Acoustic Characteristics of Vocal Tremor in Adductor Laryngeal Dystonia

Department of Otolaryngology - Head and Neck Surgery,
Boston University School of Medicine
Boston, MA

Abstract:

Patients with adductor spasmodic dysphonia (ASD) have an associated vocal tremor. How-ever, little is known about the prevalence or acoustic nature of the tremor in these patients. This study aimed to acoustically define vocal tremor in ASD patients. Voice recordings were analyzed for 30 ASD patients and 60 control group subjects (without vocal or neurologic pathology). Ab-normal vocal tremor was found in 66.7% of the ASD patients and the mean intensity tremor rate was significantly higher than for the ASD group. No significant difference in F0 tremor rates was found. Tremor extent for intensity was significantly higher and significant variability of the signal around the fundamental was noted within the ASD group.

Introduction:

Laryngeal dystonia is a devastating voice disorder of essentially unknown etiology. Current literature offers several etiologies, symptoms and treatments of/for laryngeal dystonia since it was first described in 1871 as a form of "nervous hoarseness".⁽¹⁾

For most of its clinical history, adductor spasmodic dysphonia (ASD) was considered a psy-chogenic disorder because of the apparent absence of structural or mucosal abnormalities of the vo-cal folds.⁽²⁾ Subjective classifications have sub-classified ASD and abductor types^(19)(2) and de-scribed them with strain-strangled voice, excessive tension, respiratory exertion, and voice breaks during speech but not in vegetative tasks. Alternatively, it has been noted that normal speech may accompany laughter, intoxication, and states of joy and anger.⁽⁴⁾

Electromyographic data indicated that vocal strain represented selective hyperactivity of the adductor musculature, with abnormal resting electromyographic levels in the thyroarytenoid and cricothyroid muscles. ⁽⁵⁾ Features such as enlarged, attenuated or polyphasic potentials, denerva-tion patterns, pseudomytonic discharges, reduced number of participating motor units, breaks in vo-litional activity and regular or irregular tremor electromyographic patterns have been reported in pa-tients with ASD, and provide a contradictory picture.⁽⁶⁾⁽⁷⁾ These highly deviant patterns were ab-sent in falsetto and whisper, as well as respiratory and valsalva tasks, the latter requiring forceful vocal fold collision.

Therefore, in humans, the electromyographic evidence appears to argue for an "imbalance" between the thyroarytenoid and cricothyroid muscles, more likely implicating the thyroarytenoid. The imbalance was thought to create excessive adduction and shortening of the vocal folds during phonation, resulting in strained closure leading to significantly increased subglottic pressure.⁽⁷⁾⁽⁸⁾

There is, however, a growing recognition that a substantial portion of spasmodic phonatory dysfunction is accompanied by tremor. ⁽²⁾⁽⁹⁾ A retrospective review of clinical records found an association of ASD and limb essential tremor, leading authors to propose that vocal arrests in ASD are a form of enhanced essential tremor.^(10)(11)

The chief obstacle to the study of ASD and tremor has been the means of objectively meas-uring tremor and its characteristics. ⁽⁶⁾ There have been attempts to document tremulous move-ments with videoendoscopic studies of the larynx during phonation and respiration. ⁽¹²⁾ Davis, Boone, Carroll et. al., showed tremor to occur at 5 Hz with the anterior-to-posterior axis appearing to be most affected. Their article suggested that normal proprioceptive control of arytenoid move-ments, while not fundamentally impaired, were unable to cope with a constantly moving baseline of muscle tone. Their observation, combined with electromyographic and airflow studies ⁽⁵⁾ demon-strate that tremor is a constant feature of ASD. Therefore, the physiologically based, acoustic output signal (i.e., voice) should contain information about the phonatory system in ASD patients. Previ-ous studies used acoustic analysis to identify tremor in ASD patients. However, early analysis may not have been refined enough, and may have involved too small of a population to extract low fre-quency variability from sustained phonations.

The objective of this study was to demonstrate that acoustic tremor, if present, could be ex-tracted from the signal of patients diagnosed with ASD.