Primary osteoarthritis treatments such as a total hip (THR) or knee (TKR) replacement lead to postural control changes reinforced by age. Balance tests such as standing with eyes open (EO) or closed (EC) give a possibility to calculate both linear and nonlinear indicators. This study aimed to find the group of linear and/or nonlinear measures that can differentiate healthy people and patients with TKR or THR from each other. This study enrolled 49 THR patients, 53 TKR patients, and 16 healthy controls. The center of pressure (CoP) path length, sample entropy (SampEn), fractal dimension (FD), and the largest Lyapunov exponent (LyE) were calculated separately for AP and ML directions from standing with EO/EC. Cluster analysis did not result in correct allocation to the groups according to all variables. The discriminant model included LyE (ML-EO, ML-EC, AP-EC), FD (AP-EO, ML-EC, AP-EC), CoP-path AP-EC, and SampEn AP-EC. Regression analysis showed that all nonlinear variables depend on the group. The CoP path length is different only in THR patients. It was concluded that standing with EC is a better way to assess the amount of regularity of CoP movement and attention paid to maintain balance. Nonlinear measures better differentiate TKR and THR patients from healthy controls.
Keywords: Lyapunov exponent; body balance; fractal dimension; hip arthroplasty; knee arthroplasty; older adults; osteoarthritis; postural control; sample entropy.