Speech, communication, and neuroimaging in Parkinson's disease : characterisation and intervention outcomes

Most individuals with Parkinson's disease (PD) experience changes in speech, voice or communication. Speech changes often manifest as hypokinetic dysarthria, a motor speech disorder that may result in restrictions in communicative participation and motivates rehabilitation. HiCommunication is a novel group intervention designed to improve speech and communication in PD, addressing challenges in retaining and transferring improvements obtained during intervention to situations outside the clinic.

The aim of this thesis was to characterise dysarthria and evaluate the efficacy of the HiCommunication intervention in individuals with mild to moderate PD. This comprehensive approach included acoustic and auditory-perceptual assessment of speech, disease-related measures, cognitive assessments, and neuroimaging. Additionally, it explored the feasibility of using real-time monitoring in a randomised controlled trial (RCT) to assess speech and voice changes in daily life after intervention.

This thesis consists of four studies that were part of the EXPANd trial, an RCT including 95 participants with mild to moderate PD. The aim was to study the link between behavioural changes obtained in intervention and possible structural and functional brain changes. Participants were assessed pre- and post- intervention using a comprehensive test battery that included recordings of speech and voice, clinical dysarthria assessment, a neuropsychological test battery, assessments of balance and motor impairment, and structural and functional magnetic resonance imaging (MRI). Six-month follow-up speech and voice recordings were also performed.

Paper I was a cross-sectional study based on data obtained from assessment pre-intervention. Auditory-perceptual listener ratings of text reading were performed by two experienced speech and language pathologists specialising in motor speech disorders. Participants were rated on overall speech impairment, articulation, voice function, and prosody, and classified into three groups: no dysarthria, mild dysarthria, and moderate dysarthria. Structural and resting-state functional MRI were analysed using Voxel-Based Morphometry and fractional Amplitude of Low Frequency Fluctuations, respectively. Non-parametric Kruskal- Wallis tests were used to examine differences between the dysarthria severity groups. Significant differences were observed in voice sound level, degree of dysphonia, and motor symptom severity. Exploratory analyses showed that dysarthria was associated with decreased regional spontaneous resting-state brain activity in the right postcentral gyrus and superior frontal gyrus, and increased activity in the right inferior occipital gyrus and cerebellum exterior/fusiform gyrus. Decreased gray matter volume in the right superior and middle temporal cortices correlated with lower scores on global cognition, language, and executive functions.

Papers II and III were longitudinal studies comparing HiCommunication to an active control intervention, utilising data from all participants across all time points. Acoustic analysis and auditory-perceptual listening assessment by experienced speech and language pathologists focused on speech domains affected by hypokinetic dysarthria: general impression, voice and breathing, articulation, and prosody. Additionally, resonance was assessed using only auditory-perceptual listening assessment. Spectral dynamic causal modeling and the parametric empirical Bayes methods were applied to resting-state functional MRI data to describe effective connectivity changes in a speech- related network. Statistical analyses of acoustic and auditory-perceptual outcomes included all participants (intention-to-treat) with imputed missing values and complementary per-protocol analyses of intervention completers. In intention-to-treat analysis, significant group-by-time interaction effects favouring HiCommunication were found post-intervention for the acoustic outcomes voice sound level in text reading and monologue, Acoustic Voice Quality Index, and Harmonics-to-Noise Ratio. HiCommunication completers also showed statistically significant improvements in the auditory-perceptual parameters imprecise articulation and vocal fry. However, these changes were not maintained at follow-up after 6 months. Immediately post-intervention, a majority of HiCommunication participants showed a clinically relevant increase in voice sound level. Effective connectivity analysis revealed a significant decrease in inhibitory self-connectivity in the left supplementary motor area and increased connectivity from the right supplementary motor area to the left paracentral gyrus after HiCommunication compared to the active control intervention.

Paper IV was a feasibility study including 23 participants from the EXPANd trial. Voice use in daily life was registered with the portable voice accumulator (PVA) VoxLog for 4-7 days before and after intervention. The VoxLog registers voice sound level, phonation time, and levels of environmental noise, allowing analysis of voice use in relation to different environmental noise levels. Process feasibility results indicated challenges in data collection due to low inclusion rates, compliance issues, technical problems, and participants' difficulties understanding and following instructions. Recommendations for future protocols include adding a midway check-up, more detailed activity diaries, and repetition of instructions. The need for user-friendly, smartphone-based PVAs was highlighted. Scientific feasibility results indicated that data from PVAs provide valuable information on voice use in daily life, complementing studio recordings. This data offers insights into voice use across different levels of environmental noise and information on phonation time, i.e., the amount of time the person spends using their voice.

Data from studies in this thesis show that voice symptoms are the most prominent feature of dysarthria in mild to moderate PD and that the HiCommunication intervention improves these symptoms, with clinically relevant increases in voice sound level for many participants. The findings also suggest potential functional brain changes related to dysarthria severity and altered connectivity in a speech-related network following intervention. Replication of these results in further RCTs, other settings, and broader PD populations is essential. It is important to interpret the neuroimaging results cautiously, as these studies were exploratory. The next step is to implement HiCommunication in routine healthcare. Furthermore, there is a need to assess the clinical relevance of changes through self-reported outcome measures, establish the minimal clinically important difference for acoustic outcomes, and develop novel PVAs for registering voice sound level in relation to environmental noise. Incorporating PVAs when studying intervention outcomes could enhance ecological validity by capturing real-life voice use, ensuring that observed improvements translate into meaningful changes in everyday communication.

List of scientific papers

I. Steurer, H., Schalling, E., Franzén, E., & Albrecht, F. (2022). Characterization of mild and moderate dysarthria in Parkinson's disease: Behavioral measures and neural correlates. Frontiers in Aging Neuroscience, 14, 870998. https://doi.org/10.3389/fnagi.2022.870998

II. Steurer, H., Albrecht, F., Gustafsson, J. K., Razi, A., Franzén, E., & Schalling, E. (2024). Speech and neuroimaging effects following HiCommunication: a randomized controlled group intervention trial in Parkinson's disease. Brain Communications, 6(4), fcae235. https://doi.org/10.1093/braincomms/fcae235

III. Steurer, H., Körner Gustafsson, J., Franzén, E., & Schalling, E. (2025). Auditory-perceptual analysis of speech outcomes following HiCommunication intervention in Parkinson's disease - a secondary analysis of a randomized controlled trial. [Manuscript]

IV. Steurer, H., Körner Gustafsson, J., Franzén, E., & Schalling, E. (2024) (published online 2022). Using portable voice accumulators to study transfer of speech outcomes following intervention - A feasibility study. Journal of Voice, 38(4), 965.e1-965.e13. https://doi.org/10.1016/j.jvoice.2021.10.016