A retrospective analysis of audiograms was completed for persons from 3 different patient groups. These 3 groups were (1) persons with unilateral Meniere's disease, (2) persons with unilateral acoustic tumor, and (3) persons from the general clinical population of an audiology clinic in a major medical center. As Paparella, McDermott, Luiz, and de Sousa (1982) also report, the most common audiometric configuration in the Meniere's disease group was peaked, a condition in which poorer hearing is reported in the low and high frequencies, and an island of better hearing sensitivity occurs at either 1.0 or 2.0 kHz. Some of the peaked audiograms from persons in our Meniere's disease group could be a result of a low-frequency hearing loss caused by Meniere's disease combined with a high-frequency hearing loss due to aging or other environmental factors. However, 27% (13/48) of the peaked audiograms in ears with Meniere's disease were found in persons with no hearing loss in the high frequencies of their nondiseased ear. Thus, the peaked audiometric configuration is associated with Meniere's disease, as suggested by research results from animals with experimentally induced endolymphatic hydrops. Peaked audiometric configurations were also observed in roughly 9% of the general clinical population and in 12.5% of ears of persons with acoustic tumors. A rulebased on the audiometric configurations that maximized identification of patients with Meniere's diseaseusing the general clinical population as a control did not fare as well when the tumor group was used as a control. This finding provides further evidence that peaked audiograms, although common to Meniere's disease, are not pathognomonic of this disease. Furthermore, these results indicate that the diagnosis of Meniere's disease from audiometric profiles is risky, at best.
KEY WORDS: Meniere's disease, acoustic tumor, audiogram, differential diagnosis, audiometric configurations
Submitted on August 31, 1998
Accepted on March 12, 1999