Impairments in Sensorimotor Control and Tremor in Individuals With Multiple Sclerosis

Background:

Accurate arm movements are the key to performing many necessary daily tasks. In individuals with Multiple Sclerosis (MS), the processes that control arm movement are disrupted, resulting in tremor and dysmetria. The origins of these motor symptoms are difficult to ascertain. MRI and animal studies suggest that tremor in MS may be related to lesions in the brainstem and cerebellum. However, other studies suggest that some forms of tremor are influenced by visual feedback, and can be alleviated by occlusion of arm position during goal directed movements. 

Objectives:

To distinguish between possible functional causes of tremor, we used systems identification techniques to characterize the neural control of arm movements. These techniques have been used previously to examine movements in healthy subjects, and are here extended to subjects with MS.

Methods:

Eight subjects with MS who exhibited tremor (ages 25-68, 6 women, 7 right-handed), and eight age-matched controls (ages 25-61, 6 women, 8 right-handed) performed a series of compensatory tracking tasks using flexion/extension movements about the elbow joint. They used a 1-D robotic manipulandum to respond to visual and proprioceptive perturbations to the position of a cursor relative to a stationary target that was presented on a computer display. 

Results:

Visual response delays in subjects with MS were significantly longer (p<.05) than control subjects in 6 of the 8 subject pairs. Proprioceptive response delays were not significantly different between the groups (MS: 77±19ms, Control: 81±31ms, p>.05). In control subjects, internal predictions of visual response delay, calculated from the timing of submovements, were matched to the measured delay. In subjects with MS, predicted visual delays were not significantly different (p>.05) from those of control subjects, resulting in a mismatch relative to their long visual response delays. Model fits of subject data also revealed a mismatch between actual and predicted limb kinematics (poor internal estimates of limb inertia, viscosity, and stiffness) in subjects with MS, the degree of which increased with tremor score. In simulations of some subjects' model parameters, mismatched estimates of inertia, viscosity, and stiffness were able to stabilize the system, compensating for instability and reducing tremor caused by inaccurate predictions of visual delay.

Conclusions:

The results of this study suggest that intention tremor is related to mismatches between subjects’ actual and predicted delays in visual feedback that may be compounded in some cases by a maladaptive effort to reduce instability. Rehabilitation strategies may therefore be improved by targeted retraining of adaptive mechanisms to more effectively compensate for visual delay mismatches.