Prader-Willi syndrome

The speech and language of 55 individuals (27 males and 28 females) with Prader-Willi syndrome (PWS), aged from 6 months to 42 years, were examined through standardized testing and spontaneous speech sample analysis. While great variability was noted in speech and language abilities, most subjects presented with speech sound errors characterized by imprecise articulation (85%), and oral motor difficulties (91%). Hypernasality was noted in 62% and hyponasality in 14%. Other speech characteristics included a slow speaking

rate, flat intonation patterns, abnormal pitch of the voice, and harsh/hoarse voice quality. Narrative retelling abilities were poor, with specific deficits in sequencing of story events. Individuals with PWS as a result of deletions of chromosome 15 did not differ significantly in speech and language from individuals with PWS as a result of uniparental disomy.

Key Words: syndrome, genetic, speech, language, chromosome

Prader-Willi syndrome (PWS) is a complex multisystem genetic disorder that results in infantile hypotonia, developmental delays and mental retardation, behavioral disorders, characteristic facial appearance, obesity, hypogonadism, and short stature (Cassidy, 1997). Most individuals with PWS present with some type of speech and language deficit, although there is great variability in the type and severity of the disorder, ranging from individuals who are nonverbal to those who develop normal speech and language skills by adulthood. The speech-language pathologist (SLP) may become involved as early as birth to treat feeding difficulties and may continue to serve individuals with PWS for communication disorders throughout their lifespan.

The goals of the present article are to summarize the genetic and phenotypic characteristics of PWS, to examine the existing literature on the speech and language skills of PWS, and to report on the speech and language findings of a relatively large cohort of individuals with PWS. Previous descriptions of speech and language skills have been based on a small number of individuals and have not examined the developmental course of the speech and language disorder. Many studies have not distinguished between speech sound errors due to poor oral motor skills and structural deviations and errors due to phonological deficits. Further, studies have not attempted to associate speech and language characteristics to a particular causative genetic abnormality.

Genetic Basis of PWS

PWS results from the loss of function of genes that normally function only when they are inherited from the father (see Figure 1). Genes that function differently when they are inherited from the mother or the father are said to be imprinted. Imprinted genes appear to be involved in important developmental processes. PWS was the first recognized human genomic imprinting disorder (Nicholls, Knoll, Butler, Karam, & Lalande, 1989). PWS may result from one of three mechanisms leading to the loss of function of imprinted genes on chromosome 15 in the q11-- q13 region. First, the paternally contributed PWS region of chromosome 15 may be deleted. This random event happens spontaneously and is generally not familial. It accounts for approximately 70% of the cases of PWS. Secondly, maternal uniparental disomy (UPD) of chromosome 15 can cause PWS. This results when two copies of chromosome 15 are inherited from the mother and no copy from the father. This accounts for about 25% (Nicholls, Knoll, Butler, Karam, & Lalande, 1989) to 28% (Everman & Cassidy, 2000) of individuals with PWS. Finally a familial translocation, methylation defect or other structural abnormality of chromosome 15 or an error in the process of activating paternal genes (imprinting defect) in the PWS (15q11-13) region may result in PWS in 2% (Everman & Cassidy, 2000) to 5% (Nicholls, Knoll, Butler, Karam, & Lalande, 1989) of individuals. PWS may be confirmed by molecular genetic testing through a methylation test, which distinguishes which parent contributed each copy of chromosome 15. If only maternal genetic material is present, follow-up testing by fluorescence in situ hybridization (FISH) using probes within the PWS region can be done to detect a deletion of the paternal PWS region, and other tests can be done to determine if UPD (an extra copy of the maternal chromosome 15) or a translocation on the paternal chromosome is the cause (reviewed in Cassidy, 1997).

Phenotype of PWS

The incidence of PWS is 1 in 10,000 to 15,000 individuals, with males and females affected with equal frequency (Burd, Veseley, Martsolf, & Kerbeshian, 1990). Holm et al. (1993) have described clinical diagnostic criteria for Prader-Willi syndrome. These include neonatal and infantile hypotonia, feeding problems in infancy, obesity with onset between 12 months and 6 years, distinct facial features including a narrow bifrontal (forehead) diameter, almond shaped eyes, a narrow nasal bridge, and a down-turned mouth with a thin upper lip, hypogonadism, short stature, and cognitive and behavioral abnormalities. Many features associated with PWS may contribute to speech and language difficulties (see Table 1), however, hypotonia and cognitive impairments are cited as the chief contributory factors (Kleppe, Katayama, Shipley, & Foushee, 1990). Hypotonia may result in a slow rate of speech as well as imprecise articulation. Mental retardation associated with PWS (the average full-scale IQ is 65) may influence receptive and expressive language abilities. Other factors such as structural characteristics of the mouth and jaw including a narrow overjet, narrow palatal arch, and micrognathia (small lower jaw), may also influence articulation (Akefeldt, Akefeldt, & Gillberg, 1997). Velopharyngeal movement and closure may be inadequate, resulting in hypernasality or hyponasality due to hypotonia of muscles of the soft palate. Altered growth of the larynx due to endocrine dysfunction may affect pitch and voice quality. Reduced saliva output may cause dental decay and xerostomia (dry mouth), thus impairing articulation. Also, reduced saliva output may lead the speech-language pathologist to incorrectly assume that oral motor abilities are better than they actually are, as young children with PWS do not usually demonstrate excessive drooling typically associated with poor oral motor skills. Finally, behavioral disturbances associated with PWS-including verbal and physical aggression-may impact pragmatic language skills as well as compliance with therapy.

Speech and Language Characteristic of PWS

The speech and language skills of individuals with Prader-Willi syndrome are reported to be below expectations based on intellectual levels (Branson, 1981; Edmonston, 1982; Kleppe et al., 1990; Munson-Davis, 1988). Although great variability exists in the speech/ language skills of individuals with Prader-Willi, several common features have been noted. Table 1 illustrates some of the clinical features of PWS and their potential impact on speech and language. A distinguishing feature appears to be poor speech sound development (Dyson & Lombardino, 1989). Some authors have attributed this to poor oral motor skills, especially reduced tongue elevation for speech and slower alternating movements of the articulators (Akefeldt, Akefeldt, & Gillberg, 1997; Branson, 1981; Kleppe et al., 1990). Other authors (Dyson & Lombardino, 1989) have viewed poor phonological skills as a component of a more general language deficit. Downey and Knutson (1995) report that most individuals with PWS present with delayed speech sound development characterized by phonological patterns typical of younger normally developing children. Nevertheless, some individuals demonstrate atypical patterns such as a phonological disorder or an apraxia of speech (Munson-Davis,

1988). With improved genotyping techniques, it may be possible to correlate these various phenotypes with differing genotypes.

Besides articulation errors, hypernasality resulting from poor velopharyngeal function due to hypotonia, a highpitched voice, inadequate vocal intensity, and harsh/ hoarse voice quality have also been reported (Akefeldt, Akefeldt, & Gillberg, 1997; Downey & Knutson, 1995). Fluency disorders have not been reported in individuals with PWS.

Individuals with PWS frequently demonstrate poor receptive and expressive language skills, with expressive language more impaired than receptive skills (Branson, 1981; Kleppe et al., 1990; Munson- Davis, 1988). Several authors have described patterns of cognitive strengths and weaknesses frequently observed in PWS that might impact language abilities (Dykens, Hodapp, Walsh, & Nash, 1992). Specific deficits have been reported in auditory short-term memory (Dykens et al., 1992), linear or temporal order processing, and auditory verbal processing skills (Curfs, Wiegers, Sommers, Borghgraef, & Fryns, 1991). Narrative skills of individuals with PWS have not yet been described. Nonetheless, these specific cognitive deficits may impact narrative abilities. Pragmatic deficits including problems with maintaining a topic, judging appropriate proximity to the conversational partner, and turn taking have also been reported (Downey & Knutson, 1995).

Pragmatic skills may be influenced by a number of behavioral disturbances (Thompson, Butler, MacLean, & Joseph, 1996). For example, temper tantrums, compulsive behavior, and skin picking may interfere with peer relationships. Children with PWS have difficulty with social relationships. Sullivan & Tager-Flusberg (1998) demonstrated that children with PWS were less likely than IQ-matched children,with Williams syndrome to show appropriate empathetic responses. A study of 30 adults with PWS reported perseverative speech (Clarke, Boer, Chung, Sturmey, & Webb, 1996). Such pragmatic deficits may impede progress in therapy.

Strengths associated with PWS include vocabulary knowledge and reading decoding, as well as reading comprehension (Dykens, Hodapp, Walsh, & Nash, 1992). Visual spatial skills have also been reported as a relative strength (Dykens et al., 1992). However, variability has been noted in these patterns of strengths and weaknesses. Curfs, Wiegers, Sommers, Borghgraef, & Fryns (1991) reported that 10 of the 26 subjects with PWS had performance IQs at least 15 points higher than their verbal IQ, 3 had verbal IQs at least 15 points higher than their performance IQs, and 13 did not show any discrepancy. In summary, individuals with PWS syndrome present with speech, language, and cognitive deficits that impede their communication skills. While previous studies have described many of the speech and language characteristics of PWS, few studies reflect the heterogeneity of these speech and language deficits or describe these speech and language difficulties across the lifespan. Furthermore, differences in speech and language skills associated with specific genetic abnormalities (i.e. deletions or UPD) have not been described. Finally, narrative abilities of individuals with PWS have not been studied.

The present study reports on the speech and language abilities of a relatively large cohort of individuals with PWS (N = 55) who were seen as part of a specialty clinic at University Hospitals of Cleveland, Ohio. Given the low incidence of the syndrome, 55 individuals comprise a sizable group of affected individuals. Specific goals of the study were: (a) to describe the heterogeneity of speech and language abilities, including narrative skills, found in Prader-Willi syndrome; (b) to examine speech and language skills within four age groups: infant/ toddler, preschool, school age, and adolescent/adult; and (c) to examine speech and language findings with respect to different genetic abnormalities.

Method

Participants

The speech and language skills of fifty-five individuals (27 males and 28 females) with a diagnosis of Prader-Willi syndrome were examined as part of a visit to a specialty clinic serving individuals with Prader-Willi syndrome at University Hospitals of Cleveland, Ohio. The participants ranged in age from 6 months to 42 years. All participants were examined by a medical geneticist (the last author) and met the diagnostic criteria for Prader-Willi syndrome. Genetic testing confirmed the diagnosis in all cases. The participants and their primary caregivers agreed to be seen by two speech-language pathologists at the specialty clinic for observation and assessment when appropriate.

Due to time limitations of clinical visits and varying developmental levels of participants, not all measures were administered to all participants. We chose to include participants with missing data however, as our goal was to describe the heterogeneity of speech and language skills of individuals with PWS. If we reported only on participants who completed all measures, the more severely affected individuals would have been eliminated. See Table 2 for a summary of participants' demographics.

Participants were divided into four groups based on age: infant/toddler (under 2 years), preschool (2.0 to 5.0 years old), school age (5.1 to 12.0 years old), and adolescent/ adult (over 12 years old). Five children were under 2 years of age (6, 9, 18, 18, and 23 months old). Four were boys and one was a girl. All presented with receptive and expressive language delays on the Early Language Milestone Scale-2 (Coplan, 1993). Due to the young age of the children, conversational speech samples, narrative tasks, standardized tests, and oral motor protocols were not deemed appropriate.

Twenty-one children (14 boys and 7 girls) fell within the preschool category (M = 3.4 years; SD = 0.9 years; Range = 2.0 to 4.9 years). Because three children did not exhibit any connected speech, conversational and narrative analyses were precluded. All 18 remaining children contributed to the MLU measure from conversational analysis. In contrast, conversational samples yielded measures of vocal characteristics for only 12 tol6 of the children, due to too few utterances to allow for adequate assessment of vocal characteristics. Only 8 of the 18 children participated in the narrative analysis due to time constraints, limited cooperation, and low developmental level. Nine of the original 21 children lacked the ability to perform standardized testing, and one could not perform the oral motor protocol. Time constraints prevented another seven children from completing the full test battery and protocol. Therefore, the number of preschool children who contributed to a given measure from the published battery of tests and protocols varied from 5 to 20.

Fifteen children (6 boys and 9 girls) were school age (M = 7.6 years; SD = 1.6; Range = 5.4-10.5 years). Because one child was nonverbal, all planned assessment procedures were precluded. Of the 14 remaining children, only 12 and 8 were able to participate in conversational and narrative analyses, respectively, due to time constraints. Only 11 of the 12 children who participated in the conversational analysis contributed to MLU. A sufficient number of utterances was not elicited from one participant to allow for accurate calculation of MLU. Time constraints also prevented 7 of the 14 children from completing the full battery of published tests and oral motor protocols (particularly tests of semantics and intelligence). Therefore, the number of participants who contributed to a given measure from the published battery of tests and protocols varied from 7 to 13.

Fourteen participants (3 males and 11 females) were over 12 years of age (M = 27.3; SD = 11.6; Range = 14.3 to 56.0 years old). All participants were verbal, however one participant could not be assessed due to limited cooperation. Although all 13 remaining individuals participated in the conversational analysis, only 8 were able to participate in the narrative analysis due to time constraints. Time constraints also prevented 9 of the 13 participants from completing the full battery of published tests and the oral motor protocol (particularly tests of semantics and intelligence). Therefore, the number of participants who contributed to a given measure from the published battery of tests and protocols varied from 4 to 13.

Measures

Measures were selected to assess intelligence, articulation and phonology, oral motor skills, expressive grammar, prosody and voice characteristics, lexical comprehension and expression, and narrative abilities. The battery of standardized tests included: the Goldman-Fristoe Test of Articulation (GFTA; Goldman & Fristoe, 1986), the Khan-- Lewis Phonological Process Analysis (Khan & Lewis, 1986), the Peabody Picture Vocabulary Test-R (PPVT-R; Dunn & Dunn, 1981), the Expressive One-Word Picture Vocabulary Test-Revised (EOWPVT-R; Gardner, 1990), and the Stanford Binet Intelligence Scale, 4th ed. (Thorndike, Hagen, & Sattler, 1986). Conversational speech samples were also analyzed. Because individuals received various tests based on their age and intellectual abilities, rating scales were adopted to summarize data across age and skill levels. Standardized tests, rating scales, and narrative measures are described below.

Intelligence. Full-scale IQ scores were obtained by the Stanford Binet Intelligence Scale, Fourth Edition (Thorndike, Hagen, & Sattler, 1986). A licensed psychologist administered all IQ tests.

Articulation and Phonology. The Goldman-Fristoe Test of Articulation (GFTA; Goldman & Fristoe, 1986) was used to assess production of consonant sounds in the initial, medial, and final position of single words and blends in individuals over 2 years of age. A percentile score was entered into data analysis. Errors were further classified by the Khan-Lewis Phonological Process Analysis (Khan & Lewis, 1986) and a severity rating of normal, mild, moderate, and severe was assigned based on the descriptive categories for the composite scales described in the test manual (p. 41). In the present system, normal = insignificant (a percentile rank of 81-99), mild minimal or average (a percentile rank of 41-80), moderate = moderate (a percentile rank of 21-40), and severe excessive (a percentile rank of 1-20).

Oral Motor Skills. Oral motor skills were assessed by the Oral Speech Motor Protocol (OSMP; Robbins & Klee, 1987) for children 2 years, 6 months to 7 years old, and the Test of Oral Structures and Functions (TOSF; Vitali, 1986) for individuals over 7 years old. The TOSF assigns a scale score that describes functioning as unremarkable (0-69), borderline (70-79), impaired (80-89), or markedly impaired (90-99). Based on these measures participants were assigned a rating of normal = unremarkable, mild = borderline, moderate = impaired, or severe

impairment = markedly impaired for oral motor skills. The OSMP does not assign a rating scale, per se. Based on the z-scores obtained from developmental normative data the following ratings were assigned > -1 normal, between -1 and -2 = mild, between -2 and -3 = moderate, and between -3 and -4 = severe.

Expressive grammar. Spontaneous conversational samples were obtained using technical and interlocuter procedures for free speech sampling developed by Shriberg and Kwiatkowski (1985) and language sampling methods described by Miller (1981). Each sample was audio-tape recorded and transcribed by a licensed and certified speech-language pathologist. Fifty utterances were obtained. Conversational samples were used to compute the mean length of utterance (MLU) in morphemes according to the procedures described by Chapman (in Miller, 1981, pp. 22-27).

ProsodY and Voice Characteristics. Conversational samples were also used to rate voice characteristics including pitch, quality, and resonance, employing the Voice Rating Severity Scale described by Boone and McFarlane (1988). Pitch was rated as normal, high, or low for the participant's age and gender. Resonance characteristics of the speech sample were rated as normal, hypernasal, or hyponasal. Vocal quality was rated as normal, soft, harsh, hoarse, or strained. A monotone quality to the spontaneous speech sample was also noted. The vocal characteristics of each participant were rated independently by two licensed and certified speechlanguage pathologists (the first and second authors). Interrater agreement (i.e., assignment of identical rating categories) was achieved 92% of the time for ratings of vocal quality. When there was a discrepancy, the tapes were reviewed by both speech language pathologists until consensus was achieved.

Lexical Comprehension and Expression. The Peabody Picture Vocabulary Test-R (PPVT-R; Dunn & Dunn, 1981) and the Expressive One-Word Picture Vocabulary Test-Revised (EOWPVT-R; Gardner, 1990) were selected as measures of lexical comprehension and expression. These measures were chosen because they provide normative data from preschool through adulthood and therefore individuals of different ages could be compared. Other standardized measures of overall language abilities were administered. Nevertheless, as these measures varied with the age and the cognitive abilities of the subject, group data are not presented in this report.

Narrative Abilities. Measures also included a narrative retelling task, The Fox and Bear Story (Liles, 1985). In this task a simple story is read to the participant and the participant is asked to retell it. Responses were audio-tape recorded. Narratives were coded for the inclusion of story grammar components and content items (with a maximum of 14 components and 26 items scored, respectively). Three factual and three inferential questions were asked to determine comprehension. The percent of items correct for story grammar components, content items, inferential, and factual questions were entered into data analysis.

Results

All participants reported a history of communication difficulties, current enrollment in speech-language therapy, or both. As shown in Table 2, the majority of the infant/ toddlers (60%), preschoolers (86%), and school-age participants (87%) were receiving speech-language therapy. Three adolescent/adult participants (21%) were enrolled in speech-language therapy. Scores for standardized tests appear in Table 3. Ratings based on protocols and tests appear in Table 4.

Intelligence

Full-scale IQ, as measured by the Stanford Binet Intelligence Scale, Fourth Edition (Thorndike, Hagen, & Sattler, 1986), was determined for five preschool children (M = 92; SD = 11), seven school-age children (M = 73; SD = 9), and eight adolescent/adults (M = 62; SD = 13).

Oral Motor Skills and Articulation

Oral motor skills were measured for 46 of the 55 potential participants (see Table 4). The majority of participants (91%) demonstrated some degree of motor deficit (mild to severe) including poor tongue mobility, shortness of the palate, and incoordination of the articulators. All children under the age of 12 (i.e., the infant/toddler, preschool, and school age groups) demonstrated oral motor difficulties, and only 4 of 13 adolescents/adults demonstrated normal oral motor abilities.

Articulation and phonological skills were measured for 46 of the 55 potential participants respectively. Mild to severe articulation impairment was observed in 85% (n = 39) of 46 participants, with younger participants more severely impaired (see Table 4). All preschool children were rated moderately or severely impaired in articulation. Only 3 (21 %) of the 14 school age children and 4 (31 %) of the 13 adolescent/adults who contributed a score had achieved normal articulation skills. As can be seen in Table 3, mean percentile scores on the GFTA were below the 50th percentile for all age groups (M = 24.6 percentile; SD = 37.8; Range =

Voice 1 Conversational sample analysis revealed disorders of pitch, resonance, and quality. As shown in Table 4, 20% of participants (8 of 41) presented with a high pitch and 24% (10 of 41 participants) with a low pitch relative to their age and gender. Hypernasality was noted in 62% of the participants (23 of 37) and hyponasality in 14% (5 of 37 participants). Voice quality was often judged to be harsh (26%, or 10 of 39 participants) or hoarse (13%, or 5 of 39 participants).

Language

As displayed in Table 3, on the average MLU appears low for what we expect from normally functioning individuals in these age ranges (see Leadholm & Miller, 1992, for typical values from ages 3 to 13 years based on samples of 100 utterances). Vocabulary was assessed by the PPVT-R (standard score M = 70.8; SD = 17.6; standard score range = 40-98) and the EOWPVT-R (standard score M = 87.4; SD = 21.2; standard score range = 56-131). Vocabulary was found to be impaired, although great variability was noted, with some participants scoring within the normal range. Substantial deficits were noted in narrative retelling abilities. Difficulties with inclusion of story grammar components, mention of content items and story comprehension were observed. Mean scores on the Fox and Bear narrative retelling task by age group are presented in Table 5. Children in the preschool group were essentially unable to retell the story or answer questions concerning the story. Narrative abilities improved with age but remained impaired. Older participants (the adolescent/ adult group) on the average used only 22% of the story grammar components (approximately 3 components out of 14) and 46% of the content items (approximately 12 items out of 26). Comprehension difficulties remained into adolescence or adulthood, also, with only 33% of the inferential questions and 46% of the factual questions answered correctly.

Comparisons among UPD and 15q deletions

Five one-way analysis of variance (ANOVA) were conducted with age as a covariate, to make comparisons between the UPD and deletion groups because the ages of the participants in the two groups were significantly different. The participants with the translocation chromosome type (n = 3) were not considered in these analyses due to the small number of subjects with this chromosome anomaly. Not all participants contributed scores for each of the five ANOVAs. The mean age of the deletion group contributing to at least one of the five ANOVAs (n = 24) was 7.8 years, whereas the mean age of the UPD group contributing to at least one of the five ANOVAs (n = 17) was 15.4 years [t(1, 50) = 2.89, p = .03]. As shown in Table 6, the deletion group differed significantly from the UPD on the GFTA, and MLU. Effect sizes were small. These differences may have been related to group differences in age of the participants. Individual comparisons of five participants from the deletion group to an age matched participant from the UPD group failed to show differences between genetic types.

Discussion

The first goal of this project was to describe the speech and language characteristics of a relatively large clinical population of individuals with PWS. Although this syndrome has a low incidence, it is of particular relevance to speech-language pathologists as most individuals with PWS have speech and language difficulties, some of which persist into adulthood. All participants reported a history of speech and language difficulties and most preschoolers (86%) and school-age children (87%) were enrolled in speech-language therapy. Twenty-one percent of the adolescents/adult group continued to receive therapy. Because the present project surveyed a clinical population, participants differed in age, developmental levels, and the degree to which they were able to complete standardized testing. Whereas the heterogeneity of the participants may be a limitation of the study, it also served to demonstrate the variability of skills found in this population. The findings of the present study confirm that individuals with Prader-Willi syndrome demonstrate a variety of speech and language difficulties.

Poor oral motor and articulation skills appear to be hallmarks of PWS. Most participants (91%) presented with oral motor impairments that impacted articulation skills. Half of the preschool children presented with mild oral motor impairments while the other half demonstrated moderate to severe impairments. No preschool child presented with normal oral motor skills. Impaired oral motor abilities persisted at school age, with half of the school age children presenting with mild impairments and half with moderate impairments. Less than a third of the adolescents and adults achieved normal oral motor skills.

All groups performed below the 50th percentile on the GFTA, with all preschool children exhibiting moderate (37%) to severe (63%) impairments when the KLPA was applied to the GFTA data. Few older individuals (three school age children and four adolescent/adults) achieved normal articulation. Fifty percent of the school-age children and 23% of the adolescents/adults presented with severe articulation disorders. Clinically, we have noted that errors include imprecise articulation, sound distortions and omissions, vowel errors, and difficulty sequencing syllables.

Besides articulation difficulties, many participants exhibited abnormal vocal characteristics. Disorders of pitch included both a high pitch (20%) and a low pitch (24%) for age and gender. Nevertheless, not all participants demonstrated inappropriate pitch, with 44% of preschoolers, 75% of school-age children, and 54% of adolescents/adults receiving a normal pitch rating. An unexpected finding not previously reported in the literature is that of hyponasality (14% of individuals) as well as hypernasality (62%). Unlike previous studies (Akefeldt et al., 1997) which report a distinctive voice and resonance characteristic (namely hypernasality and a high pitch) we found that some individuals presented with hyponasality and a low-pitched voice. Hypernasality was most common in the preschool group with 75% of children demonstrating hypernasality. However, hypernasality continued to be prevalent in both the school-age (58%) and adolescent/adult (54%) groups. Both hypernasality and hyponasality may be attributed to poor velopharyngeal movement due to hypotonia. Variations in pitch may be due to altered growth of the larynx. Clinically, we have noted that the use of growth hormone in the Prader-Willi syndrome population may alter vocal characteristics. Since the use of growth hormone with PWS is relatively new, additional research is needed to determine its effect on pitch and resonance.

This study also demonstrated reduced language skills for all groups. Mean standard scores for all groups were below 100 on the PPVT-R (preschool M = 79, SD = 19; school age M = 71, SD = 17; adolescent/adult M = 64, SD = 18) and the EOWPVT-R (preschool M = 81, SD = 18; school age M = 90, SD = 13; adolescent/adult M = 91, SD = 39). Nevertheless, some participants in each group scored in the average range (as indicated by the SD values for most of the groups). This suggests that language skills are variable in individuals with PWS. The mean values for MLU were also substantially reduced based on age expectations for all groups. Again, it should be noted that some individuals in each group did produce an ageappropriate MLU (as indicated by the SD values, of Leadholm & Miller, 1992). Another finding, not previously noted in the literature, was a particular weakness in narrative abilities. Children in the preschool group were essentially unable to retell the story or answer questions regarding it. School-age children only used, on the average, 2% of the story grammar components and 4% of the content items. They answered 12% of the inferential questions and 17% of the factual questions correctly. Surprisingly, even adolescents and adults with relatively good communication skills had difficulty with the narrative task, including fewer than half of the story grammar components or content items in their retelling and answering fewer than half of the questions correctly. Thus, while narrative skills continue to develop into adulthood, they appear to lag behind other language skills. Several cognitive deficits have been associated with PWS that may impact narrative abilities, including poor auditory short term memory (Dykens, Hodapp, Walsh, & Nash, 1992), deficits in linear or temporal order processing (Dykens et al., 1992), and poor auditory verbal processing skills (Curfs et al., 1991). Further research is needed to examine narrative abilities in relation to specific cognitive deficits. Because narrative skills impact conversational ability, an area that is often described by caretakers as difficult for individuals with PWS, it may be advantageous for speech-- language pathologists to incorporate work on narrative abilities into their therapy programs.

A second goal of this study was to employ a crosssectional design to examine how speech and language skills change across age groups. While skills generally improve with age, few individuals with PWS achieve normal speech and language abilities.

Although variability appears to be the rule rather than the exception in the study of speech and language skills of individuals with Prader-Willi syndrome, a typical developmental course may be proposed based on the findings of the present study and our own clinical experience. The infant/toddler (6 months to 2 years) presents with a weak cry, early feeding difficulties, hypotonia, and delayed onset of speech/language. In this study, only a few single words and no two-word combinations were noted in children between 18 months and 2 years old. During the preschool years (2 to 5 years), first words appear and the child acquires vocabulary, two-word combinations and some syntax (as reflected in PPVT-R, EOWPVT-R, and MLU scores). Oral motor skills remain poor, and speech is unintelligible in part, with many sound errors. Preschool often marks the onset of obesity and behavioral disorders. During the school years (5-12 years), residual articulation errors are noted and the child is less intelligible in connected speech. Receptive/expressive language skills lag behind those of peers (based on the present PPVT-R, EOWPVT-R, and MLU scores). Hypernasality or hyponasality may be observed as well as pitch and voice quality difficulties. Narrative abilities are poor. Behavioral disturbances may result in difficulty in social interactions. The adolescent and adult (12 years to adulthood) continue to present with some articulation errors, as well as receptive and expressive language deficits. Narrative difficulties and poor conversational skills are notable. Pitch and resonance characteristics may remain abnormal.

A final goal of this study was to compare individuals with PWS due to a deletion of 15 q11-13 and individuals with PWS due to maternal uniparental disomy (UPD) of chromosome 15. The results of this comparison must be viewed as preliminary, because the UPD group was significantly older than the deletion group. Age was employed as a covariate in the statistical analysis in an attempt to correct for group differences in age. No significant differences were found between groups on measures of IQ, the PPVT-R, or the EOWPVT-R. Differences in the GFTA and the MLU were probably related to age differences in the two groups. Clinically, we noted there was a tendency for the deletion group to demonstrate more severe articulation and oral motor impairments. Nevertheless, further research is needed with better age-matched groups to determine if phenotypic differences may be related to genotypic differences.

In conclusion, the individual with Prader-Willi syndrome will require the services of a speech-language pathologist from infancy through adulthood. It is essential that the speech-language pathologist be aware of the unique characteristics of the syndrome that may have an impact on speech-language development. The speech-- language pathologist should also understand the wide range of communication disorders that are associated with Prader-Willi syndrome and assess each individual's needs. Further research ig needed to determine if there are associations between the type of genetic abnormality and the clinical presentation. As new medical treatments are employed with the Prader-Willi syndrome population, such as the use of growth hormone and surgical procedures to reduce hypernasality, continued research is needed to determine their impact on speech and language.

Acknowledgments

We wish to express our appreciation to the individuals with Prader-Willi syndrome and their families who generously agreed to participate and to Francesca Sieg and Dianne Pacella who assisted in data collection.

References

Akefeldt, A., Akefeldt, B., & Gillberg, C. (1997). Voice, speech and language characteristics of children with Prader-Willi syndrome. Journal of Intellectual Disability Research, 41(4), 302-311.

Boone, D. R., & McFarlane, S. C. (1988). The voice and voice therapy (4th ed.). Englewood Cliffs, NJ: Prentice Hall.

Branson, C. (1981). Speech and language characteristics of children with Prader-Willi Syndrome. In V.A. Holm, S. Sulzbacher, & P.L. Pipes (Eds.), The Prader-Willi Syndrome (pp. 179-183). Baltimore, MD: University Park Press.

Burd, L., Veseley, B., Martsolf, J., & Kerbeshian, J. (1990). Prevalence study of Prader-Willi syndrome in North Dakota. American Journal of Medical Genetics, 37, 97-99.

Cassidy, S. B. (1997). Prader-Willi syndrome. Journal of Medical Genetics, 34, 917-923.

Chapman, R. S. (1981). Computing mean length of utterance in morphems. In J.F. Miller, Assessing Language Production in Children (pp. 22-27). Austin, TX: Pro-Ed.

Clarke, D. J., Boer, H., Chung, M .C., Sturmey, P., & Webb, T. (1996). Maladaptive behavior in Prader-Willi syndrome in adult life. Journal of Intellectual Disability Research, 40, 159-165.

Coplan, J. (1993). Early Language Milestone Scale-Second Edition. Austin, TX: Pro-Ed.

Curfs, L. M., Wiegers, A. M., Sommers, J. R., Borghgraef, M., & Fryns, J. P. (1991). Strengths and weaknesses in the cognitive profile of youngsters with Prader-Willi syndrome. Clinical Genetics, 40, 430-434.

Downey, D. A., & Knutson, C. L. (1995). Speech and language issues. In Greenswag, L. R. & Alexander, R.C. (Eds.), Management of Prader- Willi syndrome (2nd ed.). New York: Springer-Verlag.

Dunn, L. M., & Dunn, L. M. (1981). Peabody Picture Vocabulary Test-Revised. Circle Pines, MN: American Guidance Service.

Dykens, E. M., Hodapp, R. M., Walsh, K., & Nash, L. J. (1992). Profiles, correlates, and trajectories of intelligence in Prader-Willi syndrome. Journal of the American Academy of Child & Adolescent Psychiatry, 31, 1125-1130.

Dyson, A. T., & Lombardino, L. J. (1989). Phonological abilities of a preschool child with Prader-Willi syndrome. Journal of Speech and Hearing Disorders, 54, 44-48.

Edmonston, N. K. (1982). Management of speech and language impairment in a case of Prader-Willi Syndrome. Language, Speech, and Hearing Services in the Schools, 13, 241-245.

Everman, D. B., & Cassidy, S. B. (2000). Genetics of childhood disorders XII: Genomic imprinting: It doesn't follow the rules. Journal of the American Academy of Child and Adolescent Psychiatry, 39, 386-389.

Gardner, M. F. (1990). Expressive One-Word Picture Vocabulary Test. Novato, CA: Academic Therapy Publications. Goldman, R., & Fristoe, M. (1986). The Goldman-Fristoe Test of

Articulation. Circle Pines, MN: American Guidance Service. Holm, V. A., Cassidy, S. B., Butler, M. G., Hanchett, J. M., Greenswag, L. R., Whitman, B. Y., & Greenberg, F. (1993). Prader-Willi syndrome: Consensus diagnostic criteria. Pediatrics, 91, 398-402.

Khan, L. M. L., & Lewis, N. P. (1986). Khan-Lewis Phonological Analysis. Circle Pines, MN: American Guidance Service, Inc.

Kleppe, S. A., Katayama, K. M., Shipley, K. G., & Foushee, D. R. (1990). The speech and language characteristics of children with Prader-Willi Syndrome. Journal of Speech and Hearing Disorders, 55, 300-309.

Leadholm, B. J., & Miller, J. F. (1992). Language Sample Analysis: The Wisconsin Guide. Madison, WI: Wisconsin Department of Public Instruction.

Liles, B. Z. (1985). Cohesion in the narratives of normal and language disordered children. Journal of Speech and Hearing Research, 28, 123-133.

Miller, J. F. (1981). Assessing Language Production in Children. Austin, TX: Pro-Ed.

Munson-Davis, J. A. (1988). Speech and language development in Prader-Willi Syndrome. In L.R. Greenswag & R.C. Alexander (Eds.), Management of Prader-Willi syndrome (pp. 124-133). New York: Springer-Verlag.

Nicholls, R. D., Knoll, J. H. M., Butler, M. G., Karam, S., & Lalande, M. (1989). Genetic imprinting suggested by maternal heterodisomy in non-deletion Prader-Willi syndrome. Nature, 342, 281-285.

Robbins, J., & Klee, T. (1987). Clinical assessment of oropharyngeal motor development in young children. Journal of Speech and Hearing Research, 52, 271-277.

Shriberg, L. D., & Kwiatkowski, J. (1985). Continuous speech sampling for phonologic analyses of speech delayed children. Journal of Speech and Hearing Disorders, 50, 323-334.

Sullivan, K., & Tager-Flusberg, H. (1998). Empathetic responses of young children with Prader-Willi syndrome to the distress of others. Paper presented at the 13th annual Prader-Willi syndrome association scientific conference, Columbus, Ohio, July.

Thompson, T., Butler, M. G., MacLean, W. E., & Joseph, B. (1996). Prader-Willi syndrome: genetics and behavior. Peabody Journal of Education, 71, 187-212.

Thorndike, R. L., Hagen, E. P., & Sattler, J. M. (1986). The Stanford Binet Intelligence Scale-Fourth Edition. Chicago, IL: Riverside Publishers.

Vitali, G. J. (1986). Test of Oral Structures and Functions. East Aurora, NY: Slosson Education Publications, Inc.

Received September 6, 2000

Accepted November 12, 2001

DOI: 10.1044/1058-0360(2002/033)

Contact author: Barbara A. Lewis, PhD, Department of Pediatrics, Case Western Reserve, University, 11100 Euclid Avenue,

Cleveland, Ohio 44106-6038. E-mail: bxl@po.cwru.edu

Barbara A. Lewis

Lisa Freebairn

Rainbow Babies and Children's Hospital

Case Western Reserve University, Cleveland

Shauna Heeger

Suzanne B. Cassidy

University of California, Irvine

Copyright American Speech-Language-Hearing Association Aug 2002
Provided by ProQuest Information and Learning Company. All rights Reserved