Acoustic, perceptual and physiological studies of ten-year-old children's voices

Abstract

The aim of the present investigation was to study hoarseness and its perceptual, acoustic and physiological characteristics in ten-year-old children's voices. Fifty-eight children's voices were perceptually evaluated along 15 voice parameters. The rank ordered means of each parameter revealed a discontinuity in the distribution for all parameters except pitch, breathiness and vocal fry. This discontinuity was used as an operational borderline between normal and deviant voice characteristics. Statistical analysis showed that hyperfunction, breathiness and roughness are the main predictors of hoarseness. Pitch and intensity ranges of 60 children were recorded in voice range profiles (VRP). Vocal fold status was determined by video-microlaryngoscopy and, when possible, also by stroboscopy. Six children, all boys had vocal nodules. Twenty-five children had incomplete glottal closure. The children had somewhat compressed VRP contours reflecting a more restricted pitch range and dynamics than adults.

The occurrence of register transitions in children's voices was studied in an experiment Five voice experts perceptually identified such transitions from a tape made of the VRP recording sessions. On this tape the stimuli occurred in ascending pitch order. One transition was identified in most voices at a mean fundamental frequency (F0) of A#4. A second transition was identified in four voices at a mean F0 of A#5. SPL and subglottal pressure were measured at different pitch and loudness levels in nine children. At phonation threshold and at normal conversational loudness the children's subglottal pressures were similar to those of adult female voices.

Acoustic correlations to perceptual voice characteristics were analyzed in two investigations For the first, six children representing different degrees of hoarseness were selected. The voice sample was running speech. Two perturbation measures were examined but no correlation was found between the perceptual evaluation and these two measures. For the second investigation 50 children were chosen. The material consisted of sustained vowels. The acoustic measures were period perturbation quotient (PPQ), amplitud perturbation quotient (APQ) and two harmonics-to-noise ratio measures, NNEa and NNEb, operating in different frequencies. The results showed that hoarseness, breathiness and roughness correlated with the acoustic measures PPQ and NNEa at the p 001 Ievel of significance. Hyperfunction and instability did not correlate with any of these measurements.

Conclusions: Hoarseness in children's voices is a stable concept consisting of three main predictors; hyperfunction, breathiness and roughness. Children generally had somewhat compressed VRP recording as compared to adults. Register transitions could be identified in most children's voices, approximately 25% higher in F0 than for adults Incomplete glottal closure may be regarded as a normal finding in ten- year-old children. Subglottal pressure values in these children's voices were similar to those of adult females. In children's sustained vowel phonation hoarseness, breathiness and roughness correlated with the acoustic measures PPQ and NNEa.