Table 4.

Multivariate associations of ML COM_RMS with experimental conditions and clinical covariates

Effect	P
Task	<0.001
S3 group	0.15
Task * Group	0.47
Ke	<0.001

Multivariate associations were determined using the mixed-model ANOVA. Covariates include: age, gender, education, gait speed, CES-D-R, sensory loss, TMT-B, daily alcohol use, and fall history. K_e was considered as a covariate only in the ML direction because K_e was not affected by the dual task in AP direction. Dual task-related changes in K_e did not fully explain the changes in COM_RMS with dual task. Thus the biomechanical model does not fully explain the increase in postural sway amplitude due to dual task.