Voice features in paediatric ADHD: baseline group differences and longitudinal change under methylphenidate

Abstract Diagnosis and treatment monitoring of attention-deficit/hyperactivity disorder (ADHD) largely rely on subjective assessments, highlighting the need for objective markers. Voice features and speech embeddings represent promising candidates, as they may offer an objective, scalable and clinically valuable complement to current practice. Yet it remains largely unclear which features are most informative for distinguishing ADHD and monitoring treatment effects, and which speech tasks best elicit these differences. Twenty-seven children with ADHD and 27 age-matched neurotypical controls completed six speech tasks across two study visits. Children with ADHD were unmedicated at baseline and were assessed under prescribed methylphenidate treatment at follow-up, whereas controls underwent repeated assessment without intervention. Established acoustic voice features and high-dimensional speech embeddings were extracted and analysed using linear mixed models to examine baseline group differences and group-by-time interaction effects reflecting medication-associated change patterns. At baseline, children with ADHD showed a lower and more variable pitch (F0 mean, F0 variability), altered spectral properties (MFCC1/MFCC2 variability), and reduced rhythmic stability (variability of voiced-segment length) compared to controls. Group-by-time interaction effects indicated medication-associated modulation in the ADHD group, including reduced loudness variability and normalization of first formant bandwidth toward control values. Speech embeddings (Whisper but not WavLM) showed FDR-significant baseline and interaction effects. Free speech tasks, particularly picture description, yielded the most robust and consistent effects. These findings support further investigation of speech-based measures as potential digital biomarkers in ADHD, with picture description emerging as a particularly promising task for future clinical assessment protocols. Clinical trial registration: German Clinical Trials Registry, registration no. DRKS00030766, Registration Date: 12 December 2022. Available: https://drks.de/search/en/trial/DRKS00030766