The Early Origins of Autism New research into the causes of this bafﬂing disorder is focusing on genes that control the development of the brain by Patricia M. Rodier A utism has been mystifying scientists for more than half a century. The complex behavioral disorder encompasses a wide variety of symptoms, most of which usually appear before a child turns three. Children with autism are unable to interpret the emotional states of others, failing to recognize anger, sorrow or manipulative intent. Their language skills are often limited, and they ﬁnd it difﬁcult to initiate or sustain conversa- tions. They also frequently exhibit an intense preoccupation with a single subject, activity or gesture. These behaviors can be incredibly debilitating. How can you be included in a typical classroom if you can’t be dissuaded from banging your head on your desk? How can you make friends if your overriding interest is in cal- endars? When children with autism also suffer from mental retardation— as most of them do— the prognosis is even worse. Intensive behavioral therapy improves the outcome for many patients, but their SEVEN-YEAR-OLD WITH AUTISM symptoms can make it impossible for them to live in- reaches for a soap bubble during play- time at the Eden Institute, a school for dependently, even if they have normal IQs. children with autism in Princeton, N.J. 56 Scientiﬁc American February 2000 Copyright 2000 Scientific American, Inc. JUSTINE PARSONS Copyright 2000 Scientific American, Inc. AUTISM’S EFFECTS include changes to the brain stem, the re- gion just above the spinal cord (left). The brain stem of a per- son with autism is shorter than a normal brain stem (be- low): the structures at the junction of the pons and the CEREBRUM medulla (such as the facial nucleus and the trapezoid body) are closer to the structures of the lower medulla (the hypoglossal nucleus and the inferior olive). It is as though a band of tissue were missing. The brain THALAMUS stem of a person with autism also lacks the superior olive and has a smaller-than-normal facial nucleus. Such changes could occur only in early gestation. MIDBRAIN PONS TRAPEZOID TRAPEZOID BODY BODY FACIAL SUPERIOR OLIVE NUCLEUS FACIAL NUCLEUS HYPOGLOSSAL NUCLEUS 0.2-MILLIMETER CEREBELLUM MEDULLA 1.1-MILLIMETER SEPARATION TERESE WINSLOW SEPARATION INFERIOR OLIVE INFERIOR OLIVE BRAIN STEM NORMAL BRAIN STEM BRAIN STEM OF PERSON WITH AUTISM I became involved in the search for long sought to pinpoint exactly when in describing its symptoms. The biologi- autism’s causes relatively recently—and the disorder begins. Previous speculation cal basis for autism, however, has been almost by accident. As an embryologist, I had focused on late gestation or early elusive— an unfortunate circumstance, previously focused on various birth de- postnatal life as the time of origin, but because such an understanding could en- fects of the brain. In 1994 I attended a re- there was no evidence to back up either able researchers to identify the leading markable presentation at a scientiﬁc con- hypothesis. The connection with tha- risk factors for autism and possibly to ference on research into birth defects. lidomide suddenly threw a brilliant new design new treatments for the condition. Two pediatric ophthalmologists, Marilyn light on the subject. It suggested that By examining the inheritance of the T. Miller of the University of disorder, researchers have shown that Illinois at Chicago and Ker- autism runs in families, though not in a stin Strömland of Göteborg University in Sweden, de- At least 16 of every clear-cut way. Siblings of people with autism have a 3 to 8 percent chance of scribed a surprising outcome from a study investigating 10,000 babies are born being diagnosed with the same disor- der. This is much greater than the 0.16 eye motility problems in vic- tims of thalidomide, the with autism or one of its percent risk in the general population but much less than the 50 percent morning-sickness drug that caused an epidemic of birth related disorders. chance that would characterize a genet- ic disease caused by a single dominant defects in the 1960s. The mutation (in which one faulty gene in- study’s subjects were adults who had autism originates in the early weeks of herited from one parent is sufﬁcient to been exposed to the drug while still in the pregnancy, when the embryo’s brain and cause the disorder) or the 25 percent womb. After examining these people, the rest of its nervous system are just be- chance that would characterize a single Miller and Strömland made an observa- ginning to develop. Indeed, Miller and recessive mutation (in which a copy of tion that had somehow eluded previous Strömland’s work convinced me that the the faulty gene must be inherited from researchers: about 5 percent of the mystery of autism could soon be solved. each parent). The results ﬁt best with thalidomide victims had autism, which is models in which variants of several about 30 times higher than the rate Genetic Factors genes contribute to the outcome. To among the general population. complicate matters further, relatives of When I heard these results, I felt a shock of recognition, a feeling so pow- erful that I actually became dizzy and A t least 16 of every 10,000 babies is born with autism or one of its related disorders [see box on page people with autism may fail to meet all the criteria for the disorder but still have some of its symptoms. Although began to hyperventilate. In the effort to 60]. Since autism was ﬁrst identiﬁed in these relatives may have some of the identify autism’s causes, researchers had 1943, scientists have made great strides gene variants linked to autism— what- 58 Scientiﬁc American February 2000 The Early Origins of Autism Copyright 2000 Scientific American, Inc. Thalidomide Timeline Age of embryo (days) 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 MISSING EARS SMALL EARS AND OTHER EAR MALFORMATIONS Damage caused JOHNNY JOHNSON by thalidomide MISSING OR SMALL THUMBS THUMBS WITH AN EXTRA JOINT exposure at STUNTED ARMS this time STUNTED LEGS BIRTH DEFECTS caused by thalidomide vary depending on when the mother was ex- posed to the drug (above). A 1994 study showed that thalidomide victims with autism had ear anomalies and normal limbs, suggesting that the drug triggered the disorder 20 to PETIT FORMAT/NESTLE/SCIENCE SOURCE 24 days after conception, when the embryo’s nervous system is starting to form (right). ever they may be— for some reason the sure in utero. All their subjects— Photo Researchers, Inc. genetic factors are not fully expressed Swedish adults born in the late 1950s in these individuals. and early 1960s— exhibited some of the Studies of twins in the U.K. conﬁrm malformations for which thalidomide that autism has a heritable component is infamous: stunted arms and legs, mis- but suggest that environmental inﬂu- shapen or missing ears and thumbs, ences play a role as well. For example, and neurological dysfunctions of the if genetic factors alone were involved, eye and facial muscles. Because scien- bryo as knowing when it happened. In monozygotic (identical) twins, who tists know which organs of the embryo the case of thalidomide-induced autism, share the same genes, should have a are developing at each stage of preg- the critical period is much earlier than 100 percent chance of sharing the same nancy, they can pinpoint the exact days many investigators would have guessed. diagnosis. Instead, when one twin has when a malformation can be induced: Very few neurons form as early as the autism, the second twin has only a 60 the thumb is affected as early as day 22 fourth week of gestation, and most are percent chance of being diagnosed with after conception, the ears from days 20 motor neurons of the cranial nerves, the the same disorder. That twin also has to 33, and the arms and legs from days ones that operate the muscles of the eyes, an 86 percent chance of having some of 25 to 35. What made the new study so ears, face, jaw, throat and tongue. The autism’s symptoms. These ﬁgures indi- exciting for me was Miller and Ström- cell bodies of these neurons are located cate that other factors must modify the land’s discovery that most of the in the brain stem, the region between the genetic predisposition to the disorder. thalidomide victims with autism had spinal cord and the rest of the brain. Be- anomalies in the external part of their cause these motor neurons develop at The Embryology of Autism ears but no malformations of the arms the same time as the external ears, one or legs. This pattern indicated that the might predict that the thalidomide vic- S everal environmental risk factors are already known. In utero expo- sure to rubella (German measles) or to subjects had been injured very early in gestation—20 to 24 days after concep- tion— before many women even know tims with autism would also suffer from dysfunctions of the cranial nerves. Miller and Strömland conﬁrmed this predic- birth defect–causing substances such as they are pregnant. tion— they found that all the subjects ethanol and valproic acid increases the For embryologists, nothing tells us so with autism had abnormalities of eye chances that autism will develop. Peo- much about what happened to an em- movement or facial expression, or both. ple with certain genetic diseases, such as phenylketonuria and tuberous scle- rosis, also have a greater chance of de- veloping autism. None of these factors, however, is present frequently enough to be responsible for many cases. Fur- thermore, most exposures to diseases or hazardous substances would be like- ly to affect both members of a pair of twins rather than just one. Some of the environmental inﬂuences must be more COURTESY OF SUSAN L. HYMAN subtle than those identiﬁed so far. Re- searchers do not know how the multi- ple factors combine to make some peo- ple display symptoms while allowing others to escape them. This variation makes the search for autism’s causes es- pecially difﬁcult. CHILD WITH AUTISM is normal in appearance, at least to the untrained eye. But he In their 1994 study Miller and Ström- has a few physical anomalies characteristic of the disorder. The corners of his mouth are land added another environmental con- low compared with the center of his upper lip, and the tops of his ears ﬂop over (left). tributor to autism: thalidomide expo- His ears are a bit lower than normal and have an almost square shape (right). The Early Origins of Autism Scientiﬁc American February 2000 59 Copyright 2000 Scientific American, Inc. The Spectrum of Autism Disorders A diagnosis of autism requires that the patient exhibit abnor- mal behaviors in three categories [see list at right] and have especially notable deﬁcits in the category of social interac- opment is followed by regression to severe disability, and Rett syndrome,a progressive neurological disorder that occurs only in females. tion. In addition, clinicians have identiﬁed several related disor- Although many scientists have long known that autism is an ders that share some of the behavioral features of autism but inherited disease, recent family studies by Peter Szatmari’s have different emphases or additional symptoms.For example, group at McMaster University in Ontario suggest that it is the Pervasive Development Disorder, Not Otherwise Speciﬁed spectrum of symptoms that runs in families rather than a sin- (PDD-NOS) denotes patients who miss fulﬁlling the autism cri- gle diagnosis. For example, a child with autism may have a teria in one of the three categories. As is true of autism, PDD- brother with Asperger syndrome,or a woman with autism may NOS includes patients with the whole range of IQs. Asperger have a nephew with PDD-NOS. These family studies strongly PHOTOGRAPHS BY JUSTINE PARSONS; SOURCE FOR ”DIAGNOSTIC CATEGORIES”: DIAGNOSTIC AND STATISTICAL MANUAL OF MENTAL DISORDERS (DSM-IV) syndrome is used to describe patients with normal IQs and no suggest that at least three of the diagnoses— autism,PDD-NOS evidence of language delay. Two much rarer diagnoses are and Asperger syndrome— arise from some of the same inherit- Childhood Disintegrative Disorder,in which normal early devel- ed factors. — P.M.R. The next logical question was, “Are ed backward more than 15 degrees— fects the cranial nerves but also has sec- the cases of autism after thalidomide are more common in children with ondary effects on later brain develop- exposure similar to cases of unknown autism than in typically developing ment. That is, the injury to the brain cause, or are they different?” Aside children, children with mental retarda- stem might somehow interfere with the from their behavioral symptoms, peo- tion or siblings of children with autism. proper development or wiring of other ple with autism have often been de- Dysfunctions of eye movement had brain regions, including those involved scribed not only as normal in appear- been associated with autism before the in higher-level functions such as speech, ance but as unusually attractive. They thalidomide study, and lack of facial resulting in the behavioral symptoms of are certainly normal in stature, with expression is one of the behaviors used autism. Or perhaps the ear malforma- normal-to-large heads. The few studies to diagnose the condition. tions and cranial nerve dysfunctions are that have tested nonbehavioral features only side effects of an injury that we of people with autism, however, have The Neurobiology of Autism don’t understand. Whatever the true concluded that there are indeed minor situation may be, the anomalies in pa- physical and neurological anomalies in many cases, and they are the same ones noted in thalidomide-induced autism. I s it possible that all the symptoms of autism arise from changes in the function of the cranial nerves? Probably tients with autism of unknown cause were much the same as the anomalies in the thalidomide victims with autism. For example, minor malformations of not. It is more likely that the nerve dys- The conclusion was clear: many cases the external ears— notably posterior ro- functions in people with autism reﬂect of autism, if not all, are initiated very tation, in which the top of the ear is tilt- an early brain injury that not only af- early in gestation. 60 Scientiﬁc American February 2000 The Early Origins of Autism Copyright 2000 Scientific American, Inc. disturbances— do sound like ones more tween a number of neuroanatomical likely to originate in the brain regions as- landmarks. I was surprised to discover sociated with basic functions. Further- that my hypothesis was absolutely Diagnostic Categories more, the most consistently observed ab- wrong. Although the side-to-side mea- normality in the brains of people with sures were indeed normal, the front-to- Impairment of Social Interaction: Failure autism is not a change in the forebrain back measures were astonishingly re- to use eye contact, facial expression or but a reduction in the number of neu- duced in the brain stem of the woman gestures to regulate social interaction; rons in the cerebellum, a large process- with autism. It was as though a band of failure to seek comfort; failure to develop ing center of the hindbrain that has long tissue had been cut out of the brain relationships with peers. been known to have critical functions in stem, and the two remaining pieces had the control of muscle movement. been knit back together with no seam Impairment of Communication: Failure to One reason for scientists’ confusion where the tissue was missing. use spoken language,without compensat- about the brain regions involved in For the second time in my life, I felt a ing by gesture; deﬁcit in initiating or sus- autism may be that our assumptions powerful shock of recognition. I heard a taining a conversation, despite adequate about where functions are controlled roaring in my ears, my vision dimmed, speech; aberrant language (for example, are shaky. For example, the laboratory and I felt as though my head might ex- repeating a question instead of replying). group led by Eric Courchesne of the plode. The shock was not generated by University of California at San Diego the unexpected result but by the realiza- Restricted and Repetitive Inter- has shown that parts of the cerebellum tion that I had seen this pattern of short- ests and Behaviors: Abnormally are activated during certain tasks re- ening before, in a paper that showed intense preoccupation with one quiring high-level cognitive processing. pictures of abnormal mouse brains. subject or activity; distress over Another difﬁculty is that change; insistence on routines or the symptoms of autism rituals with no purpose; repetitive movements,such as hand ﬂapping. are so complex. If simpler behavioral abnormalities Many cases of autism, could be shown to be di- agnostic of the disorder, if not all,are initiated very BEHAVIORAL THERAPY for chil- dren with autism can help them researchers might have a better chance of identify- early in gestation. lead happier lives as adults. Instruc- tors at the Eden Institute school ing their source in the ner- carefully evaluate the symptoms of vous system [see box on each child to draw up an appropri- next page]. When I retrieved the article from the ate intervention plan. They often In 1995 our research team had the stacks of papers on my ofﬁce ﬂoor, I engage the children in stimulating opportunity to follow up on the found that the correspondence between play activities (far left). The insti- thalidomide study by examining the the brain I had been studying and the tute also provides supervised hous- brain stem of a person with autism. The mouse brains described in the article ing for adults with autism (left). tissue samples came from the autopsy was even more striking than I had re- The 37-year-old man pictured here of a young woman who had suffered membered. Both cases exhibited short- used videocassette spools to make from autism of unknown cause; she ening of the brain stem, a smaller-than- the curtain behind his bed; his in- tense interest in these objects is a had died in the 1970s, but fortunately normal facial nucleus and the absence characteristic behavior of autism. the samples of her brain tissue had been of a superior olive. Additional features preserved. When we examined the of the mice were clearly related to other woman’s brain stem, we were struck by anomalies associated with autism: they the near absence of two structures: the had ear malformations and lacked one The region of the brain implicated facial nucleus, which controls the mus- of the brain structures controlling eye by the thalidomide study— the brain cles of facial expression, and the superi- movement. stem— is one that has rarely been con- or olive, which is a relay station for au- What had altered the brains of these sidered in studies of autism or in studies ditory information. Both structures mice? It was not exposure to thalidomide of other kinds of congenital brain dam- arise from the same segment of the em- or any of the other environmental factors age, for that matter. On a simplistic lev- bryo’s neural tube, the organ that devel- associated with autism but the elimina- el, neurobiologists associate the brain ops into the central nervous system. tion of the function of a gene. These were stem with the most basic functions: Counts of the facial neurons in the transgenic “knockout” mice, engineered breathing, eating, balance, motor coor- woman’s brain showed only about 400 to lack the expression of the gene known dination and so forth. Many of the be- cells, whereas counts of facial neurons as Hoxa1 so that researchers could study haviors disturbed in autism, such as lan- in a control brain showed 9,000. the gene’s role in early development. The guage, planning and interpretation of Overall, the woman’s brain was nor- obvious question was, “Could this be social cues, are believed to be controlled mal in size; in fact, it was slightly heav- one of the genes involved in autism?” by higher-level regions of the brain, such ier than the average brain. I hypothe- The literature supported the idea that as the cerebral cortex and the hip- sized that the brain stem was lacking Hoxa1 was an excellent candidate for pocampus in the forebrain. only the speciﬁc neurons already identi- autism research. The studies of knock- Yet some symptoms common in ﬁed— those in the facial nucleus and the out mice showed that Hoxa1 plays a autism— lack of facial expression, hyper- superior olive— and to test that idea I central role in development of the brain sensitivity to touch and sound, and sleep decided to measure the distances be- stem. Groups in Salt Lake City and The Early Origins of Autism Scientiﬁc American February 2000 61 Copyright 2000 Scientific American, Inc. A Simpler Symptom of Autism S cientists at York University and the Hospital for Sick Children in Toronto have recently identiﬁed an autism-related behav- ior that is much simpler than the array of behaviors that have tradi- watched lights ﬂashing on video screens [see illustration below]. The children ranged in age from four to seven.In the ﬁrst test,each child was placed in front of a three-screen panel, and a ﬂashing tionally been used to diagnose the condition.Susan Bryson and her light appeared on the middle screen. This stimulus prompted all doctoral student Reginald Landry have found that children with the children to focus their eyes on the ﬂashes (a).Then the middle autism respond abnormally to a task involving their reactions to vi- screen went blank, and a ﬂashing light appeared on the far-right or sual stimuli.Because this mental activity is probably mediated by a far-left screen of the panel. Both groups of children shifted their primitive part of the brain— most likely the brain stem or the cere- eyes to that screen (b). In the second test, however, the lights on bellum, or both— the discovery has important implications for the the middle screen kept ﬂashing while the lights appeared on the neurobiology of autism. Bryson and Landry’s work could also help other screen.The children without autism shifted their eyes to fo- clinicians develop a simpler way to test children for the disorder. cus on the new stimulus (c),but the children with autism remained In their study Bryson and Landry observed the reactions of two “stuck”on the ﬁrst stimulus and failed to turn their eyes to the new groups of children, those with autism and those without it, as they one (d).The two tests were repeated many times for each child. a b c NORMAL CHILDREN NORMAL CHILDREN NORMAL AND CHILDREN AND CHILDREN CHILDREN WITH AUTISM WITH AUTISM DANIELS & DANIELS London had studied different knockout genes do, but most changes are likely to or do nothing at all. Deviations from strains with similar results. They found be fatal, so they are rarely passed on to the normal sequence in any part of a that the gene is active in the brain stem subsequent generations. Although many gene can affect its performance, but the when the ﬁrst neurons are forming— the other genes appear in several forms— for vast majority of disease-causing varia- same period that Miller and Strömland example, the genes that encode eye col- tions are in the protein-coding regions. had identiﬁed as the time when thalido- or or blood type—highly conserved genes Thus, we began the search for variant mide caused autism. Hoxa1 produces a are not commonly found in multiple alleles by focusing on the exons of type of protein called a transcription versions (also known as polymorphic HOXA1. Using blood samples from factor, which modulates the activity of alleles, or allelic variants). The fact that people with autism and from subjects in other genes. What is more, Hoxa1 is no one had ever discovered a variant of a control group, we extracted the DNA not active in any tissue after early em- Hoxa1 in any mammalian species sug- and looked for deviations from the nor- bryogenesis. If a gene is active through- gested that my colleagues and I might mal sequence of nucleotides. out life, as many are, altered function of have trouble ﬁnding one in cases of The good news is that we have identi- that gene usually leads to problems that autism. On the other hand, it seemed ﬁed two variant alleles of HOXA1. One increase with age. A gene active only likely that if a variant allele could be has a minor deviation in the sequence of during development is a better candi- found, it might well be one of the trig- one of the gene’s exons, meaning that date to explain a congenital disability gers for the development of the disorder. the protein encoded by the variant gene like autism, which seems to be stable af- is slightly different from the protein en- ter childhood. Zeroing in on HOXA1 coded by the normal gene. We have stud- Hoxa1 is what geneticists call a “high- ied this newly discovered allele in detail, ly conserved” gene, meaning that the se- quence of nucleotides that make up its DNA has changed little over the course T he human version of the gene, labeled as HOXA1, resides on chromosome 7 and is relatively small. It measuring its prevalence among various groups of people to determine if it plays a role in causing autism. (The other of evolution. We assume that this is a contains just two protein-coding re- variant allele is more difﬁcult to investi- characteristic of genes that are critical to gions, or exons, along with regions that gate because it involves a change in the survival: they suffer mutations as other regulate the level of protein production physical structure of the gene’s DNA.) 62 Scientiﬁc American February 2000 The Early Origins of Autism Copyright 2000 Scientific American, Inc. spectrum of autism disorders. Further- substances that women need to avoid more, the allele that we have studied in during early pregnancy. What is more, Bryson and Landry found that children with detail is variably expressed— its pres- by examining the development of these other kinds of brain damage are perfectly ence does not guarantee that autism will genetically engineered mice, we could normal in their ability to disengage from one arise. Preliminary data indicate that the learn more about the brain damage that stimulus and focus on another. Children with variant allele occurs in about 20 percent underlies autism. If researchers can de- autism, however, repeatedly fail to disengage of the people who do not have autism termine exactly what is wrong with the from the ﬁrst stimulus, even if they are highly and in about 40 percent of those who brains of people with autism, they may intelligent.Researchers suspect that this abili- do. The allele approximately doubles be able to suggest drug therapies or oth- ty is a low-level brain function because it typi- the risk of developing the condition. But er treatments that could ameliorate the cally appears in infants—as early as three to in about 60 percent of people with effects of the damage. four months after birth—and in children with autism, the allele is not present, mean- Devising a genetic test for autism— low IQs. Animals also orient themselves to- ing that other genetic factors must be similar to the current tests for cystic ﬁ- ward new stimuli, so scientists could conceiv- contributing to the disorder. brosis, sickle cell anemia and other dis- ably use a similar test in animal studies to ver- To pin down those factors, we must eases— would be a much more difﬁcult ify whether genetic manipulations or toxico- continue searching for other variants in task. Because so many genes appear to logic exposures have produced this symptom HOXA1, because most genetic disor- be involved in the disorder, one cannot of autism. —P.M.R. ders result from many different deviant accurately predict the odds of having a alleles of the same gene. Variations in child with autism by simply testing for d other genes involved in early develop- one or two variant alleles in the parents. ment may also predispose their carriers Tests might be developed, however, for to autism. We have already discovered a the siblings of people with autism, who variant allele of HOXB1, a gene on often fear that their own children will chromosome 17 that is derived from the inherit the disorder. Clinicians could same ancestral source as HOXA1 and look for a set of well-established genetic has similar functions in the develop- risk factors in both the family member ment of the brain stem, but its effect in with autism and the unaffected sibling. autism appears to be minor. Other in- If the person with autism has several vestigators are scrutinizing candidate re- high-risk alleles, whereas the sibling CHILDREN gions on chromosome 15 and on anoth- does not, the sibling would at least be WITH AUTISM er part of chromosome 7. Although re- reassured that his or her offspring searchers are focusing on alleles that would not be subject to the known risks increase the risk of autism, other alleles within his or her family. may decrease the risk. These could help Nothing will make the search for explain the variable expression of the autism’s causes simple. But every risk fac- spectrum of autism-related disorders. tor that we are able to identify takes Even a minimal understanding of the away some of the mystery. More impor- We found that the rate of the variant al- genetic basis of autism would be of tant, new data spawn new hypotheses. lele among people with autism was sig- great value. For example, researchers Just as the thalidomide results drew at- niﬁcantly higher than the rate among could transfer the alleles associated with tention to the brain stem and to the their family members who do not have autism from humans to mice, engineer- HOXA1 gene, new data from develop- the disorder and the rate among unre- ing them to be genetically susceptible to mental genetics, behavioral studies, brain lated individuals without the disorder. the disorder. By exposing these mice to imaging and many other sources can be The differences were much greater than substances suspected of increasing the expected to produce more welcome would be expected by chance. risk of autism, we would be able to shocks of recognition for investigators of The bad news is that, just as the fami- study the interaction of environmental autism. In time, their work may help al- ly studies had predicted, HOXA1 is factors with genetic background and leviate the terrible suffering caused by only one of many genes involved in the perhaps compile an expanded list of the disorder. SA The Author Further Information PATRICIA M. RODIER is professor of obstetrics Autism in Thalidomide Embryopathy: A Population Study. K. Strömland, V. and gynecology at the University of Rochester. She Nordin, M. Miller, B. Åkerström and C. Gillberg in Developmental Medicine and has studied injuries to the developing nervous sys- Child Neurology, Vol. 36, No. 4, pages 351–356; April 1994. tem since she was a postdoctoral fellow in embryol- Embryological Origin for Autism: Developmental Anomalies of the Cra- ogy at the University of Virginia, but she began to nial Nerve Motor Nuclei. P. M. Rodier, J. L. Ingram, B. Tisdale, S. Nelson and investigate autism only after hearing the results of J. Romano in Journal of Comparative Neurology, Vol. 370, No. 2, pages 247–261; the thalidomide study. Rodier has assembled a June 24, 1996. group of scientists from many disciplines at six in- Thinking in Pictures: and Other Reports from My Life with Autism. Temple stitutions to study the genetic and environmental Grandin. Vintage Books, 1996. causes of the disorder and says that working with More information on autism is available at the Web page of the National Alliance experts from other ﬁelds is rejuvenating. for Autism Research at www.naar.org The Early Origins of Autism Scientiﬁc American February 2000 63 Copyright 2000 Scientific American, Inc.
Pages to are hidden for
"autism"Please download to view full document