Background: Few tools evaluates quality of movements in individuals with knee osteoarthritis (OA). The Test for Substitution Patterns (TSP) is developed to measure the ability to perform five functional movements regarding postural control and altered movement patterns (1). TSP is validated and reliable in individuals with anterior cruciate ligament injury, but has not yet been evaluated in individuals with knee OA.

Objectives: To study the relationships between the OA modified TSP (OA-TSP) and self-reported knee function as measured with the Knee Injury and osteoarthritis Outcome Score (KOOS) and the 30-s chair stand test (30-s CST) in individuals with symptomatic knee OA. A second aim was to study the discriminative ability of the OA-TSP for unilateral knee pain.

Methods: Sixty-two individuals with symptomatic knee osteoarthritis were included using consecutive sampling. Health status was assessed with the EuroQol five dimension scale (EQ5D, 0-1 worst-best), and knee function in five subscales for KOOS (pain, symptoms, ADL, quality of life and sport/recreation, 0-100 worst-best). The 30-s CST-test measured the number of rises in 30 seconds. In the OA-TSP, substitution patterns are observed and scored from 0-3 (no substitution pattern-poorly performed) during five standardized functional movements. The maximum score is 54 points/side with score of 108 points. Median and min-max were used for all descriptive data. Spearman´s correlation and Wilcoxon signed rank test were used for analyzes. A correlation coefficient r_s ≥±0.50 is considered large, ±0.30 to < 0.50 moderate and ±0.10 < 0.30 small.

Results: The median age was 54 years (30-61), 76% were women. The median Body Mass Index was 25 (18-48) and EQ5D 0.8 (0.29-1.00). There were no significant differences between the gender regarding BMI and EQ5D. Median OA-TSP total score was 29 (10-70). Median KOOS pain was 75 (36-100), symptoms 71 (21-96), ADL 87 (30-100), and sport/rec 50 (0-100). In the 30-s CST the median was 16 raises (5-32). Moderate, significant correlations were observed between TSP total score and KOOS pain and KOOS ADL (r_s=-0.30; p=0.03, r_s=-0.35; p=0.01 respectively) and small correlations between TSP and KOOS sport/recreation and KOOS symptoms (r_s=-0.13; p=0.36, r_s=-0.22; p=0.11 respectively). There was a moderate, significant correlation between TSP total score and 30-s CST (r_s=-0.34; p<0.01). Discriminative ability for the TSP for unilateral knee pain was found to be significant worse in the painful side, with median 18 (2-36) vs. 14 (7-37) in the not painful side, p=0.001.

Conclusion: The OA-TSP could be used as a functional test to detect altered knee alignment interpreted as an early sign of knee OA and assist the physiotherapist in functional testing during the rehabilitation of individuals with symptomatic knee OA.