212 Elec troencephalography and Clinical Neurophysiology, 1980, 48 : 212--221 © Elsevier/North-Holland Scientific Publishers, Ltd. PERIOD AND AMPLITUDE ANALYSIS OF NREM EEG IN SLEEP: REPEATABILITY OF RESULTS IN YOUNG ADULTS 1 I. FEINBERG z, G. FEIN and T.C. FLOYD Veterans Administration Medical Center, San Francisco, Calif. 94121, and University of California at San Francisco, San Francisco, Calif (U.S.A.) (Accepted for publication: May 21, 1979) In a previous report, we described our methods for computer analysis of EEG wave periods and amplitudes and we applied these methods to the measurement of delta EEG (0.5--3 c/sec) in non-REM (NREM) sleep (Feinberg et al. 1978). Results were presented for 20 young adult men (median age, 19.7 years; age range, 18--23 years). Ten of these subjects (Ss) were long distance runners and 10 were age-matched, relatively sedentary, controls (see Walker et al. 1978 for details). Ss had been studied for 2 consecutive baseline nights, which immediately followed one adap- tation night. Here we compare the findings in this first sample of young adults with those obtained in a second, slightly older group, who participated in a study of sleep cycle pat- terns during and after extended sleep. Materials and methods The 20 Ss in the first (exercise) study are designated as sample 1. Sample 2 consisted of 19 adult male medical students and 2 under- graduates. They were studied for 3 consecu- tive nights: an adaptation night (in bed from 11:00 p.m. to 7:00 a.m.), a night of extended sleep (in bed 11:00 p.m. to 11:00 a.m.) and i This work was supported by research funds from the Veterans Administration. 2 Current address: Psychiatry Service, New York Veterans Administration Medical Center, First Ave. at East 24th Street, New York, N.Y. 10010, U.S.A. To whom reprint requests should be addressed. recovery night lin bed 11:00 p.m. to 7:00 a.m.). Thus, 'lights out' occurred at the same point in the circadian cycle each night. These Ss ranged in age from 20.3 to 29.3 years, with a median age of 24.0 years of age. Sample 2 was significantly older than sample 1 ~t = 8.28, P < 0.0001). Whereas only 2 Ss in sam- ple 1 were over 21 years of age, 19 Ss in sam- ple 2 exceeded this age. Recording of EEG and eye movement was carried out as previously described (Feinberg et al. 1967). NREM sleep stages 2--4 were classified according to the criteria of Recht- schaffen and Kales (1968). The C3 A1 EEG lead was recorded on a Vetter Model A tape recorder at 15/16 in. sec. The C3 A1 lead was analyzed off-line at 4 times recorded speed with a minicomputer (Digital Equipment Corp., PDP-12). Our pro- gram for wave form analysis, PANV35, was described in Feinberg et al. (1978). Briefly, it analyzes EEG waves in specifiable frequency bands on the basis of both baseline crossings and zero first derivative counts. Only results from the baseline crossing analysis will be reported here. This analysis measures the number of half-waves, the time they occupy, their integrated amplitudes and their curve lengths (sum of peak-trough amplitudes). This output as well as an identifying time-code is recorded on LINC tape for each 20 sec epoch. These data are transferred to industry-com- patible 9-track tape and subsequent analysis is carried out on an IBM 370 system. Seg-