Early speech motor and language skills in childhood apraxia of speech: evidence for a core deficit in speech motor control?

Abstract

Children with childhood apraxia of speech (CAS) present with significant speech production deficits, the effects of which often persist well into late childhood (American Speech-Language-Hearing Association, 2007; Lewis, Freebairn, Hansen, Iyengar, & Taylor, 2004). Debate has historically surrounded whether the features of CAS are the result of an impairment in linguistic or speech motor systems, or both (American Speech-Language-Hearing Association, 2007). Most research, however, has failed to explicitly consider a developmental perspective of the disorder, arguably limiting the associated interpretations that often (implicitly) assume an established underlying system (Maassen, 2002). One of the key tenets of such a developmental perspective is the possibility of an original core deficit in one system, with negative consequences for aspects of the system that subsequently develop.A mixed-methodology paradigm was employed in the present research in order to explore the core deficit in CAS. Similar paradigms have been applied to the study of dyslexia (Koster et al., 2005; Lyytinen et al., 2001; Viholainen et al., 2006) and autism spectrum disorders (Coonrod & Stone, 2004; Dawson, Osterling, Meltzoff, & Kuhl, 2000; Iverson & Wozniak, 2007), but have yet to be applied to CAS.Study 1 sought to quantify parental report of vocalisation behaviours in children with a clinical diagnosis of CAS. The parents of 20 children with suspected CAS (sCAS) completed a questionnaire focussing on the prelinguistic development of their children as infants. Responses were compared to those from parents of 20 children with Specific Language Impairment (SLI) and 20 children with typically developing (TD) speech and language development. The sCAS children were reported to be significantly less vocal, less likely to have babbled, later in the emergence of first words and later in the emergence of two-word combinations than the TD children. However, the SLI children were reported similarly on many (but not all) items. Despite this similarity, the sCAS group were unique in terms of the presence of reported babbling (35% were reported not to have babbled at all, compared to the TD and SLI children who were all recalled as having babbled in infancy), and the emergence of two word combinations (significantly later than both the TD and SLI groups). In addition, the motor milestones of age of crawling and age of walking were significantly correlated with age of emergence of two-word combinations in the sCAS group, suggesting commonly constrained speech and motor development. Overall, the results provided preliminary support for the notion of atypical prelinguistic vocal development in children with sCAS, and highlighted the importance of further research on the topic.Study 2 applied a retrospective data paradigm in exploring the prelinguistic vocal development of children with CAS. Nine clinically-ascertained children, aged 3 to 4 years and presenting with a range of speech and language profiles (including 3 with suspected CAS), were characterised in terms of operationally-defined CAS characteristics in the first stage (2A) of this study. The battery of tasks included standardised speech and language assessments as well as non-standardised tasks targeting speech production ability. A group of 21 age-matched children with typically developing speech and language skills provided comparison data for the non-standardised tasks. This phase of the study documented CAS characteristics in five of the nine clinical sample participants, with two of these children showing all five of the features investigated. Study 2B examined the early speech, language and motor development of the clinical sample children, via analysis of data available retrospectively for this unique group of children.Their infant profiles were compared to those of 205 infants who had been part of the same community program that the clinical sample had been involved in (and thus had infant data available) but who did not have identified ongoing speech and language issues. Single case comparisons (Crawford & Garthwaite, 2005) revealed that the child with the greatest number and severity of CAS features at preschool age demonstrated significantly poorer expressive skills and a significant dissociation in receptive-expressive abilities in infancy, compared to the typically developing children. Profiles for the other clinical sample children varied considerably.In the third study (Study 3), the development of infants with a family history of CAS (n = 8) was compared to that of infants with no such familial risk (n =8) to further examine the proposed core deficit in CAS. Early speech, language and motor development was tracked at 9, 12, 15, 18 and 24 months. The siblings as a group demonstrated significantly poorer expressive language, speech sound development and fine motor ability than the comparison group, consistent with the notion of a verbal trait deficit (Lewis, Freebairn, Hansen, Taylor et al., 2004). At two years of age, two siblings (and none of the comparison infants) showed clinically-important delays in speech and language development. Inspection of their profiles suggested one infant (SIB2) to present with features consistent with putative early features of CAS (Davis & Velleman, 2000); the other (SIB1) to present with language difficulties not suggestive of CAS.Analysis of their vocalisation samples revealed that while SIB2’s rate of vocalisations at 9 months was not different to that of the comparison group, the nature of the vocalisations were different. While all comparison infants were using canonical syllables at 9 months, SIB2 had not entered this important stage until 12 months, and showed a significantly reduced proportion of canonical syllables at this age (2.5% compared to the comparison infants, who averaged 17%, with none producing less than 6%). Acoustic analyses performed on prelinguistic canonical syllables showed that while duration did not differ, a restricted use of the F1:F2 planar space was noted for SIB2 compared to the typically developing infants, suggesting limited vowel production. Furthermore, a particularly strong correlation between F1 and F2 was observed, suggesting stronger coupling of the articulators. Importantly, the vocalisation data, together with data from standardised assessments, showed a dissociation between speech motor and conceptualiser areas, with a deficit in speech motor control evident in the context of intact conceptual skills for this infant. In contrast, SIB1 (who showed a language delayed profile at 2 years, with no CAS features) did not evidence the types of anomalies identified for SIB2.Taken together, the results of the present research provide support for the viability of a speech motor control deficit account of CAS, when interpreted in a developmental context. As such, they highlight the importance of the prelinguistic period and longitudinal investigations in examining the underlying core deficit in CAS, and suggest important implications for theoretical and clinical conceptualisations of the disorder.