Objective: The primary feature of adductor laryngeal dystonia (AdLD) is involuntary spasm of the laryngeal adductor muscles, with more spasms during production of voiced sounds than voiceless sounds. Co-occurring secondary symptoms of compensatory laryngeal tension and associated vocal effort are also common. Relative fundamental frequency (RFF) is an acoustic measure of small-scale changes in fundamental frequency during the transitions between voiced and voiceless sounds and can be used as an estimate of laryngeal tension. The purpose of this study was to investigate whether increased laryngeal tension in speakers with AdLD fluctuates with voicing contexts similar to the characteristic primary spasms.
Methods: Thirty-eight individuals with AdLD (33 female; aged 25–80 years) read eight sentences embedded with voiced-unvoiced-voiced tokens; four sentences were comprised of predominantly voiced sounds and the others of predominantly voiceless sounds. Participants were recorded when their voices were most symptomatic. RFF was calculated for the embedded tokens in all sentences. Sentences were manually labeled for acoustic discontinuities associated with AdLD (creak, phonatory break, frequency shift).
Results: A repeated-measures analysis of variance (ANOVA) tested the effects of condition (voiced- or voiceless-loaded sentence), onset or offset of voicing, RFF cycle (1–10), and their interactions on RFF (ST). As expected, there were medium-to-large significant effects of the voice onset/offset (p < .001), voice cycle (p < .001), and their interaction (all p < .001) on RFF values.
Percent acoustic discontinuity significantly differed by condition (Welch’s ANOVA: F = 32.03, p < .001), with more acoustic discontinuities in the voiced than in the voiceless condition. However, there was no significant effect of condition or the interaction of condition × onset/offset × cycle on RFF. Although a significant interaction between onset/offset and condition was observed (p = .002), the effect size was negligible (η2p = .01).
Conclusion: There was no significant effect of voicing context on RFF, suggesting that increased laryngeal tension in speakers with AdLD could be a stable symptom that does not fluctuate in parallel with the task-dependent primary spasms.