A multiple regression normalization approach to evaluation of gait in total knee arthroplasty patients

Highlights

• Variance in gait data may limit capacity to discern gait patterns in a cohort.

• Multiple-regression normalization decorrelates demographic factors from gait data

• TKA patients have reduced knee function but employ greater hip power

Abstract

Background

Gait features characteristic of a cohort may be difficult to evaluate due to differences in subjects' demographic factors and walking speed. The aim of this study was to employ a multiple regression normalization method that accounts for subject age, height, body mass, gender, and self-selected walking speed in the evaluation of gait in unilateral total knee arthroplasty patients.

Methods

Three-dimensional gait analysis was performed on 45 total knee arthroplasty patients and 31 aged-matched controls walking at their self-selected speed. Gait data peaks including joint angles, ground reaction forces, net joint moments, and net joint powers were normalized using subject body mass, standard dimensionless equations, and a multiple regression approach that modeled subject age, height, body mass, gender, and self-selected walking speed.

Findings

Normalizing gait data using subject body mass, dimensionless equations, and multiple regression approach resulted in a significantly lower knee adduction moment and knee extensor power in total knee arthroplasty patients compared to controls (p < 0.05). In contrast to normalization using body mass and dimensionless equations, multiple regression normalization greatly reduced variance in gait data by minimizing correlations with subject demographic factors and walking speed, resulting in significantly higher peak hip extension angles and peak hip flexion powers in total knee arthroplasty patients (p < 0.05).

Interpretation

Total knee arthroplasty patients generate greater hip extension angles and hip flexor power and have a lower knee adduction moment than healthy controls. This gait pattern may be a strategy to reduce muscle and joint loading at the knee.

Introduction

Total knee arthroplasty (TKA) is the established treatment for end-stage knee osteoarthritis when conservative treatment options have been exhausted. In 2005, approximately 500,000 TKA procedures were performed in the United States alone at a cost exceeding $11 billion (Losina et al., 2009). It is estimated that close to 3.48 million TKAs will be performed annually by 2030 (Kurtz et al., 2007). While TKA has been shown to reduce pain, improve knee function, and increase quality of life (Ethgen et al., 2004, Hatfield et al., 2011), TKA patients are known to exhibit post-operative compensatory gait patterns to minimize joint loading and pain at the affected knee. This includes walking with less knee flexion (Mandeville et al., 2007, McClelland et al., 2011), a lower knee adduction moment (Alnahdi et al., 2011, McClelland et al., 2010), and a shorter swing phase in the affected leg (Andriacchi et al., 1982). In addition, lower ankle plantar–flexor power and higher hip flexor power has been observed in TKA subjects compared to aged-matched controls (Levinger et al., 2013).

Gait characteristics of an individual, including kinematics and net joint moments, are influenced by age, height, body mass, and gender (Moisio et al., 2003, Senden et al., 2012), as well as by variations in self-selected walking speed (Andriacchi et al., 1977, Lelas et al., 2003), including fluctuations within a single testing session (Benedetti et al., 2013, Yogev-Seligmann et al., 2008). Ultimately, variability in subject demographic factors increases dispersion of gait data and may limit capacity to discern some pathological movement patterns (Boyer et al., 2008, Hof, 1996). The dimensionless normalization (DS) equations of Hof (1996) are widely used in clinical gait analysis to normalize gait data to subject height and weight, and assume proportional scaling (Pierrynowski and Galea, 2001); however, the use of DS in normalizing spatiotemporal data, ground reaction forces (GRFs), and net joint moments has been shown to result in residual correlations between the normalized gait data and subject walking speed (Carty and Bennett, 2009, Lelas et al., 2003), body mass (Wannop et al., 2012), and gender (Boyer et al., 2008, Cho et al., 2004).

Statistical techniques, including use of linear (Lelas et al., 2003, O'Malley, 1996), multiple-linear (Lee et al., 1999, Macellari et al., 1999, Senden et al., 2012), and non-linear (Wannop et al., 2012) regression models, have been employed to scale gait data to subject anthropometry; however, these models do not account for variations in subject height, body mass, age, and gender and are rarely used to account for variations in subject walking speed. To minimize the influence of self-selected walking speed variations on gait data, subjects are often ‘speed-matched’ between groups, which can be difficult and time consuming to carry out in practice. Multiple-regression approaches to data normalization have been used to predict spatiotemporal gait data from anthropometrics and walking speed (Dixon et al., 2014); however, these approaches have received little attention to date and have not been used in models of joint angles, joint moments, and joint powers.

The aim of this study was twofold. First, to employ a multiple regression normalization (MR) method to minimize variation in subject joint angles, GRFs, net joint moments, and joint powers due to walking speed, body height, body mass, age, and gender in healthy adults; and second, to use this normalization approach to compare joint angles, GRFs, net joint moments, and net joint powers between TKA patients and healthy adults. We hypothesized that MR would reveal significant joint-level differences in angles, net moments, and powers in TKA patients and controls by eliminating effects of subject anthropometry and walking speed variations.