A signal-averaging technique for the analysis of human muscle sympathetic nerve activity

We present a signal-averaging technique for analysis of human muscle sympathetic nerve activity (SNA). Nerve traffic was averaged by coupling signal acquisition to electrocardiographic R waves. The amplitude of the averaged waveform was multiplied by the number of R waves sampled to provide a measure of SNA in arbitrary units. This was compared with SNA measured by manual digitization of hard-copy records. In nine volunteers, SNA was increased or decreased with stepwise infusions of nitroprusside or phenylephrine: there were 10 5-min periods of data in each subject. Across all subjects, the correlation between manual and signal-averaged measures of SNA was excellent during both nitroprusside (r = 0.98) and phenylephrine infusions (r = 0.91) and the slopes of the regression lines were near unity. In three periods of data collection, electrical artifacts were added randomly at frequencies of 0.5 and 0.07 Hz during playback of the signal into the computer. Signal-averaged estimates of SNA were unaffected by artifacts. This technique provides reliable observer-independent measures of SNA.