Studying the effect of medication status on resting state network metrics in Parkinson’s disease
Saadnah Naidu1,2, Yue Lily Xing1,2, and Dorothee Auer2,3

1Radiological Sciences, Division of Clinical Neuroscience, University of Nottingham, Nottingham, United Kingdom, 2Sir Peter Mansfield Imaging Centre, University of Nottingham, Nottingham, United Kingdom, 3Nottingham NIHR Biomedical Research Centre, University of Nottingham, Nottingham, United Kingdom


Current drug therapies for Parkinson’s disease (PD) offer symptom control without the capability for disease modification. Furthermore, unpredictable on/off fluctuations and dyskinesias present challenges in titrating appropriate doses. Our study aims to utilise resting state functional MRI (fMRI) to determine the effect of PD medication, as preliminary step to future work to address these limitations.


The lack of a biomarker to serve as a fingerprint of drug effects produces a methodological limitation in the development of disease-modifying therapies in PD. Here, we describe analysis performed as part of a serial study which provides the opportunity to explore medication effects.


Participants: 13 patients with idiopathic PD (age OFF medication: mean±std, 68.32±7.90 years; age ON medication: 69.43±7.77; Male, 10; H&Y OFF medication: 1.875±0.77; H&Y ON medication: 2.17±1.03) were recruited. The patients were scanned OFF and ON medication on different session visits (Disease duration OFF medication: 5.58±1.25; Disease duration ON medication: 6.89±1.34). Disease severity was determined using the MDS-UPDRS score (UPDRS-III OFF medication: 28.75±19.47; UPDRS-III ON medication: 38.25±17.01).

Image acquisition: Resting state functional imaging was performed using 3T GE 750 MRI with a 32-channel head coil. TR= 2000ms; TE= 30ms. Imaging was performed with eyes open.

Analysis: Analysis was performed using the FSL toolbox (1) using a standardised preprocessing pipeline. The MRIQC algorithm (2) was used to check for motion artefact, which was corrected using an in-house preprocessing pipeline. Images co-registered to individual T1 images were normalised to MNI space and smoothed using a 5mm Gaussian kernel. Functional connectivity was derived from the following seeds (5mm radius): putamen, caudate and primary motor cortex, the coordinates for which obtained from the AAL template and the atlas function on FSL. Z score>2.3 was used to report significance in a paired t test.

To mitigate against the serial effects of the scanning visits occurring at different time points (on medication after off medication), we only studied the on>off medication contrast, in line with the hypothesised medication induced increase of functional connectivity that is unlikely to be confounded by disease progression expected to further decrease functional connectivity.


We found increased functional connectivity in all three seed regions (primary motor cortex, putamen and caudate). Regions of heightened connectivity include limbic, striatal, thalamic, lateral occipital regions, as well as the precuneus, insular cortex and lingual gyrus (Figure 1 A-C).


Whilst comparisons between ON and OFF medication states in PD have been performed by other groups, results have been inconsistent. The functions of the regions of hyperconnectivity demonstrated in our results are associated with motor and non-motor symptoms that are known to be experienced by patients with PD (e.g. limbic involvement in emotional processing, thalamus in motor signals to the cortex, precuneus in memory).

Our study is limited by the sample size and serial non-placebo controlled design and the interval between the visits introducing a possible confounding effect. Ongoing work is investigating the effects of disease progression on neural network changes in PD. Nevertheless, it is unlikely that normalisation of reduced functional connectivity in PD would occur during natural disease progression independent from medication effects.


This preliminary work suggests that resting state fMRI can detect symptomatic PD medication effects as demonstrated by increased functional connectivity in motor and some non-motor networks. This may pave the way for nested resting state fMRI studies in future randomised control trials to investigate neural fingerprints of novel drugs to provide additional mechanistic information of drug effects.


We would like to thank Parkinson's UK for funding our study.


1. Woolrich MW, Jbabdi S, Patenaude B, Chappell M, Makni S, Behrens T, et al. Bayesian analysis of neuroimaging data in FSL. Neuroimage [Internet]. 2009;45:S173–86. Available from: http://dx.doi.org/10.1016/j.neuroimage.2008.10.055 2.

2. Esteban O, Birman D, Schaer M, Koyejo OO, Poldrack RA, Gorgolewski KJ. MRIQC: Advancing the automatic prediction of image quality in MRI from unseen sites. PLoS One. 2017;12(9):1–21.


Figure 1: Connectivity maps from seed-based functional connectivity analysis of the ON>OFF medication contrast with seeds in primary motor cortex (A), putamen (B) and caudate (C).

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)