In vivo knee joint friction after total knee arthroplasty is highly dynamic and phase-dependent

Natural synovial knee joints exhibit very low friction, enabling efficient long-term function. In contrast, total knee arthroplasty (TKA) introduces artificial articulating surfaces with altered kinematic and tribological conditions. However, how in vivo friction evolves across different dynamic activities after TKA remains unclear. The aim of this study was to quantify in vivo knee joint friction after TKA and to investigate its relationship with activity-dependent loading conditions and knee joint kinematics.

In vivo joint friction was assessed in three subjects with instrumented telemetric knee implants during chair sit, squat, and level walking. Joint contact forces and moments were recorded in vivo, while synchronous fluoroscopic imaging was used to determine knee joint kinematics. Loading data were referenced to the instantaneous axis of rotation to calculate frame-by-frame knee joint friction throughout each activity.

In vivo friction after TKA was highly dynamic and activity-dependent, showing distinct temporal variations across the three different activities. During chair sit and squat, coefficients of friction (µ) of up to approximately 0.15 were observed during loaded flexion and extension phases, consistent with mixed or boundary lubrication regimes reported in vitro. In contrast, substantially higher µ values of up to 0.3 during chair sit and 0.43 during gait were observed during low-loading or rapid kinematic transition phases. Peak friction moments reached up to 5.5 Nm during chair sit, 5.0 Nm during squat, and 8.9 Nm during gait. Overall, friction behavior varied substantially throughout the movement cycle and exhibited pronounced phase-dependent and inter-individual variability with peaks during swing phases.

These findings demonstrate that in vivo knee joint friction after TKA is dynamic, activity- and phase-dependent. The observed friction dynamics illustrate that friction is only partially linked to contact load magnitude, while motion velocity and possible lubrication-related effects may also influence friction. These findings may help improve experimental and computational models used in TKA tribology and implant evaluation.