- Email: [email protected]
behavioral relations
International Journal of Psychophysiology 51 (2003) 17–25
Cortical auditory processing and communication in children with autism: electrophysiologicalybehavioral relations ´ ´ Nicole Bruneaua,*, Frederique Bonnet-Brilhaulta, Marie Gomota, Jean-Louis Adrienb, ´´ a Catherine Barthelemy a
Service Universitaire d’Explorations Fonctionnelles, et de Neurophysiologie en Pedopsychiatrie, ´ 2, Bd Tonnelle, ´ 37 044, Tours, Cedex, ´ France b Institut de Psychologie, Universite´ Rene´ Descartes, Paris V, France Received 22 May 2003; received in revised form 27 May 2003; accepted 3 June 2003
Abstract The purpose of the present study was to investigate the relations between late auditory evoked potentials (AEPs) recorded at temporal sites (the N1c wave or Tb) and verbal and non-verbal abilities in children with autism. The study was performed in 26 mentally retarded children with autism (AUT) aged 4–8 years (mean age"S.E.M.s 71"2 months; mean verbal and non-verbal developmental quotient"S.E.M.s36"4 and 48"3). The stimuli used were 750 Hz tone bursts of 200 ms duration delivered binaurally at different intensity levels (50, 60, 70, 80 dB SPL) with 3–5 s interstimulus intervals. Temporal AEPs were first compared to those of a group of 16 normal children (NOR) in the same age range (mean age"S.E.M.s69"3 months). We then focused on the AUT group and considered relations between temporal AEPs and the severity of disorders of verbal and non-verbal communication assessed using a behavior rating scale. AEPs recorded on left and right temporal sites were of smaller amplitude in the AUT group than in the NOR group. Increasing intensity-related amplitude was observed on both sides in NOR and only on the right side in AUT. The lack of intensity effect on the left side resulted in a particular pattern of asymmetry at the highest level of intensity (80 dB SPL) with greater N1c amplitude on the right than on the left side (the reverse was found in the NOR group). Electro-clinical correlations indicated that the greater the amplitude of the right temporal N1c responses, the higher the verbal and non-verbal communication abilities. This suggests a developmental reorganization of left-right hemisphere functions in autism, with preferential activation of the right hemisphere for functions usually allocated to the left hemisphere, particularly those involving the secondary auditory areas situated on the lateral surface of the superior temporal gyrus where the N1cyTb wave is generated. 䊚 2003 Elsevier B.V. All rights reserved.
Keywords: Autism; Children; Late auditory evoked potentials; N1c wave; T complex; Temporal asymmetry; Verbal and nonverbal communication
*Corresponding author. Tel.: q33-2-47-8519; fax: q33-2-47-47-3846. E-mail address: [email protected] (N. Bruneau). 0167-8760/03/$ - see front matter 䊚 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0167-8760(03)00149-1
18
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
1. Introduction Autism is a neurodevelopmental disorder characterized by verbal and non-verbal communication disorders, impaired social interactions, and limited activities and interests (DSM IV, APA, 1994). Communication with language is so strikingly impaired in autism that the brain mechanisms underlying auditory processing remain in question. Moreover, abnormal behavioral responses to auditory stimuli have often been reported in autism, leading to hypotheses regarding particular auditory processing (Hayes and Gordon, 1977; Grandin and Scariano, 1986; Grandin, 1992). Several studies have attempted to clarify this, using electrophysiological methods (auditory evoked potentials, AEPs). Although brainstem AEP abnormalities are present in a higher proportion of autistic children than in normal children, they are not a necessary condition for autism since most autistic children display normal brainstem AEPs (review in Klin, 1993). Middle latency AEPs representing initial activation of the primary auditory cortex have also been reported to be normal in adults with autism (Courchesne et al., 1985; Grillon et al., 1989), although no findings have yet been reported in children. The abnormalities are more consistent when considering late cortical auditory evoked potentials that occur between 80 and 200 ms after stimulus onset. The responses recorded in this latency period have been extensively studied in normal populations, both in adults (for reviews, ¨¨ ¨ see Naatanen and Picton, 1987; Woods, 1995) and more recently in children (Bruneau et al., 1997; Bruneau and Gomot, 1998; Ceponiene` et al., 1998; Pang and Taylor, 2000; Ponton et al., 2000, 2002). In a previous study, these late cortical auditory evoked potentials (AEPs) were studied in children with autism who were also mentally retarded (Bruneau et al., 1999). They were compared to normal children and retarded children without autism, all aged from 4 to 8 years. During this age range, the auditory responses of normal children are mainly characterized by a large and reliable negative wave culminating at bitemporal sites (Bruneau et al., 1997). This wave corresponds to N1c (McCallum and Curry, 1980; Woods, 1995) or to the negative Tb wave of the T complex
previously described by Wolpaw and Penry (1975). This temporal-recorded deflection originates from sources in the secondary auditory cortex on the lateral aspect of the temporal lobe (Wolpaw and Penry, 1975; Celesia, 1976; Wood and Wolpaw, 1982). The large N1c in 4- to 8-year-old children contrasts with the small and inconsistent negative N1b wave culminating approximately 20 ms earlier at fronto-central sites, corresponding to the ‘classical’ N1 wave; this latter peak results from activation of generators situated in the supratemporal plane of the auditory cortex (Vaughan and Ritter, 1970; Scherg and Von Cramon, 1985) and frontal areas (Giard et al., 1994). Our previous results showed bitemporal hyporeactivity to auditory stimuli in children with autism, as indicated by the N1c wave, which was of smaller amplitude and longer latency than responses in both normal and retarded children. The fronto-central N1b wave in children with autism did not differ from that of normal children. Furthermore, in our previous electrophysiological study AEP were considered according to stimulus intensity level. Whereas N1c amplitude modulation was observed bitemporally in both control groups, it was only found on the right temporal site in children with autism, with a lack of variations in amplitude on the left temporal site. An atypical pattern of asymmetry was therefore found in autism with right temporal dominance of AEP when stimulus intensity increased. Such a right-dominant pattern of activation has previously been described in response to speech stimuli; these studies were based on measurement of dichotic listening performance (Prior and Bradshaw, 1979), EEG alpha blocking, temporal AEP (Dawson et al., 1982, 1989) and cortical activation to verbal auditory stimulation recorded with PET (Muller et al., 1999), and led these authors to hypothesize that language might be preferentially processed in the right hemisphere in autism. Might the N1c responses to tones, preferentially recorded on right temporal cortical areas, be related to verbal abilities in children with autism? These relations were considered in the present work in a larger sample of subjects than previously studied. However, autism is more than a simple language deficit, and is mainly characterized by impairment
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
of both verbal and non-verbal communication involving gestures (DSM IV), both predominantly processed in the left hemisphere in normal subjects. Our aim was therefore to investigate whether the temporal auditory responses are indicative of the right hemisphere involvement for both verbal and non-verbal communication in autism. 2. Methods 2.1. Subjects A group of 26 children with autism (22 males, 4 females) aged 4–8 years (mean"S.E.M.s 71"2 months) participated in this study; all were patients attending the Department of Child Psychiatry Day-Care Unit of a University Hospital. Infantile autism was diagnosed according to DSMIV criteria (APA, 1994) by two independent experts (a child psychiatrist and a clinical psychologist). Developmental quotients of children with autism were evaluated by using mental age´ appropriate tests: the Brunet–Lezine-R develop´ mental test for infants (Brunet and Lezine, 1976) that examines psychomotor development from 1 to 30 months and EDEI-R for children (a revised form of a French scale evaluating intellectual skills) that assesses cognitive abilities from 30 months to 9 years (Perron-Borelli, 1978). These two developmental tests provide both verbal and non-verbal developmental quotients. These tests have been standardized on large samples of normal children, the mean VDQ and NVDQ being 100 (ranging from 70 to 130) in the normal population. In the group of children with autism (AUT), verbal developmental quotients (VDQ) ranged from 12 to 80 (mean"S.E.M.s36"4) and non-verbal quotients (NVDQ) ranged from 27 to 90 (mean"S.E.M.s48"3). Autistic behavior was assessed by using the Behavior Summarized Evaluation Scale (BSE-R; ´´ Barthelemy et al., 1997), a composite scale of 29 items for evaluation of autistic symptoms (poor social interaction, abnormal eye contact, verbal and non-verbal communication disabilities, bizarre responses to auditory stimuli, ritual use of objects, etc«) and associated features, such as eating, sleep disorders. The BSE-R is completed each
19
week by nurses who care for the children during the day. Raters are trained on the scale and are experienced with children with autism. For this study we retained the results of this scale rated during the week of the electrophysiological recording. Each of the 29 items is rated from 1 to 5 according to the severity of the disorder: 1, if the disorder is absent; 5 if the symptoms are prominent and always observed. A glossary explaining the ´´ significance of each item is available (Barthelemy et al., 1998). According to this glossary items 5 and 6 are defined as follows: Item 5: does not make an effort to communicate using voice andyor words Here assessment should be based on the effort towards communication and not on verbal level. A child with speech can make no effort to communicate and score a high mark (non-communicative echolalic language). A child without speech can try to make himselfyherself understood in hisy her own way (babbling, prattling) and score a low mark. Item 6: lack of appropriate facial expression and gestures The child’s face is not expressive (amimia, facial immobility). If the child can speak, heyshe does not use facial, vocal or gestural expression with normal frequency and liveliness. Heyshe cannot direct the examiner’s hand to obtain a desired object: does not wave hands in its direction: cannot indicate precisely what he wants by gesture, attitude or look. Heyshe does not show any anticipatory postural reaction when about to be picked up. A group of normal children (12 males, 4 females) aged 4–8 years (mean"S.E.M.s69"3 months), drawn from children in the normal school population constituted the normal control group (NOR). All subjects had normal hearing as assessed by brainstem auditory evoked responses (BAER) recorded before study of late auditory evoked potentials. Children with metabolic or chromosomal disease, history of substantial neurological
20
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
disorders or seizures, abnormal EEG with either slow waves or epileptiform discharges were excluded. All children were free of psychotropic medication for at least one month before the electrophysiological study. 2.2. Procedure The conditions were similar to those previously used and detailed in Bruneau et al. (1997). The subjects sat in a comfortable armchair in a dimly lit, sound-insulated room. Children were accompanied by their nurse, mother or a person known to them. During the recording session, the children were asked to remain still; this was the only instruction given. The behavior of each child was monitored by a video system throughout the session; it allowed verification that the head was correctly positioned between the speakers. The role of the nurse or the parent seated near the child was first to reassure himyher and second to control the child’s head position. The speakers were placed on small tables beside each arm of the armchair and the adult sat near a speaker, either to the right or to the left of the child. The EEG was recorded from 5 Ag–Ag Cl cup electrodes (Cz, Fz, Pz, T3, T4) with linked earlobes as reference. Eye movements were recorded from two supraorbital linked electrodes, also referred to linked earlobes. The EEG was sampled on-line at a rate of 250 Hz for 1 s duration (500 ms pre- and post-stimulus). The stimuli comprised 750-Hz tone bursts of 200 ms duration with risey fall times of 20 ms delivered through two speakers placed 50 cm directly in front of each ear. Each subject received a randomized series of tones of varying intensity. Interstimulus interval varied from 3 to 5 s. All trials contaminated by ocular movements were rejected on-line (criterion 100 mV) and the sequence was then prolonged (the stimuli always being delivered in a pseudo-random order) until a total of 40 trials for each stimulus intensity had been collected. The percentages of rejected (and then added) trials were between 5 and 25% in normal children and between 10 and 35% in patients. Data analysis was performed on auditory evoked potentials recorded at stimulus intensity levels of 50, 60, 70 and 80 dB SPL.
The negative peak occurring during the 100– 200 ms period after stimulus onset were measured at each temporal site (T3, T4). Manual scoring using a cursor program was used to measure amplitude and latency of the peak according to the mean of the 500-ms prestimulus baseline. The N1c peak culminates at approximately 170 ms at temporal sites in 4- 8-year-old children (Bruneau et al., 1997). 2.3. Data analysis Amplitude and latency data of temporal N1c peak were analyzed using repeated measures ANOVA. The between-subjects variable was group (two levels: AUT and NOR). The two within-subject variables were site (left and right temporal, i.e. T3 and T4) and intensity (four levels: 50, 60, 70, 80 dB SPL). In all ANOVAs involving repeated measures effects with more than two levels, Greenhouse–Geisser epsilon corrections were used to adjust probabilities for repeated measures F-values. The results of such ANOVA indicated successively: the uncorrected degrees of freedom, F value, the epsilon-corrected P value and the corresponding epsilon value. Newman–Keuls tests were used for post hoc comparisons. Correlations between electrophysiological and behavioral data were performed using Spearman rank correlations, which were computed between the AEP indices (latency, amplitude) and verbal and non-verbal level (QDV and QDNV) and, on the other hand, communication disabilities assessed using the BSE scale. Items selected were item 5 (does not make an effort to communicate using voice andyor words), item 6 (lack of appropriate communicative gestures and facial expression) and item 24 (bizarre responses to auditory stimuli). Significance took into account the Bonferroni correction (Bland and Altman, 1995). 3. Results Inter-group comparisons of late auditory temporal responses (N1c) 3.1. Amplitude Overall ANOVA performed on N1c amplitude indicated a significant main effect of Group
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
21
Fig. 1. Mean amplitude and S.E.M. of significant N1c in 16 normal children (NOR) and 26 children with autism (AUT) as a function of stimulus intensity level. AUT showed smaller amplitudes than NOR on both sides at all intensity levels (*P-0.002). At 80 dB SPL, responses were greater at right than at left temporal site for AUT (*P-0.006).
wF(1, 40)s14.85; P-0.0005x, and Intensity wF(3, 120)s10.66; P-0.00001; ´s0.89x and a significant Group=Hemisphere=Intensity interaction wF(3, 120)s6.78; P-0.0005; ´s0.89x. N1c amplitude was smaller on both sides and at all intensity levels in AUT than in NOR children (Fig. 1). ANOVA performed at each intensity level indicated that the Group=Hemisphere interaction was only significant at 80 dB wF(1,40)s11.36; P-002x. At this high level of intensity, planned comparisons indicated that N1c amplitude was significantly greater on the right hemisphere than on the left in AUT wF(1,40)s8.72; P-0.005x, the difference in interhemisphere amplitude being reversed (LH) RH) but did not reach significance in NOR wF(1,40)s3.87; Ps0.06x (Fig. 1). Grand average auditory responses recorded in both groups of subjects at the highest level of intensity are shown on Fig. 2.
6.58; P-0.0004; ´s0.86x but no side effect. Moreover, no significant interaction was found. Post hoc comparisons indicated that N1c latency was longer in AUT than in NOR on both sides and at all intensity levels. Latency decreased with increasing stimulus intensity in both groups and on both sides (see Table 1). 3.3. Relations between temporal AEPs and communication abilities (verbal, non-verbal) in AUT: The electrophysiological results detailed above indicated that preferential activation of the right
3.2. Latency Overall ANOVA performed on N1c latency indicated a significant main effect of Group wF(1, 40)s6.86; P-0.02x, and Intensity wF(3, 120)s
Fig. 2. Grand averaged AEP elicited by 80 dB SPL tones for normal children (NOR) and children with autism (AUT).
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
22
Table 1 Mean (S.E.M.) peak latency for N1c waves measured at temporal sites (left: T3; right: T4) in normal children (NOR) and children with autism (AUT) 50 dB T3 T4
NOR AUT NOR AUT
176 182 176 191
(3) (5) (3) (3)
60 dB 168 179 169 185
(3) (5) (5) (5)
70 dB 163 174 167 183
(3) (5) (3) (3)
80 dB 170 179 167 184
(3) (6) (4) (5)
hemisphere occurs in AUT for the highest level of stimulus intensity (80 dB SPL). This intensity level was therefore chosen to calculate electrobehavioural relations. Although the amplitude of N1c was not correlated with VDQ on the left side (rs0.30), the correlation was significant on the right side (rs0.45; P-0.02): the greater the N1c amplitude at the right temporal electrode, the higher the verbal quotient. This was confirmed by results obtained by considering relations between N1c amplitude and scores for communication skills evaluated using
items 5 (verbal communication) and 6 (use of gestures to communicate) of the BSE scale. Scores for these respective items were significantly correlated with N1c amplitude on the right side (rs y0.60; y0.58; P-0.002) whereas they were not correlated with N1c amplitude on the left side (rs y0.40; y0.21). On the BSE scale, the higher the score for each item, the greater the severity of the disturbance. Negative correlation therefore indicated that the greater the N1c wave on the right side, the lower the score, i.e. the better the communication abilities were, using verbal and non-verbal skills. The correlations obtained between N1c amplitude and the mean value of E5 and E6 were calculated for each intensity level (50, 60, 70, 80 dB). Results were, respectively, y0.37, y0.13, y0.21, y0.32 (all ns) on the left side and y0.42, P-0.05; y 0.44, P-0.02; y0.56, P-0.005; y0.61, Ps 0.001 on the right side. After Bonferroni correction, results were significant at both 70 and 80 dB. Data obtained at the highest intensity level are shown on Fig. 3.
Fig. 3. Correlation between N1c amplitude at right temporal site at 80 dB SPL and mean score for items E5 and E6 of BSE-R scale evaluating verbal and non-verbal communication disorders. The greater the N1c amplitude at T4, the smaller the scores for E5 and E6, i.e. the better the verbal and non-verbal communication skills.
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
No significant correlations were found between the item evaluating abnormal behavioral reaction to sound and N1c amplitude, either on the right (rsy0.18) or on the left side (rsy0.16). No significant correlation was found when considering N1c latency and behavioral evaluations (all r0.2). 4. Discussion These electrophysiological results concerning temporal auditory responses (N1c) in autism confirm those previously obtained in a smaller sample of children (Bruneau et al., 1999). Bitemporal electrophysiological hyporeactivity was obtained in children with autism, revealed by N1c, which was of smaller amplitude and longer latency than in normal children. Moreover, although N1c amplitude increased with increasing stimulus intensity in NOR, this effect was only found on the right side in AUT. A particular pattern of asymmetry was therefore found at the highest level of stimulus intensity in AUT, indicating greater activation of the right temporal region. One explanation might be that the homologous regions in the left hemisphere has a much higher ‘threshold’ of intensityrelated cortical response than the right, which might be due to a particular developmental neural network organization. Since N1c has been demonstrated to be generated in the associative auditory areas located in the lateral part of the superior temporal gyrus (Celesia, 1976; Wood and Wolpaw, 1982; Scherg and Von Cramon, 1986), the bitemporal electrophysiological N1c hyporeactivity might be related to hypoperfusion of both temporal lobes, involving the associative auditory cortex evidenced in PET neuroimaging studies in AUT at rest (Zilbovicius et al., 2000; Ohnishi et al., 2000) and in response to tone stimuli (Muller et al., 1999; Boddaert et al., 2001). Our results are in agreement with results obtained with SPECT on 5- to 11-year-old autistic children after activation by repeated 80 dB SPL tone stimuli (Garreau et al., 1994). Posterior associative temporal areas were activated on the left side in normal children and on the right side in children with autism. A lack of cerebral blood
23
flow modifications in response to tone stimulations had also previously been shown in autistic children using transcranial Doppler ultrasonography (Bruneau et al., 1992). Modifications of AEP observed when stimulus intensity increases, are usually interpreted as reflecting better synchronisation of single trial responses, which lead to greater mean amplitude and shorter latency of peaks. This phenomenon does not occur at all on the left side in AUT. The greater amplitude of N1c might also correspond to stronger activation that spreads over larger cortical temporal areas. Another possible explanation might be an ampliotopic effect on scalp potential distribution of N1c. If localization of the N1c generator is dependent on intensity, a more radial orientation of the generators activated at high intensity level might explain greater N1c amplitude at temporal sites. It might also be hypothesized that N1c involves two (or more) overlapping components that are differentially activated at high intensity levels. This is supported by results showing a broader peak of N1c at T4, with a small shoulder on the descending part of the N1c curve clearly evident at 80 dB (Fig. 2). These different possibilities are probably more complementary than exclusive. Further topographical studies of N1c according to intensity will be required to clarify these different hypotheses. Relations between electrophysiological and behavioral data indicated that the right hemisphere ‘reactivity’ as revealed by N1c amplitude is indicative of verbal and non-verbal communication abilities in children with autism. These results suggest reorganization of the respective roles of the left and right hemispheres in auditory information processing during early brain development rather than dysfunction of the left hemisphere alone. The present results are therefore arguments indicating that when verbal and non-verbal communication are acquired in autism, acquisition may occur via right-hemisphere compensation for early left-hemisphere dysfunction. Right-hemisphere specialization for language is not uncommon in cases of early left-hemisphere trauma (Rasmussen and Milner, 1977). Such a hypothesis would be
24
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
reinforced by data showing relations between N1c amplitude and language communication in normal children. The lack of relations between N1c amplitude and the score for item 24 (‘Bizarre responses to auditory stimuli’) must be emphasized. This item quantifies excessive, insufficient or selective sensitivity to noise, sounds and speech, as well as paradoxical responses (both hypo- and hyper-reactivity). The behaviors evaluated may be too heterogeneous on this item and may need to be separated to study relations with temporal responses. Another interpretation might be that, as shown by our results, temporal N1c is an index of activity of the cortical networks involved in the cognitive processes of verbal and non-verbal communication rather than of cortical networks involved in sensory processes such as the hyperacusis and hyporeactivity observed in autism. Acknowledgments This study was supported by INSERM U316 Dynamics and Pathology of Cerebral Development (Dir: Pr Pourcelot). We thank Doreen Raine for editing the English. References APA: American Psychiatric Association, 1994. Diagnostic and Statistical Manual for Mental Disorders (DSM IV), 4th edn, Washington, DC: American Psychiatry Press. ´´ Barthelemy, C., Hameury, L., Lelord, G., 1998. Infantile Autism: Exchange and Development Therapy. Expansion ¸ Scientifique francaise. BSE-R scale glossary, pp. 80–84. ´´ ´ Barthelemy, C., Roux, S., Adrien, J.L., Hameury, L., Guerin, P., Garreau, B., et al., 1997. Validation of the revised Behavior Summarized Evaluation scale (BSE-R). J. Autism Dev. Disord. 27, 139–154. Bland, J.M., Altman, D.G., 1995. Multiple significance tests: the Bonferroni method. Br. Med. Journal. 310, 21. Boddaert, N., Belin, P., Poline, J.B., Chabanne, N., Mouren´´ Simeoni, M.C., Barthelemy, C., et al., 2001. Temporal lobe dysfunction in autism: a children PET auditory activation study. Neuroimage 13 (6), S1028. Bruneau, N., Dourneau, M.C., Garreau, B., Pourcelot, L., Lelord, G., 1992. Blood flow response to auditory stimulations in normal, mentally retarded, and autistic children: a preliminary transcranial Doppler ultrasonographic study of the middle cerebral arteries. Biol. Psychiatry 32, 691–699. Bruneau, N., Gomot, M., 1998. Auditory evoked potentials (N1 wave) as indices of cortical development throughout
childhood. In: Garreau, B. (Ed.), Neuroimaging in Child Neuropsychiatric Disorders. Springer Verlag, pp. 113–124. ´´ Bruneau, N., Roux, S., Adrien, J.L., Barthelemy, C., 1999. Auditory associative cortex dysfunction in children with autism: evidence from late auditory evoked potentials (N1 wave—T complex). Clin. Neurophysiol. 110, 1927–1934. ´ ´´ Bruneau, N., Roux, S., Guerin, P., Barthelemy, C., Lelord, G., 1997. Temporal prominence of auditory evoked potentials (N1 wave) in 4- 8-year-old children. Psychophysio 34, 32–38. ´ ´ Brunet, O., Lezine, I., 1976. Echelle de developpement psy` enfance. In: Brunet, O., Lezine, ´ chomoteur de la premiere ´ ` I. (Eds.), Le developpement Psychologique de la Premiere Enfance. PUF, Paris 2nd ed. Celesia, G.G., 1976. Organization of auditory cortical areas in man. Brain 99, 403–414. ` R., Cheour, M., Naatanen, ¨¨ ¨ Ceponiene, R., 1998. Interstimulus interval and auditory event-related potentials in children: evidence for multiple generators. Electroenceph. Clin. Neurophysiol. 108, 345–354. Courchesne, E., Courchesne, R.Y., Hicks, G., Lincoln, A.J., 1985. Functioning of the brain-stem auditory pathway in non-retarded autistic individuals. Electroenceph. Clin. Neurophysiol. 61, 491–501. Dawson, G., Warrenburg, S., Fuller, P., 1982. Cerebral lateralization in individuals diagnosed as autistic in early childhood. Brain Lang. 15, 353–368. Dawson, G., Finley, C., Phillips, S., Lewy, A., 1989. A comparison of hemispheric asymmetries in speech-related brain potentials of autistic and dysphasic children. Brain and language 37, 26–41. ´ Garreau, B., Zilbovicius, M., Guerin, P., Samson, Y., Syrota, A., Lelord, G., 1994. Effects of auditory stimulation on regional cerebral blood flow in autistic children. Dev. Brain Dysfunction 7, 119–128. ´ Giard, M.H., Perrin, F., Echallier, J.F., Thevenet, M., Froment, J.C., Pernier, J., 1994. Dissociation of temporal and frontal components in the human auditory N1 wave: a scalp current density and dipole model analysis. Electroenceph clin Neurophysiol. 92, 238–252. Grandin, T., 1992. An inside view of autism. In: Schopler, E., Mesibov, G.B. (Eds.), High-functioning individuals with autism. Plenum Press, New York, pp. 105–126. Grandin, T., Scariano, M., 1986. Emergence: labeled autistic. Arena, Novato, CA. Grillon, C., Courchesne, E., Akshoomoff, N., 1989. Brainstem and middle latency auditory evoked potentials in autism and developmental language disorder. J. Autism Dev. Disord 19, 255–269. Hayes, R.W., Gordon, A.G., 1977. Auditory abnormalities in autistic children. Lancet 2, 767. Klin, A., 1993. Auditory brainstem responses in autism. Brainstem dysfunction of peripheral hearing loss? J. Autism Dev. Disord. 23, 15–34. McCallum, W.C., Curry, S.H., 1980. The form and distribution of auditory evoked potentials and CNVs when stimuli and responses are lateralized. In: Kornhuber, H.H., Deecke, L.
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25 (Eds.), Motivation, motor and sensory processes of the brain: Electrical potentials, behavior and clinical use, Progress in brain research, Vol. 54. Elsevier, Amsterdam, pp. 767–775. Muller, R.A., Behen, M.E., Rothermel, R.D., Chugani, D.C., Muzik, O., Mangner, T.J., et al., 1999. Brain mapping of language and auditory perception in high-functioning autistic adults: A PET study. J. Autism Develop. Disorders 29, 19–31. ¨¨ ¨ Naatanen, R., Picton, T., 1987. The N1 wave of the human electric and magnetic response to sound: a review and an analysis of the component structure. Psychophysio 24, 375–429. Ohnishi, T., Matsuda, H., Hashimoto, T., Kunihiro, T., Nishikawa, M., Uema, T., et al., 2000. Abnormal regional cerebral blood flow in childhood autism. Brain 123, 1838–1844. Pang, E.W., Taylor, M.J., 2000. Tracking the development of the N1 from 3 to adulthood: an examination of speech and non-speech stimuli. Clin. Neurophysiol. 111, 388–397. ´ ´ Perron-Borelli, M., 1978. Les echelles differentielles defficiences intellectuelles EDEI, Manuel. Editions Scientifiques et Psychotechniques, Paris. Ponton, C.W., Eggermont, J.J., Kwong, B., Don, M., 2000. Maturation of human central auditory system activity: evidence from multi-channel evoked potentials. Clin. Neurophysiol. 111, 220–236. Ponton, C.W., Eggermont, J.J., Khosla, D., Kwong, B., Don, M., 2002. Maturation of human central auditory system
25
activity: separating auditory evoked potentials by dipole source modeling. Clin. Neurophysiol. 113, 407–420. Prior, M.R., Bradshaw, J.L., 1979. Hemisphere functioning in autistic children. Cortex 15, 73–81. Rasmussen, T., Milner, B., 1977. The role of early left-brain injury in determining lateralization of cerebral speech functions. Ann. N.Y. Acad. Sci. 299, 355–369. Scherg, M., Von Cramon, D., 1985. Two bilateral sources of the late AEP as identified by a spatio-temporal dipole model. Electroenceph. Clin. Neurophysiol. 62, 32–44. Scherg, M., Von Cramon, D., 1986. Evoked dipole source potentials of the human auditory cortex. Electroenceph. Clin. Neurophysiol. 65, 344–360. Vaughan Jr, H.G., Ritter, W., 1970. The sources of auditory evoked responses recorded from the human scalp. Electroenceph Clin. Neurophysiol. 28, 360–367. Wolpaw, J.R., Penry, J.K., 1975. A temporal component of the auditory evoked response. Electroenceph Clin. Neurophysiol. 39, 609–620. Wood, C.C., Wolpaw, J.R., 1982. Scalp distribution of human auditory evoked potentials. II. Evidence for overlapping sources and involvement of auditory cortex. Electroenceph. Clin. Neurophysiol. 54, 25–38. Woods, D.L., 1995. The component structure of the N1 wave of the human auditory evoked potential. Electroenceph Clin. Neurophysiol. Suppl. 44, 102–109. ´ Zilbovicius, M, Boddaert, N., Belin, P., Poline, J.B., Remy, PH., Mangin, J.F., et al., 2000. Temporal lobe dysfunction in childhood autism: a PET study. Am. J. Psychiatr. 157, 1988–1993.
Cortical auditory processing and communication in children with autism: electrophysiologicalybehavioral relations ´ ´ Nicole Bruneaua,*, Frederique Bonnet-Brilhaulta, Marie Gomota, Jean-Louis Adrienb, ´´ a Catherine Barthelemy a
Service Universitaire d’Explorations Fonctionnelles, et de Neurophysiologie en Pedopsychiatrie, ´ 2, Bd Tonnelle, ´ 37 044, Tours, Cedex, ´ France b Institut de Psychologie, Universite´ Rene´ Descartes, Paris V, France Received 22 May 2003; received in revised form 27 May 2003; accepted 3 June 2003
Abstract The purpose of the present study was to investigate the relations between late auditory evoked potentials (AEPs) recorded at temporal sites (the N1c wave or Tb) and verbal and non-verbal abilities in children with autism. The study was performed in 26 mentally retarded children with autism (AUT) aged 4–8 years (mean age"S.E.M.s 71"2 months; mean verbal and non-verbal developmental quotient"S.E.M.s36"4 and 48"3). The stimuli used were 750 Hz tone bursts of 200 ms duration delivered binaurally at different intensity levels (50, 60, 70, 80 dB SPL) with 3–5 s interstimulus intervals. Temporal AEPs were first compared to those of a group of 16 normal children (NOR) in the same age range (mean age"S.E.M.s69"3 months). We then focused on the AUT group and considered relations between temporal AEPs and the severity of disorders of verbal and non-verbal communication assessed using a behavior rating scale. AEPs recorded on left and right temporal sites were of smaller amplitude in the AUT group than in the NOR group. Increasing intensity-related amplitude was observed on both sides in NOR and only on the right side in AUT. The lack of intensity effect on the left side resulted in a particular pattern of asymmetry at the highest level of intensity (80 dB SPL) with greater N1c amplitude on the right than on the left side (the reverse was found in the NOR group). Electro-clinical correlations indicated that the greater the amplitude of the right temporal N1c responses, the higher the verbal and non-verbal communication abilities. This suggests a developmental reorganization of left-right hemisphere functions in autism, with preferential activation of the right hemisphere for functions usually allocated to the left hemisphere, particularly those involving the secondary auditory areas situated on the lateral surface of the superior temporal gyrus where the N1cyTb wave is generated. 䊚 2003 Elsevier B.V. All rights reserved.
Keywords: Autism; Children; Late auditory evoked potentials; N1c wave; T complex; Temporal asymmetry; Verbal and nonverbal communication
*Corresponding author. Tel.: q33-2-47-8519; fax: q33-2-47-47-3846. E-mail address: [email protected] (N. Bruneau). 0167-8760/03/$ - see front matter 䊚 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0167-8760(03)00149-1
18
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
1. Introduction Autism is a neurodevelopmental disorder characterized by verbal and non-verbal communication disorders, impaired social interactions, and limited activities and interests (DSM IV, APA, 1994). Communication with language is so strikingly impaired in autism that the brain mechanisms underlying auditory processing remain in question. Moreover, abnormal behavioral responses to auditory stimuli have often been reported in autism, leading to hypotheses regarding particular auditory processing (Hayes and Gordon, 1977; Grandin and Scariano, 1986; Grandin, 1992). Several studies have attempted to clarify this, using electrophysiological methods (auditory evoked potentials, AEPs). Although brainstem AEP abnormalities are present in a higher proportion of autistic children than in normal children, they are not a necessary condition for autism since most autistic children display normal brainstem AEPs (review in Klin, 1993). Middle latency AEPs representing initial activation of the primary auditory cortex have also been reported to be normal in adults with autism (Courchesne et al., 1985; Grillon et al., 1989), although no findings have yet been reported in children. The abnormalities are more consistent when considering late cortical auditory evoked potentials that occur between 80 and 200 ms after stimulus onset. The responses recorded in this latency period have been extensively studied in normal populations, both in adults (for reviews, ¨¨ ¨ see Naatanen and Picton, 1987; Woods, 1995) and more recently in children (Bruneau et al., 1997; Bruneau and Gomot, 1998; Ceponiene` et al., 1998; Pang and Taylor, 2000; Ponton et al., 2000, 2002). In a previous study, these late cortical auditory evoked potentials (AEPs) were studied in children with autism who were also mentally retarded (Bruneau et al., 1999). They were compared to normal children and retarded children without autism, all aged from 4 to 8 years. During this age range, the auditory responses of normal children are mainly characterized by a large and reliable negative wave culminating at bitemporal sites (Bruneau et al., 1997). This wave corresponds to N1c (McCallum and Curry, 1980; Woods, 1995) or to the negative Tb wave of the T complex
previously described by Wolpaw and Penry (1975). This temporal-recorded deflection originates from sources in the secondary auditory cortex on the lateral aspect of the temporal lobe (Wolpaw and Penry, 1975; Celesia, 1976; Wood and Wolpaw, 1982). The large N1c in 4- to 8-year-old children contrasts with the small and inconsistent negative N1b wave culminating approximately 20 ms earlier at fronto-central sites, corresponding to the ‘classical’ N1 wave; this latter peak results from activation of generators situated in the supratemporal plane of the auditory cortex (Vaughan and Ritter, 1970; Scherg and Von Cramon, 1985) and frontal areas (Giard et al., 1994). Our previous results showed bitemporal hyporeactivity to auditory stimuli in children with autism, as indicated by the N1c wave, which was of smaller amplitude and longer latency than responses in both normal and retarded children. The fronto-central N1b wave in children with autism did not differ from that of normal children. Furthermore, in our previous electrophysiological study AEP were considered according to stimulus intensity level. Whereas N1c amplitude modulation was observed bitemporally in both control groups, it was only found on the right temporal site in children with autism, with a lack of variations in amplitude on the left temporal site. An atypical pattern of asymmetry was therefore found in autism with right temporal dominance of AEP when stimulus intensity increased. Such a right-dominant pattern of activation has previously been described in response to speech stimuli; these studies were based on measurement of dichotic listening performance (Prior and Bradshaw, 1979), EEG alpha blocking, temporal AEP (Dawson et al., 1982, 1989) and cortical activation to verbal auditory stimulation recorded with PET (Muller et al., 1999), and led these authors to hypothesize that language might be preferentially processed in the right hemisphere in autism. Might the N1c responses to tones, preferentially recorded on right temporal cortical areas, be related to verbal abilities in children with autism? These relations were considered in the present work in a larger sample of subjects than previously studied. However, autism is more than a simple language deficit, and is mainly characterized by impairment
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
of both verbal and non-verbal communication involving gestures (DSM IV), both predominantly processed in the left hemisphere in normal subjects. Our aim was therefore to investigate whether the temporal auditory responses are indicative of the right hemisphere involvement for both verbal and non-verbal communication in autism. 2. Methods 2.1. Subjects A group of 26 children with autism (22 males, 4 females) aged 4–8 years (mean"S.E.M.s 71"2 months) participated in this study; all were patients attending the Department of Child Psychiatry Day-Care Unit of a University Hospital. Infantile autism was diagnosed according to DSMIV criteria (APA, 1994) by two independent experts (a child psychiatrist and a clinical psychologist). Developmental quotients of children with autism were evaluated by using mental age´ appropriate tests: the Brunet–Lezine-R develop´ mental test for infants (Brunet and Lezine, 1976) that examines psychomotor development from 1 to 30 months and EDEI-R for children (a revised form of a French scale evaluating intellectual skills) that assesses cognitive abilities from 30 months to 9 years (Perron-Borelli, 1978). These two developmental tests provide both verbal and non-verbal developmental quotients. These tests have been standardized on large samples of normal children, the mean VDQ and NVDQ being 100 (ranging from 70 to 130) in the normal population. In the group of children with autism (AUT), verbal developmental quotients (VDQ) ranged from 12 to 80 (mean"S.E.M.s36"4) and non-verbal quotients (NVDQ) ranged from 27 to 90 (mean"S.E.M.s48"3). Autistic behavior was assessed by using the Behavior Summarized Evaluation Scale (BSE-R; ´´ Barthelemy et al., 1997), a composite scale of 29 items for evaluation of autistic symptoms (poor social interaction, abnormal eye contact, verbal and non-verbal communication disabilities, bizarre responses to auditory stimuli, ritual use of objects, etc«) and associated features, such as eating, sleep disorders. The BSE-R is completed each
19
week by nurses who care for the children during the day. Raters are trained on the scale and are experienced with children with autism. For this study we retained the results of this scale rated during the week of the electrophysiological recording. Each of the 29 items is rated from 1 to 5 according to the severity of the disorder: 1, if the disorder is absent; 5 if the symptoms are prominent and always observed. A glossary explaining the ´´ significance of each item is available (Barthelemy et al., 1998). According to this glossary items 5 and 6 are defined as follows: Item 5: does not make an effort to communicate using voice andyor words Here assessment should be based on the effort towards communication and not on verbal level. A child with speech can make no effort to communicate and score a high mark (non-communicative echolalic language). A child without speech can try to make himselfyherself understood in hisy her own way (babbling, prattling) and score a low mark. Item 6: lack of appropriate facial expression and gestures The child’s face is not expressive (amimia, facial immobility). If the child can speak, heyshe does not use facial, vocal or gestural expression with normal frequency and liveliness. Heyshe cannot direct the examiner’s hand to obtain a desired object: does not wave hands in its direction: cannot indicate precisely what he wants by gesture, attitude or look. Heyshe does not show any anticipatory postural reaction when about to be picked up. A group of normal children (12 males, 4 females) aged 4–8 years (mean"S.E.M.s69"3 months), drawn from children in the normal school population constituted the normal control group (NOR). All subjects had normal hearing as assessed by brainstem auditory evoked responses (BAER) recorded before study of late auditory evoked potentials. Children with metabolic or chromosomal disease, history of substantial neurological
20
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
disorders or seizures, abnormal EEG with either slow waves or epileptiform discharges were excluded. All children were free of psychotropic medication for at least one month before the electrophysiological study. 2.2. Procedure The conditions were similar to those previously used and detailed in Bruneau et al. (1997). The subjects sat in a comfortable armchair in a dimly lit, sound-insulated room. Children were accompanied by their nurse, mother or a person known to them. During the recording session, the children were asked to remain still; this was the only instruction given. The behavior of each child was monitored by a video system throughout the session; it allowed verification that the head was correctly positioned between the speakers. The role of the nurse or the parent seated near the child was first to reassure himyher and second to control the child’s head position. The speakers were placed on small tables beside each arm of the armchair and the adult sat near a speaker, either to the right or to the left of the child. The EEG was recorded from 5 Ag–Ag Cl cup electrodes (Cz, Fz, Pz, T3, T4) with linked earlobes as reference. Eye movements were recorded from two supraorbital linked electrodes, also referred to linked earlobes. The EEG was sampled on-line at a rate of 250 Hz for 1 s duration (500 ms pre- and post-stimulus). The stimuli comprised 750-Hz tone bursts of 200 ms duration with risey fall times of 20 ms delivered through two speakers placed 50 cm directly in front of each ear. Each subject received a randomized series of tones of varying intensity. Interstimulus interval varied from 3 to 5 s. All trials contaminated by ocular movements were rejected on-line (criterion 100 mV) and the sequence was then prolonged (the stimuli always being delivered in a pseudo-random order) until a total of 40 trials for each stimulus intensity had been collected. The percentages of rejected (and then added) trials were between 5 and 25% in normal children and between 10 and 35% in patients. Data analysis was performed on auditory evoked potentials recorded at stimulus intensity levels of 50, 60, 70 and 80 dB SPL.
The negative peak occurring during the 100– 200 ms period after stimulus onset were measured at each temporal site (T3, T4). Manual scoring using a cursor program was used to measure amplitude and latency of the peak according to the mean of the 500-ms prestimulus baseline. The N1c peak culminates at approximately 170 ms at temporal sites in 4- 8-year-old children (Bruneau et al., 1997). 2.3. Data analysis Amplitude and latency data of temporal N1c peak were analyzed using repeated measures ANOVA. The between-subjects variable was group (two levels: AUT and NOR). The two within-subject variables were site (left and right temporal, i.e. T3 and T4) and intensity (four levels: 50, 60, 70, 80 dB SPL). In all ANOVAs involving repeated measures effects with more than two levels, Greenhouse–Geisser epsilon corrections were used to adjust probabilities for repeated measures F-values. The results of such ANOVA indicated successively: the uncorrected degrees of freedom, F value, the epsilon-corrected P value and the corresponding epsilon value. Newman–Keuls tests were used for post hoc comparisons. Correlations between electrophysiological and behavioral data were performed using Spearman rank correlations, which were computed between the AEP indices (latency, amplitude) and verbal and non-verbal level (QDV and QDNV) and, on the other hand, communication disabilities assessed using the BSE scale. Items selected were item 5 (does not make an effort to communicate using voice andyor words), item 6 (lack of appropriate communicative gestures and facial expression) and item 24 (bizarre responses to auditory stimuli). Significance took into account the Bonferroni correction (Bland and Altman, 1995). 3. Results Inter-group comparisons of late auditory temporal responses (N1c) 3.1. Amplitude Overall ANOVA performed on N1c amplitude indicated a significant main effect of Group
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
21
Fig. 1. Mean amplitude and S.E.M. of significant N1c in 16 normal children (NOR) and 26 children with autism (AUT) as a function of stimulus intensity level. AUT showed smaller amplitudes than NOR on both sides at all intensity levels (*P-0.002). At 80 dB SPL, responses were greater at right than at left temporal site for AUT (*P-0.006).
wF(1, 40)s14.85; P-0.0005x, and Intensity wF(3, 120)s10.66; P-0.00001; ´s0.89x and a significant Group=Hemisphere=Intensity interaction wF(3, 120)s6.78; P-0.0005; ´s0.89x. N1c amplitude was smaller on both sides and at all intensity levels in AUT than in NOR children (Fig. 1). ANOVA performed at each intensity level indicated that the Group=Hemisphere interaction was only significant at 80 dB wF(1,40)s11.36; P-002x. At this high level of intensity, planned comparisons indicated that N1c amplitude was significantly greater on the right hemisphere than on the left in AUT wF(1,40)s8.72; P-0.005x, the difference in interhemisphere amplitude being reversed (LH) RH) but did not reach significance in NOR wF(1,40)s3.87; Ps0.06x (Fig. 1). Grand average auditory responses recorded in both groups of subjects at the highest level of intensity are shown on Fig. 2.
6.58; P-0.0004; ´s0.86x but no side effect. Moreover, no significant interaction was found. Post hoc comparisons indicated that N1c latency was longer in AUT than in NOR on both sides and at all intensity levels. Latency decreased with increasing stimulus intensity in both groups and on both sides (see Table 1). 3.3. Relations between temporal AEPs and communication abilities (verbal, non-verbal) in AUT: The electrophysiological results detailed above indicated that preferential activation of the right
3.2. Latency Overall ANOVA performed on N1c latency indicated a significant main effect of Group wF(1, 40)s6.86; P-0.02x, and Intensity wF(3, 120)s
Fig. 2. Grand averaged AEP elicited by 80 dB SPL tones for normal children (NOR) and children with autism (AUT).
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
22
Table 1 Mean (S.E.M.) peak latency for N1c waves measured at temporal sites (left: T3; right: T4) in normal children (NOR) and children with autism (AUT) 50 dB T3 T4
NOR AUT NOR AUT
176 182 176 191
(3) (5) (3) (3)
60 dB 168 179 169 185
(3) (5) (5) (5)
70 dB 163 174 167 183
(3) (5) (3) (3)
80 dB 170 179 167 184
(3) (6) (4) (5)
hemisphere occurs in AUT for the highest level of stimulus intensity (80 dB SPL). This intensity level was therefore chosen to calculate electrobehavioural relations. Although the amplitude of N1c was not correlated with VDQ on the left side (rs0.30), the correlation was significant on the right side (rs0.45; P-0.02): the greater the N1c amplitude at the right temporal electrode, the higher the verbal quotient. This was confirmed by results obtained by considering relations between N1c amplitude and scores for communication skills evaluated using
items 5 (verbal communication) and 6 (use of gestures to communicate) of the BSE scale. Scores for these respective items were significantly correlated with N1c amplitude on the right side (rs y0.60; y0.58; P-0.002) whereas they were not correlated with N1c amplitude on the left side (rs y0.40; y0.21). On the BSE scale, the higher the score for each item, the greater the severity of the disturbance. Negative correlation therefore indicated that the greater the N1c wave on the right side, the lower the score, i.e. the better the communication abilities were, using verbal and non-verbal skills. The correlations obtained between N1c amplitude and the mean value of E5 and E6 were calculated for each intensity level (50, 60, 70, 80 dB). Results were, respectively, y0.37, y0.13, y0.21, y0.32 (all ns) on the left side and y0.42, P-0.05; y 0.44, P-0.02; y0.56, P-0.005; y0.61, Ps 0.001 on the right side. After Bonferroni correction, results were significant at both 70 and 80 dB. Data obtained at the highest intensity level are shown on Fig. 3.
Fig. 3. Correlation between N1c amplitude at right temporal site at 80 dB SPL and mean score for items E5 and E6 of BSE-R scale evaluating verbal and non-verbal communication disorders. The greater the N1c amplitude at T4, the smaller the scores for E5 and E6, i.e. the better the verbal and non-verbal communication skills.
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
No significant correlations were found between the item evaluating abnormal behavioral reaction to sound and N1c amplitude, either on the right (rsy0.18) or on the left side (rsy0.16). No significant correlation was found when considering N1c latency and behavioral evaluations (all r0.2). 4. Discussion These electrophysiological results concerning temporal auditory responses (N1c) in autism confirm those previously obtained in a smaller sample of children (Bruneau et al., 1999). Bitemporal electrophysiological hyporeactivity was obtained in children with autism, revealed by N1c, which was of smaller amplitude and longer latency than in normal children. Moreover, although N1c amplitude increased with increasing stimulus intensity in NOR, this effect was only found on the right side in AUT. A particular pattern of asymmetry was therefore found at the highest level of stimulus intensity in AUT, indicating greater activation of the right temporal region. One explanation might be that the homologous regions in the left hemisphere has a much higher ‘threshold’ of intensityrelated cortical response than the right, which might be due to a particular developmental neural network organization. Since N1c has been demonstrated to be generated in the associative auditory areas located in the lateral part of the superior temporal gyrus (Celesia, 1976; Wood and Wolpaw, 1982; Scherg and Von Cramon, 1986), the bitemporal electrophysiological N1c hyporeactivity might be related to hypoperfusion of both temporal lobes, involving the associative auditory cortex evidenced in PET neuroimaging studies in AUT at rest (Zilbovicius et al., 2000; Ohnishi et al., 2000) and in response to tone stimuli (Muller et al., 1999; Boddaert et al., 2001). Our results are in agreement with results obtained with SPECT on 5- to 11-year-old autistic children after activation by repeated 80 dB SPL tone stimuli (Garreau et al., 1994). Posterior associative temporal areas were activated on the left side in normal children and on the right side in children with autism. A lack of cerebral blood
23
flow modifications in response to tone stimulations had also previously been shown in autistic children using transcranial Doppler ultrasonography (Bruneau et al., 1992). Modifications of AEP observed when stimulus intensity increases, are usually interpreted as reflecting better synchronisation of single trial responses, which lead to greater mean amplitude and shorter latency of peaks. This phenomenon does not occur at all on the left side in AUT. The greater amplitude of N1c might also correspond to stronger activation that spreads over larger cortical temporal areas. Another possible explanation might be an ampliotopic effect on scalp potential distribution of N1c. If localization of the N1c generator is dependent on intensity, a more radial orientation of the generators activated at high intensity level might explain greater N1c amplitude at temporal sites. It might also be hypothesized that N1c involves two (or more) overlapping components that are differentially activated at high intensity levels. This is supported by results showing a broader peak of N1c at T4, with a small shoulder on the descending part of the N1c curve clearly evident at 80 dB (Fig. 2). These different possibilities are probably more complementary than exclusive. Further topographical studies of N1c according to intensity will be required to clarify these different hypotheses. Relations between electrophysiological and behavioral data indicated that the right hemisphere ‘reactivity’ as revealed by N1c amplitude is indicative of verbal and non-verbal communication abilities in children with autism. These results suggest reorganization of the respective roles of the left and right hemispheres in auditory information processing during early brain development rather than dysfunction of the left hemisphere alone. The present results are therefore arguments indicating that when verbal and non-verbal communication are acquired in autism, acquisition may occur via right-hemisphere compensation for early left-hemisphere dysfunction. Right-hemisphere specialization for language is not uncommon in cases of early left-hemisphere trauma (Rasmussen and Milner, 1977). Such a hypothesis would be
24
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25
reinforced by data showing relations between N1c amplitude and language communication in normal children. The lack of relations between N1c amplitude and the score for item 24 (‘Bizarre responses to auditory stimuli’) must be emphasized. This item quantifies excessive, insufficient or selective sensitivity to noise, sounds and speech, as well as paradoxical responses (both hypo- and hyper-reactivity). The behaviors evaluated may be too heterogeneous on this item and may need to be separated to study relations with temporal responses. Another interpretation might be that, as shown by our results, temporal N1c is an index of activity of the cortical networks involved in the cognitive processes of verbal and non-verbal communication rather than of cortical networks involved in sensory processes such as the hyperacusis and hyporeactivity observed in autism. Acknowledgments This study was supported by INSERM U316 Dynamics and Pathology of Cerebral Development (Dir: Pr Pourcelot). We thank Doreen Raine for editing the English. References APA: American Psychiatric Association, 1994. Diagnostic and Statistical Manual for Mental Disorders (DSM IV), 4th edn, Washington, DC: American Psychiatry Press. ´´ Barthelemy, C., Hameury, L., Lelord, G., 1998. Infantile Autism: Exchange and Development Therapy. Expansion ¸ Scientifique francaise. BSE-R scale glossary, pp. 80–84. ´´ ´ Barthelemy, C., Roux, S., Adrien, J.L., Hameury, L., Guerin, P., Garreau, B., et al., 1997. Validation of the revised Behavior Summarized Evaluation scale (BSE-R). J. Autism Dev. Disord. 27, 139–154. Bland, J.M., Altman, D.G., 1995. Multiple significance tests: the Bonferroni method. Br. Med. Journal. 310, 21. Boddaert, N., Belin, P., Poline, J.B., Chabanne, N., Mouren´´ Simeoni, M.C., Barthelemy, C., et al., 2001. Temporal lobe dysfunction in autism: a children PET auditory activation study. Neuroimage 13 (6), S1028. Bruneau, N., Dourneau, M.C., Garreau, B., Pourcelot, L., Lelord, G., 1992. Blood flow response to auditory stimulations in normal, mentally retarded, and autistic children: a preliminary transcranial Doppler ultrasonographic study of the middle cerebral arteries. Biol. Psychiatry 32, 691–699. Bruneau, N., Gomot, M., 1998. Auditory evoked potentials (N1 wave) as indices of cortical development throughout
childhood. In: Garreau, B. (Ed.), Neuroimaging in Child Neuropsychiatric Disorders. Springer Verlag, pp. 113–124. ´´ Bruneau, N., Roux, S., Adrien, J.L., Barthelemy, C., 1999. Auditory associative cortex dysfunction in children with autism: evidence from late auditory evoked potentials (N1 wave—T complex). Clin. Neurophysiol. 110, 1927–1934. ´ ´´ Bruneau, N., Roux, S., Guerin, P., Barthelemy, C., Lelord, G., 1997. Temporal prominence of auditory evoked potentials (N1 wave) in 4- 8-year-old children. Psychophysio 34, 32–38. ´ ´ Brunet, O., Lezine, I., 1976. Echelle de developpement psy` enfance. In: Brunet, O., Lezine, ´ chomoteur de la premiere ´ ` I. (Eds.), Le developpement Psychologique de la Premiere Enfance. PUF, Paris 2nd ed. Celesia, G.G., 1976. Organization of auditory cortical areas in man. Brain 99, 403–414. ` R., Cheour, M., Naatanen, ¨¨ ¨ Ceponiene, R., 1998. Interstimulus interval and auditory event-related potentials in children: evidence for multiple generators. Electroenceph. Clin. Neurophysiol. 108, 345–354. Courchesne, E., Courchesne, R.Y., Hicks, G., Lincoln, A.J., 1985. Functioning of the brain-stem auditory pathway in non-retarded autistic individuals. Electroenceph. Clin. Neurophysiol. 61, 491–501. Dawson, G., Warrenburg, S., Fuller, P., 1982. Cerebral lateralization in individuals diagnosed as autistic in early childhood. Brain Lang. 15, 353–368. Dawson, G., Finley, C., Phillips, S., Lewy, A., 1989. A comparison of hemispheric asymmetries in speech-related brain potentials of autistic and dysphasic children. Brain and language 37, 26–41. ´ Garreau, B., Zilbovicius, M., Guerin, P., Samson, Y., Syrota, A., Lelord, G., 1994. Effects of auditory stimulation on regional cerebral blood flow in autistic children. Dev. Brain Dysfunction 7, 119–128. ´ Giard, M.H., Perrin, F., Echallier, J.F., Thevenet, M., Froment, J.C., Pernier, J., 1994. Dissociation of temporal and frontal components in the human auditory N1 wave: a scalp current density and dipole model analysis. Electroenceph clin Neurophysiol. 92, 238–252. Grandin, T., 1992. An inside view of autism. In: Schopler, E., Mesibov, G.B. (Eds.), High-functioning individuals with autism. Plenum Press, New York, pp. 105–126. Grandin, T., Scariano, M., 1986. Emergence: labeled autistic. Arena, Novato, CA. Grillon, C., Courchesne, E., Akshoomoff, N., 1989. Brainstem and middle latency auditory evoked potentials in autism and developmental language disorder. J. Autism Dev. Disord 19, 255–269. Hayes, R.W., Gordon, A.G., 1977. Auditory abnormalities in autistic children. Lancet 2, 767. Klin, A., 1993. Auditory brainstem responses in autism. Brainstem dysfunction of peripheral hearing loss? J. Autism Dev. Disord. 23, 15–34. McCallum, W.C., Curry, S.H., 1980. The form and distribution of auditory evoked potentials and CNVs when stimuli and responses are lateralized. In: Kornhuber, H.H., Deecke, L.
N. Bruneau et al. / International Journal of Psychophysiology 51 (2003) 17–25 (Eds.), Motivation, motor and sensory processes of the brain: Electrical potentials, behavior and clinical use, Progress in brain research, Vol. 54. Elsevier, Amsterdam, pp. 767–775. Muller, R.A., Behen, M.E., Rothermel, R.D., Chugani, D.C., Muzik, O., Mangner, T.J., et al., 1999. Brain mapping of language and auditory perception in high-functioning autistic adults: A PET study. J. Autism Develop. Disorders 29, 19–31. ¨¨ ¨ Naatanen, R., Picton, T., 1987. The N1 wave of the human electric and magnetic response to sound: a review and an analysis of the component structure. Psychophysio 24, 375–429. Ohnishi, T., Matsuda, H., Hashimoto, T., Kunihiro, T., Nishikawa, M., Uema, T., et al., 2000. Abnormal regional cerebral blood flow in childhood autism. Brain 123, 1838–1844. Pang, E.W., Taylor, M.J., 2000. Tracking the development of the N1 from 3 to adulthood: an examination of speech and non-speech stimuli. Clin. Neurophysiol. 111, 388–397. ´ ´ Perron-Borelli, M., 1978. Les echelles differentielles defficiences intellectuelles EDEI, Manuel. Editions Scientifiques et Psychotechniques, Paris. Ponton, C.W., Eggermont, J.J., Kwong, B., Don, M., 2000. Maturation of human central auditory system activity: evidence from multi-channel evoked potentials. Clin. Neurophysiol. 111, 220–236. Ponton, C.W., Eggermont, J.J., Khosla, D., Kwong, B., Don, M., 2002. Maturation of human central auditory system
25
activity: separating auditory evoked potentials by dipole source modeling. Clin. Neurophysiol. 113, 407–420. Prior, M.R., Bradshaw, J.L., 1979. Hemisphere functioning in autistic children. Cortex 15, 73–81. Rasmussen, T., Milner, B., 1977. The role of early left-brain injury in determining lateralization of cerebral speech functions. Ann. N.Y. Acad. Sci. 299, 355–369. Scherg, M., Von Cramon, D., 1985. Two bilateral sources of the late AEP as identified by a spatio-temporal dipole model. Electroenceph. Clin. Neurophysiol. 62, 32–44. Scherg, M., Von Cramon, D., 1986. Evoked dipole source potentials of the human auditory cortex. Electroenceph. Clin. Neurophysiol. 65, 344–360. Vaughan Jr, H.G., Ritter, W., 1970. The sources of auditory evoked responses recorded from the human scalp. Electroenceph Clin. Neurophysiol. 28, 360–367. Wolpaw, J.R., Penry, J.K., 1975. A temporal component of the auditory evoked response. Electroenceph Clin. Neurophysiol. 39, 609–620. Wood, C.C., Wolpaw, J.R., 1982. Scalp distribution of human auditory evoked potentials. II. Evidence for overlapping sources and involvement of auditory cortex. Electroenceph. Clin. Neurophysiol. 54, 25–38. Woods, D.L., 1995. The component structure of the N1 wave of the human auditory evoked potential. Electroenceph Clin. Neurophysiol. Suppl. 44, 102–109. ´ Zilbovicius, M, Boddaert, N., Belin, P., Poline, J.B., Remy, PH., Mangin, J.F., et al., 2000. Temporal lobe dysfunction in childhood autism: a PET study. Am. J. Psychiatr. 157, 1988–1993.