On a method to detect long-latency excitations and inhibitions of single hand muscle motoneurons in man

Summary

The peri-stimulus-time histogram (PSTH) analysis of stimulus-related neuronal spike train data is usually regarded as a method to detect stimulus-induced excitations or inhibitions. However, for a fairly regularly discharging neuron such as the human α-motoneuron, long-latency modulations of a PSTH are difficult to interpret as PSTH modulations can also occur as a consequence of a modulated neuronal autocorrelation. The experiments reported here were made (i) to investigate the extent to which a PSTH of a human hand-muscle motoneuron may be contaminated by features of the autocorrelation and (ii) to develop methods that display the motoneuronal excitations and inhibitions without such contamination. Responses of 29 single motor units to electrical ulnar nerve stimulation below motor threshold were investigated in the first dorsal interosseus muscle of three healthy volunteers using an experimental protocol capable of demonstrating the presence of autocorrelative modulations in the neuronal response. It was found for all units that the PSTH as well as the cumulative sum (CUSUM) derived from these responses were severely affected by the presence of autocorrelative features. On the other hand, calculating the CUSUM in a slightly modified form yielded — for all units investigated — a neuronal output feature sensitive only to motoneuronal excitations and inhibitions induced by the afferent volley. The price that has to be paid to arrive at such a modified CUSUM (mCUSUM) was a high computational effort prohibiting the on-line availability of this output feature during the experiment. It was found, however, that an interspike interval superposition plot (IISP) — easily obtainable during the experiment — is also free of autocorrelative features. Both mCUSUM and IISP of all 29 units showed a common reflex pattern in response to the electrical stimulation of low-threshold afferents. The first sizeable effect was a motoneuronal inhibition starting 50–70 ms after the stimulus followed by an excitatory period with an onset latency of 90–110 ms. Thereafter, a second inhibitory period with a duration of approximately 100 ms appeared beginning 140–160 ms after the stimulus. These long-lasting effects with long-latency could not be deduced from the “raw” PSTH or the unmodified CUSUM making it necessary to evaluate long-latency reflexes of single motoneurons by calculation of an mCUSUM and — in case of on-line requirements — by construction of an IISP.