Within- and between-day reliability of trunk mechanical behaviors estimated using position-controlled perturbations

Recent applications of position-controlled perturbation techniques to the human trunk have allowed separate estimation of intrinsic and reflexive trunk mechanical behaviors. These mechanical behaviors play an important role in spinal stability and have been associated with low back pain risk, yet the reliability of these measures remains unknown. Therefore, the objective of the current study was to assess within- and between-day reliability of several measures of trunk mechanical behaviors obtained from position-controlled trunk perturbations. A secondary objective was to assess if different harness designs, used to connect a participant to the perturbing device, influenced reliability. Data were analyzed from baseline measurements obtained from two previously published studies, and a third unpublished study. The total combined subject pool included 33 healthy young adults (17 M, 16 F). Relative and absolute reliability was quantified using intraclass correlation coefficients (ICCs) and standard errors of measurement (SEM), respectively. Within-day ICCs of intrinsic trunk stiffness (0.84-0.90) and effective mass (0.91-95) were excellent, and were generally higher than ICCs for reflex gain (0.55-0.85), maximum reflex force (0.65-0.85), and timing of maximum reflex force (0.48-0.86). Within-day ICCs (0.48-0.95) were consistently superior to between-day values (0.19-0.72). Improvements in harness design increased both within- and between-day reliability and reduced SEMs for most measures.