Thomas C. Floyd

Period and amplitude analysis of 0.5–3c/sec activity in NREM sleep of young adults

Abstract A computer program (PANV35) for modified period and amplitude analysis of the sleep EEG is described. This analysis requires prior visual classification of the polygraph record into REM, NREM and waking. Analysis of magnetic tape segments corresponding to NREM sleep is then carried out on a small digital computer. Baseline crossings, wave durations, integrated amplitude and curve length are obtained for specifiable frequency bands. These data are then processed on an IBM 360 system; a number of secondary measures are computed and the data are presented for all NREM sleep, by NREMPs, and for the average 20-sec NREM epoch. PANV35 was employed to analyze delta (0.5–3 c/sec) activity in NREM sleep in 20 young normal males aged 18–22.5 years, half of whom were long-distance runners in training. Measurement was carried out on 2 successive nights of recording which followed one adaptation night. Visually obtained sleep stage scores were also presented. Computer measures of delta showed excellent night-to-night reliability. Internight correlations for mean frequency within 0.5–3 c/sec were exceptionally high, indicating that this measure is an extremely stable individual characteristic. All delta measures showed highly reliable (P Stage 4 showed a significant negative correlation with age. Computer analysis revealed that this change in visually scored delta was associated with a variety of alterations in 0.5–3 c/sec activity: amplitude, total duration, number and density of delta waves declined significantly with age and mean frequency increased. These findings indicate that period and amplitude analysis of NREM sleep yields measures which are both sensitive and stable.

EEG coherence in unmedicated schizophrenic patients

We have recently shown that electroencephalogram (EEG) coherence data recorded with common reference methods, including those obtained from schizophrenics, are confounded by power and phase effects. Three published reports using bipolar recordings found that EEG coherence was higher in schizophrenics; however, only medicated patients were studied. To extend these findings, we measured EEG coherence from bipolar EEG recordings in unmedicated schizophrenics (n = 10), affective disorder patients (n = 8), and normal controls (n = 13) during resting and task conditions. Seven schizophrenics were restudied after a period of neuroleptic treatment. Schizophrenics had higher across-task interhemispheric (p less than 0.05) and intrahemispheric (p less than 0.04) coherence in the theta band and tended to have higher intrahemispheric alpha coherence (p less than 0.08). Medication treatment was associated with clinical improvement and increases in spectral power, but not with changes in coherence values. These results confirm those obtained by earlier investigations and suggest that increased coherence reflects the presence of anomalous cortical organization in schizophrenics rather than medication effects or transient states related to acute clinical disturbance.

Negative symptoms and EEG alpha in schizophrenia : a replication

The authors previously reported that negative symptoms were associated with reduced EEG alpha power and coherence in medication-free schizophrenic inpatients. These post-hoc findings were based on resting EEG data in an eyes open condition. This report describes the replication of these results in a new sample of 17 male veterans (aged 38 +/- 8) recently hospitalized with DSM-III-R schizophrenia. All patients had been free of neuroleptic medication at least 12-14 days. The relationships between resting alpha (7.5-12.5 Hz) power and coherence and symptom ratings (as measured by subscales derived from the Brief Psychiatric Rating Scale) were examined with multivariate repeated measure analyses of covariance. Results were similar to those obtained earlier, with a main effect of negative symptoms (p = 0.05) on log alpha power, a localized effect on right frontal-parietal alpha coherence (p < 0.02), and a main effect (p < 0.03) on between-hemisphere alpha coherence. There was also a trend for an asymmetrical effect on power favoring the right side in parietal leads. Negative symptoms were associated with reduced alpha power and less alpha coherence between hemispheres and between right parietal and frontal regions. The authors discuss the implications of these results on neurodevelopmental, genetic, and attentional aspects of schizophrenia.

Correlation between left frontal phospholipids and Wisconsin Card Sort Test performance in schizophrenia

The relationship between frontal lobe phospholipid measures as measured by in vivo 31phosphorus magnetic resonance spectroscopic imaging and performance on the Wisconsin Card Sort Test was examined in 16 chronic schizophrenic patients and 13 normal controls. Lower left frontal phosphomonoester levels in the schizophrenics were associated with fewer categories achieved, lower percent conceptual level, and greater total errors. No significant correlations between frontal phospholipid measures and performance on the WCST were noted in the controls. The results suggest a relationship between altered left frontal phospholipid metabolism and a specific measure of frontal lobe neuropsychological functioning.

EEG patterns during and following extended sleep in young adults.

Abstract Sleep EEG and eye movement were recorded in 21 normal males aged 20.3–29.3 years beginning at 11:00 p.m. on 3 consecutive nights with 8, 12 and 8 h in bed on adaptation (A), extended (E), and recovery (R) nights, respectively. On the E night, NREMP durations were roughly equal in the first 3 cycles, fell sharply between NREMPs 3 and 5 and were at about the same level in NREMPs 5–6 . These and other data suggest that NREMP durations approach an asymptote in the fifth NREMP. On the R night, NREMP 1–4 durations were not significantly different from the corresponding values on the E night. As expected from the work of others, the increased sleep (and decreased waking) in the 24 h period preceding the R night was associated with a reduction in stage 4 EEG. Surprisingly, this effect was limited to the first NREMP; both visually scored delta sleep and computer measurements of 0.5–3 c/sec activity showed significantly lower values in NREMP 1 on the R night with no differences between E and R nights for NREMPs 2–4 . Analysis of REM sleep confirmed previous observations by Aserinsky (1969) and Verdone (1968). In agreement with the former investigator, we found that REMP durations declined after reaching a maximum in the fourth REMP and that the late REMPs showed an increase in eye movement density. In agreement with both Aserinsky and Verdone, we found that a huge increase in REM sleep above habitual levels when sleep is extended has virtually no effect on the onset, amount or distribution of REM on the following night. In spite of almost 30 years of research since the discovery of REM sleep by Aserinsky and Kleitman, the quantitative relations between the duration of waking and the amount and temporal characteristics of EEG sleep patterns have not been fully established. We believe that the application of computer as well as visual measurement and the analysis of sleep by physiological units (NREMPs and REMPs) will prove useful tools for describing these relationships.

Flurazepam effects on slow-wave sleep: stage 4 suppressed but number of delta waves constant

Repeated administration of flurazepam reduced stage 4 sleep (high delta-wave concentration) but produced a greater increase in stage 2 duration so that total sleep time was increased. Computer analysis revealed that the increased amount of stage 2 (low delta-wave concentration) sleep provided a number and duration of delta waves sufficient to offset the loss of delta activity in stage 4. However, the amplitude of the average delta wave was reduced. These results demonstrate the value of direct quantification of delta-wave activity, the variable that underlies visual classification of slow-wave sleep into stages 2 to 4. They also give rise to new hypotheses regarding the relative absence of side effects in spite of profound stage 4 suppression by flurazepam and the mechanisms by which total sleep time is increased by this drug.

Period and amplitude analysis of NREM EEG in sleep: Repeatability of results in young adults

Abstract Period and amplitude analysis of NREM sleep was carried out in 21 young adults (sample 2) on the first 4 NREM periods (NREMPs) of a night of extended sleep to determine the repeatability of results obtained in a previous group of 20 (sample 1) Ss. The results for mean values for delta (0.5–3 c/sec) variables by NREMP were quite similar in both samples; so also were the trends in these values across NREMPs. In both groups, these trends were linear, highly significant and accounted for high proportions of the within-subject variance. Within-subject, across-night correlation coefficients were comouted in sample 2 for activity summed for NREMPs 1–4. The results confirmed the high reliability of these measures found in sample 1. However, significant reliability was not limited to delta frequencies. Substantial and statistically significant across-night correlation coefficients were found for all except the highest frequency band (>23 c/sec) analyzed by PANV35. We interpret the high reliability of period and amplitude EEG measures during NREM sleep as resulting from the fact that Ss are studied in the same biological state and that extensive measurements are obtained. Intercorrelations among age, NREM sleep stages and computer measures of delta were also computed. The strong correlations of delta measures with age found in sample 1 were not replicated in sample 2. This discrepancy may result from the different age range of the two samples or it may indicate that the initial results were due to chance. Further observations on a new group of Ss in the age range of sample 1 are required. Intercorrelations among the other variables were similar to those obtained in the first study, except for the baseline crossings measure. These results indicate that our method of period and amplitude analysis is reliable and encourage its application in studies of aging and in other conditions where sleep is altered.

EEG asymmetry in schizophrenic patients before and during neuroleptic treatment

Lateral asymmetry of electroencephalographic (EEG) spectra was assessed in schizophrenic patients compared to normal controls. Ten predominantly unmedicated schizophrenic inpatients and nine normal controls performed monitored cognitive tasks during bilateral recording of EEG from parietal and temporal sites. Lateralization of EEG power in five frequency bands was compared between the groups; separate analyses were performed for linked ears and vertex references. A subsample of schizophrenic patients was restudied after a period of neuroleptic treatment. All significant group differences were obtained with the linked ears reference only. Pretreatment schizophrenics manifested relatively less alpha power over the right hemisphere during all conditions than controls, particularly in the parietal leads. After treatment, there was a significant shift in alpha lateralization toward the control values. These latter effects were also present in the theta frequency band to a lesser extent.

Computer-Detected Patterns of Electroencephalographic Delta Activity During and After Extended Sleep

Delta (0.5 to 3 hertz) waves are the electroencephalographic hallmark of human sleep. We measured their rate of production during and following an extended night of sleep. On the extended night, we confirmed previous observations of a linear decline in delta wave production across the first four periods of non-rapid-eye-movement (non-REM) sleep. An asymptote was reached in the fifth non-REM period, perhaps signifying that sleep processes reached completion. On the day after the extended night, subjects were allowed to remain awake 3.6 hours less than normal. During the next sleep session, amplitude and number of delta waves in non-REM periods 1 and 3 were significantly reduced. These findings illustrate the value of computer analysis of electroencephalographic waveforms in sleep. Systematic measurement of the amount and distribution of these waveforms as a function of preceding waking duration should provide clues to the kinetics of the metabolic processes underlying sleep.

Topography of P300 response to target and novel stimuli in schizophrenic patients

To study frontal brain activity we measured event-related potentials (ERPs) to auditory target and novel stimuli in nine schizophrenic patients, five affective controls, and 13 normals. We predicted that schizophrenics, because of impaired frontal lobe function, would not show as much frontal augmentation of the P3 component to infrequent taskirrelevant novel stimuli as would normal subjects. Subjects were instructed to respond (finger lift) to infrequently occurring (10%) target tones (TS) while ignoring frequently occurring (80%) nontarget tones and infrequently occurring (10%) random electronic noise (NS). Ten channel EEG was collected until 30 correct and artifact-free trials to TS and NS were acquired. After transformation to average reference, the peak global field power between 276 and 450 milliseconds after stimulus onset was used to identify P300 latency. Potentials at this latency were compared across groups by repeated measures ANOVA. In agreement with previous studies, maximum positivity to NS shifted anteriorly relative to TS (p < .04). The topography of the P300 to NS varied with diagnosis (p < .007). The largest differences occurred at Cz, where the response to NS was larger than to TS across groups (p = .0003), but with smaller differences in schizophrenics (group x condition, p < .02). In fact, schizophrenics produced the shallowest potential maps (p < .04) and higher positivity in frontal relative to central leads (p < .05). Medicated schizophrenics (N = 5) differed from unmedicated schizophrenics (N = 4) only in prolonged latency (p = .002); there were no other latency differences between groups. These preliminary findings suggest a link between abnormalities of arousal and selective attention with frontal cortex functioning in schizophrenic patients.