OpenCap-based five-times sit-to-stand analysis in older Asian women with and without sarcopenia

Sarcopenia-related lower-limb functional impairment is commonly assessed using five-times sit-to-stand (5STS) completion time, which provides limited information about how performance differs within the task. This study examined two questions: the concurrent agreement of OpenCap-derived lower-limb kinematics with Vicon during repeated 5STS, and whether OpenCap-derived kinematic and temporal measures distinguished healthy and sarcopenic older Asian women.

Twenty-four participants were classified as healthy (n = 12) or sarcopenic (n = 12) according to AWGS 2025 criteria and completed concurrent OpenCap and Vicon recordings. Agreement was assessed from participant-level mean waveforms using RMSE and MAE for angular error magnitude, ICC(A,1) for absolute agreement, R² for waveform similarity, and phase-dependent bias.

Knee flexion/extension showed the lowest absolute error and high agreement (RMSE = 1.84°, ICC(A,1) = 0.998). Hip flexion/extension also showed high agreement, although absolute error was larger (RMSE = 5.42°, ICC(A,1) = 0.970). Ankle agreement was more variable. Statistical parametric mapping identified localized reductions in hip and knee flexion in the sarcopenic group, while angular velocity waveforms did not differ significantly. Sarcopenic participants also showed longer total, ascent, and descent durations and greater repetition-related slowing, while repetition-duration variability did not differ significantly.

Within this protocol, OpenCap showed strong participant-level agreement for hip and knee kinematics and provided phase- and repetition-specific descriptors of 5STS performance. The independent validity and incremental clinical value of the derived temporal measures require further evaluation.