Reduction of Rapid Eye Movement Sleep by Diurnal and Nocturnal Seizures in Temporal Lobe Epilepsy

ORIGINAL CONTRIBUTION Reduction of Rapid Eye Movement Sleep by Diurnal and Nocturnal Seizures in Temporal Lobe Epilepsy

Carl W. Bazil, MD, PhD; Luiz H. M. Castro, MD; Thaddeus S. Walczak, MD

Background: Patients with brief, complex partial sei- maintenance of wakefulness test and 2 subjective drowsi- zures frequently suffer from tiredness and decreased pro- ness tests. ductivity that continue well beyond the postictal period. A possible explanation is that seizures, even when Results: Daytime seizures reduced REM from 18% ± 1% occurring during the day, disrupt sleep the following night. to 12% ± 2% (P = .003). Night seizures reduced REM from 16%±1%to6.8%±2%(PϽ.001). Night seizures also sig- Objective: To determine the effect of temporal lobe com- nificantly reduced stages 2 and 4 while increasing stage plex partial seizures on sleep structure and daytime 1 sleep. Night seizures, but not day seizures, signifi- drowsiness. cantly reduced sleep efficiency, increased time to first REM period, and increased drowsiness as measured by the Methods: Patients with temporal lobe epilepsy were ad- maintenance of wakefulness test. mitted for video-electroencephalography monitoring. All- night polysomnography was recorded under the follow- Conclusions: Temporal lobe complex partial seizures de- ing 3 conditions: seizure free, seizure during the day before crease REM sleep, particularly when occurring during sleep the recording, and seizure during the recording. Percent- but also when occurring on the previous day. This may, age of time in each sleep stage, sleep efficiency, and time in part, be responsible for the prolonged impairment of func- to first and second rapid eye movement (REM) period tioning that some patients report following seizures. were compared for seizure vs control conditions. Day- time drowsiness was also measured, using a modified Arch Neurol. 2000;57:363-368

PILEPTIC SEIZURES typically or during the previous day were com- last less than 2 minutes; pared with recordings in the same pa- however, patients may re- tient when no seizure occurred for at least port decreased perfor- 24 hours. By using patients as their own mance for days afterward. In controls, the effects of individual sei- fact,E 2 of the most prevalent complaints zures were examined independently of of patients with epilepsy are disturbed sleep other variables. and excessive daytime drowsiness,1 which in themselves may be sufficient to inter- RESULTS fere with patients’ ability to work or go to school. These symptoms may have a num- A total of 116 nights were recorded in 34 ber of causes. Epileptic seizures may dis- patients (1-7 nights per patient). Record- rupt sleep and be responsible for daytime ing characteristics are summarized in the drowsiness.2 Anticonvulsants may dis- Table. Our analysis used those record- rupt sleep,3-5 although their effects are vari- ings (87 recordings in 21 patients) in From the Department of able and often difficult to distinguish from which control and seizure conditions Neurology, Comprehensive the effects of seizures. Finally, the under- were recorded. Seven patients had Epilepsy Center, lying disease process causing seizures may DAYSZ and NTSZ recordings. The aver- Columbia-Presbyterian Medical be responsible for changes in sleep. age time during which recordings were Center, New York, NY In our study, all-night polysomnog- obtained was 5.9 nights (range, 2-13 (Dr Bazil); Hospital das Clinicas, São Paulo, Brazil raphy was performed in patients with in- nights). The average time from daytime (Dr Castro); and Minnesota tractable epilepsy who were admitted to seizure to start of recording was 8.5 Comprehensive Epilepsy Care an epilepsy monitoring unit to determine hours. Of the 21 patients, 18 had control (MINCEP), Minneapolis, Minn changes in sleep structure. Recordings in recordings first; 4 of these also had con- (Dr Walczak). which seizures occurred during the night trol recordings after at least 1 seizure

ARCH NEUROL / VOL 57, MAR 2000 WWW.ARCHNEUROL.COM 363

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 PATIENTS AND METHODS for sleep staging were used. Seizure and postictal epochs were not included in total sleep for our purposes. This modification of the original Rechtschaffen and Kales criteria6 was To examine a relatively homogeneous population, only pa- essential, as diffuse delta typically seen after seizures is clearly tients with temporal lobe partial epilepsy (based on interic- not synonymous with any of the accepted sleep stages. Us- tal and ictal recordings) were included. Patients with known ing our method, stages 3 and 4 sleep possibly could be un- or suspected sleep disorders were excluded, although no derestimated, but it would be uncommon for patients to specific testing for sleep apnea or periodic limb move- pass directly into these stages without first entering stage ments was performed. No patient had exclusively diurnal 2. Similarly, mild slowing and attenuation seen postictally or nocturnal seizures. Patients were admitted consecu- would not be scored as stage 1 sleep, and this stage could tively to the Epilepsy Monitoring Unit at Columbia- be underestimated. Presbyterian Medical Center, New York, NY, for diagnosis The initial night of recording was not used in the analy- or surgical evaluation. Temporal onset seizures were veri- sis (to control for first-night effects)7; however, patients ad- fied by video-electroencephalographic (EEG) monitor- hered to the sleep schedule on that night. ing, and computerized seizure detection ensured that un- Control polysomnograms were defined as no seizure witnessed seizures were detected. Most patients were for at least 24 hours before sleep onset and during the re- receiving maintenance anticonvulsant therapy, although this cording. Polysomnograms following daytime seizures was typically decreased and sometimes discontinued dur- (DAYSZ condition) were defined as a seizure between 7 AM ing the admission. Patients taking or withdrawing from bar- and 11 PM on the day the recording began. Polysomno- biturate or benzodiazepine therapy were excluded (includ- grams with night seizures (NTSZ condition) were defined ing patients who received a single dose of a benzodiazepine as a seizure during the recording (after sleep onset). Pa- following a seizure cluster). Beverages containing caffeine tients were identified who had at least 1 control and 1 DAYSZ were not allowed. and/or NTSZ polysomnogram; these were used in the sub- In addition to the usual 10-20 array of scalp elec- sequent analysis. Sleep efficiency, percentage of time in trodes, patients had bilateral outer canthus electrodes and each sleep stage, time to first and second rapid eye move- chin electromyograph electrodes plus a subtemporal chain ment (REM) period, and total sleep time were calculated. of electrodes (total, 29 scalp electrodes). Seizures were di- For nights with seizures, a subanalysis was made when agnosed using video-EEG with the full scalp array. The sub- seizures occurred before or after the first REM period. All jects were not sleep deprived on the nights before record- results were compared using a paired t test vs controls. ing or during recording and were not allowed daytime naps. When more than 1 suitable recording was obtained, the Nurses instructed the patients to sleep at 11 PM and awak- results were averaged before comparisons using the ened them at 7 AM; however, precise “lights-off” times were paired t test. not recorded. Although the patients were in shared rooms Between 1 and 3 PM daily, patients were administered on a hospital ward, efforts were made to optimize sleeping 3 tests of drowsiness. Subjective measures included the Stan- conditions. During recordings, the doors of the room re- ford Sleepiness Scale (SSS),8 where the patient chooses a mained closed, the lights off, and the patients undisturbed description giving a numerical result from 1 to 7, and a lin- unless a seizure occurred. Following seizures, patients were ear analog scale of drowsiness (LASD). For the latter, pa- encouraged to return to sleep. If a benzodiazepine was given tients were asked to mark their drowsiness on a 100-mm after prolonged or repeated seizures, the night was not in- line, from alert to sleepy, yielding a result from 0 to 100. cluded in the analysis. In addition, a more objective nap test was performed that Polysomnography was scored in 30-second epochs, ac- was a modification of the maintenance of wakefulness test cording to standard technique,6 by reformatting digital EEG (MOW).9 For our MOW, patients were placed in a quiet, to polysomnographic channels and settings. Because there dark room in the supine position and instructed to stay was not always a precise lights-off time, scoring began at awake. The time to sleep was recorded (determined by 3 sleep onset (defined by 3 consecutive epochs of stage 1 or consecutive stage 1 epochs or a single epoch of another sleep 1 epoch of stage 2 sleep) and continued until awakening stage), giving a number from 0 to 20. The standard MOW in the morning by the staff. Therefore, sleep latency could includes 5 scheduled recordings in a day. The SSS and LASD not be determined, and total recording time varied some- were administered immediately before each MOW. All tests what. Sleep efficiency is generally defined as time asleep were performed on days meeting the inclusion criteria for as a percentage of time in bed (lights off to lights on). For the polysomnogram portion of the study, and therefore were this study sleep efficiency was calculated somewhat differ- not performed on the day of admission or on the follow- ently from the standard, as percentage of time asleep from ing day. The SSS, LASD, and MOW all give numerical re- sleep onset until awakening. This would likely result in sults that were compared using a paired t test. When more higher numbers, but comparisons between groups should than 1 suitable test was obtained, the results were aver- still be valid. Sleep following a seizure was scored as post- aged before comparisons using the paired t test. When data ictal sleep until the first epoch of normal stage 2 sleep or were not normally distributed, nonparametric statistics were wakefulness was seen; after that point, the usual criteria used.

recording. The mean number of suitable recordings per SLEEP STAGES patient was 2.3 (range, 1-5) for control, 1.3 (range, 1-3) for DAYSZ, and 1.3 (range, 1-2) for NTSZ conditions. Daytime seizures did not result in any significant changes Results are given as mean ± SEM. in sleep stages 1 through 4; however, REM period was

ARCH NEUROL / VOL 57, MAR 2000 WWW.ARCHNEUROL.COM 364

DAYSZ NTSZ No. of recordings 15 13 No. of male-female subjects 6:9 6:7 25

Age range (mean), y 26-56 (41) 26-64 (45) Sleep, % Total ∗

*DAYSZ indicates a seizure occurred between 7 AM and 11 PM on the day of the recording; NTSZ, a seizure occurred during the nocturnal recording. 0 significantly reduced (12% ± 2% vs 18% ± 1%; P = .003). Night seizures caused an increase in stage 1 sleep Control ∗ NTSZ 50 (15% ± 2% vs 11% ± 2%; P = .002) and decreases in stage ∗ BREM 2 sleep (47% ± 3% vs 56% ± 2%; P = .004), stage 4 sleep (0.9% ± 0.5% vs 2.4% ± 0.9%; P = .046), and REM sleep (6.8% ± 2% vs 16% ± 1%; PϽ.001) (Figure 1). For nights when the seizure occurred before the first REM period 25

Total Sleep, % Total ∗ (n = 7), REM sleep was further reduced (3% ± 1% vs ∗ 16% ± 2%; P = .002). ∗ ∗ TIME TO FIRST REM ∗ 0 1 2 3 4 REM Daytime seizures increased the time from sleep onset to Sleep Stage the first REM period (180 ± 38 vs 116 ± 17 minutes), al- Figure 1. Sleep structure after seizures. Top, Percentage of each sleep stage though this difference was not significant (P = .09). Night is shown when no seizure occurred for at least 24 hours before the recording seizures significantly increased the time to the first REM (control condition [CONT]) vs recordings when the same patients had a period (214 ± 43 vs 116 ± 37 minutes; P = .009). When seizure between 7 AM and 11 PM on the day of the recording (DAYSZ condition). Rapid eye movement (REM) sleep was significantly reduced in seizures occurred before the first REM period, the dif- DAYSZ condition. Bottom, Percentage of each sleep stage for CONT vs those ference was more striking (358 ± 30 vs 114 ± 20 min- when the same patient had a seizure during the nocturnal recording (NTSZ utes; PϽ.001) (Figure 2). The time from the begin- condition) or during the recording before the first REM period (BREM condition). Using a paired t test and comparing with controls, stage 1 sleep ning of the first REM period to the beginning of the second was significantly increased and stage 2 sleep was decreased for NTSZ and was not different for any group. BREM conditions. Stage 4 sleep was significantly decreased for NTSZ There were no significant differences in total re- condition only. Asterisk indicates a significant difference ( PϽ.05) compared cording duration for NTSZ (416 ± 12 vs 432 ± 12 min- with CONT. utes) or DAYSZ conditions (443 ± 16 vs 439 ± 9 minutes) compared with controls for each group. therefore, increased drowsiness. For DAYSZ condition, SLEEP EFFICIENCY MOW was 15 ± 2 for controls vs 17 ± 2 for seizure (n = 6; P = .56). For NTSZ condition, MOW was 16 ± 2 for con- Day seizures did not significantly decrease sleep effi- trols vs 7 ± 3 for seizure (n = 8; P = .01). For SAMESZ ciency (87% ± 4% for DAYSZ vs 91% ± 2% for control con- condition, MOW was 18 ± 2 for controls vs 12 ± 2 for sei- ditions; P = .20). Night seizures significantly decreased zure (n = 7; P = .11). Therefore, nocturnal seizures were sleep efficiency (74% ± 4% for NTSZ vs 91% ± 2% for con- the only type that significantly increased drowsiness by trol conditions; PϽ.001), and this effect was further en- this measure, although the number of suitable patients hanced when seizures occurred before the first REM pe- was small. riod (65% ± 3% vs 90% ± 2%; PϽ.001) (Figure 3). There were no significant differences in the more subjective SSS and LASD for any of these groups. DROWSINESS MEASURES Using nonparametric statistics (the Sign Test10), stage 4 sleep was not significantly different between control Three groups were compared with controls using paired and NTSZ conditions, and the MOW was not signifi- t test. Because nocturnal and diurnal seizures disrupt sleep, cantly different following night seizures. All other dif- controls for the purposes of drowsiness measures were ferences noted above remained significant (PϽ.05) us- defined as recordings with no seizures during the pre- ing this stricter test. vious 32 hours (the previous day, the previous night, or the morning of the test). These were compared with re- COMMENT cordings when a seizure occurred the previous day (DAYSZ condition), the previous night (NTSZ condi- These data suggest that complex partial seizures pro- tion), or the morning before the test (SAMESZ condition). foundly disrupt nocturnal sleep, even when they occur The results of the MOW are shown in Figure 4. during the daytime. Diurnal and nocturnal seizures de- Lower numbers represent shorter time to sleep onset and, creased REM sleep. Nocturnal seizures increased stage

ARCH NEUROL / VOL 57, MAR 2000 WWW.ARCHNEUROL.COM 365

80 ∗ 200

Minutes 70 Sleep Efficiency, % Sleep Efficiency,

0 50 Control DAYSZ Control NTSZ 400 ∗ Control 100 NTSZ BREM ∗ 90 ∗

200 80 Minutes

70 ∗ Sleep Efficiency, % Sleep Efficiency,

0 Sleep to First First to Second 50 REM Period Control BREM Control AREM

Figure 2. Time to REM period after sleep onset. Top, Time to first REM Figure 3. Sleep efficiency following seizures. Sleep efficiency is shown period and time between the beginning of the first and second REM periods during CONT vs DAYSZ and NTSZ conditions. Sleep efficiency was are shown during CONT vs DAYSZ conditions. Although time to first REM significantly reduced during NTSZ condition. Bottom, Sleep efficiency for period was increased following seizures, this was not significant. Bottom, CONT vs BREM condition or when a seizure occurred after the first REM Times shown for CONT vs NTSZ or BREM conditions. The time to first REM period (AREM condition). Sleep efficiency was significantly decreased for period was significantly increased for NTSZ and BREM conditions. There both conditions, although more pronounced for BREM. Other abbreviations were no significant differences in the time to second REM period. are given in the legend to Figure 1. Asterisk indicates a significant difference Abbreviations are given in the legend to Figure 1. Asterisk indicates a ( PϽ.05) compared with CONT. significant difference ( PϽ.05) compared with CONT.

1 sleep and decreased stages 2 and 4 sleep and sleep ef- creases16 in REM sleep in patients with epilepsy, but did ficiency. not report whether seizures occurred. There are several possible reasons for decreased REM In all of these studies, patients with epilepsy con- sleep in patients with seizures. The first is simply that tinued to receive anticonvulsants, which can alter sleep is more disrupted, with more frequent awaken- sleep.3-5,17-19 This is particularly significant when com- ings and less time asleep. This is supported by de- parisons were made with healthy controls who took no creased sleep efficiency with nocturnal seizures; how- anticonvulsants. All but 1 study did not control for the ever, sleep efficiency was not changed with diurnal first-night effect, which could affect patients with epi- seizures. A second explanation for decreased REM sleep lepsy differently than controls. Finally, all of these stud- is that seizures affect the circadian pattern responsible ies failed to control for the presence of daytime seizures, for REM, thus delaying its onset. This is supported by which (in our investigation) also affect sleep. the increased time to first REM period with all seizures Our study controlled for any effects of the under- (although this was not statistically significant for diurnal). lying disease process by using patients as their own con- Third, patients knew that they had had seizures; there- trols. This method also controls for anticonvulsant ef- fore, psychological factors could have affected sleep. Fi- fects to some extent. Although anticonvulsant therapy nally, seizures may have a direct REM suppressant ef- typically was tapered during admissions, in general, pa- fect without disruption of circadian rhythms. tients took less drug when seizures occurred than on con- Previous investigations of the effects of seizures on trol nights. Although this was not specifically recorded, sleep have given variable results, largely due to differ- the number of recordings with no anticonvulsant at thera- ences in methods.11 Baldy-Moulinier12 reported de- peutic level in our study was similar for the 3 study con- creased REM sleep only on nights with generalized or mul- ditions. Therefore, any sleep disruption due to anticon- tiple partial seizures when compared with healthy vulsant treatment would be expected to affect control controls. Besset13 reported decreases in total sleep time nights at least as much as seizure nights. Our study spe- and REM sleep when patients with seizures were com- cifically addresses seizures with temporal lobe onset; other pared with patients with no seizures. Studies of single- seizure types (such as frontal onset or primary general- night recordings performed on patients with epilepsy and ized epilepsy) could affect sleep differently. compared with healthy controls, with no mention of the Not surprisingly, patients were more drowsy on proximity of seizures, showed no change14,15 or de- days following a nocturnal seizure as measured by the

ARCH NEUROL / VOL 57, MAR 2000 WWW.ARCHNEUROL.COM 366

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 20 times were not recorded, so the important measure of Control sleep latency could not be calculated and sleep effi- DAYSZ ciency calculations were modified. As respiratory measures were not included, it is possible that undiagnosed sleep apnea was present in some patients, which could induce seizures and exacerbate baseline sleep disrup- 10 tion. As our study looked only for differences with sei- Minutes zure occurrence in the same patients, this confounding factor would only be important if seizures induced or exacerbated sleep apnea.20 This possibility could be investigated with formal sleep studies. 0 As the precise function of REM sleep is unclear, the importance of the observed selective REM deprivation 20 is unknown. Studies in animals show that prolonged REM Control sleep deprivation causes death,21 supporting an essen- NTSZ SAMESZ tial (although as yet undiscovered) function. REM sleep has also been implicated in learning and memory con- solidation.22,23 Therefore, REM sleep deprivation could ∗ 10 be responsible for the dullness and decreased produc-

Minutes tivity that many patients experience following seizures. Deprivation of REM sleep is not, however, known to cause drowsiness; decreased sleep efficiency observed in patients with nocturnal seizures is more likely responsible for this. It is also possible that changes in REM sleep are 0 Maintenance of Wakefulness Test part of an overall disruption in circadian rhythms, which include temperature and melatonin secretion. Disrup- Figure 4. Abbreviated maintenance of wakefulness test (MOW) following seizures. Top, MOW is shown for CONT vs DAYSZ conditions. No difference tion of these patterns clearly causes decreased function- was seen. Bottom, MOW for CONT vs NTSZ conditions or when a seizure ing, as demonstrated in the disorders of jet lag and shift was seen on the morning before the test (SAMESZ condition). The MOW was work.24,25 The temperature curve is difficult to displace, significantly less (and thus the patients more drowsy) for the NTSZ condition only. Other abbreviations are given in the legend to Figure 1. Asterisk and this is unlikely to occur with isolated seizures. How- indicates a significant difference ( PϽ.05) compared with CONT. ever, evidence suggests that melatonin secretion is disrupted by seizures.26,27 Overall, our study shows that temporal lobe com- MOW (although this was not significant by the Sign plex partial seizures disrupt sleep (especially REM), par- Test). This finding is clearly not simply a postictal ticularly when occurring early in a night’s sleep but even effect, as patients who had seizures during the day just when they happen many hours before sleep onset. Fur- before the test (SAMESZ condition) were not signifi- ther studies will be needed to determine whether simi- cantly drowsier than under control conditions. Multiple larly altered sleep patterns are seen in patients with other measures, as described in the standard MOW,9 would types of epileptic seizures. be more accurate; however, this was not deemed practi- cal in patients with frequent seizures in an epilepsy Accepted for publication August 30, 1999. monitoring unit. Although suggestive, our results must This project was supported in part by a grant from the be interpreted with caution and need to be confirmed Epilepsy Foundation of America, Landover, Md. by standard testing. Reprints: Carl W. Bazil, MD, PhD, Comprehensive Epi- There are limitations to our study. The environ- lepsy Center, The Neurological Institute, 710 W 168th St, ment was not a sleep laboratory, so conditions were not New York, NY 10032 (e-mail: [email protected]). optimal for sleep. Although this limits the significance of baseline data, the differences between seizure and REFERENCES seizure-free studies should be valid. There were no baseline sleep diaries, so it is possible that patients were 1. Hoeppner JB, Garron DC, Cartwright RD. Self-reported sleep disorder symp- relatively sleep deprived on admission. If this were true, toms in epilepsy. Epilepsia. 1984;25:434-437. most patients had control recordings first, since REM 2. Touchon J, Baldy-Moulinier M, Besset A, Cadilhac J. Sleep organization and epi- rebound may have contributed to relative increases in lepsy. In: Degen R, Rodin EA, eds. Epilepsy, Sleep, and Sleep Deprivation. 2nd REM on control nights. It is therefore possible that ed. New York, NY: Elsevier Science Inc;1991:73-81. Epilepsy Research Supple- patients with temporal lobe epilepsy chronically cycle ment 2. 3. Harding GFA, Alford CA, Powell TE. The effect of sodium valproate on sleep, re- between REM deprivation and rebound. Anticonvul- action times and visual evoked potentials in normal subjects. Epilepsia. 1985; sants also affect sleep. Therapy was typically tapered 26:597-601. during admissions, and specific doses were not 4. Touchon J, Baldy-Moulinier M, Billiard M, et al. Organisation du sommeil dans recorded. However, the percentage of recordings with l’e´pilepsie re´cente du lobe temporal avant et apre`s traitement par carbamaze´- pine. Rev Neurol (Paris). 1987;143:462-467. no anticonvulsant at therapeutic level in this study was 5. Wolf P, Roder-Wanner UU, Brede M, Noachtar S, Sengoku A. Influences of an- 26% (13/50) for control, 30% (6/20) for DAYSZ, and tiepileptic drugs on sleep. In: Martins da Silva A, ed. Biorhythms and Epilepsy. 29% (5/17) for NTSZ conditions. Precise lights-off New York, NY: Raven Press; 1985:137-153.

ARCH NEUROL / VOL 57, MAR 2000 WWW.ARCHNEUROL.COM 367

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 6. Rechtschaffen A, Kales A, eds. A Manual of Standardized Terminology, Tech- 16. Hamel AR, Sterman MB. Sleep and epileptic abnormalities during sleep. In: Ster- niques, and Scoring System for Sleep Stages of Human Subjects. Los Angeles: man MB, Shouse MN, Passouant P, eds. Sleep and Epilepsy. Orlando, Fla: Aca- University of California–Los Angeles Brain Information Service, Brain Research demic Press Inc; 1982:361-380. Institute; 1968. 17. Johnson LC. Effects of anticonvulsant medication on sleep patterns. In: Ster- 7. Agnew HW, Webb WS, Williams RL. The first night effect: an EEG study of sleep. man MB, Shouse MN, Passouant P, eds. Sleep and Epilepsy. Orlando, Fla: Aca- Psychophysiology. 1966;2:263-266. demic Press Inc; 1982:381-394. 8. Hoddes E, Zarcone V, Smythe H, et al. Quantification of sleepiness: a new ap- 18. Kay DC. Sleep and some psychoactive drugs. Psychosomatics. 1973;14:108- proach. Psychophysiology. 1973;10:431-436. 118. 9. Mitler MM, Gujavarty KS, Browman CP. Maintenance of wakefulness test: a poly- 19. Wauquier A, Clincke GHC, Declerck AC. Sleep alterations by seizures and anti- somnographic technique for evaluating treatment efficacy in patients with ex- convulsants. In: Martins da Silva A, ed. Biorhythms and Epilepsy. New York, NY: cessive somnolence. Electroencephalogr Clin Neurophysiol. 1982;53:658-661. Raven Press; 1985:123-135. 10. Daniel WW. Biostatistics: A Foundation for Analysis in the Health Sciences. 4th 20. Malow BA, Frommes GA, Aldrich MS. Usefulness of polysomnography in epi- ed. New York, NY: John Wiley & Sons Inc; 1987:586-595. lepsy patients. Neurology. 1997;48:1389-1394. 11. Chokroverty S. Sleep and epilepsy. In: Chokroverty S, ed. Sleep Disorders Medi- 21. Kushida CA, Bergmann BM, Rechtschaffen A. Sleep deprivation in the rat, IV: cine. Newton, Mass: Butterworth-Heinemann; 1995:429-454. paradoxical sleep deprivation. Sleep. 1989;12:22-30. 12. Baldy-Moulinier M. Temporal lobe epilepsy and sleep organization. In: Sterman 22. Crick F, Mitchison G. The function of dream sleep. Nature. 1983;304:111-114. MB, Shouse MN, Passouant P, eds. Sleep and Epilepsy. Orlando, Fla: Academic 23. Karni A, Tanne D, Rubenstein BS, Askenasy JJ, Sagi D. Dependence on REM sleep Press Inc; 1982:347-359. of overnight improvement of a perceptual skill. Science. 1994;265:603-604. 13. Besset A. Influence of generalized seizures on sleep organization. In: Sterman 24. Brzezinski A. Melatonin in humans. N Engl J Med. 1997;336:186-195. MB, Shouse MN, Passouant P, eds. Sleep and Epilepsy. Orlando, Fla: Academic 25. Sack RL, Blood ML, Lewy AJ. Melatonin rhythms in night shift workers. Sleep. Press Inc; 1982:339-346. 1992;15:434-441. 14. Bittner-Manicka M. Investigations on the mechanism of nocturnal epilepsy. J Neu- 26. Molina-Carballo A, Acuna-Castroviejo D, Rodriguez-Cabezas T, Munoz-Hoyos A. rol. 1976;211:169-181. Effects of febrile and epileptic convulsions on daily variations in plasma mela- 15. Declerck AC, Wauquier A, Sijben-Kiggen R, Martens W. A normative study of tonin concentration in children. J Pineal Res. 1994;16:1-9. sleep in different forms of epilepsy. In: Sterman MB, Shouse MN, Passouant P, 27. Schapel GJ, Beran RG, Kennaway DL, McLoughney J, Matthews CD. Melatonin eds. Sleep and Epilepsy. Orlando, Fla: Academic Press Inc; 1982:329-338. release in active epilepsy. Epilepsia. 1995;36:75-78.

Correction Correction

Typographical Error. In the December ARCHIVES a typographical error ap- peared in the first paragraph of the editorial “Proton Magnetic Resonance Spec- troscopy: An In Vivo Window to Study Neurodegenerative Disorders” by Lin- fante and Ashizawa (Arch Neurol. 1999;56:1446-1447). The sentence reading “The pulse sequence produces a signal decay in which the spins are per- cussing at a frequency determined by the local magnetic field and ultimately by their chemical relationship with different compounds” should have read “The pulse sequence produces a signal decay in which the spins are precessing at a frequency determined by the local magnetic field and ultimately by their chemical relationship with different compounds.” (Bold typeface added.) We regret the error.

ARCH NEUROL / VOL 57, MAR 2000 WWW.ARCHNEUROL.COM 368