Delay in maturation of the auditory pathway and its relationship to language acquisition disorders. (Original Article).

Abstract

We studied 81 children, mostly boys, who experienced language acquisition delay but whose audiometric thresholds were normal. We assessed the evolution of children with delayed maturation of auditory pathways by brainstem evoked response audiometry (BERA). We also used a questionnaire administered during diagnostic procedures to determine if there was a probable etiology in each patient. In addition, we further studied language evolution in 29 patients by means of a second questionnaire that was administered approximately 2 years later. Finally, we studied the evolution of the I-V interwave interval and the I/V amplitude ratio in 16 patients by performing a second BERA after a mean interval of 3 years. We observed improvement in both brainstem transmission time and language acquisition in all 81 patients. However, only a few patients achieved normal range results. Morphologic alterations, which were most common in patients who had had perinatal jaundice, remained unchanged. The most common possible risk facto rs for the delayed maturation pattern observed on BERA were parental consanguinity, prematurity, perinatal anoxia and jaundice, and postnatal seizure and infection. Some patients had more than one of these possible risk factors. We conclude that high-risk newborns and 2-year-old children who have no primitive verbal language skills should undergo BERA as well as investigation of hearing thresholds, interwave intervals, and I/V amplitude ratios. The alteration of these parameters points out the need for early intervention if there is no favorable prognosis.

Introduction

During our 10 years' experience with brainstem evoked response audiometry (BERA) in children, we have observed a high incidence of lengthening of the interval between waves I and V in patients with suspected audiologic or neurologic disturbances. This elongation is probably attributable to a delay in maturation of the auditory pathway in the brainstem. These patients frequently exhibit language acquisition disturbances despite having normal electrophysiologic thresholds and an absence of motor or auditory lesions; this disorder is more common among boys than girls. (1)

The primary acoustic nuclei of the brainstem (cochlear and olivar superior nuclei) are the first auditory relay stations to the more central regions, and they are responsible for comprehension of language symbols and reflex responses to sound. The ventral nucleus exhibits reflex acoustic activity, whereas the dorsal nucleus processes dynamic and complex auditory stimuli and acts as a central integrator under the inhibitory influence of the ventral nucleus. (2) Along the course between the cochlear nuclei and the olivary complex, auditory fibers functionally interact with reticular formation structures and form the nonspecific acoustic pathway. The rostral portion of the reticular formation responds to new acoustic stimuli, whereas the posterior portion acts as a cerebral modulator that synchronizes stimuli and functions.

Histologic studies of children with neonatal asphyxia have revealed the presence of brainstem lesions in approximately 90%. (3) The immature reticular formation is particularly vulnerable to asphyxia. (3) Cell loss of the cochlear nuclei, especially in the ventral nuclei, has also been observed in these patients, as have minor changes in the olive, the superior olivary complex, and the inferior colliculus. (4) However, it is believed that perinatal anoxia can seriously damage not only pure-tone auditory function but also more complex activity, such as speech interpretation.

The latency of wave I on BERA progressively diminishes between birth and 8 to 10 weeks following delivery. (5,6) Later, a second-phase auditory function maturation occurs, which involves a slower rate of decrease of later wave latencies, particularly of wave V. (5,7) It is not until children reach approximately 2 years and 6 months of age that their acoustic information processing becomes similar to that seen in adults. The younger the child is, the greater the latency of the waves is in each tested tone intensity.

Some authors believe that in patients with perinatal hypoxic encephalopathy, the edema might interfere with synaptic transmission. They feel that this interference might lead to an increase in wave V latency on BERA as a result of (1) the increase in the size of the intracellular spaces and (2) cell loss in the area of the cochlear nuclei. Upon reduction of the edema, BERA results might improve if there has been no permanent damage to neural tissue. (8,9) A few prospective studies of newborns at high risk for developing neurologic disturbances who were followed until they reached school age showed a high incidence of language and learning deficiencies as an expression of subtle neurologic sequelae. (10)

In view of the fact that patients diagnosed with delayed maturation of the auditory pathway who experience delayed language acquisition but who have normal auditory thresholds are generally considered to have a good prognosis, we believe that they usually receive less attention than they should from physicians and psychologists. Because we found no data in the literature regarding the evolution of these children, we conceived this study to analyze these cases. Our study had four primary goals:

* to determine if maturation of the auditory pathway and normalization of the I-V interval on BERA occurs after some years have passed

* to ascertain if language acquisition becomes normal after several years

* to identify any possible association between the evolution of these patients and their history and other symptoms--that is, to establish a cause-and-effect relationship

* to determine if there are any risk factors for failure of language development so that affected children might be monitored

Patients and methods

This study was conducted at Clinica OtoRhinus in Sao Paulo, where approximately 3,200 BERA examinations were performed in children with delayed language acquisition and/or a suspected hearing loss during the 7-year period from January 1992 through December 1998. To be eligible for the study, patients had to exhibit an elongation of the I-V latency interval ([greater than or equal to]4.5 msec) or a wave I amplitude that was disproportionate to the wave V amplitude. Also, we included only children whose BERA threshold was at least 60 dB HL in at least one ear. (11) BERA results were recorded on a Life-Tech 8101 AR device (Life-Tech; Stafford, Tex.); three needle electrodes were placed--one over the vertex (positive electrode), one on the earlobe ipsilateral to the acoustic stimuli (negative), and one on the contralateral earlobe (ground). All children younger than 6 years and those who were uncooperative were administered general anesthesia with halothane. Eighty-one children--69 boys (85.2%) and 12 girls (14.8 %), mean age 3 years and 1 month ([+ or -]1 yr 3 mo)--met the criteria for inclusion.

The guardian of each of the 81 children answered a questionnaire regarding the patient's history, including the presence of pre-, peri-, and postnatal factors that might constitute a high risk for neurologic lesions. Each guardian was also asked to convey his or her subjective impressions of the child's hearing acuity and language acquisition pattern. Finally, each patient's neuropsychomotor development was graded on the basis of when each began to sit and walk as well as other aspects of behavior.

After we reviewed the 81 questionnaires, we mailed a second questionnaire to the 53 guardians who had provided us with a complete address, and we received 29 responses. The questions on the second survey concerned diagnoses and treatments, and they solicited further data on each child's psychomotor development. The ages of the children whose guardians returned the second survey ranged from 28 months to 11 years. The 29 children were invited to undergo another BERA examination, and 16 did. The second BERA was performed between 6 months and 7 years and 1 month following the initial B ERA; we obtained data on the I/V amplitude ratio and the electrophysiologic threshold.

For our final analysis, the study population was made up of the 53 patients who were mailed the second questionnaire. They were divided into two groups:

Group I. Group I was made up of the 29 patients, mean age 5 years and 8 months ([+ or -]2 yr 6 mo), whose guardians answered the second questionnaire. This group was further divided into two subgroups: 16 patients who underwent a second BERA examination (group IA) and 13 who did not (group IB).

Group II. Group II was made up of the 24 patients whose guardians did not answer the second questionnaire.

Results

Initial assessment. We reviewed the responses to the initial 81 questionnaires (table) and the results of the first BERA examination. BERA curves showed signs of delayed maturation of the auditory pathway (enlargement of the interval I-V) in 79 of the 81 patients and a disproportionate wave I amplitude in the other two (who were given a diagnosis of a diffuse brainstem lesion). (12) The BERA analysis revealed that the electrophysiologic threshold was between 20 and 30 dB in 52 patients (64.2%), 40 dB in 12 patients (14.8%), and between 50 and 60 dB in 17 patients (2 1.0%). Latency intervals ranged from 4.0 to 5.7 msec (mean: 4.76 [+ or -] 0.23) in a normal distribution pattern. Only one patient exhibited a morphologic change in wave V; the wave I amplitude was larger than the wave V amplitude in 12 patients, with or without elongation of the I-V latency intervals.

Second assessment: Group I. Among the 29 patients whose guardians completed the second questionnaire (group I), 12(41.4%) had a richer vocabulary, while eight (27.6%) continued to be unable to speak or to exhibit only rudimentary language. A comparison of spoken language evolution between the first and second questionnaires is shown in figure 1. Moreover, 24 of the 29 children (82.8%) were undergoing phonotherapy, with or without psychotherapy. Analysis of behavior showed that with the exception of one isolated case of severe neuropsychomotor development delay and one case of suspected autism, all children were considered to be normal by their guardians in terms of showing affection. Some children were considered to be "excited," but none was aggressive. Overall, their verbal comprehension was good, although guardians mentioned that some children confused words that were similar or had some difficulty learning the names of the colors.

Second assessment: Group IA. With regard to electrophysiologic thresholds among the 16 children (31 ears; in one ear, the I-V latency interval was normal) in group IA, the second BERA examination revealed that only two experienced an improvement of 20 to 30 dB in both ears; even so, one of them still could not speak in complete phrases and the other could speak only a few words. In 20 of the 31 ears (64.5%), the I-V latency intervals ranged between 4.4 and 4.7 msec. Eight of the 16 patients (50.0%) continued to have a greater amplitude of wave I than of wave V; their I-V latency intervals ranged from 4.0 to 4.8 msec (mean: 4.4 [+ or -] 0.23), and they exhibited a normal distribution pattern. We found no relationship between the intensity of neuropsychomotor development delay and the quality of acquired language. The I-V latency interval was normal in only five of the 16 patients (31.3%), although three of these examinations were not considered normal because they featured an elongation of wave I amplitude (fi gure 2). The I-V interval was shortened but still moderately long in 17 of the 31 ears (54.8%). Only three of the 16 children (18.8%) experienced no reduction of the I-V interval on the second BERA test. In nine patients (56.3%), the abnormality was attributed to the presence of a central pathology in addition to deficient myelinization (wave I amplitude greater than wave V amplitude); in these patients, no normalization of the I-V interval was observed.

Possible etiologies. At the initial examination of 81 patients, seven were noted to have had postnatal seizures. The mean duration of the I-V latency interval in these seven patients was significantly greater than that in the other 74 patients (4.89 vs 4.74 msec; p<0.01). 1). In the 12 patients whose wave I amplitude was greater than the wave V amplitude, four had a history of perinatal jaundice, two had had postnatal seizures, one had had perinatal anoxia, and one had been born prematurely; no possible etiologic factor was apparent in the other four patients. Among the 81 patients, a history of perinatal anoxia was present in 19 (23.5%) and postnatal infection (recurrent pneumonia) in six (7.4%).

Discussion

As is the case with all studies that are based on surveys, questions arise as to how representative and valid our findings are. In analyzing the data on age, neuropsychomotor development, language acquisition, and hearing acuity provided by the guardians during the first questionnaire and BERA assessment, we did not find any differences that would suggest that there was any specific reason that the 29 guardians in group I were more interested than those in group II in answering the second questionnaire. Likewise, we did not find any specific reason why the guardians of the 16 patients in subgroup IA were more motivated to have their children undergo the second BERA examination than were those in subgroup IB.

In analyzing the changes we observed in BERA results, neuropsychomotor development, and language development patterns, we considered their relationships to several pre-, peri-, and postnatal factors, but it was still difficult to isolate simple cause-and-effect associations. (13) The most common possible risk factors were perinatal anoxia (n = 19), postnatal seizures (n = 7), postnatal infection (n = 6), perinatal jaundice (n = 5), prematurity (n = 5), and parental consanguinity (n = 5). Some patients had more than one of these possible risk factors.

Four of the five patients with perinatal jaundice exhibited a significantly greater increase in the amplitude of wave I than in the amplitude of wave V. This finding is in agreement with that of Lenhardt et al, who reported that hyperbilirubinemia can lead to an increase in axon degeneration and a greater loss of myelin than of ciliated cells. (14) None of the nine patients whose wave I amplitude was greater than the wave V amplitude demonstrated normalization of the I-V interval on the second BERA. Jiang observed that after asphyxiation, prolonged I-V intervals returned to normal more quickly than did reduced wave V amplitudes and decreased I/V amplitude ratios. (15)

Even though our patients had been examined during their childhood rather than during their neonatal period, they showed improvement in their BERA results. Knowing that clinical improvement in these patients can be slow, all possible therapies should be offered so that their improvement can take place as rapidly as possible. Because it is not possible to determine whether any specific child will or will not improve, it is worth administering stimulation therapy to all children as soon as possible. (2)

Newborns who are considered to be at risk for hearing loss must undergo BERA testing to ascertain their auditory threshold, I-V latency interval, and I/V amplitude ratio. (16) If these parameters indicate any change, the patient's development pattern and BERA results should be carefully observed because the verbal language acquisition pattern in these patients is often compromised.

Any child who has not acquired at least rudimentary verbal language skills by the age of 2 years should undergo BERA testing. An increase in the I-V latency interval requires that these patients undergo neurologic, phoniatric, and psychological evaluations so that any stimulation therapy can be started as soon as possible. In patients who exhibit an increase in wave I amplitude and/or any morphologic change in wave V, care must be intensified because brainstem anomalies in these patients tend not to resolve.

[FIGURE 1 OMITTED]

Figure 2

A: Right ear tracings obtained when the patient was 4 years and 10
months old (top) and when he was 11 years old (bottom) show that there
was an important reduction in transmission time in the brainstem from
4.7 to 4.2 msec. Wave morphology patterns on the two tracings are
similar. B: Left ear tracings obtained in another patient at the age of
3 years and 11 months (top) and at the age of 12 years (bottom) with 90
dB of stimuli shows a reduction in the latency interval from 4.6 to 4.4
msec. The wave I/V amplitude shift remains constant.

         A         B

I    1,5  1,5  1,7  1,6
II   2,6  2,6  2,8  2,6
III  4,1  3,8  4,1  3,8
IV   5,3  5,1
V    6,2  5,7  6,3  6,0

Note: Table made from line graph
Table. Information obtained from guardians as recorded on the initial
questionnaire (N = 81)

Finding                                n (%)

Presence of possible risk factors *

Prenatal period +                    16 (19.8)
Perinatal period ss                  31 (38.3)
Postnatal period n                   14 (17.3)

Subjective estimate of hearing

Normal hearing                       54 (66.7)
Impaired hearing                     19 (23.5)
Deafness                              8  (9.9)

Language pattern

Did not speak at all                 31 (38.3)
Expressed rudimentary speech         17 (21.0)
Spoke only a few words               19 (23.5)
Did not form phrases                  2  (2.5)
Misused letters                       3  (3.7)
Too young to be tested                9 (11.1)

Neuropsychomotor development

Normal                               32 (39.5)
Discrete or moderate delay           37 (45.7)
Significant delay                    12 (14.8)

* Some patients had more than one possible risk factor; some patients
had none.

+ Includes five cases of parental consanguinity.

(ss)Includes 19 cases of anoxia, five cases of jaundice, and five cases
of prematurity.

(n)Includes seven cases of seizure and six cases of infection.

Acknowledgment

We thank Maria Cecilia Lorenzi for her assistance in analyzing the data and translating the text.

References

(1.) Seeman M. Troubles du Langage chez l'Enfant. 2nd ed. Vol. 1. Berlin: Veb Verlag Volk und Gesundheit, 1965.

(2.) English GM, ed. Otolaryngology. Philadelphia: J.B. Lippincott, 1994.

(3.) Leech RW, Alvord EC, Jr. Anoxic-ischemic encephalopathy in the human neonatal period. The significance of brain stem involvement. Arch Neurol 1977;34:109-13.

(4.) Hall JE. The cochlea and the cochlear nuclei in neonatal asphyxia. Acta Otolaryngol (Stockh) 1964;Suppl 194.

(5.) Fria TJ, Doyle WJ. Maturation of the auditory brain stem response (ABR): Additional perspectives. Ear Hear 1984;5:361-5.

(6.) Gorga MP, Reiland JK, Beauchaine K, et at. Auditory brainstem responses from graduates of an intensive care nursery: Normal patterns of response. J Speech Hear Res 1987;30:311-8.

(7.) Gorga MP, Kaminski JR, Beauchaine K, et al. Auditory brainstem responses from children three months to three years of age: Normal patterns of response. II. J Speech Hear Res 1989;32: 281-8.

(8.) Kileny P, Robertson CM. Neurological aspects of infant hearing assessment. J Otolaryngol Suppl 1985;14:34-9.

(9.) Coll J, Carriere JP. [Early auditory potentials in neurologic involvement in children]. Rev Laryngol Otol Rhinol (Bord) 1988; 109:333-5.

(10.) Siegel L. The prediction of possible learning disabilities in preterm and full-term children. In: Field TM, Sostek AM, eds. Infants Born at Risk: Physiological, Perceptual and Cognitive Process. New York: Grune and Stratton, 1983:295-315.

(11.) Bento RF, Silveira JA, Ferreira MR, et al. Estudo do padrao de normalidade da audiometria de tronco cerebral (BERA) nas diversas faixas etrias, Rev Bras Otorrinolaringol 1988;54:37-41.

(12.) Starr A, Achor J. Auditory brain stem responses in neurological disease. Arch Neural 1975;32:761-8.

(13.) Murray AD. Newborn auditory brainstem evoked responses (ABRs): Prenatal and contemporary correlates. Child Dev 1988;59:571-8.

(14.) Lenhardt ML, McArtor R, Bryant B. Effects of neonatal hyperbilirubinemia on the brainstem electric response. J Pediatr 1984;104:281-4.

(15.) Jiang ZD. Maturation of peripheral and brainstem auditory function in the first year following perinatal asphyxia: A longitudinal study. J Speech Lang Hear Res 1998;41:83-93.

(16.) Meyer C, Witte J, Hildmann A, et al. Neonatal screening for hearing disorders in infants at risk: Incidence, risk factors, and follow-up. Pediatrics 1999;104:900-4.

Reprint requests: Dr. Vera Lucia R. Fuess, Faculdade de Medicina da Universidade de Mogi das Cruzes, Av. Dr. Candido Xavier de Almeida Souza, 200 Mogi das Cruzes, SP CEP 08780-210, Brazil. Phone: [phone]; fax: [phone]; e-mail: [e-mail]