The Neurocognitive Profile of Williams Syndrome: A Complex Pattern of Strengths and Weaknesses.

Abstract

* The rare, genetically based disorder, Williams syndrome (WMS), produces a constellation of distinctive cognitive, neuroanatomical, and electrophysiological features which we explore through the series of studies reported here. In this paper, we focus primarily on the cognitive characteristics of WMS and begin to forge links among these characteristics, the brain, and the genetic basis of the disorder. The distinctive cognitive profile of individuals with WMS includes relative strengths in language and facial processing and profound impairment in spatial cognition. The cognitive profile of abilities, including what is 'typical' for individuals with WMS is discussed, but we also highlight areas of variability across the group of individuals with WMS that we have studied. Although the overall cognitive abilities (IQs) of individuals with WMS are typically in the mild-to-moderate range of mental retardation, the peaks and valleys within different cognitive domains make this syndrome especially intriguing to stu dy across levels. Understanding the brain basis (and ultimately the genetic basis) for higher cognitive functioning is the goal we have begun to undertake with this line of interdisciplinary research.

INTRODUCTION

Crystal, an ambitious 14-year-old, was overheard to say "You're looking at a professional book writer. My books will be filled with drama, action, and excitement. Everyone will want to read them. I'm going to write books, page after page, stack after stack." Crystal is quite good at creating original stories on a moment's notice, most recently spinning a tale about a chocolate princess who saves her chocolate kingdom from melting by changing the color of the sun. Remarkably, her creative talents are not limited to storytelling, but extend to music as well; she has composed the lyrics and music to a song. Considering her ease with language, her creative ideas, and her unshaking enthusiasm, her ambition to become a writer may seem plausible -- however, Crystal has an IQ of 49. She fails all Piagetian seriation and conservation tasks, milestones normally attained by age 8. She has reading, writing, and math skills comparable to those of a first-grader, demonstrates visuo-spatial abilities of a 5-year-old, and cann ot be left alone without a babysitter.

Crystal has Williams syndrome (WMS), a rare (1 in 25,000) genetically based neurodevelopmental disorder. Characteristics of the syndrome include specific face and physical features; a variety of cardiovascular difficulties, including supravalvular aortic stenosis (SVAS); failure to thrive in infancy; transient-neonatal hypercalcemia; delayed language and motor development; and abnormal sensitivities to certain classes of sounds (hyperacusis) (Lenhoff, Wang, Greenberg, & Bellugi, 1997; Marriage & Scientist, 1995; Bellugi & Morris, 1995). The precise genetic underpinnings of WMS are becoming clear and are currently known to involve a submicroscopic deletion of one copy of about 20 contiguous genes on chromosome 7, including the gene for elastin (Korenberg et al., this volume; Frangiskakis et al., 1996).

Work being performed in our lab and others is mapping out the cognitive profile of WMS and its distinctive identifying features (Mervis, Morris, Bertrand, & Robinson, 1999; Karmiloff-Smith, 1998; Volterra, Capirci, Pezzini, Sabbadini, & Vicari, 1996; Bellugi, Bihrle, Neville, Jernigan, & Doherty, 1992; Bellugi, Wang, & Jernigan, 1994; Bellugi, Hickok, Jones, & Jernigan, 1996a; Bellugi, Klima, & Wang, 1996b; Bellugi & Wang, 1998; Bellugi, Lichtenberger, Mills, Galaburda, & Karenberg, 1999a, among others). Interest in Williams syndrome arises from the uneven cognitive profile that is associated with the syndrome through these studies, including specific dissociations in cognitive functions. In this paper, we will review a series of formal and informal studies of cognitive behavior of individuals with Williams syndrome, largely from our laboratory over the past decade.

In this special issue as a whole, results are described across an array of disciplines from cognitive neuroscience to molecular genetics. This chapter and the next (Jones et al., this volume) sketch out some of the findings from studies of neurocognitive and social behavior in the same set of Williams individuals. Other chapters involve studies using brain-imaging techniques including event-related potentials (Mills et al., this volume) and magnetic-resonance imaging (Reiss et al., this volume) with the same subjects. The studies include brain cytoarchitectonics (Galaburda & Bellugi, this volume) as well as molecular genetics (Korenberg et al., this volume). The special strength of the program project is that (except for the cytoarchitectonic studies), the same subjects undergo cognitive, neurophysiological, neuromorphological, and molecular genetic probes. In this way, we can begin to link phenotype and genotype, as well as to link variability at one level with variability at other levels, in a large group of well-defined subjects. The chapters in this special issue show the linkages among cognitive, neurobiological, phenotypic, and genotypic profiles in order to create an initial picture of the functional neuroarchitecture of the syndrome. This linking occurs in several ways. By drawing on known connections between neurobiological systems and cognitive functions, we have begun to match cognitive abnormalities with their probable bases in neurobiological abnormalities. In addition, individual variation within the WMS population can be capitalized upon by predicting correlations between the strength of neurophysiological markers (in terms of event related potentials -- ERPs) and performance on specific behavioral measures. Distinctive profile characteristics at the neurophysiological level (such as abnormal neurophysiological responses to face and language processing) can inform and refine our picture of aspects of the cognitive profile. Finally, we can take initial steps in the process of linking the presence or absence of copies of a small set of specific genes to the development of brain structure and function as well as to the specific cognitive profile of WMS.

Subjects

The groups of subjects with WMS who participated in the studies reported here involve a database of about 100 clinically and genetically diagnosed individuals with Williams syndrome, and are compared with Down Syndrome (DNS) individuals and normal controls, as well as with other specific disorders. Adolescent and adult individuals with WMS or DNS are part of a Program Project at The Salk Institute from the National Institute of Child Health and Human Development, "Williams Syndrome: Bridging Cognition and Gene." Infants and young children are part of a collaborative project with the Center for Neurodevelopmental studies at the University of California-San Diego (National Institute of Neurological Diseases and Stroke; and National Institute of Deafness and Other Communication Disorders). These studies include not only children with WMS or DNS syndrome, but also other projects including children with early onset focal lesions to the right or to the left hemisphere, with language impairment or autism. Subjects from the NICHD Program Project range in age from 10 and up and those from the Center for Neurodevelopmental studies range in age from birth to 10 years. Subjects are recruited for the WMS studies through our extensive contacts with families, the Williams Syndrome Association, national and regional conferences, private physicians, geneticists, cardiologists, and others that are familiar with the research in our laboratory. All are thoroughly screened prior to induction into the study and must pass a set of clear inclusionary and exclusionary criteria. Information about each subject's medical history is obtained from medical records, including a medical genetic evaluation confirming the diagnosis of WMS. In addition, a diagnostic interview, which tests for the common phenotypic features of the syndrome, is conducted with each subject and his/her caregiver. As part of this interview, a Diagnostic Score sheet (developed by the Medical Advisory Board of the Williams Syndrome Association) is completed (see Table 1 for a summary of some of the major diagnostic medical characteristics of WMS).

Molecular genetic testing (fluorescence in situ hybridization, or FISH) can now be used to confirm the deletion of one copy of the elastin gene and other surrounding genes in a small region of chromosome 7, characteristic of nearly all individuals with clinically diagnosed WMS. The studies in this paper and in subsequent papers in this volume (Jones et al., this volume; Korenberg et al., this volume; Mills et al., this volume; Reiss et al., this volume) include subjects who meet all aspects of our strict screening and diagnostic criteria for Williams or Down syndrome and, importantly, the same individuals undergo probes at these major levels.

Throughout these studies of children and adults with WMS, we report on a variety of comparison groups: Normal individuals who are variously matched with WMS individuals on chronological age, mental age, or language age, and individuals with DNS who are matched in age and Full Scale IQ to the WMS group. As contrasts to the WMS data, we additionally report data from infants or young children with language impairment, with early focal lesions, and individuals with autism. The majority of the initial results we report in this paper consist of data collected from WMS subjects and from those with DNS, the latter primarily as a comparison group, matched in age and IQ, as well as normal controls matched for chronological age or mental age. Williams and Down groups were chosen initially because both are genetically based disorders resulting in mental retardation, but other comparison groups are important as well.

General Cognitive Functioning

Across an array of standardized conceptual and problem-solving tasks (some verbal, some nonverbal in nature), subjects with WMS demonstrate a consistent, serious impairment in general cognitive functioning. On general cognitive tasks such as IQ probes, most individuals with WMS rank in the 'mild-to-moderate mentally retarded' range, with global standard scores on IQ tests ranging from 40 to 90 and a mean of around 55 (Bellugi et al., 1996b). Figure 1 contrasts the distribution of Wechsler Full Scale IQ (Wechsler, 1974; Wechsler, 1981) of 82 subjects with WMS, with a typical normal distribution of Full Scale IQ scores. We can now describe what is "average" or "typical" and what is the distribution for WMS individuals on many standardized tests. As seen in the range of performance on IQ tests shown in Figure 1, there is also some variability within WMS as a group, with the WMS group mean shifted downward from the normal distribution into the mild-to-moderate range of mental retardation. Reflecting the variabili ty in cognitive functioning, some adults with WMS live independently or semi-independently (Udwin, 1990), while others need significant help. It should be noted that arithmetic is an area of great difficulty for most individuals with WMS, but some are able to master addition and, in a small number of cases, subtraction and division as well. Reading is a challenge for some, while others have been noted to be avid readers of books, magazines, and newspapers, but often on very specific topics of interest (Howlin, Davies, & Udwin, 1998).

The typical global cognitive impairment that is seen in WMS is similar to that found in DNS in our studies.

As shown in Figure 2, adolescents with WMS and those with DNS in our studies score equally poorly across the board on IQ tests such as the Wechsler Intelligence Scale for Children-Revised (WISC-R; Wechsler, 1974) or Wechsler Adult Intelligence Scale-Revised (WAIS-R; Wechsler, 1981). In each of these subject groups, there was no clinically significant difference between Verbal and Performance IQ scores on the …