Symposium
Chairs: Lynne Vernon-Feagans and Sandy Friel-Patti
Discussants: Lynne Vernon-Feagans and Sandy Friel-Patti
This symposium will present different data and viewpoints about whether and under what conditions otitis media with effusion (middle ear disease) early in life affects development in the preschool period and later in childhood. Otitis media is of concern to those studying infant development because up to one half of infants have chronic problems with this disease in the first few years of life (Vernon-Feagans, 1999). Although the disease is rather mild it is associated with a hearing loss that can last many months after the symptoms disappear. In many cases of otitis media the child exhibits no symptoms at all, making it difficult to document the disease and its effects. The hearing loss associated with otitis media has been shown to be a mild to moderate hearing loss across the speech frequencies (Gravel & Nozza, 1997; Roberts et al., 1995, 1998). This hearing loss occurs at a time when children are learning important language and social skills and can last for many months. Initial retrospective studies suggested that otitis media may be linked to a host of developmental delays and later learning problems for children who had experienced chronic problems with this disease early in life (Zinkus & Gottlieb, 1980). Only recently have there been good longitudinal studies that document otitis media, hearing, and development. This symposium will present results from some of the major longitudinal studies of otitis media and development. The major longitudinal studies of otitis media vary considerably in the numbers of children that were followed, the SES and ethnicity of the children followed, daycare attendance, the amount and quality of hearing data obtained, and the kind of outcomes examined. The results from these studies have generally found some short term negative effects on expressive language and behavioral/social processes but the results are often mediated by hearing loss and may be more pronounced in certain groups of children. For instance some studies have found that otitis media only has severe negative effects on language and behavior for children who are in low quality daycare. High quality daycare appears to buffer children against the negative effects. The presenters in the symposium will review many of the longitudinal studies and present some new data that may shed light on a variety of theoretical and analytic approaches that may help us to better understand the effects of this disease on development. The discussion will focus on how to possibly reconcile these conflicting findings and propose future directions for research in infancy, health, and development.The discussants of this symposium will focus their comments on how to interpret the similar and different findings from these studies. This will be done from both a theoretical and methodological perspective. First, there are a variety of theoretical perspectives that can encompass the findings. Some have proposed that otitis media has direct effects on language development because of the hearing loss. Others have suggested that basic semantic and syntactic language development is rather impervious to environmental and biological insults and that more subtle aspects of language, such as discourse and narrative skills, will be most affected. Others have suggested that early hearing loss permanently alters the way children process auditory information through differing development of the auditory pathways in the brain. Still others have hypothesized that otitis media may have its greatest effects only under certain conditions (moderators such as poor quality daycare, ambient noise in the home, etc.). Second, there are many methodological flaws in even the best studies. The discussants will review the designs and analysis strategies of the major studies, including a discussion of how often and when hearing data should be obtained from infants, the kind of contextual information needed, the specification of outcome measures, and power issues. Finally, the discussants will suggest future directions for research in otitis media as well as how many basic studies of infancy might incorporate some measure of otitis media in their studies.
Details of individual items:
paper
no abstract
paper
Introduction. Despite much research, the relationship between recurrentotitis media with effusion (OME) and language development remainsinconclusive. One of the most serious flaws of previous research is thefailure to include hearing data that could explain the link betweenrecurrent OME and communicative outcomes. The current investigation wasdesigned to remedy some of the previous shortcomings by frequently assessingboth middle ear status and hearing in a large cohort of childrenprospectively followed since birth. One of the aims of this study was toexamine when the occurrence of OME would have the greatest impact and whatthe role of hearing is in the relationship. Methods. The subjects for this investigation were 114 children enrolled by2.5 months of age in the Bronx ChEARs Project. There were 55 females and 59males; the group was diverse with respect to ethnicity; 59% percent of thecohort were from middle/upper SES backgrounds. The children were scheduledfor bi-monthy middle-ear and hearing assessments from 2.5 months to 36months. Beginning at 5 months of age, OME was diagnosed on the basis of anabnormal tympanogram (tympanometric width > 275 daPa). Behavioralaudiometry was used to assess hearing sensitivity beginning at 5 months. Inthis report, we averaged the obtained hearing thresholds over testedfrequencies, and a mean was obtained for each year's data. The PreschoolLanguage Scale-3 (expressive and receptive scales) was administered yearly. Results. We categorized the sample in terms of their OME status on a yearlybasis: bilaterally free, frequent OME, and infrequent OME. One way ANOVAsindicated that the groups differed in average hearing thresholds during thefirst, second, and third years, although average hearing thresholds of allgroups tended to be lower than previously reported. Children in thebilaterally free group had the lowest thresholds, and the children withfrequent OME had the highest thresholds. We then examined 3 year languagescores as a function of OME and hearing (i.e., elevated hearing v. normalhearing). No relationship was found between OME or hearing in the first twoyears of life and the PLS scores at 3 years of age. However, significantmain effects for OME and hearing were found for 3-year language scores.Although children with elevated hearing had significantly poorer auditorycomprehension scores, children with frequent OME had higher auditorycomprehension and expressive communication scores. This difference was, inpart, related to other factors such as SES. Conclusion. In this group of carefully followed preschool children, OME wassomewhat less prevalent than in other cohorts. Although OME did result inelevated thresholds, the mild degree of hearing loss may account for thefailure to find much of an effect on standardized language scores. It alsomay be that use of global as opposed to more specific indices has obscuredthe impact on language. Other findings from this cohort will shed morelight on this issue.
paper
Cumulative risk models have been developed to examine the multiple factors which can either put a child at risk for poor outcomes or buffer a child from poor outcomes. Often they are used as a framework for studying pathological development, but it has been less common for researchers to use them to examine variability within a group of normally developing children. Some early childhood factors such as day care experience and middle ear disease (otitis media) have been studied in relation to childrens later social and language behavior, but have produced conflicting results as to their impact on development. Typically, these factors are studied singly, but in real life factors are rarely experienced in isolation. For example, children who attend day care are at a higher risk of contracting otitis media than their peers who do not attend day care. Therefore, these factors actually may be interrelated and may work in combination with each other to influence development. Examining only one of these factors may reveal an incomplete picture of how those early experiences are related to later outcomes. The present study examined whether the basic tenets of the cumulative risk model were useful for predicting normal variability in childrens social and language behavior. We were particularly interested to understand if shy and inhibited behavior and less developed language within a sample of normally developing children could be predicted from a combination of measures in early childhood. Rather than focusing on one factor in early childhood and studying it as if it occurred in isolation from other factors, the present study examined how chronic middle ear disease (otitis media), the age at which children first began child care, and the quality of child care related to childrens social and verbal behavior at seven years of age: singly, additively, or in interaction with each other. The 37 children (14 boys) in this study are part of a larger sample of children who were part of a longitudinal study of health and child care in a semirural area of the northeastern U.S.; all were at very low risk for developmental difficulties. Children were recruited into the study from child care centers (2 high quality, 1 low quality) before their first birthday and examined for otitis media weekly until they were three years old. Follow-up interviews and observations of children were conducted when children were in first grade. Multiple regression analyses indicated that children with the highest Arisks@ (low quality child care, chronic otitis media, late entry into child care) for social/behavioral problems were the most shy and withdrawn with their friends at seven years of age. Verbal behavior and ability were predicted by quality of care and experience with otitis media. Higher quality child care was related to higher verbal ability, and chronic otitis media was related to lower verbal ability. These results support the hypotheses that the examining the interaction of childrens early experiences is important for an understanding of later behavior.
paper
Otitis media, particularly otitis media with effusion (OME) poses an interesting challenge for research in child development. Usually characterised as involving a mild hearing loss of conductive type that fluctuates and resolves, it in fact embraces a diversity of presenting signs and symptoms, differing across age. Debate with an emphasis on language outcomes, has continued on whether or not developmental and educational consequences (sequelae) do or do not exist after remission. In recent years it has been found that differences in abstract reasoning, pragmatics and behaviour are at least as large as any in language. In children over about 4, it is hard to show language effects. As a marker of whether or not a child has the disease, an objective test on a single occasion is low in predictive validity. As over 80% of children have had an episode at some point, this is not a useful disease marker. It is therefore desirable that developmental studies with the above outcome variables should attempt to document total OM histories, at least at the level of a few well-tried questions that would establish general severity and persistence. Unfortunately, many longitudinal studies of language and cognition have ignored OME; in studies of risk factors for cognitive delay that are also risk factors for OME (e.g. low birthweight, non breastfeeding) this amounts to a serious methodological flaw.There is some evidence that effects of OME on development are mediated in part by illness effects on behaviour, rather than by reduced auditory input alone. Statistically, the sequelae are not long-lived. However there is sufficient evidence to justifiably regard OME as a condition potentially requiring intervention. Updating the qualification in the 'potentially' requires the completion of a current research agenda concerning the detailed processes of developmental influence and of 'co-factors'. OME, whilst not drastic on its own, may materially add or synergise with these co-factors in producing developmental delays.