Examination of muscle strength and pressure pain thresholds in knee osteoarthritis: test-retest reliability and agreement

Abstract

Background and purpose: Knee osteoarthritis (OA) is associated with reduced muscle strength and pain sensitization. The purpose of this study was to determine intrarater reliability and agreement (measurement error) of isometric knee extensor and flexor muscle strength assessed using handheld dynamometry and of pressure pain thresholds (PPT; a measure of pain sensitization) from the knee, the leg, and the forearm assessed using handheld algometry in knee OA.

Methods: A total of 20 subjects with knee OA participated in 2 sessions separated by 1 week. The highest of 4 examinations and the mean of the 3 highest examinations of muscle strength and the first and the mean of 2 PPT examinations were applied in the statistical analyses. Intrarater reliability was assessed using a 2-way random-effects model, consistency-type intraclass correlation coefficient, whereas agreement was assessed using 95% limits of agreement (LOA) as a percentage of the mean (LOA%).

Results: Intraclass correlation coefficients for muscle strength were between 0.78 and 0.91 when using the highest examination and were between 0.86 and 0.94 when using the mean of the 3 highest examinations. Intraclass correlation coefficients for PPT were between 0.53 and 0.87 when using the first examination and were between 0.84 and 0.91 when using the mean of 2 examinations. Agreement (LOA%) for muscle strength ranged from 38.3% to 47.3% when using the highest examination and from 40.4% to 53.3% when using the mean of the 3 highest examinations. Agreement for PPT ranged from 54.2% to 80.6% when using the first examination and from 50.6% to 58.9% when using the mean of 2 PPT examinations.

Discussion: A tendency toward improved intraclass correlation coefficients and LOA% (only for PPTs) was found when using the mean of more than 1 examination for both muscle strength and PPTs. Intrarater reliability was high to very high, whereas the LOA/LOA% indicated relatively high measurement errors.

Conclusions: Examination of muscle strength and PPTs in knee OA is reliable, but affected by the measurement error, which is important to consider when reporting the results of clinical trials and in clinical practice.