- Email: [email protected]
Childhood apraxia of speech and multiple phonological disorders in Cairo-Egyptian Arabic speaking children: Language, speech, and oro-motor differences
International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Childhood apraxia of speech and multiple phonological disorders in CairoEgyptian Arabic speaking children: Language, speech, and oro-motor differences Azza Adel Aziz, Sahar Shohdi *, Dalia Mostafa Osman, Emad Iskander Habib ENT Department, Phoniatric Unit, Faculty of Medicine, Cairo University, Egypt
A R T I C L E I N F O
A B S T R A C T
Article history: Received 13 November 2009 Received in revised form 30 January 2010 Accepted 2 February 2010 Available online 3 March 2010
Childhood apraxia of speech is a neurological childhood speech-sound disorder in which the precision and consistency of movements underlying speech are impaired in the absence of neuromuscular deficits. Children with childhood apraxia of speech and those with multiple phonological disorder share some common phonological errors that can be misleading in diagnosis. This study posed a question about a possible significant difference in language, speech and non-speech oral performances between children with childhood apraxia of speech, multiple phonological disorder and normal children that can be used for a differential diagnostic purpose. 30 pre-school children between the ages of 4 and 6 years served as participants. Each of these children represented one of 3 possible subject-groups: Group 1: multiple phonological disorder; Group 2: suspected cases of childhood apraxia of speech; Group 3: control group with no communication disorder. Assessment procedures included: parent interviews; testing of nonspeech oral motor skills and testing of speech skills. Data showed that children with suspected childhood apraxia of speech showed significantly lower language score only in their expressive abilities. Nonspeech tasks did not identify significant differences between childhood apraxia of speech and multiple phonological disorder groups except for those which required two sequential motor performances. In speech tasks, both consonant and vowel accuracy were significantly lower and inconsistent in childhood apraxia of speech group than in the multiple phonological disorder group. Syllable number, shape and sequence accuracy differed significantly in the childhood apraxia of speech group than the other two groups. In addition, children with childhood apraxia of speech showed greater difficulty in processing prosodic features indicating a clear need to address these variables for differential diagnosis and treatment of children with childhood apraxia of speech. ß 2010 Elsevier Ireland Ltd. All rights reserved.
Keywords: Childhood apraxia of speech Multiple phonological disorder Phonetic accuracy Syllabic accuracy Prosodic accuracy
1. Introduction Speech is a remarkable and unique motor accomplishment. It is not unusual for speech to be produced at rates of up to 6–9 syllables per second, which is faster than any other discrete human motor performance, and involves more motor fibers than any other human mechanical activity [1]. Childhood Apraxia of Speech (CAS) is a neurological childhood speech-sound disorder in which the precision and consistency of movements underlying speech are impaired in the absence of neuromuscular deficits. CAS may occur as a result of known neurological impairment, in association with complex neurobehavioral disorders or as an idiopathic neurogenic speech-sound
* Corresponding author. Tel.: +20 233850455. E-mail addresses: [email protected] (A.A. Aziz), [email protected] (S. Shohdi), [email protected] (D.M. Osman), [email protected] (E.I. Habib). 0165-5876/$ – see front matter ß 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2010.02.003
disorder. The core impairment in planning and/or programming spatiotemporal parameters of movement sequences results in errors in speech-sound production and prosody [2]. The complex of behavioral features reportedly associated with Childhood Apraxia of Speech (CAS) places a child at increased risk for early and persistent problems in speech, expressive language, and the phonological foundations of literacy as well as the possible need for augmentative and alternative communication and assistive technology [3]. Most studies of CAS have focused on describing proposed clinical correlates for differential diagnosis from other speech disorder categories. Both understanding of the underlying nature of the disorder and careful description of differential diagnostic behavioral correlates are important for developing effective treatment regimens. Children with CAS and those with Multiple Phonological Disorders (MPD) share some common phonological errors that can be misleading in diagnosis [4,5]. Review of the research literature indicates that, at present, there is no validated list of
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
diagnostic features of CAS that differentiates this symptom complex from other types of childhood speech-sound disorders, especially those primarily due to phonological-level delay [6]. Previous studies reported that along with articulation errors, children with CAS demonstrate problems in non-speech oral motor performances [6,7]. It has also been observed that children with CAS perform poorly, or analogous to children with language disorders on evaluating their language skills [5]. Therefore, it is needed to observe how these children perform compared to other MPD group as well as to normal children on a battery of language, speech and non-speech oral motor tasks. This study posed a question about a possible significant difference in language, speech and non-speech oral performances between children with CAS, MPD and normal children that can be used for a differential diagnostic purpose. Accordingly, suggestions for diagnostic criteria can be proposed, which would in turn aid in establishing a common diagnostic system. 2. Subjects and methods 30 Pre-school children between the ages of 4 and 6 years served as participants. Each of these children represented one of 3 possible subject-groups: Group 1: MPD consisting of 10 children (5 females and 5 males; with a mean age of 4 years, 11 months); Group 2: Suspected cases of CAS consisting of 10 children (4 females & 6 males; with a mean age of 5 years, 2 months); and Group 3: Control group with no communication disorder consisting of 10 children (5 females and 5 males; with a mean age of 5 years, 1 month). The 3 groups were matched in age, gender and socioeconomic distribution. All the participants were assessed in the Phoniatric Unit, Kasr El Aini hospital in the period from 2002 to 2008. Consent was taken from the parents of all children participating in the study. Routine assessment protocol for children with delayed language development was performed for all participants including: Patient interview and history taking (inquiring about child’s feeding difficulty, the age of babbling, the age of 1st word and sentence and any relevant medical problems. Cognitive assessment using Stanford Binet 4th ed. [8] Audiological assessment using immetencemetry, pure tone audiometry and confirmatory evoked response audiometry if needed. Neurological assessment: History taking: prenatal, natal, and postnatal histories, neurological problems among the family. Stigmata of any syndromatic findings Gross neurological examination: motor, sensory, reflexes, coordination, cranial nerves, back, spine, skull and gait examination. Minimal pyramidal signs (leg lag and pronator signs) Language and speech assessment: Arabic language test was used to measure overall language abilities both receptive and expressive, including semantic, syntactic abilities and pragmatic language skills [9]. Arabic Articulation test [10]: this is a picture naming test, where pictures are introduced to elicit 23 Arabic consonants in word initial, middle and final positions. Participants were then selected according to the following criteria:
Inclusion criteria Pre-school age children (4–6 years) Normal cognitive abilities (performance IQ: 80 or above) Mean length of utterance: 2 words (at least) Normal hearing abilities
579
Normal neuromuscular abilities No history of previous language or speech therapy. A team of 4 Phoniatricians made the primary diagnosis of MPD and CAS. Presence of 3 or more phonological error types will classify the child under the MPD group. On the other hand the primary diagnosis of CAS was made on the basis of a cluster of speech and language characteristics typical of the disorder, including high frequency of consonant and syllable omissions, vowel errors, prosodic abnormalities, inconsistent errors and reduced ability to imitate non-verbal speech movements as suggested by [11]. Each child completed a battery of non-speech oral motor tasks and speech tasks. These tasks were completed across three 30-min sessions to avoid possible response degradation due to fatigue from prolonged sessions. All performances were audio taped. Three examiners (from the study team) independently coded responses to the tests and the resulting data were averaged across the 3 examinations. 2.1. Non-speech oral motor tasks Assessment of volitional non-verbal oral movement: This estimates how the child moves his articulators in attempt to imitate non automatic movements and sequences of movements. The child was asked to make 4 specific movements on command with his lips or tongue; blowing a kiss, smiling with showing his teeth, protruding the tongue out, and puckering the cheeks. This was applied on 4 commands of a single movement and another 4 commands of 2 sequential movements (kissing then protruding the tongue, smiling then puckering, tongue out then smiling, puckering then tongue out). The child’s response was rated on a 3point scale, 10 points were awarded for a correct response, 5 points for an emerging response with groping (where the child responds by positioning his articulators in the wrong position and then moving to the correct position), and 0 point for an incorrect response for each movement. The 3 examination trials of each group were averaged to give the final score (out of 40) of the 4 single and the 4 sequential movements for each group. 2.2. Speech tasks All speech assessment tasks were audio taped, phonetically transcribed, and analyzed for: I- Phonetic accuracy, II- Syllabic accuracy and III-Prosodic accuracy. 2.2.1. Phonetic accuracy Two tasks were used for its assessment: A Arabic Articulation Test [10]: to assess speech-sound production first in isolation, then in initial, middle, and final word positions. 23 Arabic consonants were tested; ?, /b/, /t/, /g/, /£/, /x/, /d/, /r/, /z/ , /s/, /R/, Q/, /*/, /?/, /]/, /f/, /k/, /l/, /m/, /n/, /h/, /w/ and /j/. In addition 6 Arabic vowels were also examined /i/, /e/, /a/,/ "/, /o/ and /u/. B The connected spontaneous speech of children was assessed by asking the participants to describe 3 complex pictures that trigger 3-word sentences on the average. Then the following phonetic accuracy data were calculated: (a) Consonant accuracy Consonant accuracy was assessed by computing the percentage of consonants produced correctly compared to the number of consonants attempted during each session to get the percent consonants correct (PCC) [12]. PCC was calculated
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
580
for total consonants produced in the speech sample; first in isolation, second in different positions inside the words, and then in connected spontaneous speech. (b) Vowel accuracy This tests the child’s ability to produce the targeted vowels and calculate the percentage of correctly produced vowels per totally examined vowels (i.e. Percent Vowels Correct PVC). The vowels were tested first in isolation, in the middle of the word, and then in connected spontaneous speech. (c) Consistency of consonant and vowel errors
(b) Syllable shape accuracy
This was observed through repetition of the articulation test by the examining team on 3 different sessions to detect consistency or inconsistency of consonant or vowel errors. (d) Frequency of consonant error types
Analysis of syllable sequencing accuracy was performed using Maximum Repetition Rate (MMR) of 1 word and 2 non-words per 20 s. Participants were asked to repeat the following /ba/, /batata/, / bataka/.
Consonant errors were classified as substitution, omission, or retroflex errors. Substitutions are errors perceived by the transcriber as a different consonant sound than the target sound. Omissions are errors where no consonant was produced for an intended consonant target. Retroflex errors comprised omission or substitution of /r/. Cluster errors are well known in apraxic patients, but the Arabic language does not contain frank consonant clusters, thus words with 2 sequential consonants with short vowels in between (close-to-cluster) were used to assess this problem. Errors involved omission or substitution of one or more sounds in the produced word. Relative frequencies of error types were calculated by dividing the number of errors of a given type by the total number of consonant errors in the speech sample (in isolation, in single words and in connected spontaneous speech). 2.2.2. Syllabic accuracy Syllabic accuracy measures the participant ability to produce accurate syllable number, shape and sequencing. Analysis of syllable number and structure accuracy was done using a specially designed syllable accuracy task ‘‘The Arabic Syllable Accuracy Word Task’’ (ASAWT). Participants were asked to repeat 32 different consonant vowel (CV) structures with increasing task demands. Task levels were arranged into 8 categories guided by the hierarchy of phonotactic difficulty. [2] All participants were asked to repeat the following: (1) Simple open monosyllable (CV) as in /ba/, /ti/, /me/, /do/. (2) Reduplicated open syllables (CVCV—same syllable repeated) as in /mama/, /baba/, /nunu/, /titi/. (3) Harmonized non-reduplicated disyllabic open syllable forms CVCV as in /teta/, /duda/, /keka/, /bobi/ (4) Non-harmonized non-reduplicated disyllabic open syllable forms /toka/, /futa/, /tani/, /hati/. (5) Harmonized closed monosyllable CVC /bab/, /dad/, /tut/, /bib/. (6) Non-harmonized closed monosyllable CVC /kan/, /gab/, /bet/, / hat/ (7) CVCVC polysyllable words (non-reduplicated, non-harmonized) /dulab/, /ketab/, /Rarab/, /mezan/ (8) Words with initial, medial, and/or final close-to-clusters (using short vowels /gebna, /benti/, /wedni/, /kofta/
Syllable shape accuracy is the proportion of words produced in which the syllable shape is maintained regardless of the correctness of consonants. For example, if the target word / ketab/ is produced as /tetab/, the syllable shape is accurate even though the substitution of the posterior velar /k/ by the anterior alveodental /t/ stop is an error. The percent of correct syllable shape to the totally produced syllable was calculated to get the syllable shape accuracy.
2.2.3. Prosodic accuracy Testing the ability of identification and production of rhymes was performed through repetition of a common nursery rhyme (Happy birthday). Children were rated on a 3-point scale; 10 points were awarded for a correct response, 5 points for fair and 0 point for poor response. 3 Trials of each group were averaged to give the final score for each group. Scores from the control group were obtained employing ageadjusted normative standards. 2.3. Statistical analysis Descriptive data was expressed as mean, standard deviation and percentage. Individual comparison of both pathological conditions versus the control group, and versus each other was achieved using paired 2 tailed multiple student t test. Significance was set when p value was <0.01. 3. Results 3.1. Language abilities Developmental data collection revealed that 80% of the Children with suspected CAS were reported by their parents to have babbled very little or with very little phonetic diversity compared to 50% of children with MPD and 30% of children in the normal control group. Utterance of the first word was significantly delayed in both group 1(MPD) and group 2 (CAS) in comparison to the normal group with significantly more delay in the CAS group in comparison to MPD group (Table 1). Children with suspected CAS obtained high scores in receptive language comprehension in the Arabic Language test with poor expressive language scores, while those with MPD reported below average scores in both receptive and expressive subtests compared to the age-adjusted normative standards of the control group. Table 2 shows significant differences between CAS and the 2 other groups only in expressive language scores. 3.2. Non- speech tasks Responses from 8 tasks were used for comparison of nonspeech oral motor performances, 4 with single command and 4
The following scores were then calculated (a) Syllable number accuracy Syllable number accuracy is the proportion of words in which the correct number of syllables is maintained regardless of the correctness of consonants. For example, if 5 words should have been produced with 2 syllables and only 4 were produced with 2 syllables, then syllable number accuracy for disyllables is 4/5 = 80%.
Table 1 Age of uttering of the 1st word in months in the 3 study groups.
Mean Standard deviation * y z
MPD
CAS
Normal
23*,y 3.88
29.2z 4.87
11.9 1.19
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585 Table 2 The results of Arabic language test of the 3 study groups. MPD
Sum of receptive language items Sum of expressive language items * y z
CAS
Normal
Mean
SD
Mean
SD
Mean
SD
67.4y
9
78.8
16
84.8
11.3
11.1
27.6z
14.2
81.4
10.7
63.82*,y
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
with 2 sequential commands. However, performances for the nonspeech tasks of lip and tongue movement did not identify significant differences between the MPD and CAS groups except for tasks which required 2 sequential motor performances, where the experimental group with CAS was significantly lower than those with MPD and the control group as seen in Table 3. 3.3. Speech tasks Results of consonant accuracy revealed that PCC was significantly different in children with CAS than those with MPD and the control group for isolated consonants production. It was also significantly lower inside the words of Arabic articulation test [11] and in connected spontaneous speech production (Table 4). PVC was significantly lower in children with suspected CAS than those with MPD and the control group and it did not differ whether in isolated form production, within the word or in connected spontaneous speech, while there was no significant difference in PVC between MPD and the control group in isolation and in connected speech (Table 5).
581
As regard the consistency of consonant and vowel errors, the results revealed that the standard deviation of correctly produced consonants and vowels in the tested speech sample for children with suspected CAS at the 3 measurement times was much more higher than the MPD group. This wide range of standard deviation of results is suggestive of high session-to-session variability in children with CAS in addition to within-session variability, which was not evident among the other 2 groups as seen in Tables 4 and 5. The relative frequencies of consonant error types are shown in Table 6. Omissions and substitutions were the most frequent error types for children with suspected CAS, with mean frequency of occurrence across sessions of 50.4% and 32.1% respectively. Retroflex and close-to-cluster simplification errors were infrequent, although considerable variation in their occurrence was noted. The frequency of omissions in initial, medial and final position was further analyzed; final consonants were more often omitted than initial or medial consonants in CAS. Initial consonant deletion errors were comparatively infrequent but were more obvious in patients with MPD than with CAS at each measurement point (Table 7). Children with MPD reported higher incidence of substitution than omission with mean frequency of occurrence across sessions of 37.2% and 18.6%, respectively. Retroflex errors were also frequent and constituted 39.0% of errors. Syllable number accuracy differed significantly in the CAS group than the other 2 groups. Target words with 3 syllables or more (i.e. polysyllables) were less accurate than disyllabic and monosyllabic words for children with CAS than for those with MPD and control group. The percent of correctly produced syllable word number were 45.8%, 68.3% and 92.2% for polysyllables, disyllables and monosyllable, respectively for CAS group, while it was 77%, 93.3% and 100% for the MPD group (Table 8).
Table 3 Non-speech oral motor tasks of the 3 study groups. CAS
Normal
Single movement
Mean score/examination/child Standard deviation
MPD 39.50 1.52
38.5z 2.67
40 0
Two sequential movements
Mean score/examination/child Standard deviation
39* 2.03
12.33z 7.15
40 0
N.B. The child’s response was rated on a 3-point scale; 10 points were awarded for a correct response, 5 points for an emerging response with groping and 0 point for an incorrect response for each movement. The 3 trials of each group were averaged to give the final score (out of 40) of the 4 single and the 4 sequential movements of each group. * Means a significant difference between MPD and CAS groups, P value <0.01. z Means a significant difference between CAS and normal groups, P value <0.01.
Table 4 Consonant accuracy: percent consonant correct (PCC) for total consonants in isolation, inside the word and in connected speech samples averaged across 3 time-examinations of the 3 study groups. MPD
CAS
Normal
Correct consonants in isolation
Mean/examination/child Standard deviation Total correct/totally produced Percent consonant correct
16.1*,y 1.70 483/690 70
13.43z 3.51 403/690 58.40
23 0 690/690 100
Correct consonants inside the word
Mean/examination/child Standard deviation Total correct/totally produced Percent consonant correct
48.33*,y 2.69 1450/2070 70
34.7z 6.14 1041/2070 50.20
69 0 2070/2070 100
Correct consonants in connected speech
Mean/examination/child Standard deviation Total correct/totally produced Percent consonant correct
25.06*,y 1.77 752/1140 65.96
7.46z 3.59 224/750 29.90
43.9 2.39 1317/1317 100
* y z
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
582
Table 5 Vowel accuracy: percent vowel correct (PVC) for total vowels in isolation, inside the word and in connected speech samples averaged across 3 time-examinations of the 3 study groups. MPD
CAS
Normal
Correct vowels in isolation
Mean/examination/child Standard deviation Total correct/total trials Percent vowel correct
7.7* 0.65 231/240 96.20
3.76z 1.22 113/240 47.08
8 0 240/240 100
Correct vowels inside the word
Mean/examination/child Standard deviation Total correct/total trials Percent vowel correct
7.6*,y 0.67 228/240 95
2.73z 1.14 82/240 34.16
8 0 240/240 100
Correct vowels in connected speech
Mean/examination/child Standard deviation Total correct/total trials Percent vowel correct
11.96* 2.07 359/372 96.50
1.93z 0.90 58/210 27.61
13.26 1.50 398/398 100
* y z
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
Table 6 Frequency of consonant error types to the total consonant errors. MPD
CAS
Normal
Omission
Mean/examination/child Standard deviation Total no./total errors Percent/total consonant errors
7.56*,y 1.43 227/1215 18.60%
30.96z 7.35 929/1842 50.40%
0 0 No errors 0%
Substitution
Mean/examination/child Standard deviation Total No./total errors Percent/total consonant errors
15.06y 2.62 45/1215 37.20%
18.26z 5.94 593/1842 32.19%
0 0 no errors 0%
Retroflection
Mean/examination/child Standard deviation Total no./total errors Percent/total consonant errors
15.8*,y 2.02 474/1215 39.01%
4.8z 2.91 144/1842 7.81%
0 0 No errors 0%
Cluster reduction
Mean/examination/child Standard deviation Total no./total errors Percent/total consonant errors
2.06*,y 0.98 62/1215 5.10%
7.36z 3.02 221/1842 11.99%
0 0 No errors 0%
* y z
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
The results of syllable shape accuracy differed significantly in the CAS group than the other 2 groups, the condition that signifies a problem of syllable construction in CAS group; the simple open monosyllabic structure and the disyllabic structure with duplication or harmonized forms yielded higher correctly produced syllable structure scores than the closed ended monosyllable, nonharmonized non-duplicated disyllable, polysyllabic and the complex words with consonant close-to-cluster forms, while the other 2 groups did not show significant problem in the patterns of syllable production, except for less CVCVC and close-to-cluster accuracy in MPD group than the control group. Close-to-cluster accuracy was significantly different among the 3 groups. It was 11.60% for children with CAS and 76.6% for MPD and 95% for the control group respectively. As expected, the control group had the lowest number of errors (Table 9). Table 7 Frequency of consonant omission in initial, medial and final positions.
MPD CAS
Initial
Medial
Final
Total omission errors
40% 25%
25% 10%
35% 65%
227 929
Sequence accuracy tests revealed that children with suspected CAS reported difficulty with sequencing syllables than those with MPD and control group. Sequencing errors were noted on the repetition of monosyllable, multisyllabic non-words and words, with the lowest repetition rate across sessions was for the multisyllabic non-words for CAS group (mean 4.8 and SD 1.6), compared to MPD group (mean 11.1 and standard deviation 1.1) and control group (mean 13.3 and standard deviation 1.5) respectively (Table 10). Analysis of prosodic accuracy results revealed that experimental group with suspected CAS was significantly lower than MPD and control groups while repeating the nursery rhyme 3 times, their final scores were (mean 2.5 and standard deviation 2.5) compared to MPD group (mean 8.5 and standard deviation 2.3) and control group (mean 10 and standard deviation 0.0) respectively, where 10 is for the highest accuracy rank of the repetition, 5 is for the middle rank and 0 is for the least rank (Table 11). 4. Discussion A central issue in research on CAS is whether this disorder can be reliably distinguished from more general speech impairments
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
583
Table 8 Syllable number accuracy: percent of correct word length averaged across sessions of the 3 study groups. MPD
CAS
Normal
*
Monosyllable
Mean/examination/child Standard deviation Total correct no./totally produced Percent correct/totally produced
12 0 360/360 100%
11.06z 1.14 332/360 92.20%
12 0 360/360 100%
Disyllables
Mean/examination/child Standard deviation Total correct No./totally produced Percent correct/totally produced
11.2*,y 0.88 336/360 93.30%
8.2z 2.21 246/360 68.30%
12 0 360/360 100%
Polysyllables
Mean/examination/child Standard deviation Total correct No./totally produced Percent correct/totally produced
6.16*,y 1.36 185/240 77%
3.66z 1.56 110/240 45.80%
8 0 240/240 100%
* y z
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
Table 9 Syllable shape accuracy: percent of correct syllable shape averaged across sessions of the 3 study groups (c for consonant and v for vowel). MPD
CAS
Normal
1-CV Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
4* 0 120/120 100%
3.66z 0.47 110/120 91%
4 0 120/120 100%
2-Duplicated CVCV Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
4* 0 120/120 100%
3.16z 0.69 95/120 79.0%
4 0 120/120 100%
3-Harmonized CVCV Mean/examination per child SD Total correct syllable shape/totally produced Percent correct per totally produced
4* 0 120/120 100%
2.6z 1.00 78/120 65%
4 0 120/120 100%
4-Non-duplicated non-harmonized CVCV Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
3.76*,y 0.43 113/120 94.10%
1.73z 1.14 52/120 43.30%
4 0 120/120 100%
5-Harmonized CVC Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
4* 0 120/120 100%
2.73z 1.17 82/120 68.30%
4 0 120/120 100%
6-Non-harmonized CVC Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
4* 0 120/120 100%
2.33z 1.12 70/120 58.30%
4 0 120/120 100%
7-CVCVC Mean/examination per child SD Total correct syllable shape/totally produced Percent correct/totally produced
3.1*,y 0.75 93/120 77.50%
1.53z 0.86 46/120 38.30%
4 0 120/120 100%
8-Clusters Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
3.06*,y 0.82 92/120 76.60%
0.46z 0.57 14/120 11.60%
3.8 0.40 114/120 95%
* y z
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
in children as MPD. This study revealed relevant differences between the study groups that can be used as diagnostic markers. One of the most relevant developmental data detected from history taking was babbling. The results revealed that children
suspected to have CAS were reported by their parents to have babbled very little or with very little phonetic diversity than those with MPD. This agrees with the results of Davis and Velleman [13] who pointed to the use of early vocalizations as a tool to identify
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
584
Table 10 Syllable sequencing accuracy (the maximum repetition rate/20 s). MPD
CAS *,y
Monosyllable word
Mean rate/examination/child Standard deviation
25.16 2.35
Multisyllabic non-words
Mean rate/examination/child Standard deviation
11.13*,y 1.10
Multisyllabic words
Mean rate/examination/child Standard deviation
17.0*,y 1.31
* y z
Normal z
30.96 1.92
4.86z 1.65
13.36 1.58
18.53 3.57
9.13z 2.43
20.6 1.40
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
infants at risk for communication disorders, possibly because it reflects the child’s neurological integrity as well as specific sensory and motor function. Another relevant finding was the significant discrepancy between the receptive and expressive language age of children with CAS than those with MPD, who demonstrated a generalized delay in both receptive and expressive language age in comparison to the standard normative data of the control group, the condition that can be added to the diagnostic markers of CAS. The results of non-speech motor tasks demonstrated the importance of applying it on sequential motor performances rather than on a single performance to significantly differentiate between the 2 groups. A high rate of consonant error is frequently cited as a diagnostic indicator for CAS [11,14]. Observations in this study were consistent with previous findings of consonant errors but also reported similar error patterns found in children of MPD and normally developing children, therefore consonant errors per se are not sufficient for differential diagnosis of CAS and overlap with behavioral features identified in other childhood speech disorders as agreed by [14]. Both children with CAS and MPD presented with a speech disorder characterized by high rates of consonant omission and substitution errors, however, the study revealed that omission errors in children with suspected CAS were accounted for almost exclusively by deletion of the final consonant in words, which is representing a consistent pattern of syllabic error, while children with MPD committed omission errors of different consonant positions in words. The current study revealed session-to-session variability of the CAS group evident by their wide range of standard deviation. The inconsistent trajectory of performance in articulation testing for these children characterizes CAS as reported by [5,15]. Because of this inconsistency in performance, best practice for differential diagnosis of CAS should include assessment on multiple occasions, using both single-word testing and spontaneous sampling. Spontaneous sampling in CAS remains important since it reflects day-to-day speech production more closely than single-word articulation testing. Table 11 Prosodic accuracy.
Mean score/examination/child Standard deviation
MPD
CAS
Normal
8.5*,y 2.33
2.5z 2.54
10 0
N.B. Children were rated on a 3-point scale; 10 points were awarded for a correct response, 5 points for fair and 0 point for poor response. The 3 trials of each group were averaged to give the final score (out of 10) of each group. * Means a significant difference between MPD and CAS groups, P value <0.01. y Means a significant difference between MPD and normal groups, P value <0.01. z Means a significant difference between CAS and normal groups, P value <0.01.
Syllables were examined to explore the relationship between syllable complexity and its accuracy. Syllables were analyzed for syllable number, shape and sequencing accuracy. Children with CAS and MPD generally maintained the targeted number of syllables in their productions for monosyllables. Target words with 2 syllables or more were significantly less accurate in children with suspected CAS than the MPD group and the control group. Syllable accuracy analyses showed that children with suspected CAS had significantly lower syllabic shape accuracy scores than those with MPD. Simple open monosyllables, and open disyllables with duplication or harmony were the only syllable shapes relatively correctly produced in children with suspected CAS, suggesting that persistent omission errors likely represent a deficit in syllabic construction rather than sound-specific errors. This matched with previous results which suggested that consonant errors affecting syllable structure are more common in CAS than in common speech disorders not identified as CAS [5,14,15]. Accuracy of polysyllables and consonant close-to-clusters varied significantly between CAS group and the other 2 groups. This highlights the difficulty confronted by CAS patients in relation to certain syllabic shape that could be used as a diagnostic marker. Errors typically involved omission of one of the consonants, indicative of a more general phenomenon of syllable shape reduction rather than a consonant difficulty. This condition reflects an impaired ability to construct accurate word shape and indicates that syllable-level error pattern represents a persisting area of deficit relative to chronological age expectations for children diagnosed with CAS than those with MPD. Due to the paucity of multisyllabic productions in children with apraxia, sequencing errors are often difficult to identify in younger children. Therefore, phonetic transcripts of the multisyllabic words and non-words were examined for sequencing errors, which was applied to document problems in programming of sequential articulatory movements. The results identified a significant deficit in patients with CAS than those with MPD especially for sequencing of multisyllabic nonsense words. This is probably because non-word repetition task does not only represent the combined influence of several speech and language processes (including auditory memory and phonologic processing) but it also entails preparation and execution of a speech movement sequence, which is to say, speech motor control. The results of prosodic assessment showed significant delay in children with CAS than those with MPD across the 3 trials. Thus results of prosodic assessment can be added to the diagnostic markers of CAS. This is in agreement with the results of Marion et al., Shriberg et al., Skinder et al. and Peter and Stoel-Gammon [16–20]. This study simultaneously examined language; speech and non-speech performances of children identified with suspected CAS and MPD and compared them to normal children. This consideration has enabled identification of common character-
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
istics, as well as atypical features of the speech-disordered groups. It has also ascertained the interaction between speech and language performances among these children. The descriptive data showed that children with suspected CAS showed relatively greater difficulty for prosodic features and consonant cluster productions in addition to difficulty in processing syllabic information, indicating a clear need to address these variables for differential diagnosis and treatment of children with CAS. References [1] R. Kent, Research on speech motor control and its disorders: a review and prospective, Journal of Communication Disorders 33 (2000) 391–428. [2] S.L. Velleman, Phonotactic therapy, Seminars in Speech and Language 23 (1) (2002) 43–56. [3] American Speech Language and Hearing Association, Childhood Apraxia of Speech, Retrieved on July 1, 2007, from http://www.asha.org/docs/html/ PS2007-00277.html. [4] L. Nijland, B. Maassen, S. Meulen, Evidence of motor programming deficits in children diagnosed with DAS, Journal of Speech, Language, and Hearing Research 46 (2) (2003) 437. [5] B. Lewis, L. Freebairn, A. Hansen, S. Iyengar, H. Taylor, School-age follow-up of children with Childhood Apraxia of Speech, Language Speech and Hearing Services in Schools 35 (2004) 122–140. [6] S. Drummond, J. Ashe, Developmental verbal apraxia and phonological disorders: speech & non-speech performances, Presented at the American Speech Language Hearing Association National Convention, November 2005. [7] K. Shipley, McAfee J (Eds.), Assessment in Speech-language Pathology: A Resource Manual, second ed., Singular Publishing Group, San Diego, 1998. [8] R.L. Thorndike, E.P. Hagen, J.M. Sattler, Stanford–Binet Intelligence Scale, fourth ed., Riverside Publishing, Itasca, IL, 1986.
585
[9] M.N. Kotby, A. Khairy, M. Baraka, N. Rifaie, El Shobary, Language testing of Arabic Speaking Children, Proceeding of the XXIII World Congress Of The International Association of Logopedics and Phoniatrics 1995. [10] M.N. Kotby, S. Bassiouny, El Zomor, E. Mohsen, Standardization of an articulation test, in: Proceedings of the 9th Annual Ain Shams Medical Congress, Egypt, 1986. [11] B.L. Davis, K.J. Jakielski, T.P. Marquardt, Developmental Apraxia of Speech: determiners of differential diagnosis, Clinical Linguistics and Phonetics 12 (1) (1998) 25–45. [12] L.D. Shriberg, J. Kwiatkowski, Phonological disorders: III. A procedure for assessing severity of involvement, Journal of Speech and Hearing Disorders 47 (3) (1982) 256–270. [13] B.L. Davis, S.L. Velleman, Differential diagnosis and treatment of Developmental Apraxia of Speech in infants and toddlers, Infant Toddler Intervention 10 (2000) 177–192. [14] L.D. Shriberg, D.M. Aram, J. Kwiatkowski, Developmental Apraxia of Speech: II. Toward a diagnostic marker, Journal of Speech Language and Hearing Research 40 (2) (1997) 286–312. [15] T. Marquardt, H.M. Sussman, T. Snow, A. Jacks, The integrity of the syllable in Developmental Apraxia of Speech, Journal of Communication Disorders 35 (2002) 31–49. [16] M. Marion, H. Sussman, T. Marquardt, The perception and production of rhyme in normal and developmentally apraxic children, Journal of Communication Disorders 26 (1993) 129–160. [17] L.D. Shriberg, D.M. Aram, J. Kwiatkowski, Developmental Apraxia of Speech: I. Descriptive and theoretical perspectives, Journal of Speech, Language, and Hearing Research 40 (1997) 273–285. [18] L.D. Shriberg, D.M. Aram, J. Kwiatkowski, Developmental Apraxia of Speech: III. A subtype marked by inappropriate stress, Journal of Speech, Language & Hearing Research 40 (1997) 313–337. [19] A. Skinder, E.A. Strand, M. Mignerey, Perceptual and acoustic analysis of lexical and sentential stress in children with Developmental Apraxia of Speech, Journal of Medical Speech-Language Pathology 7 (1999) 133–144. [20] B. Peter, C. Stoel-Gammon, Timing errors in two children with suspected Childhood Apraxia of Speech (sCAS) during speech and music-related tasks, Clinical Linguistics & Phonetics 19 (2) (2005) 67–87.
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Childhood apraxia of speech and multiple phonological disorders in CairoEgyptian Arabic speaking children: Language, speech, and oro-motor differences Azza Adel Aziz, Sahar Shohdi *, Dalia Mostafa Osman, Emad Iskander Habib ENT Department, Phoniatric Unit, Faculty of Medicine, Cairo University, Egypt
A R T I C L E I N F O
A B S T R A C T
Article history: Received 13 November 2009 Received in revised form 30 January 2010 Accepted 2 February 2010 Available online 3 March 2010
Childhood apraxia of speech is a neurological childhood speech-sound disorder in which the precision and consistency of movements underlying speech are impaired in the absence of neuromuscular deficits. Children with childhood apraxia of speech and those with multiple phonological disorder share some common phonological errors that can be misleading in diagnosis. This study posed a question about a possible significant difference in language, speech and non-speech oral performances between children with childhood apraxia of speech, multiple phonological disorder and normal children that can be used for a differential diagnostic purpose. 30 pre-school children between the ages of 4 and 6 years served as participants. Each of these children represented one of 3 possible subject-groups: Group 1: multiple phonological disorder; Group 2: suspected cases of childhood apraxia of speech; Group 3: control group with no communication disorder. Assessment procedures included: parent interviews; testing of nonspeech oral motor skills and testing of speech skills. Data showed that children with suspected childhood apraxia of speech showed significantly lower language score only in their expressive abilities. Nonspeech tasks did not identify significant differences between childhood apraxia of speech and multiple phonological disorder groups except for those which required two sequential motor performances. In speech tasks, both consonant and vowel accuracy were significantly lower and inconsistent in childhood apraxia of speech group than in the multiple phonological disorder group. Syllable number, shape and sequence accuracy differed significantly in the childhood apraxia of speech group than the other two groups. In addition, children with childhood apraxia of speech showed greater difficulty in processing prosodic features indicating a clear need to address these variables for differential diagnosis and treatment of children with childhood apraxia of speech. ß 2010 Elsevier Ireland Ltd. All rights reserved.
Keywords: Childhood apraxia of speech Multiple phonological disorder Phonetic accuracy Syllabic accuracy Prosodic accuracy
1. Introduction Speech is a remarkable and unique motor accomplishment. It is not unusual for speech to be produced at rates of up to 6–9 syllables per second, which is faster than any other discrete human motor performance, and involves more motor fibers than any other human mechanical activity [1]. Childhood Apraxia of Speech (CAS) is a neurological childhood speech-sound disorder in which the precision and consistency of movements underlying speech are impaired in the absence of neuromuscular deficits. CAS may occur as a result of known neurological impairment, in association with complex neurobehavioral disorders or as an idiopathic neurogenic speech-sound
* Corresponding author. Tel.: +20 233850455. E-mail addresses: [email protected] (A.A. Aziz), [email protected] (S. Shohdi), [email protected] (D.M. Osman), [email protected] (E.I. Habib). 0165-5876/$ – see front matter ß 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2010.02.003
disorder. The core impairment in planning and/or programming spatiotemporal parameters of movement sequences results in errors in speech-sound production and prosody [2]. The complex of behavioral features reportedly associated with Childhood Apraxia of Speech (CAS) places a child at increased risk for early and persistent problems in speech, expressive language, and the phonological foundations of literacy as well as the possible need for augmentative and alternative communication and assistive technology [3]. Most studies of CAS have focused on describing proposed clinical correlates for differential diagnosis from other speech disorder categories. Both understanding of the underlying nature of the disorder and careful description of differential diagnostic behavioral correlates are important for developing effective treatment regimens. Children with CAS and those with Multiple Phonological Disorders (MPD) share some common phonological errors that can be misleading in diagnosis [4,5]. Review of the research literature indicates that, at present, there is no validated list of
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
diagnostic features of CAS that differentiates this symptom complex from other types of childhood speech-sound disorders, especially those primarily due to phonological-level delay [6]. Previous studies reported that along with articulation errors, children with CAS demonstrate problems in non-speech oral motor performances [6,7]. It has also been observed that children with CAS perform poorly, or analogous to children with language disorders on evaluating their language skills [5]. Therefore, it is needed to observe how these children perform compared to other MPD group as well as to normal children on a battery of language, speech and non-speech oral motor tasks. This study posed a question about a possible significant difference in language, speech and non-speech oral performances between children with CAS, MPD and normal children that can be used for a differential diagnostic purpose. Accordingly, suggestions for diagnostic criteria can be proposed, which would in turn aid in establishing a common diagnostic system. 2. Subjects and methods 30 Pre-school children between the ages of 4 and 6 years served as participants. Each of these children represented one of 3 possible subject-groups: Group 1: MPD consisting of 10 children (5 females and 5 males; with a mean age of 4 years, 11 months); Group 2: Suspected cases of CAS consisting of 10 children (4 females & 6 males; with a mean age of 5 years, 2 months); and Group 3: Control group with no communication disorder consisting of 10 children (5 females and 5 males; with a mean age of 5 years, 1 month). The 3 groups were matched in age, gender and socioeconomic distribution. All the participants were assessed in the Phoniatric Unit, Kasr El Aini hospital in the period from 2002 to 2008. Consent was taken from the parents of all children participating in the study. Routine assessment protocol for children with delayed language development was performed for all participants including: Patient interview and history taking (inquiring about child’s feeding difficulty, the age of babbling, the age of 1st word and sentence and any relevant medical problems. Cognitive assessment using Stanford Binet 4th ed. [8] Audiological assessment using immetencemetry, pure tone audiometry and confirmatory evoked response audiometry if needed. Neurological assessment: History taking: prenatal, natal, and postnatal histories, neurological problems among the family. Stigmata of any syndromatic findings Gross neurological examination: motor, sensory, reflexes, coordination, cranial nerves, back, spine, skull and gait examination. Minimal pyramidal signs (leg lag and pronator signs) Language and speech assessment: Arabic language test was used to measure overall language abilities both receptive and expressive, including semantic, syntactic abilities and pragmatic language skills [9]. Arabic Articulation test [10]: this is a picture naming test, where pictures are introduced to elicit 23 Arabic consonants in word initial, middle and final positions. Participants were then selected according to the following criteria:
Inclusion criteria Pre-school age children (4–6 years) Normal cognitive abilities (performance IQ: 80 or above) Mean length of utterance: 2 words (at least) Normal hearing abilities
579
Normal neuromuscular abilities No history of previous language or speech therapy. A team of 4 Phoniatricians made the primary diagnosis of MPD and CAS. Presence of 3 or more phonological error types will classify the child under the MPD group. On the other hand the primary diagnosis of CAS was made on the basis of a cluster of speech and language characteristics typical of the disorder, including high frequency of consonant and syllable omissions, vowel errors, prosodic abnormalities, inconsistent errors and reduced ability to imitate non-verbal speech movements as suggested by [11]. Each child completed a battery of non-speech oral motor tasks and speech tasks. These tasks were completed across three 30-min sessions to avoid possible response degradation due to fatigue from prolonged sessions. All performances were audio taped. Three examiners (from the study team) independently coded responses to the tests and the resulting data were averaged across the 3 examinations. 2.1. Non-speech oral motor tasks Assessment of volitional non-verbal oral movement: This estimates how the child moves his articulators in attempt to imitate non automatic movements and sequences of movements. The child was asked to make 4 specific movements on command with his lips or tongue; blowing a kiss, smiling with showing his teeth, protruding the tongue out, and puckering the cheeks. This was applied on 4 commands of a single movement and another 4 commands of 2 sequential movements (kissing then protruding the tongue, smiling then puckering, tongue out then smiling, puckering then tongue out). The child’s response was rated on a 3point scale, 10 points were awarded for a correct response, 5 points for an emerging response with groping (where the child responds by positioning his articulators in the wrong position and then moving to the correct position), and 0 point for an incorrect response for each movement. The 3 examination trials of each group were averaged to give the final score (out of 40) of the 4 single and the 4 sequential movements for each group. 2.2. Speech tasks All speech assessment tasks were audio taped, phonetically transcribed, and analyzed for: I- Phonetic accuracy, II- Syllabic accuracy and III-Prosodic accuracy. 2.2.1. Phonetic accuracy Two tasks were used for its assessment: A Arabic Articulation Test [10]: to assess speech-sound production first in isolation, then in initial, middle, and final word positions. 23 Arabic consonants were tested; ?, /b/, /t/, /g/, /£/, /x/, /d/, /r/, /z/ , /s/, /R/, Q/, /*/, /?/, /]/, /f/, /k/, /l/, /m/, /n/, /h/, /w/ and /j/. In addition 6 Arabic vowels were also examined /i/, /e/, /a/,/ "/, /o/ and /u/. B The connected spontaneous speech of children was assessed by asking the participants to describe 3 complex pictures that trigger 3-word sentences on the average. Then the following phonetic accuracy data were calculated: (a) Consonant accuracy Consonant accuracy was assessed by computing the percentage of consonants produced correctly compared to the number of consonants attempted during each session to get the percent consonants correct (PCC) [12]. PCC was calculated
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
580
for total consonants produced in the speech sample; first in isolation, second in different positions inside the words, and then in connected spontaneous speech. (b) Vowel accuracy This tests the child’s ability to produce the targeted vowels and calculate the percentage of correctly produced vowels per totally examined vowels (i.e. Percent Vowels Correct PVC). The vowels were tested first in isolation, in the middle of the word, and then in connected spontaneous speech. (c) Consistency of consonant and vowel errors
(b) Syllable shape accuracy
This was observed through repetition of the articulation test by the examining team on 3 different sessions to detect consistency or inconsistency of consonant or vowel errors. (d) Frequency of consonant error types
Analysis of syllable sequencing accuracy was performed using Maximum Repetition Rate (MMR) of 1 word and 2 non-words per 20 s. Participants were asked to repeat the following /ba/, /batata/, / bataka/.
Consonant errors were classified as substitution, omission, or retroflex errors. Substitutions are errors perceived by the transcriber as a different consonant sound than the target sound. Omissions are errors where no consonant was produced for an intended consonant target. Retroflex errors comprised omission or substitution of /r/. Cluster errors are well known in apraxic patients, but the Arabic language does not contain frank consonant clusters, thus words with 2 sequential consonants with short vowels in between (close-to-cluster) were used to assess this problem. Errors involved omission or substitution of one or more sounds in the produced word. Relative frequencies of error types were calculated by dividing the number of errors of a given type by the total number of consonant errors in the speech sample (in isolation, in single words and in connected spontaneous speech). 2.2.2. Syllabic accuracy Syllabic accuracy measures the participant ability to produce accurate syllable number, shape and sequencing. Analysis of syllable number and structure accuracy was done using a specially designed syllable accuracy task ‘‘The Arabic Syllable Accuracy Word Task’’ (ASAWT). Participants were asked to repeat 32 different consonant vowel (CV) structures with increasing task demands. Task levels were arranged into 8 categories guided by the hierarchy of phonotactic difficulty. [2] All participants were asked to repeat the following: (1) Simple open monosyllable (CV) as in /ba/, /ti/, /me/, /do/. (2) Reduplicated open syllables (CVCV—same syllable repeated) as in /mama/, /baba/, /nunu/, /titi/. (3) Harmonized non-reduplicated disyllabic open syllable forms CVCV as in /teta/, /duda/, /keka/, /bobi/ (4) Non-harmonized non-reduplicated disyllabic open syllable forms /toka/, /futa/, /tani/, /hati/. (5) Harmonized closed monosyllable CVC /bab/, /dad/, /tut/, /bib/. (6) Non-harmonized closed monosyllable CVC /kan/, /gab/, /bet/, / hat/ (7) CVCVC polysyllable words (non-reduplicated, non-harmonized) /dulab/, /ketab/, /Rarab/, /mezan/ (8) Words with initial, medial, and/or final close-to-clusters (using short vowels /gebna, /benti/, /wedni/, /kofta/
Syllable shape accuracy is the proportion of words produced in which the syllable shape is maintained regardless of the correctness of consonants. For example, if the target word / ketab/ is produced as /tetab/, the syllable shape is accurate even though the substitution of the posterior velar /k/ by the anterior alveodental /t/ stop is an error. The percent of correct syllable shape to the totally produced syllable was calculated to get the syllable shape accuracy.
2.2.3. Prosodic accuracy Testing the ability of identification and production of rhymes was performed through repetition of a common nursery rhyme (Happy birthday). Children were rated on a 3-point scale; 10 points were awarded for a correct response, 5 points for fair and 0 point for poor response. 3 Trials of each group were averaged to give the final score for each group. Scores from the control group were obtained employing ageadjusted normative standards. 2.3. Statistical analysis Descriptive data was expressed as mean, standard deviation and percentage. Individual comparison of both pathological conditions versus the control group, and versus each other was achieved using paired 2 tailed multiple student t test. Significance was set when p value was <0.01. 3. Results 3.1. Language abilities Developmental data collection revealed that 80% of the Children with suspected CAS were reported by their parents to have babbled very little or with very little phonetic diversity compared to 50% of children with MPD and 30% of children in the normal control group. Utterance of the first word was significantly delayed in both group 1(MPD) and group 2 (CAS) in comparison to the normal group with significantly more delay in the CAS group in comparison to MPD group (Table 1). Children with suspected CAS obtained high scores in receptive language comprehension in the Arabic Language test with poor expressive language scores, while those with MPD reported below average scores in both receptive and expressive subtests compared to the age-adjusted normative standards of the control group. Table 2 shows significant differences between CAS and the 2 other groups only in expressive language scores. 3.2. Non- speech tasks Responses from 8 tasks were used for comparison of nonspeech oral motor performances, 4 with single command and 4
The following scores were then calculated (a) Syllable number accuracy Syllable number accuracy is the proportion of words in which the correct number of syllables is maintained regardless of the correctness of consonants. For example, if 5 words should have been produced with 2 syllables and only 4 were produced with 2 syllables, then syllable number accuracy for disyllables is 4/5 = 80%.
Table 1 Age of uttering of the 1st word in months in the 3 study groups.
Mean Standard deviation * y z
MPD
CAS
Normal
23*,y 3.88
29.2z 4.87
11.9 1.19
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585 Table 2 The results of Arabic language test of the 3 study groups. MPD
Sum of receptive language items Sum of expressive language items * y z
CAS
Normal
Mean
SD
Mean
SD
Mean
SD
67.4y
9
78.8
16
84.8
11.3
11.1
27.6z
14.2
81.4
10.7
63.82*,y
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
with 2 sequential commands. However, performances for the nonspeech tasks of lip and tongue movement did not identify significant differences between the MPD and CAS groups except for tasks which required 2 sequential motor performances, where the experimental group with CAS was significantly lower than those with MPD and the control group as seen in Table 3. 3.3. Speech tasks Results of consonant accuracy revealed that PCC was significantly different in children with CAS than those with MPD and the control group for isolated consonants production. It was also significantly lower inside the words of Arabic articulation test [11] and in connected spontaneous speech production (Table 4). PVC was significantly lower in children with suspected CAS than those with MPD and the control group and it did not differ whether in isolated form production, within the word or in connected spontaneous speech, while there was no significant difference in PVC between MPD and the control group in isolation and in connected speech (Table 5).
581
As regard the consistency of consonant and vowel errors, the results revealed that the standard deviation of correctly produced consonants and vowels in the tested speech sample for children with suspected CAS at the 3 measurement times was much more higher than the MPD group. This wide range of standard deviation of results is suggestive of high session-to-session variability in children with CAS in addition to within-session variability, which was not evident among the other 2 groups as seen in Tables 4 and 5. The relative frequencies of consonant error types are shown in Table 6. Omissions and substitutions were the most frequent error types for children with suspected CAS, with mean frequency of occurrence across sessions of 50.4% and 32.1% respectively. Retroflex and close-to-cluster simplification errors were infrequent, although considerable variation in their occurrence was noted. The frequency of omissions in initial, medial and final position was further analyzed; final consonants were more often omitted than initial or medial consonants in CAS. Initial consonant deletion errors were comparatively infrequent but were more obvious in patients with MPD than with CAS at each measurement point (Table 7). Children with MPD reported higher incidence of substitution than omission with mean frequency of occurrence across sessions of 37.2% and 18.6%, respectively. Retroflex errors were also frequent and constituted 39.0% of errors. Syllable number accuracy differed significantly in the CAS group than the other 2 groups. Target words with 3 syllables or more (i.e. polysyllables) were less accurate than disyllabic and monosyllabic words for children with CAS than for those with MPD and control group. The percent of correctly produced syllable word number were 45.8%, 68.3% and 92.2% for polysyllables, disyllables and monosyllable, respectively for CAS group, while it was 77%, 93.3% and 100% for the MPD group (Table 8).
Table 3 Non-speech oral motor tasks of the 3 study groups. CAS
Normal
Single movement
Mean score/examination/child Standard deviation
MPD 39.50 1.52
38.5z 2.67
40 0
Two sequential movements
Mean score/examination/child Standard deviation
39* 2.03
12.33z 7.15
40 0
N.B. The child’s response was rated on a 3-point scale; 10 points were awarded for a correct response, 5 points for an emerging response with groping and 0 point for an incorrect response for each movement. The 3 trials of each group were averaged to give the final score (out of 40) of the 4 single and the 4 sequential movements of each group. * Means a significant difference between MPD and CAS groups, P value <0.01. z Means a significant difference between CAS and normal groups, P value <0.01.
Table 4 Consonant accuracy: percent consonant correct (PCC) for total consonants in isolation, inside the word and in connected speech samples averaged across 3 time-examinations of the 3 study groups. MPD
CAS
Normal
Correct consonants in isolation
Mean/examination/child Standard deviation Total correct/totally produced Percent consonant correct
16.1*,y 1.70 483/690 70
13.43z 3.51 403/690 58.40
23 0 690/690 100
Correct consonants inside the word
Mean/examination/child Standard deviation Total correct/totally produced Percent consonant correct
48.33*,y 2.69 1450/2070 70
34.7z 6.14 1041/2070 50.20
69 0 2070/2070 100
Correct consonants in connected speech
Mean/examination/child Standard deviation Total correct/totally produced Percent consonant correct
25.06*,y 1.77 752/1140 65.96
7.46z 3.59 224/750 29.90
43.9 2.39 1317/1317 100
* y z
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
582
Table 5 Vowel accuracy: percent vowel correct (PVC) for total vowels in isolation, inside the word and in connected speech samples averaged across 3 time-examinations of the 3 study groups. MPD
CAS
Normal
Correct vowels in isolation
Mean/examination/child Standard deviation Total correct/total trials Percent vowel correct
7.7* 0.65 231/240 96.20
3.76z 1.22 113/240 47.08
8 0 240/240 100
Correct vowels inside the word
Mean/examination/child Standard deviation Total correct/total trials Percent vowel correct
7.6*,y 0.67 228/240 95
2.73z 1.14 82/240 34.16
8 0 240/240 100
Correct vowels in connected speech
Mean/examination/child Standard deviation Total correct/total trials Percent vowel correct
11.96* 2.07 359/372 96.50
1.93z 0.90 58/210 27.61
13.26 1.50 398/398 100
* y z
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
Table 6 Frequency of consonant error types to the total consonant errors. MPD
CAS
Normal
Omission
Mean/examination/child Standard deviation Total no./total errors Percent/total consonant errors
7.56*,y 1.43 227/1215 18.60%
30.96z 7.35 929/1842 50.40%
0 0 No errors 0%
Substitution
Mean/examination/child Standard deviation Total No./total errors Percent/total consonant errors
15.06y 2.62 45/1215 37.20%
18.26z 5.94 593/1842 32.19%
0 0 no errors 0%
Retroflection
Mean/examination/child Standard deviation Total no./total errors Percent/total consonant errors
15.8*,y 2.02 474/1215 39.01%
4.8z 2.91 144/1842 7.81%
0 0 No errors 0%
Cluster reduction
Mean/examination/child Standard deviation Total no./total errors Percent/total consonant errors
2.06*,y 0.98 62/1215 5.10%
7.36z 3.02 221/1842 11.99%
0 0 No errors 0%
* y z
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
The results of syllable shape accuracy differed significantly in the CAS group than the other 2 groups, the condition that signifies a problem of syllable construction in CAS group; the simple open monosyllabic structure and the disyllabic structure with duplication or harmonized forms yielded higher correctly produced syllable structure scores than the closed ended monosyllable, nonharmonized non-duplicated disyllable, polysyllabic and the complex words with consonant close-to-cluster forms, while the other 2 groups did not show significant problem in the patterns of syllable production, except for less CVCVC and close-to-cluster accuracy in MPD group than the control group. Close-to-cluster accuracy was significantly different among the 3 groups. It was 11.60% for children with CAS and 76.6% for MPD and 95% for the control group respectively. As expected, the control group had the lowest number of errors (Table 9). Table 7 Frequency of consonant omission in initial, medial and final positions.
MPD CAS
Initial
Medial
Final
Total omission errors
40% 25%
25% 10%
35% 65%
227 929
Sequence accuracy tests revealed that children with suspected CAS reported difficulty with sequencing syllables than those with MPD and control group. Sequencing errors were noted on the repetition of monosyllable, multisyllabic non-words and words, with the lowest repetition rate across sessions was for the multisyllabic non-words for CAS group (mean 4.8 and SD 1.6), compared to MPD group (mean 11.1 and standard deviation 1.1) and control group (mean 13.3 and standard deviation 1.5) respectively (Table 10). Analysis of prosodic accuracy results revealed that experimental group with suspected CAS was significantly lower than MPD and control groups while repeating the nursery rhyme 3 times, their final scores were (mean 2.5 and standard deviation 2.5) compared to MPD group (mean 8.5 and standard deviation 2.3) and control group (mean 10 and standard deviation 0.0) respectively, where 10 is for the highest accuracy rank of the repetition, 5 is for the middle rank and 0 is for the least rank (Table 11). 4. Discussion A central issue in research on CAS is whether this disorder can be reliably distinguished from more general speech impairments
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
583
Table 8 Syllable number accuracy: percent of correct word length averaged across sessions of the 3 study groups. MPD
CAS
Normal
*
Monosyllable
Mean/examination/child Standard deviation Total correct no./totally produced Percent correct/totally produced
12 0 360/360 100%
11.06z 1.14 332/360 92.20%
12 0 360/360 100%
Disyllables
Mean/examination/child Standard deviation Total correct No./totally produced Percent correct/totally produced
11.2*,y 0.88 336/360 93.30%
8.2z 2.21 246/360 68.30%
12 0 360/360 100%
Polysyllables
Mean/examination/child Standard deviation Total correct No./totally produced Percent correct/totally produced
6.16*,y 1.36 185/240 77%
3.66z 1.56 110/240 45.80%
8 0 240/240 100%
* y z
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
Table 9 Syllable shape accuracy: percent of correct syllable shape averaged across sessions of the 3 study groups (c for consonant and v for vowel). MPD
CAS
Normal
1-CV Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
4* 0 120/120 100%
3.66z 0.47 110/120 91%
4 0 120/120 100%
2-Duplicated CVCV Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
4* 0 120/120 100%
3.16z 0.69 95/120 79.0%
4 0 120/120 100%
3-Harmonized CVCV Mean/examination per child SD Total correct syllable shape/totally produced Percent correct per totally produced
4* 0 120/120 100%
2.6z 1.00 78/120 65%
4 0 120/120 100%
4-Non-duplicated non-harmonized CVCV Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
3.76*,y 0.43 113/120 94.10%
1.73z 1.14 52/120 43.30%
4 0 120/120 100%
5-Harmonized CVC Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
4* 0 120/120 100%
2.73z 1.17 82/120 68.30%
4 0 120/120 100%
6-Non-harmonized CVC Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
4* 0 120/120 100%
2.33z 1.12 70/120 58.30%
4 0 120/120 100%
7-CVCVC Mean/examination per child SD Total correct syllable shape/totally produced Percent correct/totally produced
3.1*,y 0.75 93/120 77.50%
1.53z 0.86 46/120 38.30%
4 0 120/120 100%
8-Clusters Mean/examination/child SD Total correct syllable shape/totally produced Percent correct/totally produced
3.06*,y 0.82 92/120 76.60%
0.46z 0.57 14/120 11.60%
3.8 0.40 114/120 95%
* y z
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
in children as MPD. This study revealed relevant differences between the study groups that can be used as diagnostic markers. One of the most relevant developmental data detected from history taking was babbling. The results revealed that children
suspected to have CAS were reported by their parents to have babbled very little or with very little phonetic diversity than those with MPD. This agrees with the results of Davis and Velleman [13] who pointed to the use of early vocalizations as a tool to identify
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
584
Table 10 Syllable sequencing accuracy (the maximum repetition rate/20 s). MPD
CAS *,y
Monosyllable word
Mean rate/examination/child Standard deviation
25.16 2.35
Multisyllabic non-words
Mean rate/examination/child Standard deviation
11.13*,y 1.10
Multisyllabic words
Mean rate/examination/child Standard deviation
17.0*,y 1.31
* y z
Normal z
30.96 1.92
4.86z 1.65
13.36 1.58
18.53 3.57
9.13z 2.43
20.6 1.40
Means a significant difference between MPD and CAS groups, P value <0.01. Means a significant difference between MPD and normal groups, P value <0.01. Means a significant difference between CAS and normal groups, P value <0.01.
infants at risk for communication disorders, possibly because it reflects the child’s neurological integrity as well as specific sensory and motor function. Another relevant finding was the significant discrepancy between the receptive and expressive language age of children with CAS than those with MPD, who demonstrated a generalized delay in both receptive and expressive language age in comparison to the standard normative data of the control group, the condition that can be added to the diagnostic markers of CAS. The results of non-speech motor tasks demonstrated the importance of applying it on sequential motor performances rather than on a single performance to significantly differentiate between the 2 groups. A high rate of consonant error is frequently cited as a diagnostic indicator for CAS [11,14]. Observations in this study were consistent with previous findings of consonant errors but also reported similar error patterns found in children of MPD and normally developing children, therefore consonant errors per se are not sufficient for differential diagnosis of CAS and overlap with behavioral features identified in other childhood speech disorders as agreed by [14]. Both children with CAS and MPD presented with a speech disorder characterized by high rates of consonant omission and substitution errors, however, the study revealed that omission errors in children with suspected CAS were accounted for almost exclusively by deletion of the final consonant in words, which is representing a consistent pattern of syllabic error, while children with MPD committed omission errors of different consonant positions in words. The current study revealed session-to-session variability of the CAS group evident by their wide range of standard deviation. The inconsistent trajectory of performance in articulation testing for these children characterizes CAS as reported by [5,15]. Because of this inconsistency in performance, best practice for differential diagnosis of CAS should include assessment on multiple occasions, using both single-word testing and spontaneous sampling. Spontaneous sampling in CAS remains important since it reflects day-to-day speech production more closely than single-word articulation testing. Table 11 Prosodic accuracy.
Mean score/examination/child Standard deviation
MPD
CAS
Normal
8.5*,y 2.33
2.5z 2.54
10 0
N.B. Children were rated on a 3-point scale; 10 points were awarded for a correct response, 5 points for fair and 0 point for poor response. The 3 trials of each group were averaged to give the final score (out of 10) of each group. * Means a significant difference between MPD and CAS groups, P value <0.01. y Means a significant difference between MPD and normal groups, P value <0.01. z Means a significant difference between CAS and normal groups, P value <0.01.
Syllables were examined to explore the relationship between syllable complexity and its accuracy. Syllables were analyzed for syllable number, shape and sequencing accuracy. Children with CAS and MPD generally maintained the targeted number of syllables in their productions for monosyllables. Target words with 2 syllables or more were significantly less accurate in children with suspected CAS than the MPD group and the control group. Syllable accuracy analyses showed that children with suspected CAS had significantly lower syllabic shape accuracy scores than those with MPD. Simple open monosyllables, and open disyllables with duplication or harmony were the only syllable shapes relatively correctly produced in children with suspected CAS, suggesting that persistent omission errors likely represent a deficit in syllabic construction rather than sound-specific errors. This matched with previous results which suggested that consonant errors affecting syllable structure are more common in CAS than in common speech disorders not identified as CAS [5,14,15]. Accuracy of polysyllables and consonant close-to-clusters varied significantly between CAS group and the other 2 groups. This highlights the difficulty confronted by CAS patients in relation to certain syllabic shape that could be used as a diagnostic marker. Errors typically involved omission of one of the consonants, indicative of a more general phenomenon of syllable shape reduction rather than a consonant difficulty. This condition reflects an impaired ability to construct accurate word shape and indicates that syllable-level error pattern represents a persisting area of deficit relative to chronological age expectations for children diagnosed with CAS than those with MPD. Due to the paucity of multisyllabic productions in children with apraxia, sequencing errors are often difficult to identify in younger children. Therefore, phonetic transcripts of the multisyllabic words and non-words were examined for sequencing errors, which was applied to document problems in programming of sequential articulatory movements. The results identified a significant deficit in patients with CAS than those with MPD especially for sequencing of multisyllabic nonsense words. This is probably because non-word repetition task does not only represent the combined influence of several speech and language processes (including auditory memory and phonologic processing) but it also entails preparation and execution of a speech movement sequence, which is to say, speech motor control. The results of prosodic assessment showed significant delay in children with CAS than those with MPD across the 3 trials. Thus results of prosodic assessment can be added to the diagnostic markers of CAS. This is in agreement with the results of Marion et al., Shriberg et al., Skinder et al. and Peter and Stoel-Gammon [16–20]. This study simultaneously examined language; speech and non-speech performances of children identified with suspected CAS and MPD and compared them to normal children. This consideration has enabled identification of common character-
A.A. Aziz et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 578–585
istics, as well as atypical features of the speech-disordered groups. It has also ascertained the interaction between speech and language performances among these children. The descriptive data showed that children with suspected CAS showed relatively greater difficulty for prosodic features and consonant cluster productions in addition to difficulty in processing syllabic information, indicating a clear need to address these variables for differential diagnosis and treatment of children with CAS. References [1] R. Kent, Research on speech motor control and its disorders: a review and prospective, Journal of Communication Disorders 33 (2000) 391–428. [2] S.L. Velleman, Phonotactic therapy, Seminars in Speech and Language 23 (1) (2002) 43–56. [3] American Speech Language and Hearing Association, Childhood Apraxia of Speech, Retrieved on July 1, 2007, from http://www.asha.org/docs/html/ PS2007-00277.html. [4] L. Nijland, B. Maassen, S. Meulen, Evidence of motor programming deficits in children diagnosed with DAS, Journal of Speech, Language, and Hearing Research 46 (2) (2003) 437. [5] B. Lewis, L. Freebairn, A. Hansen, S. Iyengar, H. Taylor, School-age follow-up of children with Childhood Apraxia of Speech, Language Speech and Hearing Services in Schools 35 (2004) 122–140. [6] S. Drummond, J. Ashe, Developmental verbal apraxia and phonological disorders: speech & non-speech performances, Presented at the American Speech Language Hearing Association National Convention, November 2005. [7] K. Shipley, McAfee J (Eds.), Assessment in Speech-language Pathology: A Resource Manual, second ed., Singular Publishing Group, San Diego, 1998. [8] R.L. Thorndike, E.P. Hagen, J.M. Sattler, Stanford–Binet Intelligence Scale, fourth ed., Riverside Publishing, Itasca, IL, 1986.
585
[9] M.N. Kotby, A. Khairy, M. Baraka, N. Rifaie, El Shobary, Language testing of Arabic Speaking Children, Proceeding of the XXIII World Congress Of The International Association of Logopedics and Phoniatrics 1995. [10] M.N. Kotby, S. Bassiouny, El Zomor, E. Mohsen, Standardization of an articulation test, in: Proceedings of the 9th Annual Ain Shams Medical Congress, Egypt, 1986. [11] B.L. Davis, K.J. Jakielski, T.P. Marquardt, Developmental Apraxia of Speech: determiners of differential diagnosis, Clinical Linguistics and Phonetics 12 (1) (1998) 25–45. [12] L.D. Shriberg, J. Kwiatkowski, Phonological disorders: III. A procedure for assessing severity of involvement, Journal of Speech and Hearing Disorders 47 (3) (1982) 256–270. [13] B.L. Davis, S.L. Velleman, Differential diagnosis and treatment of Developmental Apraxia of Speech in infants and toddlers, Infant Toddler Intervention 10 (2000) 177–192. [14] L.D. Shriberg, D.M. Aram, J. Kwiatkowski, Developmental Apraxia of Speech: II. Toward a diagnostic marker, Journal of Speech Language and Hearing Research 40 (2) (1997) 286–312. [15] T. Marquardt, H.M. Sussman, T. Snow, A. Jacks, The integrity of the syllable in Developmental Apraxia of Speech, Journal of Communication Disorders 35 (2002) 31–49. [16] M. Marion, H. Sussman, T. Marquardt, The perception and production of rhyme in normal and developmentally apraxic children, Journal of Communication Disorders 26 (1993) 129–160. [17] L.D. Shriberg, D.M. Aram, J. Kwiatkowski, Developmental Apraxia of Speech: I. Descriptive and theoretical perspectives, Journal of Speech, Language, and Hearing Research 40 (1997) 273–285. [18] L.D. Shriberg, D.M. Aram, J. Kwiatkowski, Developmental Apraxia of Speech: III. A subtype marked by inappropriate stress, Journal of Speech, Language & Hearing Research 40 (1997) 313–337. [19] A. Skinder, E.A. Strand, M. Mignerey, Perceptual and acoustic analysis of lexical and sentential stress in children with Developmental Apraxia of Speech, Journal of Medical Speech-Language Pathology 7 (1999) 133–144. [20] B. Peter, C. Stoel-Gammon, Timing errors in two children with suspected Childhood Apraxia of Speech (sCAS) during speech and music-related tasks, Clinical Linguistics & Phonetics 19 (2) (2005) 67–87.