Influence of Posture, Viscosity, and Neck Circumference on Swallowing Acoustics: A Mixed-Effects Analysis

Neck-surface swallowing acoustics may enable accessible monitoring of swallowing, yet intensity-derived features can be influenced by both experimental manipulations and anatomical sound transmission. This study aimed to examine the effects of head-neck posture, bolus viscosity, and neck circumference on swallowing-sound duration and uncalibrated mean acoustic intensity recorded using a neck-mounted contact microphone. Sixty healthy young adults (30 female, 30 male; 18-35 years) completed four standardized 10-mL swallows comprising thin-liquid and pudding-thick swallows in neutral and chin-tuck head-neck postures (International Dysphagia Diet Standardisation Initiative [IDDSI] Levels 0 and 4). Swallow sounds were segmented in Praat to extract swallowing-sound duration and mean acoustic intensity (dB; uncalibrated; Praat-derived). Linear mixed-effects models with a participant-level random intercept tested head-neck posture, viscosity, and their interaction for both acoustic outcomes; neck circumference was included as a continuous covariate in the intensity model. Mean acoustic intensity was higher in the chin-tuck than in the neutral head-neck posture (B = 5.3 dB, p < 0.001), and a significant posture × viscosity interaction was observed (B = - 2.8 dB, p = 0.030), indicating a smaller chin-tuck-related acoustic-intensity increase for pudding than for thin liquid. Neck circumference was negatively associated with mean acoustic intensity (B = - 1.0 dB/cm, p = 0.024). Swallowing-sound duration showed a small posture-related increase (B = 0.02 s, p = 0.031); however, this effect was not robust after adjustment for multiple comparisons, and no posture × viscosity interaction was observed. Head-neck posture and viscosity systematically influenced surface-recorded mean acoustic intensity, and neck circumference accounted for meaningful between-participant variability consistent with soft-tissue attenuation. Task conditions and anatomical covariates should therefore be considered when interpreting uncalibrated acoustic-intensity measures obtained from cervical auscultation sensors.