PROPRIOCEPTION IS "CENTRAL" TO MOTOR SEQUENCE LEARNING: DIFFERENT CONSEQUENCES OF PERIPHERAL AND CENTRAL PROPRIOCEPTIVE DISRUPTION TO SEQUENCE LEARNING

It has long been a goal of the human movement research community to develop a more complete understanding of how individuals learn complex patterns of movement. This knowledge is useful to both those engaged in precision activities such as the hitting a baseball as well as those determined to improve motor function following neurologic injury such as a stroke. The following dissertation explores one component of the motor learning process: the role of sensation, specifically proprioception, in motor learning. Two questions were addressed. First, does peripheral disruption of proprioceptive sensation impair sequence-specific learning? Second, what is the relationship between sequence-specific learning and stroke related proprioceptive deficit? Because little is known about the role of proprioception in learning movement patterns the first aim of this study, presented in Chapter 2, sought to characterize skill acquisition in healthy adults when proprioceptive sensation was disrupted. Twelve participants (DIS) were randomly assigned to receive vibration to the upper arm while performing a continuous tracking task. Twelve additional subjects (CTL) served as a comparison group and received vibration to the arm that was not being used to track. Despite initially less accurate performance by the DIS group, both groups learned to track a repeated sequence of movements, indicating that accurate peripheral report of limb state may not be crucial for motor learning. Rather the motor learning system appears to be robust and capable of utilizing other sources of information for error correction and motor plan development. Vision is one likely source of information used by the central nervous system (CNS), concomitant with or in place of proprioception. Previous work has consistently demonstrated a close link between visual and manual motor performance. Further, the visual system is capable of learning stimulus regularities. However, visual and manual learning during continuous sequence tracking has not been investigated previously. To address this gap in the literature, 9 healthy young adults performed a continuous tracking task similar to that described in Chapter 2. Eye movements were recorded during manual tracking. It was found that though performance improved for both visual and manual tracking over practice, eye and motor accuracy were relatively independent. Tracking performance of either effector (i.e., arm versus eye) on individual trials was not predictive of performance on the other effector. Thus, reflecting somewhat the findings on proprioception in Chapter 2, it was concluded that though vision may be an important component of sensory information, a high degree of visual tracking accuracy is not critical for continuous motor sequence learning. In Chapter 4, the focus shifted to proprioceptive disruption following central neurological insult. Ten survivors of stroke and 9 age-matched, neurologically intact control participants engaged in a continuous tracking task similar to that presented in Chapter 2. However, no exogenous sensory disruption was applied during training. Instead, sensory disruption was an endogenous consequence of stroke. The level of central proprioceptive disruption was indexed using a limb-position matching task. The healthy participant group was able to learn sequence-specific components of the task. As a group, stroke survivors also demonstrated continuous sequence learning. However, proprioceptive accuracy was strongly correlated with sequence-specific versus general task improvement in continuous tracking within this group. These data support the animal literature that suggests central somatosensory integrity is critical for novel skill learning. Finally, the cerebellum is often identified as a substrate of motor learning. Furthermore, the cerebellum has extensive access to ascending afferent somatosensory information. To examine the use of proprioceptive information by the cerebellum, 7 individuals with cerebellar damage were recruited to perform the same continuous tracking task. Contrary to previous reports, some individuals in this study learned to perform a sequence of movements with their more involved hand. However, two exhibited no such learning. These individuals had large lesions encompassing the lateral cerebellar cortex. They also demonstrated proprioceptive deficit and motor ataxia. These results suggest that learning of movement sequences may be in part related to somatosensory discrimination and error-feedback learning performed in the cerebellar cortex. In summary, the results suggest that though body state information may be important for learning, central and peripheral disruption of proprioceptive information are not analogous in terms of the impact on continuous motor sequence learning. Rather, in addition to proprioception, other sources of afferent information may be used to learn a new skill. In some situations these sources of information, including vision and planning, can operate and encode movement sequence components independent of each other, resulting in skill learning. However, integrity of central information processing systems appears to be an important component of continuous sequence learning. Rehabilitation professionals should consider the state of a client's proprioceptive system when developing a plan of care