Sleep abnormalities in mentally retarded autistic subjects: Down's syndrome with mental retardation and normal subjects

Brain & Development 21 (1999) 548±553

Original article

www.elsevier.com/locate/braindev

Sleep abnormalities in mentally retarded autistic subjects: Down's syndrome with mental retardation and normal subjects Marina Diomedi a,*, Paolo Curatolo a, Anna Scalise a, Fabio Placidi a, Flavia Caretto b, Gian Luigi Gigli c b

a Dipartimento di Neuroscienze UniversitaÁ Tor Vergata, Roma, Italy Centro Ricerche per la DisabilitaÁ Mentale e Motoria, Associazione Anni Verdi, Roma, Italy c U.O. Neurologia-Neuro®siopatologia, Ospedale S. Maria della Misericordia, Udine, Italy

Received 15 May 1998; received in revised form 30 July 1999; accepted 2 August 1999

Abstract We compared sleep parameters in mentally retarded infantile autism (MRIA) and mentally retarded Down's syndrome (MRDS) by means of polysomnography, evaluating traditional analysis with particular attention to the phasic components in each disorder. Data were compared with those obtained in normal subjects matched for age and sex. Mental age, Intellectual Quotient and the Childhood Autism Rating Scale were performed to obtain an estimation of the neuropsychological de®cit. Abnormalities of phasic components of sleep and the presence of REM sleep components into non-REM sleep were observed in both MRIA and MRDS even if in different ways. In fact, MRDS subjects presented a reduction of REM sleep percentage and R index (number of high frequency REMs against number of low frequency REMs) and this was positively correlated to a low IQ. Unlike MRDS subjects, MRIA subjects did not show any parallelism between intellectual abilities and REM sleep de®cit. In addition, the presence of undifferentiated sleep in autistic subjects implies a maturational de®cit that is still present in adulthood. Finally, a high R index in MRIA was observed. This ®nding, which is not present in MRDS, could represent an estimation of the disorganized arrival of information caused by a dyscontrol or a reduction of inhibitor pathway. With reference to sleep mechanisms, our results suggest that the cognitive de®cit in MRIA may differ from that of MRDS subjects. A maturational de®cit of CNS with a dysfunction of brainstem monoaminergic neurons could represent the underlying mechanism. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Autism; Sleep; Down syndrome; Mental retardation; REM sleep

1. Introduction Autistic social dysfunction arises early in development and may have a broad phenotype presentation. In particular, medical literature differentiates between high functioning and low functioning subjects, depending on their social abilities and Intellectual Quotient (IQ). Moreover, even among low functioning individuals, there is evidence that certain characteristics of the cognitive patterns can differentiate them from mentally retarded subjects without autism [1,2]. Previous research attempting to understand autistic dysfunction and behavior explored anatomic, neurochemical, genetic and neuropsychological ®elds suggesting the presence of a disorder in the maturation of different neuronal systems [3]. Progress has been made especially with * Corresponding author. Clinica Neurologica, Ospedale S. Eugenio, Piazzale Umanesimo 10, 00144 Roma, Italy. Tel.: 139-6-5914436; fax: 139-6-5922086. E-mail address: [email protected] (M. Diomedi)

respect to diagnosis, but many neurophysiological aspects still remain unclear. Many authors have analyzed the importance of the relationship between REM sleep and learning or memory [4±9]. These studies demonstrated that the integration of clinical aspects and neurophysiological characteristics can be particularly helpful in understanding the underlying cognitive de®cit. The aim of this study was to analyze sleep structure in adolescents and young adult autistic subjects with a severe mental retardation, and to compare it with that of severe MRDS and normal subjects, with particular attention to the characteristics of REM sleep, for its implication in memory and learning abilities.

2. Material and methods We selected 10 mentally retarded autistic subjects (three F, seven M; mean age 18.2, range 12±24 years old) ful®lling

0387-7604/99/$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0387-760 4(99)00077-7

M. Diomedi et al. / Brain & Development 21 (1999) 548±553

the criteria of DSM-IV [10] from the Center for Mental and Motor Disabilities Anni Verdi. All subjects were affected by autism of unknown etiology. Detailed history and full physical and neurological examination were performed. The absence of associated pathology was assessed by means of karyotyping, cerebral MRI, metabolic screening, audiometry, visual acuity test and prolonged EEG monitoring to exclude the most frequently associated diseases such as epilepsy, phenylketonuria, fragile X syndrome, tuberous sclerosis, sensory impairment, and other structural alterations of the central nervous system. The severity of autistic behavior was evaluated with the Childhood Autism Rating Scale (CARS), that classi®es subjects into three different degrees of severity: #29.5 no autism, 30±36.5 light or moderate, $37 severe [11]. All autistic subjects were submitted to the Psychological Educational Pro®le (PEP) to evaluate their evolutive age, expressed in number of months [12]. Intellectual Quotient (IQ) was ,30 in all cases. In order to compare the data obtained in MRIA with the neurophysiologic characteristics of different neurocognitive pro®les, we selected also eight Down syndrome subjects matched for chronological age, mental age and gender (two F, six M; mean age 22.5, range 17±31 years old). They were studied for the presence of epilepsy and of sleep apnea syndrome with a prolonged EEG and polysomnographic recording to exclude confounding factors that could modify their sleep structure. The IQ was ,30. Subjects with history of pharmacological treatment during the last 3 months were excluded from the study. The control group included eight normal subjects, matched with the other two groups for chronological age and gender (two F, six M; mean age 20.5, range 13±20). Informed consent was obtained from the parents of all patients and from control subjects prior to entry in the study. Clinical characteristics of subjects are summarized in Table 1. All subjects underwent two consecutive overnight polysomnographic recordings in the sleep laboratory, at hours compatible with their usual lifestyle. The ®rst night was used for adaptation to sleep laboratory and the second one for statistic analysis. Polygraphic recording included 4 electroencephalographic channels (EEG): C3-A2, C4-A1, O1A2, O2-A1; one electromiographic channel (EMG): mylohyoideus; four electrooculographic channels (EOG) to avoid the detachment of the electrodes eventually due to the poor collaboration of subjects: two right and two left ocular cantus. Sleep macrostructure was scored in 30 s epochs according to Rechtschaffen and Kales' criteria [13]. The epochs characterized by the presence of REM and non-REM elements (i.e. sleep epochs characterized by saw-tooth waves, low voltage background activity, high EMG activity and absence of eye movements; or sleep epochs with the presence of rapid eye movements associated to spindles or to high EMG activity) were classi®ed as undifferentiated sleep (US) according to Lairy et al.'s

549

Table 1 Clinical characteristics of the groups

Chronological age Mental age (month) CARS a PEP b (month) IQ c

Autistic patients

Down patients

Control subjects

18.2 ^ 1.6 47.5 ^ 6.77 43.25 ^ 1.45 34.3 ^ 6.62 23.3 ^ 4.84

22.5 ^ 1.64 20.5 ^ 8.26 50.87 ^ 4.91 ± ± ± ± ± 19.18 ^ 3.48 ±

a

CARS: Childhood Autism Rating Scale. PEP: Psychological Educational Pro®le. c IQ: Intellectual Quotient. b

criteria [14]. In addition, rapid eye movements (REMs) with an amplitude $50 mV were counted and the intervals [15] between each eye movement in the same burst (i.e. with interval between any REM not longer than 3 s) were classi®ed depending on their duration in: I , 1 s, 1 s # I , 2 s, I $ 2 s. We calculated the R index as the ratio between high frequency REMs (short intervals) and low frequency REMs (long intervals) (R index: I , 1 s/I $ 2 s). In addition, REM activity (number of REMs), REM density (number of REMs/total REM sleep) and number of REM cycles (number of REM periods separated by at least 15 min of non-REM sleep) were calculated. We calculated chin twitches and their relationship with REMs in autistic and Down's syndrome subjects. For this purpose we calculated the tonic inhibition index (TII, number of phasic mental muscle activities lasting #0.5 s in REM sleep/total phasic mental muscle activities lasting ,2.0 s in REM sleep) and the phasic inhibition index (PII, rate of mental muscle activity lasting ,2 s occurring during the burst occurrence of REMs against total number of the activities in REM sleep) in accordance with Kohyama criteria [16]. Statistical analysis was performed with Mann±Whitney test for group comparisons. Linear regression was used to correlate the IQ of MRDS subjects and CARS score of MRIA subjects with REM sleep characteristics. 3. Results Compared to MRDS subjects, MRIA subjects showed a signi®cantly higher amount of REM sleep percentage (P ˆ 0:008), of REM activity (P ˆ 0:004), and US percentage (P ˆ 0:0004). Although they showed a trend to lower frequency of awakening and a higher R index, they were statistically insigni®cant. Compared to normal subjects, autistic subjects presented a signi®cant reduction of REM sleep percentage (P ˆ 0.008) and a signi®cant increase of US (P ˆ 0:0004). Their sleep continuity was disturbed by an increased presence of interspersed wakefulness (P ˆ 0:001), and an increased number of awakenings (P ˆ 0:002), with a consequent reduction of

550

M. Diomedi et al. / Brain & Development 21 (1999) 548±553

Table 2 Mean and standard deviation of sleep parameters a Sleep parameters

Autistic subjects

Down subjects

Controls

Autistic vs. Down

Autistic vs. controls

Down vs. controls

S1% S2% S3 1 S4% REM% Number of REM cycles REM activity REM density US% Wake % Number of awakenings Sleep ef®ciency index R index TII PII

7.8 ^ 0.9 36.5 ^ 3.1 24.4 ^ 3 12.2 ^ 1.5 3.9 ^ 1.1 175.3 ^ 44.7 3.14 ^ 0.65 3.4 ^ 0.5 14.7 ^ 2.2 11.4 ^ 1.4 0.81 ^ 0.02 1.6 ^ 0.28 0.721 ^ 0.1 10.92 ^ 3.88

10.6 ^ 1.6 43.8 ^ 3.4 20.8 ^ 1.6 2 ^ 0.8 1.62 ^ 1.5 35.7 ^ 18.5 2.16 ^ 0.77 1.3 ^ 0.64 21.2 ^ 4.2 18.9 ^ 2.7 0.74 ^ 0.06 0.82 ^ 0.33 0.51 ^ 0.07 4.71 ^ 3.6

6.8 ^ 0.73 41.13 ^ 2.4 29.2 ^ 2.4 20.1 ^ 0.64 4.8 ^ 0.83 145.1 ^ 10.3 2.12 ^ 0.56 0.06 ^ 0.05 2.8 ^ 1.1 4.2 ^ 0.75 0.92 ^ 0.01 1.42 ^ 0.08 0.79 ^ 0.06 2.05 ^ 0.75

NS NS NS 0.008 0.006 0.004 NS 0.0004 NS 0.002 NS NS 0.0008 0.002

NS NS NS 0.008 NS NS NS 0.0004 0.001 0.002 0.003 NS NS 0.0004

NS NS NS 0.0008 0.001 0.004 NS 0.074 0.002 0.0008 0.01 0.03 0.0008 NS

a

US, undifferentiated sleep; Sleep ef®ciency index; total sleep time/time in bed; REM activity; number REMs; REM density; number REMs/total REM sleep; R index, I , 1 s/I $ 2 s; TII, tonic inhibition index; PII, phasic inhibition index.

Sleep Ef®ciency Index (SEI, total sleep time/time in bed) (P ˆ 0:003) (Table 2). In the group of autistic subjects the analysis of phasic events in REM sleep showed an increase of dispersed REMs occurring out of the bursts of REMs, a fragmentation of REM periods, due to a frequent intrusion of non-REM sleep stage 1 and 2 and an increased amount of muscle twitches. REM activity and REM density were higher compared to both MRDS and normal subjects, but these differences were not signi®cant. Comparing with age matched normal subjects, TII revealed to be normal range (0:721 ^ 0:10), but PII showed signi®cantly higher values (10:92 ^ 3:88 vs. 2:05 ^ 0:75; P ˆ 0:0004). As compared to Down's syndrome subjects they show a signi®cantly higher TII (P ˆ 0:0008) and lower PII (P ˆ 0:002). Spearman Rank Test Correlation showed no correlation between CARS score or IQ and REM sleep % in MRIA

(R ˆ 0:42; P ˆ 0:39 and R ˆ 0:11; P ˆ 0:82 respectively). Finally, the analysis of the EEG pattern in MRIA subjects showed the frequent occurrence of diffuse theta rhythms, similar to the ``mu rhythm'', during REM sleep (Fig. 1), sporadic central spikes especially during non-REM sleep (stage 2), and a clear abundance of spindle activity (13± 20 Hz) not only during stage 2 of non-REM sleep but also during SWS and REM sleep (Fig. 2). Polygraphic recordings of Down's syndrome subjects showed a signi®cant reduction in REM sleep percentage (P ˆ 0:008 with respect to MRIA subjects; P ˆ 0:0008 with respect to controls) and in number of REM cycles (P ˆ 0:006 with respect to MRIA subjects; P ˆ 0:001 with respect to controls); a signi®cant reduction in R index was also observed (P ˆ 0:03 with respect to controls); ®nally, a signi®cant increase in US was found with respect to MRIA (P ˆ 0:0004).

Fig. 1. Polysomnographic EEG recording during nocturnal sleep. A 7±7.5 Hz activity with a characteristic arch shaped wave morphology (arrow), similar to the `mu rhythm', appears during REM sleep.

M. Diomedi et al. / Brain & Development 21 (1999) 548±553

551

Fig. 2. Polysomnographic EEG recording during nocturnal sleep. The epochs characterized by the presence of a 10.5±15 Hz activity and K complexes during a stage presenting with REM sleep features (rapid eye movements and low voltage EMG activity) are classi®ed as undifferentiated sleep.

The study of motor activity during REM sleep showed that TII is signi®cantly lower with respect to normal subjects (P ˆ 0:0008) and autistic subjects (P ˆ 0:0008) while PII is signi®cantly higher (P ˆ 0:002) only with respect to autistic subjects (Table 2). Linear regression analysis showed a signi®cant correlation between REM sleep % and IQ in MRDS subjects (R ˆ 0:9; P ˆ 0:14).

4. Discussion 4.1. Sleep pattern in mentally retarded subjects Several studies on sleep organization in mentally retarded subjects demonstrated the peculiarity of their sleep pattern [17±22]. In fact, authors reported an increase of REM sleep latency, a reduction of REM sleep percentage and number of REM cycles, and the presence of undifferentiated sleep. The sleep characteristics described above are not speci®c of an etiology of mental retardation while their severity seems to be positively correlated to the severity of mental retardation [4]. Many authors have analyzed the importance of the relationship of REM sleep to the learning ability and memory. In fact, REM sleep deprivation in animals causes a de®cit of mnemonic and learning abilities [5±7]; on the contrary, the amount of REM sleep increases with learning sessions [8,9]. REM sleep percentage can also change according to an ontogenetic pro®le, since it is high in new-borns and decreases progressively with aging [23]. On the basis of these observations, REM sleep has been considered as a neurophysiological marker of the ability of the central nervous system (CNS) to receive new information, i.e. an index of brain plasticity [24,25]. Another important neurophysiological index is the interval between two consecutive REMs in the same burst of REMs. Some studies demonstrated an increase of the R

index, after learning sessions in normal individuals [26], and even in subjects with mental retardation submitted to very structured (organized) learning sessions [27]. This index seems to re¯ect the activation of the complex neurophysiological systems that generate oculomotor activity during REM sleep (R index, I , 1 s/I $ 2 s). R index has an ontogenetic pro®le in normal subjects, that is, increases with age and decreases after 60 years. However, in MRDS subjects this index, remaining at low levels, does not follow the normal maturational increase [28] and seems to be positively correlated to IQ [22,27]. For these reasons, R index has been considered a quantitative esteem of the organizational abilities of the subject; i.e. the ability to coherently organize, during REM sleep, the information received during wakefulness. In this framework, we con®rm the characteristics previously reported in the literature [17±20]. The reduced REM sleep percentage, characteristic of mental retarded subjects, could be interpreted as a reduced opportunity to receive new information (reduced plasticity like in aging) and the reduced R index as a reduced ability to organize them (like in new-borns). Since motor activity during sleep is considered to be correlated with neural maturation for its chronological changes, the altered tonic motor suppression during REM sleep suggests an involvement of brainstem inhibitory centers with a consequent insuf®cient REM atonia. 4.2. Sleep pattern in autistic subjects Polygraphic studies of sleep in autistic subjects are less numerous and their results are controversial [15,29±32]. In particular, Aihara and Hashimoto also found a similar amount of REM sleep in autistic and normal subjects in contrast with the abnormality of phasic aspects of REM sleep in MRIA [31]. In fact, they observed an abnormal presence of 10.5±15 c/s, spindling like, EEG activity during REM sleep, and the presence of rapid eye movements

552

M. Diomedi et al. / Brain & Development 21 (1999) 548±553

during stage 1 and 2 of non-REM sleep suggesting a defect of maturational process in autistic children. A similar pattern was reported in new-borns [33]. More recently, Elia et al. [30] reported in autistic children a normal REM sleep percentage, an increased number of rapid eye movements and REM density (number of REMs/ total REM sleep), and a signi®cant increase of R index. In girls with Rett syndrome, a syndrome often associated with autistic behavior, Segawa and Nomura describe an altered circadian rhythm, an alteration of phasic and tonic components and a leakage of the components of REM sleep stage into non-REM sleep. This abnormal pattern does not change with age, persisting in adulthood, suggesting a progression of pathology; on the contrary, in autistic subjects sleep abnormalities tend to normalize [34]. The results of our study point out some important and peculiar neurophysiological characteristics in autistic subjects. 4.3. Nocturnal awakenings As previously mentioned we found a large number of awakenings in retarded subjects compared to autistic and normal subjects and in autistic subjects compared to normals. Recently, a similar ®nding was found on purely clinical basis [35]. 4.4. REM sleep % and intellectual abilities Linear regression analysis demonstrates that autistic subjects, unlike MRDS subjects, do not show any correlation between intellectual abilities and percentage of REM sleep during nocturnal recording. The abundance of studies showing a strict relationship between REM sleep and learning/intellectual abilities lead us to wonder if traditional IQ evaluation is applicable to autistic subjects. As compared to MR subjects, IQ measures of autistic subjects might be not only by their real intellectual de®cit but also by their attentive and relational problems, resulting in lower scores and therefore creating a lack of parallelism between IQ and REM sleep. Supporting this concept, in a recent study, a negative correlation of medial frontal cortical glucose metabolic rate (GMR) with attentional performance, was found in autistic subjects suggesting that neuronal inef®ciency in that region may contribute to poor performances [36]. 4.5. Undifferentiated sleep The increased percentage of US in autistic subjects suggests a possible maturational defect of EEG activity; since a similar pattern has been described during active sleep of premature and newborns [33]. In addition, recent studies report the in¯uence of monoamine systems in neural development [37]. Recently, a decreased serotonin 5-HT2 receptor binding, associated to an elevated platelet serotonin content has been demonstrated in autistic subjects [38]. An imbalance of serotoni-

nergic activity in the serotoninergic neurons of dorsal raphe nuclei could result in an insuf®cient inhibition of pontine neurons responsible for generation of PGO (ponto-geniculooccipital) waves, hence giving rise to phasic events of REM sleep during non-REM sleep. An involvement of these structures could explain the abnormalities in sleep-wakefulness cycle reported in early infantile autism. Moreover, the developmental disorder in MRDS subjects seems to be associated with an altered prenatal peak in serotonin 5-HT1A receptors in the fetal brain [39]. This suggests a possible role of serotonin in pathophysiologic mechanisms leading to sleep abnormalities. Our data demonstrate that the presence of rapid eye movements during non-REM sleep in autistic subjects, previously reported by other authors [15,29±31] and similar to MRDS subjects, persists during adolescence and adulthood at least in our individuals, with severe mental retardation. 4.6. Oculomotor phasic activity during REM sleep Similarly to Elia et al. [30] we observed a high R index in autistic subjects. The abundance of high frequency REMs in MRIA could represent a redundant, continuous, repetitive signal without an informative meaning, probably caused by an altered control or a reduction of inhibitor pathway with a consequent reduced ability to organize new information in a coherent system. This observation and the presence of ``mu rhythm'' and sporadic central spikes could be congruent with what has been observed by Niedermeyer et al. [40] during sleep in Rett syndrome. The presence of these frequent EEG abnormalities (epileptiform discharges, mu rhythm, redundant spindle activity, high R index) could be due to a ``cortical disinhibition'' and could be the neurophysiological base of some characteristic clinical manifestations (stereotypies, sensory perceptual aberrations, decreased sociability, poor attention, impaired learning). Recently a possible anatomic correlate of cortical dishinibition has been hypothesized by Bauman and Kemper, who detected an increased cell packing density in limbic circuits in autistic childrens' brains [41]. Gustafsson proposed a neural circuit theory for the dysfunctional selforganization as the cause of the disability to extract features from stimuli [42]. The higher values of PII with respect to normal subjects suggest an involvement of a particular neural system in the brainstem responsible for the control of phasic motor activity. 5. Conclusions Current genetic, biochemical, pathological and neurophysiological studies cannot give de®nite answers in autism, but the hypothesis of an early event that causes a disturbed development involving several structures and, in particular, serotoninergic and perhaps dopaminergic pathways, becomes more reasonable. It seems that, at least in severely

M. Diomedi et al. / Brain & Development 21 (1999) 548±553

mentally retarded autistic subjects, there is a de®cit of the control of phasic elements of REM sleep with an abnormal presence of brief twitches in REM sleep (higher value of PII), the presence of REMs in non-REM sleep (undifferentiated sleep), and a trend to higher R index. We think that neurophysiological research into infantile autism has a particular interest in the broader ®eld of developmental psychopathology because it could give some explanation of casual mechanisms that underlie autistic syndrome and other developmental syndromes. References [1] Rumsey JM, Hamburger SD. Neuropsychological ®ndings in highfunctioning men with infantile autism, residual state. J Clin Exp Neuropsychol 1988;10:201±221. [2] Bennetto L, Pennington BF, Rogers SJ. Intact and impaired memory functions in autism. Child Dev 1996;67:1816±1835. [3] Rapin I, Katzman R. Neurobiology of autism. Ann Neurol 1998;43:7± 14. [4] Castaldo V. Down'syndrome: a study of sleep patterns related to level of mental retardation. Am J Ment De®c 1969;74:187±190. [5] Leconte P, Bloch V. De®cit de la reÂtention d'un conditionnement apreÁs privation du sommeil paradoxal. C R Acad Sci (Paris) 1970;271:226±229. [6] Fishbein W. Disruptive effects of rapid eye movement sleep deprivation on long term memory. Physiol Behav 1971;6:279±282. [7] Pearlman C, Greenberg R. Brief REM sleep deprivation impair consolation in complex learning in rats. Psychophysiology 1972;9:109± 110. [8] Leconte P, Hennevin E. Augmentation de la dureÂe du sommeil paradoxal conseÂcutif aÁ un apprentissage chez le rat. C R Acad Sci (Paris) 1971;273:86±88. [9] McGrath MJ, Cohen DB. REM sleep faciltation of adaptive waking behaviour: a review of literature. Psychol Bull 1978;85:24±57. [10] American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington, DC: American Psychiatric Association, 1994 [11] Schopler E, Reichler RJ, De Vellis RF, Daly K. Toward objective classi®cation of childhood autism: Childhood Autism Rating Scale (CARS). J Autism Dev Disord 1980;10:91±103. [12] Schopler E, Reichler RJ, Basbford A, et al. Individualized assessment and treatment for autistic and developmentally disabled children. Psychoeducational pro®le revised (PEP-R), Vol. 1, Austin, TX: ProEd, 1990. [13] Rechtscaffen A, Kales A. A manual of standardized terminology, technique and scoring systems of sleep stages of human subjects, Washington, DC: NIH Publications, 1968 p. 204. [14] Lairy G, Barros-Ferreira M, Goldsteinas L. DonneÂes reÂcentes sur la physiologie et la physiopathologie de l'activite onirique. In: Gastaut H, editor. The abnormalities of sleep in man, Bologna: Aulo-Gaggi, 1967. pp. 275±283. [15] Ornitz EM, Ritvo ER, Brown MB, La Franchi S, Parmelee T, Walter RD. The EEG and rapid eye movement during REM sleep in normal and autistic children. Electroenceph clin Neurophysiol 1969;26:167± 175. [16] Kohyama J. Sleep as a window on the developing brain. Curr Probl Pediatr 1998;27:73±93. [17] Petre-Quadens O, Jouvet M. Sleep in the mentally retarded. J Neurol Sci 1967;4:354±357. [18] Feinberg I, Braun M, Schulman E. EEG sleep patterns in mental retardation. Electroenceph clin Neurophysiol 1969;27:128±141. [19] Castaldo V, Krynicki V. Sleep and eye movement patterns in two groups of retardates. Down,s syndrome. Biol Psychi 1974;9:231±244.

553

[20] Fukuma E, Umezawa Y, Kobayashi K, Motoike M. Polygraphic study on the nocturnal sleep of children with Down's syndrome and endogenous mental retardation. Folia Psychiatr Neurol Jpn 1974;28:333± 345. [21] Clausen J, Sersen EA, Lidsky A. Sleep pattern in mental retardation. Down's Syndrome. Electroenceph clin Neurophysiol 1977;43:183± 191. [22] Grubar JC. Sleep and mental de®ciency. Rev Electroenceph Neurophysiol Clin 1983;13:107±113. [23] Williams R, Karacan I, Hursch C. Electroencephalography of human sleep: clinical applications, London: Wiley, 1974. [24] Gigli GL, Grubar JC, Bergonzi P, Maschio MC. Long-term administration of butoctamide hydrogen succinate on nocturnal sleep of mentally retarded subjects: a polygrphic study versus placebo. Psychopharmacology 1995;117:438±442. [25] Grubar JC, Sakuma A, Gigli GL, D'Erario C, Pollicina C. De®cience mentale et pharmacologie: l'approche neuropeÂdagogique. Enfance 1990;3:323±337. [26] Spreux F, Lambert C, Chevalier B, Meriaux H, Freixa I, Baque E, et al. Modi®cation des caracteritiques de sommeil paradoxal consecutif aÁ un apprentissage chez l'homme. Cahiers Psycol Cognitive 1982;2:327±334. [27] Gigli GL, Grubar JC, Colognola RM, Amata MT, Pollicina C, Ferri R, et al. Butoctamide hydrogen succinate and intensive learning sessions: effects on night sleep of Down's syndrome patients. Sleep 1987;10:563±569. [28] Grubar JC. Approche psychophysiologique du potentiel intellectuel. Enfance 1985;1:85±90. [29] Tanguay PE, Ornitz EM, Forsythe AB, Ritvo ER. Rapid eye movement (REM) activity in normal and autistic children during REM sleep. J Autism Child Schizophr 1976;6:275±288. [30] Elia M, Ferri R, Musumeci SA, Candian C, Orviati S, Miano GM, Bergonzi P. Further considerations on REM sleep in autistic subjects. Dev Brain Dysfunct 1993;6:352. [31] Aihara R, Hashimoto T. Sleep polygraphic and neuroendocrinological studies on autistic children. Brainstem Telencephalon Behav. 1986;:29±43. [32] Richdale AL, Prior MR. The sleep/wake rhythm in children with autism. Eur Child Adolesc Psychiatry 1995;4:175±186. [33] Parmelee Jr AH, Shulte FJ, Akiyama Y, Wenner WH, Schultz MA, Stern E. Maturation of EEG activity during sleep in premature infants. Electroenceph clin Neurophysiol 1968;24:319±329. [34] Segawa M, Nomura Y. Polysomnography in the Rett syndrome. Brain Dev 1992;14:S46±S54. [35] Taira M, Takase M, Sasaki H. Sleep disorder in children with autism. Psychiatry Clin Neurosci 1998;52(2):182±183. [36] Siegel Jr BV, Nuechterlein KH, Abel L, Wu JC, Buchsbaum MS. Glucose metabolic correlates of continuous test performance in adults with a history of infantile autism, schizophrenics, and controls. Schizophr Res 1995;17:85±94. [37] Levitt P, Harvey JA, Friedman E, Simansky K, Murphy EH. New evidence for neurotransmitter in¯uences on brain development. Trends Neurosci 1997;20:269±274. [38] Cook EH, Leventhal BL. The serotonin system in autism. Curr Opin Pediatr 1996;8:348±354. [39] Bar-Peled O, Gross-Iseroff R, Ben-Hur H, Hoskins I, Groner Y, Biegon A. Fetal human brain exhibits a prenatal peal in the density of serotonin 5-HT1A receptors. Neurosci Lett 1991;127:173. [40] Niedermeyer E, Naidu SB. Attention-de®cit hyperactivity disorder (ADHD) and frontal-motor cortex disconnection. Clin Electroenceph 1997;28:130±136. [41] Bauman ML, Kemper TL. Neuroanatomic observations of the brain in autism. In: Bauman ML, Kemper TL, editors. The neurobiology of autism, Baltimore, MD: Johns Hopkins University Press, 1994. pp. 119±145. [42] Gustafsson L. Inadequate cortical feature maps: a neural circuit theory of autism. Biol Psychiatry 1997;42:1138±1147.