Characterizing respiratory aerosol emissions during sustained phonation

To elucidate the role of phonation frequency (i.e., pitch) and intensity of speech on respiratory aerosol emissions during sustained phonations.

Respiratory aerosol emissions are measured in 40 (24 males and 16 females) healthy, non-trained singers phonating the phoneme /a/ at seven specific frequencies at varying vocal intensity levels.

Increasing frequency of phonation was positively correlated with particle production (r = 0.28, p < 0.001). Particle production rate was also positively correlated (r = 0.37, p < 0.001) with the vocal intensity of phonation, confirming previously reported findings. The primary mode (particle diameter ~0.6 μm) and width of the particle number size distribution were independent of frequency and vocal intensity. Regression models of the particle production rate using frequency, vocal intensity, and the individual subject as predictor variables only produced goodness of fit of adjusted R² = 40% (p < 0.001). Finally, it is proposed that superemitters be defined as statistical outliers, which resulted in the identification of one superemitter in the sample of 40 participants.

The results suggest there remain unexplored effects (e.g., biomechanical, environmental, behavioral, etc.) that contribute to the high variability in respiratory particle production rates, which ranged from 0.2 particles/s to 142 particles/s across all trials. This is evidenced as well by changes in the distribution of participant particle production that transitions to a more bimodal distribution (second mode at particle diameter ~2 μm) at higher frequencies and vocal intensity levels.