DOI
https://doi.org/10.25772/9DG8-0R93
Defense Date
1981
Document Type
Thesis
Degree Name
Doctor of Philosophy
Department
Human Genetics
First Advisor
Walter E. Nance
Abstract
A self-administered thirteen page Hearing Loss Questionnaire (HLQ) was designed in order to systematically collect medical and family history data on deaf children and their families. Data were collected from over 400 families with one or more children enrolled in September 1979 at the Maryland School for the Deaf (MSD). Almost 70% of the parents provided pedigree and family history information by completing the detailed HLQ. Computer analyses of the collected data allowed a thorough examination of almost 200 medical and family history variables, providing useful reference data on the MSD probands. Parental responses to a four-step rating scale of proband hearing ability were compared with actual audiometric data, allowing comparison with similar data from previous studies of hearing populations. Family history data on the non-respondents were available from school records, providing a unique opportunity to assess the potential response bias in questionnaire studies of genetic disease. Segregation analysis was performed on the informative sibships ascertained by incomplete truncate selection. The pooled estimate of the ascertainment probability, π, was 0.488, with no significant evidence of heterogeneity among the respondents and non-respondents. The hypothesis of fully penetrant dominant inheritance (H0:p=0.50) was accepted in the Deaf by Hearing matings. However, the maximum likelihood estimate of the segregation ratio (p=0.257) was consistent with reduced penetrance in these families, as it also was in the Deaf by Deaf matings (p=0.31). There were no significant differences in the maximum likelihood estimates of p or of the proportion of sporadic cases, x, between respondents and nonrespondents in the Hearing by Hearing matings. Among the non-consanguineous Hearing by Hearing matings with no family history of hearing loss, the maximum likelihood estimate of x was 0.81. The removal of 46 sibships with probands born during the 1964-65 rubella epidemic reduced x to 0.71, indicating the potential value of segregation analysis for monitoring the secular trends in sporadic vs. genetic deafness. Among Hearing by Hearing matings with a family history of early onset hearing loss, a recessive hypothesis with no sporadic cases (H0:p=0.25, x=O.OO) fit the data well. However, the same hypothesis was rejected among the Hearing by Hearing matings with a family history of "presbycusis", where x=0.59. Thus, although a family history of early onset hearing loss appears to be a much more reliable index of a genetic etiology that does a family history of "presbycusis", the results of this study suggest that the latter may also be a positive risk factor. The HLQ data implied that both parents and doctors may underestimate the extent to which genetic factors contribute to childhood hearing loss, even in the presence of a positive family history. Genetic factors were estimated to account for approximately 35% of the deafness in the MSD population. In the group with genetic deafness, the estimated proportions of recessive, dominant, and X-linked deafness were 57%, 39.%, and 5% respectively. Comparison of the estimates in the respondent vs. the non-respondent groups revealed remarkable similarity between the two groups, indicating that the use of the HLQ did not further confound existing biases. This study has demonstrated the value and utility of using self-administered questionnaires in genetic research. Indeed, the HLQ may serve as a useful prototype for future large scale population based studies of deafness in man.
Rights
© The Author
Is Part Of
VCU University Archives
Is Part Of
VCU Theses and Dissertations
Date of Submission
7-14-2016