Posters

Parkinson’s disease (PD) is a neurodegenerative disorder characterised by bradykinesia, resting tremor, sleep disturbances and depression. This study presents a neural network model of the motor system using binary state neurons connected according to patterns described in the literature. The aim of the model is to reproduce the motor symptoms observed in different clinical forms of PD and simulate the effects of pharmacological and deep brain stimulation (DBS) interventions. The network included multiple areas of the motor system, such as cerebellar, basal ganglia, cortical and spinal circuits controlling an idealised joint. Motor programs and their priorities were fed into the network, whose output drove agonist and antagonist muscles to generate joint trajectories. PD motor symptomatology emerged by decreasing the levels of dopamine release. Concurrently, altering specific synaptic weights reproduced the two main subtypes of PD (tremor-dominant and postural instability and gait difficulty). Effects of pharmacological and DBS interventions were simulated, resulting in the expected symptomatology improvements as reported in the literature. Effects of levodopa and levodopa overadministration were reproduced, providing an explanation for dyskinesias observed in clinical settings. The simulations support the view that PD resting tremor may be an undesirable consequence of cerebellar compensation to alleviate bradykinesia. This compensation is driven mainly by an increased functional connectivity of the subthalamic nucleus with the cerebellum. Overall, the proposed modelling strategy offers a mechanistic tool for investigating PD pathophysiology and potential personalised treatments, enabling predictions of optimal pharmacological or DBS interventions based on patient-specific alterations in functional connectivity.

Background: Anxiety is frequent in Parkinson’s disease (PD), and may relate to pathological changes in fear and limbic circuits. The hippocampus and amygdala, key structures in these circuits affected in PD, may underlie PD-related anxiety. This study examines the changes in volume in hippocampal subregions and the amygdala in anxious PD patients.

Methods: 45 PD participants and 45 healthy controls were divided into anxious and non-anxious participants (PDA/PDN; HCA/HCN) based on GAD-7 scores. Volumes of automatically segmented hippocampal subregions and the amygdala, measured via 3T MRI as ratio of ROI volume (%00) to TIV, were compared using ANCOVAs (covariates: age, gender, total intracranial volume (TIV)). Correlations with anxiety severity (GAD-7, PAS) were assessed.

Results: The right fimbria was smaller in PDA (0.39%00) compared to both PDN (0.42%00) and HCA (0.41%00). The left fimbria was smaller in PDA (0.40%00) versus HCA (0.49%00). The right HATA was smaller in PDA (0.34%00) versus HCA (0.34%00), but larger when compared to PDN (0.31%00). In PD groups, higher anxiety severity correlated negatively with smaller left fimbria (GAD-7: r=-0.43; PAS: r=-0.35) and left HATA volumes (GAD-7: r=-0.46). No correlations between anxiety and subregion volumes were observed in controls.

Conclusion: To our knowledge, the association between reduced fimbria volume and PD-related anxiety is a novel finding. As the fimbria is a pivotal white matter pathway within the Papez circuit, its disruption may impair emotional regulation, contributing to PD-related anxiety. This supports a disease-specific pathological basis for anxiety in PD, rather than a reactive psychological response to diagnosis.