Análise combinada de variáveis biomecânicas por meio de uma rede neural artificial auto-organizada e o uso do statistical parametric mapping na capacidade de diferenciar mulheres com dor patelofemoral

Abstract

Introduction: Individuals with patellofemoral pain (PFP) experience pain during walking, however, they do not present changes in movements that increase stress on the patellofemoral joint in this task. It is not known exactly how pain and movement are related in walk and the evidence of kinematic changes in this task is limited. PFP is related to changes in muscle activity and movement patterns during functional tasks. Despite the common use of electromyography amplitude signal analysis, the evidence regarding this variable in PFP is heterogeneous and limited. Objectives: 1. To determine whether the combined analysis of kinematics and electromyography data enhances the ability to differentiate between women with and without PFP, compared to the independent analysis of kinematics and electromyography. 2. Compare walking angular kinematics between women with and without PFD using Statistical Parametric Mapping (SPM); compare walking linear kinematics between women with and without PFP; evaluate the relationship between pain intensity and angular kinematics in women with PFP in the variables that present differences between the groups. Methods: 1. Case-control study that evaluated kinematics and electromyography during the lateral Step Down (LSD) of 37 women with PFP and 34 unimpaired controls. For 3D kinematics, movements in the sagittal, frontal and transverse planes of the trunk, pelvis, hip, knee and foot were evaluated. For electromyography, signals from the adductor longus, gluteus medius, vastus lateralis and medialis, rectus femoris, biceps femoris, medial gastrocnemius and tibialis anterior muscles were used. The Movement Deviation Profile (MDP) was used to analyze kinematics, electromyography and kinematics combined with electromyography. A MANOVA of MDPmean Z-scores was used to compare the groups, considering p<0.05. 2. Cross-sectional study that evaluated the walking kinematics of 571 women with PFP and 571 unimpaired controls. For angular variables, movements in the sagittal, frontal and transverse planes of the trunk, pelvis, hip, knee and foot were considered. For the linear variables, the time of the first double support, time of the second double support, time of the single support, stride length, speed and cadence were considered. To compare linear data between the two groups, a MANOVA was used, considering p<0.05; and to compare the angular data, SPM was used. Results: 1. MANOVA showed group interaction (Lambda=0.838; F=4.329; p=0.008). There was a significant difference between groups in the Z-score only for kinematics (p=0.003). 2. MANOVA showed interaction between groups (Lambda = 0.738; F = 57.12; p < 0.001). There was a significant difference in the first double support (p=0.008), single support (p<0.001) and stride length (p<0.001). The SPM demonstrated a difference between groups for lateral inclination and trunk rotation; pelvic obliquity; hip adduction and flexion; knee flexion; dorsiflexion; and foot progression angle. Conclusion: Women with PFP exhibit altered movement patterns during the LSD task, but no change in muscular activity. Using the MDP, which can combine kinematic and electromyography variables from different segments and muscles, kinematics was the most influential in distinguishing between women with and without PFP.