Fig. 5: A higher CBR in patients with RT and asymmetry in caudate binding is sufficient to explain observations associating CBR with tremor amplitude.

A Results from a meta-analysis on studies reporting CBR in different motor phenotype subgroups of PD patients (with Tremor/Tremor Dominant and without Tremor/No-Tremor Dominant). B Correlation coefficients between RT and CBR extracted from the literature. C Correlation matrix between RT related-symptoms (Right Tremor: Sum of Right Upper - 3.17a - and Lower Limb - 3.17c - RT scores; Left Tremor: Sum of Left Upper - 3.17b - and Lower Limb - 3.17d - RT scores; Jaw tremor: 3.17e; RT: Total sum of all scores) and Right and Left Caudate and Putamen binding ratios. D An in silico model was created with 3 assumptions: (1) Patients with tremor have a higher CBR than patients without tremor, (2) Tremor amplitude (i.e, values above 0) on the reference side is not associated with CBR, (3) Ipsi and contralateral CBR are highly correlated (~0.80). Top: Simulated contralateral and ipsilateral CBR distributions for Tremor and No-Tremor groups, Bottom: Correlation of the simulated RT score and CBR. E Left: Distributions of ipsi and contralateral CBR in patients presenting tremor on the reference side and in patients presenting without tremor on the reference (PPMI data). Right: Correlation between tremor amplitude score and ipsi and contralateral CBR for the PPMI dataset. F Results from the model presented in D with distributions means defined based on the PPMI data described in G. Correlation results of a simulation of 10.000 subjects. G The same strategy used in F was applied to the data available and described in A. The scatter plot denotes the correlation of the coefficients predicted by our model with the coefficients reported in the original studies.