A Modified Method for Scoring Slow Wave Sleep of Older Subjects

Transcription

1 Sleep, 5(2): Raven Press, New York A Modified Method for Scoring Slow Wave Sleep of Older Subjects Wilse B. Webb and Lewis M. Dreblow Department of Psychology, University of Florida, Gainesville, Florida Summary: The slow wave sleep of two nights' recording of 40 older subjects (50-60 years) was scored visually. Criteria that eliminated an amplitude requirement were used. The slow wave sleep approximated that of younger subjects, and the scoring reliability by independent scorers was in the 0.80 to 0.90 range. Key Words: Sleep stages-slow wave sleep-older subjects Scoring criteria. The standard scoring criteria (1) for non rapid eye movement (NREM) sleep stages 3 and 4, or slow wave sleep (SWS), specify the presence of' 'waves of 2 cps or slower which have amplitudes greater than 75 IL V from peak to peak." An alternative system (2,3) sets criteria of "1-3 Hz" and "40 IL V peak to peak activity." The application of these criteria has led to a consistent reporting of a sharp diminution of SWS associated with aging (4-8). However Agnew et al. (4), noting the absence of any marked "biochemical, physiological, psychological or behavioral" changes in their population and a clear decline in the amplitude of their electroencephalographic (EEG) records, cited the possibility that the "measurement (of SWS) no longer reflects the process. " Agnew (9), using an amplitude integrator system on five groups of subjects ranging from months to years, clearly established the decline in EEG amplitude as a function of age. Utilizing a sleep analyzing hybrid counter system (10) that permits independent analysis of amplitude and frequency characteristics, Smith et al. (11) reported a significant decline in amplitude of delta sleep but no decline in frequency-defined delta sleep. Feinberg et al. (12) reported recently on a computer-based analysis of delta sleep (0.5-3 Hz) on a large (n = 48) sample of healthy older subjects (ages ; mean = 71.8). They affirm the sharp reduction in amplitude in comparison with a younger population (33%). In addition, they report a 13% reduction in frequency of waves and a slowing of the frequency (mean = 1.67 versus 1.51) Accepted for publication November Address correspondence and reprint requests to Wilse B. Webb, Ph.D., Room 101 SSRB, University of Florida, Gainesville, Florida

2 196 W. B. WEBB AND L. M. DREBLOW within the frequency band. They conclude ihat "... a major factor underiying the decline in visually scored stage 4 sleep is a reduction in the amplitude of delta waves in the elderly." The findings of both Feinberg et al. (12) and Smith et al. (11) indicate the substantial presence of SWS, when defined by a frequency criterion as well as a sharp reduction in amplitude, in older subjects. It is clear that the reported sharp decline in NREM sleep stages 3 and 4 in visually scored records could be a function of the combined amplitude and frequency criteria and the established decline of amplitude with aging. This is a study of the visual stage scoring of the all night EEG records of older subjects (50-60 years) that makes use of modified scoring criteria, based on frequency, for stages 3 and 4. The analyses focus on the reliability and amount of stages 3 and 4. METHODS Forty male subjects between the ages of 50 and 60 years (mean = 55.3) were selected from a population of career service employees and salaried professionals at the University of Florida after their questionnaire responses indicated normal sleep patterns, routine medical histories, and the absence of recent physical ailments. For each of four consecutive nights, two subjects at a time reported to the laboratory approximately 1 ~ h prior to their normal bedtimes (about 11 p.m.) for electrode placement. The International system was employed to locate the sites F 11F 7, P lit 5, and CJ A2 for three channels of the EEG. Two additional sites were located 1 cm above the outer canthus of the left eye and 1 cm below the outer canthus of the right eye for one channel of an electrooculogram (EOG). At the completion of the pre sleep regimen (11 p.m.), subjects retired to separate, soundproof, temperature-controlled rooms. Continuous EEG and EOG recordings were obtained for each subject by use of a Grass model VI electrophysiograph. Data from nights 2 and 3 were used for this report. Thirty-nine records were available for night 2 and forty records for night 3. Scoring procedures. The original scoring procedures used in this laboratory (2,3) for stage 3 and stage 4 sleep are based on I-min scoring epochs. The salient excerpts are as follows. Stage 3: "An epoch... contains at least 13 seconds of one through three cycles per second of 40 microvolts or better activity but less than 30 seconds... The frontal trace... will typically show obvious slow activity with a faster cps background. Frequently this background activity is obviously spindling and in other cases it is of a lower three to five cps variety... " Stage 4: "An epoch contains at least 30 seconds of one to three cycles per second 40 microvolt or better activity... the experienced scorer, once he is sure at least 30 seconds of one to tbree cps dominant activity exists in several successive epochs, can rapidly score stage 4 epochs by making sure the first 30 seconds is one to three dominant... then the remaining thirty seconds can be rapidly scanned." These instructions were modified as follows. "Stages 3 and 4 will be scored without consideration of amplitude. You will detect its presence early in the Sleep, Vol, 5, No, 2, 1982

3 SLOW WAVE SLEEP SCORING 197 record on the basis offrequency (0.5-3 cps) and that will serve as your criterion." These instructions simply summarized the orientation for the scoring of each record. The emphasis was on carefully attending to and counting the frequency of waves in the record until the slow wave pattern of that record had been detected (usually between 10 and 40 min after sleep onset). This pattern then served as a guideline in the detection of the frequency-based response for the remainder of the record. Figure 1 shows a lo-s segment of a difficult low amplitude record. The scorers scored each record independently. The sleep stage scores of an entire night were compared by a supervisor. If the percentage of agreement of minutes of all stages of a record was less than 90%, the records were rescored independently and a comparison was made. If 90% consensus was not reached after rescoring, the record was scored a third time. RESULTS The correlations between the stage 3, stage 4, and SWS (stages 3 + 4) scores of scorers 1 and 2 from nights 2 and 3, as well as their means and SD, were obtained independently. Neither the separate night correlations nor the means were significantly different from each other, and the data across the two nights were collapsed. Table 1 presents the data obtained from the first independent scoring of the records. Nine of the records from night 2 and eight of. the records from night 3 did not meet the 90% agreement criterion for the total minutes of all sleep stages across the night. These records were independently rescored. One record of each night required a third rescoring. The correlations, means, and SD which utilized the final scorings also are shown in Table 1. Record Number Year old Male Fl - F7 Eye - Eye 50uV FIG. 1. A record of a 58 year old subject displaying SWS with reduced amplitude. Sleep. Vol. 5, No

4 198 W. B. WEBB AND L. M. DREBLOW TABLE 1. Correiation between scorers and their mean scores with standard deviations Stage 3 Stage 4 SWS Scorer ± ± ± ± ± ± 24.8 Scorer ± ± ± ± ± ± 25.9 Correlation Mean scores ± SO are in min. The first figure in each category is taken from the first scoring of the records. The second figure reflects the rescoring of 17 records that did not originally meet a 90% agreement criterion for all stages. DISCUSSION The substantial correlation coefficients of Table 1 show that stage 3 and stage 4, in the records of older subjects, can be visually scored reliably by use of modified criteria that attenuate the amplitude requirement. Including the most difficult records (less than 90% agreement), stage 4 and SWS reliabilities approximated Those records (n = 62) with initial 90% total agreement had reliability coefficients of 0.79, 0.89, 0.90 for stages 3, 4, and SWS, respectively. Rescoring of the difficult records resulted in reliability coefficients approximately 0.90 for all records (Table 1). The absolute levels of the scores are scorer-dependent. Scorer 2 was significantly lower than scorer 1. However, although the differences were statistically significant, it should be noted that the bias resulted in approximately 3, 5, and 8 min difference in the population means for stages 3, 4, and SWS, respectively. We may use the SEM to estimate the lower limits of SWS. Multiplying the SEM by 1.96 and using the lower scores of scorer 2, there is a 95% probability that the true means of stages 3, 4, and SWS fall between 19.3 and 20.9 min, 47.8 and 53.4 min, and 66.9 and 72.8 min, respectively. As the mean sleep time of our subjects was 435 min, the higher sleep stage percentages are 5% and 12% for stage 3 and stage 4 sleep. These approximate closely the scores obtained with our earlier young adult population (13), which were 7% and 13%, respectively, for stage 3 and stage 4. We conclude that the previous reports, based upon visual scoring, of substantially reduced stages 3 and 4 in older subjects reflect scoring criteria that include an amplitude criterion. However, stages 3 and 4 can be scored reliably by modified criteria that emphasize frequency and such scoring results in the substantial presence of stages 3 and 4 in older subjects. We have, then, an age-sensitive measure and an age-resistant measure. Until other functional correlates of these measures are determined, clearly neither can be declared "better" or more "accurate." It would seem desirable to recommend that the records be scored by both measures. However, our experience makes us doubt this possibility. (We asked our scorers to attempt rescoring using the standard amplitude/frequency criteria for comparison purposes but they could not.) We do recommend the use of our Sleep, Vol. 5, No.2, 1982

5 SLOW WAVE SLEEP SCORING 199 modified criteria in low amplitude records for the assessment of SWS. In each such use, however, they should be identified as frequency defined to distinguish them from the standard measures of stages 3 and 4 (1) and the Florida system (2,3). Acknowledgment: This work was supported by National Institute on Aging Grant 5ROl AG The effort of the scorers, Alton R. Howard and Greg K. Graham, is also gratefully acknowledged. REFERENCES 1. Rechtschaffen A, Kales A (eds). A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Brain Information Service/Brain Research Institute, University of California at Los Angeles, Agnew HW, Webb WB. Sleep stage scoring. Journal Supplement Abstract Service, American Psychological Association, Ms. #293, May, Williams R, Karacan I, Hursch C. Electroencephalography of human sleep: clinical applications. New York: John Wiley & Sons, Agnew HW, Webb WB, Williams RL. Sleep patterns in late middle aged males: an EEG study. Electroencephalogr Clin Neurophysiol1967; 23: Feinberg I, Carlson V. Sleep patterns as a function of normal and pathological aging in man. Arch Gen Psychiatry 1968; 18: Kahn E, Fisher C. The sleep characteristics of the normal aged male. J Nerv Ment Dis 1969; 148: Kales A, Wilson R, Kales J, Jacobson A, Paulson M, Kollar E, Walter RD. Measurement of all-night sleep in normal elderly persons: effects of aging. JAm Geriatr Soc 1967; 15: Prinz P. Sleep patterns in the healthy aged: relationship with intellectual function. J Gerontal 1977; 32: Agnew HW. Integrator analysis of the sleep electroencephalogram. Electroencephalogr Clin Neurophysiol 1973; 34: Smith JR, Karacan I. EEG sleep stage scoring by an automatic hybrid system. Electroencephalogr Clin Neurophysiol 1971; 31: Smith J, Karacan I, Yang M. Ontogeny of delta activity during human sleep. Electroencephalogr CUn Neurophysiol 1977; 43: Feinberg I, Fein C, Floyd TC, Aminoff MJ. Delta (0.5-3 Hz) EEG waveforms during sleep in young and elderly normal subjects. Adv Sleep Res, in press. 13. Williams R, Agnew HW, Webb WB. Sleep patterns in young adults: an EEG study. Electroencephalogr Clin Neurophsyiol 1964; 17: Sleep. Vol. 5. No