IMU-derived static postural control performance: Validity and reliability in an externally valid stance in persons with multiple sclerosis

Background Monitoring postural control in persons with multiple sclerosis (pwMS) is vital. Inertial measurement units (IMUs) could be a valid alternative to force plates. However, underlying findings often relate to difficult stances with low external validity, which may require an emphasized hip strategy that is uncommon in everyday life. Therefore, this study examined the validity and reliability of IMUs in an externally valid stance. Methods 21 pwMS with mild to moderate disability and 19 healthy controls stood in an open parallel stance on a force plate for 20 s, holding a smartphone at the frontal waistline, twice with open and once with closed eyes. The average velocity (VEL) of the center of pressure (force plate) and the average acceleration (ACC) (smartphone) were tested for reliability and validity. Findings VEL (pwMS: ICC=0.83, p<0.001; controls: ICC=0.71, p<0.001) and ACC (pwMS: ICC=0.78, p<0.001; controls: ICC=0.79, p<0.001) were reliable. ACC did not show effects of group (pwMS vs. controls) (p=0.794), condition (eyes open vs. closed) (p=0.088, trend level ges=0.025), their interaction (p=0.610), or age (p=0.237). VEL revealed both group (p=0.004, ges=0.072) and condition effects (p<0.001, ges=0.215), but no interaction (p=0.566); age was significantly associated (p<0.001, ges=0.106). In both models, adding the respective other sway metric revealed significant associations of VEL and ACC (p=0.018, ges=0.049). Interpretation The within-session reliability of the IMUs was confirmed. However, the validity of IMUs in pwMS with mild to moderate disability in an externally valid stance was not achieved, as the waist-worn IMUs primarily measure hip rather than ankle strategy.