Modeling voice production and self-perception in noise: Understanding the Lombard effect in non-phonotraumatic vocal hyperfunction

The sensorimotor adaptation process is crucial for maintaining oral communication. Recent studies have shown that individuals with non-phonotraumatic vocal hyperfunction (NPVH) experience difficulties in sensorimotor adaptation when speaking in noise (known as the Lombard effect). However, the role of auditory and somatosensory feedback in the dynamics of adaptation to speaking in noise is still unclear. In this study, the use of a simple three-parameter mathematical model, known as SimpleDIVA model, was extended to explore the adaptation dynamics of speaking in noise among a group of participants with typical voices and NPVH. All participants were asked to utter a series of syllables under three conditions: baseline (quiet environment), Lombard (speech-shaped noise at 80 dB), and recovery (quiet environment after 5 min of rest). The results indicate that participants with NPVH did not return to baseline after exposure to speaking under noise. The SimpleDIVA model analysis reveals a diminished feedforward learning rate and reduced somatosensory feedback gain in participants with NPVH in comparison to participants with typical voices. This suggests that participants with NPVH may be using less somatosensory information when speaking in noise and may require more time to update the feedforward commands during and after speaking in noise.