The broad, long-term objectives of this project are to develop adult reference standards and methodology necessary for a subsequent study which will test the hypothesis that longitudinal follow-up of hearing levels may be an effective method of early detection of deterioration in hearing thresholds in non-industrial settings.
The specific aims are 1) to develop age-specific reference standards for hearing thresholds in adult men and women participating int he Baltimore Longitudinal Study of Aging, 2) to develop statistical methodology necessary to develop the longitudinal age- specific percentile reference curves, and 3) to develop statistical methodology for testing hypotheses regarding the variability in longitudinal studies of hearing thresholds. The objectives and aims of this project will contribute to the overall goal of reducing hearing- related disability by developing tools for early detection of changes in hearing thresholds at a point when interventions may be most effective. Longitudinal tracking of changes in hearing thresholds may also provide a clear demonstration to the individual that intervention is necessary and effective. The study design will utilize existing longitudinal data from over 756 men and 426 women, age 20-97, for pulsed pure-tone hearing thresholds collected and archived since 1965 by the Baltimore Longitudinal Study of Aging. Existing longitudinal data analysis techniques such as the linear mixed-effects regression models will be utilized and modified to allow for the construction of age-specific percentile curves which will serve as reference standards for longitudinal follow-up of changes in hearing thresholds. Furthermore, Monte Carlo studies will be conducted to compare the statistical properties of restricted maximum likelihood and maximum likelihood tests of the random-effects variance component in order to develop appropriate methods of testing hypotheses regarding the variability of longitudinal changes in hearing thresholds.
| Morrell, C H (1998) Likelihood ratio testing of variance components in the linear mixed-effects model using restricted maximum likelihood. Biometrics 54:1560-8 |
| Morrell, C H; Pearson, J D; Brant, L J et al. (1997) Construction of hearing percentiles in women with non-constant variance from the linear mixed-effects model. Stat Med 16:2475-88 |
| Morrell, C H; Gordon-Salant, S; Pearson, J D et al. (1996) Age- and gender-specific reference ranges for hearing level and longitudinal changes in hearing level. J Acoust Soc Am 100:1949-67 |
