Using speech as objective markers for major depressive disorder (MDD) has shown promise, yet their generalizability across clinical settings remains largely unvalidated. This study aimed to validate previously identified speech markers of depressive symptoms in an independent clinical cohort, thereby assessing their reproducibility and robustness for cross-site application. Speech data from two independent psychiatric cohorts (RWTH Aachen and University of Oldenburg, Germany) were analyzed, comprising 135 participants (71 healthy controls, 64 MDD patients). Participants completed a positive and a negative storytelling task, over 80 temporal, lexical, and spectral speech features were extracted from the acoustic signal. Statistical analyses assessed group differences and correlations with Beck Depression Inventory (BDI-II) scores. Machine learning models trained on the Aachen data were tested on the Oldenburg cohort. Several temporal and spectral speech features, including utterance duration, pause duration, and MFCCs, were consistently associated with MDD diagnosis and symptom severity across both cohorts. Machine learning models trained on Aachen data achieved a classification accuracy (ROC-AUC) of 0.63 on the Oldenburg sample, demonstrating above-chance but modest transfer performance. Voice quality features (shimmer, jitter) showed more variable associations: partial correlations indicated some significant effects (e.g., shimmer and jitter during positive storytelling), whereas moderation analyses revealed interaction effects, particularly for shimmer and jitter in negative storytelling, where MDD patients exhibited higher values in the Aachen cohort but lower values in the Oldenburg cohort compared to healthy controls. The study indicates that temporal and spectral markers of speech are relatively robust across independent clinical samples, whereas voice quality markers (shimmer, jitter) show site-dependent inconsistencies, acting as technical artifacts of varying recording conditions rather than robust biomarkers. While current speech-based classifiers remain less accurate than established self-report measures, their integration with clinical scores offers a more balanced trade-off between sensitivity and specificity. Future work should prioritize systematic evaluation across elicitation tasks, languages, and longitudinal settings to delineate which speech features are transferable and which are task-specific.
F. Menne, Felix Dörr, J. Tröger et al.· Annals of General Psychiatry· 0 citations
Early detection of cognitive impairment is essential for dementia prevention and timely care. However, implementation in primary care and community settings remains limited. Building dementia-prepared health systems requires scalable and adaptive pathways that integrate subjective, digital and biological indicators, while accounting for heterogeneity in risk, education and age. CogScreen I was a cluster-randomized trial conducted in Munich senior centers from March 2023 to March 2024. Recruitment focused on community-dwelling adults aged ≥ 60 years reporting subjective cognitive concerns. Centers were randomized to: (A) the Subjective Cognitive Decline Questionnaire (SCD-Q) only, (B) SCD-Q plus digital cognitive testing, or (C) SCD-Q plus digital testing plus blood biomarkers (Aβ1-42/1–40, pTau181, GFAP, NfL). The primary endpoints were feasibility and acceptability, assessed through structured questionnaires and follow-up interviews with participants and general practitioners. Secondary endpoints examined latent cognitive structures and biomarker associations. Exploratory factor and clustering analyses revealed multimodal subgroups across subjective, digital, and biological measures, which informed a hypothesis-generating three-tier adaptive detection framework (low, medium, high intensity). Among the 473 participants (mean age 74.1 ± 7.6 years; 66% female; 63% with tertiary education), both feasibility and acceptability were high: Participants described the assessments as personally relevant, clearly communicated, and medically meaningful, and appreciated the added value of digital testing and biomarkers. Digital cognitive testing and biomarkers captured distinct latent dimensions of learning/working memory, psychomotor attention, and glial and amyloid pathology. Exploratory multimodal clustering identified three dementia risk profiles—low concern, intermediate, and at risk—which informed an adaptive, tiered detection model. Notably, subjective cognitive concerns assessed by the SCD-Q were strongly associated with overall subjective symptom burden, supporting the use of brief self-report questionnaires as a pragmatic and scalable first-step stratification approach in community-based dementia detection. Community-based, tiered detection pathways for cognitive decline are feasible, acceptable, and meaningful to older adults. Exploratory multimodal profiling suggests that combining subjective, digital, and biological measures may support individualized assessment strategies. However, limited GP engagement highlights the need for stronger integration with primary care to ensure downstream diagnostics and prevention pathways. The proposed framework remains hypothesis-generating and requires prospective validation. Clinical trials registered CogScreen has been registered at clinical trials (NCT06191952, 2023–12-20).
Carolin I. Kurz, Nikola Clara-Sophie Wüsten, Paulina Tegethoff et al.· Scientific Reports· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
