International Journal of Pediatric Otorhinolaryngology, 26 (1993) I 17-127 0 1993 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0165-5876/93/$06.00
Chronic otitis media and early speech development: a case study Michael P. Robb a, Jerre L. Psak b and Glenn K. Pang-Ching ’ aJohn A. Burns School of Medicine, University of Hawaii and, Kapiolani Medical Center for Women and Children, Honolulu, HI (USA) b U.S. Army Medical Service Corps, Bayne-Jones Army Community Hospital, Audiology Section, Fort Polk, LA (USA) and ’ John A. Burns School of Medicine, University of Hawaii, Honolulu, HI (USA) (Received 17 March 1992) (Revised version received 17 June 1992) (Accepted 21 June 1992)
Key words: Development;
Abstract A case study of the speech development in a male infant with chronic otitis media is reported. The phonetic behavior characterizing the child’s vocalizations was sampled monthly between the ages of 11 and 21 months, as he progressed from pre-speech to early speech periods of language development. Results of monthly phonetic inventory analyses indicated age-appropriate types of consonants in his pre-word and later-word vocalizations. However, the child’s repertoire of consonants was considerably reduced as he first began to produce meaningful speech. Results of phonetic diversity analyses revealed an overall lack of phonetic complexity in his vocalizations throughout the course of study. In general, the child’s early sound productions were characteristic of developmental delay, closely resembling the speech patterns found among severely hearing-impaired children. Findings are discussed with respect to the probable influence of chronic otitis media with effusion on a child’s eventual speech development.
Introduction Otitis media is among the most common diseases of childhood and the single most common diagnosis made by pediatricians [l]. Based on persistence of the Correspondence to: Michael P. Robb, Ph.D., University of Hawaii, John A. Burns School of Medicine, Div. Speech Pathology and Audiology, 1410 Lower Campus Drive, Honolulu, Hawaii, 96822, USA.
disease, otitis media can be classified into three major forms. The rapid onset and short duration (i.e., less than three weeks) of signs and symptoms of inflammation in the middle ear is termed acute otitis media . Acute otitis media accompanied by middle ear effusion COME) occurs in almost every child at some time during the first decade of life [29-311. For infants and children who have frequent episodes of otitis media, the condition is termed recurrent OME. Bluestone  defines recurrent otitis media as three or more episodes occurring within a six month period. Chronic OME differs from acute OME and recurrent OME in clinical course, bacteriology, pathophysiology, and rate of complications and sequelae . Whereas, acute OME is rapid in onset, quick in resolution and usually free of significant sequelae, chronic OME is slow and insidious in its onset, tends to be persistent, and is very destructive. Perhaps the most obvious difference between the various forms of otitis media is that chronic OME involves pus behind the tympanic membrane and/or rupture of the membrane. The visible discharge of pus (otorrhea) into the external auditory canal is generally intermittent and foul-smelling. The presence of OME early in life frequently results in a fluctuating conductive hearing loss . The fluctuation in hearing sensitivity may have a significant impact on a child’s subsequent speech (i.e., phonetic) development [7,9,13,29]. The hypothesis is that children with OME are at risk for not obtaining the auditoryverbal experience necessary for acquiring speech normally. The most common method of testing this hypothesis has been to describe the speech characteristics of children known to have a history of OME [16,21-23,311. That is, children are sampled ‘after the fact’ and relationships are then drawn between the child’s present speaking abilities and past bouts with OME. Yet most children experience OME well before the onset of their first true words - the time period when infants are unable to communicate their symptoms. Few prospective studies exist examining the influence of OME on the general communication development in children across pre-speech (i.e., babbling) and early speech (i.e., single word) periods of development [7,18]. Further, there are no specific data pertaining to the impact of OME upon an infant’s speech sound acquisition. The study reported here presents a longitudinal investigation of one child experiencing chronic OME during pre-speech and early speech periods of development. Although we recognize the inherent limitations of a single-subject study, the time series sampling of this particular child afforded a unique opportunity to test the effects of chronic OME during the formative years of speech development. The objective of the present study was to examine the influence of chronic OME on the range and types of consonants produced across pre-speech and early speech periods of language development. Methods Subject
The child profiled in the current investigation was a Caucasian male named KC. The data collected for K.C. span the age range of 11 months to 22 months. The
child was initially recruited, along with seven other children, as part of an extensive longitudinal study of selected acoustic and phonetic characteristics of normally developing infants . Each child’s vocalization behavior was sampled monthly over a 12 month period, spanning linguistic periods of pre-speech and early speech development. Because of K.C.‘s chronic OME condition he was excluded from the original subject sample, although he was still sampled monthly over a 12-month period. At the beginning of the study, K.C. was producing strictly pre-speech (i.e., babbling, jargon) vocalizations. At the conclusion of the study, he had an expressive vocabulary of 20 single words. His communication abilities were formally assessed at 11 months and found to be within normal age expectations . His language abilities were reassessed at 21 months and found to be five months delayed for his chronological age. A summary of K.C.‘s general characteristics across the 12-month period are provided in Table 1. During the first recording session (11 mos), K.C. was diagnosed as presenting with chronic OME. He had already undergone bilateral myringotomy with tympanoplasty tube insertion. Throughout the course of the study K.C. received antimicrobial treatment. Bluestone  has noted that middle ear discharges (otorrhea) are a frequent complication of tympanoplasty tube placement. Not surprisingly, K.C. was found to have observable auditory canal discharges at 16 months of age. His hearing abilities were informally assessed across the course of study. During the first session, K.C. presented with a hearing handicap consistent with the concurrent chronic OME . He did not respond to speech spoken at normal conversational levels (i.e., 55-60 db SPL at 1 meter). His mother reported that K.C. was only responsive to speech which was spoken louder than normal. During the last session (21 mos), K.C. was found to be responsive to verbal requests, and localize to both speech and nonspeech (e.g., noise burst) stimuli. Standard audiological evaluations were performed periodically following completion of the study. During these follow-up tests, the child demonstrated normal hearing sensitivity at better than 20 dB HL  at octave intervals extending from 500-4000 Hz. Vocalization samples
One-hour vocalization samples were gathered from K.C. while he played with quiet, familiar toys. The child was encouraged to vocalize as he interacted with his caregiver and one of the researchers (M.R.). The goal of these recording sessions was to obtain a sample of roughly 50 comfort-state vocalizations. A comfort-state vocalization was defined as the production of vowel-like and consonant-like sounds occurring within an audible exhalation/ airstream . The child wore a condenser lapel microphone clipped onto his chest clothing to within 30 cm of the mouth. Occasionally, K.C. responded adversely to the microphone clip, in which case, the microphone was handheld by the researcher to within 30 cm of the child’s mouth. The output of the microphone was connected to a stereo cassette tape recorder (Marantz PMD-360). K.C. cried throughout the sessions at 16, 20, and 22 months, precluding the collection of comfort vocalizations during those age periods.
120 TABLE I General anatomical and linguistic growth characteristics Age
11 12 13 14 15 16 17 18 19 20 21 22
16.2 79.0 83.8 83.8 84.4 84.4 85.0 85.0 86.3 86.3 86.3 86.3
46.3 47.0 47.0 47.0 48.2 48.2 48.2 48.2 48.6 48.6 48.6 48.6
Vocalization Sample Size
12 24 63 38 crying 87 70 66 crying 48 crying
_ _ _ _ 16 _
The table also includes the language-age performance on the Receptive-Expressive guage (REEL) scale  at the beginning and conclusion of data collection.
Vocab. 0 0 0 2 3 2 10 10 20
To evaluate K.C.‘s phonetic development across pre-speech and early speech periods of development, three specific points in his word acquisition process were identified. Determination of the child’s expressive vocabulary allowed the monthly vocalization samples to be categorized as belonging to one of three linguistic periods, pre-words (0 words), early-words (l-9 single words), and later-words (10 + single words). The pre-word period (11-13 months) was believed to be most representative of K.C.‘s pre-speech vocalizations, whereby no recognizable words were produced. The early-word period (14-17 months) contained his earliest single word productions, while the later-word period (18-21 months) reflected a point at which K.C. was producing single words on a consistent basis. Phonetic analysis
All of the child’s distinguishable vowel and consonant vocalizations produced during each recording session were transcribed via headset using the symbols of the International Phonetic Alphabet (IPA). The phonetic symbols of the IPA are provided in Appendix A (p. 125). In an attempt to characterize the developmental progression of speech sound acquisition for K.C., two measures of phonetic behavior were calculated: (1) a monthly phonetic inventory and (2) a monthly estimate of phonetic diversity. Phonetic inventory. A phonetic inventory analysis involves tallying the occurrence of individual consonants or vowels produced by a child within a predetermined time frame (e.g., one hour) . The various sounds produced by the child are summarized to provide a profile of his/her individual speech sound system. Monthly phonetic inventories, listing the consonants produced by K.C. were created. In order for a consonant to be included in K.C.‘s monthly inventories, the sound had to be produced at least twice within a recording session. Adherence to
this criterion ensured that only stable, frequently occurring consonants were included. Phonetic diuersity. Phonetic diversity (or complexity) refers to a child’s use of different articulatory movements across time or within a vocalization [31. A basic form of phonetic diversity is the production of a consonant and vowel sound within a single vocalization. The ratio of consonants-to-vowels (C : V) is one measure of phonetic diversity which has been shown to summarize a child’s ‘running’ or ongoing articulation. Specifically, the proportional occurrence of consonants increases relative to vowels as a function of chronological age across the first two years of life . The present study used the C : V ratio as a measure of evaluating K.C.‘s phonetic diversity in comparison to normally developing children. The consonants and vowels comprising every comfort vocalization produced by K.C. were identified and separated into categories. The total number of vowels was then divided into the total number of consonants to create a C : V ratio for each monthly session. The monthly C: V ratios were then plotted as a function of the child’s chronological age to display the development of phonetic diversity.
Results Phonetic inventories The total number of transcribable vocalizations produced in the nine monthly sessions was 492. The 9 samples ranged in size from 12 (12 mos) to 87 vocalizations (17 mos>. Collectively, 555 consonants were identified in K.C.‘s vocalizations. The types of consonants contained in KC’s monthly samples are listed in Table II. His consonant inventories ranged in size from 2 (14 mos) to 10 (21 mos). The consonants found to occur most often throughout the sampling period were /b. d, m, h/. To assess the similarity of K.C.‘s consonant inventory data with previous reports of same-aged normal children, the monthly data were compared to the published data of Glazewski et al.  and Stoel-Gammon . In general, similarity in size and type of consonants was noted during the earliest and latest sampling periods. K.C.‘s inventory was considerably reduced at 14 and 1.5 months of age. Interestingly, the similarities and differences in inventories corresponded roughly to the three predefined periods of language acquisition. His pre-word (11-13 mos> and later-word (18-21) inventories were similar in size and type to those found for normally developing children. In fact, his inventories at 11, 13 and 21 months were larger than those reported in the literature. On the other hand, his consonant inventories during the early-word period (14-17 mos) were much smaller than those found for normally developing children. A graphic display of K.C.‘s monthly consonant inventories across the three linguistic periods is shown in Fig. 1. Phonetic dicers& The C : V ratios calculated for each of K.C.‘s monthly samples are displayed in Fig. 2. A c’ : V ratio of 1.00 would reflect an equal number of consonants and
122 TABLE II K. C. ‘s monthly consonant inventories
/b,d,t,m,hW * (b,d,t,hJ
11 mos 12 mos
/d,n,h,s/ * (d,g,h,w)
/b,t,m,h,w/ * (d,h,.i)
/m,h/ * (b,d,g,m,hj)
/m,h/ * (b,d,m,n,W
/d,km,h,w/ * * (b,d,t,m,n,h,w)
/b,d,t,p,m,n,h,z,s,w/ * * (b,d,t,m,n,h)
* Glazewski et al. (1984) data * * Stoel-Gammon (1985) data The applicable same-age inventories from Glazewski et al.  and Steel-Gammon parentheses.
 are shown in
w is b 22
AGE (months) Fig. 1. Number of different consonants produced by K.C. plotted as a function of chronological age (in months) and language level.
5 :, E
AGE (months) Fig. 2. K.C.‘s monthly C: V ratios (M) plotted as a function of chronological age. The figure also contains the corresponding C: V ratios for normally developing children (0) (taken from Robb and Bauer ). First-order polynomials are fit to the set of data points for normals (dashed line) and K.C. (solid line).
vowels produced in a vocalization sample. His monthly ratios ranged from 0.30 (15 rnos) to 1.20 (19 mos), with a majority of the samples yielding ratios lower than 1.00. In order to evaluate the strength of relationship between K.C.‘s monthly C : V ratio and increasing age, a Pearson correlation statistic was computed. The test was not significant (1. = 0.50, P > 0.05). The monthly C: V ratios for normally developing children are also shown in Fig. 2. Upon comparison to the normal developmental pattern, K.C. showed a trend toward increasing consonant use as a function of age, although his C : V ratios were consistently lower than expected. Robb and Bauer  have found normally developing children to demonstrate C : V ratios above 1.00 by 16 months of age. K.C.‘s C : V ratios did not rise above 1.00 until 19 months of age, reflecting a general lack of phonetic diversity. His vocalizations contained relatively fewer productions of consonants compared to vowels.
Discussion The consonants /b/, /d/, /m/ and /h/ were the most prevalent sounds in K.C.‘s pre-speech and early speech vocalizations. There was a marked reduction in inventory size between 14-17 months of age, which corresponded to his early-word period of language development. The reduced inventories may have been a reflection of the vocalization sample size obtained during those particular recording sessions. However, examination of his vocalization samples at earlier and later sessions showed instances in which fewer vocalizations were collected yet more consonantal types were noted. The noticeable differences between K.C. and normally developing infants may be partially related to the influence of linguistic factors on the child’s acquisition of
speech sounds. Locke  has suggested that prior to meaningful speech production, a child’s use of particular consonants and vowels is ‘biologically given.’ This notion of a physiological predisposition for early developing sounds is based on observed commonalities in infants reared in various language communities, as well as in the pre-speech vocalizations of children with Down Syndrome  and deafness . However, once a child attains first-word status, his/her sound productions begin to reflect the characteristics of the particular language the child is learning to speak [lo]. Indeed, KC.3 pre-speech productions were similar to what is found among normally developing children, thus suggesting that the role of ‘learning’ seems to be rather minimal during pre-speech productions . Controversy exists regarding the course of normal speech sound development between pre-speech and early speech periods. Jakobson [lo] believed that children undergo a reorganization of their phonetic repertoire as they begin to produce single words, resulting in a considerable reduction of the child’s phonetic inventory. On the other hand, the more commonly accepted view of children’s speech development is that continuity occurs between pre-speech and early speech periods, with similarity in the size and type of phonetic repertoire noted across both periods [12,20]. Contrary to expectation, K.C. seemed to follow a developmental course similar to what Jakobson [lo] would predict. His consonant productions were greatly reduced as he entered the early-word period of development. K.C.‘s reduction in phonetic inventory may have reflected a general difficulty in learning language. Important, however, is the fact that during K.C.‘s early-word period, especially at 16 months of age, he was demonstrating observable suppurative OME. One might assume that this period was most representative of fluctuating middle ear functioning for the child, thereby contributing to his reduced phonetic inventory. Interestingly, the finding of reduced inventory corresponds to what has been noted for same-aged severely hearing-impaired children [24,27,28]. The consonants identified as most prevalent among hearing-impaired infants are /b, p, w, m, h/ [24,271. K.C.‘s most frequent productions during the early-word period were /m, h/. Yet by 21 months of age, when K.C.‘s hearing ability was judged to be within normal limits, his phonetic inventory increased markedly. Assessment of K.C.‘s monthly phonetic diversity provided perhaps the most compelling evidence of an overall delay in his speech sound production abilities. Calculation of a monthly C: V ratio allowed for estimating the articulatory complexity of K.C.‘s vocalization behavior. While a monthly increase in phonetic diversity was found, his monthly C : V ratios were consistently lower in comparison to normally developing children. In other words, K.C. was demonstrating. development of phonetic diversity, albeit at a slower rate. The robust difference in phonetic diversity is surprising, considering his phonetic repertoire during the pre-word and later-word periods was essentially normal. In terms of overall developmental progression, K.C.‘s phonetic behavior differed from normally developing children. Both measures of phonetic ability appeared to be sensitive to the subsequent 5-month delay in language ability noted at the conclusion of the study. Although a causal relationship between phonetic behavior and chronic OME cannot be assumed, the pattern of speech development
presented in this case study is reflective of children with fluctuating middle ear function. The results suggest that a child with chronic OME is at risk for developing the American English sound system at a slower rate than normal. As such, early speech and language intervention for children presenting with similar symptoms appears warranted. The current report reminds us once again of the adverse effects of early chronic OME on subsequent speech development, and the need for further detailed investigations.
Appendix A The consonant sounds of General American English. Each sound is given by its dictionary symbol and the symbol in the International Phonetic Alphabet (IPA).
me, aim, smile
“0, pan, snap pea, map, - spill -
tea, ba!, still
key, sic&, skip
be, mob dawn, good
go, - bagvan, give
way? had, hat see, face, fast
wet, SWOOQ lawn, PO&, slap
IPA symbol m n
_ran, press, hard
joke, badge chop, each
show, fresh azure, treasure sis, think
References 1 Adessman, A., Altshuler, L., Lipkin, P. and Walco, G. (1990) Otitis media in children with learning disabilities and in children with attention deficit disorder with hyperactivity,.Pediatrics Suppl., 442-446. 2 ANSI Specifications for audiometers - S3.6, American National Standards Institute, New York (1969). 3 Bauer, H. (1988) The ethologic model of phonetic development: I. Phonetic contrast estimators. Clin. Ling. Phonet. 2, 347-381. 4 Bluestone, C. (1990) Update on otitis media: 1990. Unpublished manuscript. University of Pittsburgh School of Medicine. Pittsburgh, PA. 5 Bzoch, K. and League, M. (1977) Receptive Expressive Emergent Language Scale: For the measurement of language skills in infancy. Baltimore: University Park Press. 6 Downs, M. (1988) Contribution of mild hearing loss to auditory language learning problems. In: Roeser, R. and Downs, M. (Eds.). Auditory disorders in School Children. New York: Thieme Medical Publishers, pp. 186-199. 7 Friel-Patti, S. and Finitzo, T. (1990) Language learning in a prospective study of otitis media with effusion in the first two years of life. J. Speech Hear. Res. 33, 188-194. 8 Glazewski, B., McCune, L., Spillman, D., Betts, N., Durant, C., and Ribau, R. (1984) Consonantal characteristics of babbling between eight and sixteen months. Paper presented at the American Speech-Language-Hearing Association, Washington, DC. 9 Hasenstab, M. (1987) Language learning and otitis media. Boston: College-Hill Press. 10 Jakobson, R. (1968) Kindersprache, Aphasie und allgemeine Lautgesetae. Translated by A.R. Keller. Child Language, Aphasia and Phonological Universals. Mouton: The Hague. 11 Kent, R., Osberger, M., Netsell, R. and Hustedde, C. (1987) Phonetic development in identical twins differing in auditory function. J. Speech Hear. Dis. 52, 64-75. 12 Locke, J. (1983) Phonological acquisition and change. New York: Academic Press. 13 Menyuk, P. (1980) Effect of persistent otitis media on language and development. Ann. Otol. Rhinol. Laryngol., 89, 257-263. 14 Northern, J. and Downs, M. (1984) Hearing in Children (3rd Edn.). Baltimore: William & Wilkins. 15 Oller, D.K., Eilers, R., Bull, D. and Carney, A. (1985) Prespeech vocalizations of a deaf infant: a comparison with normal metaphonological development. J. Speech Hear. Res. 28, 47-63. 16 Paden, E., Novak, M., and Beiter, A. (19871 Predictors of phonologic inadequacy in young children prone to otitis media. J. Speech Hear. Dis. 52, 232-242. 17 Payne, E. and Paparella, M. (1976) Otitis Media. In: Northern, J. (Ed.). Hearing Disorders. Little, Brown and Co., Boston, pp. 119-128. 18 Pearce, P., Saunders, M., Creighton, D. and Sauve, R. (1988) Hearing and verbal-cognitive abilities in high-risk preterm infants prone to otitis media with effusion. J. Dev. Behav. Pediatr. 9, 346-352. 19 Robb, M. and Bauer, H. (1991) The ethologic model of phonetic development: II. The closant curve. Clin. Ling. Phonet. 5, 339-353. 20 Robb, M. and Saxman, J. (1990) Syllable durations of preword and early word vocalizations. J. Speech Hear. Res. 33, 583-593. 21 Roberts, J., Sanyal, M., Burchinal, M., Collier, A., Ramey, C. and Henderson, F. (1986) Otitis media in early childhood and its relationship to later verbal and academic performance. Pediatrics 78, 423-430. 22 Roberts, J., Burchinal, M., Koch, M., Footo, M. and Henderson, F. (1988) Otitis media in early childhood and its relationship to later phonological development. J. Speech Hear. Dis. 53, 424-432. 23 Shriberg, L. and Smith, A. (1983) Phonological correlates of middle-ear involvement in speech-delayed children: a methodological note. J. Speech Hear. Res. 26, 293-297. 24 Smith, B. (1982) Some observations concerning premeaningful vocalizations of hearing-impaired infants. J. Speech Hear. Dis. 47, 439-442. 25 Smith, B. and Oller, D.K. (1981) A comparative study of premeaningful vocalizations produced by normal and Down’s Syndrome infants. J. Speech Hear. Dis. 46, 46-52. 26 Stoel-Gammon, C. (1985) Phonetic inventories, 15-24 months: a longitudinal study. J. Speech Hear. Res. 28. 505-512.
127 27 Stoel-Gammon, C. (1988) Prelinguistic vocalizations of hearing-impaired and normally hearing subjects: a comparison of consonantal inventories. J. Speech Hear. Dis. 53, 302-315. 28 Stoel-Gammon, C. and Otomo, K. (1986) Babbling development of hearing-impaired and normally hearing subjects. J. Speech Hear. Dis. 51, 33-41. 29 Teele, W., Klein, J., Rosner, B. and The Greater Boston Otitis Media Study Group. (19841 Otitis media with effusion during the first three years of life and development of speech and language. Pediatrics 74, 282-287. 30 Teele, W., Klein, J., Rosner, B. and The Greater Boston Otitis Media Study Group. (1989) Epidemiology of otitis media during the first seven years of life in children in Greater Boston: a prospective cohort study. J. Infect. Dis. 160, 83-94. 31 Teele, W., Klein, J., Rosner, B. and The Greater Boston Otitis Media Study Group. (19901 Otitis media in infancy and intellectual achievement, speech, and language at age 7 years. J. Infect. Dis. 162, 685-694. 32 Winitz, H. (1961) Repetitions in the vocalizations and speech of children in the first two years of life. J. Speech Hear. Dis. (Monogr.) 7, 55-62.