Annotation of Article by Sharon Gretz, M.Ed.

Article: Encoding, Memory, and Transcoding Deficits in Childhood Apraxia of Speech

Clinical Linguistics & Phonetics, May 2012; 26(5):  445 – 482
Shriberg, L.D, Lohmeier, H.L., Strand, E.A., and Jakielski, K.J.

In this study, researchers were examining whether the deficits of children with CAS were limited to speech motor planning & programming (what the author calls “transcoding deficits”) or if the children also had difficulties in auditory perception (termed “encoding” in this paper) and/or memory (ability to store and access representations).  Various analyses and properties of a task called the “Syllable Repetition Task”, as well as other measures, were used to examine 369 children in four distinct groups.  The groupings of subjects were:

• typical speech-language,
• speech delayed – typical language,
• speech delayed – language impairment,
• and a group with childhood apraxia of speech

Overview

Origins of CAS

While there is extensive research about adult acquired apraxia of speech, research in CAS is still young, happening primarily in the last 20 years.  While adult AOS is due primarily to illness or injury, the origins of CAS are more complicated.  Various genes have been associated with CAS including FOXP2, FOXP1, FOXG1, CNTNAP2, ELP4, RAI1 and others.  Because the origins of CAS are by and large different than those for adult AOS, it can be expected that profiles or patterns of deficits would also show differences between adults and children.  First of all, adults who acquire apraxia of speech had a lifetime of a typical path of speech and language formation, development, acquisition and refinement.  We can contrast this to children who are identified with CAS.  Children with CAS are in the acquisition process and the system is already disrupted.  As the authors point out, the integrity and development of one system will influence and have ramifications to other related systems- each in and of themselves in the process of growth and development.  Some systems serve as building blocks for other systems.  Further, the influence of disrupted gene expression in the brain may include not only areas involved with speech processing but other related structures and neural circuits.  Classification systems used for adult motor speech may not relate well to child motor speech disorders.

Speech Processing System

Research often breaks down speech processing into four domains:

• Auditory encoding – auditory information is taken in, interpreted, and transformed into representations of sounds/phonemes, syllables, and words.
• Memory – allows one to store and access the representations
• Transcoding – takes the representations and transforms them into plans and programs for speech movements
• Execution – the neurological processes of actually performing – executing – the programmed speech movements

While speech motor planning is the “noticeable” deficit in CAS, research appears to suggest that encoding and/or memory impairments may also be present, represented by difficulties with reduced perception & production of vowels, syllables, rhymes, and phonemic sequences; problems in reading & spelling.

Transcoding deficits would reflect motor planning and programming difficulties but there is not total agreement in the research community as to what sort of speech behaviors represent planning or programming.  In current CAS research there are no firm markers that are sufficiently specific or sensitive to identify with certainty those children with apraxia of speech.  The authors of this study describe a speech behavior within a Syllable Repetition Task that they believe does tap into motor planning/programming processes.

The Syllable Repetition Task developed by Shriberg et al. includes 18 items and can be scored in 5 minutes. The procedure allows for assessment of the processes underlying nonword repetition in a simple context that can be used for speakers with mild to severe articulation errors.  Nonwords are used to remove linguistic load from the test items and can be a window through which to observe new word learning. The authors used various metrics developed and associated with SRT performance to index deficits in encoding, memory and transcoding.

Research Questions:

1. Do children with apraxia only have speech motor planning/programming deficits (transcoding) or do children with CAS have deficits in multiple domains of the speech processing system?
2. Can findings from a syllable repetition task (SRT) contribute significant diagnostic information on whether or not the individual has CAS?

In this study, researchers include 40 participants with CAS – 20 with genetic conditions and 20 subjects with idiopathic CAS (CAS of unknown origins).  The comparison groups were:

• typical speech/language children,
• children with speech delay and typical language, and
• children with speech and language delay.

As the authors write, “the CAS participants … constitute one of the largest reported samples and the most diverse in age range and diversity of origins.”  In the CAS neurogenetic group, 55% of the sample was males as compared to 65% from the idiopathic CAS group, representing roughly equal gender prevalence in the genetic group.

Notes About Measurements & Scoring Within the Study:

Below is general information about some of the measurements reported in the study:

• Competence scores denoted relative mastery.

• Encoding scores (correct auditory-perceptual representations) were calculated using the mean percentage of within-class manner substitutions that occurred in the 3 and 4 syllable tokens. Higher scores reflect higher competence in encoding.

• Memory scores reflect a ratio between percentages of sounds correct in longer three syllable-length items as compared to shorter two syllable-length items. Lower SRT memory scores point to greater difficulty in accurately repeating 3 syllable length items as compared to 2 syllable length items.

• Transcoding scores were defined as percentage of responses to each of the 18 SRT items that included one or more sound additions, with the score subtracted from 100 (so that lower scores would denote less competence in speech motor planning/programming)

Results

• Competence scores of participants with CAS were lower in both younger subjects (ages 3 – 6) and older subjects (7+ years) as compared to the comparison groups with speech delay, and especially for suprasegmentals and within the older subjects. Competence scores on segmental measures were generally lower in the CAS genetic group than the CAS idiopathic group.
• Participants in the typical speech/language group had the highest competence and processing scores and subjects in the CAS group had the lowest competence and processing scores.
• Subjects with CAS have lower SRT scores than the other three groups on 11 of 12 comparisons.
• Particularly relevant to the study questions, the adjusted mean transcoding score (purported to represent motor planning & programming) for the CAS group was significantly different than adjusted mean scores for the other three comparison groups.
• CAS subjects had significantly lower scores on all measures than the Group 1 (typical speech/language subjects); lower SRT competence and memory scores than Groups 2 & 3 (speech delay/normal language, speech delay/language delay); and significantly lower encoding scores than the Speech delay/normal language comparison group, but not the speech & language delay group 3.

Did the SRT indexes help with diagnostic accuracy of the CAS individual speakers?

Results

• Out of all of the SRT measures, the transcoding scores had the highest diagnostic accuracy for discriminating subjects with CAS (74% accurate in discriminating those with CAS from those with speech delay, with and without language impairment).
• Transcoding scores of 80% ruled out CAS with 93.9% accuracy.
• Transcoding scores of less than 80% were 8.3 times more likely from subjects with CAS than from subjects with speech delay

Transcoding score Differences within the CAS group

• Significantly more CAS subjects with idiopathic origins had low transcoding scores indicating the possible importance of etiology in CAS (neurogenic vs. idiopathic) in future explanatory accounts of the disorder.

Discussion

– Based on the comparisons in this study, CAS is more consistent with an explanation of deficits across multiple speech processing domains, with the most significant deficit in transcoding.

– Transcoding scores were moderately sensitive to CAS, yet some participants from the comparison groups also had low transcoding scores.

– SRT scores can provide significant diagnostic information for CAS, but are not diagnostically conclusive.

– No one diagnostic marker is likely to identify CAS in individuals of all ages, cognitive abilities, competencies, or etiologies.

Conclusions

• Findings from this study indicate that deficits in auditory-perceptual encoding, memory, and transcoding are core features of CAS in idiopathic and genetic contexts, although transcoding (motor planning/programming) was most significant.
• Those with CAS will have quantitative differences in speech, voice and prosody than those with other speech sound disorders rather than conclusive differences and to identify them will require a battery of standard assessment items in these areas.