Reliability assessment of a marker based methodology for the evaluation of spine mobility in ankylosing spondylitis

Abstract

s / Gait & Posture 37 (2013) S1–S31 S23 points, correspondent to 100% of the gait cycle; 3 gait cycles were selected from each session. Comparison among protocolswas done using parameters: r (Pearson’s correlation coefficient) and (mean difference). For the T3D protocol only kinematic data were used at this stage.Results: Variability in kinematics and kinetics was evaluated. The largest differences were dependent on slow walking speed. SSS variation between sessions was up to 15%. Interestingly, only (O) data showed values of hip internal rotation during stance, both with and without offset correction. Apart from slow speed trials, knee angles show good correlation only for flex/ext (r>0.99, 0.9, 0.9, <0.1W/kg).Discussion:Besides a strongvariation in joint kinematics and kinetics dependent on walking speed, differences were evident in knee ab/ad angles, which reflect in differences in the transverse plane at hip and knee. Careful marker placement improves consistency, but protocol-dependent offsets cannot be avoided. The use of the thigh wand, as alternative to thigh marker placement directly on the skin or on the trochanter, seems to be the only method able to measure hip internal rotation during terminal stance, resulting in larger differences at knee and hip moments during the stance phase. Correction of thigh rotation and tibial torsion appears less significant in affecting hip and knee kinetics. Joint powers, being not affected by joint reference system rotations, showonly speed dependent differences. Reference normality curves should therefore be specific to both the lab and the protocol. Subject’s mood might influence walking speed, even with the same lab settings, with consequent possible speed-dependent modifications in measured kinematics and kinetics.