Abstract
Background: Semi-occluded vocal tract exercises (SOVTE) improve vocal quality and capacity. Water resistance therapy (WRT), a specific form of SOVTE with a tube submerged under water, generates increased and oscillating oral pressure through bubble formation during phonation, thereby influencing transglottal pressure and vocal fold dynamics. While the physiological effects of WRT using tube-based systems have been extensively studied, the influence of vowel-specific vocal tract configurations during WRT remains unclarified. This study examined how different vowel qualities during WRT affect vocal fold oscillation using the DoctorVox(®) mask, which allows near-natural mouth opening and vowel articulation. Methods: Ten vocally healthy, untrained adults (25-50 years) performed a continuous vowel glide (/i/-/a/-/u/-/i/) at constant fundamental frequency and habitual loudness during WRT using the DoctorVox(®) mask, with the tube submerged 2 cm in water. Simultaneous recordings included transnasal high-speed videoendoscopy (20,000 fps), electroglottography (EGG), acoustic signals and intra-tube oral pressure measurements. Glottal area waveforms (GAW) were derived to calculate the open quotient (OQ(GAW)) and closing quotient (ClQ(GAW)). Analyses were conducted separately for intra-tube pressure maxima, minima and intermediate phases within the bubble cycle during WRT. Statistical analysis used Wilcoxon signed-rank tests with Bonferroni correction. Results: In the baseline condition without WRT, significant vowel-related differences were found: /u/ showed a higher open quotient than /i/ and /a/ (p < 0.05) and a higher closing quotient than /a/ (p < 0.05). During WRT, these vowel-specific differences were no longer statistically significant. A non-significant trend toward reduced OQ(GAW) during WRT was observed, most notably for /u/, while differences between pressure phases within the bubble cycle were minimal. Conclusions: WRT using the DoctorVox(®) mask reduces vowel-specific differences in vocal fold vibration patterns, suggesting that for voice therapy, vowel quality modifications during WRT have little impact on vocal outcomes.