Psychiatry Investig.  2024 Aug;21(8):822-831. 10.30773/pi.2023.0417.

Exploring the Relationships Between Antipsychotic Dosage and Voice Characteristics in Relation to Extrapyramidal Symptoms

Affiliations
  • 1Department of Brain and Cognitive Sciences, College of Natural Sciences, Seoul National University, Seoul, Republic of Korea
  • 2Department of Neuropsychiatry, Seoul National University Bundang Hospital, Seongnam, Republic of Korea
  • 3Music and Audio Research Group, Graduate School of Convergence Science and Technology, Seoul National University, Seoul, Republic of Korea
  • 4Department of Psychiatry, College of Medicine, Seoul National University, Seoul, Republic of Korea

Abstract


Objective
Extrapyramidal symptoms (EPS) are common side effects of antipsychotic drugs. Despite the growing interest in exploring objective biomarkers for EPS prevention and the potential use of voice in detecting clinical disorders, no studies have demonstrated the relationships between vocal changes and EPS. Therefore, we aimed to determine the associations between voice changes and antipsychotic dosage, and further investigated whether speech characteristics could be used as predictors of EPS.
Methods
Forty-two patients receiving or expected to receive antipsychotic drugs were recruited. Drug-induced parkinsonism of EPS was evaluated using the Simpson-Angus Scale (SAS). Participants’ voice data consisted of 16 neutral sentences and 2 second-long /Ah/utterances. Thirteen voice features were extracted from the obtained voice data. Each voice feature was compared between groups categorized based on SAS total score of below and above “0.6.” The associations between antipsychotic dosage and voice characteristics were examined, and vocal trait variations according to the presence of EPS were explored.
Results
Significant associations were observed between specific vocal characteristics and antipsychotic dosage across both datasets of 1–16 sentences and /Ah/utterances. Notably, Mel-Frequency Cepstral Coefficients (MFCC) exhibited noteworthy variations in response to the presence of EPS. Specifically, among the 13 MFCC coefficients, MFCC1 (t=-4.47, p<0.001), MFCC8 (t=-4.49, p<0.001), and MFCC12 (t=-2.21, p=0.029) showed significant group differences in the overall statistical values.
Conclusion
Our results suggest that MFCC may serve as a predictor of detecting drug-induced parkinsonism of EPS. Further research should address potential confounding factors impacting the relationship between MFCC and antipsychotic dosage, possibly improving EPS detection and reducing antipsychotic medication side effects.

Keyword

Diagnosis; Parkinsonism; Drug side effects; Antipsychotics
Full Text Links
  • PI
Actions
Cited
CITED
export Copy
Close
Share
  • Twitter
  • Facebook
Similar articles
Copyright © 2024 by Korean Association of Medical Journal Editors. All rights reserved.     E-mail: koreamed@kamje.or.kr