Correspondence to: Dr. Claire-Dominique Walker, Department of Psychiatry, McGill University, Douglas Hospital Research Centre, 6875 Lasalle Blvd., Verdun QC H4H 1R3; fax 514 762-3034; waldom@douglas.mcgill.ca

Medical subject headings: autism, infantile; autoimmunity; catecholamines; cerebellum; cerebral cortex; child developmental disorders, pervasive; genetics; mental disorders diagnosed in childhood; pregnancy complications; serotonin; temporal lobe

Submitted Aug. 31, 1998
Revised Nov. 10, 1998
Accepted Nov. 17, 1998

© 1999 Canadian Medical Association

• Résumé
• Introduction
• Genetic factors in autism
• Neurobiological aspects of autism
• Conclusion
• References

The etiology of autism is complex, and in most cases the underlying pathologic mechanisms are unknown. Autism is a hetereogeneous disorder, diagnosed subjectively on the basis of a large number of criteria. Recent research has investigated genetics, in utero insults and brain function as well as neurochemical and immunological factors. On the basis of family and twin studies, there appears to be a genetic basis for a wide "autistic syndrome." About a quarter of cases of autism are associated with genetic disorders such as fragile X syndrome or with infectious diseases such as congenital rubella. Genetic studies have shown an association between autism markers of brain development such as 3 markers of the c-Harvey-ras oncogene and the homeobox gene EN2. In some cases, autism is associated with insults early in gestation, including thalidomide embryopathy. Autism may arise from abnormal central nervous system functioning, since most autistic patients have indications of brain dysfunction, and about half of them have abnormal electroencephalograms. Similarly, the pattern of evoked response potentials and conduction time is altered in autistic children. There is substantial evidence from neuroimaging studies that dysfunctions in the cerebellum and possibly the temporal lobe and association cortex occur in autistic symptoms. Neurochemical studies have investigated the role of serotonin, epinephrine and norepinephrine, since levels of these neurotransmitters are altered in autism, although other hypotheses implicate overactive brain opioid systems and changes in oxytocin neurotransmission. Autoimmunity may also play a role; antibodies against myelin basic protein are often found in children with autism, who also have increased eosinophil and basophil response to IgE-mediated reactions. In summary, the prevailing view is that autism is caused by a pathophysiologic process arising from the interaction of an early environmental insult and a genetic predisposition.

[Contents]

[Contents]

Autism is a severe behavioural disorder that develops in the first 3 years of life. It is characterized by pervasive impairments in social interaction, deficits in verbal and nonverbal communication, and stereotyped and repetitive patterns of behaviours and interests. The prevalence in the general population ranges from 0.04% to more than 0.1%,1,2,3,4 and boys are 3 to 4 times more commonly affected than girls. While all children with autism are grossly impaired in their social relations, the severity and specific nature of cognitive impairments vary widely. The intellectual competency of children with autism ranges from profound mental deficiency to superior intelligence, although most such children suffer from mental retardation.

For more than 2 decades after Leo Kanner first described it in 1943, autism was believed to be psychogenic. However, the high incidence of mental deficiency and epileptic seizures pointed to biological determinants. Many studies have reported abnormalities in electroencephalograms (EEGs) of autistic subjects; more recently, structural abnormalities have been found in various regions of the brain.⁵ Genetic influences have also been noted; there are high concordance rates in siblings of autistic probands, as well as in monozygotic twins.⁶ Hence, it has now become generally accepted that autism has a biological basis.

Like most behavioural syndromes, autism is etiologically complex, and the underlying pathologic mechanisms are unknown in most cases. A number of causes have been proposed, including various medical conditions (fragile X syndrome, phenylketonuria, and tuberous sclerosis, among others), perinatal insult and autoimmune mechanisms. Like many other psychiatric conditions, such as bipolar disorder, panic disorder, alcoholism, Alzheimer's disease and schizophrenia,⁷ autism has been shown to have a genetic component, which is now the focus of intensive investigation. There has also been substantial work to elucidate the neurobiological correlates of the disorder. In this review, recent advances in the understanding of the genetics and neurobiology of infantile autism will be discussed.

[Contents]

Epidemiological studies have determined that the risk of autism in the siblings of an autistic child is around 3%,8 which is about 50 times higher than the prevalence rate for the whole population. Moreover, in same-sex twins, there is a much higher concordance rate in monozygotic (MZ) than in dizygotic (DZ) twins. A British twin study conducted in 1977 by Folstein and Rutter⁹ found a 36% concordance for diagnosed autism in MZ twins, compared with no concordance in DZ twins. However, when the data were later reassessed using a broader phenotype of cognitive and social abnormalities, the concordance increased to 82% in MZ twins and 10% in DZ twins.¹⁰ These results were confirmed recently in a large investigation by Bailey et al,⁶ which also revealed that milder symptoms were more common in first-degree relatives of probands with autism than in relatives of nonautistic mentally retarded individuals (with Down syndrome). Indeed, a significant proportion of parents of autistic children exhibit impaired executive function, as evidenced by poor planning skills and attentional flexibility.¹¹ Studies also reported an aggregation of a variety of personality characteristics, language abnormalities and psychiatric disorders in families with autism. It has been hypothesized that this constellation of symptoms may be a manifestation of an underlying genetic susceptibility to autism.^10,12 Thus, the bulk of twin and family studies points to a genetic basis for an autistic spectrum extending beyond classic diagnosed cases.

A variety of mechanisms have been proposed to explain the etiology of autism, including an association with a known medical syndrome. A recent systematic review concluded that autism related to a medical disorder accounts for an average 24.4% of cases across studies.¹³ Such disorders include genetic disorders such as fragile X syndrome (the chromosomal abnormality most frequently associated with autism), tuberous sclerosis, phenylketonuria and Rett syndrome, as well as nongenetic conditions such as infectious diseases (e.g., congenital rubella, acute encephalopathy, cytomegalovirus).^14,15 Although the frequency of these disorders in the population with autism is far greater than would be expected by chance, none of these disorders is consistently associated with typical or atypical autism, highlighting the heterogeneity of the autistic spectrum. Association of autism with genetic diseases could result from close linkage of the 2 genes that, separately, cause both disorders.¹⁶ In this perspective, the variety of genetic defects associated with autism seems to indicate multiloci inheritance of the disorder, and the frequent association of fragile X syndrome with autism supports at least partial X-linked inheritance of a genetic predisposition to autism.^17,18 An association has also been reported between chronic intestinal inflammation (ileal lymphoid nodular hyperplasia) and cognitive and behavioural regression in boys with a history of normal development.¹⁹ Onset of autistic-like symptoms has also been reported by parents to coincide with a vaccination or infection, highlighting the possible importance of inflammatory or immunological responses in the association between autism and other medical conditions.

It has been hypothesized that one of the underlying causes of autism is the brain abnormalities that accompany some of these medical conditions. If this is the case, clinical presentation and severity of autistic symptoms must depend on the brain systems specifically affected.^20,21 In support of this hypothesis, it was recently reported that some children with fetal alcohol syndrome (FAS), who are generally most impaired in the area of interpersonal skills,²² have been given the diagnosis of autism. Interestingly, animal models of FAS exhibit cerebellar abnormalities,²³ which are a key finding in the neuropathology of autism.

In cases not related to a known medical condition, various mechanisms have been put forward to explain the patterns of familial aggregation. These include autosomal recessive inheritance, X-linked inheritance, and multifactorial and polygenic models.¹⁰ Some authors have reported that autism sometimes presents characteristics of genetic anticipation -- increased symptom severity and decreased age of onset with successive generations.^24,25 This indicates that transmission of autism in some cases may involve "unstable DNA" -- that is, genes susceptible to trinucleotide repeat expansion. Interestingly, it was found that autistic individuals have increased expression of autosomal fragile (folate-sensitive) sites compared with control children.²⁶ Segregation analysis in families multiplex for autism (with at least 2 affected members) has suggested autosomal recessive transmission patterns; hence, differential penetrance in males and females could account for the unbalanced sex ratio in autism (with almost 4 times more males than females are affected).¹⁶ However, reported recurrence rates do not seem to fit with single-gene models of inheritance.^27,28 Based on data from twin studies, Bailey et al⁶ proposed that inheritance of the genetic susceptibility to autism is best explained by a multifactorial threshold model involving more than 1 locus, with a higher threshold in females. This type of model predicts a greater frequency of autistic behaviours in relatives of the least affected sex. Most epidemiological findings do not show a greater frequency in families of female autistic probands;¹⁶ however, a recent study reported a higher rate of verbal anomalies in the male relatives of autistic girls,²⁸ providing partial support for the multifactorial inheritance model.

The mode of inheritance proposed by Bailey et al could explain familial aggregation in some pedigrees, but it remains highly probable that autism is genetically heterogenous. Environmental factors, in concert with genetic susceptibility, may play an important role in the determination of the autistic phenotype. Folstein and Rutter⁹ suggested as early as 1977 that genotypic autism is usually expressed as a language disorder, and becomes clinical autism only with the co-occurrence of a secondary factor, such as perinatal brain damage.⁹ Adoption studies have been valuable in determining the relative importance of environmental versus genetic factors in other conditions. The lack of such studies in autism may be explained by the rarity of the disorder and the intrinsic difficulties associated with adoption studies.

Genetic linkage studies

Genetic evidence from case reports points to a variety of chromosomal abnormalities that are occasional causes of autism.^14,20 The most common abnormalities arise from duplication of chromosomal material, frequently chromosome 15 or the sex chromosomes. The 15q region from the maternally derived chromosome was found to be the most commonly affected in autistic individuals. Of possible etiological relevance, this region has also been implicated in the genetic basis of dyslexia, one of the features of the autistic spectrum. As will be discussed later, potential candidate genes from this site include genes coding for 3 *-aminobutyric acid (GABA)-A receptor subunits.^29,30,31

A full genome screen was recently performed in 99 pairs of affected relatives to identify loci associated with autism.³² There were significant association scores for regions on chromosomes 4, 7, 10, 18, 19 and 22, with the highest significance for region 7q. A more specific study investigated linkage of autism with a large number of red cell markers and serum enzymes, but found no positive association.³³ Because of the frequent association of autism with fragile X syndrome, a genetic linkage between autism and the gene responsible for fragile X (FMR-1) has been investigated in families multiplex for autism.³⁴ No involvement of the FMR-1 region was found in autistic individuals who had no cytological evidence of fragile X expression. However, recent findings concerning other markers on the X chromosome support involvement of X-related factors in autism.³⁵ One potential problem with linkage analysis in autism is that such analysis requires homogeneous samples, along with a clear distinction between affected and unaffected individuals. This is clearly not the case in autism; hence, other associative approaches have been attempted. In particular, the association method investigates allele frequency of individual loci ("candidate genes") chosen for their potential "mapping" of specific features of the disorder.^7,36 For instance, autism has often been associated with biochemical abnormalities in the metabolism of serotonin, and drugs that target (5-HT)_T have been found to be effective in relieving autistic symptoms.³⁷ Recent results reveal an association between the autistic phenotype and variants of the serotonin transporter (5-HT)_T gene.³⁷ Similarly, Hérault et al³⁸ tested for an association between autism and a number of candidate genes, including the serotonin (5-HT)_2A receptor gene and genes coding for tyrosine hydroxylase (TH), dopamine ß-hydroxylase and tryptophan hydroxylase. Frequencies of these genes in autistic individuals were no higher than in normal children.³⁹ The lack of association for the TH gene, situated on the short arm of chromosome 11, was later confirmed by Comings et al⁴⁰ using a more variable gene polymorphism. In further investigations of the short arm of chromosome 11, no association between autism and the genes for insulin and insulin-like growth factor 2 were found, but a positive association was shown between autism and the c-Harvey-ras oncogene,⁴¹ and this was confirmed by other investigators.⁴² A subsequent study supported an association of autism with 2 more c-H-ras markers.³⁹ Important functions of ras protein in cell growth, signal transduction, cell architecture and intracellular transport suggest that a mutation of c-H-ras could contribute to autism. Furthermore, a marker of c-H-ras associated with autism was also found to be related to the expression of obsessive­
compulsive and phobic symptoms in patients with Tourette's syndrome.⁴² These association studies make a strong case for a general involvement of the 11p15.5 region in the etiology of childhood autism and other psychiatric disorders. It is noteworthy that the D₄ dopaminergic receptor gene has been localized to this same region of chromosome 11. Research shows that this gene is not related to schizophrenia,⁴³ but its relation to autism has not yet been studied.

Another line of research targeted genes involved in cell growth and development. Based on the hypothesis that a deficiency in adenylosuccinate lyase (ADSL), which is responsible for de novo purine synthesis, may be an etiological factor in certain cases of autism, investigators performed a preliminary genetic screening of 119 autistic patients for point-mutations in the ADSL gene.⁴⁴ However, they found no occurrence of mutations in this particular gene, and it was concluded that ADSL deficiency is, at best, a rare cause of autism. Allele frequencies were also studied for 2 probes of the homeobox gene EN2, a gene located on chromosome 2q13-q21 and involved in cerebellar development.⁴⁵ Significant differences were found between autistic and control subjects for 1 polymorphic site, which may prove especially meaningful in light of recent neuropathological findings of cerebellar abnormalities in autism.

Pre-, peri- and neonatal insults in the etiology of autism

As for other neuropsychiatric disorders, most notably schizophrenia, earlier lines of research focused on the influence of early environmental variables. It was repeatedly suggested that complications during pregnancy and delivery could be causal factors in autism. However, reviews of pre-, peri- and neonatal complications did not prove any consistent or specific association with the autistic syndrome,⁴⁶ although positive associations were found in some cases. For instance, various studies have suggested that autistic symptoms may be associated with events that occur in the first trimester of gestation. There is a solid association between autism and congenital rubella occurring before maturation of the fetal immune response.¹⁴ Similarly, the frequent occurrence of autism in thalidomide-associated embryopathy (in 4% to 5% of cases) is probably related to damage to the central nervous system in the early stages of gestation.⁴⁷ In addition, in pairs of twins discordant for autism, the autistic twin is more frequently afflicted with minor congenital anomalies than the nonautistic twin,⁶ which may reflect some early congenital insult. In addition, elevated maternal age was once considered a risk factor, but recent studies have found no significant differences in maternal age between autistic and control groups.⁴⁸

Patients with many psychiatric conditions -- including schizophrenia, schizoaffective disorder, major depression, bipolar disorder and autism -- have a seasonal birth distribution.⁴⁹ Although the link between this seasonal birth excess and autism has not been firmly established, the effects of weather, differences in nutrition or maternal infections during pregnancy may be involved in the development of autism.⁵⁰

In summary, autism seems to arise from environmental factors interacting with a genetic predisposition -- as yet unidentified -- to a broad "autistic syndrome." The most promising genetic associations thus far have involved genes that code for the serotonin transporter and the c-Harvey-ras oncogene. In addition to genetic investigation, studying the neurobiological correlates of autism could provide important insights to elucidate the etiological and therapeutic aspects of this severely debilitating disorder.

[Contents]

Autism seems to profoundly affect the normal decoding of sensory information. Evoked response potentials (ERPs) were studied to determine whether autistic children have defects in immediate sensory processing at the brain stem level (evidenced by early ERP waves [I-IV]), or whether abnormalities lie in the delayed sensory processing in the cortex (evidenced by long latency potentials and their negative components). No differences between autistic children and controls were found in ERP waves (I-IV), but the amplitude of long latency auditory waves was reduced, and the negative component was absent in autistic subjects.¹⁴ Since both types of affected waves originate in the association cortex, these findings argue for a defect in secondary sensory processing. In addition to differences in the amplitude of the ERP, their conduction time (CCT) can be altered in autism. Indeed, McClelland et al51 found that the CCT was prolonged in autistic children aged 14 years and older compared with age-matched normal controls, and that the CCT was normal in younger autistic children. This prolonged duration of nervous conduction in older autistic children could be explained by a defect in myelination, since, in normal children, the process of myelination leads to progressive shortening of CCT with age.⁵² A myelination defect could also account for the late occurrence of epilepsy in some autistic patients. Interestingly, 58% of autistic individuals, compared with only 9% of control subjects, have antibodies reactive to myelin basic protein (anti-MBP) in their serum.⁵³ An early anti-MBP autoimmune response could result in poor myelination or abnormal function of myelin. Myelination abnormalities could also be caused by perinatal insults, such as anoxia, and this could be a critical factor in the development of the autistic syndrome.⁵³ However, a more precise electrophysiological recording technique revealed anomalies in conduction latency in most autistic individuals, regardless of age,⁵⁴ which puts into question the validity of the abnormal myelination hypothesis.

Neuroimaging and neuropathological studies

The investigation of central nervous system pathology in autism has intensified as a result of the increased availability of structural and functional neuroimaging techniques and detailed histopathology. Early neuroimaging studies have reported various anomalies, including dilation of cerebral ventricles, abnormalities of the basal ganglia, and diverse cortical malformations.⁸ The heterogeneity and nonspecificity of these observations tended to disqualify them as causes or necessary correlates of autism. Although it remains difficult to integrate findings from different neuroimaging studies, the emerging concept is that dysfunctions in autism occur at several levels, including the brain stem, cerebellum, limbic system and association cortex.⁴⁵

Cerebellum: There have been interesting findings from the study of cerebellar size and morphology in autistic individuals. Using magnetic resonance imaging (MRI), Courchesne et al⁵⁵ repeatedly showed hypoplasia of vermian lobules VI and VII in groups of autistic individuals, although these results could not be replicated by others.^56,57,58 A further meta-analysis revealed, however, that one subgroup of autistic individuals consistently shows significant vermian hypoplasia (in approximately 87% of patients), while the remaining 13% of patients show vermian hyperplasia.^59,60 Although abnormality of the cerebellar vermis is a consistent and well-replicated finding in autism, it is difficult to assign morphological changes to the single pathology of autism, since hypoplasia of cerebellar vermian lobules is also seen in various neurogenetic syndromes, with or without autistic behaviours.⁶¹

Postmortem studies of autistic brains have failed to detect any kind of gross pathology. However, recent investigations have reported alterations in neuronal size, density and dendritic branching in the cerebellum and limbic structures of autistic individuals.^62,63 The most consistent finding is a reduction in granule and Purkinje cell density in the neocerebellum.⁶⁰ Cell loss was not associated with gliosis, which indicates that the putative insult must have occurred in the early stages of embryonic development. Cerebellar abnormalities in autism appear to be progressive and age-
related. An autopsy study by Bauman and Kemper⁶⁴ revealed age-dependent abnormalities in neuronal cell size in nuclei of the cerebellum and brain stem in autistic children. The authors speculated that early loss of Purkinje cells in the cerebellum may cause the persistence of fetal circuitry (large cells) in the olivary and cerebellar nuclei of younger subjects. In older autistic children, this circuitry would be progressively lost but not replaced by the adult neuronal pattern, leading to reduced cell size and number. Hashimoto et al⁶⁵ also observed a correlation between area of the vermis and age in autistic but not in control subjects, suggesting an abnormal pattern of vermian growth in autism.

Substantial evidence supports the involvement of the cerebellum in functions affected in the primary symptoms of infantile autism. Animal and human studies have indicated that the cerebellum, and more specifically the cerebellar vermis, is implicated in affect, motivation, social interaction, learning, and in the processing and modulation of sensory and motor information.⁶⁶ A recent neuroanatomical comparison of MZ twins discordant for strictly defined autism showed a markedly smaller caudate, amygdaloid and hippocampal volume, together with smaller cerebellar vermian lobules VI and VII, in the most severely affected twin, compared with his brother.⁶⁷ Finally, neurophysiological studies have revealed that cerebellar damage is associated with impairments in rapid mental attention-shifting in autism.⁶⁸ Positron-emission tomographic studies in autistic individuals have also shown deficiencies in coordinated interaction between the cortical and subcortical systems involved in directed attention.⁶⁹ Since coordination of attentional processes is important in the acquisition of social communication skills,⁷⁰ dysfunctions in these systems may explain the profound impairments in socialization associated with autism.

Temporal lobe: Research on nonhuman primates has demonstrated that insults to the temporal lobe during infancy lead to persistent socio-emotional abnormalities and selective deficits in learning and memory analogous to those seen in autism. There have also been a number of case reports of autistic behaviours in humans following temporal lobe damage,^71,72 including focal temporal lobe abnormality in patients with tuberous sclerosis.⁷³ Another line of evidence implicating temporal lobe dysfunction in autism is the high incidence of temporal lobe epilepsy, as well as the finding of temporal lobe enlargement in autistic individuals.⁷⁴ Various authors^71,75,76,77 have postulated that a developmental lesion in the medial temporal lobe, and more specifically the amygdala, may be a cause of autism. According to this model, precise location and size of the lesion could determine the nature and severity of symptoms.

Association cortex: It is believed that the autistic syndrome cannot be fully explained by localized lesions to subcortical sites, since some autistic dysfunctions, such as social relatedness and complex reasoning, are governed by higher cortical functions. In contrast to other mental disorders, in which a degenerative course is often accompanied by brain atrophy, in autism about 25% of those affected have increased head circumference. This marker appears to be specific to autism, as it is not seen in normal mental retardation or in language disorders.¹⁵ Megalencephaly is typically caused by thickening of the cerebral cortex, and a recent MRI study analyzing the size of the various cortical lobes concluded that autistic subjects exhibit a specific sex-dependent pattern of cortical lobe enlargement, accounting for increased brain size in autistic males.⁷⁴ MRI studies also showed a high incidence of gyri malformations in autistic subjects, while none were detectable in controls.⁷⁸ These cortical malformations, in the absence of gliosis, are thought to result from a defect in the migration of neurons during the first 6 months of fetal development. However, they are not specific to autism, having been reported in association with various other developmental disorders. It is noteworthy that similar migration anomalies have been observed in schizophrenia, in which it is hypothesized that misconnection of brain regions may underlie the neuropathology.⁷⁹ Indeed, some of the manifestations of schizophrenia are closely related to autistic symptoms (disturbed sense of self, social withdrawal and emotional detachment, formal thought disorder including repetitions and stereotypies); and, in fact, autism was once viewed as a form of childhood-onset schizophrenia.

In autism, abnormalities could exist within the frontal cortex or affect the functional linkage between frontal and parietal association cortices.^8,80 One preliminary study using magnetic resonance spectroscopy provided evidence of abnormal phosphorus metabolism and increased degradation of neuronal membranes in the frontal lobe of autistic subjects.⁸¹ This study also showed that the amount of breakdown products from membranes increased as test performance decreased in the autistic group, but not in normal subjects. These findings are consistent with anatomical observations of truncated dendritic branching.^82,83 In addition, there are similarities in cognition, affect and behaviour between autistic individuals and patients with acquired lesions of the frontal lobes.⁸ Frontal lobe processes are required for the formation of second-order cognitive representations, which are dramatically impaired in autism, as well as other high-
order cognitive processes.⁸⁴ However, findings of frontal lobe abnormalities in autism are rare, which has led to the suggestion that there may be transient abnormalities during postnatal development. Regional cerebral blood flow (CBF) was studied in age-matched autistic and normal children at 3 to 4 and 6 to 7 years of age. At first examination, corresponding to the age of frontal maturation in normal children, the autistic group had lower CBF values in the frontal cortex. Three years later, however, their CBF values were not different from those of normal children of the same age.⁸⁴ A functional imaging study in normal volunteers performing specific story-comprehension tasks also provided a glimpse into the neural substrates of autism.⁸⁵ It pointed to a specific pattern of activation associated with normal mental state attribution (which is strikingly affected in autism), including the temporal poles, the left superior temporal gyrus, and the posterior cingulate cortex. These are likely to reflect areas that have abnormal structure or function in autistic brains, and are valuable targets for future research.

Neurochemical studies

The neurochemistry of childhood autism has been the focus of extensive research. The most often replicated finding is that about one-third of autistic individuals have hyperserotonemia.⁸⁶ However, investigation of various other hormone and neurotransmitter systems, as well as drug trials, provide indirect evidence of several other neurochemical abnormalities. Elevated serotonin levels are not specific to autism, being found as well in other neuropsychiatric disorders, such as chronic schizophrenia, and in severe mental retardation.⁸⁶ While an in-depth review of this literature is beyond the scope of this paper, the following discussion gives a brief overview of the salient points of neurochemical investigation in autism.

Serotonin: The interest in assessing serotonergic function in autistic individuals stems from the role of serotonin in the perception and filtering of sensory signals, as well as in social attachment. (Animal research has indicated that hyperserotonemia could reduce the drive for social attachment by inhibiting separation distress.⁸⁷) Several groups have reported high levels of serotonin in the blood and platelets in a subgroup of autistic patients,⁸⁸ and platelet serotonin levels have been correlated with certain autistic behaviours.⁸⁹

However, Hérault et al³⁸ recently questioned the notion of elevated serotonin transmission; their study found evidence of reduced whole-blood serotonin levels in autistic children. As well, there were high urinary levels of 5-hydroxyindoleacetic acid (5-HIAA) in autistic compared with control individuals,³⁸ whereas another study found little or no difference in 5-HIAA concentrations in cerebrospinal fluid (CSF).⁹⁰ Furthermore, the ratio of 5-HIAA to homovanillic acid in CSF was shown to reflect functional competence in autistic children.⁹¹ Todd and Ciaranello⁹² suggested that autistic children may have an autoimmune dysfunction affecting brain serotonin receptors, since 7 out of 13 autistic children, but no control subjects, had CSF antibodies directed against brain 5-HT receptors. However, a more recent study found no differences in brain-tissue (5-HT)_1A and (5-HT)₂ receptor inhibition, on th e basis of immunoglobulins isolated from the plasma of autistic subjects compared with unrelated controls.⁹³ Although disparate, these findings seem to argue for a serotonergic dysfunction in autism, with highly variable manifestations. Further evidence implicating serotonin in the pathology of autism was derived from clinical pharmacological trials. Administration of fenfluramine, a potent serotonin depletor, was associated with clinical
improvement in about one-third of the patients.⁹⁴ Risperidone, a blocker of both serotonin and dopamine receptors, was also effective at relieving the behavioural symptoms of autism, including self-injury.⁹⁴ A recent report showed that tryptophan depletion (with a resultant reduction in serotonin precursor availability) exacerbates autistic symptoms in most patients.⁹⁵ By contrast, trials of selective serotonin reuptake inhibitors, which enhance central serotonergic transmission, also had favourable outcomes: both fluoxetine⁹⁶ and clomipramine⁹⁷ provided promising results in reducing repetitive and compulsive behaviours.

Catecholamines: Animal research has highlighted various functions of the noradrenergic system. In particular, noradrenergic cells from the locus ceruleus have been implicated in attentional processes, filtering of irrelevant stimuli, arousal, anxiety, and learning, which are all impaired functions in autistic patients.⁹¹ Various independent studies have demonstrated that plasma epinephrine and norepinephrine levels are increased in autistic subjects. However, urinary metabolite determinations were much less consistent, and, unlike plasma levels, platelet concentrations of norepinephrine and epinephrine were found to be decreased in autism.⁸⁶ Similarly, dopamine levels were found to be elevated in plasma, but decreased in platelets, of autistic patients.⁹⁸ Investigations of plasma and urinary levels of homovanillic acid (the end-product of the catecholamine degradation pathway) were inconclusive, showing levels ranging from normal to elevated in autistic individuals.⁹⁰ Other evidence of a dopaminergic dysfunction in autism is a blunted or delayed growth-hormone response to levodopa administration,^99,100 which indicates decreased central dopamine transmission.

In terms of plasma monoamine levels, Martineau et al¹⁰¹ found a significant age effect for dopamine, homovanillic acid, 3-methyltryptamine, norepinephrine, epinephrine and serotonin in autistic children. Based on these findings, the authors suggested that autism may be characterized by abnormal maturation of monoamine systems. A few studies have investigated key enzymes of monoaminergic metabolism, but, in general, did not find them to be affected.⁸⁶

Other neurotransmitters and hormones: Preclinical studies in animals suggest that social attachment, an area of functioning severely impaired in autism, is mediated, at least in part, by serotonin, melatonin and opioids.⁸⁷ The opioid hypothesis of autism, supported by studies showing elevated plasma and CSF levels of opioid peptides in autistic individuals, states that overactive brain opioid systems may underlie symptoms of autism, such as lack of socialization and decreased sensitivity to pain.¹⁰² This theory was the impetus for pharmacological trials of opiate blockers. Short-term administration of naltrexone acts to improve hyper-
activity, stereotypies and social withdrawal, while both naltrexone and naloxone reduce self-injurious behaviours.^87,103 It is well known that oxytocin, in addition to opioids, is involved in the formation and maintenance of social bonds. Two independent groups have proposed that social deficits in autism could be mediated by oxytocin hypoactivity¹⁰⁴ or excess.¹⁰⁵ In accordance with the hypoactivity hypothesis, Modahl et al¹⁰⁶ recently found that autistic children have significantly lower plasma oxytocin concentrations than normal children, and do not show the normal increase in oxytocin secretion with age.¹⁰⁶

There are other hypotheses of neurotransmitter alterations. Cohen et al¹⁰⁷ suggested that GABA could mediate social avoidance behaviour in autistic individuals. This could be particular to GABAergic transmission in the Purkinje cells of the cerebellum, which has repeatedly been shown to be affected by a degenerative process in autistic individuals.¹⁰⁷

Investigations into the neuroendocrinology of autism have provided inconclusive data. Cases of improvement after therapy with triiodothyronine (T₃) have been reported, and a high rate of abnormal reponse to thyrotropin-releasing hormone challenge has been demonstrated.⁸⁶ Marginal changes in circadian rhythms of thyroid-stimulating hormone and an abnormal circadian rhythm of melatonin have been seen in severely autistic young adults.¹⁰⁸ Despite normal basal hormonal levels, subgroups of autistic children have been shown to have a blunted or delayed growth hormone response to levodopa,^99,100 a blunted prolactin response to fenfluramine^109,110 and a high rate of cortisol nonsuppression after dexamethasone administration.¹¹¹ Chamberlain and Herman⁸⁷ have proposed a cascade model of biochemical dysfunctions in autism, starting with hypersecretion of pineal melatonin, and alternations in the serotonin system.

Studies of the immune system

Considering the high incidence of autoimmune disorders or abnormalities of the immune system in the autistic population, autoimmunity has been proposed as a possible mechanism underlying autism. Autistic children have been shown to exhibit many anomalies in cell-
mediated immunity, including abnormal T cell activation,¹¹² decreased relative numbers of helper-inducer lymphocytes, and a lower helper-suppressor ratio.¹¹³ These last 2 measures were inversely correlated with severity of autistic symptoms. Autistic children also demonstrate increased urinary levels of native neopterin, a marker of cellular immunity activation,^114,115 and abnormally elevated *-interferon levels in the plasma, which could influence cell growth, activation and proliferation.¹¹⁶ Interestingly, withdrawal and loss of interest, resembling symptoms of autism, have been observed in children receiving interferon therapy for cancer.¹¹⁷ Furthermore, as discussed earlier in this review, antibodies against myelin basic protein are frequently found in children with autism and might be implicated in abnormal myelination, leading to symptoms of autism.⁵¹ One study reported that autistic subjects and their first-degree relatives have antibodies to neuron-axon filament proteins, and that treatment with immunomodulants resulted in symptom improvement in 6 of 8 autistic patients.¹¹⁸

Autoimmune diseases are also frequently triggered by micro-organisms, and it is thought that an immune response to brain antigens resulting from pre- or postnatal viral infection plays a role in the development of the autistic phenotype.⁵³ Since autistic subjects have an increased frequency of the null allele for the complement component 4-binding protein (C4bp),¹¹⁹ partial complement deficiency could result in prolonged immunological stimuli and eventually trigger an autoimmune reaction. Allergic reactions may also play a role in autistic symptoms. Autistic children appear to have an increased eosinophil and basophil response to IgE-mediated reactions, which could explain the high prevalence of allergic disorders among autistic chidren.^120,121 Investigators have recently suggested that food allergy and toxicity of food peptides affecting the central nervous system may be involved in physiopathological processes leading to autism. Lucarelli et al¹²² showed that antibody reactivity to certain common foods is higher in autistic than in normal children, and elimination of these foods from the diet of autistic subjects significantly improved their behavioural symptoms.¹²²

While the significance of these immunological findings in the development of autism is unclear, evidence is mounting that some manifestations of an impaired immune response might contribute to some aspects of the symptoms. More efforts are required to understand the nature of the interactions between the immune and neurobiological systems in autism and to design treatment strategies that account for altered immune function in autistic individuals.

To summarize, it should be noted that the causes underlying neurochemical abnormalities in autism are rarely known and that the results of neurochemical studies can often be obscured by the various medications administered to relieve autistic symptoms. It is also important to keep in mind that abnormalities in neurochemical transmission during brain development could have complex and lasting imprinting effects. Thus, research would benefit from increased emphasis on the timing and localization of neurochemical dysfunctions, as well as relating these anomalies to neuropathological findings.

[Contents]

Autism is a developmental disorder involving changes in neural organization. Evidence from neuropathological as well as epidemiological studies suggests that autism is best viewed as a convergent behavioural manifestation of various brain dysfunctions with different causes. In spite of our limited knowledge regarding the etiology of autism, we can hypothesize that the pathophysiological process is initiated with the interaction of a highly variable early environmental insult and a general genetic predisposition. A very similar causal mechanism has been proposed to explain the etiology of schizophrenia.⁷⁹ A general liability to neuropsychiatric disorders could encompass a variety of diagnoses and give rise to defined phenotypes when paired with a lesion at a specific site and at a specific time during central nervous system development. The epidemiological link between autism and thalidomide embryopathy,⁴⁷ as well as neuropathological reports of cerebellar cell loss in the absence of gliosis,⁶⁰ suggest that the precipitating insult to the developing central nervous system likely occurs in early pregnancy. In fact, based on neuroanatomical data of cerebellar abnormalities, some investigators have proposed that the fifth week of pregnancy is a period of increased vulnerability.123

Although certain findings may be integrated into coherent models, heterogeneity remains a central feature of autism. The diagnosis of autism is and has always been subjective and relies on meeting a defined number of criteria out of a large set. No single manifestation is pathognomonic, and some symptoms are also seen in other mental disorders, such as mental retardation, developmental dysphasia, obsessive­compulsive disorder, and learning disabilities. It has been suggested that future research should focus on identifying homogenous groups in order to derive more conclusive results.¹²⁴ However, it might prove difficult to define boundaries between different "autistic diseases," be they defined by neuropsychological testing, degree of genetic loading, specific brain pathology, or other more or less artificial criteria. Identification of efficient strategies for the prevention and cure of the disorder is hampered by the substantial variation in autistic manifestations among patients. Further work is expected to provide glimpses into neuronal substrates and genetic determinants that will explain the heterogeneity of autism and thus provide hope of alleviating, if not curing, this deeply debilitating lifelong disorder.

Send a letter to the editor Envoyez une lettre à la rédaction

1. Gillberg C, Steffenburg S, Schaumann H. Autism — epidemiology: Is autism more common now than 10 years ago ? Br J Psychiatry 1991;158:403-9. [MEDLINE]
2. Gillberg C, Coleman M. Autism and medical disorders: a review of the literature. Dev Med Child Neurol 1996;38(3):191-202. [MEDLINE]
3. Sugiyama T, Takei Y, Abe T. The prevalence of autism in Nagoya, Japan. In: Naruse H, Ornitz EM, editors. Neurobiology of infantile autism. Amsterdam: Excerpta Medica; 1992. p. 181-4.
4. Bryson SE, Clark BS, Smith IM. First report of a Canadian epidemiological study of autistic syndromes. J Child Psychol Psychiatry 1988;29:433-45. [MEDLINE]
5. Kemper TL, Bauman ML. The contribution of neuropathologic studies to the understanding of autism. Neurol Clin 1993;11(1):175-87. [MEDLINE]
6. Bailey A, Le Couteur A, Gottesman I, Bolton P, Simonoff E, Yuzda E, et al. Autism as a strongly genetic disorder: evidence from a British twin study. Psychol Med 1995;25(1):63-77. [MEDLINE]
7. Crowe RR. Genetics. In: Bloom FE, Kupfer DJ, editors. Psychopharmacology: the fourth generation of progress. New York: Raven Press; 1995. p. 1821-33.
8. Bailey A. The biology of autism. Psychol Med 1993;23(1):7-11. [MEDLINE]
9. Folstein S, Rutter M. Infantile autism: a genetic study of 21 twin pairs. J Child Psychol Psychiatry 1977;18:291-321. [MEDLINE]
10. Folstein SE, Piven J. Etiology of autism: genetic influences. Pediatrics 1991;87(5 Pt 2):767-73. [MEDLINE]
11. Hughes C, Leboyer M, Bouvard M. Executive function in parents of children with autism. Psychol Med 1997;27(1):209-20. [MEDLINE]
12. Bolton P, Macdonald H, Pickles A, Rios P, Goode S, Crowson M, et al. A case–control family history study of autism. J Child Psychol Psychiatry 1994;35(5):877-900. [MEDLINE]
13. Gillberg C, Coleman M. Autism and mental disorders : a review of the literature. Dev Med Child Neurol 1996;38(3):191-202. [MEDLINE]
14. Lotspeich LJ, Ciaranello RD. The neurobiology and genetics of infantile autism (review). Int Rev Neurobiol 1993;35:87-129. [MEDLINE]
15. Bailey A, Philips W, Rutter M. Autism: towards an integration of clinical, genetic, neuropsychological, and neurobiological perspectives. J Child Psychol Psychiatry 1996;37(1):89-126. [MEDLINE]
16. Smalley SL. Genetic influences in autism. Psychiatr Clin North Am 1991;14(1):125-39. [MEDLINE]
17. Hunt A, Shepherd C. A prevalence study of autism in tuberous sclerosis. J Autism Dev Disord 1993;23(2):323-39. [MEDLINE]
18. Lida-Pulik H, Basquin M. Dysfonctionnement cognitif et manifestations psychiatriques dans le syndrome de l'X fragile. Autisme et X fragile. Psychiatr Enfant 1993;36(1):5-26. [MEDLINE]
19. Wakefield AJ, Murch SH, Anthony A, Linnell J, Casson DM, Malik M, et al. Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children. Lancet 1998;351(9103):637-41. [MEDLINE]
20. Steffenburg S. Neuropsychiatric assessment of children with autism: a population-based study. Dev Med Child Neurol 1991;33(6):495-511. [MEDLINE]
21. Rapin I. Autistic children: diagnosis and clinical features. Pediatrics 1991;87(5 Pt 2):751-60. [MEDLINE]
22. Thomas SE, Kelly SJ, Mattson SN, Riley EP. Comparison of social abilities of children with fetal alcohol syndrome to those of children with similar IQ scores and normal controls. Alcohol Clin Exp Res 1998;22(2):528-33. [MEDLINE]
23. Nanson JL. Autism in fetal alcohol syndrome: a report of six cases. Alcohol Clin Exp Res 1992;16(3):558-65. [MEDLINE]
24. Ross CA, McInnis MG, Margolis RL, Li SH. Genes with triplet repeats: candidate mediators of neuropsychiatric disorders. Trends Neurosci 1993;16(7):254-60. [MEDLINE]
25. Margolis RL, Stine OC, McInnis MG, Ranen NG, Rubinstein DC, Leggo J, et al. cDNA cloning of a human homologue of the Caenorhabditis elegans cell fate-determinig gene mab-21: expression, chromosomal localization, and analysis of a highly polymorphic (CAG)n trinucleotide repeat. Hum Mol Genet 1996;5(5):607-16. [MEDLINE]
26. Arrieta I, Nunez T, Gil A, Flores P, Usobiaga E, Martinez B. Autosomal folate sensitive fragile sites in an autistic Basque sample. Ann Genet 1996;39(2):69-74. [MEDLINE]
27. Debraekeleer M, Tremblay M, Thivierge J. Genetic analysis of genealogies in mentally retarded autistic probands from Saguenay Lac-Saint-Jean (Québec, Canada). Ann Genet 1996;39(1):47-50. [MEDLINE]
28. Plumet MH, Goldblum MC, Leboyer M. Verbal skills in relatives of autistic females. Cortex 1995;31(4):723-33. [MEDLINE]
29. Baker P, Piven J, Schwartz S, Patil S. Brief report: duplication of chromosome 15q11-13 in two individuals with autism. J Autism Dev Disord 1994;24(4):529-35. [MEDLINE]
30. Hotopf M, Bolton P. A case of autism associated with partial tetrasomy 15. J Autism Dev Disord 1995;25(1):41-9. [MEDLINE]
31. Schroer RJ, Phelan MC, Michaelis RC, Crawford EC, Skinner SA, Cuccaro M, et al. Autism and maternally derived aberrations of chromosome 15q. Am J Med Genet 1998;76(4):327-36. [MEDLINE]
32. Bailey A. A full genome screen for autism with evidence for linkage to a region on chromosome 7q. Hum Mol Genet 1998;7(3):571-8. [MEDLINE]
33. Spence MA, Ritvo ER, Marazita ML, et al. Gene mapping studies with the syndrome of autism. Behav Genet 1985;15:1-13. [MEDLINE]
34. Hallmayer J, Pintado E, Lotspeich L, Spiker D, McMahon W, Petersen PB, et al. Molecular analysis and test of linkage between the FMR-1 gene and infantile autism in multiplex families. Am J Hum Genet 1994;55(5):951-9. [MEDLINE]
35. Petit E, Hérault J, Martineau J, Raynaud M, Cherpi C, Perrot A, et al. X chromosome and infantile autism. Biol Psychiatry 1996;40(6):457-64. [MEDLINE]
36. Lombroso PJ, Pauls DL, Leckman JF. Genetic mechanisms in childhood psychiatric disorders (review). J Am Acad Child Adolesc Psychiatry 1994;33(7):921-38. [MEDLINE]
37. Klauck SM, Poutska F, Benner A, Lesch KP, Poutska A. Serotonin transporter (5-HTT) gene variants associated with autism. Hum Mol Genet 1997;6(13):2233-8. [MEDLINE]
38. Hérault J, Petit E, Martineau J, Cherpi C, Perrot A, Barthélémy C, et al. Serotonin aud autism: biochemical and molecular biology features. Psychiatry Res 1996;65(1):33-43. [MEDLINE]
39. Hérault J, Petit E, Martineau J, Perrot A, Lenoir P, et al. Autism and genetics — Clinical approach and association study with two markers of the HRAS gene. Am J Med Genet 1995;60(4):276-81. [MEDLINE]
40. Comings DE, Gade R, Muhleman D, Sverd J. No association of a tyrosine hydroxylase gene tetranucleotide repeat polymorphism in autism, Tourette syndrome, or ADHD. Biol Psychiatry 1995;7(7):484-6. [MEDLINE]
41. Hérault J, Martineau J, Petit E, Perrot A, Sauvage D, Barthélémy C, et al. Genetic markers in autism: association study on short arm of chromosome 11. J Autism Dev Disord 1994;24(2):233-6. [MEDLINE]
42. Comings DE, Wu S, Chiu C, Muhleman D, Sverd J. Studies of the c-Harvey-Ras gene in psychiatric disorders. Psychiatry Res 1996;63(1):25-32. [MEDLINE]
43. Maier W, Schwab S, Hallmayer J, Ertl MA, Minges J, Ackenheil M, et al. Absence of linkage between scizophrenia and the dopamine D4 receptor gene. Psychiatry Res 1994;53(1):77-86. [MEDLINE]
44. Fon EA, Sarrazin J, Meunier C, Alarcia J, Shevell MI, Philippe A, et al. Adenylsuccinate lyase (ADSL) and infantile autism: absence of previously reported point mutation. Am J Med Genet 1995;60(6):554-7. [MEDLINE]
45. Petit E, Hérault J, Martineau J, Perrot A, Barthélémy C, Hameury L, et al. Association study with two markers of a human homeogene in autism. J Med Genet 1995;32(4):269-74. [MEDLINE]
46. Nelson KB. Prenatal and perinatal factors in the etiology of autism. Pediatrics 1991;87(5 Pt 2):761-6. [MEDLINE]
47. Strömland K, Nordin V, Miller M, Åkerström B, Gillberg C. Autism in thalidomide embryopathy: a population study. Dev Med Child Neurol 1994;36:351-6. [MEDLINE]
48. Mouridsen SE, Rich B, Isager T. Brief report: parental age in infantile autism, autistic-like conditions, and borderline childhood psychosis. J Autism Dev Disord 1993;23(2):387-96. [MEDLINE]
49. Torrey EF, Miller J, Rawlings R, Yolken RH. Seasonality of births in shizophrenia and bipolar disorder: a review of the literature (review). Schizophr Res 1997;28(1):1-38. [MEDLINE]
50. Bolton P, Pickles A, Harrington R, MacDonald H, Rutter M. Season of birth issues, approaches and findings for autism. J Child Psychol Psychiatry 1992;33:509-31. [MEDLINE]
51. McClelland RJ, Eyre DG, Watson D, Calvert GJ, Sherrard E. Central conduction time in childhood autism. Br J Psychiatry 1992;160:659-63. [MEDLINE]
52. Houston HG, McClelland RJ. Age and gender contributions to intersubject variability of the auditory brainstem potentials. Biol Psychiatry 1985;20:419-30. [MEDLINE]
53. Singh VK, Warren RP, Odell JD, Warren WL, Cole P. Antibodies to myelin basic protein in children with autistic behavior. Brain Behav Immun 1993;7(1):97-103. [MEDLINE]
54. Thivierge J, Bédard C, Côté R, Maziade M. Brainstem auditory evoked response and subcortical abnormalities in autism. Am J Psychiatry 1990;147:1609-13. [MEDLINE]
55. Courchesne E, Yeung-Courchesne R, Press GA, Hesselink JR, Jernigan TL. Hypoplasia of cerebellar vermal lobules VI and VII in autism. N Engl J Med 1988;318:1349-54. [MEDLINE]
56. Garber HJ, Ritvo ER. Magnetic resonance imaging of the posterior fossa in autistic adults. Am J Psychiatry 1992;149:245-7. [MEDLINE]
57. Holttum JR, Minshew NJ, Sanders RS, Phillips NE. Magnetic resonance imaging of the posterior fossa in autism. Biol Psychiatry 1992;32:1091-101. [MEDLINE]
58. Piven J, Nehme E, Simon J, Barta P, Pearlson G, Folstein SE. Magnetic resonance imaging in autism: measurement of the cerebellum, pons and fourth ventricle. Biol Psychiatry 1992;31:491-504. [MEDLINE]
59. Courchesne E, Saitoh O, Townsend JP, Yeung-Courchesne R, Press GA, Lincoln AJ, et al. Cerebellar hypoplasia and hyperplasia in infantile autism. Lancet 1994;343(8888):63-4. [MEDLINE]
60. Courchesne E, Townsend J, Saitoh O. The brain in infantile autism: posterior fossa structures are abnormal. Neurology 1994;44(2):214-23. [MEDLINE]
61. Schaefer GB, Thompson JN Jr, Bodensteiner JB, McConnel JM, Kimberling WJ, Gay CT, et al. Hypoplasia of the cerebellar vermis in neurogenetic syndromes. Ann Neurol 1996; 39:382-5. [MEDLINE]
62. Bauman ML, Courchesne E, Denckla MB, Folstein SE, James LS, Minshew NJ, et al. An update on autism: a developmental disorder — introduction. Pediatrics 1991;87(5 Pt 2):v-vi.
63. Raymond GV, Bauman ML, Kemper TL. Hippocampus in autism — a Golgi analysis. Acta Neuropathol (Berl) 1996;91(1):117-9. [MEDLINE]
64. Bauman ML, Kemper TL. Abnormal cerebellar circuitry in autism ? Neurology 1989;39(Suppl 1):186.
65. Hashimoto T, Tayama M, Miyazaki M, Murakawa K, Shimakawa S, Yoneda Y, Kuroda Y. Brainstem involvement in high functioning autistic children. Acta Neurol Scand 1993;88(2):123-8. [MEDLINE]
66. Schmahmann JD, Sherman JC. The cerebellar cognitive affective syndrome. Brain 1998;121(4):561-79. [MEDLINE]
67. Kates WR, Mostofsky SH, Zimmerman AW, Mazzocco MM, Landa R, Warsofsky IS, et al. Neuroanatomical and neurocognitive differences in a pair of monozygous twins discordant for strictly defined autism. Ann Neurol 1998;43(6):782-91. [MEDLINE]
68. Courchesne E, Townsend J, Akshoomoff NA, Saitoh O, Yeung-Courchesne R, Lincoln AJ, et al. Impairment in shifting attention in autistic and cerebellar patients. Behav Neurosci 1994;108(5):848-65. [MEDLINE]
69. Courchesne E. Neuroanatomic imaging in autism. Pediatrics 1991;87(5 Pt 2):781-90. [MEDLINE]
70. Lenti C, Perruzi C, Bianchini E. Brief report: the association betwen autism and fragile X syndrome: a case report. J Autism Dev Disord 1995;25(6):655-63. [MEDLINE]
71. Bachevalier J. Medial temporal lobe structures and autism: a review of clinical and experimental findings. Neuropsychologia 1994;32(6):627-48. [MEDLINE]
72. Bachevalier J. Brief report: medial temporal lobe and autism: a putative animal model in primates. J Autism Dev Disord 1996;26(2):217-21. [MEDLINE]
73. Bolton PF, Griffiths PD. Association of tuberous sclerosis of temporal lobes with autism and atypical autism. Lancet 1997;349:392-5. [MEDLINE]
74. Piven J, Arndt S, Bailey J, Andreasen N. Regional brain enlargement in autism — a magnetic resonance imaging study. J Am Acad Child Adolesc Psychiatry 1996;35(4):530-6. [MEDLINE]
75. Hoon AH Jr, Reiss AL. The mesial-temporal lobe and autism: case-report and review. Dev Med Child Neurol 1992;34(3):252-9. [MEDLINE]
76. Gillberg IC. Autistic syndrome with onset at age 31 years: herpes encephalitis as a possible model for childhood autism. Dev Med Child Neurol 1991;33(10):920-4. [MEDLINE]
77. Ghaziuddin M, Tsai LY, Eilers L, Ghaziuddin N. Brief report: autism and herpes simplex encephalitis. J Autism Dev Disord 1992;22(1):107-13. [MEDLINE]
78. Piven J, Berthier ML, Starkstein SE, Nehme E, Pearlson G, Folstein S. Magnetic resonance imaging evidence for a defect of cerebral cortical development in autism. Am J Psychiatry 1990;147(6):734-9. [MEDLINE]
79. Weinberger DR, Lipska BK. Cortical maldevelopment, anti-psychotic drugs, and schizophrenia: a search for common ground. Schizophr Res 1995;16:87-110. [MEDLINE]
80. Minshew NJ. Indices of neural function in autism: clinical and biologic implications. Pediatrics 1991;87(5 Pt 2):774-80. [MEDLINE]
81. Minshew NJ, Goldstein G, Dombrowski SM, Panchalingam K, Pettegrew GW. A preliminary 31P MRS study of autism: evidence for undersynthesis and increased degradation of brain membranes. Biol Psychiatry 1993;33:762-73. [MEDLINE]
82. Lotspeich LJ. Autism and pervasive developmental disorders. In: Bloom FE, Kupfer DJ, editors. Psychopharmacology: the fourth generation of progress. New York: Raven Press; 1995. p. 1653-63.
83. Minshew NJ. Brief report: brain mechanisms in autism: functional and structural abnormalities. J Autism Dev Disord 1996;26(2):205-9. [MEDLINE]
84. Zilbovicius M, Garreau B, Samson Y, Remy P, Barthelemy C, Syrota A, et al. Delayed maturation of the frontal cortex in childhood autism. Am J Psychiatry 1995;152(2):248-52. [MEDLINE]
85. Fletcher PC, Happe F, Frith U, Baker C, Dolan RJ, Frackowiak RSJ, Frith CD. Other minds in the brain — a functional imaging study of theory of mind in story comprehension. Cognition 1995;57(2):109-28. [MEDLINE]
86. Cook EH. Autism: review of neurochemical investigation. Synapse 1990;6:292-308. [MEDLINE]
87. Chamberlain RS, Herman BH. A novel biochemical model linking dysfunctions in brain melatonin, proopiomelanocortin peptides, and serotonin in autism. Biol Psychiatry 1990;28:773-93. [MEDLINE]
88. Anderson GM, Freedman CX, Cohen DJ, Volkmar FR, Hoder EL, McPhedran P, et al. Whole blood serotonin in autistic and normal subjects. J Child Psychol Psychiatry 1987;28:885-900. [MEDLINE]
89. Kuperman A, Beeghly JHL, Burns TL, Tsai LY. Association of serotonin concentration to behavior and IQ in autistic children. J Autism Dev Disord 1987;17:133-40. [MEDLINE]
90. Narayan M, Srinath S, Anderson GM, Meundi DB. Cerebrospinal fluid levels of homovanillic acid and 5-hydroxyindoleacetic acid in autism. Biol Psychiatry 1993;33:630-5. [MEDLINE]
91. Young JG, Leven LI, Newcorn JH, Knott PJ. Genetic and neurobiological approaches to the pathophysiology of autism and the pervasive developmental disorders. In: Meltzer HY, editor. Psychopharmacology: the third generation of progress. New York: Raven Press; 1987. p. 825-36.
92. Todd RD, Ciaranello RD. Demonstration of inter- and intraspecies differences in serotonin binding sites by antibodies from an autistic child. Proc Natl Acad Sci U S A 1985;82:612-6. [MEDLINE]
93. Cook EH Jr, Perry BD, Dawson G, Wainwright MS, Leventhal BS. Receptor inhibition by immunoglobulins: specific inhibition by autistic children, their relatives, and control subjects. J Autism Dev Disord 1993;23(1):67-78. [MEDLINE]
94. Aman MG, Kern RA. Review of fenfluramine in the treatment of the developmental disabilities. J Am Acad Child Adolesc Psychiatry 1989;28:549-65. [MEDLINE]
95. McDougle CJ, Naylor ST, Cohen DJ, Aghajanian GK, Heninger GR, Price LH. Effects of tryptophan depletion in drug-free adults with autistic disorder. Arch Gen Psychiatry 1996;53(11):993-1000. [MEDLINE]
96. Cook EH Jr, Rowlett R, Jaselskis C, Leventhal BL. Fluoxetine treatment of children and adults with autistic disorder and mental retardation. J Am Acad Child Adolesc Psychiatry 1992;31(4):739-45. [MEDLINE]
97. Gordon CT, State RC, Nelson JE, Hamburger SD, Rapoport JL. A double-blind comparison of clomipramine, desipramine, and placebo in the treatment of autistic disorder. Arch Gen Psychiatry 1993;50:441-7. [MEDLINE]
98. Israngkun PP, Newman HA, Patel ST, Duruibe VA, Abou-Issa H. Potential biochemical markers for infantile autism. Neurochem Pathol 1986;5(1):51-70. [MEDLINE]
99. Deutsch SI, Campbell M, Sachar EJ, Green WH, David R. Plasma growth hormone response to L-DOPA in infantile autism. J Autism Dev Disord 1985;15:205-12. [MEDLINE]
100. Realmuto GM, Jensen JB, Reeve E, Garfinkel BD. Growth hormone response in autistic children to DA and NE agonists. In: Proceedings of the American Academy of Child and Adolescent Psychiatry. Los Angeles (CA): American Academy of Child and Adolescent Psychiatry; 1986. p. 35.
101. Martineau J, Barthélémy C, Jouve J, Muh JP, Lelord G. Monoamines (serotonin and catecholamines) and their derivatives in infantile autism: age-related changes and drug effects. Dev Med Child Neurol 1992;34(7):593-603. [MEDLINE]
102. Gillberg C, Terenius L, Lönnerholm G. Endorphin activity in childhood psychosis. Arch Gen Psychiatry 1985;42:780-3. [MEDLINE]
103. Campbell M. Drug treatment of infantile autism: the past decade. In: Meltzer HY, editor. Psychopharmacology: the third generation of progress. New York: Raven Press; 1987. p. 1225-31.
104. Modahl C, Fein D, Waterhouse L, Newton N. Does oxytocin mediate social deficits in autism? J Autism Dev Disord 1992;22(3):449-51. [MEDLINE]
105. Panksepp J. Commentary on the possible role of oxytocin in autism. J Autism Dev Disord 1993;23(3):567-9. [MEDLINE]
106. Modahl C, Green L, Fein D, Morris M, Waterhouse L, Feinstein C, et al. Plasma oxytocin levels in autistic children. Biol Psychiatry 1998;43(4):270-7. [MEDLINE]
107. Cohen IL, Sudhalter V, Pfadt A, Jenkins EC, Brown WT, Vietze PM. Why are autism and the Fragile-X syndrome associated? Conceptual and methodological issues. Am J Hum Genet 1991;48:195-202. [MEDLINE]
108. Nir I, Meir D, Zilber N, Knobler H, Hadjez J, Lerner Y. Brief report: circadian melatonin, thyroid-stimulating hormone, prolactin and cortisol levels in serum of young adults with autism. J Autism Dev Disord 1995;25(6):641-54. [MEDLINE]
109. McBride PA, Anderson GM, Hertzig ME, Sweeney JA, Kream J, Cohen DJ, et al. Serotonergic responsivity in male young adults with autistic disorder. Arch Gen Psychiatry 1989;46:205-12. [MEDLINE]
110. McBride PA, Anderson GM, Mann JJ. Serotonin-mediated responses in autism. Biol Psychiatry 1989;25:183A.
111. Jensen JB, Realmuto GM, Garfinkel BD. The dexamethasone suppression test in infantile autism. J Am Acad Child Psychiatry 1985;24:263-5. [MEDLINE]
112. Warren RP, Yonk J, Burger RW, Warren WL. DR-positive cells in autism: association with decreased plasma levels of the complement C4b protein. Neuropsychobiology 1995;31(2):53-7. [MEDLINE]
113. Denney DR, Frei BW, Gaffney GR. Lymphocyte subsets and interleukin-2 receptors in autistic children. J Autism Dev Disord 1996;26(1):87-97. [MEDLINE]
114. Harrison KL, Pheasant AE. Analysis of urinary pterins in autism. Biochem Soc Trans 1995;23:603S. [MEDLINE]
115. Messahel S, Pheasant AE, Pall H, Ahmedchoudhury J, Sungumpaliwal RS. Urinary levels of neopterin and biopterin in autism. Neurosci Lett 1998;241(1):17-20. [MEDLINE]
116. Stubbs G. Interferonemia and autism . J Autism Dev Disord 1995;25(1):71-3. [MEDLINE]
117. Hill NO, Pardue A, Khan A, Aleman C, Dorn G, Hill JM. Phase 1 human leucocyte interferon trials in cancer and leukemia. J Clin Hematol Oncol 1981;11:23-5.
118. Singh VK, Fudenburg HH, Emerson D, Coleman M. Immunodiagnosis and immunotherapy in autistic children. Ann N Y Acad Sci 1988;540:602-4. [MEDLINE]
119. Warren RP, Singh VK, Cole P, Odell JD, Pingree CB, Warren WL, et al. Increased frequency of the null allele at the complement C4b locus in autism. Clin Exp Immunol 1991;83(3):438-40. [MEDLINE]
120. Bidet B, Leboyer M, Descours B, Bouvard MP, Benveniste J. Allergic sensitization in infantile autism (letter). J Autism Dev Disord 1993;23(2):419-20. [MEDLINE]
121. Renzoni E, Beltrami V, Sestini P, Pompella A, Menchetti G, Zapella M. Brief report: allergological evaluation of children with autism. J Autism Dev Disord 1995;25(3):327-33. [MEDLINE]
122. Lucarelli S, Frediani T, Zingoni AM, Ferruzzi F, Giardini O, Quintieri F, et al. Food allergy and infantile autism. Panminerva Med 1995;37(3):137-41. [MEDLINE]
123. Courchesne E. Brainstem, cerebellar and limbic neuroanatomical abnormalities in autism. Curr Opin Neurobiol 1997,7(2):269-78. [MEDLINE]
124. Spiker D, Lotspeich L, Kraemer HC, Hallmeyer J, MacMahon W, Petersen PB, et al. Genetics of autism: characteristics of affected and unaffected children from 37 multiplex families. Am J Med Genet 1994;54(1):27-35. [MEDLINE]