MEG signals help predict symptom progression in Parkinson’s disease, new study shows

A new study published in npj Parkinson’s Disease suggests that simple, non-invasive brain recordings may help forecast how motor symptoms will evolve in people with Parkinson’s disease. Using resting-state magnetoencephalography (MEG), a team of researchers from Karolinska Institutet, Stockholm, Sweden, tracked 27 individuals with Parkinson’s disease and 30 healthy volunteers over an average of four years to explore how changes in brain activity relate to worsening symptoms.

By analysing brain activity region by region, the team identified subtle shifts in neural signals that were linked to specific aspects of disease progression. One key finding was that an increase in a particular background signal component, known as the aperiodic exponent, in the left sensorimotor cortex was associated with growing rigidity over time. They also observed that elevated beta-band activity in posterior brain regions appeared to be linked to milder movement slowing at baseline, a relationship that weakened as symptoms advanced. This pattern may reflect early compensatory mechanisms that later break down.

Importantly, the researchers used these MEG features to predict future motor deterioration in an independent group of patients. Their model explained around 20% of the variability in long-term symptom progression, highlighting the potential of MEG-based markers to support prognosis and personalised monitoring in Parkinson’s disease.

The baseline MEG dataset from the study is available through the EBRAINS platform as the NatMEG-PD BIDS dataset, contributing to open science by supporting data sharing and reuse in Parkinson’s disease research. The corresponding data descriptor has been published in Scientific Data.

Read the article: 

Magnetoencephalography-based prediction of longitudinal symptom progression in Parkinson’s disease

Josefine Waldthaler, Igori Comarovschii & Daniel Lundqvist.