Journal of Pediatric Psychology, Vol. 25, No. 1, 2000, pp. 5-13
© 2000 Society of Pediatric Psychology
Effects of Early Middle Ear Effusion on Child Intelligence at Three, Five, and Seven Years of Age
1 University of Houston, 2 University of Texas Medical Branch at Galveston
All correspondence should be sent to Dale L. Johnson, Department of Psychology, University of Houston, P.O. Box 1687, Houston, Texas 77204-5341. E-mail: dljohnson{at}uh.edu .
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionReferences |
|---|
Objective: This study tested the hypothesis that children with prolonged middle ear effusion (MEE) during the first 3 years of life are at risk for cognitive delays or deficits.
Method: A prospective study enrolled 698 children from diverse backgrounds and controlled for eight demographic and environmental factors. Participants were recruited at birth and monitored for ear status frequently in the home; 379 children were assessed for cognition with the Stanford-Binet, 4th ed., at 3 years of age, 294 at 5 years, and 198 at 7 years.
Results: Using the SAS General Linear Models (GLM) procedure, we found a significant direct relation between duration of bilateral MEE and Stanford-Binet Composite and Nonverbal Reasoning/Visualization Factor scores at age 3, but not at age 5 or age 7. Statistical clustering analysis revealed four groups with different temporal patterns of MEE: Low MEE, Early MEE (peaking at 0-6 months), Later MEE (peaking at 6-12 months), and High MEE. GLM analyses revealed no direct effects, but several moderated effects, of MEE cluster on cognitive development at 3 years, but none at 5 or 7 years. In general, children in the Later MEE and High MEE groups appeared to be more adversely affected by bilateral MEE at 3 years, but effects were moderated in complex ways by socioeconomic status or home stimulation. Growth curve modeling across the three assessment periods showed no effects of total duration of MEE but did indicate that children in the Later MEE cluster had low scores at age 3 but caught up at ages 5 and 7.
Conclusions: Prolonged MEE, especially between 6 and 12 months, may put children at risk for cognitive delays at 3 years, but the risk effect is not strong and effects are no longer detectable at 5 or 7 years.
Key words: cognitive delays; middle ear effusion.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
This report focuses on the effects of early middle ear effusion (MEE) on cognitive development, measured at 3, 5, and 7 years of age. Researchers in the field generally agree that cognitive development is a function of experience combined with biological predispositions that help to organize that experience (Feldman & Gelman, 1986
). Infants actively seek experiences that accord with their
existing cognitive structures, and if their range of sensory experiences is
limited, their potential for cognitive development is thereby reduced. Hence,
in theory, the child whose auditory stimulation is sufficiently restricted by
middle ear disease in the early years of life would show measurable delays or
deficits in cognitive development.
Despite many theoretical and empirical studies, however, the question of
whether persistent MEE has lasting effects on child development is unresolved.
The problem may lie in the nature of the hearing impairment caused by MEE; it
is generally intermittent rather than permanent, and partial rather than
total. Feldman and Gelman
(1986) have argued that
children afflicted with MEE may overcome the difficulties posed by partial,
intermittent hearing loss without lasting impairment. However, Eimas and
Clarkson (1986) hold that the
effects of MEE would "make it exceedingly difficult for the child to
learn just how the parental language maps very fine acoustic differences onto
different, language-specific categories." For the child with only a few
episodes of MEE, hearing acuity may be slightly to moderately diminished
(about 10 dB of hearing loss) for only a few days or weeks, but then return to
full acuity. Such a brief experience would not be expected to have measurable
effects on language or cognitive development. However, children with very
frequent or prolonged episodes of MEE may be at risk for hearing loss for a
large proportion of their early lives. We have previously shown that MEE-prone
children receiving conventional therapy from their physicians spent a mean of
38% and a maximum of 70% of their first three years with MEE
(Owen, Baldwin, Luttman, & Howie,
1993). Even with only partial hearing loss, these children might
indeed be expected to suffer developmental consequences.
Stool et al. (1994) have
reviewed many studies that report the effects of early MEE on language,
intelligence, and behavioral development and school achievement. Most of these
studies, except for the three discussed below, have used retrospective,
case-control designs and were subject to several kinds of sampling error.
Three prospective studies have related MEE to intelligence in infants and
children followed to age 5 and into the school years. Prospective designs in
these studies corrected many of the flaws of the studies described above. Two
prospective investigations of relatively small samples found no relations
between MEE duration and scores on intelligence tests at ages 3 and 4
(Gravel & Wallace, 1992;
Roberts, Campbell, Footo, & Burchinal,
1991; Roberts, Burchinal,
& Campbell, 1994). The third study, by Teele and associates,
investigated a larger sample. They found a significant effect of MEE duration
on scores for the Peabody Picture Vocabulary Test (PPVT) at three years
(Teele et al., 1984), and
significant associations with Wechsler Intelligence Scale for Children-Revised
(WISC-R) Full Scale, Performance and Verbal scores at 7 years of age
(Teele et al., 1990).
The generalizability of the Teele et al. studies is somewhat limited
because all children studied were from Euro-American, middle-class families.
The Roberts et al. studies
(1991,
1994), on the other hand,
investigated mostly African American and lower-class children. It is difficult
to determine whether the different findings of these two groups are
attributable to ethnic differences or differences in sample size. Thus, these
three prospective studies, despite their merits, have not resolved the
difficult question of whether or to what extent MEE may influence
developmental outcomes in children who represent the spectrum of ethnic and
socioeconomic groups found in our society.
The prospective study reported here included a large sample of children with diverse socioeconomic and ethnic characteristics and controlled for a broad range of potentially confounding demographic and environmental factors. The children were recruited at birth, monitored for the presence of MEE frequently in the home, and assessed at 3, 5, and 7 years of age with measures of cognition, language, and behavioral problems. In this report, we focus on associations between duration of MEE in the first 3 years of life and multiple measures of cognitive status at 3, 5, and 7 years. We hypothesized that communication restrictions associated with MEE would result in reductions in general intelligence, not just verbal aspects of intelligence, and that this reduction would be present at each of the outcome years assessed. Results of the language and behavior problems assessments carried out in this study will be reported separately.
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Participants
A sample of 698 healthy term infants was recruited at birth between 1984 and 1989 from the three newborn nurseries in Galveston County, Texas. Written informed consent was obtained from the parents. Enrollment was offered to all infants who met eligibility criteria without respect to gender, ethnicity, socioeconomic status (SES), or primary medical care provider. Because of the planned cognitive and language testing, children were excluded if English was not the primary language in the home. Other exclusion criteria included lack of a home telephone and plans to move from the area within 5 years. Medical exclusion criteria were birthweight of less than 2.5 kg, known perinatal asphyxia or an Apgar score of less than 7, admission to a Level III nursery, perinatal administration of ototoxic antibiotics, total bilirubin level above 17, and presence of craniofacial anomalies. Children with later evidence of sensory-neural hearing loss were also excluded.
MEE Monitoring and Treatment Groups
The MEE status (i.e., the presence of middle ear effusion independent of
other symptoms) of all subjects was monitored using tympanometry at home
visits every 2 to 4 weeks for the first 3 years of life, regardless of the
presence or absence of symptoms. Tympanometry, which measures motility of the
ear drum as an indicator of fluid in the middle ear, was performed by trained
technicians with automated screening tympanometers using a 226 Hz probe tone
(Owen, Baldwin, et al., 1993).
At 30% of the visits, acoustic reflectivity was also measured with an acoustic
otoscope (Endeco Medical, Marion, MA). Based on the literature regarding
prediction of MEE with tympanometry (ASHA, 1990) and acoustic reflectivity
(Lampe, Weir, Spier, & Rhodes,
1985; Schwartz & Schwartz,
1986), the following criteria for diagnosis of MEE were adopted:
(1) purulent otorrhea (pus draining from ear) visible without an otoscope, (2)
an acoustic reflectivity 
5, or (3) a type B tympanogram (i.e., compliance
of 0.0 or 0.1, or compliance of 0.2 or 0.3 only if the absolute gradient was
<0.1 ml). For children with tympanostomy tubes, MEE was diagnosed by the
presence of purulent otorrhea or a Type B tympanogram in the presence of an
external ear canal volume indicating that the tube was not patent. At home
visits, each ear was evaluated separately by these criteria and diagnosed as
either "normal" or "MEE." To minimize false positives,
the test was repeated if an abnormal reading was obtained. If a subsequent
normal reading was obtained, the ear was considered "normal."
Duration of MEE was calculated as described in Owen, Baldwin, et al.
(1993). All analyses were with
children who had bilateral MEE.
The original design called for an experiment in which children who had 6
weeks of continuous MEE at any time prior to one year of age were assigned
randomly to receive typanostomy tube placement or conventional treatment. The
design was abandoned when some children in the conventional therapy and
minimal MEE groups received tubes while parents of some children in the
treatment group refused tubes. Details of this experiment are reported in
Owen, Baldwin, et al.
(1993).
Measures
All of the following factors were included in the study to control for
their potential confounding effects on MEE duration or the outcome variables
selected for study. These control variables were selected because prior
research has shown their importance as factors influencing MEE or
intelligence.
At the time of enrollment, the child's gender, ethnicity, and birth rank
were obtained by parental report. Then, SES, level of educational stimulation
in the home environment, and mother's IQ were determined when each child was 2
years of age. The SES was measured with the Hollingshead Four Factor method
(Hollingshead, 1975), which
takes into account the education and occupation of both parents. The mother
was interviewed and observed with her child in her home using the Home
Observation of the Measured Environment (HOME) scale
(Caldwell & Bradley, 1984);
HOME is a measure of the amount of educational stimulation provided in the
home. Mothers' IQ was assessed with the Shipley Institute of Living Scale
(Zachary, 1987), a brief test
of vocabulary and conceptualization which correlates well (r =.76)
with the Wechsler Adult Intelligence Scale.
Dates were recorded for the start and cessation of breast-feeding, formula feeding, and feeding with other foods. The duration of breast-feeding of any amount (with or without other foods) was calculated and analyzed as a continuous variable. At the time of recruitment, we recorded the number of packs of cigarettes smoked, without regard to setting, per day by each member of the household (mother, father, and others). Because our previous research has indicated that mother's smoking is more predictive of the development of MEE in children than other smoking in the home, we used the total number of packs of cigarettes smoked per day by the mother.
To ensure that a subject's hearing was normal at the time of cognitive
testing, and also to exclude children with sensory-neural hearing loss, we
assessed the hearing of children at 6, 12, 24, and 36 months by brainstem
auditory evoked potential (BAEP) tests, and at 12, 24, and 36 months by
sound-field audiometry. At the time of hearing tests, middle ear status was
evaluated by pneumatic otoscopy by one of the investigators or a research
assistant. The method of BAEP testing has been described previously
(Owen, Norcross-Nechay, & Howie,
1993). Sound-field audiology was performed in a soundproof room by
a licensed audiologist.
Outcome measures were administered to children within one month of 36, 60, and 84 months of age by one of six psychology graduate students specially trained in the tests used. Examiners did not know the MEE history of the children tested. All assessments were carried out in the project's research offices in quiet, nondistracting rooms, with parents nearby, but not present. All children had hearing within the normal range at the time of cognitive testing.
The Stanford-Binet, 4th ed. (SBIV), administered at all three ages,
assesses a wide range of cognitive abilities
(Thorndike, Hagen, & Sattler,
1986). The Composite score provides a measure of general
intelligence. In addition, two factor scores, Verbal Comprehension (VC) and
Nonverbal Reasoning/Visualization (RV), were used at ages 3 and 5, and a third
factor, Memory, was included at age 7. The Kaufman Assessment Battery for
Children (KABC) achievement subtests
(Kaufman & Kaufman, 1983),
considered good predictors of later school performance, were used at age 5
only.
Design and Analysis
The original experimental design of this study was set aside in favor of a
correlational analysis using general linear models, including analysis of MEE
clusters. The primary research question was whether early MEE is related to
developmental outcomes. We analyzed our data in terms of the duration of
bilateral MEE experienced by each child and the temporal pattern of bilateral
MEE, regardless of group assignment or therapy. Hence, the study was analyzed
as a prospective cohort study.
Our first step in statistical analysis was to evaluate duration of
bilateral MEE during the first 3 years in relation to cognitive outcomes,
using the SAS General Linear Models (GLM) procedure
(Freund & Littell, 1981).
Included in this statistical model were a number of environmental/biological
factors: SES, ethnicity, HOME, gender, mother's intelligence, time spent in
day care, mother's smoking in the home, and breast-feeding history.
Next, we used cluster analyses to group participants according to the
temporal pattern of their bilateral MEE experience over the first 3 years of
life. The purpose of a cluster analysis is to use statistical methods to
identify groups of participants who are more similar to each other than to
other groups of participants with respect to certain specified parameters
(Anderberg, 1973). The
parameters used in our analysis were the proportion of days of bilateral MEE
experienced at 0-6 months, 6-12 months, 12-18 months, 18-24 months, and 24-36
months of age.
We performed the cluster analysis in two stages. The first stage employed a
hierarchical, agglomerative method, using Ward's minimum variance technique
(Ward, 1963). Pseudo
F statistics were examined to estimate the number of clusters
represented by the data. Four to six clusters were apparent. Because
hierarchical methods do not allow for reassignment of participants to other
clusters once they have been assigned, the results were followed by a
nonhierarchical K-means analysis (Ward,
1963) specifying a maximum of six clusters with a minimum of 20
participants per cluster. This sequential clustering method, which is most
efficient for studies of large data sets, identified four significant clusters
with a pseudo F statistic of 178.1.
Figure 1 displays the mean percentage of time with bilateral MEE for each cluster in 6-month age blocks. Once clusters were identified, we used a GLM procedure to analyze relations between MEE cluster and developmental outcomes by a correlational design. MEE cluster was the main variable; also included in the model were cognitive outcome variables, potential confounding environmental/biological variables, and interactions between the environmental/biological variables and cognitive outcomes. The final model, which excluded potential confounding variables that proved insignificant, included control variables for gender, ethnicity, SES, mother's smoking, mother's intelligence, and breast-feeding.
|
In the third step of analysis, we used growth curve modeling with SBIV Composite mental age scores available at all three time points. In this technique, individual scores are considered to be representatives of an underlying growth continuum that results in the differences in scores over time. Depending on the nature of the growth model, this procedure allows estimation of individual growth parameters that can then be modeled as outcome variables. The complexity of the growth curve depends on the number of available assessments. Given that only three time points were available for the current study, only the intercept and slope could be estimated for each individual; however, it was possible to test for possible curvature in the model by including a quadratic parameter as a fixed, rather than a random, effect.
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Sample Characteristics
The originally recruited sample consisted of 698 infants; 379 subjects were still active in the study at 3 years of age, 294 at 5 years, and 198 at 7 years. Attrition was greatest between enrollment and 18 months of age, before our measures of SES, HOME, and mother's intelligence were administered. Children who dropped out of the study early (0-3 years) were significantly more likely to be of African American ethnicity,
2(2) = 8.2,
p <.02, and had experienced relatively little MEE,
2(1) = 27.1, p =.001, compared to other
participants.
The frequency of drops by age is shown in
Table I. The drop category was
not related to the gender, ethnicity, or birth order of the child or to the IQ
of the child at age 3, although the latter test had no data for those who
dropped before age 3. There were also no differences in the MEE cluster,
mother's IQ, HOME score, parental smoking behavior, or SES by drop category.
However, mothers who did not breast-feed were significantly more likely to
drop before the child was 3, 2(3) = 21.5, p <.01,
and participants with minimal MEE were more likely to drop, 2
(6) = 33.6, p <.01. Also, those participants who spent more time
in daycare were more likely to remain in the study, F(3, 694) = 89.2,
p <.01.
|
The majority of the sample was Euro-American (56%); 30% were African American and 14% Hispanic (English-speaking only). The children were almost equally divided by gender (51% female). The mean score for SES was 37.4 (SD = 12.5), with values ranging from 9 to 66. Hence, SES scores were quite heterogeneous, with the mean in the middle of the scale. Scores of the families on the HOME scale were also typical of the normative group for this test, with a mean of 39.1 (SD = 4.8) and a range from 18 to 45. Results from the Shipley measure of mothers' intelligence also suggest the group was fairly typical of the normative sample, with a mean of 50 (SD = 10.1) and a range from 18 to 68.
Mothers smoked an average of 0.16 packs of cigarettes per day (SD =.42). These data also represent a skewed distribution; only 18.5% of the mothers reported any smoking. Fifty-four percent of the mothers reported breast-feeding their infant at least briefly; mean duration of breast feeding was 5.3 months.
Repeated hearing assessments were performed on all children, and records of these tests were later reviewed by one of the physician investigators (D.P.M.). One child had evidence of sensory-neural hearing loss and was dropped from the study. Mild conductive hearing losses (0-24 dB in one or both ears) were sometimes observed in association with the presence of pressure equalization tubes, draining pus from the ear, middle ear effusion (type B tympanogram), or negative middle ear pressure (type C tympanogram). Twelve percent of children experienced at least one episode of moderate conductive hearing loss (threshold >24 dB in one or both ears). All children with transient mild or moderate conductive hearing loss were documented to have returned to normal at a later date; hence, no permanent hearing loss was detected in any child.
Duration of Bilateral MEE and Intelligence Outcomes
SBIV Composite scores at ages 3, 5, and 7 years were 100.6 (SD =
12.2), 98.3 (SD = 12.6) and 98.0 (SD = 14.6), respectively,
suggesting a sample of children with a normal IQ distribution.
The associations between bilateral MEE and intelligence outcomes were evaluated by GLM procedures, using models that included controls for gender, SES, HOME, mother's intelligence, ethnicity, and mother's smoking. At age 3 we found that duration of MEE from 0-36 months was inversely related to SBIV Composite scores, t(336) = -1.79, p <.05, one-tailed. The result was significant for the SBIV RV factor, t(335) = -1.86, p <.05, one tailed, but not for the SBIV VC factor, t(336) = -0.86, p =.39.
MEE Clusters (Temporal Pattern of MEE) and Intelligence Outcomes
Cluster Analyses. The results of the bilateral MEE cluster
analyses described in the Method section are shown in
Figure 1. The Low MEE group
includes participants with consistently low levels of MEE across all time
points. Two groups typically had intermittent episodes of MEE during the first
0 to 12 months of life: the Early MEE group, whose MEE peaked before 6 months
of age, and the Later MEE group, whose MEE peaked between 6 and 12 months. The
High MEE group had relatively higher levels of MEE than the other clusters
across all time periods.
We analyzed relations between MEE cluster and cognitive outcome measures at 3, 5, and 7 years in several sets. In each set, the first analysis included interactions with the control variables of SES, HOME, ethnicity, mother's intelligence, and mother's smoking. If some of the interactions proved insignificant, a second analysis that eliminated these interactions was performed.
SBIV at 3 Years. When models that included control variables were tested, we found no direct relations between MEE cluster and any cognitive outcome variables at any age. When potential moderating effects were included, the difference between MEE clusters on the Composite scores at age 3 was significantly moderated by SES, F(3, 313) = 3.24, p <.05, and HOME score, F(3, 313) = 2.68, p <.05.
The effect of having consistently high levels of MEE or having high levels during the 6-12 month period was moderated by SES; that is, in the High and Later MEE clusters only, children with higher SES tended to have higher Composite scores. On the other hand, a stimulating environment was more positively related to Composite scores in the Early MEE group. Hence, children whose bilateral MEE peaked during the first 6 months appeared to be most able to benefit from a more stimulating home environment, while children with higher levels of MEE, or MEE which peaked between 6 and 12 months, appeared more adversely affected if they came from lower SES families. Similar results were found for the RV factor, but not for the VC factor.
SBIV at 5 and 7 Years. There were no significant relations for total duration of MEE or for clusters at 5 or 7 years.
Kaufman Assessment Battery for Children at 5 Years. There were no relations of MEE cluster to KABC scores at age 5 for the Achievement subscales, nor were interactions with the environmental/biological variables detectable.
Growth Curve Modeling of SBIV Scores
The initial model indicated no curvature in the growth model since the
fixed parameter for the quadratic did not differ from zero, t(318)
=.46, p >.05, nor was the variance of the curvature parameter
significant. The curvature term was dropped from the model. The model
including the intercept and slope as random effects revealed that, while there
was a significant individual variance in the intercept (z = 10.91,
p <.05), there was no significant individual variation in the
slope parameter (z =.39; p >.05), nor was the intercept
significantly related to the slope. Tests of the fixed effects indicated that
the slope, t(509) = 77.74, p <.05, and the intercept,
t(373) = 136.24, p <.05, did differ significantly from
zero. Thus, the final random coefficient model kept the intercept as random
but included the slope as a fixed effect.
The next step was to include the predictors into the model. All predictors were included in estimating relations to both the intercept and slope. Slopes were considered first, and variables that did relate significantly to the growth in mental age were dropped. The only variables that did relate to the growth were SES, F(1, 486) = 4.45, p =.04, and daycare, F(1, 486) = 4.61, p =.03. Next, those variables that were not related to the intercept were dropped. Those that remained were ethnicity, F(2, 486) = 11.74, p <.01; SES, F(1, 486) = 25.33, p =.01; mother's IQ, F(1, 486) = 7.92, p <.01; packs of cigarettes smoked per day by the mother, F(1, 486) = 4.80, p =.03; whether or not the mother breast-fed her infant, F(1, 486) = 3.65, p =.05; and the birth order of the child, F(1, 486) = 8.88, p <.01.
Next, the total duration of bilateral MEE in the first 3 years of life was added to the model, but this variable did not differentiate the intercept or the slope. When MEE cluster was added, it did not relate to the intercept, but there was a relation to slope, F(1, 482) = 4.42, p <.01. Examination of the parameter estimates indicated that the Early and Low clusters had significantly slower growth in mental age than did the Later cluster, t(482) = -3.63, p <.01; t(482) = -2.23, p =.03, respectively, and there was a tendency for the High cluster to also have a less steep slope, t(482) = -1.82, p =.07. There was no indication that the relations of MEE to growth in mental age were moderated by other factors.
Although the fact that the Later MEE cluster tended to show the most rapid growth in mental age may seem contradictory, this was the cluster that had the lowest adjusted IQ scores at age 3. Therefore, the more rapid increase in the mental age of these children may reflect developmental "catch-up" that would explain why the difference in clusters was not significant at 5 or 7 years of age.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
The effects of MEE in the first 3 years of life were assessed with multiple measures of cognition, with eight background variables controlled. Data were analyzed cross-sectionally and longitudinally, and analyses included total duration of bilateral MEE through 3 years of age as well as temporal pattern of MEE (clusters). Cluster assignment indicated whether a child's exposure to MEE came mainly in the first or second half of the first year, or was consistently low or high over the 3-year period.
Considering first the cross-sectional analyses, with significant background variables included in the models (SES, HOME, ethnicity, mother's intelligence, and mother's smoking), we found two significant direct effects. The significant relation between total MEE duration and 3-year-old SBIV Composite scores, which is our most reliable and valid measure of child intelligence, suggests that persistent MEE may have harmful effects in the early years of child cognitive development. We estimate that a reduction of amount of MEE by 50% during the first 3 years would be associated with an increase of the child's intelligence by 3.5 SBIV Composite points at 3 years of age. This amount is not large, but is still considerable. For the sake of comparison, one year in a Head Start program is estimated to raise the intelligence of low-income children by 5 IQ points, on average. The presence of a statistically significant result for a nonverbal (RV) aspect of the SBIV as well as the Composite IQ may suggest that MEE has broad associations with cognitive abilities in general, not just with verbal abilities.
The absence of significant effects at age 5 and 7 for any of the cognitive measures suggests that negative effects of MEE may disappear over time. The longitudinal analysis using growth curve modeling made it possible to test this hypothesis. We found no effects of total MEE duration in these analyses, but an examination of the temporal pattern of MEE using clusters showed evidence consistent with a "catch-up" phenomenon. Children in the Later MEE cluster had lower IQ scores at age 3 than the Low or Early MEE clusters, but did not differ significantly from the other clusters at ages 5 and 7. A trend toward the same "catch up" pattern was observed in the High MEE cluster.
Our cross-sectional cluster analyses demonstrated no significant direct effects of temporal pattern of MEE on intelligence outcomes at any age, when environmental variables were controlled. However, we observed a number of complex moderated effects on cognitive development at age 3. Moderated effects indicate that intelligence outcomes were associated with MEE only when certain other conditions were present. These effects are complex because typically they involve an interaction between MEE cluster and one or more factors that are themselves related to IQ, particularly, SES and HOME. In many of our models, SES and HOME had stronger effects on measured cognitive outcomes than bilateral MEE.
In general, the effects of MEE on cognitive outcomes, although relatively
weak, appeared to be stronger for the Later MEE cluster (children who had most
of their bilateral MEE between 6 and 12 months of age) than for the other
three clusters, including the High MEE group. For none of the outcome measures
were children in the Low or Early MEE groups negatively affected. It is
unclear why the period from 6 to 12 months might be one of special
vulnerability to transitory hearing problems. As the Later MEE group was more
adversely affected than the High MEE group, this result might suggest that
interruption of hearing acuity during this period of language learning is more
significant to developmental outcomes than the more frequent interruptions of
hearing acuity over a longer period experienced by the High MEE group. Teele
et al. (1984) reported
stronger effects for children who experienced a higher proportion of days with
ear effusion in the first year of life than for children with more MEE in the
second and third years. These results appear to be somewhat similar to our
finding of greater effects of bilateral MEE for the Later MEE cluster than
other MEE clusters, although our method of defining time periods was different
from that of Teele and associates.
SES had a moderating influence on the effects of MEE. Children in the Later MEE cluster who were also from low SES backgrounds had lower Composite IQ scores. In general, high SES appeared to blunt the potentially harmful effects of MEE. HOME stimulation was more positively related to intelligence in the Early MEE cluster than in the Later or High MEE clusters. This finding suggests that a stimulating home environment may be more advantageous if one does not have much MEE between 6 and 12 months or prolonged MEE throughout the first 3 years of life.
What do the results of this study contribute to the ongoing controversy about the long-term effects of MEE on child cognitive development? Many investigators have shown that most children who have MEE experience only a temporary period of mild-to-moderate hearing loss, with a return to normal acuity after about 28 days. Even children with frequent and prolonged MEE, like several of the children in our sample, still spend most of their time in the first 3 years without MEE and without significant continuing hearing loss. However, prolonged bilateral MEE, especially in the second half of the first year, may put children at more risk for cognitive delays than MEE at other times. Infants at this time are developing an awareness of language. As a child's awareness of language grows and the child becomes more involved with language communications between 6 and 12 months, even temporary disruption of these powers may be felt more strongly than continuous attenuation of hearing acuity over a longer period of time. The results of our analyses suggest that for children with prolonged middle ear effusion, interventions to limit MEE should be undertaken, if at all, during the first year of life.
Our finding of significant associations between MEE and intelligence at 3
years of age differs from the reports of Gravel and Wallace
(1992) and Roberts et al.
(1991,
1994), who found no evidence
suggesting MEE effects on intelligence scores early in the child's life. Teele
et al. (1984) did obtain
significant results at 3 years of age with one measure, the Peabody Picture
Vocabulary Test, which is sometimes regarded as a measure of verbal
intelligence. Our study appears to be the first prospective study to report
evidence of a negative association between MEE duration and a broad measure of
child cognition like the SBIV Composite score at 3 years of age. That Teele et
al. (1990) obtained
significant MEE effects at 7 years of age using the WISC-R, and we did not
using the SBIV, raises the question of method differences. We think these are
minimal; both tests are standard measures of child intelligence and have
similar psychometric properties. Teele et al. had positive results to verbal
and nonverbal measures, and we had negative results with similar verbal and
nonverbal measures. This suggests that the differences were not a function of
type of test, but of other factors, most likely sample selection. It should
also be noted that Roberts et al.
(1994) did not obtain OME
effects on intelligence using the WISC-R ages 6.5 and 8.
We found that children with prolonged bilateral MEE were at significant risk for mild cognitive delay at 3 years. Those whose MEE peaked between 6 and 12 months appeared to be at most risk, particularly if they had other unfavorable risk factors for cognitive impairment such as low SES and limited home stimulation. However, these modest effects of MEE appeared only with the large sample at 3 years and were not evident with the smaller samples at 5 and 7 years of age. Growth curve modeling done with children available at all three assessment times showed no significant effects of total duration of bilateral MEE. However, longitudinal analysis with clusters suggested that children most affected by early MEE (the Later MEE cluster) caught up with children in the other clusters by age 7. Overall, our study suggests that children with prolonged MEE in their early years may be temporarily delayed, but that the effect is small and discernible only with a large sample.
|
Acknowledgments |
|---|
This project was funded in part by the National Institute of Child Health and Human Development, grant no. HD20988.
Received August 11, 1997; revision received December 19, 1997; revision received April 15, 1998; accepted August 12, 1998
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Anderberg, M. R. (1973). Cluster analysis for applications. New York: Academic Press.
Caldwell, B. M., & Bradley, R. H. (1984). Home observation for measurement of the environment. Little Rock, AR: Center for Child Development and Education, University of Arkansas at Little Rock.
Eimas, P. D., & Clarkson, R. L. (1986). Speech perception in children: Are there effects of otitis media? In J. F. Kavanagh (Ed.), Otitis media and child development (p. 157). Parkton, MD: York Press.
Feldman, H., & Gelman, R. (1986). Otitis media and cognitive development: Theoretical aspects. In J. F. Kavanagh (Ed.), Otitis media and child development (pp. 27-44). Parkton, MD: York Press.
Freund, R. J., & Littell, R. C. (1981). SAS for linear models: A guide to ANOVA and GLM procedures. Cary, NC: SAS Institute.
Gravel, J. S., & Wallace, I. F. (1992). Listening and language at 4 years of age: effects of early otitis media. Journal of Speech and Hearing Research, 35, 588-595.
Hollingshead, A. B. (1975). Four-factor index of social status. Unpublished manuscript, Department of Sociology, Yale University, New Haven, CT.
Kaufman, A. S., & Kaufman, N. L. (1983). K-ABC: Kaufman assessment battery for children. Circle Pines, MN: American Guidance Service.
Lampe, R. M., Weir, M. R., Spier, J., & Rhodes, M. F.
(1985). Acoustic reflectivity in the detection of middle ear
effusion. Pediatrics, 76,
75-8.
Owen, M. J., Baldwin, C. D., Luttman, D., & Howie, V. M. (1993). The universality of otitis media with effusion detected by tympanometry on frequent home visits in Galveston, Texas. Proceedings of the Fifth International Symposium on Recent Advances in Otitis Media, 17-20.
Owen, M. J., Norcross-Nechay, K., & Howie, V. M. (1993). Brainstem auditory evoked potentials in young children before and after tympanostomy tube placement. Journal of Pediatric Otorhinolaryngology: 25, 105-117.
Roberts, J. E., Campbell, F., Footo, M., & Burchinal, M. (1991, April). Otitis media from birth to five and later intellectual development and academic performance. Paper presented at the Biennial Meeting of the Society for Research in Child Development, Seattle, WA.
Roberts, J. E., Burchinal, M. R., & Campbell, F.
(1994). Otitis media in early childhood and patterns of
intellectual development and later academic performance. Journal of
Pediatric Psychology, 19,
347-367.
Schwartz, D. M., & Schwartz, R. H. (1986).
Validity of acoustic reflectometry in detecting middle ear effusion.
Pediatrics, 79,
739-742.
Stool, S., Berg, O. A., Berman, S., Carney, C., Cooley, J., Culpepper, L., Eavey, R., Feagans, L., Finitzo, T., Friedman, E., Goetz, J., Goldstein, A., Grundfast, K., Long, D., Macconi, L., Melton, L., Roberts, J., Sherrod, J. & Sisk, J. (1994). Otitis media with effusion in young children. AHCPR Publication No. 94-0622. Washington, DC: U.S. Department of Health and Human Services.
Teele, D. W., Klein, J. O., Chase, C., Menyuk, P., Rosner, B. A., & the Greater Boston Otitis Media Study Group. (1990). Otitis media in infancy and intellectual ability, school achievement, speech, and language at age 7 years. Journal of Infectious Diseases, 162, 685-694.[Web of Science][Medline]
Teele, D. W., Klein, J. O., Rosner, B. A., & the Greater Boston
Otitis Media Study Group. (1984). Otitis media with effusion
during the first three years of life and development of speech and language.
Pediatrics, 74,
282-287.
Thorndike, R. L., Hagen, E. P., & Sattler, J. M. (1986). Guide for administering and scoring the Stanford-Binet Intelligence Scale: Fourth Edition. Chicago: Riverside.
Ward, J. H. (1963). Hierarchical grouping to optimize an objective function. Journal of the American Statistical Association; 58, 236-244.
Working Group of Acoustic Immittance Measurements and the Committee on Audiologic Evaluation, American Speech-Language-Hearing Association. (1990). Guidelines for screening for hearing impairment and middle-ear disorders. ASHA, 32(suppl. 2), 17-24.
Zachary, R. A. (1987). Shipley Institute for Living Scale. Los Angeles, CA: Western Psychological Services.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. E. Roberts, M. R. Burchinal, and S. A. Zeisel Otitis Media in Early Childhood in Relation to Children's School-Age Language and Academic Skills Pediatrics, October 1, 2002; 110(4): 696 - 706. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Perrin Should We Operate on Children with Fluid in the Middle Ear? N. Engl. J. Med., April 19, 2001; 344(16): 1241 - 1242. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||