Procedural learning deficit in children with Williams syndrome
Introduction
Williams syndrome (WS) is a genetic condition. It was first identified in 1961 by Williams and his colleagues. WS is typified by a number of severe medical anomalies: mental retardation, facial dysmorphology, abnormalities of the cardiovascular system, as well as of the renal, musculoskeletal, endocrine and other organ systems. At birth, WS infants are below average weight and often show no spontaneous sucking, with frequent vomiting during the first few months. WS children also suffer from hyperacusis, an unusual sensitivity to certain environmental sounds [20], [21], [24], [32], [34], [38], [41], [44]. They are extremely social and show marked friendliness to adults including strangers. Their sensorimotor development is very slow but then they reach an acceptable level of motor, cognitive and linguistic competence, and some of them are able to live autonomously as adults [53]. Incidence of this syndrome is estimated to be one in 20 000 live births. Until very recently it often went undiagnosed until adulthood [21]. With the use of new diagnostic tools, it may be found that the incidence of WS is actually higher than the above estimate. WS is genetic in origin and is associated with a microdeletion on chromosome 7q11.23 [8], [19], [20], [66], [67].
Previous studies claimed that the syndrome not only presents specific clinical features but also a major dissociation in cognitive functioning: (a) severe cognitive deficits co-occurring with relatively spared language; and (b) severe difficulties in spatial cognition and numbering but excellent facial recognition [4], [5], [6], [56], [76]. Other researchers suggest a more complex neuropsychological profile with non-homogeneous development not only across but also within cognitive domains [7], [28], [33], [49], [52]. In particular, some dissociation was demonstrated in the verbal domain, i.e. preserved phonological processing of words but severe impairment of grammatical and semantic aspects of language [41], [52], [70], [75], as well as in visuo-spatial abilities, where visual perception is relatively preserved and spatial processing is more severely impaired [1], [3], [6].
The dissociation in WS subjects between more preserved verbal competencies than visuo-spatial ones is also reflected in short-term memory. Wang and Bellugi [76] described a double dissociation in the memory span exhibited by a group of WS and a group of mental-age matched Down syndrome subjects. In fact, WS subjects had a larger verbal than a spatial span; the reversed performance pattern was observed in Down subjects [74]. However, verbal short-term memory is not completely preserved in WS. Vicari et al. [74] found normal word-length and phonological similarity effects but a null word frequency effect in the verbal span of these children. This documents a normally functioning articulatory loop but a poor contribution of lexical-semantic knowledge to the short-term maintenance of verbal stimuli. Instead, the visuo-spatial/verbal dichotomy was not respected in the long-term memory domain. Vicari et al. [73] reported impaired episodic retrieval of both verbal and visuo-perceptual stimuli in a group of WS children.
The present study was aimed at evaluating implicit memory processes in WS subjects. A memory is considered implicit when a previous experience or previously acquired information influences current behavior without the subject consciously recollecting the previous experience or information. Implicit memory is generally opposed to explicit memory, which, instead, underlies the conscious retrieval of previous experiences. According to Squire [63], implicit memory manifests in at least three different ways. The first, repetition priming, is reflected by an advantage (in terms of performance speed or accuracy) in reprocessing recently experienced stimuli over not experienced stimuli (for a review, see Schacter et al. [60]). The second type of implicit memory, procedural learning, is reflected in performance improvement (also in this case, in terms of speed or accuracy) in a particular task as a function of previous experience with the same task, also when stimuli different from those previously experienced are utilized (for a review, see Salmon and Butters [57]). Finally, conditioning refers to the formation of simple associations in which a neutral stimulus comes to elicit a certain motor response as a function of repeated reinforcement (e.g. McDonald and White [40]).
A problem frequently encountered in the experimental literature dealing with implicit memory is the definition of operational criteria to establish the real ≪implicit≫ nature of a particular memory task. In fact, although in typical implicit tasks the test instructions do not make reference to the fact that currently presented stimuli have been experienced in a recent past, the subjects may become aware of this overlap, and may decide to rely on explicit retrieval strategies for recovering the relevant information. Among the various operational criteria proposed (see Schacter et al. [59] for a review), the one that most directly addresses the issue of the independence of a memory task from the intentional retrieval of previous experiences is the demonstration that groups of subjects that differ in the efficiency of their explicit memory perform at the same level in a supposedly implicit task. This criterion has been extensively applied to neuropsychological patients with explicit memory disorders. The fact that adult amnesic patients, who are typically impaired in traditional memory tests of recall and recognition (requiring conscious access to stored memory representations), are generally normal in repetition priming and procedural learning tests [11], [14], [15], [57] is usually considered a straightforward demonstration that explicit memory plays a negligible role in these tasks.
The same kind of logic as that used for interpreting the dissociation between the performance of explicit and implicit memory tasks in adult amnesic patients can also be applied to developmental studies documenting discrepant maturation rates in the performance of explicit and implicit memory tasks in normal children. For example, it is widely documented that from 4 to 10 years of age, accuracy in explicit recall and recognition steadily improves; nevertheless, in the same age range the level of repetition priming, as revealed by Fragmented Pictures Identification [18], [47] or Stem Completion [13], [51] tests, remains substantially unchanged. Moreover, in studies dealing with mentally retarded individuals, a dissociation between poor scores on explicit memory measures and normal facilitation in re-processing visual or verbal stimuli [12], [36], [78] and in re-performing particular visuo-motor or cognitive tasks [42], [62], [69] has been frequently reported. On this line, Vicari, Bellucci and Carlesimo [72] recently investigated implicit and explicit memory abilities in subjects with Down syndrome. Consistent with previous findings in different etiology mentally retarded children [12], [50], the Down syndrome subjects had poor explicit memory but demonstrated repetition priming (both in the verbal and in the visuo-perceptual domain) and procedural learning (of both visuo-motor and cognitive skills) similar to that displayed by mental-age matched normal children. Thus, mental retardation does not seem to be a clinical condition able per se to cause difficulty in implicit memory. The possible finding of an impairment in implicit memory in some etiological group of mentally retarded people would suggest a specific deficit in cognitive architecture, probably as a reflection of some peculiar characteristic of their anomalous brain development, rather than of a generic effect of global mental impairment.
The evidence of normal performance in implicit memory tasks by adult amnesic patients demonstrates that mesio-temporal structures and diencephalic nuclei, which are typically lesioned in these patients, do not play any role in implicit memory processes [64]. Instead, based on neuropsychological evidence in adult neurological patients and on functional neuroimaging data in healthy humans, a specific role of basal ganglia and of cerebellum in the learning of visuo-motor and cognitive skills [29], [57] and of associative neocortex of posterior hemispheric regions in repetition priming [58] has been proposed. Basal ganglia would also be implicated in the formation of new associations underlying conditioning [57].
The investigation of brain morphology in WS and Down syndrome subjects by means of the Magnetic Resonance Imaging (MRI) technique revealed remarkable differences in the rate of development of some cerebral structures known to be involved in skill learning. In particular, in the context of a comparable overall reduction of brain volume, Down children show a disproportionately severe atrophy of cerebellum while WS children have markedly atrophic basal ganglia structures [30], [31]. Moreover, Rae et al. [54] recently reported MRI spectroscopy evidence of a decrease in the neuronal marker N-acetylaspartate in cerebellum of WS subjects. Interestingly, the cerebellar concentration of N-acetylaspartate in these subjects correlated with performance scores on a variety of tests exploring general intelligence, vocabulary and attention. To summarize, WS subjects suffer both from a morphological abnormality of basal ganglia and from a biochemical alteration of cerebellum, the brain areas most directly involved in skill learning. Therefore, these subjects seem to be particularly interesting for evaluating the role of basal ganglia and cerebellar development in the maturation of procedural learning.
Section snippets
Participants
Twelve persons with WS (five males, seven females), ranging from 11 to 19 years of age and with a mean mental age of 6 years and 5 months (SD=0.8) measured by the L-M form of the Stanford-Binet Intelligence Scale [9], were included in our study. Informed consent was obtained from patients and parents. All subjects took part in the larger longitudinal study on medical and neuropsychological features of WS conducted at the Children's Hospital Bambino Gesù in Santa Marinella, Rome with the support
Implicit memory
Performance scores of the two groups on the first and second execution of the Tower of London test (Fig. 2) were analyzed by means of a two-way Group×Block (first and second) mixed ANOVA. The Group effect was significant, F(1, 22)=18.55; P<0.001 as well as the Block effect, F(1, 22)=25.5; P<0.001 and the Group×Block interaction, F(1, 22)=5.4; P<0.05). Overall, performance of WS subjects was poorer than that of normal controls (M=20.2; SD=6.6 vs. M=28.1; SD=2.02). In the whole sample, a significant
Discussion
Consistent with previous reports on different etiology mentally retarded children [50], [65], [72], [78], in the present study WS subjects showed a level of repetition priming similar to that of mental-age matched normal controls. In fact, previous exposure to unfragmented stimuli facilitated normal and WS children to the same extent in the subsequent Word Stem Completion and Fragmented Pictures Test.
In contrast, WS children differed from normal controls in the two procedural learning tasks.
Acknowledgements
The financial support of Telethon-Italy (Grant no. E.C. 685) is gratefully acknowledged. We also wish to thank the children who participated in the study and their parents and two anonymous reviewers for their precious comments on an earlier version of this paper.
References (79)
- et al.
Expression analysis and protein localization of the human HPC-1/syntaxin 1A, a gene deleted in Williams syndrome
Genomics
(1999) Perceptual and conceptual priming in amnesic and alcoholic patients
Neuropsychologia
(1994)- et al.
Long term memory in mental retardation. Evidence for a specific impairment in subjects with Down's syndrome
Neuropsychologia
(1997) - et al.
Developmental dissociation between visual and auditory repetition priming: the role of input lexicons
Cortex
(2000) - et al.
Operant conditioning after temporal lobe lesions in man: conditional and simple discrimination learning
Cortex
(1993) - et al.
Procedural and declarative memory: a developmental study
Brain and Cognition
(1994) - et al.
Lim-Kinase-1 hemizygosity implicated in impaired visuospatial constructive cognition
Cell
(1996) Acquisition and transfer of declarative and procedural knowledge by memory impaired patients: a computer data-entry task
Neuropsychologia
(1992)- et al.
Pictorial priming and cued recall in Alzheimer's and Huntington's disease
Brain and Cognition
(1990) - et al.
Ontogeny is different for explicit and implicit memory in humans
Neuroscience Letters
(1993)