Bilateral soleus H-reflexes in humans elicited by simultaneous trains of stimuli: symmetry, variability, and covariance

Experiments using electrical and mechanical activation of spinal reflexes have contributed important results toward the understanding of neuronal and synaptic dynamics involved in spinal neural circuits as well as their response to different inputs. In this work, data obtained from the simultaneous stimulation of both legs are analyzed to provide information on the degree of symmetry of the respective spinal reflex circuits and on the characteristics of reflex variability. H-reflexes recorded from relaxed muscles show a frequency-dependent amplitude depression when elicited by a train of stimuli. This effect has been attributed to homosynaptic depression. Soleus H-reflexes were recorded in response to trains of simultaneous stimuli applied to both legs in right-handed subjects that were sitting in a relaxed state. The first objective was to verify the existence of asymmetries in H-reflex parameters obtained from the two legs. We measured the mean, variance, and coefficient of variation of the depressed H-reflex amplitudes and the time constant of decay toward the depressed plateau. The second objective was the analysis of the time correlation of subsequent H-reflex amplitudes in a long train of responses recorded from a given leg. The statistical dependence of H-reflex amplitudes in the long trains recorded from both legs was also investigated. Data obtained from preliminary experiments showed that there was no effect of a given stimulus on the contralateral leg applied simultaneously or 1 s before, therefore validating the simultaneous stimulation paradigm. Paired t-tests indicated that several parameters measured bilaterally from soleus H-reflex trains of right-handed subjects were not statistically different in the overall, although individually there were statistically significant asymmetries, toward either the right or left leg. Sequences of H-reflex amplitudes, as measured by the auto-covariance, were either white or had a memory ranging from 2 up to 50 s. This indicates that the random fluctuations in presynaptic inhibition and/or postsynaptic inputs to motoneurons may have either fast or slow time courses. The average auto-covariance sequences of the right and left legs, computed from all subjects, were practically superposable. The cross-covariance between the bilateral H-reflex amplitudes showed a statistically significant peak at zero lag in some experiments, suggesting a correlation between the synaptic inputs to the Ia-motoneuron systems of the soleus muscles of both legs.