Speech Characteristics of Children . After Otitis Media n Alice
The purpose of this investigation was to obtain preliminary descriptions of the speech characteristics of children with histories of otitis media. Speech samples were collected from 20 children between the ages of 3 years, 6 months and 5 years, 5 months who had had at least three episodes of otitis media. Results indicated that most of the children performed below age norms on at least two of the speech characteristics studied. Identification of such deficient speech patterns may aid the practitioner in recognizing and referring children who require evaluation of their speech and language development. Implications for further research with this population were suggested. J PEDIATR HEALTH CARE.
L hronic otitis media, a common childhood disease, ofien results in a temporary conductive hearing loss. For some children, the first years of life are plagued with recurrent otitis media accompanied by fluctuating hearing acuity. Otitis media is second only to upper respiratory infections as a reason for physician visits. The highest incidence of otitis media occurs early in childhood, declining after 6 years of age (Pashley, 1984). Ruebin and Hanson (1979) suggested that children with histories of recurrent otitis media often have more difficulty with auditory processing skills than their otologically normal peers. Menyuk (1980) divided the development of speech processing abilities into four distinct periods, extending from infancy into the early school years. The accomplishments of each period of development were said to be dependent on the accomplishments of the previous period. The child who has a stable auditory system, particularly during the critical period of speech development, may best develop the ability to produce and process verbal stimuli with rules learned through experience and perception. However, the child with recurrent otitis media during these years often lacks this stability. All health care professionals working with young children should be aware of the potential for speech problems in this population. Appropriate referrals
Alice T. Dyson is an Assistant Professor University of Florida, Gainewille. Alice E. Holmes is an Associate University of Florida, Gainesville.
at the Department
at the Department
Deanna V. Duff&t IS a speech-language pathologist at the James Paton Memorial Hospital, Gander, Newfoundland, Canada. Reprint requests: Alice T. Dyson, University of Florida, Gainesville, JOURNAL
Ph.D., Department FL 3261 1.
of Speech, 455 ASB,
allowing for early recognition and intervention may prevent persistent communication disorders. Recently researchers have attempted to determine whether speech delay in children having chronic otitis media can be predicted from medical history, specific speech characteristics, or other factors. A longitudinal study by Paden, Novak, and Kuklinski (1985) evaluated 40 children from 18 to 36 months of age. These children were referred to otolaryngologists and were diagnosed as having chronic otitis media. Evaluations of these children at 3-month intervals revealed five variables that may be useful in predicting whether or not normal speech skills will be achieved by 3 years of age. These variables included audiologic factors, speech characteristics, and the insertion of myringotomy tubes. Speec:h characteristics occurring with unusual frequency in children with histories of otitis media were also reported by Shriberg and Smith (1983). Specifically, patterns of errors affecting word-initial consonants and nasal consonants occurred at significantly higher levels in the otitis media children than in children with no histories of otitis media. Equivocal results such as those reported by Teele et al. (1984) and discussed by Paradise and Rogers (1986) suggest the need for further research in this area. If the speech development characteristics of this population can be identified, health care professionals could feel more confident in screening and referring children with possible deviations for speech-language pathology services. The purpose of this study was to obtain preliminary descriptions of the speech characteristics of a group of children with histories of otitis media before 3 years of age. Particular attention was given to those characteristics that might help practitioners distin261
. TABLE sound
1 Age ranges
Volume 1, Number 5 September-October 1987
‘May still be somewhat distorted until 6 or 7 years of age. tAll sounds are produced but may not be produced correctly
guish the speech development of children in this group from normally developing speech systems. n
1983). These patterns are usually present to some extent m normally developing children during the early speech acquisition period between 2 and 5 years of age. However, when they persist at high levels in older children who use longer and more complex language constructions, intelligibility is difficult and a diagnosis of “phonological disorder” is often made. Seven deficient patterns were examined. 1. Cluster reductkm: omission of one or more sounds from a cluster of consonants (e.g., spoon + poon or cups -+ cus). Clusters of consonants are usually considered more difficult for children to produce than single consonants. Normal children first simplify clusters by omitting one or more sounds and then gradually learn to produce them correctly, typically between 4 and 5 years of age. By the age of 4 years, most children include both consonants in the cluster, although they may not be produced correctly (Grunwell, 1982).
Subjects were 20 children, 12 boys and 8 girls, between the ages of 3 years, 6 months and 5 years, 1 month (mean 4 years, 3 months). The subjects were obtained from two sources, the outpatient ear, nose, and throat clinic of a university medical center and day-care facilities in the same community. All subjects had three or more medically documented episodes of bilateral otitis media within a 12month period before the age of 3 years. Only children whose reported ages of sitting, walking, first words, and combining words were within normal limits on the Denver Developmental Screening Test (Frankenburg, Dodds, & Fandal, 1973) were selected. None of the children had received speech remediation. Their parents reported no known physical disabilities that would be expected to interfere with speech and/or language development. Ten of the children had received myringotomy tube placement at least once before the age of 3 years, whereas the other 10 had never been treated with myringotomy tubes. These two halves of the group were closely matched as to age and sex. All evaluations were completed in a university clinic. Hearing testing was conducted by a certified audiologist with the use of a calibrated diagnostic audiometer (model 170 1, Grason-Stadler, Inc., Littleton, MA). Pure tone and speech reception thresholds were within normal limits for all subjects (less than or equal to 25 dl3 hearing level bilaterally). Taped speech samples were collected while the children named a set of 55 one-syllable words selected to optimize the po.ssible occurrence of seven deficient speech patterns often found to characterize the speech of unintelligible children (Hodson & Paden,
t can be seen that liquid deviation and cluster reduction were the most frequently occurring patterns (of deficient speech), followed by stridency deletion, final consonant deletion, and nasal deviation.
2. Final consonant omission: omission of a single consonant following the vowel in a syllable, usually at the end of a word (e.g., boat + boa). Normal children under 2 years of age are frequently reported to omit final consonants, but this pattern occurs only on occasional words after the age of 3 years (Dyson & Paden, 1983; Grunwell, 1982). 3. Initial sound change: omission or substitution of a glottal sound /h, 3i for a consonant or consonants preceding the vowel in a syllable (e.g., boat * oat or hoat). This unusual pattern has never been reported to be common among normally developing children. 4. Stridency deletion: omission of one of the strident consonants /s, z, f, v, S ,3 , t/‘ 4 / or substitution of another (nonstrident) consonant for a strident consonant (e.g., bus + bu, or bus -+ but). Stidency refers to the high intensity noise produced as air strikes a hard surface, such as the back of the teeth. Although normal children may distort strident sounds (especially /s/) until 6 or 7 years of age, these sounds are seldom omitted after 4 years and seldom replaced by nonstrident sounds after 3 years of age (Hodson & Paden, 1983; Khan & Lewis, 1986). A pronouncing key for phonetic symbols is included in the appendix.
lournal oi Pediatric
H TABLE 2 Percentages of occurrence occurrence reported in two normative
than norms than norms
of seven deficient studies
patterns for 20 subjects compared
by Haelsig and Madison (1986). by Khan and Lewis (1986).
5. Velar~ontin~:: substitution of a consonant produced farther forward in the mouth for a velar consonant (e.g., back -+ bat). Velars /k, g, :, / are produced correctly by most normal children before the age of 3 years (Dyson, 1986). 6. Lipid deviation: errors in production of the liquid consonants /II, r/. Liquids before a vowel may be omitted, replaced by a glide /w, j/ (called gliding) or a stop /p, b, t, d, k, g/ (called stopping) (e.g., lamb + amb, or wamb, or tamb). Liquids after a vowel may be replaced by a vowel or omitted (e.g., car + cauh or ca) (Hodson & Paden, 1983). Liquids are usually mastered by normal children by the age of 5 years, althoughi many children learn them much earlier (Hodson & Paden, 1983; Grunwell, 1982). 7. Nasal deviation: omission of a nasal consonant /m, n, 3 I, substitution of one nasal for another, or addition of a stop consonant immediately before or after a nasal (e.g., might + ight, night, or mbight) (Shriberg & Smith, 1983). This pattern has not been
reported to be common among normally developing children. Age ranges of normal acquisition of these types of sounds are summarized in Table 1. As indicated, normal children show considerable variation in the development of these sounds. Two experienced listeners trained in transcription of children’s speech samples transcribed the audiotapes with the use of a consensus procedure to ensure validity of transcription (Shriberg, Kwiatkowski, & Hoffman, 1984). The transcribed samples were examined for occurrence of each of the seven deficient patterns of interest. The percentage of occurrence of each of the seven patterns was computed for each subject by calculating the number of occurrences and then dividing it by the total number of possible occurrent es. Percentage of occurrence scores for each subject were compared with normative data derived from two reported studies of normal children (Haelsig &
Madison, 1986; Khan & Lewis, 1986). Care was taken to replicate procedures described in the normative studies as closely as possible in collection and analysis of the data. Children in both normative studies were selected from preschools and day-care centers and were screened and selected in the same manner used in the present study. Samples were collected with very similar tape recording equipment under similar conditions. ,4fter collection, the samples were transcribed and scored according to procedures common to both normative studies. The norms available from these two sources are reported by age ranges and allow comparison of each subject’s performance with his or her age peers.
I t is tempting commonly deviation, consonant pinpointing for normal
Volume 1, Number 5 September-October 1987
to suggest that the three most occurring patterns-liquid cluster reduction, and final deletion-might have value in children with a poor prognosis speech development.
’ RESULTS The percentages of occurrence of each of the seven deficient patterns examined for each of the 20 subjects are presented in Table 2. Because of the small sample size in this study, statistical treatment of the data was limited to descriptive analysis. It can be seen that liquid deviation and cluster reduction were the most frequently occurring patterns, followed by stridency deletion, final consonant deletion, and nasal deviation. Initial sound change and velar fronting were uncommon in this group of children; each was used frequently by only one child. The range of scores was quite large for the most frequently occurring patterns as there was always at least one child who did not use the pattern at all. For example, the scores for liquid deviation and cluster reduction ranged from zero to 100% and from zero to 75%, respectively. The 1986 study by Haelsig and Madison of 50 children offered normative data for four of the patterns studied: cluster reduction, final consonant deletion, velar fronting, and liquid deviation (liquid gliding and liquid vocalization). Khan and Lewis (1986) reported norms based on 852 preschool children for five of the patterns included in this study, the four patterns mentioned for Haelsig and Madison as well as stridency deletion. It can be seen in Table 2 that when norms from both studies were available, they agreed in all but two instances, liquid deviations for subjects 3 and 6. In both cases the subjects exceeded
the norms listed by Khan and Lewis but not those of Haelsig and Madison. Comparison of the individual patterns with the norms indicated that 16 children exceeded the norm of at least one study on liquid deviation. Twelve children exceeded the norm of both studies on cluster reduction, eight children on final consonant deletion, and one child on velar fronting. Seven children exceeded the norm on the one study available on stridency deletion. w DISCUSSION Clearly most children in this study were using the deficient patterns at higher levels than those demonstrated by their age peers in the two normative studies. Thirteen of the 20 children studied demonstrated percentages of occurrence higher than at least one set of norms on two or more patterns; 11 of these scored higher on three or more patterns. Subjects 1, 4, and 11 exceeded the norms on four patterns. It is interesting to note that four of the children-subjects 2, 9, 10, and 16-performed within or better than the age norms despite their histories of otitis media. No clear differences were indicated between children who had received myringotomy tube insertions and those who had not. It is tempting to suggest that the three most commonly occurring patterns-liquid deviation, cluster reduction, and final consonant deletion-might have value in pinpointing children with a poor prognosis for normal speeeh development. Two of these patterns-cluster reduction and final consonant deletion-were also found by Paden et al. (1985) to be useful in distinguishing between children with histories of otitis media who would reach normal speech levels by the age of 3 and those children whose speech delay would continue beyond that age. Because the cross-sectional nature of the present study did not allow observation of these patterns over time, it is not clear that the high levels found would persist as the children continued to develop. However, children at all ages studied (Table 2) showed higher than expected levels on the three patterns. Although norms are not currently available on the expected occurrence of the two deficient patternsnasal deviation and initial sound change-described by Shriberg and Smith (1983), nasal deviation was observed at least twice (9% or more) in 9 of the 20 children. If, as indicated by Shriberg and Smith, this pattern is extremely uncommon in normally developing children at any age, it would appear that its high occurrence in these subjects is significant. This study may be interpreted as offering evidence of the value of this pattern in predicting speech difficulty
Journal of Pediatric Health
in children with histories of otitis media. However, the second pattern, initial sound change, was seen in only three children.
the cross-sectional nature of the present study did not allow observation of these patterns over time, it is not clear that the high levels found would persist as the children continued to develop.
It is important to note the great variability in speech skills found in this study. The mere presence of otitis media does not appear to indicate that a child will be delayed in speech development, but higher than expected occurrences of the deviant speech patterns included in this investigation were noted in most children. Speech patterns, such as those assessed in this study, may help to alert the practitioner to potential speech delays that may require further evaluation and intervention by a speech-language pathologist. The results of this study may have been influenced by external factors. Neither the current study nor the two normative studies controlled for socioeconomic status or geographic location. Further studies that control such factors and the duration of episodes of otitis media (Paradise & Rogers, 1986; Teele et al., 1984) with larger sample sizes are needed. The purpose of the present investigation was to obtain preliminary descriptions of the speech characteristics of a group of childr en with histories of otitis media. Future investigation of this topic is warranted.
Dyson, A., Kr Paden, E. (1983). Some phonological acquisition strategies used by two-year-olds. Journal of Childhood Communicatiun D~ordeq 7, 6-18. Frankenburg, W. K., Dodds, J. B., & Fandal, A. W. (1973). Denver Develqmental Screening Test. Boulder, CO: University of Colorado Medical Center. Grunwell, P (1982). Clinical phonology. Rockville, MD: Aspen Systems. Haelsig, I’., &Madison, C. (1986). A study of phonological processes exhibited by 3-, 4-, and 5year-olds. LanJtiage, Speech, and Hearing Sties in +hols, 17, 107-114. Hodson, B. W., & Paden, E. P. (1983). Target& intell~ible speech. San Diego, CA: College-Hill Press. Khan, L., & Lewis, N. (1986). The Khan-Lewkphonolo~kl analris. Circle Pines, MN: American Guidance Service. Menyuk, P. (1980). Effects of persistent otitis media on language development. Annah of Otology, Rhinology, and La+ynaolom (Suppl 68), 89, 257-263. Paden, E. P., Novak, M. A., & Kuklinski, A. L. (1985). Predictors of phonological inadequacy in young children prone to otitis media. Presented at the Annual Convention of the American Speech-Language-Hearing Association, Washington, DC. Paradise, J. I-., & Rogers, K. D. (1986). On otitis media, child development, and tympanostomy tubes: New answers or old questions? Pediutiu, 77, 88-92. Pashley, N. R. (1984). Otitis media. In Northern, J. L. (Ed.), HearingDisorders (2nd ed), pp. 103-110. Boston: Little Brown &co. Ruebin, R. J., & Hanson, D. G. (1979). Summary of discussion and recommendations made during the workshop on otitis media and development. Annals of Otology, Rbinolo~, and LarynHoloHy (Suppl 60) 88, 107-111. Shriberg, L., Kwiatkowski, J., & Hoflinan, K. (1984). A procedure for phonetic transcription by consensus. Journal of Speech and Hearing Research, 27, 456-465. Shriberg, L. D., & Smith, A. J. (1983). Phonological correlates of middle-ear involvement in speech-delayed children: A meth odological note. Journal of Speech and Hearing Research, 26, 293-297. Teele, D. W., Klein, J. O., Rosner, B.A., &The Greater Boston Otitis Media Study Group. (1984). Otitis media with effusion during the first three years of life and development of speech and language. Pediahics, 74, 282-287.
This investigation of the speech characteristics of 20 children (mean age 4 years, 3 months) with histories of otitis media indicated that most performed below age norms on at least two deficient patterns. The presence of such deficient speech patterns may aid the practitioner in identifying and referring children who require evaluation of their speech and language development. n REFERENCES Dyson, A. (1986). Development of velar consonants mal two-year-olds. Joumrd of Speech and Hearing 493-498.
among norResearch, 29,
/ml as in my In! as in no ltji as in sing lpl as in pie lbl as in boy It/ as in tie Id/ as in did lkl as in kite igl as ingo Xl as in f;n /vi as in”van
Is/ as in soup lzl as in zoo lsf as in shoe I31 ItJI
as in vision as in chair I as in jump as in like as in run as in win as in you as in uh oh as in he