Auditory dysfunction may accompany the motor manifestations of Parkinson’s disease (PD), but high-frequency auditory encoding in PD remains insufficiently characterised. This pilot study compared the frequency-following response (FFR), a sustained response reflecting neural synchronisation to periodic sound, between participants with PD and healthy controls (HC) during passive listening, and examined its clinical associations. Stimulation at 333 Hz was selected to favour predominantly subcortical contributions and establish a baseline response for future studies of auditory-motor interactions. Electroencephalographic (EEG) activity was recorded using 64 channels during binaural presentation of 200 ms pure tones at 80 dB SPL with alternating polarity. Of 26 enrolled participants, 23 (13 HC and 10 PD) were retained after exclusions due to poor recording quality or unreliable triggers. PD participants were assessed in the medication OFF-state. Data were processed in MATLAB and FieldTrip. Phase locking was assessed with the Rayleigh test, complemented by a spectral F-test. The signal-to-noise ratio (SNR) of channels selected following review of the Rayleigh results was analysed using linear mixed-effects models with participant-specific random intercepts, with and without age adjustment. Age-adjusted PD-only models examined associations with Parts I-III of the Movement Disorder Society-sponsored revision of the Unified Parkinson’s Disease Rating Scale (MDS- UPDRS). Simulation-based power analysis assessed the sensitivity of the study design. All retained participants showed significant phase locking at 333 Hz. Adding age improved model fit (p = 0.005), but the adjusted SNR difference between PD and HC was not statistically significant (PD minus HC: -0.924 dB; 95% Wald confidence interval: -2.908 to 1.061; p = 0.362). No significant associations were found with MDS-UPDRS Parts I-III. Assuming the pilot variance estimates and an effect equal to the age-adjusted group coefficient, simulations estimated approximately 17% power at the available sample size.The non-significant findings do not establish equivalence between groups. The study provides a methodological and physiological baseline for subsequent investigations of auditory-motor interactions, while the clinical rele- vance of the 333-Hz FFR remains unresolved. The small sample, absence of hearing-threshold adjustment and partly visual channel selection indicate the need for larger cohorts and more reproducible measurements.
La malattia di Parkinson (PD), oltre ai disturbi motori, può comportare alterazioni nell’elaborazione uditiva. In questa popolazione, le risposte neurali ad alta frequenza risultano tuttavia ancora poco studiate. Questa tesi ha esaminato la frequency-following response (FFR), una risposta neurale che traccia nel tempo la periodicità dello stimolo acustico. Lo studio pilota ha confrontato la FFR tra persone con PD e controlli sani (HC) durante l’ascolto passivo, esplorandone le possibili associazioni con i punteggi clinici. Sono state effettuate registrazioni elettroencefalografiche (EEG) a 64 canali durante l’ascolto di toni puri a 333 Hz, una frequenza scelta per privilegiare il contributo delle strutture sottocorticali. I toni (durata di 200 ms, intensità di 80 dB SPL) sono stati presentati a entrambe le orecchie alternando la polarità tra una ripetizione e la successiva. Dei 26 partecipanti iniziali, 23 (13 HC e 10 PD) sono stati inclusi nell’analisi, mentre i restanti sono stati esclusi per qualità insufficiente del segnale o problemi nella sequenza dei trigger. I pazienti affetti da PD sono stati esaminati in condizione farmacologica OFF (in assenza di farmaci). I segnali sono stati elaborati in MATLAB tramite il toolbox FieldTrip. Il phase-locking è stato valutato mediante il test di Rayleigh, affiancato da un F-test per l’analisi spettrale. Sui canali selezionati in base ai risultati del test di Rayleigh è stato quindi calcolato il rapporto segnale/rumore (SNR). Il confronto tra gruppi si è basato su modelli lineari ad effetti misti con intercetta casuale per partecipante, stimati sia inserendo sia escludendo l’età come covariata. Nel solo gruppo PD, ulteriori modelli hanno esaminato le associazioni con i punteggi delle parti I–III della scala MDS-UPDRS, aggiustando per l’età. La potenza statistica dello studio è stata infine valutata tramite simulazioni. In tutti i partecipanti analizzati è stata rilevata una risposta con phase-locking significativo a 333 Hz. Il modello che includeva l’età descriveva meglio i dati (p = 0,005), ma anche tenendo conto di questa variabile la differenza di SNR tra i gruppi non è risultata statisticamente significativa (PD vs HC: -0,924 dB; 0,924 dB; intervallo di confidenza di Wald al 95%: da −2,908 a 1,061; p = 0,362). Non sono inoltre emerse associazioni significative con le parti I-III della scala MDS-UPDRS. Per il campione disponibile, le simulazioni hanno indicato una potenza di circa il 17%. Le simulazioni si basa- vano sulle varianze stimate nel pilota e ipotizzavano una differenza reale tra gruppi pari a quella ottenuta con il modello che includeva l’età. L’assenza di significatività statistica non dimostra tuttavia che i due gruppi abbiano risposte equivalenti. Il lavoro offre un riferimento fisiologico e metodologico per futuri studi sulle interazioni tra udito e movimento, mentre resta da chiarire l’effettiva utilità clinica della FFR a 333 Hz. Il campione ridotto, il mancato inserimento delle soglie uditive nei modelli e la selezione dei canali basata in parte sull’ispezione visiva limitano l’interpretazione dei risultati, rendendo necessari studi su campioni più ampi e con misure maggiormente riproducibili.
The Frequency-Following Response as a Candidate Marker of Brainstem Integrity in Parkinson's Disease: A Pilot EEG Study
GUTIERREZ, NICOLA
2025/2026
Abstract
Auditory dysfunction may accompany the motor manifestations of Parkinson’s disease (PD), but high-frequency auditory encoding in PD remains insufficiently characterised. This pilot study compared the frequency-following response (FFR), a sustained response reflecting neural synchronisation to periodic sound, between participants with PD and healthy controls (HC) during passive listening, and examined its clinical associations. Stimulation at 333 Hz was selected to favour predominantly subcortical contributions and establish a baseline response for future studies of auditory-motor interactions. Electroencephalographic (EEG) activity was recorded using 64 channels during binaural presentation of 200 ms pure tones at 80 dB SPL with alternating polarity. Of 26 enrolled participants, 23 (13 HC and 10 PD) were retained after exclusions due to poor recording quality or unreliable triggers. PD participants were assessed in the medication OFF-state. Data were processed in MATLAB and FieldTrip. Phase locking was assessed with the Rayleigh test, complemented by a spectral F-test. The signal-to-noise ratio (SNR) of channels selected following review of the Rayleigh results was analysed using linear mixed-effects models with participant-specific random intercepts, with and without age adjustment. Age-adjusted PD-only models examined associations with Parts I-III of the Movement Disorder Society-sponsored revision of the Unified Parkinson’s Disease Rating Scale (MDS- UPDRS). Simulation-based power analysis assessed the sensitivity of the study design. All retained participants showed significant phase locking at 333 Hz. Adding age improved model fit (p = 0.005), but the adjusted SNR difference between PD and HC was not statistically significant (PD minus HC: -0.924 dB; 95% Wald confidence interval: -2.908 to 1.061; p = 0.362). No significant associations were found with MDS-UPDRS Parts I-III. Assuming the pilot variance estimates and an effect equal to the age-adjusted group coefficient, simulations estimated approximately 17% power at the available sample size.The non-significant findings do not establish equivalence between groups. The study provides a methodological and physiological baseline for subsequent investigations of auditory-motor interactions, while the clinical rele- vance of the 333-Hz FFR remains unresolved. The small sample, absence of hearing-threshold adjustment and partly visual channel selection indicate the need for larger cohorts and more reproducible measurements.| File | Dimensione | Formato | |
|---|---|---|---|
|
Gutierrez_Nicola.pdf
accesso aperto
Dimensione
2.87 MB
Formato
Adobe PDF
|
2.87 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/119611