We have presented the baseline neurodevelopmental data from the AS Natural History study. Our analysis of the BSID-III developmental quotients (DQs) demonstrated that the non-deletion participants were more advanced than the deletion participants in all aspects of development except expressive language skills, which is consistent with the findings from previous studies.11–15
We also noted significant differences in the VABS-II language and motor standard scores between molecular subtypes where again, non-deletion patients scored higher as compared to deletion patients. It is notable that overall, the most consistent finding in our study is the uniformly poor expressive language skill in all participants regardless of their molecular sub-class. Since the single common molecular mechanism in all participants is the lack of UBE3A protein in parts of the brain where the UBE3A
gene is imprinted, we hypothesize that UBE3A may be essential in the development of expressive language skills. Although expressive language delay is common in syndromes associated with intellectual disabilities, AS is unusual in that all individuals have extremely poor expressive language skills that are out of proportion to their overall level of intellectual disability and their receptive language skills (as shown by the BSID-III analyses herein), and even the highest-functioning adult individuals are not expected to develop more than a few short phrases. In our collective experience, the vast majority of AS individuals have little use of spoken words or phrases.
As found in previous studies, and consistent with the previously reported behavioral phenotype in AS,7
our study shows that AS children have a relative strength in their social skills (as indicated by the socialization domain in VABS-II).
A previous study on the developmental profile of 20 AS children between ages 5 months and 10 years using the BSID-II found that these children had a developmental age of no greater than 17 months, based on the Mental Development Index.7
In contrast, approximately 10% of our subjects had a developmental age of greater than 18.9 months (up to a maximum of 27 months), based on the Cognitive Scale of the BSID-III. One of the differences in these scores may be that the BSID-II Mental Development Index also included language-based items, whereas the BSID-III has separate cognitive and language domains. Thus, the cognitive scale of the BSID-III may be a more sensitive measure of cognition (as opposed to the BSID-II Mental Development Index that is more heavily weighted with language-based items).
There is little published data on changes in the developmental profile of AS individuals over time. It is also unclear whether the neurodevelopment in AS individuals continues throughout life or whether it inevitably stagnates at a pre-determined chronological or developmental age. We anticipate that longitudinal data from our AS natural history study will eventually help to answer these questions. Meanwhile, we performed exploratory regression analyses using our current baseline data to find out if there might be differences in the developmental profile between the younger and the older participants in our cohort. Our data suggest that the cognitive skills in children with deletions do not appear to develop much beyond 13 months while those in other molecular sub-classes continue to develop over time, albeit at a slower rate compared with a normally developing child. This is reflected both in the overall decline in the VABS adaptive behavior composite standard scores, which indicates that these children fall increasingly behind their age-matched peers in their neurodevelopment, and the widening difference in the VABS adaptive behavior composite standard scores between the deletion and non-deletion participants.
Although AS children are often regarded as being highly sociable, over time, their socialization skills lag increasingly behind age-matched peers, as seen in the decline in the VABS-II socialization domain standard scores as the chronological age of the participants increases. It is important to recognize that the assessment of socialization skills in the VABS-II is partly dependent on the linguistic capacity of the participant, suggesting that the decline in the socialization domain standard scores in the older participants may be a result of their inability to communicate well. Therefore, interventions to actively encourage communication using all modalities including gestures, sign language, and assisted communicative devices may help to promote the development of social skills in this population. We recognize the limitations of these regression analyses, including the use of a cross-sectional sample, the limited age range (up to only 60 months of age), and the relatively wide scatter in our data points. Nevertheless, we believe that these preliminary data suggest that the developmental profile of these individuals may change as they age, and hence the importance of obtaining longitudinal developmental data on AS individuals across a wide age range for at least 5 – 10 years. Although we acknowledge that most individuals with developmental delays will probably show a decline in their DQ and standard scores on standardized assessments with increasing age, it will be useful to know the rate and pattern of decline in these scores for any given condition because such information may have some prognostic value. In a longitudinal study on 55 boys with Fragile X syndrome aged between 8 and 68 months, Roberts et al. found that although the rate of increase in the developmental age, as measured using the Mullen Scales of Early Learning, lagged behind that of the chronological age, the rate of developmental progress was constant over that age range with no evidence of any decline or stagnation in development.34
An earlier study on the adaptive behavior skills of 80 children with Down syndrome between the ages of 1.08 to 11.5 years found that the rate of increase in their “adaptive age” was constant until about 7 years of age when there appeared to be a “plateau” in the development of their adaptive skills; however, the data points for participants older than 6 years of age were widely scattered suggesting that some individuals continued to progress while others stagnated or even regressed.28
In contrast, the rate of development in our cohort of AS children appeared to be decreasing or stagnating before the age of 4 years. Future studies examining the developmental trajectories of children with other genetic syndromes associated with intellectual disabilities may help elucidate the extent to which our preliminary observations are unique to AS. Moreover, such information provides useful and important baseline data for the development of future therapeutic trials that aim to reverse or slow this relative decline.
Analyses of the ABC scores confirmed that most of the participants were very hyperactive, which is consistent with frequent anecdotal parental reports of hyperactivity in these children. This finding is important because excessive activity may impede a child’s ability to concentrate and to learn new skills; hence, hyperactivity in these children may need to be actively managed. As has been reported by parents and observed by other clinicians, AS individuals with deletions had lower irritability (aggression) compared to those without deletions. This finding has also been noted in Prader-Willi syndrome in which those without deletions are more likely to have psychiatric difficulties and problem behaviors.35
In contrast, lethargy (withdrawal) was not a common problem in our population, which is consistent with a previous study in which participants with AS were found to be less lethargic when compared to individuals with developmental disabilities due to other etiologies using the ABC.33
Our participants did not demonstrate as many “problematic behaviors” based on the ABC as that observed in the aforementioned study, possibly because our participants are younger (mean age: 2 years 10 months vs. 9 years old) and may not have developed sufficiently to demonstrate many of the problem behaviors observed in the older AS population. Another possible hypothesis is that caregivers may find it easier to control maladaptive behaviors in younger children and hence do not perceive these behaviors as being “problematic”.
Strengths and Weaknesses of Study
Identifying an overall neurodevelopmental profile of young children with AS has been challenging due to the dearth of standardized instruments that can be used to evaluate subjects in the lowest ranges of cognitive functioning. One of the strengths of this study is the use of standardized and structured developmental instruments such as the BSID-III and VABS-II that allowed for the assessment of children with profound developmental delays. These instruments allowed us to objectively compare the developmental profile of children with AS to age-matched peers of all developmental levels. However, the VABS-II and the ABC rely on reports and opinions by parents or caregivers to assess the capabilities of the child. The subjective nature of this information may be inadvertently biased by the way in which parents and caregivers report behaviors or skills.
One of the very few standardized instruments that allows for discrimination among subjects with very low cognitive abilities is the BSID-III, which can discriminate among subjects performing at a developmental age up to 42 months old. Like most instruments that allow for discrimination in the low-functioning population, the BSID-III is not designed to be used in older children and adults. Nevertheless, since none of the subjects in our study were found to have a developmental age beyond 34 months in any of the developmental domains assessed by the BSID-III, we believe that this is a valid measure of developmental skills in this population. Moreover, the majority of our participants under 42 months of age scored at or near the floor of the test, yielding little to no variability in standard scores; hence DQs were calculated for all of our participants. Although it is recognized that developmental assessments in infancy and early childhood are not highly predictive of intelligence in later life due to the wide variability in that population,26,36
some authors have found that the predictability of developmental assessments increases with increasing severity of intellectual disabilities (i.e., they are more predictive of long-term intelligence among children who are more severely delayed).37
The use of DQ when the norms of a developmental instrument are not applicable to the population being assessed is a well-accepted practice, as DQ allows for the comparison of children at different ages and at different times28,38
, and it also provides for a broader range of possible scores.31
Moreover, BSID-III DQ scores have been found to correlate well with subsequent scores on the Stanford-Binet Intelligence Scale in a group of children with intellectual disabilities.29
Although this is one of the largest studies on the neurodevelopmental profile of individuals with AS, there are relatively few subjects with molecular etiologies other than deletions, which limits our ability to perform comparisons and genotype-phenotype correlations among the non-deletion sub-populations. We also recognize that the younger mean age of our deletion participants compared to the non-deletion participants may have confounded our analysis of the gross and fine motor abilities in these two sub-groups, because abnormal muscle tone can lead to delays in the acquisition of motor skills, and 73% of our participants up to the age of 24 months had abnormalities in muscle tone compared to 43% of those who were at least 42 months old.
Future studies should focus on the development of psychometric instruments that can discriminate among subjects who function at very low levels in each of the commonly-tested developmental domains (e.g. communication, motor skills, social skills, and activities of daily living), and the validation of such instruments in a wide age group from early infancy to adulthood. It is only with the availability and acceptance of such tools that accurate assessment of the developmental profile of children and adults with AS and other developmental disorders will be possible. There is also a need for large-scale longitudinal studies with subjects in a broad age group over an extended period of time, such as the ongoing AS Natural History study from which these baseline data were obtained. Such longitudinal evaluations will provide insights into the changes in the developmental profile that occur over time, and facilitate the creation of developmental growth curves for the cognitive, language, and motor functioning in AS individuals. This will enable providers and caretakers to understand the developmental trajectories of AS individuals and to anticipate level of functioning in adulthood.