Reliability and minimal detectable change of knee mechanics during gait and squatting, using markerless motion...

Meaningful changes in joint mechanics within or across sessions require observed differences greater than measurement error. While markerless motion capture systems have achieved comparable gait parameters to marker-based methods, there is a knowledge gap for markerless approaches applied to postures adopted in the workplace in terms of reliability and minimal detectable change (MDC).

To compare within-session reliability, intersession reliability, and MDCs of peak knee angles and moments between walking and squatting in the workplace.

Three-dimensional markerless motion capture and force plates recorded gait and squat trials for twenty-one female early childhood educators [age: 41.6 yrs (10.7), BMI: 29.3 kg/m² (7.7); mean (SD)] before, and immediately following, their work shift. Intraclass correlation coefficients (ICC_2,2) for reliabilities and MDCs were calculated from two trials or averages each session for peak knee flexion and peak knee adduction angles and moments.

Within-session reliability was poor to excellent for gait (ICC_2,2=0.24-0.94) and moderate to excellent for squats (ICC_2,2=0.52-0.99). Intersession reliability was moderate to excellent for gait (ICC_2,2=0.59-0.96) and poor to excellent for squats (ICC_2,2=0.05-0.96). MDCs for peak knee joint angles and moments ranged from 1.6 to 9.9° and 0.09-0.84 Nm/kg for gait, and 3.7 to 10.2° and 0.05-0.24 Nm/kg for squats, respectively.

Within-session and intersession reliability for peak knee outcomes was higher for squatting compared to gait, and peak knee moments had reduced within-session reliability compared to the peak knee angles. Squat peak knee flexion and adduction angles exceeded the 5° MDC threshold, suggesting that larger differences are required to be interpreted as meaningful.