Voice Characteristics of Individuals with Dementia Due to Suspected Alzheimer’s Disease: Three Case Studies

Abstract

Previous studies have described communication changes to the speech and language of individuals with Alzheimer’s Disease (AD); however, other facets of communication, such as voice, have been left largely under-researched in this population. Perceived voice quality changes have not been reported as part of the disease process and reports of acoustic or aerodynamic correlates to vocal pitch, loudness, and quality are limited. The present study is a series of three case reports with the aim of gathering a preliminary dataset that can be used to identify key measures or differences that should be measured in a larger study. To this end, voice evaluation was completed in three people with suspected AD and results were compared to a control group of age- and gender- matched adults with typical cognition. Measures included a patient-reported quality of life scale, perceptual ratings of voice and speech, videostroboscopic ratings, cepstral peak prominence (CPP), low/high spectral ratio (L/H Ratio), mean fundamental frequency (fo), maximum phonational frequency range, mean airflow, and maximum inspiratory pressure (MIP). Co-variates included medical history and frailty screening. All three participants reported dysphonia. Endoscopic imaging for Participant 2 showed a bulky lesion of the right vocal fold. This lesion is likely the primary source of his dysphonia rather than AD. Imaging could not be obtained for participant 1 (P1) or 3 (P3). P1’s voice was rougher and breathier than the controls and P3’s voice was more severely impaired (overall) and breathier than controls. Acoustic results included normal CPP and low L/H Ratio for P1 and P3. Aerodynamic results included lower than expected MIP for all three participants and elevated mean airflow only for P3. These results provide preliminary evidence that disordered voice quality can exist in people with AD. It is clear that there are many possible causes of voice disorder that will need to be considered when designing the larger study. These include medical co-morbidities, age, frailty, cognitive and general motor status, hearing loss, and frequent interaction with people with hearing losses. While L/H Ratio was more sensitive to perceived voice quality differences than CPP for these participants, the measure should be retained in future studies. A more natural speech sample is also recommended to eliminate potential effects of reading on speech production and to assess prosodic changes in typical conversation.