Reference intervals for knee functions specific to outpatients with knee osteoarthritis: a cross-sectional study
Hideyuki Ito1, Tetsuya Amano2, Kiyoshi Ichihara3
1Department of Rehabilitation, Faculty of Wakayama Health Care Science, Takarazuka University of Medical and Health Care, 2252 Nakanoshima, Wakayama, Wakayama-Pref.640-8392, Japan. ito5273@honey.ocn.ne.jp.
Background:
Reference values (RVs) for knee function tests have been reported in perioperative patients with knee osteoarthritis (KOA); however, such values for practical use in outpatient setting has yet to be determined. Therefore, we aimed to establish the reference intervals (RIs) for outpatients with mild to moderate KOA.
Methods:
This cross-sectional study enrolled 202 outpatients with KOA from 8 Japanese orthopedic clinics and measured knee extensor/flexor muscle strength (MS) and knee extension/flexion range of motion (ROM). We used multiple regression analysis to evaluate the sources of variation, including sex, age, body mass index, Kellgren-Lawrence (K-L) classification, bilateral KOA, and exercise habits. Magnitude of between-subgroup differences is expressed as standard deviation ratio (SDR) based on a three-level nested analysis of variance, with SDR ≥ 0.4 as the threshold for requiring RIs specific for subgroups. RIs were calculated parametrically using two-parameter Box-Cox formula if Gaussian transformation of RVs was successful, otherwise calculated nonparametrically.
Results:
Partitioning was required by sex for extensor and flexor MS (SDR = 0.65, 0.57, respectively) and by K-L classification for flexion ROM (SDR = 0.54). RIs were determined parametrically for extensor MS as 0.27-2.09 (male) and 0.27-1.54 (female) Nm/kg and for flexor MS 0.18-1.20 (male) and 0.13-0.79 (female) Nm/kg. On the other hand, RIs for extension and flexion ROM were determined nonparametrically due to discrete nature of their RVs. The RIs determined for extension ROM were -15°-0° and for flexion ROM were 105°-150° (for K-L grade I/II) and 95°-140° (for K-L grade III/IV).
Conclusions:
The ranges of RIs determined specifically for patients with mild to moderate KOA were in-between those of age-matched healthy controls and pre-surgical KOA patients, both of which we had reported for use in physiotherapeutic management of KOA patients undergone total knee arthroplasty. The newly derived RIs will provide an objective benchmark for physiotherapy targeting outpatients with mild to moderate KOA.
