Reward insensitivity is associated with dopaminergic deficit in rapid eye movement sleep behaviour disorder

Abstract

Idiopathic rapid eye movement sleep behaviour disorder (iRBD) has now been established as an important marker of the prodromal stage of Parkinson's disease and related synucleinopathies. However, although dopamine transporter single photon emission computed tomography (SPECT) has been used to demonstrate the presence of nigro-striatal deficit in iRBD, quantifiable correlates of this are currently lacking. Sensitivity to rewarding stimuli is reduced in some people with Parkinson's disease, potentially contributing to aspects of the neuropsychiatric phenotype in these individuals. Furthermore, a role for dopaminergic degeneration is suggested by the fact that reward insensitivity can be improved by dopaminergic medications. Patients with iRBD present a unique opportunity to study the relationship between reward sensitivity and early dopaminergic deficit in the unmedicated state. Here, we investigate whether a non-invasive, objective measure of reward sensitivity might be a marker of dopaminergic status in prodromal Parkinson's disease by comparing with SPECT/CT measurement of dopaminergic loss in the basal ganglia. Striatal dopaminergic deficits in iRBD are associated with progression to Parkinsonian disorders. Therefore, identification of a clinically measurable correlate of this degenerative process might provide a basis for the development of novel risk stratification tools. Using a recently developed incentivized eye-tracking task, we quantified reward sensitivity in a cohort of 41 patients with iRBD and compared this with data from 40 patients with Parkinson's disease and 41 healthy controls. Patients with iRBD also underwent neuroimaging with dopamine transporter SPECT/CT. Overall, reward sensitivity, indexed by pupillary response to monetary incentives, was reduced in iRBD cases compared with controls and was not significantly different to that in patients with Parkinson's disease. However, in iRBD patients with normal dopamine transporter SPECT/CT imaging, reward sensitivity was not significantly different from healthy controls. Across all iRBD cases, a positive association was observed between reward sensitivity and dopaminergic SPECT/CT signal in the putamen. These findings demonstrate a direct relationship between dopaminergic deficit and reward sensitivity in patients with iRBD and suggest that measurement of pupillary responses could be of value in models of risk stratification and disease progression in these individuals.

Keywords: Parkinson’s disease; REM sleep behaviour disorder; dopamine; reward.

Figure 1

Oculomotor paradigm schematic representation of the eye tracking task (adapted from Muhammed et al.). Participants heard an auditory cue that informed them of the maximum reward available for each trial: ‘0p, 10p or 50p maximum’. After a variable delay of 1400, 1500 or 1600 ms, the central fixation disc disappeared and a new target disc appeared. Participants were rewarded according to reaction time, with the reward obtained displayed within the target disc in pence.

Figure 2

DaT SPECT/CT imaging in RBD patients. (A) Example striatal ROI registered to DaT SPECT/CT. (B) ROI used in the calculation of striatal SURs. Regions used were putamen [posterior ROIs within the striatal outline, shown in red (left putamen) and orange (right putamen)] and occipital reference (dark blue). (C) ROI superimposed on an example DaT SPECT/CT image. (D and E) Example DaT SPECT/CT images from two RBD patients with normal (D) and abnormal (E) imaging. The abnormal image (E) demonstrates asymmetric signal loss in the putamen, typical of Parkinson’s disease.

Figure 3

Pupillary response to reward in controls, RBD and Parkinson’s disease patients. (A) Proportional pupil changes at each reward level across healthy controls, RBD and Parkinson’s disease patients, with each group normalized to the 0p baseline level to demonstrate the relationship between reward sensitivity slopes. Error bars indicate standard error of the mean difference observed between each reward level and the 0p baseline. (B) Pupil reward sensitivity across the groups, calculated as the difference in pupil response between 50p and 0p rewards on offer. Box and whisker plots indicate median (line within box), mean (+), interquartile range (box outline) and maximum and minimum values (whiskers).

Figure 4

Pupillary response to reward in controls, Parkinson’s disease patients and iRBD patients split according to abnormal and normal DaT SPECT/CT imaging. (A) Proportional pupil changes at each reward level across healthy controls, Parkinson’s disease patients and RBD patients divided into DaT SPECT/CT outcome, with each group normalized to the 0p level. Error bars indicate standard error of the mean difference observed between each reward level and the 0p baseline. (B) Pupil reward sensitivity (pRS) across the groups including RBD subgroups, calculated as the difference in pupil response to 50p and 0p. Box and whisker plots indicate median (line within box), mean (plus sign), interquartile range (box outline) and maximum and minimum values (whiskers). (C–E) Mean pRS (pupil change to 50p reward minus response to 0p) plotted over time in RBD patients (C), Parkinson’s disease patients (D) and controls (E). (C) In RBD patients, a significant difference in pRS between those with normal (red, asterisk) and abnormal (purple, plus sign) dopaminergic imaging occurred from ∼1300 ms to the end of the trial, indicated by the grey bar (P < 0.05). (D) In Parkinson’s patients, there was a significant reduction in pRS when off dopaminergic medication (blue, number sign) versus on (green, triangle). Parts D and E adapted from previously published data (Muhammed et al.). Shaded areas indicate standard error of the mean.

Figure 5

Association between average putamen DaT SPECT/CT signal, pRS and average pupillary arousal level. (A) Association between pRS and mean putamen DaT imaging. Linear mixed-effects modelling was used to encompass reward sensitivity scores while including 0p, 10p and 50p data. A significant positive association between pRS and mean putamen DaT SPECT/CT signal was demonstrated (t = 2.2, P = 0.03). Purple shaded area indicates 95% CI of the best fit line. (B) General pupil arousal was measured as the average change in pupil response across all reward levels to the cue over the 1400–2400 ms period of interest. This was correlated against an individual’s mean putamen DaT SPECT/CT signal and no significant effect was found.