CHAPTER 19

In 1943 Leo Kanner eloquently described 11 children with innate “autistic disturbances of affective contact” characterized by profound lack of reciprocal social engagement and interaction; disturbances in communication ranging from mutism to echolalia, pronoun reversal, and literalness; and unusual responses to the environment, including insistence on sameness that was not explained by general cognitive impairment.1 More than 7 decades later, diagnostic criteria and assessment instruments remain remarkably influenced by Kanner’s original description.

Definition and Classification: DSM-5

In the American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), published in 2013, the DSM-IV umbrella term pervasive developmental disorders and diagnostic classifications autistic disorder, Asperger’s disorder, childhood disintegrative disorder, Rett’s disorder, and pervasive developmental disorder not otherwise specified (PDD-NOS) were replaced by a single diagnostic term, autism spectrum disorder (ASD).² These changes were prompted by research demonstrating that the subclassifications had little scientific justification and were not used reliably, even by experts.³ The International Classification of Diseases (ICD) criteria, the 11th revision of which are scheduled to be published in 2018, are expected to also include these changes and to be very similar to the DSM-5 criteria.⁴

The DSM-5 diagnostic criteria for ASD are presented in Box 19.1, and the severity level classification, which is new to the DSM, is described in Box 19.2.² ASD diagnosis requires meeting all 3 of the social communication and interaction criteria and at least 2 of 4 restricted and repetitive behavior criteria. The symptoms must cause clinically significant impairment in current functioning and must have been present in the early developmental period, although they may not become fully manifest until social demands exceed limited capacities, or they may be masked by learned strategies in later life. To make a diagnosis of ASD in an individual with intellectual disability (ID), social communication and interaction skills must be below what would be expected based on general developmental level.

Box 19.1. DSM-5 Diagnostic Criteria for Autism Spectrum Disorder^a (299.00; F84.0)

^a Reprinted with permission from the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, (Copyright © 2013). American Psychiatric Association. All Rights Reserved.

Social (pragmatic) communication disorder (SCD) was introduced as a diagnostic classification in DSM-5.2,4 The goal was to more accurately recognize individuals who have early-onset, persistent difficulty using verbal and nonverbal communication for social purposes, leading to significant functional limitations in effective communication, social participation, development and maintenance of social relationships, academic achievement, and/or occupational performance.² The diagnostic criteria for SCD are provided in Box 19.3.² By definition, the communication deficits in SCD are not explained by low general cognitive ability or low abilities in the domains of word structure and grammar, and they occur in the absence of the repetitive and restricted behaviors and interests that characterize ASD; in fact, ASD must be ruled out for SCD to be diagnosed. Diagnosis of SCD with other communication disorders is permitted. However, it is expected that diagnosis of SCD would seldom be made before the age of 4 to 5 years, when structural language abilities are often adequate to determine that deficits in pragmatic language are discrepant from structural language abilities.⁴

Box 19.3. DSM-5 Diagnostic Criteria for Social (Pragmatic) Communication Disorder (315.39; F80.89)

SCD is likely to apply to some individuals who would have been classified as having PDD-NOS by DSM-IV-TR criteria but do not meet the DSM-5 criteria for ASD due to lack of restricted repetitive behaviors.4,5 The DSM-5 SCD concept expands upon research on pragmatic language impairment by including nonverbal communication, and further evaluation of the validity of the diagnosis is needed.⁴ Preliminary evidence suggests that this diagnosis identifies individuals who have marginally subthreshold autistic traits, rather than being qualitatively distinct from ASD.⁵

Epidemiology

The Centers for Disease Control and Prevention (CDC) Autism and Developmental Disabilities Monitoring Network (ADDM) 2012 estimate of ASD prevalence in 8-year-old children was 1.46%, or 1 in 68.⁶ Analysis of a general population cohort of more than 3 million children using a commercial managed health care administrative database yielded a similar prevalence of 1.25% in the United States.⁷ Somewhat higher prevalence estimates have been described in survey studies, which are based on parent report of ASD diagnosis. For example, parents reported a current diagnosis of ASD in 2.0% of children aged 6 to 17 years in the 2011–2012 National Survey of Children’s Health,8 and the rate of parent-reported lifetime diagnosis of ASD was 2.24% of children aged 3 to 17 years in the 2014 National Health Interview Survey.⁹ The estimated global prevalence of ASD is lower: 0.76% in the Global Burden of Disease Study 2010,¹⁰ although recent estimates based on national registry data in Sweden, Denmark, and Iceland were greater than 1%,^11–13 and the highest recent prevalence estimate, 2.64% for 7- to 12-year-old children in 2005–2009, was from South Korea.¹⁴ Males are more commonly affected than females, by a ratio of 4:1 to 4.5:1 in most studies,⁶ although a 2017 meta-analysis of studies conducted since the introduction of DSM-IV and ICD-10 criteria concluded that the true ratio is closer to 3:1.¹⁵

Substantial increases in the estimated prevalence of ASD in the United States over the last 3 decades have been found in studies using special education and developmental services administrative data,^16–18 national family surveys,^8,9,19 and active public health surveillance conducted by the CDC ADDM Network.⁶ Similarly, studies using registry data in Denmark,²⁰ Norway,²¹ Iceland,¹² Sweden,^11,22 Israel,²³ and other countries have reported an increase in ASD prevalence. However, incidence data adequate for determining time trends are scarce, and there is controversy about whether the available prevalence data actually establish that there has been an increase in ASD.^10,24–26 For example, a 2015 systematic review of epidemiological studies with Bayesian meta-regression analysis found that after accounting for methodological variations, there was no clear evidence of a change in estimated worldwide prevalence of DSM-IV-TR– defined autism spectrum disorders (autistic disorder, Asperger disorder, and PDD-NOS) between 1990 and 2010.¹⁰

If there has been an increase in prevalence, it is not clear whether there has been a true secular increase in incidence due to etiologic factors in addition to the clear contribution of nonetiologic factors such as changes in identification and diagnostic practices, variation in study methods, and increased public awareness and availability of services.^{6,10,20,22,25} Some studies provide evidence of diagnostic recategorization toward ASD and away from other diagnoses such as ID and language disorders.^18,27–31

Etiology/Pathophysiology

ASD is etiologically heterogeneous and, perhaps unsurprisingly, a unifying pathophysiology has not been identified for either the disorder as a whole or its core behavioral components. Although research aimed at elucidating the biological basis of ASD and related neurodevelopmental disorders is in its early stages, much progress has been made in the areas of genetics, neuroimaging, neurophysiology, and neuropathology.

Genetics and Genomics

Although clinically and etiologically heterogeneous, ASD is heritable, as demonstrated by family studies showing increased risk of ASD in siblings of affected individuals and twin studies documenting substantially higher concordance rates for monozygotic twins than for dizygotic twins.32–36 For example, a large population-based family study involving more than 2 million Swedish children, including 14,516 diagnosed with ASD, yielded a heritability estimate of 0.83.³⁶ Heritability estimates range from 64% to 91%, according to a meta-analysis of twin studies involving 6,413 twin pairs.³⁵

The complex genetic architecture of ASD is currently thought to involve the following^37–42:

1.A large number of rare variants, each conferring significant risk

2.Polygenic risk conferred primarily by a large number of common variants, or polymorphisms, which individually confer very small risk but collectively contribute substantially to the phenotype and to heritability

3.Some combination of genotype-by-genotype interactions (epistasis), genotype-by-environment interactions, and epigenetic effects

Although common variants, epistasis, and environmental modification of genotype effects are thought to be important, these aspects of the genetics of ASD are still poorly understood.^39,42,43 Genome-wide association studies and candidate gene association studies have not yet implicated replicable risk loci, likely due to small effect sizes and lack of statistical power.^38,41,43 Epistasis is difficult to detect because of the need for very large sample sizes, and there is little evidence so far for environmental modification of genotype effects.³⁹

In contrast, advances such as the development of microarray and next-generation sequencing technologies have resulted in identification of large-effect variants that appear to be causally associated with ASD, including copy number variants (CNVs), which are microdeletions or microduplications ≥1,000 base pairs in size that alter the dosage of genes, short insertions and deletions (indels), and single-nucleotide variants (SNVs).^40–42,44 Pathogenic rare variants may arise de novo or be inherited as autosomal dominant, autosomal recessive, or X-linked mutations. Microarray and exome sequencing studies have established that although de novo and inherited rare variants of large effect size are collectively common, no individual pathogenic variant accounts for more than 1% of nonsyndromic cases of ASD.^37,45–48

It is important to note that no specific mutation has been identified that is unique to ASD. Chromosomal microarray analysis (CMA) and whole exome sequencing (WES) studies in ASD cohorts and other populations have made it clear that there is substantial genetic overlap between ASD and other neurodevelopmental disorders, including ID, epilepsy, and schizophrenia. These disorders all are associated with increased CNV and single gene loss-of-function mutation burden, and individual pathogenic variants are not specific to one categorical diagnostic phenotype—the same pathogenic CNVs and SNVs have been detected in individuals with ASD, ID, epilepsy, schizophrenia, and other clinical presentations.^49–55

In clinical practice, a genetic etiologic diagnosis may be suspected clinically and confirmed by targeted genetic testing, or, more commonly, it may be revealed by CMA or WES completed as a routine part of the evaluation of a patient with ASD in the absence of a clinically recognizable syndrome. Examples of some of the most common genetic syndromes and mutations, including SNVs and recurrent CNVs, that are causally associated with ASD are presented in Table 19.1.^{40,42,44,56–60} Proteins encoded by genes that have been implicated in ASD have a variety of biological functions, but a pattern of convergence on functional pathways involved in synaptic structure and function and transcription regulation/chromatin remodeling has emerged.^38,40,42,59

Increased maternal and paternal age are independently associated with ASD risk, and there is evidence for a combined parental age effect as well; ASD risk increases when the difference between the parental ages is 10 years or more.^61–63 At least part of the ASD risk associated with advancing paternal age is explained by an age-related increase in the rate of de novo loss-of-function mutations in the male germ cell; potential mechanisms mediating the effect of advancing maternal age are under investigation.64

^a In the case of NRXN1, a very large gene (1.1 Mb, 24 exons), large intragenic deletions may be detected by chromosomal microarray analysis.

Abbreviations: CNV, copy number variant; SNV, single-nucleotide variant.

Gene names: ADNP, activity-dependent neuroprotector homeobox; ANK2, ankyrin 2; ARID1B, AT rich interactive domain 1B (SWI1-like); ASH1L, ASH1 (absent, small, or homeotic)-like; CACNA1C, calcium channel, voltage-dependent, L-type, alpha-1c subunit; CHD2, chromodomain helicase DNA binding protein 2; CHD7, chromodomain helicase DNA binding protein 7; CHD8, chromodomain helicase DNA binding protein 8; DHCR7, 7-dehydrocholesterol reductase; DYRK1A, dual-specificity tyrosine-(Y)-phosphorylation regulated kinase 1A; FMR1, fragile X mental retardation 1; GRIN2B, glutamate receptor, ionotropic, N-methyl-D-aspartate 2B; HDAC8, histone deacetylase 8; KATNAL2, katanin p60 subunit A-like 2; MECP2, methyl-CpG-binding protein 2; NF1, neurofibromatosis 1; NIPBL, nipped-B-like; NRXN1, neurexin 1; POGZ, pogo transposable element with ZNF domain; PTEN, phosphatase and tensin homolog; RAD21, double-strand-break repair protein rad21 homolog; SCN2A, sodium channel, voltage-gated, type II, alpha subunit; SHANK3, SH3 and multiple ankyrin repeat domains 3; SMC1A, structural maintenance of chromosomes protein 1A; SMC3, structural maintenance of chromosomes protein 3; SUV420H1, suppressor of variegation 4–20 homolog 1; SYNGAP1, synaptic Ras GTPase activating protein 1; TBR1, T-box brain, 1; TSC1, tuberous sclerosis 1; TSC2, tuberous sclerosis 2; UBE3A, ubiquitin protein ligase E3A.

Environment

There is evidence that aspects of the prenatal environment such as exposure to maternal medications (eg, valproate), short interpregnancy interval, multiple gestation, maternal obesity, gestational diabetes, gestational bleeding, and infections (eg, rubella, cytomegalovirus, influenza) may be associated with increased risk of ASD.^65–67 The case for in utero selective serotonin reuptake inhibitor exposure as a causal risk factor independent of the maternal conditions (eg, depression and anxiety) for which they are prescribed has not yet been proven.⁶⁷ Perinatal factors such as prematurity, low birth weight, fetal growth restriction, intrapartum hypoxia, and neonatal encephalopathy also may be associated ASD risk.^67–69 However, more research is required to determine whether these are independent risk factors because genetic susceptibility may be associated with both obstetric suboptimality and neonatal encephalopathy.

Although toxic environmental chemicals have not been proven to cause ASD, associations between some compounds such as organophosphates and certain other pesticides, metals, and volatile organic compounds and ASD risk at the population level have been published.^66,67,70 Several US studies have suggested an association between ASD and prenatal exposure to traffic-related air pollutants,^70–72 but it is difficult to exclude confounding from sociodemographic or other factors; a Swedish twin study and a collaborative study of 4 European population-based cohorts found no such association.^73,74

Epidemiological studies have definitively refuted the idea that toxic effects of either the measles-mumps-rubella vaccine, mercury exposure via thimerosal-containing vaccines, or excessive burden on the immune system caused by a larger number of vaccines is responsible for the increased prevalence of ASD, and the weight of all of the available scientific evidence overwhelmingly favors rejection of the hypothesis that there is a causal association between immunizations and ASD.^70,75–80

Epigenetics

The potential role of epigenetic factors in the etiology of ASD is also an active area of investigation and speculation.^81,82 Epigenetic modifications, such as DNA cytosine methylation and posttranslational histone modification, produce heritable changes in gene expression that do not involve a change in the DNA sequence. Abnormal patterns of DNA methylation and chromatin structure can result in altered morphology, physiology, and behavior, and some genetic disorders associated with ASD involve dysregulation of epigenetic processes during development. Examples include syndromes caused by mutations in genes encoding epigenetic regulators (eg, MECP2 in Rett syndrome and CHD7 in CHARGE syndrome) and syndromes caused by mutations in genes that are sensitive to alterations in their epigenetic regulation (eg, 15q duplication, Angelman, Prader-Willi, and fragile X).81,83 Epigenetic mechanisms are responsible for genomic imprinting, the process whereby certain genes are epigenetically marked during gametogenesis in a parent of origin specific manner so that gene expression is restricted to only one of the parental alleles.^83,84 Imprinted genes tend to be particularly sensitive to changes in their epigenetic status, and alteration of the gene dosage often results in pathology.⁸⁴ One of the most common cytogenetic abnormalities in patients with ASD is maternally derived duplication of the imprinted domain on chromosome 15q11-q13.⁸⁵

It has been speculated that epigenetic alteration of gene expression by environmental factors may play a role in the etiology of ASD, but replicated examples of environmental epigenetic marks and evidence of causality are currently lacking.^82,83 A small number of candidate gene studies and genome-wide analyses of DNA methylation in a variety of brain regions and peripheral tissues have provided preliminary evidence of alteration of DNA methylation in association with ASD or ASD-related traits. However, methodological issues and the absence of replication studies limit the conclusions that can be drawn from the existing literature.^82,86,87

Because epigenetic modifications can be influenced by environmental factors, such as prenatal exposures and postnatal experience, they represent one interface between genes and environment. Although appealing because they provide a tangible link between genes and environment, changes to the epigenome are not the only mechanism by which environmental effects are mediated, and epigenetics should not be conflated with the broader category of environmental effects.^83,88 In fact, biochemical cause-and-effect should be demonstrated to establish epigenetics as the mechanism of a proposed gene by environment interaction.^83,88 Despite the limitations of the existing evidence, exploration of the role of epigenetic and other nongenetic modifications that alter gene activity without changing the DNA sequence is likely to be an increasingly active and potentially important area of etiologic research in ASD.

Neuropathology

No uniform neuropathology has been identified, which is not surprising given the etiologic and phenotypic heterogeneity of ASD. The systematic neuropathological studies of postmortem brain tissue from individuals with ASD have been summarized in Supplemental Table 1 of the 2015 review by Chen and colleagues,³⁸ available at http://www.annualreviews.org. Dysplasia, altered neurogenesis, and neuronal migration abnormalities have been described within the cerebral neocortex, limbic system structures including the hippocampus and amygdala, basal ganglia, thalamus, brainstem, and cerebellum.^38,89 The vast majority of abnormalities described originate prenatally.^89,90

Focal disruption of the normal laminar architecture, minicolumnar abnormalities, and variations in neuronal density are among the most common cerebral cortical abnormalities described.^89,90 Reduced neuronal number and/or size, increased cell-packing density, and decreased complexity of the neuropil have been found in the hippocampus and amygdala, although not in all individuals.³⁸ Microglial infiltration and activation and astrocytosis have been observed in cerebral cortex and cerebellum, and postmortem transcriptome studies have reported upregulation of genes enriched in activated microglia and astrocytes, implicating dysregulation of neuron-glia signaling and neuroinflammation.38,42 However, it is not yet clear whether these findings represent primary pathology or a reactive or secondary process.

Biomarkers

Biomarkers are objectively measured biological characteristics that influence or predict the incidence or outcome of disease. Proposed biomarkers of ASD include head circumference growth trajectory; structural and functional magnetic resonance imaging (MRI) markers; electroencephalographic characteristics; eye tracking markers; genetic and biochemical markers in blood, urine, or brain tissue; placental pathology; and maternal autoantibody profiles. However, no biomarker has yet been shown to be clinically valuable.^91–93 Despite the limitations of the existing data, the future of biomarker research in ASD is promising due to increasing availability of bioresources linked to phenotypic data, advances in laboratory and neuroimaging technology and data management, and increasingly collaborative discovery and validation efforts.⁹³

– Early Brain Overgrowth and Neuroimaging Markers

Cross-sectional and longitudinal studies suggest that as a group, children later diagnosed with ASD have average or below-average head circumferences at birth, with abnormal acceleration in brain growth during the first 1 to 2 years of life leading to significantly above-average head circumferences and MRI brain volumes in toddlers, followed by a plateau in brain growth resulting in brain volumes that are not significantly different from controls in adolescence and adulthood.^94,95 Although this raises the possibility that atypical patterns of brain growth might be used for early identification, rate of head growth did not predict which infants developed ASD in the first 3 years of life in a large, prospective study of high-risk infants.⁹⁶

True brain overgrowth, defined as abnormally large brain size and abnormally rapid rate of brain growth, occurs in a minority of children with ASD. A systematic literature review and meta-analysis of 27 studies concluded that the prevalence of macrocephaly, defined as head circumference above the 97th percentile, in individuals with ASD was 15.7%.⁹⁵ Some recent studies have suggested that when individuals with ASD are compared to locally recruited controls rather than commonly used historical growth reference norms such as those provided by the CDC, the data are equivocal with regard to the finding of early brain overgrowth.⁹⁷ Additional research is needed to clarify the relationship between early brain overgrowth and ASD, whether it distinguishes ASD from other neurodevelopmental disorders, and to what extent it is part of generalized somatic overgrowth.^97,98

Although there are many conflicting findings, structural MRI volumetric studies of infants and young children have most consistently demonstrated increases in total brain volume, cortical gray and white matter volume (particularly frontal, temporal, and cingulate cortex), extra-axial cerebral spinal fluid volume, and amygdala volume in groups of young children with ASD relative to controls.99–101 Longitudinal studies suggest that neuroimaging findings are dynamic, and most of these volumetric differences do not persist into adolescence or adulthood.^100,101 Mean cortical thickness is somewhat increased in childhood in ASD but declines more rapidly than the typical decline that occurs in adolescence, resulting in decreased mean cortical thickness in adulthood, particularly in the frontal lobes.^101,102 Diffusion tensor imaging studies suggest age-dependent abnormalities of white matter microstructure, organization, and integrity based on metrics such as fractional anisotropy and mean diffusivity.^99,101 Cross-sectional studies suggest that reduced fractional anisotropy and other measures of white matter microstructure persist into adolescence and adulthood, particularly in the corpus callosum, cingulum, and white matter tracts connecting aspects of the temporal lobe.^100,103

Functional MRI (fMRI) studies have shown that ASD is associated with differences in activation of various brain regions relative to controls on tasks assessing motor skills, visual processing, executive functions, language, and basic and complex social processing skills.^99,104 The authors of a meta-analysis and systematic review noted some common themes across task domains, including lack of preference for social stimuli, lack of modulation in response to task demands or intensity of stimuli, and decreased activation of areas of the prefrontal cortex generally recruited for executive functions (visual-spatial systems are favored instead).¹⁰⁴ Most studies have identified evidence of task-dependent and resting state functional hypoconnectivity across a wide variety of brain areas in association with ASD, but other studies provide evidence of short range hyperconnectivity, particularly between subcortical regions and sensory cortices.^104,105 The evidence for functional underconnectivity is strongest for long-range, frontal-posterior networks.^99,104,105

Machine learning strategies including multivariate pattern classification have demonstrated that individuals can be classified categorically as having ASD with statistically significant accuracy using structural MRI and fMRI, and there is some evidence that they can predict the severity of symptoms (a quantitative or dimensional classification), but these methods have not yet been validated for clinical use.^106,107

– Electrophysiological Markers and Eye Tracking

Electrophysiological techniques such as auditory brainstem response, electroencephalography (EEG), and event-related potentials allow precise temporal discrimination, and magnetoencephalography allows fine resolution in both time and space. Although the evidence has not yet established the utility of any marker in clinical practice, studies using these techniques have identified electrophysiological correlates of abnormalities in low-level and higher cognitive auditory and visual processing (including language processing and face processing), somatosensory response, multisensory integration, recognition memory, selective attention, attentional shifting, and neural connectivity in association with ASD.^108–111 Resting-state and task-related quantitative EEG measures such as spectral power, complexity, and coherence are continuous and relate to typical development, making them good candidates for biomarker status.109,110,112 EEG spectral coherence loading patterns, for example, have been shown to differentiate between children with ASD and typically developing children with a high degree of success, but their ability to discriminate ASD from other neurodevelopmental disorders and their clinical relevance remain to be investigated.¹¹³

ASD is characterized by early emerging impairments in social attention, and eye tracking technologies have been used to demonstrate decreased fixation on the eyes’ region of the face and increased fixation on the mouth and background elements, as well as preliminary evidence that infants later diagnosed with ASD exhibit a decline in eye fixation between the ages of 2 and 6 months.^114–117

– Biochemical, Gene Expression, and Other Tissue Markers

The biochemical markers most consistently reported to differentiate groups of individuals with ASD from controls include increased platelet serotonin level, decreased plasma melatonin and urine melatonin sulphate, indicators of increased oxidative stress or altered redox status, and markers of altered immune response such as irregular cytokine profiles and central nervous system microglial activation.^91,92,93,118 Unlike variations in DNA sequence, which are largely fixed across tissues throughout the life span, the amount of RNA transcribed from each gene is tissue specific and varies developmentally and in response to environmental changes. Studies of gene expression in ASD are in their infancy and the results have varied widely, but several have reported upregulation of genes involved in immune and inflammatory responses in blood and postmortem brain tissue.^119,120 Efforts to classify ASD risk based on gene expression profiling are preliminary, and it has been noted that the genes included in the classification panels derived by 2 different research groups do not overlap.^121,122

Increased frequency of trophoblast inclusions relative to controls has been documented retrospectively in the placentas of children with ASD and prospectively in the placentas of children at risk for ASD based on having an affected sibling.^123,124 However, trophoblast inclusions also have been identified in association with chromosome aneuploidy and placenta accreta, and further investigation is needed to determine the predictive value in regard to ASD and other neurodevelopmental outcomes.^93,125 The pathogenic role of circulating maternal antibodies directed to fetal brain tissue and the potential value of maternal antibody panels as biomarkers of ASD also are active areas of investigation, but clinical value has not been established.^92,126–130

Clinical Features

ASD is defined clinically by the core impairments, which contribute to the DSM-5 diagnostic criteria in Box 19.1. However, no single specific behavior or deficit is pathognomonic for ASD and, although not part of the diagnostic criteria, other features such as cognitive and motor deficits, maladaptive behaviors, and associated medical problems have a dramatic impact on adaptive functioning and are also important aspects of the clinical presentation.

ASD is characterized by early onset, clinically significant impairment in social communication and other reciprocal social behavior and restricted and repetitive patterns of behavior, interests, and activities (Box 19.1).² By definition, these symptoms interfere substantially with functioning and do not simply reflect coexisting ID. Clinical features of ASD, as elaborated in Boxes 19.4 and 19.5, vary with age, developmental level/ intellectual ability, and severity of the condition.

– Social Communication and Social Interaction

Pervasive and persistent impairment in reciprocal social behavior that is not accounted for by developmental level (or mental age) is a defining feature of ASD; deficits in the development and understanding of social and emotional reciprocity, nonverbal communicative behaviors, and interpersonal relationships are required for diagnosis² (Box 19.1).

As infants and toddlers, children with ASD may not smile responsively or adopt a posture conveying anticipation and readiness or desire to be picked up. Young children with ASD typically lack developmentally appropriate joint attention (JA), which is coordinating attention with another individual to an external focus, showing social engagement and awareness of the partner’s mutual interest.^131,132 JA behaviors can be receptive, (eg, responding by following the direction of eye gaze and gestures of others) or expressive (eg, initiating JA by using eye contact and gestures to direct the attention of others to an object or event of interest for the purpose of sharing their experience). Both are commonly delayed in children with ASD; however, initiation of JA for the purpose of sharing interest and enjoyment/positive affect rather than requesting is particularly atypical.¹³¹ JA ability is an important precursor and predictor of language and theory of mind, an important component of empathy.^131,133 Reciprocal social behaviors such as seeking to share enjoyment with others, recognizing and reacting to the mental states of others, and forming reciprocal friendships that go beyond classroom or parent-arranged interactions are limited, even in older children.¹³⁴ Even highly functioning adolescents and adults with ASD typically have difficulty navigating social interactions when it is necessary to make accurate inferences about other people’s beliefs, motivations, goals, and emotional states.^135,136

Wing and Gould¹³⁷ described 3 types of social impairment exhibited by individuals with ASD: aloof, passive, and active but odd. Individuals in the first group, whose social impairment is most easily recognized, are generally aloof and indifferent to others, especially peers, but may accept physical affection from familiar people and enjoy physical play. Those in the passive group do not socially interact spontaneously but passively accept and even appear to enjoy approaches from others. The active but odd group includes people who make active social approaches that are naïve, odd, inappropriate, and one-sided. Stability of the social interaction subgroup classification varies with level of cognitive ability. Children with higher levels of cognitive ability may move, for example, from the aloof group in early childhood to the passive or active but odd group in adolescence, whereas those with severe or profound ID are more likely to remain aloof and indifferent to others.

The severity of language impairment ranges from profound (eg, verbal auditory agnosia) to relatively mild (eg, pragmatic impairment), and the clinical features vary with age and developmental level (Box 19.4).^138–140 Communicative speech is often delayed or absent in young children with ASD, and comprehension is impaired.¹³⁹ Nonverbal communication is also impaired, so there is often little or no attempt to compensate by using facial expressions, gestures, or pantomime, especially in young children with ASD. Approximately 20% to 30% of individuals remain nonverbal or minimally verbal.^139,141

Although structural language skills are extremely variable and may be normal in some individuals, echolalia, jargon, neologisms, and other unusual semantic and syntactic error patterns are more frequent among children with ASD than typically developing children or children with ID.139,142 Even among individuals with higher verbal skills, the language profile that emerges in childhood and persists into adulthood is characterized by unevenness, including poor comprehension relative to expressive language, persistence of certain morphological errors, semantic processing anomalies despite normal performance on vocabulary tests, and idiosyncratic word usage despite relatively intact articulation and syntax.^139,142

Deficits in pragmatics, which is the ability to use language in context for social purposes, are universal in individuals with ASD.^139,140 Even the most able communicators with ASD typically exhibit significant atypicality in conversational discourse, including poor topic maintenance, difficulty with turn-taking, failure to respond adequately to questions or comments, inclusion of non-contextual or socially inappropriate utterances, and difficulty recognizing and repairing a misunderstanding or unclear referent.¹³⁹ Other common deficits involve understanding figurative and metaphorical language, making inferences, and resolving ambiguous language.^139,140

Children with ASD typically exhibit deficits in functional and symbolic play, including delayed onset, decreased frequency, reduced diversity, and qualitative atypicality.¹⁴³ Qualitatively, the play of children with ASD tends to be solitary, repetitive, stereotyped, and excessively focused on sensorimotor manipulation of objects rather than interactive, varied, flexible, imaginative, and creative (Box 19.4).

– Restricted, Repetitive Behavioral Repertoire

Restricted and repetitive behaviors (RRBs) include repetitive sensorimotor actions (eg, stereotyped motor mannerisms and repetitive manipulation of objects), insistence on sameness/behavioral rigidity (eg, resistance to change, compulsive adherence to nonfunctional rituals or routines), and circumscribed interests (eg, fascination/ preoccupation with one subject or activity that is abnormal in intensity or focus; Box 19.5).^144,145 It is important to recognize that repetitive behaviors, including stereotyped movements and behaviors related to desire for sameness, occur in normal development and in individuals with other developmental disabilities or psychiatric conditions, although most are more prevalent and/or severe in children with ASD.^146,147

In general, younger children are more likely to engage in repetitive sensorimotor actions, whereas older children are more likely to exhibit repetitive behaviors based on insistence on sameness/behavioral rigidity and circumscribed interests.^144,145,148 As a group, RRBs commonly decline in adolescence and adulthood, although there is considerable variability.¹³⁵ Stereotyped movements are more persistent in adolescents and adults with ASD who also have ID, whereas behaviors based on insistence on sameness and circumscribed interests tend to be more independent of age and IQ.^144,149 However, even in highly functioning individuals, unusual preoccupations and circumscribed interests often interfere with social interaction. Conversely, in addition to being sources of pleasure and motivation, special interests may provide themes around which social activities can be developed and, potentially, opportunities for employment.150

Atypical reactivity to sensory input and unusual interest in sensory aspects of the environment were incorporated into the DSM diagnostic criteria for ASD for the first time in DSM-5; symptoms in this category can fulfil 1 of the 4 restricted and repetitive behavior criteria, 2 of which are required for diagnosis (Box 19.1).² Sensory modulation symptoms have been defined as difficulties in regulating and organizing the type and intensity of behavioral responses to sensory input to match environmental demands and are categorized as follows: (a) over-responsivity, which is exaggerated, rapid onset, and/ or prolonged reactions to sensory stimulation; (b) under-responsivity, which is unawareness or slow response to sensory input; and (c) maladaptive sensation seeking, which describes craving of and interest in sensory experiences that are prolonged or intense. Examples include apparent indifference to pain or temperature (under-responsivity), distress and ear-covering in response to certain sounds (over-responsivity), and atypical fascination with visual inspection of lights or spinning objects (sensation-seeking), as well as others (Box 19.5). Unusual sensory responses are described more commonly in children with ASD than in typically developing children; the effect size of the difference is greatest for under-responsivity, followed by over-responsivity and then sensation seeking.¹⁵¹ Although they may diminish in later childhood and adolescence, atypical reactivity to and/or interest in environmental sensory input may persist through adulthood.^151,152

Autistic Regression

Population-based studies suggest that approximately 20% of children with ASD experience loss of previously established language skills and/or social interest and engagement skills at a mean age of 21 to 24 months.^153,154 Clinic-based samples and survey studies yield higher rates of reported regression (35%–40%).¹⁵³ However, even the more accurate population-based studies are plagued by problems with variable and sometimes vague case definition and reliance on parent-reported skill loss without verification by rigorous clinical information gathering, and the prevalence of true regression may be substantially lower.¹⁵⁵

Autistic regression may be gradual or sudden and may occur following typical development or be superimposed on preexisting atypical development.^153,156 Some studies suggest that regression is associated with increased likelihood of ID and higher rates of certain RRBs (eg, stereotyped speech, insistence on sameness, sensory interests and/or behaviors), but there is no definitive consensus.^154,156 Although there is evidence that epilepsy is more common in individuals with ASD who have a history of regression, this association appears to be driven by lower IQ.^157,158

Neuropsychological Features

Historically, ID was thought to be present in 70% to 75% of children with ASD, but many recent studies suggest that the prevalence of ID among individuals with ASD is substantially lower (32%–55%).^6,11,12,159 A systematic review and meta-analysis of male to female ratio in ASD included 24 epidemiological studies published between 1996 and 2015 that provided sufficient information to determine that the mean proportion of participants with IQ <70 was 48%.15

Individuals with ASD often display an unusual degree of unevenness in cognitive abilities, frequently including a significant discrepancy between verbal and nonverbal abilities, which can be in either direction (verbal<nonverbal or verbal>nonverbal), and/or isolated strengths that significantly exceed not only their own general level of ability but general population norms as well.¹⁶⁰ A strong correlation between uneven intellectual development and autistic-like symptoms, including those outside of the clinical range for ASD, has also been reported.¹⁶¹ Although a high degree of variability between domains and subtests is common, a prototypical profile of cognitive strengths and weaknesses does not exist.^160,162

IQ testing to identify verbal and nonverbal cognitive abilities and specific strengths and weaknesses is useful in educational planning for children with ASD, as it is for children with other disabilities. Although IQ is one of the most robust predictors of outcome for individuals with ASD, it is important to recognize that low IQ is more reliably predictive of a poorer outcome than high IQ is of a better outcome and that IQ, especially among higher-ability individuals, often significantly exceeds adaptive functioning in the community.¹⁶⁰

Performance profiles also vary on standardized measures of academic achievement. Skills requiring primarily rote, mechanical, or procedural abilities are typically intact, but skills requiring more abstract reasoning, conceptualization, or interpretive analysis are often deficient.^163,164 For example, a child with ASD may have very good phonological awareness and decoding abilities but poor reading comprehension. Because the types of skills emphasized in the early elementary grades play to their strengths in memorization and concrete tasks, such as decoding and numerical operations, young children with ASD may perform well academically. In later elementary and middle school grades, when comprehension, analytic interpretation, and abstract reasoning skills are emphasized, these children often struggle academically and require more supports.^164,165

Even high-functioning children with ASD often have weaknesses in graphomotor skills, attention, and processing speed that interfere substantially with academic performance.^164,166

Neuropsychological theories that help to explain the behavior and cognitive performance profile of individuals with ASD are described in Table 19.2.^167–171 Some neuropsychological theories suggest that cognitive impairments or biases, such as executive functioning and information-processing differences, underlie the behavioral manifestations of ASD throughout development.¹⁶⁷ In contrast, social motivation and social cognition theories emphasize early emerging, core social deficits that precipitate a developmental cascade of disrupted social and communicative development. Although no single theory adequately explains all aspects of ASD, recognition of the fundamental deficits in social motivation, perspective-taking, cognitive flexibility, and the ability to process complex information and form a coherent global picture is very helpful for understanding the behavior of children with ASD and explaining it to parents and teachers.

In addition to motor stereotypies, children with ASD often exhibit less distinctive motor abnormalities, including impairment in basic aspects of motor coordination (eg, postural control, gait, attainment of gross and fine motor skills), performance of skilled motor gestures (dyspraxia), and motor learning activities.172–174 A core feature of motor dysfunction in ASD seems to be excessive reliance on proprioceptive feedback and weak integration of visual feedback.¹⁷⁴ Clinically, parents may be concerned about delayed attainment of motor milestones, clumsiness, toe-walking, or difficulty with handwriting. Neurological examination may reveal hypotonia, decreased postural stability, poor motor imitation abilities, and other subtle, neurological signs, such as slow speed, dysrhythmia, and increased overflow movements on timed movements of the hands and feet and stressed gait maneuvers. Tic disorders, including Tourette disorder, also occur with increased frequency in children with ASD and in their siblings.175–177

– Epilepsy

ASD is associated with increased risk of epilepsy, with bimodal peaks in age of onset in early childhood and adolescence.^158,178 All seizure types occur in individuals with ASD, but complex partial seizures, with or without secondary generalization, are reported to be more common than primary generalized seizures.¹⁷⁸ The reported prevalence of epilepsy in ASD ranges widely, with the largest studies reporting ranges of 12.5% to 37% in clinic-based samples^157,179,180 and 6.6% to 22.5% in population-based samples.^181–183 In a large cross-sectional study combining 3 clinical samples and 1 population-based sample (n=5,815), the best estimate of the cumulative prevalence of epilepsy in ASD through age 17 years was 26%.¹⁵⁷

The most robust risk factors for epilepsy in children and adolescents with ASD are low IQ and older age.^158,178,184 Although the prevalence is highest in individuals with severe ID and those with identified genetic etiologies, even the subgroup with idiopathic, high-functioning autism has a rate of epilepsy that is substantially higher than the rate in the general population.¹⁷⁸ The rate of medically refractory epilepsy may be increased in this population as well.¹⁸⁵ In addition, many children with ASD have interictal epileptiform activity on EEG, especially during sleep, even though they have not had definite clinical seizures.^178,186 The clinical significance of the increased rate of interictal EEG abnormalities is unclear, however, and very little information has been published to support the idea that epilepsy or epileptiform EEG abnormalities cause ASD or have treatment implications.^{158,178,186,187}

– Neurobehavioral Symptoms/Coexisting Psychiatric Disorders

Neurobehavioral features that are commonly associated with ASD and cause significant impairment and distress include irritability, anger outbursts, aggression, self-injurious behavior, property destruction, elopement or wandering, sleep disturbance, mood instability, anxiety, hyperactivity, impulsivity, inattention, and other disruptive behaviors or manifestations of emotional dysregulation. Although these problem behaviors are not core features of ASD, they commonly interfere with functioning in school, at home, and in the community, and they contribute substantially to the burden on families.^188–194 In some cases, the diagnosis of a coexisting psychiatric disorder can be reasonably made, although modifications of the diagnostic criteria may be necessary when there is significant intellectual impairment. Psychiatric diagnoses reported to co-occur with ASD in children and adults include attention-deficit/hyperactivity disorder (ADHD) (28%–44%), anxiety (42%–56%), depression (12%–70%), psychotic disorders (12%–17%), tic disorders (14%–38%), and oppositional defiant disorder (16%–28%).¹⁹⁴

Sleep problems are reported in 40% to 80% of children with ASD (versus 25%–40% of typical controls), and they are common at all ages and levels of intellectual ability.¹⁹⁵ Sleep disturbances are burdensome to children with ASD and their family members due to their prevalence, chronicity, and impact on emotional and behavioral problems. Insufficient sleep is associated with exacerbation of core ASD symptoms (RRBs and social communication and interaction difficulties) as well as tantrums, self-injurious behavior, aggression, and other disruptive behaviors.195 Children with ASD experience a variety of sleep problems, including bedtime resistance, prolonged sleep-onset latency, decreased sleep efficiency, increased waking after sleep onset, reduced total sleep time, and daytime sleepiness.^195,196 Behavioral factors, including inadequate sleep hygiene, maladaptive sleep-onset associations (eg, being held until asleep), and problems with limit-setting, are common, but ASD-associated melatonin or GABA abnormalities, coexisting neurological or psychiatric disorders (eg, epilepsy, ADHD, and anxiety), other medical problems (eg, gastroesophageal reflux), and adverse effects of medications may also contribute to disordered sleep.^195,196 Occasionally, other sleep disorders, such as obstructive sleep apnea, parasomnias, and periodic limb movements of sleep, are identified.

– Gastrointestinal and Feeding Problems

A meta-analysis of 15 methodologically suitable studies involving 2,215 children indicated that children with ASD experience significantly more general gastrointestinal (GI) symptoms, diarrhea, constipation, and abdominal pain than comparison groups.¹⁹⁷ The available data were not sufficient to determine whether symptoms often suggestive of organic GI pathology, such as gastroesophageal reflux, gastroenteritis, inflammatory bowel disease, or food allergies, occur more frequently in children with ASD than in controls, and there is no evidence suggesting a unique GI pathophysiology in ASD.^197,198 The contribution of behavioral factors such as delayed bowel training, rigidity about toileting routine, sensorimotor issues, and food selectivity to the observed higher prevalence of symptoms in children with ASD is unclear but plausibly substantial.¹⁹⁷

These symptoms may impact behavior substantially. For example, in a well-characterized sample of children who participated in 2 Research Units on Pediatric Psychopharmacology Autism Network medication trials, the subgroup of children with GI problems (23%) had greater symptom severity on measures of irritability, anxiety, and social withdrawal than those without GI problems, and their irritability was less likely to respond to treatment with risperidone.¹⁹⁹ GI problems may not be obvious in some children with ASD due to communication deficits and/or under-responsivity to sensory input but should be considered in cases of unexplained irritability/agitation, food refusal, self-injury, aggression, sleep disturbance, or other problem behavior that is new or represents a significant exacerbation of baseline status.^197,198

Feeding problems, including food selectivity (often based on texture, color, or temperature), rituals around food presentation, and compulsive eating of certain foods, are common, as are behaviors associated with food refusal (eg, holding food in the mouth, volitional gagging, emesis), rumination, and pica.^200–202

Identification

Surveillance is the ongoing process of actively identifying children who may be at risk for developmental disorders, and screening is the use of standardized tools at specific intervals to support and refine risk. The American Academy of Pediatrics (AAP) has established guidelines for developmental surveillance and screening (see Chapter 9, Developmental and Behavioral Surveillance and Screening Within the Medical Home).203 Surveillance is a continuous process that should occur at every preventive care visit during childhood and include the following components: eliciting and attending to the parents’ concerns, maintaining a developmental history, making accurate and informed observations of the child, identifying the presence of risk factors and protective factors, and documenting the process and findings.²⁰³

The AAP recommends general developmental screening using a broadband measure for all children at ages 9, 18, and 24 to 30 months or at any time concerns are raised by a caregiver or primary pediatric health care professional.²⁰³ In addition, specific screening for ASD is recommended at the 18- and 24-month visits and whenever parents’ concerns are raised or other risk factors are identified through general developmental surveillance and screening.¹³⁸ The 2007 recommendation for universal screening for ASD in toddlers was based on the recognition that symptoms can be observed in very young children and mounting evidence that early identification leads to early intervention and improved outcomes.^138,204 The US Preventive Services Task Force (USPSTF) completed a literature review and concluded that the current evidence is insufficient to assess the balance of benefits and harms of screening for ASD in young children (ages 18–30 months) for whom no concerns of ASD have been raised.²⁰⁵ This is not a recommendation against universal screening in this age group, but it is also not a recommendation for the practice. The AAP has affirmed its recommendation for universal screening for ASD at the ages of 18 and 24 months, and several strong commentaries have supported the approach and criticized aspects of the USPSTF interpretation of the data.^206–209

A variety of general developmental (broadband), language-specific, and autism-specific screening tools have been shown to identify children of different ages who are at risk for ASD. Some examples of parent-completed screening measures that may be appropriate for use in primary care are described in Table 19.3. The Infant-Toddler Checklist, which is part of the Communication and Symbolic Behavior Scales Developmental Profile, is a screen for communication deficits in 6- to 24-month-olds that has been shown to identify ASD, language delay, and other developmental delay at age 12 months.²¹⁰ The Modified Checklist for Autism in Toddlers, Revised, with Follow-Up (M-CHAT-R/F) is the most commonly used tool for screening toddlers for ASD. It consists of a 20-question caregiver form as the first stage and, when the score is elevated, a brief, structured interview administered by a health care professional. Among children who fail the M-CHAT-R/F in community primary care settings, 50% are ultimately diagnosed with

ASD and 98% have an actionable developmental concern.207,211 If a screening test is positive (“failed”), suggesting risk of ASD, the child should be simultaneously referred to the early intervention program or school evaluation team for services and to a specialist or team for comprehensive diagnostic evaluation.¹³⁸ Audiological evaluation is often warranted, particularly in a young child with delayed receptive and expressive language development.

Diagnostic Evaluation

The diagnostic evaluation will vary depending on the availability of local resources and clinician preferences. An interdisciplinary team specializing in ASD and other developmental disabilities may be ideal, but often these are not available outside of major medical centers, and waiting lists may be very long because of high demand. Many communities will have at least one expert who is capable of interpreting and integrating information from various disciplines and comfortable with making the diagnosis of ASD even if there is not an interdisciplinary team that meets in person. This may be a developmental-behavioral or neurodevelopmental pediatrician, child and adolescent psychiatrist, pediatric neurologist, or pediatric psychologist, for example.

The clinical standard of care for ASD diagnosis is expert clinician evaluation and application of the current DSM or ICD criteria (DSM-5, ICD-10; ICD-11 [2018]). There is considerable overlap in symptoms among developmental and behavioral diagnoses, especially in very young children and those with ID or language impairment, and it is important for the diagnostician to consider the alternate explanations for observed symptoms. Standardized rating scales, such as the assessment instruments presented in Table 19.3, which are completed by parents, teachers, or other intervention professionals, may be used to gather and quantify information. In the hands of trained clinicians, diagnostic instruments such as those presented in Table 19.4 help to operationalize the diagnostic criteria and inform clinical judgment.^138,212,213

The diagnostic evaluation should ideally incorporate the following elements^{138,212–214}:

1.Caregiver interview. This includes a thorough developmental and behavioral history (including milestone attainment and report of current abilities), medical history, social history, and family history (at least 3 generations). Structured interview tools (eg, Autism Diagnostic Interview-Revised, Diagnostic Interview for Social and Communication Disorders) are informative but are often impractical in clinical settings because of time constraints. Standardized rating scales completed by parents/caregivers and teachers or other professionals, such as the assessment instruments described in Table 19.3, are commonly used to quantify ASD symptoms and compare them to normative populations.

2.Review of pertinent existing data. The pertinent medical and educational records should be reviewed, especially any available standardized testing completed by psychologists, speech-language pathologists, occupational therapists, early intervention evaluators, or others. Although not necessary for categorical diagnosis, it is helpful to note the results of any previous etiologic investigations and other pertinent tests that have been completed (eg, genetic testing, audiometry, central nervous system imaging).

3.Direct clinical assessment. This includes standardized developmental and psychological testing (appropriate for age and level of ability) and neurobehavioral observation/ elicitation. Receptive and expressive language, nonverbal intellectual ability, and functional adaptive behavior should be measured. Administration of standardized ASD-specific direct observation/elicitation instruments (eg, Autism Diagnostic Observation Schedule, Childhood Autism Rating Scale) is desirable. When the evaluating diagnostician is a physician or advanced care practitioner, physical examination targeting the neurological examination and evaluation for dysmorphic features and other clues to etiology should be included.

The goal should not be to simply decide whether the child has ASD or not but to determine the child’s developmental diagnoses, whatever they may be, and the diagnostic summary should include domain-specific information regarding the child’s level of functioning/severity of impairment.212,213 Children who have been diagnosed with ASD also require periodic formal reevaluation because their strengths, weaknesses, and educational needs often change even if the ASD diagnosis does not.²¹²

Management

The role of primary pediatric health care professionals does not end with diagnosis. Comprehensive management of patients after the diagnosis of ASD includes conducting an etiologic investigation; providing genetic counseling; promoting general health and well-being through effective longitudinal health care within a medical home; guiding families to effective educational, behavioral, and medical interventions; and alleviating family distress by providing support, education, and access to resources.^214,215 Primary pediatric health care professionals may take part in any or all of these duties in addition to making referrals and coordinating subspecialty care and other services.

Etiologic Evaluation

Currently, there is no single consensus approach to etiologic evaluation for children newly diagnosed with ASD. In recent years the availability of new methods of genetic testing has progressed so rapidly that the publication of state-of-the-art guidelines by professional organizations lags behind. Therefore, the recommendations in this chapter reflect the authors’ assessment of the current literature rather than strict adherence to guidelines published previously by various professional organizations.

– Genetic Testing

Identifying a genetic etiologic diagnosis may allow clinicians to provide more accurate prognostication and recurrence risk counseling, identify and treat or prevent medical comorbidities, guide patients and families to condition-specific resources and supports, and, in some cases, refine treatment options.^216–224 For these reasons, there is consensus agreement that a genetic evaluation should be offered to every person with ASD.^225–228 However, surveys suggest that that only about 33% of children with ASD in the United States have undergone any genetic testing.^229–233

Chromosomal Microarray Analysis and Next-Generation Sequencing

CMA reveals a definitively pathogenic CNV in 5% to 14% (median 9%) of individuals with ASD in clinical cohorts.46,234–242 These numbers do not reflect variants of uncertain significance, many of which are likely to subsequently be determined to be pathogenic; when variants of uncertain significance are included, the reported CNV detection rate in these studies is 17% to 24%. Large, clinical whole exome sequencing studies have consistently identified a molecular diagnosis in 26% to 29% of individuals for whom neurodevelopmental disorders were the primary indication for testing.^243–245 Studies restricted to clinically ascertained samples of patients with ASD and analyses of ASD subgroups within the large, laboratory-based samples have reported WES yields of 8% to 20%.^46,48,245 There is emerging evidence that WES is not only diagnostically useful but also cost-efficient in the etiologic evaluation of children with neurodevelopmental disorders, including ASD.^246–249

Fragile X Testing

The combined yield of fragile X testing in 6 recent large studies was 9 full mutations in 1,984 individuals tested (0.45%), and 22% of those found to have fragile X syndrome were female.^{237,238,241,250–252} Because males and females with fragile X syndrome are identified fairly frequently in ASD cohorts, testing is relatively inexpensive, and the condition has important genetic counseling implications due to always being inherited, it is reasonable to routinely test both males and females with ASD for fragile X syndrome, at least until more data become available to clarify the issue. DNA testing for fragile X syndrome is ordered as a separate test because pathogenic repeat expansions such as the CGG trinucleotide repeat expansion that is responsible for fragile X syndrome are not detected on CMA or WES.

Approach to the Genetic Etiologic Evaluation

When the diagnosis of ASD is made by a physician or a diagnostic team that includes a physician, the physician may also take responsibility for the etiologic evaluation. If the diagnosis is made by a psychologist or by a physician who does not provide this service, referral to a medical geneticist is appropriate. In some cases, the primary pediatric health care professional may take on this role. In any case, clinicians should only order tests that they are capable of explaining (ie, providing pretest counseling and obtaining informed consent) and interpreting, although full interpretation of abnormal or equivocal results may involve referral to an appropriate subspecialist (eg, a medical geneticist). Pretest counseling involves describing the nature of the test, including potential outcomes, and addressing points such as possible impact on other family members. Most pediatric subspecialists who diagnose ASD, and many general pediatricians, are comfortable with explaining, ordering, and interpreting fragile X testing and CMA. WES is typically ordered by a medical geneticist or, in some cases, another pediatric subspecialist with the appropriate expertise, and a genetic counselor is likely to be involved in the vast majority of cases.

An etiologic search begins with a careful medical and family history and physical examination. Important aspects of the physical examination include assessment for abnormal growth (including head circumference), major and minor congenital anomalies, evidence of visceral storage, skin manifestations of neurocutaneous disorders or mosaicism, and neurological abnormalities.228 When a specific syndrome or metabolic disorder is suspected, the next step is to proceed with the appropriate targeted testing or refer to medical genetics. For example, a child presenting with ASD and microcephaly, ptosis, anteverted nares, and cutaneous 2–3 toe syndactyly should have a 7-dehydrocholesterol level measured to test for Smith-Lemli-Opitz syndrome. Similarly, marked macrocephaly and pigmented macules on the glans penis in a child diagnosed with ASD would warrant PTEN sequencing and deletion/duplication analysis. If there is no strong suspicion of a particular diagnosis or testing for suspected diagnoses has been completed and is negative, CMA and fragile X DNA analysis are recommended. If these are unrevealing, clinical WES is recommended. These steps are outlined in Box 19.6.

Box 19.6. Genetic/Genomic Etiologic Investigation in Patients With Autism Spectrum Disorder

Stay updated, free articles. Join our Telegram channel

Join