Predictors of Speech Ability in Down Syndrome

Speech is a critical aspect of the human experience and usually develops in a "seemingly automatic process that continues from birth through adolescence and underlies many related abilities" (e.g., language and reading, see National Academy of Medicine Report on Speech and Language Disorders, 2016). Many individuals with Down Syndrome (Trisomy 21, DS) struggle to communicate and participate more fully in human communication and educational learning experiences because their speech is difficult to understand. The purpose of the proposed project is to measure speech-articulation accuracy and speech intelligibility, and their proposed primary predictors at study entry in 20 children with DS age 6;0 to 13;11). A validated treatment, speech recast intervention (see Yoder, Camarata & Woynaroski, 2016) will be used to drive growth in speech intelligibility as a means of evaluating changes in potential sequelae of change. This integrated study will include measures of speech-articulation accuracy, speech-prosody, general cognitive ability, receptive vocabulary skills, and clinical assay of oral-motor ability as potential predictors of speech intelligibility growth in DS. The Investigators will also be measuring suprasegmental and rhythmic factors associated with growth.

In most children, speech development progresses in a "seemingly automatic manner that continues from birth through adolescence"1 and supports related communication and reading abilities. But many individuals with Down Syndrome (DS, about 1 in 700 births in the US) struggle to acquire speech and participate more fully in human communication and gain access to learning experiences, in part, because their speech is often difficult to understand. This persistent speech disorder arises because people with DS must acquire and produce speech using structurally-different articulators and poor speech oral-motor abilities. Moreover, there is considerable variability in the speech intelligibility of people with DS and it is unclear why some people with DS communicate relatively proficiently whereas others are unable to clearly communicate. The Investigator's recent research provides a unique combination of findings that, when integrated into a prospective project, can potentially provide new insights into how people with DS improve speech intelligibility. This assertion is grounded in the long-standing and replicated finding that speech intelligibility in DS is not strongly correlated with nor solely attributable to measures of phoneme production, so that improvements in intelligibility must rely not only improvements in speech accuracy (phoneme production) but also upon changes in "nonphonemic" factors such as cognitive-linguistic parameters and suprasegmental features such as prosody and speech rate that are currently poorly understood. For example, in a recent paper (Wilson, Abedduto, Camarata & Shriberg, 2019), the Investigators describe cognitive/linguistic and speech-motor factors that relate to speech intelligibility in DS and provide preliminary data on nonphonemic parameters that are significantly correlated to intelligibility. In the cognitive linguistic domain, the Investigators found that measures of cognitive ability and receptive language abilities were directly related to speech intelligibility in people with DS. The Investigators also found that measures of speech-motor ability related to speech intelligibility. Importantly, in another series of studies, the Investigators demonstrated that individualized lexically based phonological recast intervention resulted in improvements in speech intelligibility in school age children with DS and that these gains were NOT uniquely attributable to changes in speech accuracy (phoneme production).6,65 Thus, the Investigators propose to induce growth in speech intelligibility in order to study cognitive-linguistic, suprasegmental and speech motor sequelae of change and as an important step towards a) developing more effective speech interventions, and b) gaining a better understanding of parameters that drive improved speech intelligibility in DS in addition to phoneme production. This latter point is especially important because the Investigator's previous research and the extant literature indicate that traditional measures of speech accuracy (e.g., phoneme articulation) do not adequately capture or predict improvements in speech intelligibility in DS. The proposed research would thus test assumptions underlying current models of speech disorder (and intervention) in DS. It is noteworthy that the Investigator's past work supports the hypothesis that nonphonemic factors such as general cognitive level and vocabulary comprehension level are predictors of later speech intelligibility in school age students with DS. The Investigator's previous research has indicated that the Investigators can drive change in speech intelligibility without directly targeting speech accuracy, providing an opportunity to study intelligibility while controlling for speech accuracy as an untreated factor. Thus, an important first step in identifying factors that influence variation in speech intelligibility in addition to speech accuracy is to identify predictors of these speech abilities following intervention induced gains. Therefore, in keeping with the goals of an exploratory/developmental research project proposal (R21), the Investigators propose a mixed effects regression predictor design that measures speech intelligibility, and it's posited sequelae within the context of a controlled pre-post speech recast intervention study. Participants will be 20 school-age children (6 to 14 years) with DS with varied cognitive ability levels (minimum of 60 with no ceiling). The following specific aims will be pursued: Aim 1: Estimate the relative contribution of (a) general cognitive level and (b) vocabulary comprehension level, controlling for speech accuracy in accounting for variance in pre-post levels of speech intelligibility growth. Aim 2: Estimate the relative contribution of oral-motor skills in accounting for variance in speech intelligibility growth while controlling for speech accuracy. Aim 3: Estimate the relative contribution of suprasegmental prosodic parameters of (a) acoustic prominence (frequency (F0), duration, and intensity) of pitch accented target words; (b) metrical structure of rhythmic speech (rhythmic variability and accent structure) and (c) speech rate in accounting for change overall pre-post growth in speech intelligibility while controlling for speech accuracy. Impact: The results of this study within the context of a pre-post intervention design will provide important preliminary information on factors that contribute to intelligible speech in DS; thereby informing models of speech intelligibility and speech accuracy and future larger scale clinical trials. The proposed project will provide preliminary data to guide future longitudinal studies of value-added predictors of speech outcomes, and ultimately, improve assessment and intervention for speech deficits in individuals with DS.

Speech recasts are likely to improve speech intelligibility in Down Syndrome. The goal of this study is to induce change in speech intelligibility in order to study phonological, acoustic and suprasegmental sequelae of improvements in speech.

a child's spontaneous or elicited production containing phonological errors is immediately followed with a clinician model that corrects the error(s) at the word level rather than the isolated phoneme level. As an example, a speech recast of child's production of the word bake as [be] would be the whole lexeme bake [bek] rather than production drill on [k] in isolation.

Inclusion Criteria: previous medical identification of Down Syndrome chronological age 6-14 years regular use of 3-word utterances (ascertained via parent report during screening and confirmed during assessments) pass a hearing screening at the time of enrollment in the project. All will pass a hearing screening (25 dB @1000, 2000, 4000 Hz). Exclusion Criteria: A history of or parent report of seizures, diagnosed ADHD, apraxia secondary to a diagnosed neurological disorder stuttering, Autism Spectrum Disorder (Autism Diagnostic Observation Scale-2nd Edition score above the ASD cut-off) or severe disruptive behavior that would prevent participation in testing or treatment.

Yoder PJ, Camarata S, Woynaroski T. Treating Speech Comprehensibility in Students With Down Syndrome. J Speech Lang Hear Res. 2016 Jun 1;59(3):446-59. doi: 10.1044/2015_JSLHR-S-15-0148.

Yoder PJ, Woynaroski T, Camarata S. Measuring Speech Comprehensibility in Students with Down Syndrome. J Speech Lang Hear Res. 2016 Jun 1;59(3):460-7. doi: 10.1044/2015_JSLHR-S-15-0149.

Wilson EM, Abbeduto L, Camarata SM, Shriberg LD. Speech and motor speech disorders and intelligibility in adolescents with Down syndrome. Clin Linguist Phon. 2019;33(8):790-814. doi: 10.1080/02699206.2019.1595736.