# Effects of load carriage methods on fall risk and gait variability during stair ascent: a functional data analysis approach

## TL;DR

This study examines how different ways of carrying loads affect gait and fall risk during stair climbing, using advanced data analysis to reveal how the body adapts.

## Contribution

The study introduces functional data analysis to explore gait variability during loaded stair ascent, revealing specific neuromuscular adaptations.

## Key findings

- Shoulder load increases fall risk compared to hand load due to an elevated and asymmetric center of mass.
- Functional principal component analysis identified significant differences in hip, knee, and ankle joint angles across loading conditions.
- Stability in both frontal and sagittal planes is compromised under shoulder load, requiring more extensive gait adaptations.

## Abstract

Based on the existing research that predominantly focuses on loaded level walking or employs discrete-point methods to analyze stair negotiation, this study utilizes functional data analysis to systematically investigate the effects of three load carriage methods on gait variability during upstairs walking in healthy adult males, aiming to elucidate the specific neuromuscular adaptation strategies induced by different loading conditions.

Nineteen healthy young male participants were recruited for this study. Kinematic and kinetic data were collected during stair walking under three load carriage conditions using a three-dimensional motion capture system and force plates. Gait parameters, center of pressure (COP) trajectories, and lower-limb joint angle time series in the sagittal and frontal planes for the hip, knee, and ankle joints were extracted. Functional principal component analysis (fPCA) was employed to reduce the dimensionality and process the joint angle curves, aiming to identify the dominant modes of variability throughout the entire gait cycle. One-way analysis of variance (ANOVA) was subsequently applied to compare between-group differences in gait parameters and COP measures.

Significant differences were observed across different load carriage conditions in step length, single support time, and second double support time (P < 0.05). The center of pressure (COP) trajectory in the mediolateral direction also showed significant differences (P < 0.05). Regarding joint kinematics, functional principal component analysis revealed significant between-condition differences in the sagittal plane hip angle for principal component 1 (PC1), as well as in PC1 and principal component 3 (PC3) for the frontal plane hip angle (P < 0.05). For the knee joint, a significant difference was found in PC1 of the frontal plane angle time series (P < 0.05). At the ankle joint, significant differences were identified in PC3 of the sagittal plane angle and in PC1 of the frontal plane angle (P < 0.05).

By employing a functional data analysis framework, this study provides a more nuanced understanding of phase-specific compensatory mechanisms during loaded stair ascent, revealing that the shoulder load poses a greater fall risk than hand load. This elevated risk is primarily due to an elevated and asymmetric center of mass, which induces a forward trunk inclination and compromises stability in both the frontal and sagittal planes, necessitating more extensive gait adaptations.