Acoustics Features of Voice in Adolescent Females With Anorexia Nervosa

The process of human voice production is a complex physiological and acoustic phenomenon that depends on many structural, physical, and hormonal factors, systemic diseases as well as emotional states. All these factors can be present in eating disorders. However, studies on eating disorders and voice problems have usually been evaluated in terms of bulimia. Chronic starvation and emotional problems in the course of anorexia nervosa (AN) appear to be under-researched, despite various biochemical, metabolic, and hormonal changes.

The purpose of this study was to evaluate voice quality, specifically acoustic analysis, in adolescent female with AN from the point of view of the possible influence on the function and structure of the larynx, low body mass accompanying AN, as well as energy deficiency, hormonal and emotional disturbances.

A total of 84 girls diagnosed with AN (Gr.A) (15.32 years, SD = 2.12; range 12-19, BMI = 14.11 ± 1.72) were assessed for the condition of the voice such as perceptual voice evaluation on the GRBAS scale, maximal phonation time (MPT), laryngoscopy, with special attention to voice acoustic analysis - Multi-Dimensional Voice Program (MDVP). The control group (Gr.C) included 62 girls without eating disturbances (aged 12-19, mean age 15.41 ± 2.40, BMI = 21.60 ± 1.92). Perceptual voice assessment, aerodynamic test MPT, and acoustic parameters were analyzed according to girls' age.

Total GRBAS scale was higher in girls with AN compared to the control group mainly for two parameters: breathiness (B) (P = 0.00015) and asthenia (A) (P < 0.05). The MPT for Gr.A was significantly shorter compared to Gr.C (15.40 ± 3.51 seconds vs. 23.19 ± 5.17 seconds) (P < 0.001), and a correlation of MPT values with the age of the adolescent female was observed: Spearman's coefficient for Gr.A = (-)0.5378, for Gr.C = 0.5516 (P = 0.0012). Acoustic analysis revealed the decrease in the basic frequency F0 in Gr.A compared to Gr.C (231.08 Hz vs. 242.30 Hz), and narrowing of the voice scale was observed, resulting mainly from a reduction in the upper limit. Significant differences were found for measures of frequency perturbations (Jita, Jitter, RPA, PPQ, sPPR), with Gr.A scoring significantly higher than Gr.C (P < 0.05 for all). Significant changes in voice acoustic analysis parameters were found with age. Negative correlations were found for measures of F0 for Gr.A to a much greater extent compared to Gr.C. Positive correlations were found with measures of tremor assessment (SPI, FTRI, ATRI) for Gr.A.