The association between anhedonia and speech features in depression: A cross-sectional study

Abstract

Background

Anhedonia is a core symptom of depression. Most anhedonia assessments rely on self-reporting, which does not accurately reflect hedonic capacity and is biased by individual subjectivity. Therefore, objective indicators are needed. Anhedonia may result in different speech features among depressive patients. Thus, speech features have become an emerging objective indicator in depression assessment. This study aims to investigate the relationship between anhedonia and speech features in individuals with depression by comparing the speech features of patients with and without anhedonia in a multitasking paradigm.

Methods

A total of 166 patients with depression were recruited for the study. Voice data were collected through the Verbal Fluency Test, Word reading, Video description, and Semi-structured Interviews. The primary analysis was performed using analysis of covariance (ANCOVA) and partial correlation analysis. We grouped patients based on the severity of depression and performed post-hoc t-tests or Mann-Whitney U tests for comparisons. The Benjamini-Hochberg method was used to control for False Discovery Rate in multiple comparisons.

Results

After adjustment for anxiety severity (as measured by the 7-item Generalized Anxiety Disorder Scale, GAD-7), no significant differences in speech features were observed between patients with or without anhedonia. Similarly, after controlling for depression severity (as measured by the 17-item Hamilton Depression Scale, HAMD-17), no significant correlation was found between speech features and the degree of anhedonia. Post hoc analyses showed that seventeen speech features were correlated with depression severity (|r| < 0.3, small effect sizes), but no differences in speech features were found between patients with anhedonia and those without anhedonia within each subgroup based on depression severity.

Conclusion

Speech features do not differ significantly between patients with or without anhedonia at any level of depression severity. However, speech features were independently correlated with depression severity. Future studies may refine research methodology by optimizing speech task modules or assessing multidimensional prosodic features for more in-depth analysis.