Normative F wave values and the number of recorded F waves.

F waves were recorded from the right abductor pollicis brevis of 11 normal subjects following 100 supramaximal stimuli. F wave data from sequential groups of 10 stimuli were analyzed. "True" values were considered those based on the results following all 100 stimuli. The F wave data evaluated were: minimal, mean, and median latencies; persistences (the percentage of F waves present in a series of stimuli); chronodispersion (CD) (the difference between minimal and maximal F latencies); mean F/M wave (mF/M) amplitudes; and repeater waves. F latencies appear normally distributed, and there is no statistical justification for using either median latencies or amplitudes. Following 10 stimuli, however, mean latencies are more reproducible than minimal values. In these normal data, results following 10 stimuli give latency measurements within 1 ms of "true" as well as for persistences; 20 stimuli will provide mean latencies within 0.5 ms of "true" as well as probable reasonable values (i.e., 80% of "true" or greater, for mF/M and the percentage of repeater waves; for CDs 80% of "true" or greater, 50-60 stimuli are needed; for the number of individual repeater waves, data from all 100 stimuli would be required. Analysis of repeater waves amplitudes would support a preferential activation of larger motor units in F waves.