ±3.8 kg).For the CG, inclusion criteria were: presence of no more than two PFPS signals in the functional evaluation and absence of pain verified through the Visual Analogue Scale (VAS) during the last month.For the PFPSG, inclusion criteria were: three signals indicating PFPS observed during functional evaluation, pain in at least 3 cm in VAS during the last month, and report of pain in at least two functional activities (Cowan, Bennell, Crossley, Hodges, & McConnell, 2002;Cowan, Bennell, Hodges, Crossley, & McConnell, 2001;Powers, Ward, Fredericson, Guillet, & Shellock, 2003).Exclusion criteria for both groups were: history of lesion or surgery in the musculoskeletal system, hip, knee, or ankle and individuals with neurological, cardiovascular, or rheumatologic diseases.
All participants had the right leg as the dominant one, and participants in the PFPSG identified their right leg as the one with pain.All procedures of this study were approved by the Ethics Committee in Research of Hospital and Clinics School of Medicine at Ribeirão Preto, University of São Paulo (HC-FMRP-USP), Ribeirão Preto, SP, Brazil, n 8089/2007.

Procedures
The staircase used in this study had three steps (height: 20 cm for each step) with a force plate (AMTI OR6-7-1000) embedded in the first step (detailed description of the staircase can be found in Saad et al., 2011).Each participant stood in front of the first step and walked up stairs (i.e., three steps) and stop.After that, they turned around and walked down stairs (i.e., three steps) and stopped when they reached the floor.Participants performed this procedure three times starting with the right leg and three times starting with the left leg.The order of the leg used to start the task was completely randomized.Participants should walk up and down stairs following the cadence of 30 steps/min which were paced to the beats of a metronome.
Force plate registered the forces and moments of force in three directions: anterior-posterior (AP), medial-lateral (ML), and vertical.Data was sampled with a frequency of 100 Hz.Force values were normalized by body weight for each participant.
The transition between braking and propulsive impulses was based on the zero crossing point of the AP component of the GRF (Moraes et al., 2007).As illustrated on Figure 1, after the initial contact with the force plate there was a small oscillation of the AP GRF component around zero, followed by a consistent negative direction of the AP force (i.e., braking phase).After that, force direction changes to positive values, and the instant when this happens (i.e., zero crossing) defines the beginning of the propulsive phase.Braking impulse corresponds to the area under the curve from heel contact on the force plate to zero-crossing.Propulsive impulse corresponds to the area under the curve from zero-crossing to toe-off.

Statistical analysis
Comparisons were made between the dominant and nondominant legs of each group.To analyze and compare the effect of tasks (walk up and down stairs), legs (dominant and non dominant), and groups (control and PFPS), a linear regression model with mixed effects (random and fixed effects) was used (McLean, Sanders, & Stroup, 1991).Significant levels were defined for α < .05,and a confidence interval (CI) of 95% was calculated.The analysis was performed using the SAS 9.0 software (PROC MIXED function).For the comparisons, we used the posttest by orthogonal contrasts.

Results
The comparison between groups for the support and propulsive duration did not reveal any significant difference.For the braking duration, there was a significant difference between control and PFPS groups for the non-dominant limb in the walk down task (p = .05).Braking duration reduced for the PFPS group (Table 1).
The comparison of the braking and propulsive impulses in both vertical and AP directions between the groups is shown in Tables 2 and 3, respectively.No significant differences were observed between groups for any of the comparisons (legs and tasks).

Discussion
Studies assessing the impulses generated by individuals with PFPS while walking up and down stairs are scarce in the literature.This study demonstrated that individuals with PFPS did not differ from the healthy individuals with respect to the impulses generated while ascending and descending the stairs.However, this result contradicted the hypothesis that PFPS in-dividuals would reduce braking and propulsive impulses when walking down stairs.When walking up stairs, our results agree with the hypothesis that impulse would not change between groups.Remember that these hypotheses were based on the study by Saad et al. (2011).They assessed the peak value of vertical component of the GRF while walking up and down stairs and showed a reduction in both active and passive peaks for individuals with PFPS when walking down stairs.There was no difference between groups when walking up stairs.
PFPS individuals displaced their weight with less intensity onto the affected limb as a way of protecting their limb and avoiding pain when walking down stairs (Saad et al., 2011).Thus, when individuals walked down stairs, neither the task requirement nor their level of pain reported at the time of the examination within a range of mild to moderate-was enough to affect GRF peak (Saad et al., 2011).Moreover, the activity and the level of pain were not important enough to change the GRF peak when walking up stairs.
The fact that individuals with PFPS were able to reduce the peak value of the vertical GRF and, at the same time, maintain impulse generation at the same level as the control individuals

Figure 1 .
Figure 1.Vertical and anterior-posterior (AP) ground reaction forces for the walking up (left side) and walking down (right side) tasks showing the areas corresponding to braking (dark gray) and propulsive (light gray) impulses for one participant of the present study.BW: body weight.

Table 1 .
Mean support, braking, and propulsive duration for control and patellofemoral pain syndrome (PFPS) groups for both walking down and up tasks when using dominant and non-dominant limbs as the leading limb.

Table 2 .
Mean vertical impulse (braking and propulsive) for control and patellofemoral pain syndrome (PFPS) groups for both walking down and up tasks when using dominant and non-dominant limbs as the leading limb.

Table 3 .
Mean anterior-posterior impulse (braking and propulsive) for control and patellofemoral pain syndrome (PFPS) groups for both walking down and up tasks when using dominant and non-dominant limbs as the leading limb.
CI: Confidence Interval | BW: body weight