Neurophysiological methods for studies of the motor system in freely moving human subjects (Fragment)

Charles Capaday. Centre de recherche en Neurobiologie, Uni6ersite´ La6al, Que´bec City, Canada.


In this paper, the following experimental methods for studies of the motor system in freely moving human subjects will be considered: (i) eliciting the H-reflex and understanding its use as a test response, (ii) methods to measure reciprocal inhibition between antagonist muscles, (iii) methods to measure presynaptic inhibition of Ia-afferent terminals in the spinal cord, (iv) certain aspects of the interpretation of peri-stimulus time histograms (PSTH) of single motor unit discharge, and finally, (v) stimulation of the motor cortex and the measurement of response parameters that may reflect task dependent changes. Two closely related ideas bearing directly on these methods will be emphasized—the influence of the background level of motor activity on input–output properties of the neural pathway investigated and the operating point on the input–output curves at which the experimental variable is measured. Finally, in the discussion a simple model that is easily understandable in geometric terms is presented to help predict and interpret the outcome of these sorts of experiments. Introduction The study of the motor system in freely moving humans during natural motor activities, such as walking, running and postural adjustments allows for an understanding at the systems level of how neural circuits actually work. For example, the input–output properties of several spinal cord neural circuits have been shown to change as a function of the motor task (see Stein and Capaday, 1988 for stretch reflexes; Evans et al., 1989 for cutaneous reflexes). These changes are quantitative in nature, i.e. they involve changes of the parameters of input–output relations. It has been suggested that they serve to adapt the motor system to the biomechanical requirements of the task (Stein and Capaday, 1988; Capaday, 1995). Additionally, it may be possible that neural circuits subserving motor control are modified in a more fundamental manner, that is atthe level of the architecture of the circuit (Pearson, 1985; McCrea, 1994; Pearson and Collins, 1993; Capaday et al., 1995). It can be argued that the basic principles of human motor control are best determined by studying freely moving subjects during natural motor tasks, because the motor system is inherently designed to control such tasks. Finally, this research has clear applications to understanding the basis of pathophysiological conditions following damage to the CNS (e.g., spasticity following stroke or spinal cord injuries). This article contains a review of several neurophysiological methods that are commonly used to study the mechanisms of motor control in humans. I will specifically address how these methods can be used to study the neural basis of motor control in freely moving human subjects. These methods offer clear advantages that will be duly described in each section; but as with any scientific method, the limitations must also be clearly understood. Two ideas will be emphasized throughout: (i) the importance of controlling the background level of motor activity upon which a particular measurement is made and (ii) the need to understandthe nature of the input–output properties of the pathway under study. Much of the theoretical and experimental foundations of these methods go back to the pioneering studies of Renshaw (1940), Lloyd (1941), and Lundberg (1966; 1970; 1975). The reader should refer to these classic papers as an introduction to the physiological basis of the methods described herein. Sherrington’s seminal discoveries and conceptual influence must also be acknowledged (Sherrington, 1906). It must be recognized, however, that the initial development of these methods was done on reduced animal preparations in what may be called the ‘resting state’ (i.e., lacking background motor activity). Therefore, proper application of these methods in behaving human subjects and interpretation of the results require particular attention to the above two propositions (Capaday et al., 1990).