The assessment of the impact of anorexia nervosa on the vocal apparatus in adolescent girls - A preliminary report

Unveiling vocal profiles in adolescent anorexia nervosa: a Software Based, Multiparametric Analysis

Dysphonia, characterized by disturbances in voice quality and modulation, has been sporadically observed in individuals with Anorexia Nervosa (AN), potentially stemming from both organic and psychopathological factors. This study seeks to employ software-based voice analysis to compare the voices of girls with AN to those of female healthy controls (HC). Case-control study adopting "Praat" software to assess voices. Various parameters, including Acoustic Voice Quality Index (AVQI), Fundamental Frequency (F0), Yanagihara's Spectrographic Dysphonia Classifications, and "GIRBAS" perceptual qualitative voice rating, were investigated. Participants completed questionnaires for Vocal Fatigue Index (VFI) and the Reflux Symptoms Index (RSI). Puberty-related voice spectrum changes were considered, and Bonferroni-corrected BMI-adjusted Analyses of Covariance (ANCOVAs) were conducted. The study enrolled 15 girls with AN and 23 girls with HC. AN patients demonstrated greater impairment in voice tiredness/voice avoidance (VFI-1, p < 0.001), vocal physical discomfort (VIF-2, p = 0.002), and rest as alleviation (VFI-3, p = 0.012). Reflux-related scores were higher in AN (p < 0.001). Differences were observed in voice quality (AVQI) (p = 0.001), and GIRBAS scales showed alterations in multiple parameters. Spectrograms documented more frequent pathological findings in AN patients (p = 0.021). No difference was observed in Fundamental Frequency. These group (AN/HC) differences were independent of weight measures. This study is the first to connect voice irregularities in AN by employing standardized, non-invasive tools and accounting for weight-related factors. Young AN patients demonstrated substantial voice quality changes and heightened self-reported symptoms. Future research should expand on these findings with prospective designs and invasive investigations.

Effects of Age-Dependent Hormonal Changes and Estrogen Supplementation on Voice in Girls with Anorexia Nervosa-Preliminary Report

INTRODUCTION

Human development includes lots of physical and emotional changes. The human voice depends on age. Voice production is a complex physiological and acoustic phenomenon that depends on many factors such as structure, hormone level, degree of fatigue or nutrition and hydration of the body, systemic diseases, and emotional state. All these factors can be present in anorexia nervosa (AN), such as excessive weight loss, generated hydro-electrolytic changes, nutritional deficiencies, hormonal disturbances in the function of the hypothalamic-pituitary-adrenal axis, the hypothalamic-pituitary-thyroid axis, the hypothalamic-pituitary-ovarian axis, and emotional distress. The prevalence of AN ranges between 0.3% and 3%, and it is the third most common chronic disease affecting adolescent girls. However, voice changes related to AN have not been fully investigated.

OBJECTIVE

The purpose of this study was to evaluate the impact of AN on age-related changes in the voice of adolescent women-before and after puberty, particularly through acoustic analysis. An additional objective was to evaluate estrogen substitution in female patients with AN in order to investigate their effect on voice condition.

MATERIALS AND METHODS

126 girls diagnosed with AN (15.32 ± 2.12 years, range 12-19, BMI = 14.38 ± 1.67), 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 (B) included 93 girls without eating disturbances (aged 12-19, mean age 15.52 ± 2.40, BMI = 21.50 ± 1.54). Perceptual voice assessment, aerodynamic test MPT, and acoustic parameters were analyzed in age groups (≤16 years and >16 years). The human vocal tract is sensitive to sex hormones, so the analysis was carried out in the group up to the age of 16 and above 16 to check possible effects.

RESULTS

GRBAS scale was higher in girls with AN compared to the control group for breathiness (B) (P = 0.0002) and asthenia (A) (P < 0.05). The median GRBAS scale for the older group of anorexic women was the highest (2.0). The mean MPT for group A was significantly lower (15.40 ± 3.51 seconds). Comparing age subgroups there was a prolongation of MPT in the healthy group (in groups ≤16 years and >16 years respectively 21.13 seconds versus 25.40 seconds) and a shortening in the anorectic group (≤16 years versus >16 years: 17.06 seconds versus 14.17 seconds). There was no difference between groups A and C up to 16 years of age, but above 16 years of age appeared (14.17 seconds versus 25.40 seconds). Acoustic analysis revealed lower F0 values in group A and C in older subgroups (215,85 Hz versus 236,01 Hz-statistically significant), as well as between subgroups both groups (A: 251,38 Hz versus 215,85 Hz; C: 248,20 Hz versus 236,01 Hz). A narrowing of the vocal range in girls over 16 years in group A was observed. There were no statistically significant differences in F0 between subgroups ≤16 years in groups A and C (251.38 Hz versus 248.20 Hz). The ENT study found that more than half of the girls (54.55%) over the age of 16 who took hormone supplementation manifested laryngeal structure that was normal for their age, there was no effect of hormone supplementation on any of the MDVP parameters between the drug-taking and non-drug-taking groups.

CONCLUSIONS

The acoustic results of the voice in MDVP measurements in adolescent women with AN are not within the normal range and do not mimic the normal developmental changes of the voice. The most important acoustic characteristics of the voice are changes in the fundamental frequency F0 and the range of the voice tended to be more severe in anorectic women >16 years of age and to increase with age, indicating a possible cumulative effect of malnutrition-related disorders as well as hormonal dysfunctions. MDVP can be considered a simple, non-invasive method of assessing the voice organ in AN. MPT differentiated the study groups well: statistically significant differences were noted both between the groups, as well as between age groups. There was no significant effect of oral hormone supplementation on any parameters of the voice. In conclusion, body mass and fat volume in AN may be related to voice production/physiology, affecting voice quality, voice acoustic parameters, voice aerodynamics, and phonatory range in an age-dependent manner. Future studies are needed to assess the long-term efficacy of estrogen treatment in AN.

The Relation Between Eating Disorders and Voice Disorders

OBJECTIVE

The purpose of this study is to determine the relation between specific eating disorder diagnoses/purging behaviors and voice disorders.

METHOD

One hundred-nine participants with eating disorders completed a survey inquiring about eating disorder symptoms, purging behaviors, and voice disorder symptoms. Participants also completed the Eating Disorders Examination Questionnaire, Voice Handicap Index, and the Reflux Symptom Index.

RESULTS

The prevalence of voice disorders among the group with eating disorders was 21.88%. Of those with both eating disorders and voice disorders, anorexia nervosa appeared to be more prevalent in this group than bulimia nervosa. In addition, purging behaviors of exercise presented with a higher prevalence of voice problems than vomiting.

CONCLUSION

Individuals with eating disorders seem to be at a higher risk for voice disorders than the general population. Anorexia nervosa and exercise as a purging method were identified as the highest risk factors for voice disorders.

Automated assessment of psychiatric disorders using speech: A systematic review

OBJECTIVE

There are many barriers to accessing mental health assessments including cost and stigma. Even when individuals receive professional care, assessments are intermittent and may be limited partly due to the episodic nature of psychiatric symptoms. Therefore, machine-learning technology using speech samples obtained in the clinic or remotely could one day be a biomarker to improve diagnosis and treatment. To date, reviews have only focused on using acoustic features from speech to detect depression and schizophrenia. Here, we present the first systematic review of studies using speech for automated assessments across a broader range of psychiatric disorders.

METHODS

We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. We included studies from the last 10 years using speech to identify the presence or severity of disorders within the Diagnostic and Statistical Manual of Mental Disorders (DSM-5). For each study, we describe sample size, clinical evaluation method, speech-eliciting tasks, machine learning methodology, performance, and other relevant findings.

RESULTS

1395 studies were screened of which 127 studies met the inclusion criteria. The majority of studies were on depression, schizophrenia, and bipolar disorder, and the remaining on post-traumatic stress disorder, anxiety disorders, and eating disorders. 63% of studies built machine learning predictive models, and the remaining 37% performed null-hypothesis testing only. We provide an online database with our search results and synthesize how acoustic features appear in each disorder.

CONCLUSION

Speech processing technology could aid mental health assessments, but there are many obstacles to overcome, especially the need for comprehensive transdiagnostic and longitudinal studies. Given the diverse types of data sets, feature extraction, computational methodologies, and evaluation criteria, we provide guidelines for both acquiring data and building machine learning models with a focus on testing hypotheses, open science, reproducibility, and generalizability.

LEVEL OF EVIDENCE

3a.