Maintaining balance against force perturbations: impaired mechanisms unresponsive to levodopa in Parkinson's disease

Abstract

There is evidence that postural instability associated with Parkinson's disease (PD) is not adequately improved by levodopa, implying involvement of nondopaminergic pathways. However, the mechanisms contributing to postural instability have yet to be fully identified and tested for their levodopa responsiveness. In this report we investigate balance processes that resist external forces to the body when standing. These include in-place responses and the transition to protective stepping. Forward and backward shoulder pulls were delivered using two force-feedback-controlled motors and were randomized for direction, magnitude, and onset. Sixteen patients with PD were tested OFF and ON levodopa, and 16 healthy controls were tested twice. Response behavior was quantified from 3-dimensional ground reaction forces and kinematic measurements of body segments and total body center-of-mass (CoM) motion. In-place responses resisting the pull were significantly smaller in PD as reflected in reduced horizontal anteroposterior ground reaction force and increased CoM displacement. Ankle, knee, and hip moments contributing to this resistance were smaller in PD, with the knee extensor moment to backward pulls being the most affected. The threshold force needed to evoke a step was also smaller for PD in the forward direction. Protective steps evoked by suprathreshold pulls showed deficits in PD in the backward direction, with steps being shorter and more steps being required to arrest the body. Levodopa administration had no significant effect on either in-place or protective stepping deficits. We conclude that processes employed to maintain balance in the face of external forces show impairment in PD consistent with disruption to nondopaminergic systems.

Keywords: Parkinson's disease; balance; dopaminergic pathways; levodopa; postural instability; pull test.

Copyright © 2016 the American Physiological Society.

Figures

Fig. 1.

Apparatus. Participants stood on two force platforms and wore a safety harness. Two motors delivered 1-s-long pulls of different force and direction (backward and forward) via kite strings attached at shoulder level. Participants were asked to resist the pull unless one or more steps became necessary to maintain balance.

Fig. 2.

Whole body in-place responses. A: group mean (n = 16) responses for PD (black lines) and controls (gray lines) to 10-N pulls in the backward direction for session 1 (PD OFF; dashed lines) and session 2 (PD ON; solid lines). Top, perturbation force; middle, center of mass (CoM) displacement normalized by body mass; bottom, anteroposterior (AP) ground reaction force. B: group mean (±SE) AP reaction force measured between 0 and 500 ms after pull onset for PD (squares, black lines) and controls (circles, gray lines). Data from force levels of 10 N (open symbols) and 20 N (filled symbols) are shown separately for each session (S1, OFF; S2, ON) and each direction (backward, forward). C: same conventions as B for the mean normalized CoM displacement measured at the end of perturbation.

Fig. 3.

In-place joint moments. Combined left and right joint moments in the sagittal plane were measured between 0 and 500 ms after pull onset. Moments were normalized by pull force and averaged across 10- and 20-N pulls and across sessions. Group means (±SE) are shown separately for backward (left) and forward (right) perturbations in PD (squares, black lines) and controls (circles, gray lines). Direction of moments is shown as dorsiflexion (DF) or plantarflexion (PF) for ankle and extension (E) or flexion (F) for knee and hip.

Fig. 4.

Stepping threshold. A: example of sigmoid curve using all trials collected in one session from a representative participant for calculation of stepping threshold (before normalization). B: group mean (±SE) stepping threshold normalized by body mass in each session (S1, OFF; S2, ON) for PD (black squares) and controls (gray circles) shown separately for backward (solid lines) and forward (dashed lines) pulls.

Fig. 5.

Stepping responses. A: representative PD and control participants for 29-N backward pull. Top, perturbation force (solid black lines); bottom, AP position of CoM (dashed black), and heels (R, right, solid dark gray lines; L, left, solid light gray lines). Although the pulls were identical and the body displacement (CoM trajectory) and foot-lift time were similar, the PD patient took multiple short steps, whereas the control participant took a single longer step. B: group mean (±SE) length of first step for PD patients (squares, black lines) and controls (circles, gray lines) for each session (S1, OFF; S2, ON) in the backward (left) and forward (right) directions. C: same conventions as B for group mean (±SE) number of steps taken per trial.