This page was translated automatically from Turkish. Read the original

Bu sayfa, 26. Dünya Psikiyatri Kongresi'nde sunulan çalışmanın yapay zekâ desteğiyle hazırlanmış Türkçe özetidir. Çalışmanın sahibi sunan yazarlardır; atıf için orijinal kaynağa başvurun.

SO038

Voice-based digital biomarkers in attention-deficit/hyperactivity disorder: a meta-analysis of pitch-related acoustic features

VOICE BASED DIGITAL BIOMARKERS IN ATTENTION DEFICIT HYPERACTIVITY DISORDER: A META-ANALYSIS OF PITCH RELATED ACOUSTIC FEATURES

Pitch-related acoustic measures are significantly lower in children and adolescents with attention-deficit/hyperactivity disorder than in controls (SMD −0.56; p = 0.004).

The study aimed to quantify differences in pitch-related acoustic measures between individuals with attention-deficit/hyperactivity disorder (ADHD) and controls. Searches of Cochrane, PubMed, Scopus, Embase, and Web of Science identified 471 records for screening; after 278 duplicates were removed, 193 records remained. Based on titles and abstracts, 270 records were excluded. Of the 8 studies assessed in full text, 5 were excluded because they did not include the relevant outcome measure, and 3 observational studies involving children and adolescents were included. As one study provided two independent age-defined subgroups, a total of 4 datasets were available. The analysis was conducted in RStudio 2025.05.0 using an inverse-variance approach with a multilevel random-effects model and restricted maximum likelihood (REML) estimation. A total of 109 participants (52 with ADHD and 57 controls) were examined, with individual effect sizes ranging from −0.41 to −1.02. The pooled analysis showed a significant reduction in pitch-related acoustic measures in ADHD: standardized mean difference (SMD) −0.56, 95% confidence interval −0.95 to −0.18 (−0.95 to −0.17 in the table accompanying the forest plot), p = 0.004 (Z = −2.84). No heterogeneity was detected (chi-square = 1.21, df = 3, p = 0.7575; I² approximately 0%), but statistical power was limited by the small number of studies. According to the authors, the findings support the potential of voice features as candidate digital biomarkers, while speech-based phenotyping offers opportunities for objective assessment and scalable monitoring but requires further validation.

Tüm WCP 2026 özetleri