Interictal quantitative EEG in epilepsy.

The interictal EEG is often normal in epilepsy patients, particularly with partial seizures of extratemporal origin. Quantitative techniques of EEG analysis may increase the yield of diagnostic abnormality in such patients. Thirty patients with partial seizures of frontal or temporal origin had EEG recorded from left frontal (F7-C3), right frontal (F8-C4), left posterior (T5-O1), and right posterior (T6-O2) derivations. Four-second epochs were used to compute power in the delta (0.25-4.0 Hz), theta (4.25-8.0 Hz), alpha (8.25-13 Hz), and beta (13.25-30 Hz) bands. The ratio of high (8.25-30 Hz) to low (0.25-8 Hz) power on the left and the right was measured, as was the ratio between the left and the right hemisphere total power. The mean frequency deviation in the alpha band between the left and the right hemispheres was also measured, and spectral mobility was determined in the right and the left frontal regions. These values were also calculated in normal subjects and tension headache patients with normal EEGs. Seizure patients with abnormal interictal EEGs had decreased ratios of high to low power, greater asymmetry of total power and alpha frequency, and reduced spectral mobility on the side of their EEG foci. Epileptics with normal interictal EEGs had lower ratios of high to low power, greater alpha frequency asymmetry, and lower spectral mobility than did headache patients or normal controls. Power and frequency measurements, and determination of spectral measures such as mobility, can be done with commercially available digital EEG equipment. They may demonstrate otherwise obscure asymmetries in the interictal EEG and thereby aid in epilepsy diagnosis and classification.

Multiple sleep latency tests in epilepsy.

Seizure patients often complain of sleepiness or disturbed sleep. Although susceptible of medication effect, the multiple sleep latency test (MSLT) may quantify daytime sleepiness and help to establish whether qualitative sleep disturbance accompanies epilepsy. In order to measure daytime sleepiness in epilepsy patients, 30 patients with newly diagnosed or presently untreated complex partial seizures had MSLT after an overnight sleep EEG that showed no sleep deprivation or nocturnal seizures. Four 20-minute naps were undertaken at 09:00, 11:00, 13:00, and 15:00, and sleep latency was recorded along with 8 channels of EEG. Twenty of 30 seizure patients reported subjective sleepiness. Eight patients had average sleep latencies less than 8 minutes, and 3 had latencies less than 5 minutes. No sleep onset REM or respiratory disturbance was noted. Twenty-five patients had EEG abnormalities but none had ictal seizures. Right temporal epileptiform activity correlated with sleepiness. MSLT may quantify sleepiness in epilepsy patients, which is common but may be subjective or psychophysiological. Some patients with partial seizures have persistent daytime sleepiness independent of medication, possibly related to residual medication effects or non-specific effect of their epileptogenic foci.

Neurophysiologic studies in nightmare sufferers.

Nightmares have long attracted neurologic and psychiatric attention, yet little is known of their pathophysiology. We recorded 17-channel electroencephalograms (EEGs), brainstem auditory evoked potentials (BAEPs), long-latency auditory event-related potentials (AEPs), and overnight cassette sleep EEGs (AEEGs) in 10 individuals with recurrent nightmares. They were all nocturnal sleepers, took no medications, do not abuse alcohol or drugs, and had no known medical or psychiatric illnesses. Five patients were being evaluated for other complaints, 3 reported disturbed nocturnal sleep and daytime sleepiness, and 2 sought attention chiefly for nightmares. All 10 patients had normal EEGs and BAEPs. BAEP latencies did not differ significantly from control subjects. Latencies and amplitudes of AEPs were not significantly different in nightmare sufferers and controls, but the former had higher amplitude N100, P160, and N200. Those patients with sleep complaints had on overnight AEEG, less sleep, decreased slow-wave sleep, and more awakenings than those without sleep complaints, but nightmares did not occur during the AEEG recordings.

Auditory evoked potentials in Gilles de la Tourette syndrome.

Gilles de la Tourette syndrome (TS) has been increasingly studied neurophysiologically as well as clinically. Obsessive-compulsive disorder (OCD) and attention deficit disorder (ADD) have been recognized to be part of the continuum of TS. We recorded brainstem auditory evoked potentials (BAEPs) and long-latency auditory event-related potentials (ERPs) in 20 patients with TS, 10 of whom had ADD and 6 OCD. TS patients with and without OCD and ADD did not differ in BAEP latencies, and no differences were found from normal controls. AEP latencies did not differ between TS patients and controls. TS patients with ADD had longer N100 and N200 latencies than TS patients without ADD, and TS patients with OCD had shorter N200 and P300 latencies. These findings suggest that TS is neurophysiologically heterogenous, and that TS patients with OCD or ADD may differ from those without.

Computerized EEG frequency analysis in Gilles de la Tourette syndrome.

EEG abnormality has been reported in Gilles de la Tourette Syndrome but not confirmed in later studies. We carried out computerized EEG frequency analysis in 30 patients with the disorder, using Nicolet Pathfinder II frequency analysis software, versions 1.2 and 3.1 EEG was recorded from 01-A1+A2, 02-A1+A2, Fz-A1+A2, F7-C3, F8-C4, T5-01, and T6-02 in Tourette Syndrome patients and controls. Controls were taking no medications, and drug therapy for Tourette Syndrome had been stopped or not yet initiated in the patient group. Modal alpha frequency (MAF), maximal alpha frequency (MxAF), and spectral edge frequency (SEF) was measured in occipital and frontal derivations in 24 patients and controls. Left frontal (MOLF) and right frontal (MORF) mobility was calculated in F7-C3 and F8-C4 in 21 patients and controls. No significant differences were found between Tourette Syndrome patients and controls by two-tailed t-test. These findings are in accord with recent evidence of little or no EEG abnormality in Tourette Syndrome patients as compared to normals.

Sleep spindles in epilepsy.

Spindles are a ubiquitous phenomenon in sleep, but their physiology and the effects of neurologic disorder on their frequency and amplitude are incompletely understood. We compared the incidence of three commonly defined spindle types (14-15 Hz, 12-13 Hz, and 10 Hz) and the frequency and amplitude of spindles during Stage II sleep in 50 patients with complex partial, partial and secondarily generalized, and primary generalized seizures, with and without interictal behavioral symptoms. All patients had 12-13 Hz and 14-15 Hz spindles of symmetric character in C3-A1 and C4-A2 derivations during prolonged sleep-deprived EEG recordings, which were normal except for partial or generalized epileptiform activity. Seventy-one per cent of complex partial seizure patients had 10 Hz spindles, and they occurred in 50% of the other two groups, predominating among those with interictal behavioral symptoms in all groups. Spindle frequency was significantly less in patients with generalized epilepsy than with partial seizures, and patients with complex partial seizures and partial seizures with secondary generalization differed significantly in spindle frequency. Spindle frequency was significantly lower with polypharmacy than with monotherapy. Patients whose regimens included phenobarbital had significantly lower spindle frequencies and spindle frequencies differed significantly between phenytoin and carbamazepine. Differences in spindle frequency may be due to residual medication effects, underlying encephalopathy or physiological differences between partial and generalized epilepsy.

Auditory evoked potentials in anxiety disorder.

The pathophysiology of anxiety has received much recent attention. EEG findings in anxiety are nonspecific, and some changes in psychophysiological measures have been reported. We recorded short-latency brainstem auditory evoked potentials (BAEPs) and long-latency auditory event-related potentials (AEPs) in 12 patients with generalized anxiety disorder. All 12 patients had BAEP latencies within clinical norms, but I-V interpeak latencies were significantly longer in patients with anxiety than controls. N1, N2, P2, and P3 AEP components were within normal limits; N1 and P2 were reduced in amplitude in anxiety patients, but differences from controls were not significant. The BAEP findings may suggest altered brain-stem function in anxiety, which has been implied by biochemical studies of anxiety and depression. AEP differences may be related to difficulties in concentration and attention direction reported by anxious patients.

Auditory evoked potentials in narcolepsy and sleep terrors.

Dysfunction of brainstem reticular activating centers has been suggested in some sleep disorders, including narcolepsy and sleep terrors. Previous studies have suggested normal brainstem auditory evoked potentials (BAEPs) in narcolepsy and enhancement of long-latency auditory event-related potentials (ERPs) in sleep deprivation and conditions of pathological somnolence. Sleep terrors have not to date been studied neurophysiologically. We recorded early latency BAEPs and long-latency auditory ERPs in 8 patients with narcolepsy and 5 individuals with sleep terrors, and compared them to 10 normal controls. Narcolepsy patients and controls did not differ significantly in absolute or interpeak latency of BAEPs. Sleep terror patients had significant prolongation relative to controls of III-V and I-V interpeak latencies. The N1, N2, and P3 AEP components were prolonged in latency in narcoleptic patients as compared to controls, while sleep terror patients did not differ from controls. No significant differences in amplitude were found. These findings suggest that a disturbance of integration of brainstem centers subserving wakefulness and sleep may play a role in the disordered arousal of sleep terrors, but suggest no specific abnormality in brainstem function in narcolepsy. The AEP changes in narcolepsy may be a manifestation of pathological sleepiness.

Auditory evoked potentials in vertebrobasilar transient ischemic attacks.

Brainstem auditory evoked potentials (BAEPs) are affected by stroke or migraine in the vertebrobasilar arterial system. Some studies have reported BAEP changes in vertebrobasilar transient ischemic attacks (TIAs), but others have shown no alterations. We recorded BAEPs in 35 patients with TIAs in the vertebrobasilar system who did not have a stroke, other neurologic disease or significant hearing loss. Thirty patients were recorded after resolution of symptoms, while five individuals still had some resolving signs or symptoms. TIA patients as a group had longer interpeak latencies, but I-III, III-V, and I-V latencies were not significantly longer than in controls. Wave V was significantly longer in latency and lower in amplitude in TIA patients, however. The patients whose TIAs had resolved at absolute and interpeak latencies were within normal limits, but three of five had interpeak latencies at or above three standard deviations beyond the normal mean in the still symptomatic group. One of these was later tested and found to be within normal limits. BAEPs after subsidence of symptoms may add little to the evaluation of vertebrobasilar ischemia, but further AEP analysis may show more definitive differences of diagnostic use. The occasional BAEP abnormality during the resolving transient ischemia supports the recently suggested continuum between ischemia and infarction in the vertebrobasilar territory.

Visual and auditory evoked potentials in migraine.

We recorded visual (VEP) and brainstem auditory (BAEP) evoked potentials in 50 patients with clinically diagnosed common migraine attended by visual obscuration or sensory symptoms but no neurologic deficit. VEPs were recorded from Oz, 01, and 02 referenced to Fz, with replication of 200 repetitions of 1.88 per second checkerboard stimuli subtending a 56 minute retinal arc. Analysis time was 250 ms., and filter band pass was 1-250 Hz. BAEPs utilized rarefaction stimulation at 70 dB SL, with 150-3,000 Hz filter band pass and 10 ms. analysis time. Two thousand averages were recorded and replicated from Cz-A1 and Cz-A2. VEP N1, P1 and N2 latencies were longer in migraine patients than in controls, and VEP amplitudes were minimally greater. No significant differences were found between patients and controls, however. BAEP I-V and III-V interpeak latencies were significantly prolonged in migraine patients, and the degree of prolongation was greater on the left. Neither VEPs nor BAEPs exceeded clinical norms in migraine patients. VEPs and BAEPs are likely to add little to the clinical assessment of headache patients. BAEP differences may indicate dysfunction of brainstem centers, possibly related to endorphin or serotonin neurotransmission, and possibly related to the pathogenesis of migraine. The left sided asymmetry has been described previously and is of uncertain significance, but may also support a central mechanism for migraine.

Long-latency auditory event related potentials in migraine.

The pathophysiology of migraine and its associated perceptual symptoms remains controversial. We recorded long-latency auditory event related potentials (AEPs) in 30 unmedicated patients with common migraine, and compared them to 20 controls. 1,000 and 3,000 Hz tones were presented in an 80:20 ratio at 75 dB SL. 200 responses were recorded and replicated from Cz-A1 + A2, with filter band-pass of 1-100 Hz, analysis time of 1,000 ms., and interstimulus interval of 1.1 second. N100, P200, and N200 ERP components did not differ in latency or amplitude between migraine patients and controls. P300 was longer in latency among those with common migraine, and P300 amplitude was significantly reduced (P greater than 0.05). These findings suggest that migraine may have a central mechanism, and that migraineurs may have a difference in task involvement or perception which may influence their clinical course and response to therapy.

Effect of anti-epileptic drug monotherapy and polypharmacy on visual and auditory evoked potentials.

Previous reports have suggested that some anticonvulsants may prolong somatosensory and auditory evoked potential latencies. We compared pattern-reversal visual and brainstem auditory evoked potentials in normal controls, patients on monotherapy, and patients taking polypharmacy. Visual evoked potential amplitudes were less in seizure patients, and P1 latencies were longer in epileptics on polypharmacy than controls. Absolute latencies of brainstem auditory evoked potentials were longer in polypharmacy patients than in controls or monotherapy patients. I-III, III-V, and I-V interpeak latencies were greater in polypharmacy patients than in those on monotherapy or controls. These findings suggest that anticonvulsants may affect conduction along visual and auditory pathways, and that antiepileptic drug polypharmacy and monotherapy may differ in their effects.

EEG frequency analysis in conversion and somatoform disorder.

Patients with "hysterical" neurologic symptoms have long puzzled neurologists and psychiatrists. "Hysteria" has recently been subdivided into conversion and somatoform disorder. We applied computerized EEG frequency analysis to 10 patients with diagnosed conversion disorder, 10 somatoform disorder patients, and 10 control subjects. One minute EEG samples composed of 4 sec epochs were recorded from left frontal (F7-C3), right frontal (F8-C4), left posterior (T5-01) and right posterior (T6-02) derivations. Spectral power was calculated for 4 bands (0.25-4HZ, 4.25-8Hz, 8.25-13Hz, and 13.25-30Hz), and for the full frequency range. Low (0.25-8Hz) and high (8.25-30Hz) frequency bands were compared to determine high/low power ratios on the left and right (PHLL) and (PHLR), ratios of left/right front (PLRF) and posterior (PLRP) power, mean alpha frequency deviation frontally (FLRF) and posteriorly (FLRP), and mobility of left and right frontal power (MOLF and MORF). No significant differences were found between conversion disorder patients and controls in PHLL, PHLR, PLRP, and MORF. PLRF, FLRF and MOLF differed significantly between patients and controls. Power and frequency ratios of right frontal mobilities suggested a decrease in high frequency power, reduction in mean alpha frequency, and lower predominant frequency in the right frontal area in somatoform disorder patients as compared to controls, but significance was not reached. Somatoform and conversion disorder patients differed significantly in these spectral measures. These observations suggest that the two "hysterical" disorders may have distinct pathophysiology, and may be due to eventually identifiable cerebral dysfunction in some cases.