Altered subcortical emotional salience processing differentiates Parkinson's patients with and without psychotic symptoms

Abstract

Objective: Current research does not provide a clear explanation for why some patients with Parkinson's Disease (PD) develop psychotic symptoms. The 'aberrant salience hypothesis' of psychosis has been influential and proposes that dopaminergic dysregulation leads to inappropriate attribution of salience to irrelevant/non-informative stimuli, facilitating the formation of hallucinations and delusions. The aim of this study is to investigate whether non-motivational salience is altered in PD patients and possibly linked to the development of psychotic symptoms.

Methods: We investigated salience processing in 14 PD patients with psychotic symptoms, 23 PD patients without psychotic symptoms and 19 healthy controls. All patients were on dopaminergic medication for their PD. We examined emotional salience using a visual oddball fMRI paradigm that has been used to investigate early stages of schizophrenia spectrum psychosis, controlling for resting cerebral blood flow as assessed with arterial spin labelling fMRI.

Results: We found significant differences between patient groups in brain responses to emotional salience. PD patients with psychotic symptoms had enhanced brain responses in the striatum, dopaminergic midbrain, hippocampus and amygdala compared to patients without psychotic symptoms. PD patients with psychotic symptoms showed significant correlations between the levels of dopaminergic drugs they were taking and BOLD signalling, as well as psychotic symptom scores.

Conclusion: Our study suggests that enhanced signalling in the striatum, dopaminergic midbrain, the hippocampus and amygdala is associated with the development of psychotic symptoms in PD, in line with that proposed in the 'aberrant salience hypothesis' of psychosis in schizophrenia.

Fig. 1

Task design (displayed for face stimuli only). During this visual oddball paradigm, participants are presented with a random series of greyscale images of faces and outdoor scenes. 66.6% of these were ‘standard’ images. The remaining 33.4% consisted of four types of rare or contextually deviant events, which were randomly intermixed with the standard stimuli; each occurred with a probability of 8.3%. These deviant events were: neutral stimuli that required a motor response (‘target oddball’); stimuli that evoked a negative emotional response (‘emotional oddball’, angry face or image of car crash); novel stimuli (‘novel oddball’, different neutral images that appear only once); and neutral stimuli (‘neutral oddball’, neutral image of face or scene). In the current study, we were only interested in the contrast between negative emotional and neutral oddball stimuli.

Fig. 2

Display of masks used for the region of interest analysis: the bilateral striatum (blue, 3039 voxels), the hippocampus (red, 1033 voxels), the amygdala (green, 505 voxels), and the substantia nigra/VTA (yellow, 645 voxels). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

Bar chart shows mean contrast (COPEs, FSL) values extracted from all voxels of each region of interest and significant group effects (values uncorrected for covariates). Error bars show ± 1 SE. *p < .05. PD-Psychosis: PD patients without psychosis, PD + Psychosis: PD patients with psychosis.

Fig. 4

Scatter plots show correlations that yielded a significant result in at least one group. In PD patients with psychotic symptoms, we found a positive correlation between LED and the BOLD activation in the ROIs (A: bilateral amygdala: r = 0.603, p = .023, B: bilateral hippocampus: r = 0.560, p = .037, C: substantia nigra/VTA: r = 0.631, p = .016, D: striatum: r = 0.573, p = .032). LED was positively correlated to BDI score (E: r = 0.591, p = .025) and apathy score (F: AES, r = 0.849, p = .004); and BDI score was positively correlated to resting CBF bilaterally in the amygdala (G: marginally significant: r = 0.515, p = .059), the hippocampus (H: r = 0.631, p = .015) as well as the substantia nigra (I: r = 0.646, p = .013). In patients without psychotic symptoms, we found that the BDI score was positively correlated to resting CBF in the substantia nigra (I: r = 0.450, p = .036). Only in controls we found a significant correlations between BOLD activation and resting CBF; BOLD signal in the striatum was significantly correlated to bilateral resting CBF in the hippocampus (J: r = 0.567, p = .018).