Child Neurology
7th Edition

Developmental disorders have a wide variety of causes, and few have a specific, identifiable genetic or neurobiologic etiology. Different etiologies converge on identical
behavioral phenotypes. The diagnostic yield is much greater if there is associated microcephaly, antenatal toxin exposure, or focal findings (19). In their absence, a thorough search for diagnostic clues includes family history with a three-generation pedigree and information from the prenatal, perinatal, and early postnatal periods. The development of motor skills; comprehension of speech; emergence of words and formed, intelligible sentences; and quality of social play skills are often relevant. The physical examination includes measuring head circumference as well as body size and growth; assessment of vision, hearing, and ocular funduscopy; searching for facial, skeletal, somatic, and dermatoglyphic anomalies and asymmetries; visceromegaly; neurocutaneous lesions and depigmentations; and other signs of chronic illness. Hand preference and pencil grip are noted as well as any abnormal postures and involuntary movements. Clinical clues in the evaluation of developmental disorders are presented in Table 18.2.

The mental status examination assesses the child’s orientation, relatedness to, and interest in the examiner, caregiver, siblings, peers, and other people in the child’s environment; the quality of reciprocal social exchange in verbal and nonverbal communication (e.g., eye contact, gestures, facial expression), ability to engage in reciprocal imaginative play with representational toys; presence of perseverative, stereotypic, and ritualistic behavior; and presence of problematic behaviors such as attention difficulties, overactivity, aggression, and self-injury.

The laboratory evaluation of choice for developmental disorders continues to evolve as increasingly detailed understanding of the human genome leads to improved delineation of syndromes and advances in cytogenetic,
molecular, and neuroimaging techniques. Routine test batteries are not effective. The incidence of identifiable metabolic disorders in children with developmental delay is low, ranging from 0% to 5% (20,21). In addition, nonspecific, nondiagnostic, or false-positive “abnormalities” are common in routine nondirected laboratory testing. This can lead to further futile laboratory pursuits. Most infants and children with inborn errors of metabolism show signs or symptoms of their metabolic disorder such as hepatosplenomegaly, failure to thrive, intolerance of certain food groups, intermittent emesis, recurrent unexplained illnesses, seizures, intermittent somnolence, or fluctuating hypotonia. The relevant metabolic tests are reviewed in Chapter 1. In patients with developmental disorders who lack signs or symptoms of a metabolic disorder, metabolic screening should be deferred (22,23,24).

Many individuals with mental retardation have associated behavioral, emotional, and psychiatric disorders and nonneurologic congenital anomalies. A composite pattern recognition or cytogenetic testing may lead to a specific diagnosis (25). In these circumstances, the diagnostic label refers to the most salient aspect of the condition or is noted by an acronym, such as CHARGE association (of congenital anomalies) or eponym (e.g., Angelman syndrome), or by the underlying chromosomal abnormality, such as at chromosome 22q11.2. A specific molecular cytogenetic analysis, such as fluorescence in situ hybridization (FISH), may be indicated if the child has features of a known mental retardation syndrome (e.g., Williams or fragile X syndrome) or autism (e.g., chromosome 15q11 duplication) (see Chapter 4).

In most individuals with mild mental retardation, there are no specific neuropathologic correlates. Neuroimaging has a low diagnostic yield unless there are suspicious physical findings or localized neurologic deficits (26). Delayed myelination visualized on magnetic resonance imaging (MRI) is nonspecific in infants and children with neurodevelopmental delays (27). The diagnostic yield of neuroimaging increases considerably if the mental retardation is severe or complicated by microcephaly or macrocephaly, major motor abnormalities, features of a genetic syndrome, or seizures. MRI may reveal migrational disorders (e.g., lissencephaly), midline defects (e.g., holoprosencephaly or septo-optic dysplasia), or other brain malformations and disruptions (e.g., schizencephaly) (see Chapter 5). When the history suggests metabolic insult, vascular compromise, infection, or trauma, neuroimaging may reveal evidence of destructive central nervous system (CNS) changes.

There is no consensus on which children with developmental delay should be studied by imaging, and there are considerable variations in technique, methodology, demographics, and interpretation between the few available studies of computed tomography (CT) and MRI in children with mental retardation. Correspondingly, the reported frequency of abnormal neuroimaging findings in persons with mental retardation ranges from 9% to 80% (28,29,30,31). Positron emission tomography (PET), single photon emission CT (SPECT), functional MRI (fMRI) (32), and volumetric and morphometric MRI (33) provide insights into specific conditions but have limited overall application in the infant or child with a nonspecific developmental disorder. Formal audiologic testing is required to rule out hearing loss in all children who have language delay and may have isolated high-frequency hearing losses. Children with hearing impairment should be followed longitudinally for speech therapy and possible speech amplification.

Electroencephalography (EEG) is indicated if the history suggests seizures or regression or plateau in language acquisition (see Chapter 14). EEG can assist in the diagnosis of certain children with developmental disorders (34). Seizure disorders are often in their own right associated with diminishing IQ, as is their treatment with antiepileptic drugs (see Chapter 14). Memory may be deficient when the seizures are generalized. In complex partial seizure disorder, verbal and nonverbal memory may be impaired when there are left and right foci, respectively (35). Epileptic children as a group have inferior attention and general slowing of mental processes (36) as well as a varying degree of lag in language and reading ability (37). Children with frontal epileptic foci exhibited a relatively greater degree of impulsivity and disordered planning, verbal fluency, and motor coordination than generalized and temporal comparison groups (37). A comprehensive treatment of the neuropsychology of childhood epilepsy is available (38).

In habilitation, rehabilitation techniques are applied to individuals who, rather than having lost previously mastered skills, failed to master new skills at the expected age. This applies to children with developmental delay. For purposes of habilitation, the physician is less concerned about the level of function, and more with the implementation of evidence-based effective and enabling therapies, regardless of etiology. Schools are the primary venues of such therapies, and the consensus is swinging toward ever stricter demand for evidence-based justifications for such educational practices. For example, the Federal Department of Education maintains a “What Works Clearinghouse” that periodically reviews and reports educational efficacy evidence that meets high standards (39).

Children who are considered to have cognitive deficiency may come to attention because of poor academic achievement, abnormal behavior, or both. Is the lack of academic achievement due to limited cognitive ability, or does it represent underachievement relative to the child’s potential for other (e.g., psychosocial or emotional) reasons? If the child appears to be achieving at his or her
maximum level, consistent with that child’s specific profile of cognitive or attentional strengths or weaknessess, then special needs are identified, and an individualized educational plan (IEP) is implemented. If the child is underachieving, the causes are sometimes found in the child’s school and social setting or emotional well-being. In only a few developmental syndromes, notably attention-deficit hyperactivity disorder (ADHD), can cognitive potential be enhanced by medical means. In particular, there is as yet no evidence that learning disabilities, distinct from any comorbid attentional disorder, are pharmacologically treatable.

Most disorders of mental development are long lasting, if not permanent. Whether the child’s condition declares itself at or before birth or during early childhood, the parents have to adjust their previous expectations of a perfectly healthy child to the current reality. This adjustment is traumatic, and not all parents navigate its difficulties with success. The problem is amplified if, as is commonly the case, a specific diagnosis as well as a specific prognosis cannot be formulated. Such opinions as “developmental disability” convey confusion and are no substitute for the clarity and credibility of a specific diagnosis. When the disability is chronic and not fully explained—for instance, as in mental retardation and in autism—the quality of the parent–child relationship may suffer for lack of clarity about what to expect in the future. The clinician cannot always persuade the parents of what appears to be the reality of the situation, or even of the diagnosis, when that can be ascertained. Continued uncertainty is detrimental to parent–child attachment as well as to the functioning of the family unit. Under such circumstances, encouragement to seek a second opinion can with advantage be accompanied by referral for psychological counseling.

The Education for All Handicapped Children Act (EHA) was renamed the Individuals with Disabilities Education Act (IDEA) in 1990. This legislation requires schools to provide an appropriate individualized educational plan to students with disabilities, to be implemented in the “least restrictive environment.” The IDEA mandates a nondiscriminatory assessment, active involvement of parents in the educational process with due-process rights and hearing, and access to indicated ancillary services such as physical therapy, counseling, and transportation. The reauthorization of IDEA requires that students with disabilities be educated with their nondisabled peers to the extent possible, and that schools actively plan for student transitions (and that traumatic brain injury and autism be considered separate categories). The word handicapped in the original version of the law was replaced with the word disabled. Inclusion refers to the effort to include students with disabilities in general classrooms while providing special services outside the general classroom as needed. Currently, approximately 70% of all students with disabilities attend mainstream classes during part of the school day.

In infancy and early childhood, receptive and expressive language and play skills are the best indices of cognitive level. The prognostic reliability of cognitive testing improves when children reach an age at which some language competence is expected and a wider range of testing modalities becomes available. Language development up to age 3 years can initially be documented in the clinical setting with the Early Language Milestone Scale (40). Early cognitive and language developmental milestones in infancy and childhood are summarized in Table 18.1.

The diagnosis of a cognitive deficiency requires that a clinical impression be validated by standard psychometric assessment or referral report. The assessment that is appropriate for infants and children with developmental delays or disabilities depends on the child’s age, level of impairment, and the presence of additional sensory deficits. For clinical screening up to age 6 years, the Denver-II, formerly the Denver Developmental Screening Test–Revised (41), has been commonly used. The Denver-II draws on parent history, testing, and observation and takes approximately one-half hour to administer. However, for predictive value, the Battelle Developmental Inventory Screening Test (BDIST), which screens children up to age 8 years, is preferred (42) though it is not designed to identify specific learning disabilities. A trained infant psychometrist is likely to adminster the Bayley Scales of Infant Development II, which estimate the level of cognitive and motor deficits in infants (43). However, because infant cognition is limited, only sensorimotor functions can be evaluated accurately for normative purposes. These functions are nonspecifically correlated to an infant’s later emerging cognitive skills, the lack of which may result in a developmental disability in later years. More specific information about selective deficits in cognitive operations can be gleaned from the results of neuropsychologic testing. A cognitive profile is established for diagnoses and functional evaluation, the monitoring of further development, and rehabilitation planning. Some commonly used measures of cognitive and adaptive behavioral development are summarized in Table 18.3.

Administered with proper reservation, an IQ test is by far the most valid means toward establishing a child’s level of functioning—knowledge that has wide clinical application. Regardless of disagreements as to whether IQ is a valid measure of intellectual or educational potential (58), the IQ test result is unequivocably the best guide in estimating overall current levels of functioning. IQ within the normal range by definition eliminates mental retardation

as the reason for preschool or school achievement below the expected level. It does not exclude the possibility that specific learning disabilities are compromising the acquisition of some academic skills. Therefore, IQ has been withdrawn from the necessary definition of learning disability in the 2004 IDEA. Furthermore, no single IQ test measures all cognitive operations, and various subtests involve multiple cognitive processes. Therefore, a low score on a subtest would be unlikely to isolate the basic cognitive deficit that prevents the child from learning to read, write, or calculate.

Some “scatter” between the levels of subscale or subtest scores on an intelligence scale is normal. Only when it is statistically excessive might it reflect a disparity in the level of development of different intellectual skills. Verbal-performance discrepancy scores are often reported. Children with mild mental retardation tend to score somewhat lower on the Wechsler verbal subscale than on the performance subscale. A much lower verbal than performance IQ accompanies selective language delay; the converse discrepancy is often found in association with developmental difficulties in spatial orientation and visuomotor control, and to some extent in Asperger disorder.

The Stanford-Binet scale (46) and the Wechsler Preschool and Primary Scale of Intelligence (WPPSI) (44) are most commonly used for preschoolers. The Wechsler Intelligence Scale for Children (WISC-III) is available for school-aged children (45). These are well-standardized test batteries that call for the use of a range of different cognitive skills. For deaf children, the WISC-III performance scale and the Hiskey-Nebraska Test of Learning Aptitude have been specifically standardized (59). On such tests, deaf children without additional disabilities score close to the population average. The Perkins-Binet (60) and the Blind Learning Aptitude Test (61) are suitable for blind children. However, the most widely used psychometric instruments lack sensitivity in the mental retardation range. Even simple tests that use single words, such as the Peabody Picture Vocabulary Test (56), cannot be used if the patient is nonverbal. The Leiter International Performance Scale is useful for evaluating low functioning individuals (50). The interpretation of psychometric findings in mentally retarded subjects is complicated by uncertainties about the child’s test orientation and motivation. Because IQ levels in children younger than 3 years are unstable, the low-scoring child should receive a more definitive evaluation at a later date.

The developmental history and physical findings in early infancy are unreliable bases for prognosis as to developmental outcome. For example, Apgar scores of 0 to 3 at 15 and 20 minutes are predictive of high mortality and high probability of disability; however, many of these infants score normally in later developmental assessment (see Chapter 6). Intelligence tests in infancy yield a developmental quotient (DQ) that is necessarily heavily loaded with factors that relate to motor development. Within a normal population, the DQ is a poor predictor of individual differences in cognitive development. It has some success, however, in distinguishing a normally functioning from a mentally retarded population. Infant testing is highly specialized and also calls for the child’s cooperation. If this cannot be enlisted, the mother can be asked standard questions about her child’s social development (53).

Taken literally, the IQ scores can underestimate the child’s cognitive potential. Specific adaptive limitations often coexist with strengths in other adaptive skills or other personal capabilities. Limitations in adaptive skills may reflect the context of the child’s community. Motivation to perform during testing cannot be taken for granted, particularly when the child’s cultural or ethnic background has little understanding of and interest in cognitive testing and its uses. Anxiety can block the child’s reasoning processes, alienation and withdrawal can render the child unavailable for the task, thought disorders can intrude, and impulsive behavior can disrupt test performance. Lapses of attention owing to subclinical absence seizures can interrupt concentration, and psychotropic and antiepileptic medication may impair some cognitive processes. The limited attention span of patients with ADHD can degrade performance on tasks that demand sustained attention and cognitive effort. In young children, even gross impairment of vision, hearing, or touch can go unrecognized, and the child’s imperfect response to instruction or questions can be misattributed to defiance or cognitive deficiency. Other factors that complicate the interpretation of mental test performance are extremely low or high parental educational level, poor motivation with failure to expend mental effort on the test, depression, fatigue, and fear of failure with low self-esteem. Such potential sources of misinterpretation, which almost always lead to underestimation rather than overestimation, can never be fully ruled out, but conventional best practices expect the psychologist’s report, usually in writing, to provide an estimate of the reliability of the reported score, taking such factors into consideration.

The terminology that is used to characterize persons with cognitive and adaptive behavioral disabilities has evolved, in repeated efforts to minimize social stigma. Terms such
as cognitive deficiency, intellectual disability, and learning disability have been suggested to replace mental retardation, a term that misleadingly implies delay and therefore the possibility that the child will catch up. A similar term is developmental delay, which is nonspecific and is more of a chief complaint or a symptom complex than a diagnosis (62). Whereas children with mental retardation may learn new skills as they mature and are taught, most functional neurologic impairments are associated with intractable deficits of intellectual capacity. Mental retardation is a compendium term that does not refer to particular groups of diseases, syndromes, or medical disorders. It is not applied to cognitive deficiency that is a result of neurodegenerative and progressive neurometabolic diseases, or is secondary to psychiatric disorders.

The World Health Organization (ICD-10) characterizes developmental disabilities as due to impairment of the central nervous system, causing a functional disability, which consequently handicaps the individual in activities of daily living (63). It characterizes mental retardation as “incomplete or insufficient general development of mental capacities.” Formal definitions of mental retardation that are currently recognized emphasize a descriptive diagnosis of persons with significant disability because of “subaverage intellectual functioning” and “concurrent deficits or impairments in present adaptive functioning [i.e., how effective persons are in meeting the standards expected for their age by their cultural group]” (64,65). In addition, the manifestation of brain dysfunction must originate during the developmental period of life. The three formal definitions differ slightly in their approach to the diagnosis because of differences in emphasis on adaptive skill symptom menus or underlying etiologic factors.

Because mental abilities are on a continuum, the quantitative definitions of mental retardation and its subdivisions based on the IQ are necessarily arbitrary cut-off points. The mental age (the age equivalent at which the child functions on the test) divided by the chronologic age is the IQ. The mean I.Q. for the general population is set at 100. Higher IQ predicts more academic success in such terms as higher grades, more years completed, and higher standardized test scores. It also predicts positive life outcomes, such as better mental health, lower divorce rate, lower level of criminality, and greater occupational success (66,67).

The standard deviation (SD) of a test refers to the distribution of scores around the mean. Mild, moderate, severe, and profound mental retardation have been traditionally associated with cutoff points in standard deviations below the mean of –2, –3, –4, and –5, respectively (68) (roughly corresponding to IQs of 70, 55, 40, and 25). IQ scores between 70 and 85 are sometimes referred to as borderline. But measured intelligence does not offer specific information about the individual’s adaptive skills (Table 18.4). There is ongoing controversy on whether to emphasize IQ level or adaptive behavioral deficits as the central defining characteristic of mental retardation (69). The extent to which persons with mental retardation succeed in life is determined by their level of adaptive skills rather than by their IQ. However, there is no single measure of
adaptive behavior. Newer efforts to define mental retardation have emphasized the amount of support needed for an individual to succeed or maintain basic activities of daily living, instead of focusing on global degrees of impairment (70). Support intensities are subdivided into four levels: intermittent, limited, extensive, and pervasive. Support functions are classified into eight categories: teaching, befriending, financial planning, behavioral support, in-home living assistance, community and school access and use, and health assistance. This categorization acknowledges that the diagnosis of mental retardation is useful only as a pointer toward providing the individual with additional supports for activities of everyday living beyond those that are considered to be customary for the child’s age.

Further insight into the child’ strengths and weaknesses is afforded by neuropsychological testing, which attempts to tease out of the cognitive spectrum different domains of functioning that can be differentially affected by brain damage, depending on which areas have borne the brunt of the insult. These techniques are most frequently used with normal functioning children who have school problems but also can be applied to mildly mentally retarded children. The resulting information may be helpful with respect to diagnosis, prognosis, and educational and vocational planning (e.g., 71). IQ labeling should not replace repeated educational and achievement assessments. Not only is the IQ score unreliable below age 3 years, but also deficits in visual, spatial, and motor function, as well as expressive language, may influence early testing. Underprivileged and understimulated children may perform poorly on language-based tests for reasons that are sociocultural rather than neurobiological. Such a child’s level of functioning is better reflected in the results of nonverbal testing, as on the Culture-Fair Intelligence Test, Scale 2 (72), Raven’s Progressive Matrices (73), and the Universal Nonverbal Intelligence Test (74). The interpretation of IQ test results should take into account cultural and linguistic diversity, differences in communication ability, and attentional and behavioral factors.

A syndrome is a recognizable and consistent pattern of multiple manifestations that are known to have a specific etiology (75). More than 1,000 defined syndromes involve multiple congenital abnormalities (76). Most feature some pattern of impaired mental development. The degree of mental retardation is quite variable in most of these syndromes and is not indicated by the severity of neurobiologic concomitants, such as degree of abnormality of brain structure as visualized on MRI. Also, the deficit is often compounded by social factors. Dysmorphic facial features may give a misleading impression of low intelligence, and severe neurologic handicap may limit the children’s ability to express their intellect. Striking instances are athetosis and Lesch-Nyhan syndrome. Teachers may underestimate the child’s learning potential and offer a restricted number of learning opportunities. Nonetheless, some syndromes are characterized by specific behavioral phenotypes—of cognition, personality, and psychopathology—over and above the generally diminished level of functioning and any negative effects of the child’s environment.

Behavioral phenotypes have been characterized as follows, “a number of conditions, recognizable by a common physical phenotype, single gene defect or chromosomal abnormality, seem also to have a constellation of behaviors or cognitive anomalies which are characteristic” (77). Although mentally retarded children are by definition impaired in all or most cognitive domains, certain mental retardation syndromes present striking cognitive and behavioral dissociations. Selective neuropsychological deficits can occur against a background of generally deficient intellect as well as in otherwise normally functioning children. The adverse influence, whether genetic or due to intercurrent insults, is widespread, but lowering the level of intelligence below normative expectations, is not homogeneous, perhaps reflecting unequal impact on areas of the brain that differ in functional specialization. Examples of dissociations follow.

Mentally retarded children with autism are disproportionately handicapped in the language and social domains, domains that are relatively spared in children with Williams syndrome (78 and see Chapter 4). Children with Williams syndrome speak with appropriate phonology and syntax, although their comprehension and pragmatics are weak (see Chapter 4). However, they are especially deficient in visual-spatial construction and number sense. Conversely, visuospatial skills are a strong point for children with Prader-Willi syndrome, who display a specific flair for jigsaw puzzles (79). Down syndrome children are particularly impaired in receptive and still more in expressive language, long-term memory (80,81), and gross and especially fine motor skills (82). Particularly with respect to language, these impairments become more obvious after the initial relatively rapid development of the first three years has decelerated. Some Down children become autistic, especially if there is an autistic spectrum disorder family history (83). Girls with Down syndrome as a group consistently perform better than boys (84). Children with hydrocephalus tend to be disproportionately strong in speech production, producing rapid uninformative “cocktail chatter,” though not in speech comprehension and pragmatics. They are weak in face recognition and social cognition (85). Among X-linked chromosomal aberrations, fragile X syndrome (85a) is particularly handicapping in language development and social skills (86). Speech development is delayed, perseverative, and echolalic, and the children are
shy, anxious, and show many autistic features. There is fine and gross motor delay. The language and social domains also are problematic for children with XXX (87), whereas Turner syndrome (XO) is associated with visuospatial, arithmetic, and memory deficits (88) as well as deficient recognition of faces and facial expressions (89), while language is intact. XYY syndrome is characterized behaviorally by language disability, lack of sociability and aggressiveness, and Klinefelter syndrome (XXY) boys tend to exhibit low verbal abilities, clumsiness, and introversion (90). In Cornelia de Lange syndrome, nonverbal abilities may be selectively impaired (91). Untreated, phenylketonuria results in severe mental retardation. Even when effectively treated, children with phenylketonuria exhibit impaired performance on executive function tests (92), apparently implicating dopaminergic projections to dorsolateral prefrontal cortex (93). The characterization of the selective deficit as involving executive functions is disputed, however (94). Untreated children with congenital hypothyroidism are severely mentally retarded. After early treatment, deficits are confined to the domains of visuospatial skills, attention, and fine motor control (95). In galactosemia, language and executive impairments as well as fine motor deficits have been reported. Some of the children are mentally retarded (96).

It is apparent from the highly diverse listing presented above that deficits in cases of atypical development are not interpretable in terms of selective loss with respect to a single behavioral module or a single functional domain. Rather, there is a suite of more impaired domains, and a suite of better functioning domains, in every type of affected child. The cognitive phenotypes associated with a specific genetic abnormality also can be quite diverse, as for instance in Williams syndrome (97). Therefore, the phenotypes even of genetically determined syndromes appear to be multiply determined, with some effect of environment, in the broad sense that encompasses both prenatal and postnatal factors, interacting with the genetically determined susceptibility. Notably, language ability does not fall out or remain intact in one piece (98). The classical dispute about whether language skill is modular and unique among cognitive skills (99) is not resolved by the study of atypical development.

Persons with mental retardation are at increased risk for psychiatric disorders. Whereas the incidence of psychiatric disorders in the general childhood population is 7% to 10% (100), approximately one-third of all children with developmental disorders also have psychiatric disorders (100,101,102,103). In the severely retarded subpopulation, the frequency of psychiatric disorders is much higher and approaches two-thirds by adolescence (104). The broad spectrum of psychopathology includes affective disorders (105), anxiety disorders, autisticlike behaviors (106), conduct disorders (107), maladaptive behaviors (repetitive self-stimulation, self-injurious behavior, pica) (108,109), and psychosis. These emotional and behavioral disturbances are caused by multiple factors, both primary psychopathologies and neuropsychologic disorders and the secondary effects of illness, dependency, environmental deprivation, frustration, and low self-esteem (110).

The behavioral phenotypes of certain genetic syndromes are associated with typical personality characteristics and reproducible behavioral mannerisms (Table 18.5). Though otherwise often of even temperament, Down syndrome children may be irritable and inattentive because they are prone to obstructive sleep apnea (120), which in turn is a separate risk factor for neurocognitive (especially right hemisphere) deficits (121) and for depression (122). Specific personality differences also may occur in mental retardation syndromes, such as unusual sociability and distractibility in Williams syndrome and compulsivity, excessive talkativeness, impulsivity, and low activity level as well as compulsive eating and foraging for food in Prader-Willi syndrome (79).

Children with velocardiofacial syndrome have on average a borderline IQ, concrete thinking, bland affect, and little social interaction. They are highly susceptible to
psychiatric disorders, particularly schizophrenia, with which they may have a frequently occurring deletion on chromosome 22 in common (123). Twenty-five percent of children with velocardiofacial syndrome present with schizophrenia by early adulthood. A higher proportion still presents with ADHD in childhood and responds favorably to stimulant therapy (124). Girls with Turner syndrome tend to be unassertive and compliant and maintain poor and sparse social relationships (125). Children with Lesch-Nyhan syndrome have a unique form of compulsive behavior, which is expressed by an apparently involuntary drive toward self-injury and aggression. The children apologize and ask to be restrained in the act of attacking others or themselves (126). Abnormal dopamine functioning during early development may be implicated (126).

Since the frequency with which a disorder arises in a population during a stated period of time is its incidence, and since its prevalence describes the amount within a population at a given time, the prevalence of the condition is the product of its incidence and its duration. The prevalence of a developmental disorder, therefore, may change over the years because of prenatal diagnosis and more frequent termination of pregnancy, changes in the maternal population, and changes in health care. For example, the prevalence of Down syndrome increased from the 1920s through the 1960s, as persons with this condition began to live much longer because of improvements in medical care. However, since 1980, further increases have been slight (127,128).

The prevalence of mental retardation is 6 to 20 per 1,000 (129,130,131,132). The published estimates vary on account of regional differences, ascertainment biases, and variations in diagnostic criteria and study methodology. Mild mental retardation is 10 to 12 times more common than severe retardation. Mental retardation is more common among males because of X-linked syndromes involving mental retardation, especially fragile X syndrome. The male to female ratio ranges from 1.3:1 to 1.9:1 (133). Morbidity and mortality are greater in persons with severe mental retardation, not due to the mental retardation itself, but because of the severe cerebral palsy that is often associated with it. Total immobility and feeding by nasogastric tube are the features that are most predictive of curtailed life span (134,135).

As a psychopathology, autism is profoundly internalizing. The children are deeply preoccupied with their internal states of mind and feeling and only minimally engage the outside world for purposes of activities of daily living. The majority of children with autism are to a varying degree atypical even at an early age, moving and crying little, averse to being held, and content to be alone. Scales that assist earlier diagnosis are available, and home videos reviewed retrospectively can reveal a previously unsuspected onset during infancy (315,316). The children may reject solid foods, and toys elicit little interest or are held onto with unusual obstinacy. Their motor development is usually normal, but minor physical anomalies are more frequently found in autistic children than in controls (317). During the second year of life, children with autism are typically unusually sensitive to auditory, visual, tactile, and movement stimulation, and repetitive mannerisms or gestures begin to appear. The older child with autism has restricted and stereotyped behaviors and activities and markedly impaired creativity, social relations, language, and perception. Disturbances in social relations are manifested by poor or absent eye contact and a lack of interest in people, whom the child uses instrumentally rather than interactively. Toys are handled bizarrely or are dropped when handed to the child. Mannerisms include repetitive, stereotyped movements, such as hand flapping, ear flicking, or head banging (318). Toe walking, whirling, and rocking also are common. These movements, often complex, increase when the child is anxious or is confronted with a novel situation (319).

Language disturbances can consist of complete lack of speech, failure to communicate by pointing, failure to imitate (320), perseveration, echolalia (repeating words or phrases out of communicative context), and lack of reciprocity in dialogue (321). If speech does develop, its prosody is often monotonal, pronouns are used poorly, and humor is not understood. Speech in children with autism differs qualitatively from the speech of children with mild mental retardation. Within the limits of a restricted vocabulary, children with mild mental retardation typically speak normally, whereas children with autism use abnormal prosody, syntax, semantics, and pragmatics (322). In some children, termed hyperlexic, mechanical reading skills develop out of proportion to spontaneous verbal expression or reading comprehension (323,324).

Children with autism may be oversensitive to stimuli, notably self-created sounds or spinning objects (325). This oversensitivity can alternate with periods of nonresponsiveness to speech, objects, or pain. Affected children frequently ignore sounds and are “aloof,” provoking testing for possible hearing loss. Some have prolonged attention spans during self-initiated activity (326). Observational studies demonstrate hyperselectivity of attention in children with autism (319). They focus exclusively on a single distinctive aspect of a display, while ignoring its other features. They are impaired in broadening the spatial spread of visual attention (326a). Conversely, the excel as compaud to their typical peers on tests that calls for local perceptual processing, in which context has got to be discounted. An example is the Embedded Figures Task (326b). The propensity of some people with autism to achieve savant status on very narrowly defined skells (326c) also reflects their narrow focus. Children with autism also have been shown to be impaired in perceiving faces, but not objects (327). People’s clothing was more salient to autistic children then their emotional expressions (328). They fail as compaud to age-matched peers in infering other
people’s perspectives and mental states, focusing narrowly on their own (326d).

Some children exhibit transitional or milder degrees of autistic behavior, within the PDD spectrum. After age 5 to 6 years, such behaviors may merge with those observed in mental retardation. Others deviate into childhood disintegrative disorder (329) or Asperger disorder (330). The natural history and outcome of autism is variable. Some 80% manifest a pronounced impairment in intelligence. Approximately one-third of children with autism do not develop communicative speech. Others develop rudimentary language, although their communication remains literal and concrete and their affect flat. In yet other affected children, characteristics of organic brain disease become apparent over subsequent years, including auditory or visuoperceptual impairment. A small subset of children with autism develops bizarre thoughts and even delusions. About 75% of children who fail to show language skills by age 5 years fail to adjust personally or socially. Other children with autism make a fairly adequate social adjustment but retain certain peculiarities of personality, lack of humor, and unawareness of social nuances (331). Neuropsychologic testing reveals deficits on problem-solving tasks (332). Standard diagnostic instruments include the ADI (332a) and the ADOS (332b). Early diagnosis is facilitated by the use of standard instruments such as the Checklist for Autism in Toddlers (CHAT) (333). It consists of nine questions asked of the parents, and five items that call for observation by the clinician. The latter include pointing and looking to where a parent is pointing and pretend play, activities that are available to a normal 18 month old. Approximately one-half of the population with autism have an IQ above 50, and one-fourth to one-third have an IQ above 70. Children who regressed into autism after initially developing normally generally are ultimately more impaired cognitively (312,334). Even when they are enrolled in treatment programs, only some children with ASD improve significantly. If children improve, they usually do so during the second half of their first decade.

Based on twin studies, autism is thought to have the heaviest genetic loading of the chief developmental disorders, with more than 50% hereditability for the narrowly defined autistic disorder (335), and as much as 90% for the wider range of ASD (336). Moreover, autistic individuals rarely reproduce (337). Autism would therefore be expected to have a stable incidence. On the contrary, its prevalence across the ASD spectrum has risen from 2 to 4 in 10,000 children in the 1970s to 60 in 10,000 children in recent years, that is, 0.6% of the population (338,339,340,341,342,343,344,345). The rate of increase in autism is highest in developed countries, in the United States among children of American rather than immigrant mothers, and for autistic children who are high-functioning (346). It has been argued that the rise in prevalence is due to relaxation in diagnostic criteria, diagnostic substitution [e.g., ASD diagnosis substituted for Developmental Language Disorder (DLD) diagnosis—(346)], changes in and increasing availability of special education services, mandatory notification of ASD diagnoses, and increasing awareness and earlier recognition of ASD. However, the rise in incidence is so worldwide that it cannot be attributed to regulatory changes in just one country. Potential alternative causes and a true rise in incidence are not mutually exclusive, and the explanatory power of the alternative factors is open to challenge (347,348). No published research has managed to determine how much of the variance in the spectacular increase in ASD diagnoses is attributable to each of the various factors that have been suggested. Pending such information, it is unwarranted to assume that the true incidence of ASD is not increasing.

Classical autistic disorder is one of the five subtypes of pervasive developmental disorders listed in the DSM-IV. The four other subtypes are Rett syndrome (see Chapter 3), Asperger syndrome, childhood disintegrative disorder, and pervasive developmental disorder not otherwise specified (PDD-NOS). However, these subgroups may be better conceived as arranged along a spectrum of severity of autistic disorders (419).

High-resolution chromosome analysis leads to a diagnosis in approximately 5% of children with autistic traits. Brain imaging studies are indicated only if there are indications of progressive loss of function, localized neurologic deficits, or persistent focal seizures. An EEG is indicated if there is a suggestion of seizures. A spinal tap can assist in the differential diagnosis of new onset seizures or autistic regression. Approximately one-third of patients with autism experience one or more epileptic attacks during the first two decades of life. Various seizure types have been described. Of these, the most common are complex partial seizures (312). The question of subclinical seizures is frequently raised in the autistic subtype in which an acquired loss of receptive and communicative abilities occurs, as in Landau-Kleffner syndrome (see below and Chapter 14). For this reason, sleep electroencephalography has been advocated for children whose autistic symptoms develop after 2 to 3 years of age, in the hopes that antiepileptic drug therapy might improve their language and social skills (430,431). Magnetoencephalography appears to be more successful than electroencephalography in revealing the subclinical epileptiform discharges (432).

Metabolic studies, including serum copper and ceruloplasmin levels, are indicated if autistic symptoms develop in late childhood or are progressive. Cognitive, language, educational, and behavioral assessments establish the level of severity of the child’s impairments and serve as a baseline for IEPs and additional therapies.

Outcomes are most favorable in the least severely affected children, who have normal or near normal intelligence and speak before they are 5 years old (433). There is a strong impression among clinicians that applied behavior analyses (ABA), has improved the chances of children with autistic spectrum disorder to function independently in the mainstream school and society. Whether such outcomes draw upon brain plasticity, or are strictly due to improved behavioral strategies, is unknown. In their adult years, high functioning autistic or Asperger patients usually achieve normal understanding of speech. However, their semantic and pragmatic deficits are lifelong and make it hard for them to converse normally while letting the listener take his turn or introduce a new topic (434).

Autistic spectrum disorder and developmental language disorder are classified as distinct and separate entities. However, deficient or absent language skills are a major characteristic of ASD. Is autistic language similar to or distinct from language in DLD?

Until recently, autistic language has been assumed to be different from language in DLD. Whereas DLD is characterized by impairment in understanding and producing phonology (speech sounds) and syntax (grammar), the language deficit in autism was thought to be at the “pragmatic” level, at which one understands what is said and uses language communicatively (e.g., DSM-IV). However, a majority of ASD and DLD children exhibit both receptive and expressive language difficulties. Some children with ASD even exhibit these difficulties in their extreme form, verbal auditory agnosia (VAA) (433,434,435). Those whose language problems are limited to pragmatics are actually in the minority. DLD children do not have pragmatic disorder, whereas ASD children do not have expressive speech problems in isolation. Thus, most ASD children have a language
impairment that is comparable to the receptive/expressive language impairment that stands alone in DLD.

The similarity in pattern of language deficit is accompanied by a similarity in the size and shape of the various parts of the cerebral hemisphere. Previously only seen at autopsy, these parameters can now be measured by structural MRI. On average, both ASD and DLD children have larger than expected brains in childhood, excess of cerebral white matter, and anomalous asymmetries between the right and left hemispheres. The volumetrics of ASD and DLD resemble each other more than either resembles the normal control state (436,437). So, the macroscopic structure of the brain in ASD and in DLD is quite similar.

The close relationship between DLD and ASD originates from closely associated abnormalities in the genome. Susceptibility loci on distal chromosome 7q are associated both with ASD and specific language impairment (SLI). When there are abnormalities in this chromosomal region, individuals with ASD, SLI, or both are characteristically found in the same extended family tree (438,439,440,441,442). Therefore, individuals with an abnormality on distal 7q have a genetic predisposition to both ASD and SLI and may express either one. The as yet unknown intervening variables that transform a risk into the reality of a language disorder might also be able to transform an ASD risk into overt ASD.

At least 5% to 7% of the school-aged children in the United States (aged 5 to 21 years) have serious deficits of speech or deficits of hearing, with which a speech disorder is associated. An additional 5% have relatively minor speech impairments (478,479). Approximately twice as many boys as girls are affected. Speech and language disorders overlap with academic underachievement and behavior problems in school-aged children (480,481).

Lesion, stimulation, and metabolic activation effects show that in almost all right-handed individuals, the left hemisphere is primarily concerned with the neural process underlying language as well as the decoding of symbols and the programming of motor sequences. Some deficits in tasks that are not overtly verbal have been reported in children with language delay, notably discriminating the order of presentation of rapidly successive stimuli in various modalities (488). With respect to other nonverbal functions, such as spatial orientation and spatial organization of nonverbal patterns, there is a less striking but definite asymmetry in favor of the right hemisphere. As the child matures and new skills emerge, the laterality conforms to that of the child’s precursor skills.

Left-handers are either genetic or pathologic (488a) in origin. The latter are thought to have incurred mild
prenatal left hemisphere injury, not sufficient to result in an overt neurologic injury, but enough to impair right-hand dexterity sufficiently for preference to shift to the left hand. Conversely, those who incurred early right brain damage became unusually right-handed. Regardless of whether they are genetic or pathologic left-handers, about one-half are left lateralized for language, a third bilateralized, and a sixth right lateralized. A person’s hand preference, right, left, or mixed, has little or no predictive value for cognitive development.

Although a newborn’s cerebrum is barely functional, most newborns more often spontaneously turn their heads and eyes to the right rather than to the left, a tendency that subsequently comes under control of the left hemisphere. Congenital left hemispheric maldevelopment and injury in the first year do not invariably lead to the expected gross language disorder or delay (489a); the right hemisphere assumes language representation (489b). Preschoolers show transitory language deficits after left hemisphere injury, but usually only after the first decade does the language disorder assume the characteristic severity and persistence of adult aphasia (490). In very early lateralized lesions of either hemisphere, there is often contralateral compensation, but age, lesion site, lesion size, presence of seizures, and use of antiepileptic drugs or hemispherectomy contribute to variable outcomes (491,492).

Language lateralization was believed to be progressive, from bihemispheric origins toward the end-point of fully established left hemispheric dominance for language, only near the end of childhood. The weight of evidence is against this theory, however. Instead, it now appears that language processes and their antecedents are lateralized from the beginning, in conformity with how they will be lateralized in their mature adult state (491,493). This inference was directly validated when the location and extent of the language areas in young children were probed by intracranial electrical stimulation and found to be much like those of adults (494). Children are credited with more plasticity of the nervous systems than adults, in which their higher levels of brain glucose metabolism (495) and ongoing synaptic pruning (496) may be involved. Specifically with respect to language (497,498,499), young children show more recovery than older children and adults after unilateral brain damage. At any age, recovery after unilateral hemisphere injury can rely on adjoining ipsilateral areas (500), or homologous contralateral areas (501), to take over the compromised function (see also section on Acquired Childhood Aphasia, below).

The nonexistence of progressive lateralization invalidates a host of theories offered over the years that attribute a variety of language disabilities to delay in the “lateralization process.” Because no such process exists, therapy programs that purport to jump-start or accelerate it by manipulating the side of entry of stimulation, the child’s positioning, and which hand and eye the child is permitted to use (e.g., 475), are irrational.

Left hemispheric control of language involves inhibition of the potential language capability of the right hemisphere. Massive left hemispheric lesions release this inhibition. Disconnection of the hemispheres by section of the corpus callosum reveals that the right hemisphere can decode at least simple speech messages. The right hemisphere’s ability to comprehend speech also emerges when the left hemisphere is temporarily anesthetized by intracarotid injection of amobarbital (502). The patient remains responsive to verbal commands, as indicated by left facial and left hand movements. This residual ability to comprehend and respond perhaps explains why permanent total receptive aphasia is so rare and hardly ever seen in children. Right hemisphere compensation for language disorder is more complete for language decoding than encoding (503). When the right hemisphere is involved in language from the start, rather than because the left hemisphere was malfunctioning, as happens in pathologic non–right-handers, language nonetheless develops in the normal manner. There is no basis for attributing language deficits, as in Down syndrome, to right-sided language lateralization per se. Agenesis of the corpus callosum yields few of the signs of hemispheric disconnection exhibited by patients who have undergone callosal section. Whether compensation for a prenatal injury involves brainstem commissures, redundant functional specialization in each cerebral hemisphere, or both, is not clear (504). Early right hemisphere damage, like later damage, results in sensory inattentiveness and motor impersistence, as well as negative affect, consonant with the early right-sided specialization for negative emotion (505).

Recent evidence also has implicated the cerebellum, and specifically its posterior lateral area, in developing language. A cerebellar cognitive affective syndrome has been proposed (506) and related to early language development by studies of the aftermath of cerebellar excisions during childhood (507). Transitory mutism is replaced by dysarthria and impaired spontaneity of speech. Neuropsychological deficits, including impaired executive functioning, are more severe with right-sided cerebellar involvement, perhaps reflecting the crossed organization of cerebral-cerebellar connections (508). In contrast, damage to the vermis is credited with causing disordered behaviors ranging from irritability to withdrawal (509). Congenital cerebellar hypoplasia may result in a similar profile of deficits, but obscured by the profound mental retardation that typically characterizes these children (510). The pattern of cognitive and emotional impairment described in children was previously established in adult postoperative cerebellar patients.

The prevalence of speech disorders is closely related to age. Speech disorders occur most commonly in young children
and reach an incidence of 15% in kindergartners (511). Approximately 90% of children younger than 8 years of age outgrow their speech disorders, but those who still have a speech impairment by 14 years of age can expect improvement only with speech therapy.

The frequency of production of sounds normally increases up to 2.5 years of age, at which time the speech sound pattern closely resembles that of the adult. Children’s phrases average 1.5 words at 18 months, 1.8 words at 2 years, 3.1 words at 2.5 years, and 4.1 words at 3 years. Ninety percent of children can correctly articulate all vowel sounds by 3 years of age. For boys, a 90% success rate for consonant sounds comes later; by age 3 years, they can pronounce p, b, m, h, w, d, n, t, and k; by age 4 years, ng; by age 5 years, y; by age 5 to 6 years, j; by age 6 years, zh and wh; and by age 7 years, f, l, r, sh, ch, s, z, th, and v. Girls achieve some of these sounds slightly sooner. Speech sounds are more accurately spoken when they are in the initial rather than a middle position. They are least well articulated at the end of the word. Impaired intelligibility is prognostically less ominous than depressed level of language use, and referral to speech therapy on its account can be delayed until 4 years of age. If speech also is sparse, however, as in the 2 year old who does not yet utter two-word phrases, earlier referral is justified. Even so, progress is slow, and a special effort by the parents is needed to offer the child experience by talking to him or her more than is usual.

Whereas immature central control of articulation is by far the most common cause of speech disorders in children, definable neurologic deficits contribute in some cases. Quadriplegia or double hemiplegia associated with suprabulbar palsy causes speech to be slurred, and athetosis renders it jerky, explosive, and indistinct. Congenital suprabulbar palsy is a major impediment of speech. A hyperactive jaw reflex distinguishes congenital suprabulbar palsy from acquired nuclear and infranuclear paralysis of muscles of articulation. Buccolingual apraxia affects voluntary tongue, lip, and palate movements but spares automatic movements (512).

Stuttering presents between the ages of 2 and 4 years, with maximal incidence at age 3 years. Many children stutter only transiently, but in some, the stutter persists. The underlying cause of stuttering in childhood is unknown. Rarely, it appears secondary to organic brain disease, such as focal dystonia. When stuttering is a symptom of an extrapyramidal disease, the patient is more likely to repeat an entire word, rather than a single sound (palilalia). Relatively more boys and left-handed persons stutter. The notion that stuttering is the result of left hemispheric dysfunction is supported by positron emission tomographic studies (513,514). Language representation may be bilateral in children who stutter, with rivalry between the hemispheres for control over speech (515). Stuttering cannot be a disorder of auditory feedback because speech, like any highly practiced and automatized process, is not under continual feedback control. However, stutterers do find it difficult to disengage their attention from their own speech, which then cannot be automatic. Stuttering is relieved by noise that is sufficiently loud to mask speech sounds, and it is rare among deaf persons.

The child who stutters has difficulty in passing smoothly from phoneme to phoneme, especially at the beginning of sentences, and with words of more than five letters. The child who stutters explosively reiterates a single sound or blocks speech completely. Moments of self-consciousness and embarrassment yield maximal dysfluency, whereas distraction can result in temporarily normal speech. Singing is spared. Speech therapies include awareness training, regulated breathing, and social support. Treatment outcomes are variable (516).

Distinct from stuttering is cluttering, in which speech is fast, slurred, and dysrhythmic, with omission, reduplication, and transposition of speech sounds and words. Like stuttering, it can rarely result from brain lesions (517).

The DSM-IV differentiates developmental language disorders from those secondary etiologies such as epilepsy, mental retardation, autism, and severe brain injury that have other prognosis and treatment implications (48). Secondary language disorders may be associated with risk factors such as low socioeconomic status and other family and environmental adversities or with learning disabilities (518).

Twins who share a secret language, that sounds coherent though incomprehensible, rapidly transfer to normal language when they are separated. Blind children of normal intelligence tend to be slow to begin to learn to speak and often pass through an echolalic phase. Subsequently, their vocabulary grows at a normal rate, but they lack imitative gestures. Some “electively mute” children withhold speech, either totally or only outside the home (428), but comprehend well and otherwise act normally. Although some of them may be emotionally disturbed, the problem may be quite superficial, and the seriousness of the prognosis should not be overestimated. In autism, speech is not only limited but is pervaded by echolalia, stereotypic utterances, and avoidance of the personal pronoun I. Children with autism have less difficulty with articulation and more with comprehension than do children with DLD who are matched for nonverbal intelligence. Bilingual children, who have difficulty in acquiring the second language, are ones whose ability with the first language also is limited. A bilingual environment may explain some idiosyncratic uses modeled on the first language, but it does not explain delayed language development (519).

The language of culturally deprived children reflects local speaking patterns that are better described as different than impaired. On formal tests, such children score poorly relative to mainstream norms. The normally functioning
child who has been totally deprived of language experience does not speak but rapidly learns once the opportunity is offered (unless the child is secondarily emotionally disturbed). Extreme instances of children isolated from language, such as “wolf children” (520), have been reported, but even 6 years of total deprivation can supposedly be overcome (521), although a striking counterexample exists (522). Evidence for a critical period for first language development is fragmentary. It is argued that second language learning results in lower ultimate competence the later it is begun, up to age 16 years. Beyond this age, there is no further relationship between age and ultimate level of proficiency in the second language (523). It also has been argued that children generally recover faster than adults from brain injury, including injury that causes aphasia, presumably because the brain is more “plastic” and better able to reorganize before maturity (524, but see 169).

Three million American children have hearing deficits; 0.1% of the school population is deaf and 1.5% hard of hearing. Of cases of hearing deficits, the majority are congenital sensorineural, and related to genetic susceptibility factors. Children who are born with hearing loss in excess of 70 dB (severe) or even 90 dB (profound) cannot hear conversational speech and therefore do not learn to talk unassisted. Ostensibly, conflicting approaches to early intervention are the auditory-verbal and the bilingual-bicultural philosophies. When young deaf children are taught by the oral method, acquisition proceeds through the expected stages but falls far short of the expected eventual level of proficiency (525). In contrast, young children learn American Sign Language (ASL) spontaneously when exposed to it by their deaf parents. Their ultimate proficiency is an inverse function of the age at which learning began (526). Thus, it is essential to begin early. Consistent with the status of ASL as a natural language, its use depends on left hemisphere function, even though in execution it is visuospatial (527). However, ASL may not be particularly well suited to learning to read, since its gestures largely represent whole words, with less emphasis on individual phonemes. So-called “cued speech,” wherein the individual gestures often represent individual parts of words, seems to provide more support for learning the underlying structure of the language, thereby facilitating learning to read. “Total communication” aims use both oral and sign approaches flexibly, depending on the needs of the child (528). It has become the most commonly used teaching method for deaf children (529).

Deafness often goes unnoticed, even by observant parents, particularly when a bright child uses contextual cues to compensate for not hearing words. It can be suspected, however, as early as 4.5 to 6 months of age, if a child fails to look toward a sound source beyond his or her visual field. In young children, free-field audiometry is used for definitive diagnosis. This study should take place as early as possible so that the child does not go without appropriate language training (after the fitting of a hearing aid, if it is shown to help) and does not acquire the habit of ignoring auditory stimuli. Rarely, the auditory processing disorder is central, calling for tests of auditory attention and discrimination of speech modified in various ways. Performance on central auditory processing tests also is impaired in children with ADHD and improves after stimulant therapy (530). Children with recurrent otitis media can be at risk for language delays (531).

Congenitally deaf children with the usual high-frequency deafness adopt a harsh “deaf tone.” When children become deaf abruptly (usually owing to meningitis), the effect depends on the stage of their language development. Children who become deaf when younger than 5 years gradually lose their ability to control articulation and voice production, sounding increasingly like congenitally deaf persons. Children with as little as 1 year of language experience are substantially easier to train in language skills than the congenitally deaf, whereas children who become deaf before age 2 years become indistinguishable from the congenitally deaf. The early provision of hearing aids and training for all deaf children improves their language progress. The language skills and educational achievements of prelingually profoundly deaf children approximated the hearing population norms after cochlear implants (532).

In the 3–10% (532a) of children who have developmental language disorder (DLD), alternatively named specific language impairment (SLI) (533), language develops abnormally slowly. The selective language delay results in speech that notably develops late, but also is semantically and syntactically impoverished. The diagnosis of developmental language disorder excludes secondary causes, unless the coexisting condition is unable to account for the severity of the language problem. Thus, an individual with mild mental retardation who is disproportionately impaired in verbal abilities can be considered to have developmental language disorder as well. However, the language impairment must not be explicable by other deficits, physical abnormalities, or disease processes, or by social or emotional deprivation. It is therefore a diagnosis of exclusion. Nonetheless, children with DLD exhibit subtle associated deficits in cognition, emotion, and motor performance (534).

The language of 2% to 3% of 3-year-old children is deficient in expression, reception, or both. Unlike speech in mental retardation, speech in DLD also is characterized by several articulatory disorders. These conditions differ from speech or articulation disorders such as stuttering, cluttering, lisping, and cleft palate nasality in that they are more severe, affect a wider range of phonemes, and involve a far
greater difficulty in making the transition from phoneme to phoneme than in pronouncing the phonemes individually. This difficulty results in grossly curtailed word formation. Short binding words (“function” rather than “content” words) are frequently omitted.

In contrast to the relatively fixed deficits of speech disorder, delayed language is sensitive to context, and sound combinations that can be uttered at one time are unavailable at another. Greater mental effort often results in worse rather than better performance. The disorder is not limited to the spoken language; it affects writing, and even lip reading, manual alphabet, sign language, and Braille. Language delay invariably involves expressive speech. Verbal memory is impaired as well. In one subtype, speech comprehension can be relatively spared, although affected children have difficulty in discriminating speech sounds. In a less common subtype, semantic as well as syntactic comprehension is severely deficient. Rapin (535) distinguishes five subtypes:

When comprehension is severely affected, the problem is often initially mistaken for peripheral deafness. The impression that auditory acuity fluctuates, which leads parents to doubt that the learning disorder is genuine, is borne out by audiometry. Stable hearing thresholds are hard to obtain, and repeated testing is apt to lead to radically different audiometric profiles. Recording psychogalvanic skin responses to sound is informative. These instabilities are caused by inconstant focusing of attention on sound, a function that normally depends on the integrity of auditory cortex.

The claim has been made that a defect in auditory perception exists in some developmental language disorders, specifically, a difficulty in discriminating the patterns of transition from speech sound to speech sound (536). This problem is most common in verbal auditory agnosia as well as when phonology is permanently impaired. Affected persons may have variable degrees of difficulty in auditory, visual, and linguistic processing. Alternative mechanisms for the underlying nature of DLD remain controversial, however (537).

The neuropathology of developmental language disorder is poorly understood but it has been associated with a mild neuronal migrational disorder in left inferior frontal cortex (538). Children with semantic-pragmatic disorder sometimes exhibit autistic features that meet PDD criteria (429) and, like autistic children, perform poorly on tests of social cognition (539). Semantic-pragmatic disorder and autistic disorder also share neuropsychological features of right hemisphere dysfunction (540). The language characteristics of children with hydrocephalus and of children with Williams syndrome often fall into the semantic-pragmatic category.

A genetic susceptibility to language disorders is indicated by the aggregation of cases within extended families (541) and their co-occurrence in monozygotic twins (542). A common genetic basis for motor immaturity and DLD has been suggested (543). An abnormal gene, FOXP2 on chromosome 7q31, is associated with dyspraxic speech and FOXP2 may be fundamental to human expressive language ability (543a).

Developmental language disorder is associated with a substantially increased incidence of behavioral disorders (544,545). The psychiatric symptoms may occur secondary to social isolation. This isolation may be compounded by relative lack of early experience of language, when parents become discouraged and relinquish their attempts to converse with the child. In other cases, the language and behavioral disorders may arise in parallel from the same neurodevelopmental deficiency.

Aphasia is considered to be “acquired” when it results from a condition that occurs after language development has begun, generally after age 2 years (546). Whereas bacterial infections causing encephalopathy used to be the most common causes of acquired aphasia in children, traumatic lesions currently predominate (490). Stroke in childhood is another cause of acquired aphasia, when the dominant hemisphere is affected (547).

Acquired aphasia is almost always nonfluent, especially in the youngest children, and loss of spontaneous speech or even mutism occurs more commonly than in adults. Other syndromes analogous to those in adults also have been described in children, and with analogous lesion location. Right hemisphere lesions cause aphasia in children no more frequently than they do in adults (1% to 4%). When they precede the onset of language, lesions of either hemisphere can temporarily retard both language and spatial development (548). Of 10 children with extensive early left hemisphere insults, five showed bilateral or right language reorganization, whereas the other five compensated
within the damaged hemisphere. All but two of the children were left-handed, indicating more frequent transfer of hand preference to the lift than of language to the right brain (549).

Unless the lesion is bilateral, recovery of fluent speech is more likely in children with acquired aphasia than in adults, and disorders of comprehension occur less commonly. Residual minor language deficits and consequent school problems are common, however (550,551).

A rare type of acquired aphasia in children, acquired epileptic aphasia (AEA), otherwise known as Landau-Kleffner syndrome (LKS) (552) in addition to the seizure, is discussed in Chapter 14, under the heading of nonconvulsive status epilepticus. There is a severe comprehension deficit as well as an agnosia for environmental sounds. Oral expression is poorer than written, although both are impaired. Nonverbal intelligence is normal. Children with AEA typically develop normally until they are about 4 to 8 years old, when they fairly rapidly and quite unexpectedly cease to understand what is said to them and soon also lose the ability to express themselves in words. At its onset, the language regression is associated with some seizures and temporal lobe spiking on the EEG, but epilepsy rarely becomes severe in AEA. The language deterioration is often accompanied by the development of an autistic syndrome, perhaps in relation to abnormal temporal lobe metabolism (553). Perhaps temporal lobe dysfunction is why DLD and ASD are frequently associated in AEA.

Seventy percent of children with AEA have an associated severe behavioral disorder, featuring hyperactivity, impulsivity, and oppositional behavior. The severity of disturbed behavior follows fluctuations in the severity of the language disorder (554). Long-term outcome is poor despite attempts at cure by antiepileptic therapy and intravenous immunoglobulin (IVIG).

Normal intellect in domains separate from language is assumed in speech and language disorders. However, it is sometimes difficult to demonstrate, on account of the great contribution of language to overall intellectual performance. Tests that are verbally based, such as the WISC verbal subscale and the Stanford-Binet, cannot be used to estimate the intellect of children with speech and language disorders. Nonverbal alternatives include Ravens Colored Progressive Matrices for Children (73), the Leiter International Performance Scale (34), the Test of Nonverbal-Intelligence (TONI), and the Hiskey-Nebraska Test of Learning Aptitude (46). While discrepancy between a child’s language and nonverbal developmental levels is integral to the developmental language disorder diagnosis, there is no consensus on how great a differential is critical and how it might vary with age (555). Audiometry excludes cases secondary to high-frequency deafness, and it can reveal the fluctuating impairments of pure tone sensory threshold that are typical of cortical deafness. The language behavior itself is evaluated by one of the standard language test batteries, such as the Clinical Evaluation of Language Fundamentals (CELF). Physical examination reveals any mechanical or motor deficits of the vocal apparatus. In the rare instances in which receptive aphasia and deafness resist clinical differentiation, the ability to elicit auditory-evoked potentials confirms that the auditory projections to the auditory cortex are intact.

Electroencephalography and neuroimaging are often used to seek structural intracranial abnormalities but as a rule yield little information of practical use. Rarely, subclinical seizures thought to disrupt cognitive processes can be detected by EEG (556). In such cases, effective anticonvulsant management, and less often excision of circumscribed foci defined by electrocorticography, can improve cognition. Brain imaging techniques such as PET and functional MRI that index regional metabolic activity yield insights in language dysfunction (557,558), but their routine clinical value is still to be clarified. Molecular cytogenetic analysis can explain the susceptibility to developmental language disorders in a small minority of affected children (559,560).

Language delay with an organic basis resists remediation. Treating the commonly associated impairments of dexterity, attention, and impulse control has little effect on the language problems (561,562). Speech therapy is at its most effective in improving a child’s articulatory skills, although such skills also are apt to improve spontaneously in most children with increasing age. However, the central problem of impoverished vocabulary and syntax and the resulting difficulty in thinking in words is harder to address, and the so-called enrichment techniques are at best marginal or inconsistent in the improvement they can provide. Syndrome-specific remediation may be more effective (563,564). A meta-analysis revealed that speech and language therapy is effective for children with phonologic and expressive language difficulties, mixed in outcome when there are expressive syntactic problems and ineffective for receptive language disorders (564). The parents’ expectations need to be reconciled with the realities of what is often an enduring problem, and the children’s education should be organized to enable them to learn optimally at their admittedly limited rate. Individual instruction is usually required as well as family therapy and behavioral modification on account of secondary emotional disorder, which involves the whole family.

When language skills remain inadequate, augmentation and alternative communication methods (AAC) may
succeed (565). Beyond manual signing as used by the deaf, these methods are applied to children with disordered oral and verbal development caused by cerebral palsy, bulbar palsy, and dysarthria; developmental language disorder; autism; and mental retardation. The need first to verify that intensive speech training does not succeed is balanced against the advantage of starting AAC training early (566), keeping in mind that AAC training may even facilitate verbal production rather than merely replace it. AAC subdivides into unaided (manual signing, finger spelling) and aided (communication board displaying pictures, symbols, or words). Depending on the child’s motor capabilities, the child responds with finger, head, or eye indicating; yes and no indicating; or through electronic devices. A frequently used symbol system is Blissymbolics (567). The additional placement of another’s hand on the child’s hand or arm, combined with an expectation of sophisticated latent communication skills, characterizes the method of facilitated communication (568). This treatment lacks both scientific basis and empirical validity (467,569).

Parents often seek clinical assessment when their child’s performance in school is unexpectedly poor, based on the prevalent but inaccurate assumption that a child who deals intelligently with some topics also should be able to do so with others. Some children do not benefit as much as expected from conventional methods of education in specific subject areas. This failure of response to instruction (RTI) has now itself become one of the allowable defining criteria of learning disability in the 2004 IDEA.

The term learning disability conventionally refers to a developmentally determined impairment in acquisition of certain mental skills that are necessary for academic learning in the early grades. Historically, it was assumed that such skill deficits should be defined by reference to a significant discrepancy between observed low performance on an academic achievement test and significantly higher intellectual level, as estimated by the child’s age or IQ (570,570a). An increasingly influential alternative formulation invokes depressed reading level alone, regardless of IQ (571,572,573,574). The educational prognosis and response to instruction is poorly, if at all, predicted from IQ. Therefore, the requirement for IQ to be significantly higher than achievement has not been empirically supported. In the 2004 reauthorization of IDEA, the federal policy has responded by permitting schools to disregard discrepancy from IQ in their definition of learning disability. Descriptors variously imply a neurologic basis (minimal cerebral dysfunction), a perceptual deficit (word blindness), or the isolated character of the disorder (selective reading disability or dyslexia). However, while a central nervous system role in the expression of the symptoms of dyslexia is certain, a preoccupation with the brain as the basis of the difficulty can lead to investigations that do not help in planning for the child’s future. The isolated character of the deficit implied by the terms selective, specific, or pure also can be misleading, if it precludes a rational search for exactly what each affected child finds particularly difficult. Most dyslexic children have a scattering of other, nonreading disorders—often including ADHD—that are part of the overall constellation of academic difficulties. In particular, ADHD occurs along with dyslexia far more often than expected by chance alone (561,576), and it can have a significant independent role in diminishing long-term school achievement. Selective arithmetic disability, absent a coexisting reading (577) or attentional problem, is uncommon. Selective drawing disabilities, sometimes overinterpreted as “constructional apraxia,” are more common but are much less frequently brought to clinical attention, unless the drawing disability includes poor handwriting and misshaping of letters.

Because social pressure is concentrated on reading, and schools use printed texts as vehicles for all forms of learning, especially as the child grows older, an early reading disability that is left uncorrected can broaden out into a general school and vocational failure. This is further aggravated by coexisting socioeconomic disadvantage. Such generalized failure retains its more specific etiology (see the genetic review below) and must therefore be distinguished from the consequences of global mental retardation, attentional problems, and severe hearing impairment causing severe delays in both reading and math achievement even for those who are skilled users of American Sign Language (578,579).

In addition to ADHD, other psychopathologies, including internalizing ones, are prominent and sometimes serious comorbid accompaniments of dyslexia and related learning disabilities (575). Anxiety is common (580,581,582,583), including phobic reactions. Depression, including suicidality, is the more serious and equally common comorbidity (584,585,586,587,588,589,590), and it is poorly recognized by parents and by teachers (585,591). For that reason, it behooves clinicians to monitor their dyslexic patients, from as early as age 8 or 9, for hitherto unexpressed or unrecognized symptoms of emerging depression. Children are often surprisingly frank to acknowledge depressive feelings to outsiders. Such feelings, coinciding with progressive withdrawal from school or life activities, as well as classic vegetative symptoms such as altered sleep and appetite, warn of potentially significant but treatable depression.

Approximately 11% of school-aged children (aged 6 to 21 years) in the United States are classified as disabled, and
52% (2.4 million) of these students have learning disabilities (592). Another 10% to 20% of all students have learning or behavioral problems that are not severe enough to be classified as disabilities. Other studies estimate that 7% to 8% of early grade-school students have reading disabilities (593).

Models of reading have progressed from the theoretical to the empirical, which has produced a recent consensus account—the National Reading Panel (594). Word decoding ability accounts for much variance in reading skills at all levels (596,597). Reading difficulties are not confined to any particular orthographic system; they are approximately equally prevalent among readers on all continents (598). Although selective reading disability was originally thought to be more frequent in males (599), current evidence favors little or no gender difference in prevalence. However, it remains controversial whether to regard children with reading disability as simply on the low end of a normally distributed function (593) rather than a distinctive group within the childhood population.

More advanced reading also requires fluent successive visual fixation and appropriate direction of eye movement as well as knowledge, explicit or implicit, of the phonologic structure of the language and the ability to construct meaning from the words in their context. Thus, a deficit that limits beginning readers, once overcome, does not prevent further acquisition of reading skill in a normal manner, whereas a deficit that limits advanced reading might not be foreshadowed in the early stages. Failing readers in early and later stages can show different patterns of cognitive deficit. Phonologic and fluency tests particularly describe the earliest stage (kindergarten through second grade); vocabulary increases in importance until the third grade, when it assumes a major role. In some cases, difficulty is limited to spelling (600).

Numerous cognitive components, acting together, represent the larger domain of reading and writing. Each component process is potentially subject to developmental immaturity. Thus, while phonologic, fluency, and vocabulary deficits remain central to dyslexia, slow readers are routinely reported to have numerous limitations. None is necessary or sufficient for dyslexia itself, but rather the overall academic difficulties are compounded. These implicate visual and auditory memory, the ability to store memories in terms of speech sounds and their linkage to visual information, left and right discrimination, auditory synthesis, analysis of words into speech sounds, rhyming judgments, temporal order judgment, fast sequential processing, and nonword reading (601,602,603,604,605). In the long-term, most individuals with reading disabilities face persistent global deficits in academics, underemployment, and problems with behavior, social skills, and emotional adjustment (606).

Many minor neurologic abnormalities have been uncovered in children with learning disabilities, but these seem haphazard in their occurrence. Soft signs of neuromotor immaturity are frequent in these children but do not predict problems in any particular cognitive or academic domain (607). Soft signs are less prevalent after puberty but remain in excess of those found in normally developed adults. MRI indicates a tendency toward decreased volume of the left posterior perisylvian region in children with learning disabilities. Topographic mapping of EEG activity has demonstrated differences between dyslexic and normal boys. These differences are localized not only to the cortical speech areas, but also in both supplementary motor areas (608). Regional cerebral blood flow is reportedly decreased in both parietal lobes (609). Additionally, multiple foci of microdysgenesis of cortical and thalamic architecture have been documented in the brains of dyslexic individuals (610). Widespread in the left hemisphere, they also extend into the right frontal territory. In contrast, CT scans are rarely revealing, even if there are abnormalities on neurologic examination (611). Structural imaging reveals a scattering of abnormalities, consistent for inferior parietal lobule, inferior frontal gyrus and cerebellum, but varying in degree (611a). In short, at none of the three levels of analysis (behavioral, physiologic, and anatomical) is the notion of “pure” dyslexia validated. Rather, a more widespread abnormality is indicated, particularly in dyslexic adults (612) and can be demonstrated by neuropsychologic studies of adult dyslexics (613) as well as by studies that document the various comorbid deficits that rather randomly accompany dyslexia (614). Still, even when adults are defined as selectively reading disabled by their childhood psychometric records, a notable degree of residual specific reading-related deficit is still observable.

Functional neuroimaging has proven unenlightening about the etiology of dyslexia and has yet to be related to the genetic evidence. Nonetheless, the studies are revealing with respect to how the disorder works: for example, by underactivity of the visual-verbal crossmodal association areas, particularly Brodmann’s Area 37 in the vicinity of the left occipitotemporo incisura, and neighboring regions including anterior Area 19 on the ventral cerebrum near the fusiform gyrus (615). More superiorly on the lateral cortex, the angular gyrus has been implicated, sometimes as an abnormality of excess (616), sometimes as one of deficiency (617; see also 618,619,620,621,622). The abnormality of excess may reflect higher level cognitively or semantically based crossmodal associative processes that are being deployed to compensate for lower level perceptual deficits. Conversely, when the function being imaged calls for a higher level of semantic crossmodal processing, the angular gyrus may be underactive. Since the extant functional imaging studies describe a neurological
accompaniment to the behavioral deficit, functional imaging studies also might reflect the improved functioning that results from successful remediation. In a study on adults with psychometrically documented childhood dyslexia, remediation as well as deficit was imaged.

Ultimately dyslexia is an educational, not a medical, problem. Because the diagnosis of a learning disability is predicated on some specificity of deficit (thereby differing from global mild mental retardation), an array of psychometric tests is given to document the preserved and impaired functions. When these tests establish disproportionate difficulty in real word reading (both by phonologic decoding (sounding out) or sight word recognition, reading disability is appropriately diagnosed, either by the psychologist reporting the testing or by the responsible school committee. Since the highest correlations between IQ and reading achievement tests tend to range between 0.5 and 0.75, whereas the correlations between specific reading risk batteries and achievement can range above 0.8, specific skill-related testing is the sine qua non of reading disability diagnosis. Estimates of reading skill so derived are conventionally reported in standardized scores and in percentile ranks. It is when the percentile ranks range downward from about the 15th percentile, and especially when they descend below the tenth, that true disability is commonly diagnosed. However, many authorities recommend a threshold of educational concern at the 30th percentile, reflecting the fact that even children in this range have skill inefficiencies that, unless remediated, could substantially impair their future academic progress (624). If the child’s motivation is inadequate, the reason may be a behavior deviancy that often involves the whole family, an impulsive or hurried approach to the task (as in ADHD), or a cultural alienation from middle-class educational aspirations. The lack of motivation also can derive from poor self-image and a sense of hopelessness secondary to cumulative failure. Psychologic or neuropsychologic assessment is expected to differentiate the motivational or attentional factors from those involving cognitive skill deficit, even when both are present, but the distinction can never be completely assured. Caution in the diagnosis of reading disability is always appropriate, particularly since the vast majority of children with reading disability, especially in the earliest grades, improve with proper remediation (624,625).

For over a century (626), reading disability has been known to be familial, at the least; and explicit models of genetic transmission extend back a full half-century (627). A large number of genetic studies have since confirmed a genetic transmission of risk, and clinicians can be entirely confident that a child has substantially heightened risk for dyslexia if a parent is affected (629,630)—especially so if both parents are affected. Mostly within the last decade and a half, a considerable range of specific genetic linkage and association studies have been done, suggesting multiple genes conferring risk for dyslexia. The best current evidence implicates loci on chromosomes 1, 2, 3, 6, 15, 18, and 21 (631,632,633,634,635). There is no consistent evidence for sex-linked transmission, and correspondingly, population-based studies (636) tend not to show a disproportionate gender prevalence in reading disability. In the genetic as well as the epidemiologic studies, phonologic decoding and its underlying skill of phonemic awareness are routinely implicated. However, the loci on chromosome 6 and 15 may implicate a risk profile that extends to include or even to emphasize fluency or single word reading, respectively (637). Fluency deficit alone is not usually considered an identifiable subtype of reading disability, but when reading problems coexist with fluency deficit, the prognosis is particularly guarded. Fluency is the single best predictor of later outcome for poor readers (615). Single word reading by definition includes some nonphonologic (sight word) skills, so it is possible that there are phenotypic differences across the various genetic loci that have been implicated. The distinctiveness of the phenotype is most prominent in the chromosome 2 studies, which appear to implicate a relatively uncommon but quite significant language disorder syndrome that includes reading disability in its phenotypic presentation (638).

Selective verbal or performance impairments on the Wechsler scales, for example, are neither necessary nor sufficient conditions for reading problems. In children with DLD, the impact on reading—once general language skill is accounted for—is somewhat unpredictable (638a). Children with nonverbal Wechsler deficit are especially unpredictable as to their reading skills. It has been proposed that a nonverbal learning disability explains a variety of symptoms ranging from arithmetic problems to social skills deficit (639). Arithmetic problems commonly occur with reading problems, and clinical experience suggests that their occurrence in isolation is often due only to attention problems. But specific arithmetic disability, apart from ADHD, is well known, with its own subtypes (640; see below). The assumed visuospatial component of some arithmetic disability may be plausibly related to social skills deficit, since social information processing particularly requires global integrative perception, often considered a particularly right hemisphere–related function.

Historically, visuospatial subtypes of reading disability were recognized (641), but these are confounded with comorbid ADHD and are not generally diagnosed in current practice. However, a specific version of this proposal is the subject of intense current research (642,643). Surprisingly, the dorsal stream of visual information, which is
usually associated with visuospatial rather than with linguistic functioning, may be implicated in some aspects of reading disorder. This is perhaps because of the disproportionate involvement of the magnocellular visual system (644,645), or because this pathway specializes in visual “look ahead” anticipatory processing, hence fluency. The evidence does not implicate deficits in this pathway as sufficient to explain the reading disorder however, and is controversial (645a,645b).

The decoding needs not only to be accurate, it needs to be fluent. The ability to automatize retrieval of names is quantitated by the Rapid Naming test (RAN). There is a correlation of about 0.6 between RAN performance and reading rate (646,647). Scores in this test have substantial predictive value for poor readers (648).

Rare in the United States, a presumably left hemisphere–related disorder that implicates arithmetic is the developmental version of the Gerstmann syndrome, in which children make errors in manipulations (such as subtraction) that involve the relative position of digits, and in spelling that involve mistakes of letter sequence rather than letter choice (649). Failure on tests of “finger order sense” is similarly accounted for by difficulty in making use of information about relative position (650). Right-left discrimination difficulties may occur in severe cases, since typically developing children can distinguish between their right and left sides by about 7 years of age.

Another pattern is that of the child, often left-handed, who has a normal psychometric profile, but who occasionally reads or writes from right to left. Although these directional mistakes attract much attention, they are transient and of little consequence. Mirror-image reversals of letters are made by typical children who are not yet ready to learn to read.

As a group, slow readers have frequently been reported to show an excess of left-handedness, mixed-handedness, and inconsistency of side of preference for hand, foot, and eye. Some large series have shown no such relationship, however. In general, samples that yield a high proportion of left-handed subjects come from clinical sources, whereas samples that do not are drawn from the general school population (651). Samples from clinical sources are more likely to include children who have suffered early brain damage, which leads to pathologic left-handedness (489). The association between left-handedness and dyslexia also can encompass a vulnerability to autoimmune disease (652), and poor readers are more likely to have mothers with immune dysfunction (653), but current evidence distinguishes separable genetic bases for dyslexia and immune dysfunction (654). Eye preference, usually inferred from sighting dominance (655), bears no close relationship to hemisphere dominance. It is influenced by minor disparities of visual acuity between the eyes that are not neurologically significant. Anomalous lateral preference is too inconstant to be useful in diagnosis, and there is certainly no rationale for trying to correct a learning disability by interfering with a child’s established limb preference or by guiding a younger child toward right-handedness. The evidence for the relationship between laterality and learning disability is tenuous (656).

Investigating reading disability with structural imaging such as MRI rarely yields useful information. The EEG is helpful only when absence seizures momentarily interrupt the child’s attention often enough to interfere with the child’s ability to follow trains of thought in class. On average, trains of generalized 3 per second spike-wave discharges that last 5.5 seconds or more can interfere with mental activity (see Chapter 14). Brain electrical activity mapping (608), using average evoked potentials, or mapping of brain activity using MEG (magnetoencephalography) reveal interesting group differences but are not suitable for individual diagnosis. This is partly because of the heightened sensitivity of these techniques to a wide variety of other variables that affect the evoked response but are less clearly related to the clinical syndromes themselves. Structural imaging reveals reversed hemisphere asymmetry no more frequently than in controls (657; but see reference 658).

Certain obvious genetic disorders impact reading other than by causing dyslexia. For example, children with Klinefelter syndrome have a language-based learning disability as well as executive dysfunction (659). Children with Duchenne muscular dystrophy have impaired verbal memory spans, which are detrimental to classroom learning (660). More broadly, gross deficits of hearing and vision must be ruled out as contributing to the problem; however, reading is only impaired when visual acuity is substantially reduced. Orthoptic investigation may reveal poor vergence control and stereoacuity, especially in children who report that the text is subjectively unstable (661). Six months of monocular occlusion reportedly results in better reading (662). Vestibular dysfunction is no more common among children with learning disabilities than in a control population (663).

Treatment is at present altogether nonmedical. In order, it focuses first on phonemic awareness, then on matching phonemes to visual letters for decoding of words, then on sight words, and then on semantic aspects of the reading process. For early reading, the combination of phonemic awareness training (664) and direct instruction in alphabetic coding is the most effective and well-documented. Fluency training (665) is gaining a research base as well; vocabulary and comprehension training are earlier in development.

Ineffectual treatments for reading disability include optometric exercises (666), improvement of neuromuscular control (e.g., sensory integration therapy) (667,668),
labyrinthine stimulation, attempts to change peripheral or central laterality (669), psychoactive drug treatment (670), and anti–motion sickness medication (671,672).

Physical training with a view to improving spatial orientation and body image has been claimed to improve mental performance. However, these techniques involve teaching the child accomplishments far removed from the area of the child’s difficulty and giving him or her unaccustomed attention. This encouraging experience could improve self-concept and motivation, without specifically improving reading (673,674). Efforts to change hand preference do not address the basic deficits in learning disability and are ineffective. Eye-movement training ignores the fact that, except in rare cases of conjugate gaze palsies, rate of eye movement shift from fixation to fixation is not a limiting factor in reading skill. Visuomotor training is based on the view that early motor ability influences and predicts later intelligence. No evidence supports the claim that either visual training or perceptual training can improve the academic outcome of children with reading disabilities (675). A motor program based on “patterning” of locomotion (676) is not only ineffective but may be harmful and cause unnecessary expenses, delay in appropriate educational intervention, and engender a false sense of security. In general, the notion that cognitive deficits can be remediated by training ontogenetically earlier perceptuomotor skills has been abandoned.

Noneducational remedial methods have not been shown to have specific benefits for reading readiness. A systematic, analytic approach to reading itself is preferable, in which the information is presented stepwise, while distractions are minimized. Usually, this can only be achieved by individual instruction, to conform to the individual learning requirements of the child. “Language experience” approaches are generally unsuccessful.

To meet these requirements, some educators “teach to weakness” in the hopes of improving the efficiency of the process responsible for reading deficiency; others regard it as more realistic to circumvent (“bypass”) the deficiency by teaching to residual strengths. In fact, neither position is supportable in isolation. Initial attempts must be to teach to weakness, since all but a very few children can and do learn the skills of reading if given the appropriate additional instruction and practice in the very things (phoneme-grapheme correspondence, for example) they initially find difficult. In this regard, the formal randomized prospective trials (677,678,679) are unambiguous, and there is therefore no rationale for withholding such instruction and practice from children who need it. Such instruction can, however, take time that in severe cases is measured in years, and the goal of education during this period must also be to develop the child’s strengths. It is only rarely necessary or even helpful to remove the child totally from the regular classroom, though in severe cases such placements in “self-contained classrooms” of children with similar levels of need do maximize the instructional opportunity and can promote peer relationships with other similarly struggling students. There is the countervailing risk of negative self attributions by the child, who may complain that he or she has been put in the “dummy class.” When otherwise indicated by the educators on the scene (who routinely avoid these placements when they can for cost reasons), children can usually be counseled and helped to focus on the prospect of improved reading and learning skills—which they will often value above the discomfort of the placement. Such counsel, of course, implies that the clinician also will follow the child’s progress and expect evidence of improved learning from the schools, thereby to encourage the child and facilitate re-placement in a regular classroom when appropriate. In this context, the rare child who needs continued placement in a special class will become obvious to all parties. The more common case is that of a child needing individual or small-group services from specialist staff at the school, outside the main classroom. Sometimes termed pull out services, these occasions usually total at most a few hours a week. At times, children object to the adverse inference that they believe is implied by such services, but the prospect of improving their reading skills will override their social reservations.

Beyond identifying the child who needs individualized reading instruction, some educators attempt to subtype the reading disability and deploy special treatments for special needs ranging from visual recognition problems to auditory discrimination problems. If the diagnoses are about different degrees of phonologic, fluency, vocabulary, and sight word and text comprehension skills, then they can be useful in guiding efficiently individualized instruction in small groups and sometimes individually. However, suitably controlled randomized prospective efficacy evidence is not available to support allegedly specific but theoretically imprecise forms of sensory training for conditions such as “visual learner,” “left hemisphere learner,” or even “auditory discrimination deficit” (unless it involves actual deafness that prevents hearing of normal spoken discourse). The teaching of reading, like that of any other skill, has no shortcuts: The relevant underlying processes (phonemic awareness, phonologic letter-sound coding including spelling, fluent letter-sound-word associations, vocabulary, and text comprehension strategies) cannot be avoided and must not be replaced by expensive and unproven quick remedies. The state of the art in education now mirrors that in medicine: proper prospective randomized controlled studies of efficacy are required before new treatments can be endorsed. A recent review of the extant evidence is provided by the National Reading Panel Report (594).

Finally, lexical knowledge does not guarantee fluency. The child who has mastered phonics may nonetheless emerge as a halting, dysfunctional reader. Fluency training
is a necessary sequel. Different children are led through these stages at different rates, encountering different problems on the way. This exacting, long-term effort is not substantially facilitated by any form of diagnostic testing at the time of the original assessment.

Socioeconomic circumstances ultimately exert the most influence on success in learning to read (595). Holding these constant, preschool prediction of reading failure remains difficult, though less so by the second half of kindergarten (596). The predictive success of reading readiness tests naturally increases as early reading achievement itself—for example, knowledge of the alphabet—develops. Predictions from first grade forward therefore tend to be stronger than those predicted from kindergarten. Readiness testing is best performed by entry to first grade, soon enough to enable any necessary modifications in the classroom. Longitudinal studies indicate that learning disability persists into adulthood and even broadens into vocational and personal problems, and occasionally delinquency and depression, more often than generally realized.

Children who are so clumsy that they have difficulty coping with the demands of everyday tasks (such as dressing, feeding, writing, gym and playground activities) have been variously characterized as being physically awkward, developmentally dyspraxic, or subject to sensory integrative or perceptual motor dysfunction—the clumsy child syndrome (705). Named developmental coordination disorder (DCD) both in the DSM-IV and ICD-10, it was defined as existing when the child lacks the motor coordination necessary to perform age appropriate tasks, although intellectually normal and otherwise neurologically intact.
Previously believed to represent a temporary immaturity or lag that resolves with increasing age, it is now thought to persist at least into adulthood and probably to be lifelong, bringing in its wake health and fitness problems, disordered behavior, low self-esteem, and deficient social interactions (706). DCD is thought to be present in 5% to 6% of schoolchildren (707).

Movements are generally slow; DCD may involve timing difficulties related to dysfunction of the cerebellum. The child finds it particularly hard to inhibit moving toward salient external stimuli (708). Visuospatial deficits have been documented even in the absence of movement. When it is found in combination with ADHD, DCD persists into the adult years (709). The coordination deficits in ADHD children could not be attributed only to inattention (710). The poorly legible writing and copying of text found in writing disorder that is due to DCD differs from that in dysgraphia based on linguistic problems, in which spelling is poor but copying is preserved (711). When diverse physical interventions to remediate DCD are critically evaluated, it appears that any gains are slight and more probably accounted for by increased self-confidence than improvement in motor control (712).

Attention-deficit hyperactivity disorder (ADHD) refers to the covariation of inattention, hyperactivity, and impulsivity. The DSM-IV describes three ADHD subtypes: (a) inattention alone, (b) hyperactivity-impulsivity alone, and (c) a combined type with significant inattention, hyperactivity, and impulsivity. Some 80% have the combined type, 15% the inattentive, and 5% hyperactivity and impulsivity only. Children may be restless during one stage of development but not subsequently. This two-factor model is a departure from the previous unidimensional model, and though it is far from validated, it does have some empirical basis (713). Correspondingly, hyperactivity/impulsivity is associated with externalizing psychopathology (e.g., conduct disorder, oppositional defiant disorder) and the inattentive subtype with internalizing psychopathology (e.g., anxiety, depression) (714). The DSM-IV does not include developmental qualifiers among its criteria for ADHD, but the listed behaviors, and particularly hyperactivity, typically decrease as the child grows older. This does not imply that the disorder grows milder with increasing age, but that its form of presentation changes during development. A synopsis of the current state of ADHD diagnosis and treatment is available (715).

According to the DSM-IV-TR (text revision), the major clinical manifestations of ADHD—namely developmentally maladaptive attention, activity, and impulse control—must be sufficient to cause impairment in social, academic, or occupational functioning (48). Signs of ADHD generally begin before age 7 years and should persist for at least 6 months in two or more settings (e.g., home, school, or play) before the diagnosis is made. However, some children may have a later onset of the signs of ADHD (716).

There is consensus on the behavioral characteristics of ADHD that are listed in DSM-IV criteria. The most frequently reported primary manifestations of ADHD include cognitive disorganization, distractibility, inattention, impulsivity, and hyperactivity. The most commonly reported secondary manifestations include disruptive behaviors, poor social skills, emotional immaturity, fidgeting, poor academic performance, and excessive talking.

The relationship between the level of motor activity in infants and subsequent hyperactivity is unknown. Mothers often observe in retrospect that their child with hyperactivity was unusually active from birth. Feeding difficulty and sleep disturbances in infancy and the preschool period are commonly reported precursors (717,718). A prospective study found the second year of life to be the earliest in which ADHD symptoms could be detected and age 3 to 5 years to be the peak time of onset (719). Young children with hyperactivity explore their environment with unusual persistence, which accounts for the increased frequency of accidental poisonings (720) and traumatic brain injury (721) in ADHD. In older children, gross motor exuberance can give way to a less flagrant but continual wriggling restlessness. More troublesome is the child’s impulsive, distractible, and often antisocial behavior. The hyperactivity becomes less a source of adults’ irritation at home and more an occasion for disapproval from teachers at times when schoolchildren are expected to sit still and pay attention. This disapproval may be partly responsible for the high stealing and truancy rates among hyperactive children. Girls and boys with ADHD are similar in their impulsivity, academic performance, social functioning, fine motor skills, parental education, and parental depression (722). Girls tend to be less hyperactive and externalizing, but more inattentive and cognitively impaired and more subject to rejection by peers (722,723).

Gross body movements are not uniformly more frequent in children with hyperactivity. In unstructured situations, many typical children are just as active, and children with hyperactivity do not stand out. In structured situations, differences appear both in the amount and the relevance of the activity. In a 24-hour study, an ADHD group did exhibit more overall activity than controls during the night and throughout the day (724). Recorded over as long as 1 week, boys with ADHD were significantly more restless than controls at ages 6 to 11 years, but rarely so at older ages. Some children manage to continue to be attentive (e.g., to a television program) while moving around a room. Mostly, however, each movement signals a more
comprehensive reorientation of mental set. In infancy, and until the child walks, this continual reorientation appears not to cause concern. Subsequently, complaints gather as the child disrupts the orderly home or classroom while failing to pay the expected degree of attention to the teacher, the instructional materials, and homework assignments.

Young children with ADHD initially show no signs of distress, but negative reactions from adults and peers gradually engender feelings of inadequacy, and this can lead to withdrawal or aggression. Hyperactive children are seen as quarrelsome, irritable, defiant, untruthful, and destructive by their classmates as well as by adults (725). The discipline imposed by the school and the need to repeat grades further contribute to learning failure and social maladjustment. Nonhyperactive children with attention deficit are misperceived as undermotivated or lazy, and the diagnosis is often missed.

Schoolwork demands increasing effort and organization as of fourth grade (726), and the intermittency of attention interferes more with learning and becomes more troublesome to teachers. Some 10% to 50% of children with ADHD underachieve in reading (727,728). Whether this is due to the attention deficit, which can impair word attack (729), or to comorbid dyslexia is hard to determine in the individual case (730). Children with ADHD are not lower in IQ than attentionally normal children.

Neither restlessness nor inattention is the core of the ADHD syndrome. That ADHD is primarily a disinhibitory disorder (731) is questionable (732). A recent study failed to confirm a selective executive/inhibitory deficit in ADHD boys (733). There is experimental evidence that ADHD behavior is more variable than that of normal children and of children under stimulant control (734,735). Hyperactive children attend as efficiently as normal children to tasks that they find intrinsically interesting or rewarding. It is when the task is tedious and the incentive for doing it is remote in the future that their inattention is maximal. The hyperactive child understands consequences and professes the usual concern about them. However, the consequences make little impression on the child’s behavior. But if the task is attractive or of interest to the child (i.e., intrinsically motivating), then the ADHD child performs as well as a child who is attentionally normal (736).

There are no definitive diagnostic tests for ADHD. Rather, the diagnosis depends on determining (a) that the child is inattentive and/or impulsive and hyperactive to a degree that is excessive for the child’s age and expected level of developmental maturity, and (b) that there is consequent impairment in social and school functioning. Corroborating information can be obtained through child, parental, and teacher interviews; standardized rating scales (737); direct observation; behavioral laboratory tests; and formal psychologic and educational evaluations. The child’s history, physical, and neurologic examination should be complete enough to rule out any identifiable genetic or medical conditions that might simulate ADHD. Absent such conditions, and although it may uncover nonspecific soft signs, the neurological examination rarely contributes to the diagnosis. Neither physiological nor biochemical measures reliably identify ADHD children (738). Psychiatric consultation is indicated in children with more complex behavioral problems or who have comorbid depression or thought disorders. It is not known whether the three phenotypes presented in the DSM-IV-TR have different neurobiologic bases or are variants of the same underlying pathophysiology.

ADHD is the most common chronic behavioral disorder in children, with a prevalence that ranges between 3% and 5% of school-aged children worldwide (739). It is reported to be highest in inner city populations (740,741). It occurs far more frequently in boys than girls in clinic samples (742) as well as in approximately 3 to 1 ratio in population studies (722). Through the evolution of the diagnostic criteria and changes in the public’s attitude toward the disorder, the prevalence of ADHD has gradually increased, particularly in girls, as has the use of stimulant therapy (743). In a more recent study, primary care providers identified behavioral problems in approximately one in five children, and attentional/hyperactivity problems in 9.2% of the entire study sample, but the diagnosis was commonly made solely by parental interview and without the benefit of standardized questionnaires (742).

Although it is most frequently diagnosed in North America, crosscultural studies have shown comparable levels of prevalence in all countries studied, even in the Third World (739), not only of ADHD, but also of internalizing and externalizing psychopathology in general (744). In Britain, the diagnostic label hyperkinetic disorder (745) is reserved for the most severe cases. Other children who would have been considered to have ADHD in the United States are termed conduct disordered in the United Kingdom (746). This difference is not only semantic. It reflects a difference in management, being that stimulant therapy is used sparingly in the United Kingdom. Whether ADHD and conduct disorder are truly separable is questionable (747). Those who use aggressive conduct disorder as a primary diagnosis find a majority of their patients to be hyperactive as well (747). Comorbidities affect both prognosis and treatment (748,749). In particular, aggressiveness in early childhood is a negative prognostic sign. The home environment plays a role. Aggressive children as a group tend to be ones reared by depressed mothers who offer less positive caregiving. Family income is lower than for unaggressive
children, and harsh parenting is more likely. Aggressive children continue to have more academic and social difficulty, and are seen as being angry, hostile, antisocial, and oppositional (750).

There is no single underlying cause of ADHD. It occurs both in the absence of any identifiable risk factors, and in association with numerous other childhood conditions such as motor dyspraxia, tics, learning problems, speech and language disorders, sleep disorders, oppositional behavior, enuresis, and encopresis. In addition, overactive and socially disruptive behavior is common in children who have evidence of injury from infections, head trauma (751), toxic exposures (752), and extreme prematurity (753,754). Minor physical anomalies are frequent (755).

In most cases, hyperactivity is already evident in infancy, although it is not necessarily recognized as such, and it frequently persists into adult life in the form of attentional and social problems (756). Thus, it has the characteristics of a stable personality trait or temperament, and for which there is a genetic link in most cases. Families of ADHD/conduct-disordered children are notable for the high incidence of sociopathy and alcoholism (757,758). ADHD is highly heritable, probably based on multiple genes, each with small effect size. Parents of children with ADHD have elevated ratings of ADHD symptoms (759). Compared with the general population, first-degree relatives are at a 4.6- to 7.6-fold risk of developing the disorder (760). Second-degree relatives also are at increased risk (761). Several linkage studies implicate a 7-repeat polymorphism of the dopamine receptor D4 (DRD4) gene, mapped to chromosome 11p 15.5 (762), and also a polymorphism of a dopamine transporter (DAT1) (763). The transporter is a target of many of the drugs used to treat ADHD (764). The density of dopamine transporters in the striatum is markedly increased, as measured by single photon emission computed tomography (SPECT) using labeled altropane, a specific ligand for dopamine transporter (765). This disorder in dopaminergic function results in a disinhibited sensation-seeking behavioral style (766).

Right frontostriatal circuitry has been invoked with respect to defective response inhibition. The right frontal lobe is reportedly smaller than normal in ADHD structural imaging (767), and the striatum has abnormal morphology (768,769,770). The cerebellar, temporal gray matter and total cerebral volume are smaller in ADHD individuals than attentionally normal controls, and these size variables correlate significantly with ratings of severity of ADHD behaviors (771). Corpus callosum size is reduced (772,773) as well as the inferior cerebellar vermis, posterior lobe (773a). Regional cerebral blood flow is diminished in striatum and frontal lobes in ADHD, a deficiency that is partly reversible by methylphenidate (774, but see 775). Power spectrum analysis indicates prefrontal underactivation (776) and cerebral glucose metabolism is diminished generally, but most notably, in the prefrontal cortex (777). Some neuropsychologic deficits are consistent with prefrontal underactivity. The quantified EEG in ADHD has abnormal features Lazzaro and colleagues, 1998. Event-related potential (ERP) studies suggest the presence of abnormalities in the right frontal (779) and parietal (780) regions. Some neurobiological clues to the etiology of ADHD are summarized in Table 18.11.

Restlessness and short attention span can be features of mental retardation and may conform to the child’s developmental quotient. The presence of mental retardation can itself be hard to verify. If the validity of the intelligence testing is uncertain, the educational psychologist may qualify the IQ score with a warning that it could under-represent the child’s cognitive potential, particularly for measures that require sustained attention rather than merely a quick response. Retesting after initiation of stimulant therapy can yield a more realistic estimate. Improvement can be conveniently documented by teachers and parents by means of the Conners rating scales (781). However, even correcting the attentional problem can leave the educational difficulty unresolved, given the 30% to 40% overlap between ADHD and learning disabilities (782).

When cognitive deficits do not explain the attentional problem, an alternative possibility is overanxious disorder. The anxious child is most restless under emotional stress, as in the classroom and consulting room, and least restless when unstressed, as on weekends or on vacation. The child is visibly unhappy, either characterologically or because of some maladjustment within the family. The hyperactive child, on the other hand, is typically in good spirits until he or she comes up against frustration and adverse attitudes from others. Whereas the time of onset of anxiety can usually be specified, hyperactivity often dates from early infancy. A favorable response to a stimulant supports the diagnosis of ADHD as opposed to anxiety. Anxiety is comorbid with ADHD in about one-third of cases. Its presence did not impair stimulant response in a large-scale study (783). Favorable response can be documented in the laboratory on a time-response and dose-response basis by use of a paired-associated learning task (784). The findings contribute to the design of an appropriate drug regimen for the child who responds favorably. Other tasks that can be used in a similar way are the Continuous Performance Test (CPT) (785) and the Test of Variables of Attention (TOVA) (786). CPT performance variables are significantly related to ADHD symptoms (787).

Overfocusing (326) is sometimes comorbid with ADHD and sometimes distinct. Children with overfocusing have difficulty in shifting mental set associated with perseveration and isolating tendencies. In addition to the social
withdrawal that is often manifest, overfocusing classically includes aversion to novelty, complexity, and time pressure. It bears some resemblance to high-level autistic behavior, though it is milder (787a). A contrasting temperament of uncertain relation to overfocusing is exhibited by the behaviorally inhibited child (788). Irritable in infancy; shy and fearful as a toddler; and cautious, quiet, and introverted in grade school, these children are thought to have increased limbic-sympathetic tone and to be anxiety prone. Disordered sleep, especially caused by obstructive sleep apnea, can render a child inattentive and restless. Attention deficits are often found among children of schizophrenic parents, who are themselves at high risk for developing schizophrenia. They perform poorly on vigilance tasks but are not impulsive (789). Another potentially overlooked differential diagnosis of ADHD is occult mood disorder, notably bipolar disorder, particularly when it presents in adolescents with chronic rather than episodic symptomatology. The combination of dysphoric and irritable mood with aggressive conduct that is resistant to stimulant therapy as well as subjective report of elation, grandiosity, racing thoughts, flight of ideas, and decreased need for sleep (790) should prompt an examination of the family for bipolar disorder (791). However, a broad phenotype also is recognized: Children exhibit chronic nonepisodic disregulation, with severe irritability, hyperarousal, and increased reactivity to negative emotional stimuli (792). In one study of juvenile bipolar disorder, comorbidity with ADHD was 87% (793). Episodic loss of control is attributed to dysfunction in the limbic cortex (794). Whereas impulsive acts of aggression are common in ADHD, episodic explosive behavior is not. The restless movements of Sydenhams chorea are associated with ADHD in one-third of cases and are often accompanied by emotional lability and
obsessive-compulsive symptoms (795). In addition to multiple motor and vocal tics, children with Tourette syndrome (TS) often exhibit symptoms that are diagnostic of ADHD. However, these symptoms represent comorbidity and are not inherent in TS. Learning disabilities and neuropsychologic deficits (796) have been reported in children with TS. With respect to executive functions, children with TS but without comorbid ADHD were impaired only in tests that call for response inhibition (797). The neurology of TS is considered in Chapter 3.

Whereas the overt restlessness of hyperactive children diminishes in adolescence, their impulsiveness and emotional lability usually persist, with a correspondingly mixed prognosis for long-term adaptive outcome. Long-term follow-up of children with ADHD shows an unfavorable outcome in many but not all. Adults who had ADHD in childhood often continue to show functional impairment (861). When aggressiveness is a feature, it particularly tends to persist (862) and appears to bear some association with early onset alcoholism (863). Impulsive aggressiveness is associated with low concentration of the serotonin metabolite 5-HIAA in cerebrospinal fluid (864). Thus, the ADHD prognosis is intermediate between that
of control patients and the graver outlook for children with frank psychiatric disorder in childhood. Schizophrenia is not a major ADHD outcome. However, children of schizophrenic mothers, at high risk for adult schizophrenia, have been found to be prone to attentional dysfunction and poor social competence (789,865).

Attempts to relate the long-term prognosis of ADHD to psychoactive drug therapy have been so poorly controlled as to be uninterpretable. In the medium term, stimulant therapy is clearly beneficial for educational progress.