- Email: [email protected]
Infant vocalizations and the early diagnosis of severe hearing impairment
Infant vocalizations and the early diagnosis of severe hearing impairment Rebecca
E. Eilers, PhD, a n d D. K i m b r o u g h Oiler, PhD
From the Mailman Center for Child Development and the University of Miami Ear Institute, Departments of Psychology, Pediatrics, and Otolaryngology, University of Miami, Miami, Florida
To determine whether late onset of canonical babbling could be used as a criterion to determine risk of hearing impairment, we obtained vocalization samples longitudinally from 94 infants with normal hearing and 37 infants with severe to profound hearing impairment. Parents were instructed to report the onset of canonical babbling (the production of well-formed syllables such as "da," "na," "bee," "yaya"). Verification that the infants were producing canonical syllables was collected in laboratory audio recordings. Infants with normal hearing produced canonical vocalizations before 11 months of a g e (range, 3 to t0 months; mode, 7 months); infants who were deaf failed to produce canonical syllables until tt months of a g e or older, often well into the third year of life (range, tt to 49 months; mode, 24 months). The correlation between a g e at anset of the canonical stage and a g e at auditory amplification was 0.68, indicating that early identification and fitting of hearing aids is of significant benefit to infants learning language. The fact that there is no overlap in the distribution of the onset of canonical babbling between infants with normal hearing and infants with hearing impairment means that the failure of otherwise healthy infants to produce canonical syllables before 1-1 months of a g e should be considered a serious risk factor for hearing impairment and, when observed, should result in immediate referral for audiologic evaluation. (J PEDIATR1994;124:199203)
Undiagnosed hearing loss in infancy is not uncommon; the average age of diagnosis in the United States approaches 3 years. 1 Referral by pediatricians for audiologic evaluation sometimes occurs only after repeated requests by parents, or after recognition of late onset of meaningful speech during the second or third year of life. Consequently, infants often are not provided with hearing aids or other habilitatire devices during a key period in which language is naturally acquired. It is possible, however, to identify a small Supported by grants from the National Institutes of Health/National Institute of Child Health and Development (No. 5 R01 DC00484; Dr. Oller) and from the U.S. Department of Education/ National Institute on Disability and Rehabilitation Research (No. H133 G10118; Dr. Eilers). Submitted for publication June 22, 1993; accepted Aug. 30, 1993. Reprint requests: Rebecca E. Eilers, PhD, Mailman Center for Child Development,PO Box 016820, Miami, FL 33101. Copyright ® 1994 by Mosby-Year Book, Inc. 0022-3476/94/$3.00 + 0 9/20/51165
group of infants at very high risk for severe hearing impairment during the first year of life. Infants with severe to profound hearing losses fail to produce canonical (well-formed) babbling before I 1 months of age, whereas infants with normal hearing almost always babble canonically before 11 months of age. 2 The absence of this easily recognized vocal behavior can serve as a diagnostic marker of hearing impairment years before the failure of the appearance of normal speech. Infants with normal hearing produce vocalizations in stages of development that have been described by Stark 3 and Oller. 4, 5 These stages reveal step-by-step progress in the vocal capacities required for the production of speechlike syllables. Four clear stages can be discerned, the last of which, the canonical stage, marks the onset of mature (well-formed) syllables that appear commonly between 5 and 8 months of age. Once the infant reaches the canonical stage, he or she typically produces strings of syllables that are so speechlike that parents often assert that the infant is
199
200
Eilers and Oiler
talking. These syllables are not used referentially (i.e., in association with word meaning) or instrumentally during the middle of the first year of life. However, the syllables do meet the phonetic requirements of well-formed speech sounds in any of the world's spoken languages. Early claims that auditory experience has little effect on early vocal behavior of infants 6 have been replaced by evidence that auditory experience appears to be critical to the development of mature syllables. Infants who are deaf have significant delay in the onset of canonical stage vocalizations.2, 7-10 Furthermore, there appear to be consistent differences in precanonical vocalizations of deaf infants and hearing infants, ll as well as different sound production preferences once the canonical stage is reached. 1°, 12 The differences observed between the vocalizations of hearing infants and deaf infants are not subtle; rather, they are salient and striking. The onset of the canonical stage virtually always occurs by 10 months of age in normally developing hearing infants; no deaf infant has yet been found to begin the canonical stage before 11 months, and often the onset is much later. 2 Other infants who have near-normal audition but pervasive developmental delay (e.g., infants with Down syndrome) have a slight delay in the onset of the canonical stage but overlap considerably with normally developing infants in age at onset. 13 Healthy infants born prematurely, as well as infants in families of low socioeconomic status, have no delay in onset of the canonical stage. 14 Despite the striking differences between normally developing and hearing-impaired infants, few deaf infants have been studied extensively and relatively small numbers of hearing infants have been evaluated longitudinally for comparison purposes. In our earlier work we reported a longitudinal evaluation of 21 hearing infants and 7 deaf infants who were followed during the first 2 years of life, providing enough detail to determine accurately the age at onset of the canonical stage. 2 Since that time we have enlarged our sample considerably to include additional deaf infants and hearing infants, many of whom were born prematurely and others of whom were from families of low socioeconomic status. We hypothesized that the onset of canonical babbling is robust, appearing reliably before 11 months of age in hearing infants, regardless of gestational age at birth or socioeconomic status. The specific purpose of the study was to determine whether late onset of canonical babbling could be used as a criterion to determine risk status for hearing impairment during routine well-baby visits during the first year of life. METHODS
Subjects Normally hearing subjects. Seventy-three infants were followed longitudinally from the age of 2 months. Infants
The Journal of Pediatrics February 1994 were recruited by mail solicitation to participate in a study of vocal development. Only infants with unremarkable perinatal and postnatal histories were studied. Twenty-one infants were preterm. Selection criteria for the preterm infants included birth weight between 1400 and 2100 gm and absence of either congenital defects or prenatal or postnatal complications. Infants were excluded if they had evidence of intraventricular hemorrhage, hypoxia, respiratory distress syndrome, severe hyperbilirubinemia, or other medical complications requiring significant medical intervention, or if their 5-minute Apgar score was less than 7. All ages for preterm infants were corrected for duration of gestation. Gestational age estimates were based on postnatal examinations15, 16 and yielded a mean age at birth of 33 weeks (range, 28.5 to 35 weeks). The mean birth weight of the preterm infants was 1820 gm (range, 1446 to 2070 gm). Subjects with hearing impairments. Thirty infants with varying levels of hearing impairment ranging from profound (> 100 dB better-ear pure-tone average) to severe (80 to 90 dB better-ear pure-tone average) were recruited from an infant-parent educational program for the hearing impaired. Of the 30 infants, 27 had been identified at birth through a screening program for infant hearing 17 (Infant Hearing Impairment Program, Children's Medical Service, State of Florida, administered through the University of Miami and Jackson Memorial Hospital). For the youngest children, hearing loss estimates were based on auditory brain-stem response testing alone, with absence of a response at 95 dB nHL (normal hearing level) assumed to indicate a loss in the profound range. Additional subjects. In addition to the subjects previously described, seven infants with hearing impairments and 21 normally developing infants from our previous study 2 were included in the analyses. Data from these infants were collected in a fashion similar to data from the new group of infants. Because of the rarity of detailed records of deaf and hearing infants' vocalizations, it seemed important to describe the results for as many infants as possible. (A table listing the cause, age at identification of hearing impairment, age at fitting with hearing aid, and level of hearing loss of both groups of hearing-impaired subjects is available from the authors.) Sampling and recording procedures. The term infants with normal hearing were seen every other week, and the preterm infants were seen monthly by project staff members. In addition, each infant had laboratory evaluation on 5 consecutive days after the parent-designated onset of the canonical stage to ensure accuracy of age at onset assigned to each infant. 3, 18, 19 Parents had been taught to recognize canonical babbling and had been instructed to call the project staff as soon as it was noted. Canonical syllables were described as instances of infants' "talking" but with-
The Journal of Pediatrics Volume 124, Number 2
Eilers and Oller
® • 0 @ • O0O O0O 0O0 0@0 000 000
201
I! Legend -'/ • Full term hearing infant
000 000 000 000 000 @000
Preterm hearing infant Deaf infant Deaf nfant not yet canon cal)
00000 00000
@0000 00000
00000 00000 0000000 0000000
o e e e e e O e l l I I I | I 2 4 6 8 10
• • • • [] • • [] II I I i I I I I I I I D I I I I I I 14 16 18 20 22 24 26
[]
[]
I 12
I 28
II II I 30
• I I I I I I 32 34
I 36
i 38
•
I 40
i 42
i I 44 46
•
• I I 48 50
Months of Age Fig. t. Onset of canonical babbling in severely to profoundly hearing-impaired infants and in hearing infants. Data on preterm infants are shown at corrected ages.
out meaning, using well-formed syllables or sequences of syllables. Examples of "dada," "meme," "baba," "ma," and "ya" were given. After notification by a parent or identification of canonical babbling during a monthly visit, the infant's vocalizations were recorded for stage verification. Each recording session was 20 to 30 minutes in duration and consisted of the infant's interacting with project staff and accompanying family members in a sound-treated room. An attempt was made to obtain a minimum of 50 utterances per recording. If the 50-utterance minimum was not reached, or if the parent reported that the vocalizations were atypical of the infant, the session was rescheduled. Infants were designated to be in the canonical stage if, and only if, repeated occurrences of canonical syllables were observed during 5 consecutive days after onset. In all but two cases, parental reports were verified in the laboratory evaluations. In two cases, canonical babbling was documented by project staff at a later time than at the initial report. Parents proved to be reliable and timely informants concerning their infants' use of canonical syllables. The vocalizations of infants with hearing impairments were also recorded monthly by the same procedures as previously outlined. In addition, most of the infants with hearing impairments were enrolled in either a parent-infant educational program for the hearing impaired or a full-day educational program. In either case, the infants were seen either weekly or daily by a speech therapist. Vocalization samples were collected in the context of these programs, and the speech therapists reported the onset of canonical sylla-
bles as soon as they occurred. The specification of the onset of the canonical stage was more difficult for the infants with hearing impairments than for the infants with normal hearing. Five of the infants with hearing impairments had not reached the canonical stage by the time of this report, although they were older than 12 months of age. Some infants with hearing impairments had an erratic onset of the canonical stage, producing canonical syllables on some days but failing to produce them in subsequent samples, or producing them with lower frequency than infants with normal hearing in the canonical stage. Nevertheless, for the purposes of this study, once canonical syllables occurred, even erratically, infants with hearing impairments were said to be in the canonical stage. Thus the criterion for stage assignment was more stringent for the infants with normal hearing than for the infants with hearing impairment. RESULTS Figure 1 illustrates that each of the 94 preterm or term infants with normal hearing began canonical babbling before 11 months of age. The onset of canonical babbling for infants with normal hearing ranged from 3 to 10 months (mode, 7 months). In contrast, none of the 37 infants with hearing impairment had an age at onset of the canonical stage before 11 months of age. The onset of canonical babbling for infants with hearing impairment ranged from 11 to 49 months (mode, 24 months). Thus there was no overlap in the distribution of onset of canonical babbling for deaf infants and hearing infants, despite the fact that the crite-
202
Eilers and Oller
The Journal of Pediatrics February 1994
50
o) r-
_Q _0 O O 0 cO 6-
40-
A • 30
•
o C)
20-
[]
c-
f~•
o
10 ~ O)
<
0 0
10
Legend Profound HearingLossI Severeto Profound / Severe )
20 Age Aidedin Months
30
40
Fi9. 2. Relation between age at onset of canonical babbling and age at auditory amplification in hearing-impaired infants and young children. rion for assignment was more lax for infants with hearing impairment and that these infants were fitted with hearing aids and received consistent, high-quality early intervention. Figure 2 illustrates the relation between the age at onset of canonical babbling and the age at amplification in 28 of the 37 infants with hearing impairment in the study. Nine infants were not included in this analysis, five because they have not yet reached the canonical stage, three (including one with meningitis-induced deafness) because the onset of the canonical stage preceded amplification, and one because of late-onset deafness caused by meningitis. The illustration shows that there was a systematic relation between age at onset and age at amplification. A statistical analysis of the relationship yielded a Pearson product moment correlation of 0.68 (p <0.0001). DISCUSSION The key results of this study are as follows: (1) infants with severe to profound hearing impairment are significantly delayed in the onset of canonical babbling, a precursor to the development of meaningful speech; (2) there is no overlap in onset distribution of canonical babbling between hearing-impaired and normally hearing infants, and the failure of canonical babbling to start may be taken as a marker of extreme risk of severe to profound hearing impairment; (3) parents are good informants regarding their infants' ability to produce canonical syllables; and (4) the attainment of the canonical stage in infants with hear-
ing impairment is correlated with age at amplification and intervention, suggesting that early identification may be critical to successful language acquisition. Although universal newborn hearing screening has been recommended by the National Institutes of Health Consensus Conference, 1 hearing screening is still limited, and typically only high-risk infants are screened. Fully half of all infants with hearing impairments do not have neonatal or historical risk factors at the time of screening and are thus missed.1 The results of this study suggest that infant vocalizations may play an important role in the assessment of infant hearing during the first year of life, because a large number of infants receive no hearing screening, and the pediatrician may use the lack of canonical vocalizations as an indicator to refer infants with suspected severe to profound hearing loss by the time they reach 11 months of age. Infants who do not, by parental report, produce canonical syllables by 11 months should be referred for audiologic evaluation. Such referral may result in identification of hearing impairment nearly 2 years earlier than the average age at diagnosis in the United States. 1 Pediatricians may delay referrals for audiologic assessment even when given parental reports of an infant's failure to develop early language (first words and primitive sentences). The pediatrician's hesitation is often based on knowledge of the large individual differences in language development in normally developing infants and young children. Language delays may, in fact, be based on a wide variety of factors in addition to hearing loss. To our knowl-
The Journal of Pediatrics Volume 124, Number 2
edge, delays in the onset of canonical babbling beyond 11 months of age, however, rarely occur in otherwise normally developing infants and are of sufficient concern to warrant immediate audiologic evaluation by means of conditioned behavioral techniques such as visual reinforcement audiometry20, 21 and, if necessary, auditory brain-stem response audiometry.20. 21 Prompt identification by the first birthday will allow the child to benefit from early amplification and intervention, and will maximize the child's potential for oral language development by aiding the child during the period in which language is normally learned. Although early infant vocal behavior often provides a window from which to observe severe to profound hearing loss, it is not yet clear how less severe losses might affect canonical babbling. It is probable that less severe losses are less disruptive of normal vocal development; thus the infant with moderate hearing impairment may not appear to have delayed onset of babbling and yet may have significant delays in speech and language development. 1 Similarly, infants with progressive, slow-onset losses (such as those often associated with cytomegalovirus infection) or late-occurring losses (such as those associated with meningitis) may not be identified if only canonical babbling is used to assess risk. Consequently, it is important for all infants to be screened for hearing loss early in the second year of l i f e - - o r sooner if risk factors associated with hearing loss have been identified. We thank T. J. Balkany, MD, and E. Bancalari, MD, for their helpful comments, and Lynn W. Miskiel and Barbara J. Lewis for their help in assessing the infants with hearing impairment. REFERENCES
1. National Institutes of Health Consensus Development Conference. Early identification of hearing impairment in infants and young children. Bethesda, Maryland: National Institutes of Health, 1993. 2. Oiler DK, Eilers RE. The role of audition in infant babbling. Child Dev 1988;59:441-9. 3. StarkRE. Stages °f speech devel°pment in the first year °flife" In: Yeni-Komshian G, Kavanagh J, Ferguson C, eds. Child phonology; vol 1. New York: Academic Press, 1980173-90. 4. Oiler DK. Infant vocalization and the development of speech. Allied Health and Behavioral Sciences 1978;1:523-49. 5. Oiler DK. The emergence of the sounds of speech in infancy. In: Yeni-Komshian G, Kavanagh J, Ferguson C, eds. Child phonology; vol 1. New York: Academic Press, 1980:93-112.
Eilers and Oiler
203
6. Lenneberg EH, Rebelsky GF, Nichols IA. The vocalizations of infants born to deaf and hearing parents. Hum Dev 1965;8:2337. 7. Kent RD, Osberger M J, Netsell R, Hustedde CG. Phonetic development in identical twins who differ in auditory function. J Speech Hear Dis 1987;52:64-75. 8. Oiler DK, Eilers RE, Bull DH, Carney AE. Prespeech vocalizations of a deaf infant: a comparison with normal metaphonological development. J Speech Hearing Res 1985;28:47-63. 9. Stoel-Gammon C. Prelinguistic vocalizations of hearing-impaired and normally hearing subjects: a comparison of consonantal inventories. Journal of Speech and Hearing Disorders 1988;53:302-15. 10. Stoel-GaInmon C, Otomo K. Babbling development of hearing-impaired and normally hearing subjects. Journal of Speech and Hearing Disorders 1986;51:33-41. 11. Oller DK, Bull DH. Vocalizations of deaf infants. Presented at the International Conference on Infant Studies, New York, 1984. 12. Stark RE. Some features of the vocalizations of young deaf infants. In: Bosma JF, ed. The mouth of the infant. Springfield, Illinois: CharLes C Thomas, 1972:431-41. 13. Steffens ML, Oiler DK, Lynch MP, Urbano RC. Vocal development in infants with Down syndrome and infants who are developing normally. Am J Ment Retard 1992;97:235-46. 14. Eilers RE, Oller DK, Levine S, Basinger D, Lynch MP, Urbano R. The role of prematurity and socioeconomic status in the onset of canonical babbling in infants. Infant Behavior and Development 1993;16:297-315. 15. Dubowitz LM, Dubowitz V, Goldberg C. Clinical assessment of gestational age in the newborn infant. J PEDIATR 1970; 77:1-10. 16. Ballard JL, Novak KK, Driver M. A simplified score for assessment of fetal maturation of newly born infants. J PEDIATR 1979;95:769-74. 17. Filer R, Gonzales L. Children's Medical Service Report: University of Miami Infant Hearing Impairment Screening Program, July 1991. 18. Zlatin M. Explorative mapping of the vocal tract and primitive syllabification in infancy: the first six months. Presented at the American Speech and Hearing Association Convention, Washington, D.C., 1975. 19. Smith BL, Oiler DK. A comparative study of premeaningful vocalizations produced by normal and Down's syndrome infants. Journal of Speech and Hearing Disorders 1981 ;46:4652. 20. Eilers RE, Ozdamar (5, Steffens ML. Classification of audiograms by sequential testing: reliab!lity and validity of an automated behavioral hearing screening algorithm. J Am Acad Audiol 1993;4:172-81. 21. Wilson WR, Moore JM, Thompson G. Sound-field auditory thresholds of infants utilizing visual reinforcement audiometry (VRA). Presented at the American Speech and Hearing Association Convention, Houston, Texas, 1976.
E. Eilers, PhD, a n d D. K i m b r o u g h Oiler, PhD
From the Mailman Center for Child Development and the University of Miami Ear Institute, Departments of Psychology, Pediatrics, and Otolaryngology, University of Miami, Miami, Florida
To determine whether late onset of canonical babbling could be used as a criterion to determine risk of hearing impairment, we obtained vocalization samples longitudinally from 94 infants with normal hearing and 37 infants with severe to profound hearing impairment. Parents were instructed to report the onset of canonical babbling (the production of well-formed syllables such as "da," "na," "bee," "yaya"). Verification that the infants were producing canonical syllables was collected in laboratory audio recordings. Infants with normal hearing produced canonical vocalizations before 11 months of a g e (range, 3 to t0 months; mode, 7 months); infants who were deaf failed to produce canonical syllables until tt months of a g e or older, often well into the third year of life (range, tt to 49 months; mode, 24 months). The correlation between a g e at anset of the canonical stage and a g e at auditory amplification was 0.68, indicating that early identification and fitting of hearing aids is of significant benefit to infants learning language. The fact that there is no overlap in the distribution of the onset of canonical babbling between infants with normal hearing and infants with hearing impairment means that the failure of otherwise healthy infants to produce canonical syllables before 1-1 months of a g e should be considered a serious risk factor for hearing impairment and, when observed, should result in immediate referral for audiologic evaluation. (J PEDIATR1994;124:199203)
Undiagnosed hearing loss in infancy is not uncommon; the average age of diagnosis in the United States approaches 3 years. 1 Referral by pediatricians for audiologic evaluation sometimes occurs only after repeated requests by parents, or after recognition of late onset of meaningful speech during the second or third year of life. Consequently, infants often are not provided with hearing aids or other habilitatire devices during a key period in which language is naturally acquired. It is possible, however, to identify a small Supported by grants from the National Institutes of Health/National Institute of Child Health and Development (No. 5 R01 DC00484; Dr. Oller) and from the U.S. Department of Education/ National Institute on Disability and Rehabilitation Research (No. H133 G10118; Dr. Eilers). Submitted for publication June 22, 1993; accepted Aug. 30, 1993. Reprint requests: Rebecca E. Eilers, PhD, Mailman Center for Child Development,PO Box 016820, Miami, FL 33101. Copyright ® 1994 by Mosby-Year Book, Inc. 0022-3476/94/$3.00 + 0 9/20/51165
group of infants at very high risk for severe hearing impairment during the first year of life. Infants with severe to profound hearing losses fail to produce canonical (well-formed) babbling before I 1 months of age, whereas infants with normal hearing almost always babble canonically before 11 months of age. 2 The absence of this easily recognized vocal behavior can serve as a diagnostic marker of hearing impairment years before the failure of the appearance of normal speech. Infants with normal hearing produce vocalizations in stages of development that have been described by Stark 3 and Oller. 4, 5 These stages reveal step-by-step progress in the vocal capacities required for the production of speechlike syllables. Four clear stages can be discerned, the last of which, the canonical stage, marks the onset of mature (well-formed) syllables that appear commonly between 5 and 8 months of age. Once the infant reaches the canonical stage, he or she typically produces strings of syllables that are so speechlike that parents often assert that the infant is
199
200
Eilers and Oiler
talking. These syllables are not used referentially (i.e., in association with word meaning) or instrumentally during the middle of the first year of life. However, the syllables do meet the phonetic requirements of well-formed speech sounds in any of the world's spoken languages. Early claims that auditory experience has little effect on early vocal behavior of infants 6 have been replaced by evidence that auditory experience appears to be critical to the development of mature syllables. Infants who are deaf have significant delay in the onset of canonical stage vocalizations.2, 7-10 Furthermore, there appear to be consistent differences in precanonical vocalizations of deaf infants and hearing infants, ll as well as different sound production preferences once the canonical stage is reached. 1°, 12 The differences observed between the vocalizations of hearing infants and deaf infants are not subtle; rather, they are salient and striking. The onset of the canonical stage virtually always occurs by 10 months of age in normally developing hearing infants; no deaf infant has yet been found to begin the canonical stage before 11 months, and often the onset is much later. 2 Other infants who have near-normal audition but pervasive developmental delay (e.g., infants with Down syndrome) have a slight delay in the onset of the canonical stage but overlap considerably with normally developing infants in age at onset. 13 Healthy infants born prematurely, as well as infants in families of low socioeconomic status, have no delay in onset of the canonical stage. 14 Despite the striking differences between normally developing and hearing-impaired infants, few deaf infants have been studied extensively and relatively small numbers of hearing infants have been evaluated longitudinally for comparison purposes. In our earlier work we reported a longitudinal evaluation of 21 hearing infants and 7 deaf infants who were followed during the first 2 years of life, providing enough detail to determine accurately the age at onset of the canonical stage. 2 Since that time we have enlarged our sample considerably to include additional deaf infants and hearing infants, many of whom were born prematurely and others of whom were from families of low socioeconomic status. We hypothesized that the onset of canonical babbling is robust, appearing reliably before 11 months of age in hearing infants, regardless of gestational age at birth or socioeconomic status. The specific purpose of the study was to determine whether late onset of canonical babbling could be used as a criterion to determine risk status for hearing impairment during routine well-baby visits during the first year of life. METHODS
Subjects Normally hearing subjects. Seventy-three infants were followed longitudinally from the age of 2 months. Infants
The Journal of Pediatrics February 1994 were recruited by mail solicitation to participate in a study of vocal development. Only infants with unremarkable perinatal and postnatal histories were studied. Twenty-one infants were preterm. Selection criteria for the preterm infants included birth weight between 1400 and 2100 gm and absence of either congenital defects or prenatal or postnatal complications. Infants were excluded if they had evidence of intraventricular hemorrhage, hypoxia, respiratory distress syndrome, severe hyperbilirubinemia, or other medical complications requiring significant medical intervention, or if their 5-minute Apgar score was less than 7. All ages for preterm infants were corrected for duration of gestation. Gestational age estimates were based on postnatal examinations15, 16 and yielded a mean age at birth of 33 weeks (range, 28.5 to 35 weeks). The mean birth weight of the preterm infants was 1820 gm (range, 1446 to 2070 gm). Subjects with hearing impairments. Thirty infants with varying levels of hearing impairment ranging from profound (> 100 dB better-ear pure-tone average) to severe (80 to 90 dB better-ear pure-tone average) were recruited from an infant-parent educational program for the hearing impaired. Of the 30 infants, 27 had been identified at birth through a screening program for infant hearing 17 (Infant Hearing Impairment Program, Children's Medical Service, State of Florida, administered through the University of Miami and Jackson Memorial Hospital). For the youngest children, hearing loss estimates were based on auditory brain-stem response testing alone, with absence of a response at 95 dB nHL (normal hearing level) assumed to indicate a loss in the profound range. Additional subjects. In addition to the subjects previously described, seven infants with hearing impairments and 21 normally developing infants from our previous study 2 were included in the analyses. Data from these infants were collected in a fashion similar to data from the new group of infants. Because of the rarity of detailed records of deaf and hearing infants' vocalizations, it seemed important to describe the results for as many infants as possible. (A table listing the cause, age at identification of hearing impairment, age at fitting with hearing aid, and level of hearing loss of both groups of hearing-impaired subjects is available from the authors.) Sampling and recording procedures. The term infants with normal hearing were seen every other week, and the preterm infants were seen monthly by project staff members. In addition, each infant had laboratory evaluation on 5 consecutive days after the parent-designated onset of the canonical stage to ensure accuracy of age at onset assigned to each infant. 3, 18, 19 Parents had been taught to recognize canonical babbling and had been instructed to call the project staff as soon as it was noted. Canonical syllables were described as instances of infants' "talking" but with-
The Journal of Pediatrics Volume 124, Number 2
Eilers and Oller
® • 0 @ • O0O O0O 0O0 0@0 000 000
201
I! Legend -'/ • Full term hearing infant
000 000 000 000 000 @000
Preterm hearing infant Deaf infant Deaf nfant not yet canon cal)
00000 00000
@0000 00000
00000 00000 0000000 0000000
o e e e e e O e l l I I I | I 2 4 6 8 10
• • • • [] • • [] II I I i I I I I I I I D I I I I I I 14 16 18 20 22 24 26
[]
[]
I 12
I 28
II II I 30
• I I I I I I 32 34
I 36
i 38
•
I 40
i 42
i I 44 46
•
• I I 48 50
Months of Age Fig. t. Onset of canonical babbling in severely to profoundly hearing-impaired infants and in hearing infants. Data on preterm infants are shown at corrected ages.
out meaning, using well-formed syllables or sequences of syllables. Examples of "dada," "meme," "baba," "ma," and "ya" were given. After notification by a parent or identification of canonical babbling during a monthly visit, the infant's vocalizations were recorded for stage verification. Each recording session was 20 to 30 minutes in duration and consisted of the infant's interacting with project staff and accompanying family members in a sound-treated room. An attempt was made to obtain a minimum of 50 utterances per recording. If the 50-utterance minimum was not reached, or if the parent reported that the vocalizations were atypical of the infant, the session was rescheduled. Infants were designated to be in the canonical stage if, and only if, repeated occurrences of canonical syllables were observed during 5 consecutive days after onset. In all but two cases, parental reports were verified in the laboratory evaluations. In two cases, canonical babbling was documented by project staff at a later time than at the initial report. Parents proved to be reliable and timely informants concerning their infants' use of canonical syllables. The vocalizations of infants with hearing impairments were also recorded monthly by the same procedures as previously outlined. In addition, most of the infants with hearing impairments were enrolled in either a parent-infant educational program for the hearing impaired or a full-day educational program. In either case, the infants were seen either weekly or daily by a speech therapist. Vocalization samples were collected in the context of these programs, and the speech therapists reported the onset of canonical sylla-
bles as soon as they occurred. The specification of the onset of the canonical stage was more difficult for the infants with hearing impairments than for the infants with normal hearing. Five of the infants with hearing impairments had not reached the canonical stage by the time of this report, although they were older than 12 months of age. Some infants with hearing impairments had an erratic onset of the canonical stage, producing canonical syllables on some days but failing to produce them in subsequent samples, or producing them with lower frequency than infants with normal hearing in the canonical stage. Nevertheless, for the purposes of this study, once canonical syllables occurred, even erratically, infants with hearing impairments were said to be in the canonical stage. Thus the criterion for stage assignment was more stringent for the infants with normal hearing than for the infants with hearing impairment. RESULTS Figure 1 illustrates that each of the 94 preterm or term infants with normal hearing began canonical babbling before 11 months of age. The onset of canonical babbling for infants with normal hearing ranged from 3 to 10 months (mode, 7 months). In contrast, none of the 37 infants with hearing impairment had an age at onset of the canonical stage before 11 months of age. The onset of canonical babbling for infants with hearing impairment ranged from 11 to 49 months (mode, 24 months). Thus there was no overlap in the distribution of onset of canonical babbling for deaf infants and hearing infants, despite the fact that the crite-
202
Eilers and Oller
The Journal of Pediatrics February 1994
50
o) r-
_Q _0 O O 0 cO 6-
40-
A • 30
•
o C)
20-
[]
c-
f~•
o
10 ~ O)
<
0 0
10
Legend Profound HearingLossI Severeto Profound / Severe )
20 Age Aidedin Months
30
40
Fi9. 2. Relation between age at onset of canonical babbling and age at auditory amplification in hearing-impaired infants and young children. rion for assignment was more lax for infants with hearing impairment and that these infants were fitted with hearing aids and received consistent, high-quality early intervention. Figure 2 illustrates the relation between the age at onset of canonical babbling and the age at amplification in 28 of the 37 infants with hearing impairment in the study. Nine infants were not included in this analysis, five because they have not yet reached the canonical stage, three (including one with meningitis-induced deafness) because the onset of the canonical stage preceded amplification, and one because of late-onset deafness caused by meningitis. The illustration shows that there was a systematic relation between age at onset and age at amplification. A statistical analysis of the relationship yielded a Pearson product moment correlation of 0.68 (p <0.0001). DISCUSSION The key results of this study are as follows: (1) infants with severe to profound hearing impairment are significantly delayed in the onset of canonical babbling, a precursor to the development of meaningful speech; (2) there is no overlap in onset distribution of canonical babbling between hearing-impaired and normally hearing infants, and the failure of canonical babbling to start may be taken as a marker of extreme risk of severe to profound hearing impairment; (3) parents are good informants regarding their infants' ability to produce canonical syllables; and (4) the attainment of the canonical stage in infants with hear-
ing impairment is correlated with age at amplification and intervention, suggesting that early identification may be critical to successful language acquisition. Although universal newborn hearing screening has been recommended by the National Institutes of Health Consensus Conference, 1 hearing screening is still limited, and typically only high-risk infants are screened. Fully half of all infants with hearing impairments do not have neonatal or historical risk factors at the time of screening and are thus missed.1 The results of this study suggest that infant vocalizations may play an important role in the assessment of infant hearing during the first year of life, because a large number of infants receive no hearing screening, and the pediatrician may use the lack of canonical vocalizations as an indicator to refer infants with suspected severe to profound hearing loss by the time they reach 11 months of age. Infants who do not, by parental report, produce canonical syllables by 11 months should be referred for audiologic evaluation. Such referral may result in identification of hearing impairment nearly 2 years earlier than the average age at diagnosis in the United States. 1 Pediatricians may delay referrals for audiologic assessment even when given parental reports of an infant's failure to develop early language (first words and primitive sentences). The pediatrician's hesitation is often based on knowledge of the large individual differences in language development in normally developing infants and young children. Language delays may, in fact, be based on a wide variety of factors in addition to hearing loss. To our knowl-
The Journal of Pediatrics Volume 124, Number 2
edge, delays in the onset of canonical babbling beyond 11 months of age, however, rarely occur in otherwise normally developing infants and are of sufficient concern to warrant immediate audiologic evaluation by means of conditioned behavioral techniques such as visual reinforcement audiometry20, 21 and, if necessary, auditory brain-stem response audiometry.20. 21 Prompt identification by the first birthday will allow the child to benefit from early amplification and intervention, and will maximize the child's potential for oral language development by aiding the child during the period in which language is normally learned. Although early infant vocal behavior often provides a window from which to observe severe to profound hearing loss, it is not yet clear how less severe losses might affect canonical babbling. It is probable that less severe losses are less disruptive of normal vocal development; thus the infant with moderate hearing impairment may not appear to have delayed onset of babbling and yet may have significant delays in speech and language development. 1 Similarly, infants with progressive, slow-onset losses (such as those often associated with cytomegalovirus infection) or late-occurring losses (such as those associated with meningitis) may not be identified if only canonical babbling is used to assess risk. Consequently, it is important for all infants to be screened for hearing loss early in the second year of l i f e - - o r sooner if risk factors associated with hearing loss have been identified. We thank T. J. Balkany, MD, and E. Bancalari, MD, for their helpful comments, and Lynn W. Miskiel and Barbara J. Lewis for their help in assessing the infants with hearing impairment. REFERENCES
1. National Institutes of Health Consensus Development Conference. Early identification of hearing impairment in infants and young children. Bethesda, Maryland: National Institutes of Health, 1993. 2. Oiler DK, Eilers RE. The role of audition in infant babbling. Child Dev 1988;59:441-9. 3. StarkRE. Stages °f speech devel°pment in the first year °flife" In: Yeni-Komshian G, Kavanagh J, Ferguson C, eds. Child phonology; vol 1. New York: Academic Press, 1980173-90. 4. Oiler DK. Infant vocalization and the development of speech. Allied Health and Behavioral Sciences 1978;1:523-49. 5. Oiler DK. The emergence of the sounds of speech in infancy. In: Yeni-Komshian G, Kavanagh J, Ferguson C, eds. Child phonology; vol 1. New York: Academic Press, 1980:93-112.
Eilers and Oiler
203
6. Lenneberg EH, Rebelsky GF, Nichols IA. The vocalizations of infants born to deaf and hearing parents. Hum Dev 1965;8:2337. 7. Kent RD, Osberger M J, Netsell R, Hustedde CG. Phonetic development in identical twins who differ in auditory function. J Speech Hear Dis 1987;52:64-75. 8. Oiler DK, Eilers RE, Bull DH, Carney AE. Prespeech vocalizations of a deaf infant: a comparison with normal metaphonological development. J Speech Hearing Res 1985;28:47-63. 9. Stoel-Gammon C. Prelinguistic vocalizations of hearing-impaired and normally hearing subjects: a comparison of consonantal inventories. Journal of Speech and Hearing Disorders 1988;53:302-15. 10. Stoel-GaInmon C, Otomo K. Babbling development of hearing-impaired and normally hearing subjects. Journal of Speech and Hearing Disorders 1986;51:33-41. 11. Oller DK, Bull DH. Vocalizations of deaf infants. Presented at the International Conference on Infant Studies, New York, 1984. 12. Stark RE. Some features of the vocalizations of young deaf infants. In: Bosma JF, ed. The mouth of the infant. Springfield, Illinois: CharLes C Thomas, 1972:431-41. 13. Steffens ML, Oiler DK, Lynch MP, Urbano RC. Vocal development in infants with Down syndrome and infants who are developing normally. Am J Ment Retard 1992;97:235-46. 14. Eilers RE, Oller DK, Levine S, Basinger D, Lynch MP, Urbano R. The role of prematurity and socioeconomic status in the onset of canonical babbling in infants. Infant Behavior and Development 1993;16:297-315. 15. Dubowitz LM, Dubowitz V, Goldberg C. Clinical assessment of gestational age in the newborn infant. J PEDIATR 1970; 77:1-10. 16. Ballard JL, Novak KK, Driver M. A simplified score for assessment of fetal maturation of newly born infants. J PEDIATR 1979;95:769-74. 17. Filer R, Gonzales L. Children's Medical Service Report: University of Miami Infant Hearing Impairment Screening Program, July 1991. 18. Zlatin M. Explorative mapping of the vocal tract and primitive syllabification in infancy: the first six months. Presented at the American Speech and Hearing Association Convention, Washington, D.C., 1975. 19. Smith BL, Oiler DK. A comparative study of premeaningful vocalizations produced by normal and Down's syndrome infants. Journal of Speech and Hearing Disorders 1981 ;46:4652. 20. Eilers RE, Ozdamar (5, Steffens ML. Classification of audiograms by sequential testing: reliab!lity and validity of an automated behavioral hearing screening algorithm. J Am Acad Audiol 1993;4:172-81. 21. Wilson WR, Moore JM, Thompson G. Sound-field auditory thresholds of infants utilizing visual reinforcement audiometry (VRA). Presented at the American Speech and Hearing Association Convention, Houston, Texas, 1976.