Speech, Language, and Reading Skills After Early Cochlear Implantation

Objective  To examine whether age at cochlear implantation or duration of implant use is associated with speech, language, and reading skills exhibited at age 8 to 9 years in children who underwent implantation by age 5 years.

Design  Performance outcomes in speech perception, speech production, language, and reading were examined in terms of the age at which children first received a cochlear implant (2, 3, or 4 years), the age they received an updated (Spectra) processor, and the duration of use of an implant and an updated processor.

Setting  Data collection was conducted at summer research camps held over 4 consecutive years to maximize the number of children available at a specific age (8-9 years). Children were tested individually by experienced examiners, and their parents and therapists provided background and educational history information.

Participants  A total of 181 children from 33 different states and 5 Canadian provinces who received a cochlear implant by age 5 years were tested. A subsample of 133 children with performance IQ scores of 80 or greater and onset of deafness at birth were selected for the age-at-implantation analysis. Another subsample of 39 children with deafness acquired by age 3 years was also examined.

Outcome Measures  A battery of tests of speech perception, speech production, language, and reading was administered to each child and reduced to a single factor score for each skill.

Results  Correlation coefficients between age at implantation and duration of use did not reach significance for any of the outcome skills measured. Age at which the updated speech processor (Spectra) was fitted was significantly related to speech production outcome (earlier use of an updated processor was associated with greater speech intelligibility) but not to any other skill area. However, more of the children who underwent implantation at age 2 years (43%) achieved combined speech and language skills commensurate with their age-matched peers with normal hearing than did children who underwent implantation at age 4 years (16%). Furthermore, normal speech and language skills were documented in 80% of children who lost hearing after birth and who underwent implantation within a year of onset of deafness.

Conclusions  For children who receive a cochlear implant between the ages of 2 and 4 years, early cochlear implantation does not ensure better speech perception, speech production, language, or reading skills. However, greater speech and language proficiency may be expected from children who exhibit normal hearing for even a brief period after birth and receive a cochlear implant shortly after losing their hearing. Further research examining the benefits of cochlear implantation before age 2 years will help families and clinicians better understand the time-sensitive nature of the decision to conduct cochlear implant surgery.

The advent of cochlear implants has had a dramatic effect on the achievements of young, profoundly deaf children. Spoken language competence is now possible for many children who previously depended primarily on sign language for communication. Children who receive an implant early in life, followed by a period of appropriate rehabilitation, achieve speech and language skills that exceed levels observed in profoundly deaf children with hearing aids. However, there continue to be large differences in the performance outcomes of individual children, and many do not achieve speech and language skills that are commensurate with their age-matched peers with normal hearing. There are at least 2 reasons for the observed lags in development. First, the auditory information some children receive from the implant may be insufficient for normal speech development. Second, the period of profound deafness before the child receives an implant may make speech so inaccessible that a critical period for spoken language development is lost.

There is ample evidence for the critical period hypothesis. Prelingually deaf children who undergo implantation at younger ages reportedly achieve greater speech perception skills than those who undergo implantation at a later age.¹ The critical age for receiving a cochlear implant has been variously reported at 6 years,² 5 years,^3,4 and 3 years.^5,6 These findings suggest that receiving auditory stimulation before age 6 years may be critical for auditory and speech development. Such findings have prompted a reduction in earliest age of approved cochlear implantation by the US Food and Drug Administration from 24 months in 1990 to 18 months in 1998 to 12 months in 2000. However, none of these studies have examined whether age at implantation plays a role in the performance of children who receive an implant only during the critical infant and preschool years. Kirk and colleagues⁷ found a tendency for those who undergo implantation at age 2 years to close the language gap with their peers with normal hearing at a faster rate than those who undergo implantation at age 3 years, but concluded that early implantation may not be crucial for the development of spoken word recognition skills.

The present investigation examines whether there is urgency to decide whether a child should receive an implant as soon as possible after diagnosis of hearing loss. The following recent events have increased the importance of this question to parents and clinicians alike: (1) Universal newborn hearing screening programs are dramatically reducing the age at diagnosis of deafness. (2) Improvements in implant technology are resulting in a more normal rate of spoken language development in early-implanted children. (3) New diagnostic procedures are improving our ability to predict the benefits of implantation in infants. On the other hand, the following obstacles remain for infants to undergo implantation at the earliest possible age: (1) Families need time to come to terms with the diagnosis and to learn about educational and lifestyle options for their child before making the decision to proceed with implant surgery. (2) Obtaining the behavioral responses needed for creating an optimum implant map is more difficult at young ages. (3) Very early implantation may be associated with some increased surgical risk.

Parents report that the primary reason for getting a cochlear implant for their child is to allow them to develop spoken language.⁸ If providing a cochlear implant early in the child's life greatly increases the probability that the child will develop speech and language skills commensurate with his or her peers with normal hearing, parents should be made aware of the importance of an early decision to achieve this outcome. However, if receiving an implant early (ie, at any time before age 5 years) produces similar results, then the sense of urgency may be somewhat relaxed.

The institutional review board at Central Institute for the Deaf, St Louis, Mo, approved the subject selection, recruitment, and testing for this study. Over a 4-year period, 181 children from 33 different states and 5 Canadian provinces came to St Louis to attend a cochlear implant summer research camp. Potential candidates for this study were selected from the patient database of the Cochlear Corporation (Englewood, Colo). Children who met the age and duration of use criteria were identified to participating cochlear implant centers, which invited families to apply for the study. All qualifying families who applied were accepted into the study. These 181 children are likely representative of the 8- to 9-year-old children across North America who received a cochlear implant under age 5 years between 1990 and 1996. The participants did not represent any single program or method, but rather came from a variety of educational settings in 32 states and 5 Canadian provinces.⁹ Most children underwent implantation at age 2, 3, or 4 years and were 8 or 9 years old when they were tested. Children had used an implant for 4 to 6 years, so presumably the effects of duration of use were minimized. If early implantation provides a lasting advantage in terms of auditory, speech, and language development, then this should be apparent in their test scores at ages 8 and 9 years. We should expect to see more of the children who underwent implantation before age 3 years performing above average for the group.

Each of the 181 children was administered the Performance Scale of the Wechsler Intelligence Scale for Children (third edition).¹⁰ Of these children, 9 achieved IQ scores below 80. Because this factor could contribute to poorer than expected postimplantation outcome, these children were eliminated from the sample when examining age-at-implantation effects, reducing the group to 172 children with measured intelligence within or above the average range.

Although most of the children were reportedly deaf from birth, almost one fourth of them had some known cause of deafness after birth. Since the postimplantation performance of these children could be affected by their period of normal hearing, however brief, they were eliminated from this analysis, bringing the total number of children to 133. Forty-four of these congenitally deaf children received a cochlear implant before age 3 years, 52 children underwent implantation at age 3 years, and 37 underwent implantation after they turned 4 years of age. Characteristics of this subject group are summarized in Table 1.

A comparison group of 24 children aged 8 to 9 years with normal hearing were tested on some of the measures that did not provide age-appropriate normative data for the cochlear implant group. These children were recruited from a local elementary school and participated in hearing, speech, and language screenings that reflected normal development.

Postimplantation outcome was assessed using a battery of tests designed to provide a comprehensive estimate of each skill. The tests included in outcome batteries (Table 2) are described in detail elsewhere (along with mean scores for the entire sample and test intercorrelations): speech perception,¹¹ speech production,¹² language,¹³ and reading.¹⁴ The relatively high correlation coefficients obtained among the measures in each area allowed reduction to a single standardized score using principal components analysis. This approach is motivated by the belief that a collection of highly correlated measures all tap the same ability and that a single summary score would be more economical than multiple scores.¹⁵ Principal components analysis forms this summary score by creating a weighted linear combination of the original variables. The factor score listed by each test in Table 2 reflects its relative weight in the linear combination. Tests are listed in the order of their contribution to the outcome score. In the analyses to follow, each outcome variable is represented by a standardized score in which 0 represents average performance for the entire group, positive values represent above-average performance, and negative values represent below-average performance.

All speech perception tests were administered at 70 dB sound pressure level using recorded stimuli. The speech production battery included an estimate of overall speech intelligibility based on recognition of key words in sentences by naive listeners from audio recordings of 36 spoken sentences developed by McGarr.¹⁶ Percentage of correct phoneme production was established for consonants and vowels by comparing a phonetic transcription of these sentences with the targeted production. Effective conversational use of speech was measured by the percentage of time spent in communication breakdown in a videotaped conversation using the Dyalog analysis procedure.¹⁷ Ratings of both speech perception and speech production skills in everyday situations were obtained from the child's parent who completed the Auditory Responsiveness Questionnaire and the Use of Speech Questionnaire.

Total language competence was assessed using a battery of measures administered in total communication to all children, regardless of whether they knew sign language. One of the measures used to measure language development was the Index of Productive Syntax (IPSyn).¹⁸ This is a rating system for quantifying the child's use of English syntax based on transcription of a spontaneous language sample. The 2 reading subtests from the Peabody Individual Achievement Test were administered: Reading Recognition and Reading Comprehension.¹⁹ The Word Attack subtest from the Woodcock Reading Mastery Test was used to determine grade equivalent scores for phonic and structural analysis skills.²⁰

The correlations obtained between the composite outcome scores for each skill and age at first receiving a cochlear implant and duration of implant use are summarized in Table 3. None of the coefficients reached statistical significance (P<.05). Multiple regression analyses reported elsewhere^9,11-14 indicated that age at implantation accounted for insignificant added variance in outcome scores even when other contributing factors (eg, performance IQ score, family socioeconomic status, sex, and implant characteristics) were held constant.²¹ Scatterplots of age at implantation with 2 different highly weighted measures included in the factor scores for speech production (McGarr intelligibility) and language syntax (IPSyn total score) are presented in Figure 1 to illustrate the low correlations obtained among these variables. Table 4 reports the percentage of children who underwent implantation at ages 2, 3, and 4 years and scored within the range of their 8- to 9-year-old peers with normal hearing on the speech measure, the language measure, and both measures. Although correlations of age at implantation with each set of outcomes did not reach significance, more children who underwent implantation at age 2 years achieved scores that were within expectation for their peers with normal hearing when both speech and language measures were considered together.

It is possible that an advantage for those who underwent implantation at the youngest ages was not apparent because of insufficient auditory input during the first months of using the device. Most of these children were first fitted with a Nucleus 22 implant (Cochlear Corporation) with the Multi-Peak speech processing (M-Peak) strategy. This type of processor has been shown to be inferior to the newer Spectra (Cochlear Corporation) processor.²² Only 22 of the children in the study received their implant at a time when Spectra was the first and only processor they used. Perhaps those who used M-Peak at the youngest ages did not receive enough speech information to facilitate language development.

The age at which the updated Spectra processor was fitted is included in the subject characteristics listed in Table 1. Nineteen children received Spectra at age 2 years, 24 at age 3 years, 30 at age 4 years, and 52 did not receive the updated processor until 5 years or older. Table 3 presents correlations between outcome factor scores and age at receiving an updated processor for those 125 children who either received the newer processor at initial implantation or were switched over at some point before the time they were tested. The 8 children who were still using the older M-Peak processor at the time of data collection were eliminated from this analysis. Only speech production showed a significant advantage for early Spectra use (r = −0.24). Children who received Spectra at younger ages had more intelligible speech at age 8 to 9 years. Duration of Spectra use was unrelated to other outcome factor scores.

Thirty-nine children in the study had been excluded from the previous analyses because they had some brief period of normal hearing before they became deaf. A separate analysis of these 39 children was conducted to examine outcomes achieved in relation to the age at which they had acquired a hearing loss. Fourteen children became deaf before they were 1 year old, 17 when they were 1 year old, and 8 when they were 2 years old. A "duration of deafness" value was obtained by subtracting a child's age at onset of deafness from his or her age at implantation. Significant correlations were observed for speech perception (r = −0.44), speech production (r = −0.36), and language (r = −0.35), indicating an advantage for shorter duration of deafness. Reading outcome was not significantly correlated with duration of deafness (r = −0.28). Almost 80% of those receiving implants within a year after the onset of deafness achieved speech and language scores within the range of their peers with normal hearing.

For children who receive a cochlear implant between ages 2 and 4 years, age at implantation is not strongly associated with speech perception, speech production, language, or reading skills exhibited at age 8 or 9 years. There are a number of possible reasons why a stronger relation between age at implantation and postimplantation outcome was not apparent, such as the following: (1) Age 2 years is not young enough to show the advantage of early input. The first 2 years of life are most important for language development in normal children, and the loss of these formative years of auditory input may not be recoverable. (2) There may be an advantage for early implantation that is no longer apparent by age 8 years. (3) The speech coding strategies available at the time of this study, even with the updated speech processor, do not present enough information for most children to achieve normal speech and language development no matter how young they were when they received the device.

Although there are not strong correlations between age at implantation and outcome measures of performance, more of the children who underwent implantation at 2 years (43%) achieved combined speech and language skills commensurate with their peers with normal hearing than did children who underwent implantation at age 4 years (16%). Higher speech intelligibility scores were obtained from children who received an upgraded speech processor at a younger age. Furthermore, normal speech and language skills were documented in 80% of children who lost hearing after birth and received an implant within a year of onset of deafness. These findings indicate that normal speech and language development is possible for many children who experience only a short period of auditory deprivation during the critical language learning years. It is likely that an even greater proportion of children who receive auditory stimulation before a language delay has been established (ie, prior to age 2 years), will exhibit normal spoken language when they reach elementary school age. Further research comparing the benefits of cochlear implantation at ages 1, 2, 3, and 4 years will help families and clinicians better understand the time-sensitive nature of the decision to conduct cochlear implant surgery.

Corresponding author and reprints: Ann E. Geers, PhD, Department of Otolaryngology–Head and Neck Surgery, Southwestern Medical School, University of Texas at Dallas, 5323 Harry Hines Blvd, Dallas, TX 75390-9035 (e-mail: ageers@earthlink.net).

Submitted for publication August 25, 2003; final revision received November 24, 2003; accepted November 25, 2003.

This study was supported by grant DC03100 from the National Institute on Deafness and Other Communication Disorders (NIDCD) of the National Institutes of Health to Central Institute for the Deaf (CID).

This study was presented at the Ninth Symposium on Cochlear Implants in Children; April 26, 2003; Washington, DC.

We gratefully acknowledge the families of children with implants from across the United States and Canada who enthusiastically participated in this study as well as the staff of Cochlear Corporation and implant centers across the United States and Canada for their role in disseminating information about the study to potential participants. Christine Brenner, MA, coordinated subject recruiting, testing, and data analysis. Emily Tobey, PhD, conducted the speech intelligibility analysis and interpretation.