Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Nov 12;24(1):794.
doi: 10.1186/s12888-024-06253-6.

The voice of depression: speech features as biomarkers for major depressive disorder

Felix Menne  1 , Felix Dörr  2 , Julia Schräder  3   4 , Johannes Tröger  2 , Ute Habel  3   4 , Alexandra König  2   5 , Lisa Wagels  3   4
Affiliations

The voice of depression: speech features as biomarkers for major depressive disorder

Felix Menne et al. BMC Psychiatry. .

Abstract

Background: Psychiatry faces a challenge due to the lack of objective biomarkers, as current assessments are based on subjective evaluations. Automated speech analysis shows promise in detecting symptom severity in depressed patients. This project aimed to identify discriminating speech features between patients with major depressive disorder (MDD) and healthy controls (HCs) by examining associations with symptom severity measures.

Methods: Forty-four MDD patients from the Psychiatry Department, University Hospital Aachen, Germany and fifty-two HCs were recruited. Participants described positive and negative life events, which were recorded for analysis. The Beck Depression Inventory (BDI-II) and the Hamilton Rating Scale for Depression gauged depression severity. Transcribed audio recordings underwent feature extraction, including acoustics, speech rate, and content. Machine learning models including speech features and neuropsychological assessments, were used to differentiate between the MDD patients and HCs.

Results: Acoustic variables such as pitch and loudness differed significantly between the MDD patients and HCs (effect sizes 𝜼2 between 0.183 and 0.3, p < 0.001). Furthermore, variables pertaining to temporality, lexical richness, and speech sentiment displayed moderate to high effect sizes (𝜼2 between 0.062 and 0.143, p < 0.02). A support vector machine (SVM) model based on 10 acoustic features showed a high performance (AUC = 0.93) in differentiating between HCs and patients with MDD, comparable to an SVM based on the BDI-II (AUC = 0.99, p = 0.01).

Conclusions: This study identified robust speech features associated with MDD. A machine learning model based on speech features yielded similar results to an established pen-and-paper depression assessment. In the future, these findings may shape voice-based biomarkers, enhancing clinical diagnosis and MDD monitoring.

Keywords: Depression; Machine learning; Precision psychiatry; Speech biomarkers.

PubMed Disclaimer

Conflict of interest statement

Declarations Ethics approval and consent to participate The study was approved by the internal ethics committee of the university hospital Aachen, Germany (Ethik-Kommission an der Medizinischen Fakultät der RWTH Aachen; vote number EK 045 − 19). All study procedures were conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to participation in the study. Consent for publication Not applicable. Competing interests FM, FD, JT and AK are employees of ki: elements GmbH. JS, LW and UH have no competing interests to declare.

Figures

Fig. 1
Fig. 1
Lollipop plots depicting results of Spearman rank sum correlations (r) between speech features and the BDI-II. Variables are colour-coded according to the captions, based on the categories defined in Table 1. For brevity, only features with adj. p < 0.05 are depicted here. All results including effect sizes and p values can be found in Supp. Table 5
Fig. 2
Fig. 2
Receiver operating characteristic curves and areas under the curve (ROC-AUC) for HC vs. Depressed classification models. Legend: HC = healthy controls; baseline = linear model consisting of demographic and clinical data; all_features = SVM speech model; bdi_control = SVM model with BDI-II as the only variable
Fig. 3
Fig. 3
Receiver operating characteristic curves and areas under the curve (ROC-AUC) for Mild vs. Moderately Depressed classification models. Legend: baseline = Decision Tree Model consisting of demographic and clinical data; all_features = Speech Linear Model; bdi_control = Extra Trees model with BDI-II as the only variable
See this image and copyright information in PMC

References

    1. de la Torre JA, Vilagut G, Ronaldson A, Serrano-Blanco A, Martín V, Peters M, et al. Prevalence and variability of current depressive disorder in 27 European countries: a population-based study. Lancet Public Health. 2021;6(10):e729–38. - PMC - PubMed
    1. Greenberg PE, Fournier AA, Sisitsky T, Simes M, Berman R, Koenigsberg SH, et al. The economic burden of adults with major depressive disorder in the United States (2010 and 2018). PharmacoEconomics. 2021;39(6):653–65. - PMC - PubMed
    1. IsHak WW, Mirocha J, James D, Tobia G, Vilhauer J, Fakhry H, et al. Quality of life in major depressive disorder before/after multiple steps of treatment and one-year follow-up. Acta Psychiatr Scand. 2015;131(1):51–60. - PMC - PubMed
    1. Iancu SC, Wong YM, Rhebergen D, van Balkom AJLM, Batelaan NM. Long-term disability in major depressive disorder: a 6-year follow-up study. Psychol Med. 2020;50(10):1644–52. - PubMed
    1. Marx W, Penninx BWJH, Solmi M, Furukawa TA, Firth J, Carvalho AF, et al. Major depressive disorder. Nat Rev Dis Primer. 2023;9(1):1–21. - PubMed

LinkOut - more resources