[Median nerve somatosensory evoked potentials--studies on latency variability as a function of subject height and age].

Somatosensory evoked potentials (SEP) to median nerve stimulation were studied in a group of 61 healthy volunteers (36 men, 25 women) between 15 and 79 years of age. The absolute peak latencies of peripheral N9, Erb, cervical N11 and N13, cortical N20, N32 and N60 were analyzed as a function of height and age. Three interpeak latencies, Erb-N13, Erb-N20 and N13-N20, were equally studied. There was no gender difference in the latencies and interpeak latencies of SEP in multiple correlation analysis involving height and age. Multiple regression analysis showed that except for N60 peak latency, the subject's height and age provided better prediction of other SEP latencies, and Erb-N13 and Erb-N20 interpeak latencies than simple and polynomial [correction of polynominal] linear regression. Multiple regression analysis was found to provide better optimal clinical upper limits for individual SEP latencies and interpeak latencies, and may increase the diagnostic sensitivity. The t-value of the multiple regression coefficient for standardized height and standardized age showed that the effects of age on N32 and N60 peak latencies, and on Erb-N13 and Erb-N20 interpeak latencies were a more important factor than those of height. A negative correlation was found between height and age in our subjects. Although this may be a characteristic of modern Japanese, it suggests the necessity of renewing regression models at intervals of 5 or 10 years.