Just a matter of indexes: symmetry in walking

Abstract

Gait symmetry is related to differences between the left and right side of the body and provides information about functional limitation due to pathology/injury [1]. Many metrics exist to quantify gait symmetry, differing for the considered data source and mathematical approach [1]. Although the body center of mass (BCoM) is the point that globally represents an individual’s movement, very few studies have focused on the quantification of gait symmetry from its trajectory (Simmetry Index, SI) [2]. Conversely, the use of the Harmonic Ratio and of its improved version (iHR) has gained great popularity in recent years [3], being extracted from trunk 3D accelerations easily measured by wearable inertial devices. The relationship between the SI and iHR, however, has not been quantified yet. The aim of this study is thus to investigate this relationship in healthy adults during walking at different speeds. Five healthy adults (2 women, 1.69±0.08 m, 29.8±2.8 y) participates to this study (CAR 101/2021). They walked for 60 s on a treadmill at six different speeds (from 0.55 to 1.95 m/s with step of 0.28 m/s). The trajectories of 18 markers located on the main joint centers [4] were collected by an optoelectronic system (Vicon, 200 Hz). Simultaneously, lower trunk 3D linear accelerations were measured by an inertial sensor (APDM Opal, 200 Hz). The 3D trajectory of the BCoM was estimated from marker trajectories as the weighted mean of an 11–segment model [4]. Gait symmetry was then quantified for each stride and each direction (antero-posterior AP, medio-lateral ML, cranio-caudal CC) using two indices: the Symmetry Index (SI) from the BCoM trajectory [2] and the iHR from the trunk accelerations [3]. Both iHR and SI are based on the harmonic content of the relevant signal and range from 100% (perfect symmetry) to 0% (total asymmetry). To assess the effect of walking speed on SI and iHR values, a two-way repeated measure ANOVA was performed after verifying for normal data distribution. A total number of 1572 strides were analyzed, with a minimum of 288 strides for each participant. SI and iHR significantly differed in all three directions for each velocity (p<0.01), except at 1.39, 1.67 and 1.95 m/s in the ML direction (Fig. 1). When considering the effect of gait speed on symmetry values, the latter significantly changed only for the AP and CC components (p<0.0001) ( Fig. 1 ). SI and iHR may provide substantial different information about the level of gait symmetry, with iHR characterized by greater values in AP and CC directions and smaller values in ML direction. Interestingly, gait symmetry increases with speed, at least up to 1.67 m/s, in AP and CC directions. This is not the case for the ML component: whereas SI slightly decreases with speed (p<0.05), the high variability of iHR, already discussed in the literature [4], does not allow to identify any clear trend. It is suggested to carefully describe the methodology used to assess gait symmetry, as the type of data and the method used may strongly influence the final outcome and, thus, its clinical interpretation.