Objective: Investigating whether resting-state brain connectivity can predict prefrontal cortex activity (PFCA) during walking for patients with Parkinson’s disease (PD) and reflect the benefits of treadmill training (TT).

Background: Most studies in PD examine motor impairments and dysfunctions based on task paradigm. Functional near-infrared spectroscopy (fNIRS) has been used to examine PFCA during walking, reporting elevated activity in PD patients¹. TT has been shown to decrease PFCA in PD patients, improving the automaticity of walking². However, the potentials of resting-state connectivity as an indicator to access the PFCA during walking remain to be accomplished.

Method: Nineteen PD patients were recruited for 6 weeks of TT (n = 19, 74.21±6.33 years, 68% men, PD duration 10.78±6.66 years). PFCA during usual and dual-task (DT) walking was measured before and after TT using fNIRS. The functional connectivity (FC) of PFCA were compared with HC (n = 19, 69.31±5.56 years, 52% men). All participants underwent MRI scans acquiring T1-weighted and rs-fMRI datasets at enrolment and after completion of TT. Effective connectivity (EC) in rs-fMRI were assessed by time-resolved partial directed coherence. EC network differentiating the cohorts and predicting FC were identified by support vector machine (SVM) with classification and regression methods.

Results: After TT, PD patients exhibited FC reduction during usual walking compared to DT (P<0.01). SVM regression applied on EC identified brain networks predicting FC for PD_pre, PD_post, and HC with R²: 0.42, 0.32, and 0.29 respectively. SVM classifier revealed a specific EC network in distinguishing PD from HC with 87.14% accuracy. Within this brain network, the frontal area played an integral role as the network hub by holding connections between subcortical and cerebellum regions, contributing the highest predictive power to the classifier.

Conclusion: The changes in the level of FC and EC after TT showed a trend towards the level of HC suggesting a positive sign of the training effects. The resting-state EC network exhibited the potential to predict PFCA during walking. Interregional connectivity with the frontal region may be the crucial feature in distinguishing PD from HC. Fronto-subcortical and fronto-cerebellar connections may provide a different perspective of the neuropathological consequences of PD.

References: [1] Maidan, I., Nieuwhof, F.,Bernad-Elazari, H., Reelick, M.F., Bloem, B.R., Giladi, N., Deutsch, J.E.,Hausdorff, J.M., Claassen, J.A. and Mirelman, A., 2016. The role of the frontallobe in complex walking among patients with Parkinson’s disease and healthyolder adults: an fNIRS study. Neurorehabilitation and neural repair, 30(10),pp.963-971.  [2] Maidan, I., Nieuwhof, F.,Bernad-Elazari, H., Bloem, B.R., Giladi, N., Hausdorff, J.M., Claassen, J.A.and Mirelman, A., 2018. Evidence for differential effects of 2 forms ofexercise on prefrontal plasticity during walking in Parkinson’s disease.Neurorehabilitation and neural repair, 32(3), pp.200-208.

« Back to MDS Virtual Congress 2021

MDS Abstracts - https://www.mdsabstracts.org/abstract/resting-state-connectivity-predictors-for-prefrontal-cortex-activity-during-walking-in-parkinsons-disease/