Early Studies

Douglas⁶ reported on 163 infants with birth weights of 2000 g or less born in the United Kingdom during a single week in 1946. Some of the babies were born at home and some in the hospital. Of those cared for in a hospital, 18 received oxygen, and 11 were in incubators. None of the infants weighing less than 1000 g, whether born at home or in the hospital, survived; only 32% of the infants weighing 1001 to 1500 g lived. Of the infants weighing 1500 to 2000 g who did survive, none had handicaps. Of the infants weighing less than 1500 who survived, 17% had significant physical, neurological, mental, or behavioral problems. In 1958, Dann and associates⁷ described the outcomes for 73 of 116 infants born in the New York City area between 1940 and 1952 with birth weights of 1000 g or less or whose weight dropped below 1000 g during their hospitalization. The infants were kept in incubators, and most received oxygen and meticulous but nonintrusive medical support. The children were evaluated between 1950 and 1957. All 73 studied, who were among the 116 survivors, were found to have generally good physical health with few neurological defects. Most had achieved normal height, but often not until after 4 years of age. However, the IQs of 84%, while in the average range, were below those of their full-term siblings. Sixteen percent had IQs below 80. After considering variables such as birth weight, gender, race, and socioeconomic status, Dann and associates found that the infants with the highest IQs were from families with higher socioeconomic status.

Both of these studies are unique in that they preceded, by approximately two decades, the establishment of modern neonatal intensive care. As a result, they provide a historical perspective and also demonstrate that even without neonatal intensive care, some infants with low birth weight did survive and did well. With the introduction of new methods of care, survival increased and outcomes improved, although other issues have emerged.^61–63 In the next section of this chapter, we review later follow-up studies on the infants with VLBW and those with ELBW.

Studies from 1979 to the Early 1980s

Studies published after 1979^1,64–70 documented the results of the emergence of the modern age of neonatology and the progress in the evolution of care for the infant with VLBW. With technological advances and recognition of the importance of continuous and comprehensive assessment of outcomes, findings extend beyond morbidity, mortality, and medical issues to include such issues as the psychosocial, neurodevelopmental, educational, and behavioral sequelae of premature birth for children and their families.

Studies in the Late 1980s

In 1989, Hack and Fanaroff¹ reported on the outcome of infants with birth weights lower than 750 g born between 1982 and 1988. Ninety-eight infants were born between July 1982 and June 1985 (period 1), and 120 infants were born between July 1985 and June 1988 (period 2). There was some increase in survival from period 1 to period 2 among infants with gestational ages between 25 and 27 weeks (52% vs. 71%), but the overall rates of neonatal morbidity in the two groups were similar. The neurodevelopmental outcomes were also similar. Period 1 children had Bayley motor and mental scores of 90 ± 17 and 88 ± 14, respectively, at 20 months of corrected age. The period 2 children were seen at 8 months of corrected age and had motor and mental scores of 77 ± 25 and 81 ± 30. There was more aggressive intervention with the period 2 children who had many complications, including bronchopulmonary dysplasia, septicemia, retinopathy of prematurity, intraventricular hemorrhage, and deficits in neurodevelopmental function.

O’Callaghan and coworkers⁷⁸ reported on the 2-year outcome of 63 children with ELBW born between 1988 and 1990 and cared for in a neonatal intensive care unit. Findings provide some insight into how more recent cohorts of children with ELBW may be functioning at 2 years of age. Investigators compared the children to full-term matched controls by using a cognitive function measure, a neurosensory motor developmental assessment, and a medical assessment. Furthermore, they studied these children as a whole group and as a subset, a low-risk group, which included children with no intracranial hemorrhage, periventricular leukomalacia, or chronic lung disease (i.e., bronchopulmonary dysplasia). The interesting findings very much mirrored those of earlier studies. The total ELBW group differed significantly from the control group (children born at term) with regard to cognitive and personal-social functioning, although they scored in the average range. The low-risk ELBW group did not differ from the control group. There were more striking differences with review of the neurosensory motor findings. Both the total ELBW group and the low-risk ELBW group had poorer total scores than did the control group, as well as poorer gross and fine motor subscale scores.

In the past, there was interest primarily in the IQs of children born very early. Most investigators assessed early neurodevelopmental functioning and found that, as a group, the infants with VLBW did not do as well as the older and heavier premature infants or a matched control group of infants born at term. However, investigators are now able to assess more subtle aspects of central nervous system function that have an effect on cognitive functioning.

Herrgaard and colleagues⁷⁹ undertook a 5-year neurodevelopmental assessment of 60 children born before 32 weeks of gestation. These children were matched with 60 full-term controls. Assessment tools used included a standardized neurological examination, a neuropsychological assessment, an audiological examination, and an ophthalmological examination. Included in the preterm group were children thought to be handicapped (children with cerebral palsy, mental retardation [IQ < 70], bilateral hearing loss, visual impairment, and epilepsy) and those not disabled. With regard to IQ, there were significant differences between the entire preterm group and the control group, as well as significant differences between the handicapped and nonhandicapped preterm groups. The control group had the highest IQs, the nonhandicapped preterm group had lower IQs, and the handicapped group had the lowest IQs. The neurodevelopmental profile was composed of eight functional entities: gross motor, fine motor, visual-motor, attention, language, visual-spatial, sensorimotor, and memory skills. The investigators noted several interesting findings. First, all of the children born preterm had difficulty with gross, fine, and visual-motor skills. They also had difficulty with language, sensorimotor, visual-spatial, and memory skills. Second, the nonhandicapped children with minor neurodevelopmental difficulties had a similar spectrum of problems, although their IQs were in the average range, with some even in the exceptional range.

These findings are similar to those of Sostek⁸⁰ in her study of children born before 33 weeks of gestational age and with a mean birth weight of 1358 g, in comparison with children born at term. None of the premature children had lung disease, intracranial hemorrhage, or other medical problems. Although these children had normal IQs, they were compromised with regard to perceptual-motor integration and recognition, perceptual performance tasks, quantitative tasks, memory, and visual-motor skill and were found to be more distractible and to have poorer attention and less readiness for kindergarten than were full-term controls. These findings emphasize the importance of assessing neurodevelopmental profiles, rather than relying on global measures of intelligence.

Teplin and associates⁸¹ assessed the neurodevelopmental, health, and growth status at 6 years of age in 28 children with birth weights lower than 1001 g. In comparison with 26 control children born at term, the children with ELBW had significantly more mild or moderate-to-severe neurological problems (61% vs. 23%) including cerebral palsy; abnormalities of muscle tone; and immaturities of balance, speech, and articulation. In cognitive function, the controls scored significantly higher than the children with ELBW. However, more than half of the children with ELBW with normal IQs had mildly abnormal neurological findings, whereas the controls with normal IQs had normal neurological findings. When they determined the overall functional status, the investigators found that 46% of children with ELBW were normal, 36% were mildly disabled, and 18% were moderately to severely disabled; in comparison, 75% of the controls were normal, only 4% were significantly disabled, and the remainder had some mild degree of abnormality. In contrast to other reports, attentional disturbances were not a problem for the preterm groups described in these two studies.^82,83

Halsey and colleagues⁸⁴ conducted another provocative and important study on children with VLBW when they were in preschool. They studied 60 white, middle-class children with VLBW and compared them with a matched peer group. They used a general developmental scale and a scale of visual-motor integration. They found that the VLBW group’s mean scores were significantly lower than those of the controls, although they were still within one standard deviation of the mean. Of the children with VLBW, 23% were clearly disabled, 51% obtained borderline scores, and 26% were average. The control group had cognitive scores 15 to 18 points higher than those of the VLBW group and were 2.5 times more likely to have normal development. The authors were reluctant to make any predictions on the basis of these data but expressed concern that this pattern of performance placed the children with VLBW at higher risk for later difficulties. A subsequent study, to be discussed later,⁹⁰ confirmed that these data are indeed predictive of later difficulties. Thus, follow-up studies suggest that premature infants with VLBW, despite relatively intact cognitive skills as evidenced by normal IQs, appear to have neuropsychological and neuromotor disturbances that can adversely affect their school performance, self-esteem, and behavior.

The studies of the 1980s, in which children were monitored to only the preschool years, reinforced the concerns of the 1970s. Although investigators did note an increase in survival and found that many children did well, they also noted that these children were functioning at lower levels than their peers and with a variety of neurodevelopmental problems that earlier studies had not identified. These problems included deficiencies in their perceptual skills, social skills, and level of maturity.

We have thus far reviewed reports on infants with VLBW evaluated after only 2 to 6 years. However, among the most important indicators of successful outcome are the child’s social-emotional adaptation and how well the child does in school. Studies have acknowledged that many infants with VLBW have significant difficulties that persist throughout their lives. Although such children may have IQs in the average range, they do not perform as well as controls on measures of fine and gross motor and visual-motor tasks and display so-called “minor disabilities” that become more apparent in school. An important question, then, is what effect these difficulties have on school performance and peer relationships. Eilers and associates⁸⁵ studied a group of children with birth weights of 1250 g or lower who were born between July 1974 and July 1978. There were 43 survivors, 33 of whom were studied at 5 to 8 years of age. Of the 33 children, 16 were functioning at an age-appropriate level, 3 had major handicaps, and 14 were in regular classes but needed remedial help. The authors noted that 51.5% of this group required special education support, in comparison with 21.4% of the general school population.

Vohr and Garcia Coll⁸⁶ reported on a 7-year longitudinal study of children with birth weights lower than 1500 g who were born in 1975. Of their original population, 62 (51.2%) survived, and 42 (67%) were monitored. The investigators evaluated patterns of neurological and developmental functioning at 1 year of age and compared them with normal functioning children at age 7. Using a classification of “normal,” “suspect,” and “abnormal,” they found that the patterns at 1 year were significantly related to those at 7 years and that 54% of the total sample required special education or resource help at 7 years. Furthermore, those who had abnormal findings at 1 year were most likely to have difficulties at 7 years. This was less clear for the groups with suspect and normal functioning. Based on their identification at age 1 year, 27% of the children with normal patterns, 50% of the children with suspect patterns, and 87% of the children with abnormal patterns required special educational services by age 7. The investigators also noted that 45% of the children with normal patterns, 75% of those with suspect patterns, and 100% of those with abnormal patterns had visual-motor disturbances.

Another study⁸⁷ revealed that even among a relatively normal group of children with birth weights of 1500 g or lower, there was an increased incidence of visual-motor problems. Klein and coworkers⁸³ found that a group of children with VLBW scored lower at 9 years of age on tests measuring general intelligence, visual or spatial skills, and academic achievement than did full-term controls. Klein and coworkers found that a subset of children with VLBW but normal IQs showed significant deficits in mathematics skills. Crowe and associates⁸⁸ reported on 90 children born between 24 and 36 weeks of gestation who participated in a longitudinal follow-up program; children with such major neurological impairments as cerebral palsy were excluded from study. Crowe and associates found that motor development at 4½ years of corrected age was relatively intact, but children with birth weights of about 1000 g displayed significantly poorer motor skills. Moreover, such children with symptomatic intracranial hemorrhage also had significantly poorer motor performance.

Saigal and associates⁸² conducted a longitudinal, regionally based study over many years and reported on the cognitive and school abilities at 8 years of a cohort of relatively socioeconomically advantaged infants with birth weights of 501 to 1000 g who were born between 1977 and 1981. The investigators compared the children’s intellectual, motor, visual-motor, and adaptive capabilities and their teachers’ perceptions to those of a matched group of children born at term. They found that the majority of children with ELBW had IQs in the normal range but significantly lower than those of the controls. This was true even when handicapped children were excluded from the analysis. Moreover, the ELBW group was significantly disadvantaged on every measure. Furthermore, the teachers rated the ELBW group as performing below grade level. Interestingly, neurologically normal children also performed below the normal range on tests of visual-motor and motor abilities.

Hack and coworkers⁸⁹ reported on the 8-year neurocognitive abilities of a group of 249 infants with VLBW born between 1977 and 1979, in comparison with 363 randomly selected normal children born at the same time. The investigators administered a neurological examination and tests of intelligence, language, speech, reading, mathematics, spelling, visual and fine motor abilities, and behavior. Twenty-four (10%) of the children with VLBW had a major neurological abnormality. None of the controls had such a finding. With the exception of speech and total behavior scores, the VLBW group scored significantly more poorly than did the controls on all tests. Even neurologically intact children with VLBW but normal IQs had significantly poorer scores than did the controls in expressive language, memory, visual-motor function, fine motor function, and measures of hyperactivity. When the investigators controlled for social risk as a significant determinant of poor outcome, VLBW still had an adverse affect on functioning, with the exception of verbal IQ. The investigators concluded that prematurity may contribute only minimally to the negative effect of a poor psychosocial environment in this area. In contrast, biological factors may have a greater effect on the deficits of more advantaged children, in comparison with their peers.

In a more recent study, Hille and associates⁹⁰ assessed the school performance at 9 years of age of children with VLBW born in the Netherlands. They were able to gather data on 84% (N = 813) of the survivors from an almost complete birth cohort at 9 years of age. Nineteen percent were in special education programs, half of whom had been placed since 5 years of age for identified problems. Of the children with VLBW in mainstream classes, 32% were in a grade below their age level, and another 38% required special assistance. Of the children who were retained, 60% required special assistance, in comparison with 28% of children in an age-appropriate grade. The authors identified a number of factors at 5 years of age that were predictive of school difficulties at 9 years. These included developmental delays, speech and language delay, behavioral problems, and low socioeconomic status, which confirmed the findings of Hack and coworkers⁸⁹ and Halsey and colleagues.⁸⁴

A final issue to consider with this group of children is the possible effect of VLBW on behavior. We noted previously that many of these children have significant problems with hyperactivity and attention. Weisglas-Kuperus and colleagues⁹¹ addressed the issue of behavior problems in this group of children. In a study of 73 children with VLBW who were compared with 192 full-term children at 3½ years of age, the authors found a significant degree of behavioral disturbance in the VLBW group. Problems included depression and internalizing difficulties.

Studies in the 1990s

Follow-up studies proliferated throughout the 1990s. These studies focused not only on the improved survival of the infant with VLBW as a result of further technological advances but also, of even more importance, on developmental, cognitive, academic, and social outcomes. Investigators noted the improved survival rates among infants with VLBW, as well as among infants with ELBW, during this period, in comparison with the 1970s and 1980s. As a group, the studies addressed cognition (IQ), academic performance, behavioral issues and social competence, health, language development, and visual and fine motor capacities. These follow-up studies were of longer term than previous studies, extending up to 11 to 13 years of age. As in the 1980s, the investigators divided the infants into groups of those with birth weight lower than 750 g, 750 to 1000 g, 1001 to 1499 g, and 1500 to 2500 g and used as controls children with birth weights higher than 2500 g, who were also matched for gender and socioeconomic status.

The prenatal and perinatal factors with the greatest effect on outcome included birth weight, gestational age, whether the infant was born in or outside of a special care center, and the nature and degree of the complications of premature birth. These complications included chronic lung disease and the need for oxygen, the presence of intraventricular hemorrhage and its complications, and the presence of seizures. Of note, many of these children had significant infections and gastroenterological problems, including necrotizing enterocolitis and undernutrition. In addition, many of these children had recurrent ear infections, which often necessitated myringotomy and tubes retinopathy of prematurity. The lighter and more immature the infant, the more prevalent were complications and so the higher was the risk for a more adverse outcome. Thus, the smallest infants who survived, those with birth weight lower than 750 g, had the worst outcomes. As a group, they had an increased incidence of cerebral palsy, mental retardation, autism, attention-deficit/hyperactivity disorder, and learning disability and had lower IQs than their peers. In addition, these children were less socially adept than their heavier or full-term peers. This same pattern appeared with other premature infants of greater birth weights.

Ross and associates⁹² measured the academic and social competence at 7 to 8 years of age of boys and girls with birth weights lower than 1501 g. They found that, as group, these children had lower scores than their full-term peers on measures of social competence and cognitive functioning and had a greater incidence of conduct disorders. Differences were greatest for children from the lower socioeconomic groups and for boys.

Investigators in Canada have been effective in capturing regional cohorts. Saigal and associates⁸² examined the 8-year outcome of somewhat socioeconomically advantaged children with birth weights of 501 to 1000 g and compared them with a matched group of children born at term. They found that the majority of children with ELBW had IQs in the normal range but lower than those of the full-term controls. Moreover, 8% to 12% of the children with ELBW scored in the “abnormal range,” in comparison with only 1% to 2% of the controls. Even the children with ELBW who were neurologically “normal” were performing below grade level, according to their teachers’ ratings, and had difficulties with visual-motor tasks. In a later study, Saigal and associates⁹³ evaluated children with birth weights lower than 1500 g and compared them to full-term children at ages 8 to 9 years. Very few of the children with ELBW had no functional limitations, and significant numbers of these children had cognitive problems and difficulties with mobility and the processing of sensory information.

In the United States, the studies of Hack and coworkers are of particular interest because their follow-up program has continued for many years, entails evaluation of infants admitted to a single tertiary care unit, and has had excellent subject retention. Hack and coworkers⁸⁹ compared children with birth weights lower than 1500 g to full-term children at ages 8 to 9 years. They found that 10% of the infants with VLBW had major neurological deficits and an additional 21% had IQ scores lower than 85. Although the neurologically intact infants with VLBW had IQs similar to those of full-term controls, they had significantly poorer scores on tests of expressive language, memory, and visual—fine motor skills and had a higher incidence of hyperactivity. These differences persisted even after the investigators controlled for social risks.

In another study, Hack and coworkers⁹⁴ evaluated a small group of children with birth weights lower than 1000 g and found that those with birth weights lower than 750 g did much worse in school than did premature children with higher birth weights. In turn, the latter performed more poorly than did matched full-term controls. Interestingly, abnormal head ultrasonograms and prolonged oxygen dependence were associated with mental retardation and cerebral palsy. In a similar study, Halsey and colleagues⁹⁵ monitored 210 children with birth weights lower than 800 g into the school years and found that although many of these children scored in the cognitively normal range, their scores were significantly lower than those of matched full-term children. In addition, 20% of this group had disabilities, including cerebral palsy, mental retardation, autism, and learning problems, and half of the children with ELBW required special educational services. Similar patterns were reported by Taylor and associates⁹⁶ and LaPine and coworkers.⁹⁷

Kilbride and Daily⁹⁸ performed an 8-year follow-up study on 114 children with birth weights of 500 to 750 g. Of this group, 30% were considered normal at 3 years and 89% were in regular classes without educational assistance. Fifty percent had suspect IQ scores (69 to 83) and motor quotients at age 3 years. Of importance, 20% of these children were in special education classes and 33% were held back a grade and were receiving learning support. Forty-six percent were functioning in an age-appropriate class, although only 15% were not receiving additional services. (Twenty percent were abnormal at 3 years.) Seventy-five percent of there children with combined cognitive-motor concerns were in special remedial classes. This study revealed that performance at 3 years of adjusted age was predictive of functioning at 8 years. This pattern of outcome is described by a number of other reports from different centers.^{82,90,92–126}

These studies from the 1990s documented an increase in survival of children with VLBW and ELBW in comparison with earlier decades, but they also documented significant morbidity. Even among children who seem “normal,” there are ongoing major academic and behavioral challenges necessitating educational interventions and supports.

More recent studies, published between 1999 and 2005, focused primarily on infants with VLBW or ELBW.^127–167 Although most earlier studies were conducted in the United States, Canada, and Australia, later studies documented outcomes from Germany, as well as other European countries. In the Bavarian Longitudinal Study,¹²⁷ investigators reviewed the outcomes from multiple centers in Germany, assessing at 6 years of age children born with gestational ages of less than 32 weeks and comparing them with matched, full-term controls. The investigators found that the children born before term scored significantly lower on cognitive, language, and prereading skills than did the controls and were more likely to have deficits in simultaneous processing. Preterm birth had a greater effect on outcome than did socioeconomic status.

Investigators in the Epidemiological Project for ICU Research and Evaluation (EPICure) study^129,130 evaluated children with gestational ages of less than 25 weeks when they were 30 months old and then at 6 years of age. The investigators found that severe disability at 30 months was predictive of outcome at 6 years. At the 6-year pediatric visit, 46% of the 78% of surviving children who had participated at 30 months had cognitive and neurological impairments. Twenty-one percent had moderate to severe cognitive impairments in comparison to test norms, 41% had moderate to severe impairments in comparison with their classmates, 22% had severe developmental disability, 24% had moderate disability, 34% had mild disability, and 12% had disabling cerebral palsy and cognitive deficits. Thirty-eight percent whose impairments were classified as “other disability” at 30 months of age had severe disability at 6 years. Twenty-four percent who had been classified as having no disability at 30 months had significant disability at 6 years of age. Vohr¹³² reported the same pattern of outcome for a large multicenter cohort of children with birth weights of 501 to 1000 g. Significant numbers of these children had neurodevelopmental disorders, cerebral palsy, and Bayley scores lower than 70, in addition to hearing and vision impairments.

Follow-up studies from this period in which investigators evaluated very premature children and children with VLBW at 7 to 12 years of age reveal significant, previously undetected deficits in social functioning, academic performance, and attention.^133,134 The increasing survival of more immature and lighter babies is evident in comparisons of outcomes with earlier time periods. However, the consequences of this survival are increasing numbers of children with significant neurodevelopmental problems and the emergence during the school-age years of previously undetected social, academic, and behavioral difficulties.

Chromosomal Aneuploidy

Down syndrome, formerly known as “mongolism,” which most commonly results from trisomy 21, is the best-known example of chromosomal aneuploidy. Early theories of its etiology drew from the mistaken biological and racist theories of the time (that ontogeny recapitulated phylogeny and that “Mongols” represented an evolutionarily primitive stage of development); awareness of the trisomy etiology of Down syndrome is now common in the general populace. Trisomy 13 and trisomy 18 are other examples of autosomal aneuploidies that are sometimes compatible with fetal and neonatal survival, but they carry devastating implications for behavior, development, and all aspects of health.² Turner syndrome (45,XO), Klinefelter syndrome (47,XYY), and other less common disorders associated with sex chromosome aneuploidy are associated with phenotypes that are medically less severe than the autosomal aneuploidies but that have important developmental and behavioral implications.² Chromosomal aneuploidy typically results from meiotic nondisjunction, which is progressively more likely to occur as maternal age increases. Although all chromosomes can be affected by meiotic nondisjunction, it is believed that aneuploidy of chromosomes other than 21 (the smallest chromosome) almost invariably leads to fetal demise. Chromosomal aneuploidy is easily detected by conventional karyotype testing.

Other Chromosomal Anomalies

Some sporadic cases of developmental disabilities and organ malformation are associated with visible deletions of a chromosome. The loss of such substantial amounts of chromosomal material implies that these conditions are associated with the deletion of many genes, ranging in number possibly into the hundreds, depending on the specifics of the deletion. Conversely, some disorders of development and behavior may result from duplication of a portion of a chromosome, resulting in “overdosage” of the genes on the duplicated segment of chromosome. Chromosomal anomalies such as ring chromosomes, derivative chromosomes (in which portions of multiple chromosomes re-form into a single atypical chromosome), and balanced and unbalanced translocations can be associated with developmental and behavioral disorders when chromosomal material (i.e., multiple genes) is lost in formation of the new chromosomes or when the “breakpoint” of the original chromosome or chromosomes occurs in a gene that becomes dysfunctional as a result of the break. All these types of chromosomal anomalies can be detected on a conventional karyotype.

Telomeric Abnormalities

The telomeres and the subtelomeres, which are the ends of each chromosome and the regions immediately adjacent, are unique chromosomal regions that contain long stretches of DNA but do not contain genes. The understanding of telomeric function and telomeric molecular biology is still developing, but it is already appreciated that chromosomal rearrangements in the subtelomeres can be associated with cancer and with perturbations to the processes of cell senescence.

Rearrangements in the subtelomeric regions appear to be responsible for 5% to 10% of cases of moderate and severe mental retardation.³ Most cases of subtelomeric rearrangement are associated with novel or unnamed syndromes of disability. Subtelomeric rearrangements can include both deletions and duplications of chromosomal material and are very difficult to detect with conventional karyotyping. However, these subtelomeric rearrangements can be identified with specialized diagnostic methods. It is believed that individuals with subtelomeric rearrangements typically have evidence of dysmorphology and/or congenital malformations, in addition to their neurodevelopmental symptoms. As the understanding of telomeric function and dysfunction expand, their role in development and behavior will also become better understood.

Contiguous Gene Deletion Syndromes

Williams syndrome⁴ and velocardiofacial/DiGeorge syndrome⁵ are two well-known examples of contiguous gene deletion syndromes. In these genetic disorders, a submicroscopic portion of a chromosome is deleted, at 7q11.23 and 22q11.2, respectively, resulting in the deletion of all of the genes that are physically contiguous on that portion of the chromosome. In a majority of cases, these deletions occur spontaneously, but similar deletions occur in unrelated persons because of a common peculiarity of the chromosome in these regions. In Williams syndrome, for example, the deleted region is bounded on both ends by repetitive chromosomal segments that are duplicates of each other. As a consequence, these homologous regions can mispair during meiosis, which results in the deletion of the intervening region of chromosome. The same mechanism is believed to operate in velocardiofacial/DiGeorge syndrome. The risk of recurrence in affected families appears to be greater than the incidence in the general population, perhaps because these families have a particularly high degree of homology between the bounding segments.

Persons affected by contiguous gene deletion syndromes are haploid for the genes that are deleted; that is, they are missing one copy of the deleted genes. However, because humans have two copies of each autosomal chromosome, the affected persons still have one copy of these genes on the unaffected member of the chromosomal pair. Contiguous gene deletion syndromes thus highlight the concept of haploinsufficiency. Whereas normal development and physiological function may be possible with only one copy of some genes, two copies may be necessary for other genes. When only one copy of a gene from the latter category exists, the affected person is said to be haploinsufficient. In contiguous gene deletion syndromes, it is the haploinsufficiency of certain genes that contributes to the atypical phenotype. Research is active on Williams, velocardiofacial/DiGeorge, and other contiguous gene deletion syndromes to determine which of the deleted genes are haploinsufficient and thereby contribute to the phenotypes associated with these syndromes.

As is true for almost all genetic disorders, contiguous gene deletion syndromes can produce large variability in the phenotype of individual patients. Researchers initially hypothesized that this variability was related to variability in the number of genes that were deleted in each patient; that is, all patients with a specific syndrome would have a core group of genes deleted, but some patients would have a larger chromosomal deletion, with differences in phenotype accounted for by the exact number and identity of the additional genes deleted. In fact, it is now understood that the large majority of patients with a specific syndrome have exactly the same gene deletions. For example, about 95% of patients with Williams syndrome have a common deletion involving at least 23 genes at chromosome 7q11.23,⁶ and about 70% of patients with velocardiofacial/DiGeorge syndrome have a common deletion involving about 15 genes at chromosome 22q11.2.⁵ Variability in these syndromes arises from the specific alleles that the patient carries for the other copy of the deleted genes, from “background genetic effects” (i.e., the effects of the remainder of the person’s genome), and from other biological and experiential factors.⁷

The variability in phenotype for patients with a common genetic disorder sometimes led historically to a proliferation of diagnoses despite unitary etiologies. For example, the “conotruncal anomaly face syndrome,” DiGeorge syndrome, velocardiofacial syndrome, and some cases of “Opitz G/BBB” syndrome were thought to be distinct diagnoses, but all are now known to result from a contiguous gene deletion at chromosome 22q11.2.⁸

In contiguous gene deletion syndromes, the deleted chromosomal segment is typically too small to be detected by conventional microscopic karyotyping. Instead, the deletion must be probed for and found absent, through the fluorescent in situ hybridization (FISH) test. As previously stated, these disorders most commonly occur spontaneously (de novo), but vertical transmission can occur.

Single-Gene Disorders

The effects of single-gene abnormalities on development and behavior can be as powerful clinically as the effects of chromosomal disorders, whether the single-gene abnormality results from gene deletion, base pair mutation, or other genetic mechanisms. For example, Lesch-Nyhan syndrome results from a mutation in the gene HGPRT and is associated with a phenotype of severe mental retardation and self-injurious behavior. The study of Lesch-Nyhan syndrome led to the first known usage of the term behavioral phenotype, referring to the concept that genetic differences can be associated with specific phenotypes of behavior.⁹ This statement was one of the earliest and most powerful medical refutations of the concept of tabula rasa that had been championed by behavioral psychology in the first half of the 20th century.

The etiology of the fragile X syndrome, another single-gene disorder, was elucidated more recently. The clinical syndrome was described in 1943, but the specific underlying genetic mechanism, an extended repeat of three base pairs (“triplet repeat”) on the X chromosome, was not discovered until 1991. The triplet repeat mechanism of genetic disease had never been described in any condition before then, but it is now known to underlie not only the fragile X syndrome but also Huntington disease, myotonic dystrophy, several spinocerebellar ataxia syndromes, and other disorders.¹⁰

It is now appreciated that the fragile X syndrome is the most common single-gene etiology for mental retardation and the most common heritable cause of mental retardation. Down syndrome is the most common genetic etiology but is rarely transmitted vertically, whereas the fragile X syndrome is less common overall but typically results when a mother has a slightly expanded triplet repeat, known as a “premutation,” that expands into a “full mutation” in gamete formation. The phenomenon of triplet repeat expansion manifests itself clinically in the greater phenotypic severity of patients with a full mutation in comparison with patients with a shorter premutation. Research continues on the molecular implications and clinical correlates of the premutation in the fragile X syndrome, as well as on the full range of clinical manifestations associated with the full mutation.¹¹

Single-gene disorders can be diagnosed only through directed testing, triggered by clinical suspicion, and cannot be detected through nonspecific tests such as karyotyping. The specific tests for a single-gene disorders range from metabolic testing of urine or serum to gene sequencing and to other molecular diagnostic assessment, such as the test commonly used to determine triplet repeat length in the fragile X syndrome. For neurofibromatosis type 1 and tuberous sclerosis, as well as for many other single-gene disorders, diagnoses are typically made clinically, with DNA testing reserved for confirmatory testing, prenatal diagnosis, or other uncommon situations. In most cases other than the fragile X syndrome, the diagnostic testing for single-gene disorders is directed by a genetic or metabolic specialist.

Mitochondrial Genes

In addition to the nuclear genomes of both the sperm and the egg, all human embryos receive a genetic endowment from the mitochondria of the ovum. (Mitochondria from the sperm do not survive in the zygote, and thus no mitochondria from the father are passed on to the child.¹²) The mitochondria contain a genome that is much smaller than the genome that is in the nucleus of cells. The mitochondrial genome contains only about 37 genes, and all of these genes appear to be important for mitochondrial function. (There are other genes in the nuclear genome that also are important for mitochondrial function.) In this age of assisted reproductive technologies, it is useful to note that children who are conceived from donor eggs are genetically related to the egg donor through both the nuclear genome and the mitochondrial genome of the donor egg.

Despite the small number of genes in the mitochondrial genome, mutations in these genes appear to be very likely to affect brain development or function, perhaps because the high energy demand of the brain makes it particularly dependent on mitochondrial function. Examples of disorders resulting from mutations in mitochondrial DNA include the syndromes of mitochondrial myopathy, encephalopathy, lactic acidosis, and strokelike episodes (MELAS) and of myoclonic epilepsy associated with ragged-red fibers (MERFF). Some clinical syndromes, such as MELAS, may be attributable to mutations in one of several different genes. These disorders are typically diagnosed when clinical suspicion leads to pediatric neurological evaluation, followed by specific testing that sometimes requires muscle or other tissue biopsies.¹³ For some of these mitochondrial disorders, DNA testing also is available for confirmatory or prenatal testing.

Multiple-Gene Disorders and Allelic Differences

In contrast to the chromosomal disorders and single-gene disorders such as Lesch-Nyhan and the fragile X syndromes, which typically result in complete inactivation or overdose of one or more genes and have severe effects on development and behavior, many genes exert their influences on development and behavior through more subtle, additive effects. Reading disabilities^14,15 and autistic disorder¹⁶ are examples of this. The complex behavioral phenomena of these disorders are believed to result from the effects of multiple genes. Mutations or allelic differences in only one of the implicated genes may have only minor effects or no effect at all. However, if several of these genes are abnormal or are present in the form of a “pathological” allele, then full-fledged dyslexia or autism may result. If the abnormalities are present in only one or a few of these genes, then they may manifest only as a shy temperament or a tendency toward perseveration. Abnormalities that are limited to another few genes might manifest as a restricted range of interest. When a critical number of the implicated genes are abnormal or are present in the pathological allele, then the full clinical picture of autism may emerge.

Allelic gene differences are known to influence the susceptibility of an individual to a phenotypic disorder. One of the best known examples of this, from outside the field of DBP, is the susceptibility to breast cancer that is associated with mutations in the gene BRCA1.¹⁷ These mutations do not directly cause breast cancer, but they raise the risk of breast cancer substantially, because the mutations affect the ability of the BRCA1 protein to regulate the cell cycle. Other genes are believed to influence development and behavior through their effects on complicated pathophysiological processes that involve other genetic and biological factors, as well as environmental factors. Some examples of susceptibility genes and their interaction with environmental factors are described later in this chapter.

Chapter 16 reviews some of the genes that have been linked to attention-deficit/hyperactivity disorder (ADHD).¹⁸ It is not yet understood how certain alleles of these genes act to increase the likelihood that a person will have ADHD, but it is clear that these genetic differences alone do not completely prevent or cause ADHD by themselves, Instead, they presumably exert their influence through some interaction with other genes and/or with environmental factors. Background genetic effects are likely to have a strong influence on the phenotypic expression of “susceptibility genes,” just as they do on contiguous gene deletion syndromes and other genetic mechanisms affecting development and behavior.

The Epigenome

Discussions of genetic effects on phenotype typically focus on the genome—that is, on the specific genes on the chromosomes and in the mitochondrial DNA—and on the sequence of noncoding base pairs that are found between genes. There is a growing appreciation, however, that the “epigenome” also exerts important effects on the phenotype. The epigenome is defined as the entire array of gene expression states imposed by chromatin and nonhistone regulators on the genome. The two examples of epigenomic regulation that are most familiar to developmental-behavioral pediatricians are gene inactivation by methylation and X chromosome inactivation through the normal process known as lyonization.

Genomic imprinting, in which a cell can “tell” whether a specific gene is from the maternally inherited chromosome or the paternally inherited one, is commonly implemented through a process of methylation, in which a methyl group becomes bonded to the DNA in the promoter region of certain genes. One example of the significance of imprinting is found in the social skills of girls with Turner syndrome. Such girls who inherit their single X chromosome from their father exhibit social skills that are relatively superior in comparison to those of such girls who inherit their X chromosome from their mother.^18a

Another example of imprinting is found in Angelman and Prader-Willi syndromes, which result from deletions at chromosome 15q11-q13 on the maternally derived chromosome (Angelman) or the paternally derived chromosome (Prader-Willi). Alternatively, either of these disorders may result from uniparental disomy, the condition in which both copies of a particular chromosome are derived from the same parent, rather than one from the mother and one from the father.

In the fragile X syndrome, the inactivation of the mutated FMRP1 gene is associated with methylation of the FMRP1 gene. There are reports of rare cases in which there is a “full mutation” of the FMRP1 gene, but the gene is still at least partially expressed because it has somehow escaped methylation. Individuals in which this occurs exhibit less severe phenotypic effects than in most cases of full mutation, in which gene methylation results in complete absence of gene expression.¹¹

Lyonization is an early embryonic process by which one of the two X chromosomes in females is inactivated in each cell of the embryo, which results in the formation of the Barr body. For female patients who are carriers of an X-linked disorder, the random process of lyonization sometimes results in the inactivation of the normal allele of a disease gene in an unusually high (or low) percentage of cells. In the fragile X syndrome, the severity of the clinical phenotype in female carriers of the full mutation is correlated with the lyonization ratio for the abnormal versus the normal X chromosome.¹¹ Analogous findings have been reported for female carriers of the gene for Rett syndrome.¹⁹

The effects of epigenomic phenomena are thus evident in multiple disorders relevant to DBP, including Prader-Willi and Angelman syndromes and X-linked disorders such as Rett syndrome, Turner syndrome, and the fragile X syndrome. The possible effects of environmental factors on the epigenome are discussed in greater detail as follows.

Genomics and Proteomics

Although the sequencing of the entire human genome was a major research achievement, that accomplishment highlighted the fact that knowing the order of the billions of base pairs in the genome is only an intermediate milestone in understanding human molecular biology. The task that researchers now face is known as genomics research: elucidating the function and interactions of the more than 25,000 genes that every human being carries.²⁰ The role of regulatory genes, the poorly understood significance of gene introns and of the “junk DNA” that is found between genes, the extent to which alternative splicing occurs to create different transcripts from a single gene, and other mysteries all remain to be explored as part of genomics research. A potentially even more complex challenge is the understanding of proteomics: how the myriad protein translation products of the genes interact with each other in the physiological processes that ultimately manifest themselves in the typical or atypical growth, development, and behavior of children and adults.²⁰ Beyond genomics and proteomics lies the even larger challenge of understanding how environmental factors interact with these biological processes.

Medical Phenotypes

Genetic disorders and differences appear to be associated with a very wide variety of medical phenotypic manifestations. The most obvious are the congenital malformations that characterize many chromosomal disorders. In these conditions, malformations may affect any organ system and often affect multiple organ systems in a single individual. Some of these malformations may be considered major, in that they are incompatible with life or necessitate surgical intervention to establish a normal range of function, whereas other malformations may be considered minor or cosmetic.

When the best medical care is accessible, advances in medical, surgical, and chronic care have led to enormous improvements in the life expectancy and the quality of life of many individuals with chromosomal and other genetic disorders that are associated with major medical morbidity. For example, as recently as the 1960s, the congenital malformations that affect most newborns with Down syndrome, especially cyanotic cardiac defects, implied a very short life expectancy. Advances in cardiac diagnosis and surgical intervention now enable almost all such infants to survive and often to thrive with full physiological repairs of their malformations. Whereas it used to be common and acceptable to allow such neonates to die without heroic medical intervention, it is now considered unethical in many parts of the industrialized world to withhold medical intervention unless there are at least two major congenital malformations. Thus, it can be fairly asserted that in affluent communities, the effect of medical phenotypes on and across the lifespan has changed dramatically. Advances in educational and related therapeutic services also have had a dramatic effect on the development and quality of life of older children and adults with Down syndrome.²¹

Research on medical phenotypes continues to have real clinical significance nonetheless. As a direct result of the improvements in and greater availability of treatment for congenital malformations, and consequent increases in life expectancy, one area in which research is most active is the late-adult phenotype associated with genetic disorders. In Down syndrome, for example, the early onset of Alzheimer-like dementia was not recognized until sufficiently large numbers of individuals with Down syndrome began to survive into the middle-adult years.

The fragile X syndrome provides a vivid example of how new aspects of the medical phenotype are still being brought to light for a disorder that was recognized before 1950. (In this case, the new recognition is not simply the result of increased survival, but of persistent inquiry and keen clinical acumen, combined with advances in genetic diagnostics.) A new syndrome of tremor and ataxia has been identified in adults with premutations of the FMRP1 gene, and female carriers of the premutation appear to have an increased incidence of premature ovarian failure.¹¹ In many other genetic disorders, increasing survival into the adult years has called attention to the need for careful medical surveillance and systematic study of possible late manifestations of the medical phenotype.

Pharmacogenomics

Although the topic of pharmacogenomics—genetically based differences in drug pharmacology—usually is not included in DBP reviews of medical phenotype, it is a topic of enormous clinical relevance, for which the knowledge base is likely to grow very quickly.²² Researchers have begun to identify specific genes that affect both pharmacokinetics (the effects of the body on the drug: namely, absorption, distribution, metabolism, and elimination) and pharmacodynamics (the effects of the drug on the body). In the case of drug metabolism, the genetic basis for individual differences in the rate of metabolism of certain drugs has been found. Patients who are “poor metabolizers,” “extensive metabolizers,” and “ultra-rapid metabolizers” of various antidepressants and antipsychotics are now known to have different genotypes for the cytochrome P-450 2D6 enzyme (CYP2D6). The poor metabolizers carry alleles for this gene that code for a relatively low-activity version of the enzyme, whereas the ultra-rapid metabolizers carry a higher activity allele and have extra copies of these genes.

Behavioral Phenotypes

Lesch-Nyhan syndrome is the vivid exemplar for which the term behavioral phenotype was first coined; the behavioral phenotypes associated with other genetic disorders and differences are, in general, less striking.^23–25 These phenotypes span the range from other dramatic but circumscribed behaviors to temperamental characteristics, profiles of cognitive ability, and trajectories of cognitive development. Other examples of phenotypes that comprise circumscribed behaviors include the hand-wringing associated with Rett syndrome and the spasmodic “self-hug” that has been described in Smith-Magenis syndrome.²⁶ On the border between circumscribed behaviors and more generalized behavioral and temperamental phenotypes are the hyperphagia associated with Prader-Willi syndrome and the extreme difficulty with sleep seen in many individuals with Smith-Magenis syndrome (which may be related to an abnormal circadian rhythm for melatonin secretion²⁷). Temperamental or personality phenotypes have been described for several genetic disorders, including Williams, Prader-Willi, Angelman, and velocardiofacial syndromes.^4,28,29 Well-recognized (if not fully understood) developmental trajectories include the degenerative patterns of Rett syndrome and of Down syndrome (in adulthood), the onset of hyperphagia in Prader-Willi after early hypotonia and motor delays improve, and the apparent acceleration of language development often seen in toddlers and preschoolers with Williams syndrome (although this may be analogous to the language burst that many typically developing children seem to show around ages 2 and 3).³⁰ Some genetic conditions also appear to be associated with an increased risk of psychiatric disorders. Examples include the possible increased incidence of internalizing disorders in adults with Williams syndrome⁴ and the complex relationships between the fragile X syndrome and autism¹¹ and between velocardiofacial disorder and both mood and psychotic disorders.^30a Table 10B-2 depicts the different facets of behavioral phenotypes that are relevant to DBP.

TABLE 10B-2 Facets of Genetically Based Phenotypes

Cognitive profiles are probably the most extensively investigated type of behavioral phenotype in children and adults with genetic disorders. Down syndrome, Williams syndrome, velocardiofacial/DiGeorge syndrome (del 22q11.2), the fragile X syndrome, neurofibromatosis type 1, Turner syndrome, and Duchenne muscular dystrophy are among the disorders that have been the subject of extensive psychoeducational and cognitive investigation. Progress in the understanding of their cognitive behavioral phenotypes often follows a similar pattern: early publications report on the typical IQ, academic achievement, and functional profiles, whereas in later studies, investigators attempt to elucidate the fundamental neuropsychological or cognitive profiles that lie at their heart. In many cases, there also are neuroanatomical and other biomedical studies (including structural and functional neuroimaging, as well as postmortem neuropathological characterization) in which researchers examine the biological basis for the behavioral phenotypes, such as the discovery that the sleep difficulties in Smith-Magenis syndrome are associated with abnormal melatonin physiology. There seems to be little reason to doubt that progress in understanding brain-behavior relationships will eventually erase, or at least blur, the distinction between medical and behavioral phenotypes.

The psychiatric literature contains an analogous line of investigation that uniquely employs the term endophenotype, which refers to traits that are believed to be at the core of the complex processes that ultimately manifest as psychiatric disease or risk for psychiatric disease.³¹ Most commonly, the endophenotypic traits are neuropsychological or neurophysiological in nature. For example, verbal short-term memory, eye blink conditioning, and saccadic eye movements all have been studied as endophenotypes for various psychiatric diagnoses, and the genetics of these traits are studied as possible clues to the genetics of psychiatric disease. Implicit in the research on endophenotypes is the recognition that psychiatric disorders are extremely complex phenomena that probably result from developmental pathology or deviance in multiple underlying processes.

A number of caveats need to be considered carefully in interpreting the research on behavioral phenotypes.³² These caveats pertain both to the description of the phenotypes and to the “immutability” that is often incorrectly ascribed to them because of their genetic basis.

Advances in medical diagnostics also account for changes in typical levels of cognitive ability in individuals with other disorders. Disorders in higher functioning individuals who do not have major medical stigmata are now diagnosed by FISH or other techniques, whereas they would not have been diagnosed previously. Thus, for disorders such as Williams, the fragile X, and Smith-Magenis syndromes, the availability of modern diagnostics has shifted the range of individuals who receive diagnoses, resulting in a shift of our understanding of the typical behavioral phenotype, whereas in Down syndrome, the phenotype itself has shifted as a result of improvements in the environment that these children typically encounter during the course of their development.

Approach to Diagnostic Evaluation

Until every newborn receives exhaustive testing for all possible genetic disorders and differences, the diagnosis of these conditions will remain driven largely by clinical suspicion; that is, genetic conditions can be diagnosed only if they are suspected and if appropriate diagnostic tests are then requested. Although the most common genetic cause of mental retardation, Down syndrome, is in most cases easily recognized and is diagnosable by routine karyotype, specific testing is required for diagnosis of most of the other recognizable patterns of human malformation. For conditions that involve gene deletion, diagnosis is typically by FISH testing, through use of a molecular probe that is specific for the suspected condition. Thus, the diagnosis is not made unless a specific clinical suspicion engenders specific testing. Other conditions that result from defined molecular genetic abnormalities can be diagnosed with directed testing of other sorts, but these tests also must be specifically requested on the basis of a specific suspicion. Only a few commonly occurring disorders besides Down syndrome (notably, the sex chromosome aneuploidies) are diagnosable by routine karyotype.

Rapid advances in molecular biological technology may soon make suspicion-driven diagnosis obsolete. So-called microarray methods are able to assess the expression of hundreds or thousands of genes simultaneously, with very small samples of blood or other tissue.³ It may soon be possible for the developmental-behavioral pediatrician to send a single blood specimen for genetic testing, with a note on the clinical context for testing, and for the laboratory to screen rapidly and inexpensively for any of thousands of genetic disorders or differences that are known to be associated with that clinical history.

The decision on when to request genetic testing is one whose answer is evolving as well. Classically, the only children who underwent genetic testing were those with multiple congenital anomalies or with dysmorphology, because they were the only ones in whom testing was likely to yield informative results. As diagnostic methods advanced, and as the spectrum of diagnosable genetic differences widened, the population of patients who might benefit from genetic testing grew very quickly as well. Many authorities now recommend that all children with significant developmental delays of unknown etiology should be considered for testing that includes a karyotype, subtelomeric probes, and DNA testing for the fragile X syndrome, because abnormal findings on any of these tests can be associated with seemingly nonspecific developmental impairment.³ For children with autistic spectrum disorders, additional testing might be recommended, as discussed in Chapter 15. Of course, a neurodegenerative history or the presence of certain other neurological signs and symptoms is an indication for testing for various metabolic and storage diseases that manifest in these types of presentation and are discussed in Chapter 10C. Here again, advances in diagnostic methods that allow simple and inexpensive testing for many conditions simultaneously may soon make an encyclopedic knowledge of genetic disorders obsolete at the diagnostic stage of care. One algorithm for the approach to diagnostic testing is illustrated in Figure 10B-3.

FIGURE 10B-3 Approach to the clinical genetics evaluation for developmental disabilities and mental retardation. FISH, fluorescent in situ hybridization; MRI, magnetic resonance imaging.

Rights were not granted to include this figure in electronic media. Please refer to the printed book.

(From Moeschler JB, Shevell M, American Academy of Pediatrics Committee on Genetics: Clinical genetic evaluation of the child with mental retardation or developmental delays. Pediatrics 117:2304–2316, 2006.)

For conditions and genetic disorders whose phenotypes are less pervasive than for the chromosomal and similar disorders, diagnosis will probably remain driven by specific clinical questions, at least for the near future. As an example, if and when pharmacogenomic research is able to determine who will respond best to which drugs, it will be the physician’s responsibility to request the appropriate genetic testing to guide prescribing practice.

Some patients and families are unmotivated to undergo or even resistant to diagnostic testing. In these cases, the possibility of diagnostic testing should be revisited at a later time because of possible implications for reproductive decisions by the patient or by other family members and because the benefits associated with establishing a diagnosis are likely to increase as research advances.

Treatment Implications

With relatively few exceptions, genotypic diagnosis does not currently lead to specifically effective recommendations for developmental and behavioral intervention. Research remains focused on phenotypic description, which despite its challenges, remains a much simpler task than the identification of diagnosis-specific treatments and behavioral interventions.

Special education and neuropsychology are two disciplines that have led in the development of diagnosis-specific treatment strategies. For example, children with Down syndrome typically show particular impairment in auditory-based phonetic and phonological skills, which results in significant compromise in the development of their spoken language abilities.^47,48 When the communicative abilities of these individuals are below the level of their general cognitive abilities, behavioral consequences may result. Many experts therefore recommend the extensive use of sign language for young children with Down syndrome, taking advantage of their better preserved capacities for learning a nonverbal, nonauditory language. Although some are concerned that the use of sign language will delay or impair the development of spoken language skills, the studies that exist suggest that this concern is generally unwarranted and that the early introduction of sign language is associated with lasting benefits to the communicative and social skills of children with Down syndrome.⁴⁹

Other examples of therapeutic interventions that are diagnostically specific include the use of “verbal mediation” techniques for children with Williams syndrome and the use of intensive oral drills for teaching arithmetic to children with velocardiofacial/DiGeorge syndrome. In the case of Williams syndrome, this educational/therapeutic approach is driven by the neuropsychological profile that is associated with the syndrome, in which verbal and auditory skills are strengths that can be used to support other functions that are not as strong.⁴ In the case of velocardiofacial/DiGeorge syndrome, the suggestion to use a specific educational approach is driven largely by repeated anecdotal reports from parents. In both cases, the suggested therapeutic approaches remain open to rigorous validation, but they appear to be promising with regard to the potential for diagnostically driven intervention.

As discussed in the section on phenotypic variability, care providers and parents must remember that every child is unique and that the therapeutic suggestions intended to benefit most children with a specific diagnosis may not be valid for a specific child. For the purpose of educational programming and related therapeutic intervention, a thorough psychoeducational and/or neuropsychological evaluation of the individual child is the most helpful diagnostic assessment that can be made. Knowledge of the typical phenotype associated with the genetic diagnosis is best used to guide the psychological assessment and to supplement its results.

Gene therapy for neurodevelopmental disorders remains entirely hypothetical. Because a genetic diagnosis must be made before any specific genetic therapy can be instituted, and because genetic disorders affecting development and behavior are rarely diagnosed before the brain has completed all of its prenatal and much of its postnatal development, it is not clear what effect genetic therapies could have. Many genetic disorders affect brain function, as well as brain development, but it seems naive to believe that postnatal therapies could reverse developmental abnormalities that have already been completed. In the event that a genetic disorder is diagnosed in the course of prenatal genetic testing, the opportunity to intervene in the processes of brain development may be present, but there is currently no paradigm for prenatal genetic therapy.

Down Syndrome⁵²

Fragile X Syndrome^11,53

Muscular Dystrophy (Duchenne and Becker)

Neurofibromatosis Type I⁵⁶

Prader-Willi Syndrome

Rett Syndrome

Sex Chromosome Aneuploidies (Besides Turner syndrome)

Smith-Magenis Syndrome

Tuberous Sclerosis

Turner Syndrome⁶⁶

Velocardiofacial/DiGeorge Syndrome

Williams Syndrome⁷¹

Disorders in Energy Metabolism

These diseases are related to the ability to generate adenosine triphosphate or other energy substrates that are vital for normal cellular function. Among the best known of this group are the mitochondrial diseases. The majority of proteins involved in mitochondrial energy metabolism are encoded by nuclear rather than the small mitochondrial genome.² These diseases typically result in poor energy production and overproduction of lactic acid.^3,4 A second type of energy problem involves glucose metabolism.⁵ Conditions such as glycogen storage defects and gluconeogenic defects result in decreased supplies of energy substrate to the brain.^6,7 The disorders of fatty acid oxidation can also result in a limited glucose supply with similar effects.^8,9 This can rapidly result in depletion of both adenosine triphosphate and energy.

Disorders of Biomolecule Conversion

This group consists of defects in systems that convert one molecule to another. Examples include phenylketonuria, homocystinuria, hyperglycinemia, tyrosinemia, and the organic acidemias. In these diseases, failure of molecule conversion results in an oversupply of the one or more precursor metabolites and an undersupply of the products. Phenylketonuria is an excellent example of this concept. Absence of phenylalanine hydroxylase results in the accumulation of phenylalanine, which at high levels is toxic to neurons.^10–12 The inability to convert phenylalanine to tyrosine results in the conversion of tyrosine to an essential amino acid. Galactosemia is another example in which galactose cannot be isomerized to glucose.¹³ The unconverted galactose results in neurotoxicity and hepatotoxicity, and the potential energy from the glucose is lost.

Disorders of Biomolecule Clearance

These diseases are somewhat similar to the disorders of molecule conversion but occur in systems specifically designed to clear molecules that are toxic in large quantities. The urea cycle is an excellent example.^14,15 With turnover and dietary intake, the body must clear waste nitrogen on a regular basis. This nitrogen appears in the blood stream as ammonia, levels of which can elevate rapidly if the urea cycle cannot convert it to the readily excreted molecule urea. Another example is the breakdown of the simplest amino acid, glycine. Failure of the glycine cleavage complex results in its accumulation in the central nervous system, where glycine’s neurotransmitter properties interfere with function at many levels.¹⁶ In the purine metabolic pathway, defects in clearance result in toxic buildups in Lesch-Nyhan syndrome.^17,18 Disorders of biomolecule clearance also include the lysosomal storage disorders, which remove relatively inert molecules such as mucopolysaccharides and oligosaccharides.¹⁹ These molecules accumulate over time, and their occupation of cellular space leads to their toxicity.

Disorders of Cellular Function

Defects in basic cellular function constitute this group of disorders. The defects in aspartate transport (citrin deficiency, or citrullinemia type II) result in poor metabolism of both glucose and ammonia.^20,21 These disorders affect cerebral development and also result in long-term toxicity. The peroxisomal disorders, such as Zellweger syndrome, represent whole organelle failure, which results in damage to the brain and other organs.^22,23 These disorders are characterized by their effects on a wide range of cellular metabolic functions.

Acute

Some metabolites cause rapid damage only during an acute crisis episode. Examples of this damage are the rapid cerebral edema caused by ammonia in urea cycle defects and the acute hypoglycemia resulting from fatty acid oxidation defects such as medium-chain acyl—coenzyme A dehydrogenase deficiency.^8,14 Patients are often normal until the insult and then remain stable afterwards unless another crisis occurs. Long-term damage results from neuronal injury during the acute episode and can often produce strokelike residual problems.

Progressive

Patients with progressive disorders start out with normal development, but then their developmental trajectory is lost and they lose acquired skills. Examples of this scenario include the neurotoxic metabolites of phenylalanine and homocysteine, long-term exposure to mildly elevated ammonia levels, and the lysosomal storage disorders. The clinical course for these patients reflects the ongoing neurotoxicity and cell death from the toxins. Disorders in chronic energy metabolism (such as mitochondrial metabolism) also manifest in this manner.²⁴

Prenatal Onset

We also find patients whose developmental problems precede birth, such as those with hyperglycinemia or peroxisomal disorders.^16,22 The metabolites in these disorders accumulate beyond the ability of the placenta’s biofilter capacity or result in intracellular toxicity. These disorders continue to cause neurological deterioration after birth and are usually refractory to treatment.

Combined

Some disorders combine these features. Patients with urea cycle disorders develop acute hyperammonemia with damage to the brain but also have long-term neurotoxic effects from the chronic elevations of ammonia and other cycle intermediates such as arginine.^25,26 Patients with disorders resulting in lactic acidosis (such as pyruvate dehydrogenase) can have acute episodes of lactic acidosis but also suffer from chronic loss of neuronal tissue.⁵

A key point is that a downward progression of development and the loss of milestones is very suggestive of an ongoing toxic metabolic process, as is a change in development after an unexplained medical crisis.

Timing and Progression

As noted in the previous section, the timing of the developmental delay can often provide clues to the type of problem. A developmental course with steady deterioration of function is suggestive of a chronic neurotoxic metabolite, whereas a sudden onset of delay is produced by a different set of disorders. As a rule, fixed developmental delays without progression are less likely to represent metabolic disease than are progressive disorders.

Family History

Although all of the metabolic diseases discussed result from genetic defects, there is often no family history of problems, because most genetic defects are autosomal recessive. Several exceptions should be remembered. Patients with disorders on the X chromosome typically have a family history of affected male relatives over several generations. Examples include the urea cycle disorder ornithine transcarbamylase and the purine metabolic disorder Lesch-Nyhan syndrome. With both of these disorders, there is a substantial number of female relatives partially or even fully affected as a result of nonrandom X chromosome inactivation.²⁷ The disorders of mitochondrial dysfunction resulting from mutations in mitochondrial DNA manifest with a family history of maternal lineage, because mitochondria are passed on only through the ovum and not through the sperm. Finally, the family should be asked about any type of consanguinity that may exist. This circumstance dramatically increases the risk of rare autosomal recessive metabolic diseases.

Common Illnesses

The course of diseases accompanied by common illnesses such as colds and ear infections should be obtained. Patients with partial defects in metabolic systems typically become symptomatic with other illnesses.²⁶ Prolonged lethargy, slow recovery from common viral illnesses, and increased frequency of infections are all suggestive of an underlying metabolic disease. The onset of seizures with a mild illness is also highly suggestive of metabolic disease.

Seizures

The development of seizures in a patient with developmental delay is often suggestive of the presence of a toxic metabolite, which is the cause of both.²⁸ Prolonged seizures can be particularly related.

Neonatal History

The neonatal period is one of the most stressful metabolic periods for everyone. Lethargy, excessive sleeping, poor feeding, abnormal behaviors, seizures, and temperature irregularities are all indicators that a toxic metabolite may have been present. Most of these patients undergo a workup for sepsis that reveals no pathogen. By the time the sepsis workup is complete, the metabolic disease may have stabilized and become dormant.

Specific Findings in the History That Suggest Metabolic Disease

General Examination

Poor growth and an appearance of chronic illness are suggestive of an underlying process in developmentally delayed patients. Disrupted cellular metabolism from toxins or absence of vital substrate results in poor growth of both brain tissue and is also reflected in short stature and an emaciated appearance.

Head, Ears, Eyes, Nose, and Throat (HEENT)

Both microcephaly and macrocephaly are features of metabolic diseases. In the amino acid metabolism defects and in diseases with sudden onset of metabolic crisis, microcephaly most commonly results from loss of neuronal tissue. In disorders of glutaric acid metabolism, macrocephally is present and is often associated with abnormal ventricles visible on imaging.⁴⁵ A large fontanelle is present in patients with the peroxisomal disorder such as Zellweger syndrome.²³ The presence of clouding of the cornea is suggestive of a storage disorder, and cataracts often result from deposits of metabolites (this is fairly common in galactosemia).¹⁹ In patients with homocystinuria, displacement of the lens results from breakage of the fibrils holding the lens in place.⁴⁶ Examination under slit-lamp conditions by an ophthalmologist is often of particular use because the retina can also be examined for the cherry-red spots that occur in several storage disorders. Chewing of the lip is observed in patients with decreased pain sensation, as in Lesch-Nyhan syndrome.

Hair

The hair is useful in medical assessment of metabolic disorders. Hair growth requires protein manufacture and collagen crosslinking. In disorders of amino acid metabolism, the growth may be slowed as a result of deficiencies.^30,38 In the urea cycle disorders (particularly argininosuccinic lyase deficiency), the patients cannot make arginine and develop a characteristic finding of trichorrhexis nodosa. Microscopic examination reveals a bamboo appearance of the hair with numerous fragile points in the hair, which makes it breakable.³¹ Hair that has a very coarse or kinky texture is indicative of a defect in copper metabolism, which affects collagen crosslinking.⁴⁷ Microscopic examination reveals twists in the hair known as pili torti. This most closely resembles the twists seen in a cocktail stirring stick. Hair without pigment or very little pigment is suggestive of phenylketonuria or a defect in metabolism of tyrosine, which makes pigment.

Abdomen

The presence of chronic hepatomegaly is suggestive of a stored product. Glycogen storage disorders can manifest with this finding, and when accompanied by splenomegaly, it is suggestive of a storage disorder.

Skin

The presence of generalized rashes can indicate a lack of a key metabolic building block such as an amino acid (lysinuric protein intolerance, for instance) or buildup of an unprocessed metabolite, such as fatty acids in some of the fatty acid oxidation defects.^48,49

Neurological

Few specific findings in the neurological examination point toward a specific metabolic disease. The presence of hypotonia or poor muscle development in excess of the overall neurological state is suggestive of a mitochondrial myopathy, but it is difficult to distinguish from generalized hypotonia.

Common Laboratory Measurements

Readily Available but Not Commonly Drawn Laboratory Measurements

Radiological Tests

Although few radiographic findings are specific for metabolic diseases, some findings are more frequent than others. Partial or total agenesis of the corpus callosum, in addition to overall maldevelopment of the brain, is noted in patients with hyperglycinemia.⁵⁰ Enlargement of the ventricles with macrocephaly occurs in patients with glutaric aciduria. Disorders of energy metabolism with lactic acidosis affect the basal ganglia early in the course of disease and are often described as having a Swiss cheese appearance on magnetic resonance imaging. This is often described as Leigh encephalopathy by radiologists. Generalized loss of neuronal tissue observed on magnetic resonance imaging is also suggestive of an underlying metabolic toxin or process. Enlargement of the liver or spleen visible on abdominal imaging is suggestive of a storage disorder. Cardiomegaly or hypertrophy is also suggestive of either a metabolic myopathy or a storage disorder such as Pompe disease.

Assessments Specific for Metabolic Disease

These tests are conducted in specialized laboratories that require careful quality control. The tests are usually offered at major university hospitals and a select number of reference laboratories. Interpretation of the results is often difficult and requires the assistance of a specialist in biochemical genetics. Tests employing paper chromatography are less reliable and should be avoided. At the time of publication, a number of companies are offering “comprehensive” metabolic profiles to consumers, who are typically families with a child with a developmental disability. The same companies typically find a number of metabolic abnormalities that often lead to recommendations for a product or service that they also provide. Use of these companies should be discouraged because it can often delay a correct diagnosis or lead to treatments for the patient that may be harmful. The Society for the Study of Inborn Errors of Metabolism lists Clinical Laboratory Improvement Amendments (CLIA)—certified laboratories on its Web site (http://www.ssiem.org), as well as clinics specializing in metabolic diseases. Another good rule is that in the absence of overwhelming levels of a metabolite, a repeat sample is needed for confirmation. It is often best to conduct these tests when the patient is ill from the suspected underlying defect, because a higher diagnostic yield is obtained. However, acute illness from any cause can result in elevations or depressions of many of the tested elements and should always be considered. Finally, the use of parenteral nutrition can affect the levels of metabolites in these tests. As much information as possible about these issues should be provided to the testing laboratory, so that a proper interpretation can be made in context of the patient’s condition. Another helpful piece of information to include on the test request is the name and telephone number of a contact individual for discussion of abnormal results and suggestions for additional testing.

Disease-Specific Tests

These tests are typically difficult to perform, are expensive, and require carefully processed testing material (often biopsy material). It is recommended that they be performed by a biochemical geneticist or in consultation with another highly experienced individual.

Neonatal Screening Testing

As more and more states move to the expanded neonatal screening, a powerful tool has become available to the diagnostician. The tandem mass spectrometry employed in neonatal screening detects organic acids and most of the compounds involved in fatty acid oxidation that bind to carnitine. It also detects phenylketonuria and tyrosinemia through their secondary metabolites. This test can be used in the clinical setting as a cost-effective screening tool, with an understanding of what it does not detect. It does not detect storage disorders, mitochondrial defects, glycogen storage defects, most urea cycle defects, amino acid transport defects, long-chain fatty acids, glycoprotein defects, and many other disorders. It otherwise is an excellent part of a general workup if it was not performed on the patient in the neonatal period. If the neonatal information is available, it can be very useful to the clinician during workup of the developmentally delayed patient. A careful check of which metabolites are reported by the specific state should be made because the test detects more compounds than many state programs report.