Objectively measured physical activity: an important functional outcome following knee arthroplasty
Author(s)
Type
Journal Article
Abstract
Introduction: The Oxford Knee Score (OKS) is known to demonstrate a ceiling effect after knee arthroplasty, which can limit its ability to distinguish among high-functioning patients. Objective physical activity (OPA), derived from accelerometry, captures a different and complementary domain of postoperative function, namely free-living activity behaviour. This study examines if OPA can capture variation in activity following knee arthroplasty, including among patients scoring highly on the OKS.
Methods: This analysis used data from a prospective randomised controlled trial comparing total knee arthroplasty and bi-unicompartmental knee arthroplasty. Participants with paired OPA and OKS datasets were included. OPA outcomes included physical activity energy expenditure (PAEE), sedentary time (ST), light (LPA) and moderate-to-vigorous activity (MVPA), and proportional intensity contributions to PAEE. Outliers were removed (median ±1.5×IQR), after which kurtosis and skewness were calculated.
Results: Forty-nine preoperative and sixty-one postoperative participants were included. Preoperatively, OPA measures were normally distributed with no floor or ceiling effects, except for MVPA. Preoperative OKS was normally distributed with no ceiling effect. Postoperatively, all OPA distributions, bar MVPA, were normally distributed, including in high-functioning patients (OKS ≥44). Postoperative OKS was not normally distributed and showed a ceiling effect.
Conclusion: OPA was normally distributed following knee arthroplasty, and maintained this distribution in a high-functioning subgroup of patients (OKS ≥44). OPA is therefore a valuable outcome measure in knee arthroplasty which can complement PROMs, especially for patients with high functional expectations.
Methods: This analysis used data from a prospective randomised controlled trial comparing total knee arthroplasty and bi-unicompartmental knee arthroplasty. Participants with paired OPA and OKS datasets were included. OPA outcomes included physical activity energy expenditure (PAEE), sedentary time (ST), light (LPA) and moderate-to-vigorous activity (MVPA), and proportional intensity contributions to PAEE. Outliers were removed (median ±1.5×IQR), after which kurtosis and skewness were calculated.
Results: Forty-nine preoperative and sixty-one postoperative participants were included. Preoperatively, OPA measures were normally distributed with no floor or ceiling effects, except for MVPA. Preoperative OKS was normally distributed with no ceiling effect. Postoperatively, all OPA distributions, bar MVPA, were normally distributed, including in high-functioning patients (OKS ≥44). Postoperative OKS was not normally distributed and showed a ceiling effect.
Conclusion: OPA was normally distributed following knee arthroplasty, and maintained this distribution in a high-functioning subgroup of patients (OKS ≥44). OPA is therefore a valuable outcome measure in knee arthroplasty which can complement PROMs, especially for patients with high functional expectations.
Date Acceptance
2026-08-17
Citation
Bone & Joint Research
ISSN
2046-3758
Publisher
The British Editorial Society of Bone & Joint Surgery
Journal / Book Title
Bone & Joint Research
Copyright Statement
Copyright This paper is embargoed until publication. Once published the Version of Record (VoR) will be available on immediate open access.
License URL
Publication Status
Accepted