Selective effects of dopaminergic and noradrenergic degeneration on cognition in Parkinson’s disease
======================================================================================================

* Sophie Sun
* Victoria Madge
* Jelena Djordjevic
* Jean-François Gagnon
* D Louis Collins
* Alain Dagher
* Madeleine Sharp

## Abstract

The substantia nigra and locus coeruleus are among the first brain regions to degenerate in Parkinson’s disease. This has important implications for early cognitive deficits as these nuclei are sources of ascending neuromodulators (i.e., dopamine and noradrenaline) that support various cognitive functions like learning, memory, and executive function. However, because most studies of the relationship between patterns of degeneration and cognition have either studied these neuromodulator systems in isolation or studied specific cognitive domains in isolation, it is unknown if degeneration in the substantia nigra and degeneration in the locus coeruleus independently and selectively contribute to different cognitive deficits in Parkinson’s disease.

To address this gap, we tested people with Parkinson’s disease and older adults on tasks of positive reinforcement learning, attention/working memory, executive function, and memory to measure performance in domains of cognition specifically thought to be related to dopaminergic and noradrenergic function. Participants also underwent neuromelanin-sensitive magnetic resonance imaging which provides a measure of degeneration of dopamine neurons in the substantia nigra and of noradrenergic neurons in the locus coeruleus. Brain-behaviour relationships were evaluated by separate linear regressions predicting cognitive performance in each domain from substantia nigra and locus coeruleus neuromelanin signal intensities controlling for age, sex, and education.

As expected, Parkinson’s disease patients had significantly slower learning from positive feedback and lower performance on tests of attention/working memory, executive function, and memory than controls. Parkinson’s patients also had lower neuromelanin signal intensity in the substantia nigra and locus coeruleus. Examining brain-behaviour relationships, we found that reduced neuromelanin signal in the substantia nigra in Parkinson’s disease patients was independently associated with impaired positive reinforcement learning, controlling for changes in the locus coeruleus, but was not associated with other domains of cognition. In contrast, reduced neuromelanin signal in the locus coeruleus was independently associated with impairments in attention/working memory and executive function, controlling for changes in the substantia nigra, but not with reinforcement learning performance. These results show that substantia nigra degeneration and locus coeruleus degeneration independently and selectively contribute to cognitive deficits and therefore suggests that individual differences in the degree of neurodegeneration in these nuclei could explain the significant heterogeneity that exists in the cognitive and behavioural manifestations of Parkinson’s disease. These findings also highlight the potential value of leveraging known brain-behaviour relationships to develop performance-based measures of cognition that reflect underlying patterns of neurodegeneration.

Keywords
*   reinforcement learning
*   attention
*   executive function
*   neuromelanin
*   locus coeruleus
*   substantia nigra

## Introduction

The dopaminergic substantia nigra (SN) and noradrenergic locus coeruleus (LC) play crucial roles in modulating a wide range of cognitive processes.6–9 In Parkinson’s disease, degeneration of both these nuclei starts early.10–14 For instance, studies using magnetic resonance imaging (MRI) sensitive to neuromelanin, a by-product of catecholamine metabolism,15,16 have shown that neuromelanin in the SN and LC is already significantly reduced in the early stages of Parkinson’s disease.2,17,18 Correspondingly, impairments of the cognitive processes that depend on dopaminergic and noradrenergic signalling are also often already present in this early period.14,19–23 However, whether specific aspects of early cognitive dysfunction in Parkinson’s disease can be selectively attributed to degeneration in either one, or both, of these systems is less clear because they have largely been studied in isolation of one another or studies have focused on singular domains of cognitive performance rather than the various cognitive domains they are known to modulate. Meanwhile, delineating their independent contributions to cognition in Parkinson’s disease is essential for understanding the heterogeneity that exists in the cognitive phenotype of Parkinson’s disease and for developing methods to track clinical outcomes associated with early neurodegeneration.

There is a large body of work showing that pharmacological manipulations of dopamine state in Parkinson’s patients can remediate certain cognitive deficits,20,24–34 supporting the notion that degeneration in the SN plays an important role in the early cognitive deficits of Parkinson’s disease.35 Meanwhile, attempting to relate the severity of degeneration in the dopaminergic system to cognitive impairments has proven more difficult.1,36–42 One limitation of this work is that it has focused on standard domains of clinical neuropsychological testing (i.e., those recommended for classifying cognitive impairment in Parkinson’s disease)43 and may have failed to capture specific aspects of cognition known to depend on dopamine. Indeed, one of the most extensively studied dopamine-dependent cognitive processes in Parkinson’s disease that falls outside of standard neuropsychological testing is reinforcement learning.44,45 Reinforcement learning refers to the ability to update expectations on the basis of recent feedback (often measured as the computationally-defined learning rate). In Parkinson’s patients, reinforcement learning rate is believed to be impaired on the basis that positive feedback signals, carried by dopamine, are dampened.30,31,46–53 Though reinforcement learning has strong neurobiological grounding as a measure that could be used to track the cognitive consequences of SN degeneration, it has primarily been studied in small samples and in the context of dopaminergic medication manipulations, which are rarely feasible in larger scale clinical research. Whether reduced positive reinforcement learning is associated with the severity of SN degeneration in medicated Parkinson’s patients is unknown.

Another explanation for the mixed results linking SN degeneration with cognition is that there may be confounding effects of concurrent degeneration in the noradrenergic system. Indeed, degeneration of the noradrenergic LC is associated with attention, working memory, declarative memory, and executive function in Parkinson’s disease.1,2,54–56 There is also an important overlap in the cortical regions receiving projections from the SN and from the LC, as well as an overlap in cognitive functions reported to be sensitive to dopaminergic and noradrenergic modulation.57 Therefore, establishing a clear account of the independent effects of SN and LC degeneration on cognition will require consideration of these possible confounding effects.

Age-related brain changes also overlap with those of Parkinson’s disease. In particular, loss of LC integrity occurs with aging and is associated with cognitive impairments in older adults,58 yet only two studies examining LC integrity and its relationship to cognition in Parkinson’s disease accounted for the effects of age.1,2 It therefore remains unclear to what degree the relationship between LC degeneration and cognition observed in Parkinson’s patients is truly specific to the neurodegeneration of Parkinson’s disease.

The objective of the present study was to determine whether SN and LC neurodegeneration in Parkinson’s disease are independently associated with specific cognitive deficits. In a large sample of people with Parkinson’s disease and older adults we measured cognitive processes known to be related to these nuclei: positive reinforcement learning for the SN, and executive function, attention/working memory, and declarative memory for the LC. A subset of these participants also underwent neuromelanin MRI to extract measures of SN and LC neurodegeneration. First, we predicted that people with Parkinson’s disease would show impairments across all these measures compared to the older adults. Second, we predicted that SN neurodegeneration would be selectively associated with impaired learning from positive feedback while controlling for effects of LC degeneration and of aging and, conversely, that LC degeneration would be selectively associated with impairments in attention/working memory, executive function, and declarative memory while controlling for effects of SN degeneration and of aging.

## Methods

### Participants

One hundred and thirty-five people with Parkinson’s disease (PD) and 72 older adults (Controls) were recruited from the Quebec Parkinson Network (QPN) – a registry of patients with Parkinson’s disease who have expressed interest in research and who are referred to the network by their neurologist – or from the Montreal community in the case of older adults. Other than age (ranging from 50-90), there were no exclusion criteria to ensure a representative sample. The sample characteristics are summarized in **Table 1**. Patients completed the study assessments in their usual dopaminergic medicated “ON” state (only three patients were not taking any dopaminergic medication at the time of testing), but due to an error in data collection, the details of their dopaminergic medications are not available. All participants completed the study in their language of preference (English or French) at the Montreal Neurological Institute, provided written informed consent, and were paid $25 for their participation. The study was approved by the McGill University Health Centre Research Ethics Board.

View this table:
[Table 1.](http://medrxiv.org/content/early/2024/09/16/2024.09.16.24313753/T1)

Table 1. Sample Characteristics

Participants completed questionnaires about demographic information and mood (Questionnaire for Impulsive-Compulsive Disorders in Parkinson’s disease and Geriatric Depression Scale), MRI, standard neuropsychological testing, and additional neurocognitive testing. Patients with Parkinson’s disease completed the Unified Parkinson’s Disease Rating Scale. Each element of the protocol is described in further detail below. The testing took place over one or two days within 90 days of each other, depending on scheduling. Due to the toll of the testing protocol, not all participants completed all assessments and therefore some of the analyses were conducted on different but largely overlapping sub-samples to maximize sample sizes and statistical power for each set of analyses. The exact sample size for each analysis is indicated in the relevant results section and the participant characteristics for each subsample are summarized in **Supplementary Table 1**. Importantly, the sample characteristics were similar across subsamples and the key group differences observed in the overall sample (for sex distribution, MoCA and depression score) were also observed in each subsample.

### Cognitive testing

#### Cognitive measure of SN degeneration

Reinforcement learning performance was measured using an adapted probabilistic learning task48,59 with a focus on the learning phase of the task given our interest in computationally modeling the reinforcement learning rate (**Fig. 1**). The reinforcement learning task was administered using PsychoPy version 2020.2.10 (Open Science Tools Ltd) 60. In the learning phase, three abstract visual stimulus pairs (AB, CD, EF) were presented, and participants learned to choose one of the two stimuli in the pair based on the feedback provided after each trial. Each pair was presented 50 times for a total of 150 trials. The order of stimulus pair presentation was randomized, as was the side of the screen on which each stimulus of a pair was presented. Feedback was probabilistic such that for the AB pair, the probability of receiving positive feedback for choosing stimulus A was 0.8, the probability of stimulus C being rewarded in the CD pair was 0.7, and the probability of stimulus E being rewarded in the EF pair was 0.6. Stimulus images were randomly assigned to each probability pair for each participant. Written feedback was presented on screen after each trial and consisted of “Correct!” or “Incorrect!”. To enhance engagement, participants were provided a task narrative: they were told that the visual stimuli represented store logos and that their goal was to accumulate as many items as possible by learning which stores were most likely to reveal the item category of interest. Participants were randomly assigned to one of two task versions: shopping for food items or shopping for household items. To support the task narrative, at the time of feedback, images of either food or household items59,61 were shown depending on whether the trial was rewarded or not. On trials where a participant was not rewarded, an image from the alternate category was shown.

![Figure 1](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/09/16/2024.09.16.24313753/F1.medium.gif)

[Figure 1](http://medrxiv.org/content/early/2024/09/16/2024.09.16.24313753/F1)

Figure 1 Reinforcement learning task.
Participants were instructed to shop for items in a given category (i.e., food or household items) by learning which stores would most likely yield a target item through trial-and-error. **(A)** On each trial, participants had to choose between two visual stimuli, which represented the store logos. Upon selection of one of the stimuli, participants were provided with trial-unique visual feedback of a target item if they were correct or a non-target item if they were incorrect. **(B)** The stimuli had varying probabilities of yielding a correct target item with a range between 0.2 and 0.8. Shown is an example of stimulus pairs with associated probabilities.

#### Cognitive measures of LC degeneration

The measures of interest related to LC degeneration were extracted from the neuropsychological testing that participants completed. This testing protocol was designed to align with the Movement Disorder Society recommendations for assessment of cognition in Parkinson’s disease and tested five domains,43,62 but, because of their previously demonstrated relationship to LC, the domains of interest here were: attention/working memory, executive function and memory.12,39,60–62 Attention/working memory were assessed using the Digit Span test (forward and backward) and the Trail Making Test Part A (TMT A). Executive function was assessed using the TMT Part A subtracted from TMT Part B (TMT B-A), the Delis-Kaplan Executive Function (D-KEFS) Color-Word Interference Test (CWIT) using response time on the interreference condition (condition 3: Inhibition), and the Brixton Spatial Anticipation Test (BSAT; number of errors). Memory was assessed using the Hopkins Verbal Learning Test (HVLT; total on trials 1, 2, 3 and delayed recall) and the Rey Complex Figure Test (RCFT; immediate and delayed recalls). To standardize the direction of the performance scores such that greater scores represent better performance, TMT A, TMT B-A, D-KEFS CWIT, and BSAT scores were multiplied by -1.

#### Other cognitive measures

To demonstrate the specificity of the relationships of the above cognitive tests to degeneration in the SN and LC, we also examined performance in the domains of visuospatial function and language. Visuospatial function was assessed using the Clock Drawing Test (CDT) and the RCFT Copy trial. Language was assessed using the Letter Verbal Fluency (F, A, S; 1 minute per condition), Semantic Verbal Fluency (animals, actions; 1 minute per condition), and Boston Naming Test (errors without hints out of 60 items). The Montreal Cognitive Assessment (MoCA) was administered as a measure of global cognition.

### Neuroimaging

Participants underwent an MRI protocol which included 3D T1-weighted Magnetization-Prepared Rapid Acquisition Gradient Echo (T1w) and 2D T1-weighted Fast Spin Echo (neuromelanin-sensitive) sequences using a 3T Siemens Prisma scanner with a 32-channel head coil. Imaging parameters are listed in **Supplementary Table 2**. The present study aimed to investigate the structural integrity of the SN and LC using neuromelanin-sensitive images.

Raw neuromelanin-images underwent slice-by-slice normalization to remove inhomogeneities from the image using the MINC Toolkit 63 before being brought to stereotaxic space using transformations from each patient’s respective T1w image. Both sequences were acquired in the same session, with the assumption that they are co-registered in raw space. The raw T1w images had underwent pre-processing via the NeuroImaging & Surgical Technologies Longitudinal Pipeline64 and were non-linearly registered to stereotaxic space using the PD126 template65 as the registration target.

In standard space, the LC and SN regions of interest were defined using a conservatively thresholded probabilistic atlas,66 and expert manual segmentations on an in-house template, respectively. Neuromelanin signal intensity form these regions was calculated as a contrast-to-noise ratio, using the pontine tegmentum and cerebral peduncles as background intensity, respectively (1): ![Formula][1]</img>  

### Analyses

#### Reinforcement learning drift diffusion model

Reinforcement learning was evaluated by modelling dual learning rates.30,48,67 Computational modelling of the probabilistic learning task data was performed using the HDDM toolbox in Python.68 After removing omission errors from the dataset, we applied a reinforcement learning drift diffusion model (RLDDM), which uses the drift diffusion model as the choice policy (as opposed to typical softmax policies) within a reinforcement learning framework (i.e., Delta learning) allowing us to also account for cognitive processes related to decision-making.69,70 Evidence accumulation parameters were estimated using the drift diffusion model given that binary decision-making processes are required to engage in our reinforcement learning task, i.e., deciding between two stimuli on each trial, which provides more behavioural resolution than learning models using softmax or greedy choice policies that do not account for response time and thus speed-accuracy trade-offs.71

Briefly, in the RLDDM, learning is modelled using the delta learning rule with dual learning rates for positive and negative prediction errors, while the choice policy is modelled by the drift diffusion model. Drift rate (*vi*) is the speed of evidence accumulation, which is described as the per trial (*i*) difference between expected reward values between the two stimuli (*Qupper, i – Qlower, i*) multiplied by a parameter (*v*) representing the degree of exploration or exploitation (analogous to the inverse temperature parameter in softmax choice policies). Decision boundary (*a*) is the distance between the two responses such that a greater distance represents greater response caution, i.e., slower but more accurate responses. Non-decision time (*t*) is the time associated with stimulus encoding and motor execution.

The RLDDM uses hierarchical and Bayesian estimation to fit the model. The hierarchical approach estimates an individual and group-level distribution of parameters. We specified group (i.e., Parkinson’s disease vs. Control) as a condition for each parameter such that each group had their respective distribution of parameters. There was an exception for non-decision time for which we estimated one group-level distribution spanning patients and controls together since this parameter does not represent a cognitive feature of behaviour. The Bayesian approach estimates a posterior distribution of parameter values using the Markov Chain Monte Carlo method. We generated 15,000 samples and discarded the first 5,000 samples. Convergence was assessed by visual inspection of the trace, autocorrelation, and histogram of the posteriors, and by removing participants with a Gelman-Rubin convergence diagnostic greater than 1.1.

Positive learning rate was our main parameter of interest as it represents the speed of updating value from positive feedback. Negative learning rate, drift rate and decision boundary served as control parameters; they measure different behavioural processes within the reinforcement learning task that were not expected to be related to dopamine neuron loss.

#### Statistical analyses

To evaluate group differences in demographic characteristics we conducted independent samples t-tests (two-tailed) for continuous, normally distributed variables, Mann-Whitney U tests for continuous, not normally distributed variables, and chi-squared analyses for categorical variables. To examine group differences in RLDDM-derived behavioural parameters on the reinforcement learning task, we computed a Bayesian p-value by calculating the proportion of each parameter’s posterior distribution that overlapped between groups. To examine group differences in neuromelanin signal intensities, we conducted ordinary least squares regressions with group as an independent variable (effect coded with Controls as the reference group) and controlled for age, sex, and years of education.

To investigate brain-behaviour relationships, we conducted ordinary least squares regressions on cognitive measures with SN and LC neuromelanin signal intensities as independent variables controlling for age, sex, and years of education. Reinforcement learning and decision-making behaviour was summarized by behavioural parameters from the RLDDM. Cognitive performance on the neuropsychological tests was summarized by five composite scores representing attention/working memory, executive function, memory, visuospatial function, and language. Performance in each neuropsychological task was z-scored (separately for Parkinson’s disease and Control groups) and domain composite scores were computed by averaging across z-scores that fell within respective domains. Each RLDDM-derived behavioural parameter and each cognitive composite score was evaluated in separate models as the dependent variable and separate analyses were conducted in Parkinson’s patients and Controls. We performed exploratory analyses when either the SN or LC neuromelanin signal intensity significantly predicted cognitive performance to investigate whether there was an interaction between them by repeating the analysis with the additional interaction term in the regression model. We also performed exploratory analyses decomposing the composite scores for attention/working memory, executive function, and declarative memory into their individual components to understand the relative contribution of each measure to the relationship with neuromelanin scores. To do this, we repeated the regressions replacing the dependent variable with each individual score.

For all regressions, sex was effect coded with males as the reference group and all continuous independent variables were z-scored (i.e. neuromelanin signal intensity scores, age and education), such that we could directly compare the magnitudes of the standardized beta estimates. Significance was determined by an alpha-level of less than 0.05.

## Results

### Reinforcement learning performance is impaired in PD

We compared reinforcement learning parameters between 117 participants with Parkinson’s disease (PD) and 65 Controls who completed the probabilistic learning task (**Supplementary Table 1**). As expected, PD participants had lower positive learning rates (*MPD* = 0.085, *MCTRL* = 0.227, *p* = 0.01) compared to Controls (**Fig. 2**). We also found lower drift rate in PD (*MPD* = 2.421, *MCTRL* = 3.598, *p* = 0.01) but no significant differences in negative learning rate (*MPD* = 0.005, *MCTRL* = 0.007, *p* = 0.345), or decision boundary (*MPD* = 1.922, *MCTRL* = 1.956, *p* = 0.282) relative to Controls.

![Figure 2.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/09/16/2024.09.16.24313753/F2.medium.gif)

[Figure 2.](http://medrxiv.org/content/early/2024/09/16/2024.09.16.24313753/F2)

Figure 2. Reward learning differences between groups.
**(A)** Posterior distributions of the positive learning rates demonstrate that participants with Parkinson’s disease (PD) are estimated to have lower positive learning rates compared to Controls (CTRL). **(B)** The raw data is depicted as an average of correct or incorrect optimal choices (0 or 1) across 10 blocks of 15 trials each.

### Cognitive performance on measures of LC degeneration is impaired in PD

A subset of 61 PD participants and 25 Control participants who completed all neuropsychological tests (for a total of 15 measures) were included in the computation of the domain composite scores (**Supplementary Table 1**). Overall, PD participants had numerically worse performance on all cognitive measures relating to attention/working memory, executive function and memory, but statistically significant differences were found on memory (the HVLT total and delayed) and executive function (TMT B-A, D-KEFS CWIT Inhibition) scores (**Table 2**). PD participants also had numerically worse performance on average in the tests of visuospatial function and language, but group differences were not statistically different.

View this table:
[Table 2.](http://medrxiv.org/content/early/2024/09/16/2024.09.16.24313753/T2)

Table 2. Group performance on neuropsychological assessments

### SN and LC neuromelanin signal intensities are reduced in PD

We investigated the group differences in neuromelanin signal intensities in 81 PD participants and 33 Controls who completed neuroimaging and the reinforcement learning task and/or all 15 neuropsychological measures to be able to relate MRI findings to cognitive performance. We found that PD participants had lower SN neuromelanin signal than Controls (*βgroup* = -0.87, *t* = - 4.46, *p* < 0.001, **Fig. 3**). There was also an effect of sex where men overall had lower SN neuromelanin signal than women (*βsex* = 0.46, *t* = 2.51, *p* = 0.01), but no effects of age or education (*βage* = 0.06, *t* = 0.78, *p* = 0.43; *βeducation* = 0.05, *t* = 0.65, *p* = 0.52). PD participants also had lower LC neuromelanin signal than Controls (*βgroup* = -0.52, *t* = -2.55, *p* = 0.01). Additionally, older age was associated with lower LC neuromelanin signal (*βage* = -0.27, *t* = -3.20, *p* = 0.002) but education and sex were not (*βeducation* = 0.05, *t* = 0.54, *p* = 0.59; *βsex* = 0.34, *t* = 1.78, *p* = 0.08). Overall, PD participants had evidence of greater SN and LC degeneration than Controls, with LC degeneration being additionally predicted by age.

![Figure 3.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/09/16/2024.09.16.24313753/F3.medium.gif)

[Figure 3.](http://medrxiv.org/content/early/2024/09/16/2024.09.16.24313753/F3)

Figure 3. Neuromelanin signal intensity differences between groups.
Participants with Parkinson’s disease (PD) have lower neuromelanin signal intensities compared to Controls (CTRL) in **(A)** the substantia nigra (SN) and **(B)** the locus coeruleus (LC). Box and whisker plots represent the median, interquartile interval, minimum, and maximum.

### Positive reinforcement learning is associated with SN degeneration

To understand the unique effect of SN degeneration in PD on reinforcement learning, we performed a linear regression on positive learning rate with both SN and LC neuromelanin signal intensities as independent variables and age, sex, and education as covariates. A subset of 62 PD participants and 25 Controls completed relevant measures and were included in the analysis (**Supplementary Table 1**). We found that in participants with PD, lower SN neuromelanin signal intensity was associated with lower positive learning rates (*βSN* = 0.41, *t* = 2.46, *p* = 0.02; **Fig. 4A**), but there was no association between LC neuromelanin and positive learning rate (*βLC* = 0.02, *t* = 0.14, *p* = 0.89; **Fig. 4A**). Age, sex, and education were not significant predictors of positive learning rate (*βage* = -0.08, *t* = -0.57, *p* = 0.57; *βsex* = -0.36, *t* = =1.04, *p* = 0.30; *βeducation* = -0.10, *t* = 0.72, *p* = 0.47). An exploratory analysis found no interaction between SN and LC signal intensity on positive learning rate *(βSNxLC* = -0.02, *t* = -0.09, *p* = 0.93, **Supplementary Table 5**). To demonstrate the selectivity of the relationship between SN neuromelanin and positive learning rate, we repeated analyses relating SN and LC neuromelanin to the other cognitive parameters derived from the RLDDM. SN and LC neuromelanin signal intensities were not associated with negative learning rate, drift rate, nor decision boundary (p > 0.29, **Supplementary Table 3**).

![Figure 4.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/09/16/2024.09.16.24313753/F4.medium.gif)

[Figure 4.](http://medrxiv.org/content/early/2024/09/16/2024.09.16.24313753/F4)

Figure 4. Brain-behaviour relationships in Parkinson’s participants.
**(A)** Substantia nigra (SN), but not locus coeruleus (LC), signal intensity was associated with positive learning rate. LC, but not SN, signal intensity was associated with **(B)** attention/working memory (WM) and **(C)** executive functioning. **(D)** Neither SN nor LC signal intensity was associated with memory. Figures demonstrate the standardized beta coefficients after accounting for neuromelanin signal intensity in both the SN and LC, age, sex, and years of education. All variables are standardized.

To contextualize these results, we conducted an exploratory analysis in the smaller sample of Controls (*n* = 25). Surprisingly, higher SN neuromelanin signal intensity was associated with lower positive learning rates (*βSN* = -.49, *t* = -2.44, *p* = 0.02). There was no association between LC neuromelanin signal intensity and positive learning rate (*βLC* = -0.22, *t* = -1.16, *p* = 0.26; **Fig. 4A**) controlling for age, sex, and education (*βage* = 0.24, *t* = -1.16, *p* = 0.25; *βsex* = 0.35, *t* = 0.72, *p* = 0.48; *βeducation* = 0.26, *t* = 1.51, *p* = 0.15). However, the posterior distribution of positive learning rates in this smaller subsample of Controls was not representative of the posterior distribution of positive learning rates in the overall sample of Controls, thus this result should be interpreted with caution (**Supplementary** Fig. 1). Separate regressions also showed that SN and LC neuromelanin signal intensities were not associated with negative learning rate, drift rate, or decision boundary in Controls (p > 0.24, **Supplementary Table 3**).

In summary these results demonstrate that greater SN degeneration was selectively associated with worse positive reinforcement learning in PD participants, and that, importantly, there is no significant association of age with reinforcement learning.

### Attention/working memory and executive function are associated with LC degeneration

Next, we examined the unique relationship between LC neurodegeneration and performance in the three domains of cognition hypothesized to be related to LC in 61 PD participants and 25 Controls who completed all neuropsychological tests and neuroimaging. Models controlled for SN degeneration, in addition to age, sex and education, as above. In PD participants, as predicted, lower LC neuromelanin signal intensity was associated with worse attention/working memory performance (*βLC* = 0.20, *t* = 2.11, *p* = 0.01, **Fig. 4B**) and worse executive function performance (*βLC* = 0.22, *t* = 2.64, *p* = 0.01, **Fig. 4C**), but not with memory (*βLC* = 0.04, *t* = 0.41, *p* = 0.69, **Fig. 4D**). SN neuromelanin was not associated with performance in these three domains (attention/working memory: *βSN* = -0.07, *t* = -0.68, *p* = 0.50; executive function: *βSN* = 0.001, *t* = 0.02, *p* = 0.99; declarative memory: *βSN* = 0.01, *t* = 0.06, *p* = 0.95).

To further confirm the selective effects of SN and LC degeneration on cognition we also examined their relationship to visuospatial function and language. Neither SN nor LC degeneration were associated with performance in these domains (visuospatial: *βSN* = -0.15, *t* = -1.36, *p* = 0.18; *βLC* = 0.09, *t* = 0.87, *p* = 0.39; language: *βSN* = 0.08, *t* = 0.81, *p* = 0.42; *βLC* = 0.17, *t* = 1.70, *p* = 0.10) performance in PD participants. Across all five domains we found that older age was a predictor of worse performance (attention/working memory: *βage* = -0.28, *t* = -3.04, *p* = 0.004; executive function: *βage* = -0.40, *t* = -4.92, *p* < 0.001; memory: *βage* = -0.47, *t* = -4.82, *p* < 0.001; visuospatial: *βage* = -0.38, *t* = -3.54, *p* = 0.001; language: *βage* = -0.31, *t* = -3.23, *p* = 0.002). In all models, the standardized beta estimate for the effect of age on cognitive performance was larger than that of the effect of LC degeneration on cognitive performance. There were no effects of sex nor education on any of the cognitive domains (p > 0.16, see **Supplementary Table 4** for these results). To contextualize these results, we repeated analyses in Control participants. Neither SN nor LC neuromelanin signal intensity were significant predictors of any cognitive domain (p > 0.14, **Supplementary Table 4**).

We also performed exploratory analyses in PD participants predicting performance on each individual neuropsychological measure that fell into a cognitive domain that we hypothesized would relate to LC integrity, i.e., attention/working memory, executive function, and declarative memory. We posited that perhaps the individual scores were not equal in their representation of cognitive performance within their respective domains as is assumed in our composite score. Lower LC signal intensity was associated with worse performance on the D-KEFS CWIT Inhibition condition (*βLC* = 0.23, *t* = 2.01, *p* = 0.05), whereas SN signal intensity did not predict performance (*βSN* = -0.08, *t* = -0.71, *p* = 0.48). However, neither SN or LC neuromelanin signal intensity had a significant effect on Digit Span Forward, Digit Span Backward, TMT A, TMT B-A, BSAT errors, HVLT total and delayed recall, or RCFT immediate and delayed recalls performance (p > 0.07, see **Supplementary Table 6**).

Given that LC, but not SN, signal intensity was associated with performance on attention/working memory and executive function tasks, we performed exploratory analyses investigating whether there was additionally an interaction between LC and SN signal intensity on cognitive performance. We did not find an interaction between LC and SN signal intensity on attention/working memory (*βSNxLC* = -0.03, *t* = -0.29, *p* = 0.77) nor or executive function (*βSNxLC* = -0.08, *t* = -0.83, *p* = 0.41; see **Supplementary Table 5**).

In summary, we found that LC degeneration was selectively associated with impaired attention/working memory and executive function, but not, memory. However, it is also interesting to note that older age was a stronger predictor of worse cognitive performance than the degree of LC degeneration. This was not the case for the relationship between SN degeneration and reinforcement learning where age was not significantly associated with performance.

### Sensitivity analyses

All above analyses assessing brain-behaviour relationships were repeated in the 42 PD participants who completed all assessments, i.e., the reinforcement learning task, the complete neuropsychological assessment, and neuroimaging (**Supplementary Table 1**). The associations between SN neuromelanin and positive reinforcement learning (*βSN* = 0.28, *t* = 1.52, *p* = 0.14) and between LC neuromelanin and attention/working memory (*βLC* = 0.21, *t* = 1.87, *p* = 0.07) and executive function (*βLC* = 0.16, *t* = 1.81, *p* = 0.08) were positive, as above, suggesting that greater loss of neuromelanin was associated with worse performance. However, these associations did not reach significance, possibly due to lack of power in the smaller sample size.

## Discussion

Ascending neuromodulatory systems originating in the SN and LC play a pivotal role in supporting cognitive functions like attention, learning, memory, and executive function. In Parkinson’s disease, both SN and LC degeneration occur early. In the present study, we investigated the independent contributions of degeneration in each of these systems to the early cognitive deficits observed in this population. We found that in a sample of patients tested on their usual dose of dopaminergic medications, loss of neuromelanin in the SN – a marker of degeneration in dopamine neurons – was selectively associated with impaired learning from positive feedback, whereas loss of neuromelanin in the LC – a marker of degeneration noradrenergic neurons – was selectively associated with impairments in attention/working memory and executive function. Neither positive reinforcement learning performance nor SN degeneration were associated with age, whereas both LC degeneration and performance on LC-associated cognitive functions were strongly predicted by age in people with Parkinson’s disease. These results extend findings from pharmacological studies by showing that the severity of neurodegeneration in the SN and LC can be used to predict individual cognitive phenotypes in Parkinson’s disease. Furthermore, these results suggest that the brain-behaviour relationship linking SN degeneration to positive reinforcement learning provides a window into the Parkinson’s-specific and age-independent effect of neurodegeneration on cognition. Given increasing recognition that age-associated pathologies are prevalent in Parkinson’s disease and contribute to cognitive outcomes, these results highlight the potential value of measuring brain-behaviour relationships that are informed by mechanisms to begin to parse the multiple causes of cognitive impairment that likely play a role in Parkinson’s disease.

### SN degeneration is selectively associated with impaired positive reinforcement learning

First, we found that the severity of SN degeneration, as measured by loss of neuromelanin signal, was associated with reduced learning from positive feedback in participants with Parkinson’s disease. This relationship is consistent with the extensive literature demonstrating that reduced dopamine signalling is associated with impaired positive reinforcement learning and that there is a beneficial effect of dopamine replacement therapy on performance, though not all studies show this.30,48,51,53,72 The current findings extend our current understanding of the role of dopamine depletion in cognitive impairment in a few important ways. First, rather than studying the effects of a fixed dose of dopamine replacement, which cannot provide insight into the severity of the underlying dopamine deficit, we used neuromelanin MRI, one of the most promising biomarkers of degeneration in Parkinson’s disease, to establish the severity of degeneration.73–76 Our results therefore suggest that positive reinforcement learning could be used as a behavioural performance-based assessment to track SN degeneration, complimentary to the way motor assessments also correlate with neuromelanin signal.1,40,74,77,78 Second, the relationship between SN neuromelanin signal and positive reinforcement learning was independent of age, and, perhaps even more critical, age was not a predictor of the computationally-derived measure of learning performance we used. This suggests that positive reinforcement learning performance could be useful as a performance-based metric to track the specific Parkinson’s-related, rather that age-related, effects of neurodegeneration.

Of note is that positive reinforcement learning performance was impaired in Parkinson’s participants even though they were on dopaminergic medication at the time of testing. Several previous studies, but not all,49,59,79,80 have shown some degree of remediation of learning deficits in the medicated state.30,31,48–53 One possible explanation for the presence of a learning impairment in the medicated Parkinson’s participants in our sample is that they have somewhat more advanced disease than the typical samples of participants recruited to medication manipulation studies, where the requirement of completing an OFF medications session necessarily biases the sample towards less affected participants. Indeed, this is also consistent with the fact that, apart from patients with very mild disease, dopamine replacement rarely completely remediates the dopamine-sensitive motor symptoms and therefore should not be expected to completely remediate any dopamine-sensitive cognitive symptoms. Demonstrating that reinforcement learning performance captures individual differences in SN neurodegeneration even in medicated patients is a strength as it suggests this measure could be useful in the clinical research setting. Future research could additionally include a measure of the total dose of dopaminergic medications in the models to account for possible confounding effects and obtain a more precise measurement of the relationship between SN degeneration and positive reinforcement learning.

There was no relationship between SN neuromelanin and any other cognitive performance measure, including the other computationally derived measures from the reinforcement learning task. This is largely in keeping with the few other studies that have investigated the relationship between SN degeneration and cognitive performance in participants with Parkinson’s disease, though some results have been inconsistent.1,39,40 For instance, one study found a relationship between SN integrity and working memory39 and another study found a relationship between anteromedial and superior SN integrity and attention.40 These contradictory results may be explained by different working memory and attention tasks used across studies which may be differentially sensitive to deficits. In particular, the Digit Span task we used did not differentiate groups and has not been shown to be sensitive to dopamine replacement, raising questions about whether a relationship to SN integrity is to be expected.81,82 However, most notably, given the strong relationship between LC degeneration and cognitive performance, prior studies have not controlled for the jointly occurring LC degeneration in analyses of the impact of SN degeneration on cognition.

### LC degeneration is selectively associated with attention, working memory and executive function

We found that LC degeneration was associated with impaired attention and executive function in Parkinson’s disease participants, but not with declarative memory, visuospatial, language, nor positive reinforcement learning. This is largely consistent with previous studies1,2,5,54,55 and has been proposed to reflect the fact that frontal cortical regions like the prefrontal cortex (PFC) – that are thought to play a critical role in top-down attention, working memory, and executive function83–87 – are also projection sites of LC neurons,88 suggesting that the neural mechanism by which Parkinson’s disease participants experience these impairments may be, at least in part, via the loss of innervation from the degenerating LC to the PFC. We did not replicate previously reported relationships between LC degeneration and declarative memory deficits, though this has primarily been observed in older adults.1,3,4,89,90 While this could reflect the fact that our memory measures may not have been sufficiently selective (for instance, the RCFT is a measure of visual memory that also engages visuospatial function), this also raises the possibility that memory deficits in Parkinson’s disease primarily reflect involvement of other structures – such as the hippocampus, amygdala, or nucleus basalis of Meynert.91–93

### Limitations

An important limitation of our study was that we did not obtain all measurements in all participants, which limited our ability to examine interactions between both regions of interest and their related domains of cognition. In particular, the sample of older adult controls with complete testing was small thereby limiting our ability to perform meaningful contrasts and to fully account for the effects of age in the Parkinson’s disease participants. The effect of age is especially difficult to estimate in Parkinson’s disease given its relationship to disease duration but is critical to consider because age-associated brain changes also occur in Parkinson’s disease and appear to contribute to the cognitive phenotype in ways that overlap with the effects of neurodegeneration. A second limitation is that we had only one positive reinforcement learning measure. Given that the literature examining dopaminergic medication effects on positive reinforcement learning has revealed conflicting results that might depend on the type of task used, it will be important to replicate the association we found between SN degeneration and positive reinforcement learning. This would also be an opportunity to develop a shorter task, which would be more practical for implementation in clinical studies.

### Summary

The goal of this study was to identify aspects of the cognitive phenotype of Parkinson’s disease that can be selectively attributed to SN and LC degeneration. We found that SN degeneration was selectively associated with positive reinforcement learning impairments and that LC degeneration was selectively associated with impairments in attention and executive function. Identifying dissociable roles of SN and LC degeneration on different cognitive functions in early Parkinson’s disease that reflect the effects of disease above and beyond the effects of age has important implications for developing performance-based measures of cognitive function in Parkinson’s disease that specifically reflect Parkinson’s-related neurodegeneration. These findings also pave the way towards using neurobiologically-grounded performance-based measures of brain function to begin to identify different sources underlying the heterogeneity in the cognitive and behavioural phenotype of Parkinson’s disease.

## Supporting information

Supplementary Materials [[supplements/313753_file03.pdf]](pending:yes)

## Data availability

De-identified participant data is available upon request.

## Funding

This study was supported by funding from the Canadian Institutes of Health Research (MS), Fonds de Recherche Québec – Santé (MS), Healthy Brains for Healthy Lives (SS), and Parkinson Canada (SS). JFG holds a Canada Research Chair in Cognitive Decline in Pathological Aging.

## Competing interests

The authors report no competing interests.

## Acknowledgements

The authors would like to thank the Quebec Parkinson Network and all participants without whom this study would not be possible.

*   Received September 16, 2024.
*   Revision received September 16, 2024.
*   Accepted September 16, 2024.


*   © 2024, Posted by Cold Spring Harbor Laboratory

This pre-print is available under a Creative Commons License (Attribution-NonCommercial-NoDerivs 4.0 International), CC BY-NC-ND 4.0, as described at [http://creativecommons.org/licenses/by-nc-nd/4.0/](http://creativecommons.org/licenses/by-nc-nd/4.0/)

## References

1.  1.Nieoullon A, Coquerel A. Dopamine contribution to cognitive processes: evidence from lesion experiments. Curr Opin Neurol. 2003;16(Suppl 2):S3–S9. [http://journals.lww.com/co-neurology](http://journals.lww.com/co-neurology)
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/00019052-200312002-00002&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=AMBIGUOUS 15&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

2.  2.Cools R. Chemistry of the Adaptive Mind: Lessons from Dopamine. Neuron. 2019;104(1):113–131. doi:10.1016/J.NEURON.2019.09.035
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.neuron.2019.09.035&link_type=DOI) 

3.  3.Berridge CW, Waterhouse BD. The locus coeruleus-noradrenergic system: Modulation of behavioral state and state-dependent cognitive processes. Brain Res Rev. 2003;42(1):33–84. doi:10.1016/S0165-0173(03)00143-7
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S0165-0173(03)00143-7&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=12668290&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000182526300003&link_type=ISI) 

4.  4.Chamberlain SR, Robbins TW. Noradrenergic modulation of cognition: Therapeutic implications. Journal of Psychopharmacology. 2013;27(8):694–718. doi:10.1177/0269881113480988
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1177/0269881113480988&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23518815&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000321826100003&link_type=ISI) 

5.  5.Braak H, Del Tredici K, Rüb U, De Vos RAI, Jansen Steur ENH, Braak E. Staging of brain pathology related to sporadic Parkinson’s disease. Neurobiol Aging. 2003;24(2):197–211. doi:10.1016/S0197-4580(02)00065-9
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S0197-4580(02)00065-9&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=12498954&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000180616100001&link_type=ISI) 

6.  6.Hawkes CH, Del Tredici K, Braak H. A timeline for Parkinson’s disease. Parkinsonism Relat Disord. 2010;16(2):79–84. doi:10.1016/j.parkreldis.2009.08.007
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.parkreldis.2009.08.007&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=19846332&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000275617700001&link_type=ISI) 

7.  7.Goedert M, Spillantini MG, Del Tredici K, Braak H. 100 years of Lewy pathology. Nat Rev Neurol. 2013;9(1):13–24. doi:10.1038/nrneurol.2012.242
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/nrneurol.2012.242&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23183883&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

8.  8.Halliday GM, Leverenz JB, Schneider JS, Adler CH. The Neurobiological Basis of Cognitive Impairment in Parkinson’s Disease. Mov Disord. 2014;29(5):634. doi:10.1002/MDS.25857
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/mds.25857&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=24757112&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

9.  9.Aarsland D, Batzu L, Halliday GM, et al. Parkinson disease-associated cognitive impairment. Nat Rev Dis Primers. 2021;7(1):1–21. doi:10.1038/s41572-021-00280-3
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41572-021-00280-3&link_type=DOI) 

10. 10.Cassidy CM, Zucca FA, Girgis RR, et al. Neuromelanin-sensitive MRI as a noninvasive proxy measure of dopamine function in the human brain. Proc Natl Acad Sci U S A. 2019;116(11):5108–5117. doi:10.1073/pnas.1807983116
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoicG5hcyI7czo1OiJyZXNpZCI7czoxMToiMTE2LzExLzUxMDgiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyNC8wOS8xNi8yMDI0LjA5LjE2LjI0MzEzNzUzLmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 

11. 11.Betts MJ, Kirilina E, Otaduy MCG, et al. Locus coeruleus imaging as a biomarker for noradrenergic dysfunction in neurodegenerative diseases. Brain. 2019;142(9):2558–2571. doi:10.1093/brain/awz193
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/awz193&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=31327002&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

12. 12.Li Y, Wang C, Wang J, et al. Mild cognitive impairment in de novo Parkinson’s disease: A neuromelanin MRI study in locus coeruleus. Movement Disorders. 2019;34(6):884–892. doi:10.1002/mds.27682
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/mds.27682&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=30938892&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

13. 13.Reimão S, Pita Lobo P, Neutel D, et al. Substantia nigra neuromelanin magnetic resonance imaging in de novo Parkinson’s disease patients. Eur J Neurol. 2015;22(3):540–546. doi:10.1111/ENE.12613
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/ene.12613&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=25534480&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

14. 14.Wang J, Li Y, Huang Z, et al. Neuromelanin-sensitive magnetic resonance imaging features of the substantia nigra and locus coeruleus in de novo Parkinson’s disease and its phenotypes. Eur J Neurol. 2018;25(7):949–e73. doi:10.1111/ENE.13628
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/ene.13628&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=29520900&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

15. 15.Owen AM, James M, Leigh PN, et al. FRONTO-STRIATAL COGNITIVE DEFICITS AT DIFFERENT STAGES OF PARKINSON’S DISEASE. Brain. 1992;115(6):1727–1751. doi:10.1093/BRAIN/115.6.1727
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/115.6.1727&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=1486458&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1992KJ44600007&link_type=ISI) 

16. 16.Cools R, Barker RA, Sahakian BJ, Robbins TW. Enhanced or Impaired Cognitive Function in Parkinson’s Disease as a Function of Dopaminergic Medication and Task Demands. Cerebral Cortex. 2001;11(12):1136–1143.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/cercor/11.12.1136&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=11709484&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000172538400004&link_type=ISI) 

17. 17.Marsh L, Biglan K, Gerstenhaber M, Williams JR. Atomoxetine for the treatment of executive dysfunction in Parkinson’s disease: A pilot open-label study. Mov Disord. 2009;24(2):277–282. doi:10.1002/MDS.22307
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/mds.22307&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=19025777&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

18. 18.Weintraub D, Chahine LM, Hawkins KA, et al. Cognition and the course of prodromal Parkinson’s disease. Mov Disord. 2017;32(11):1640–1645. doi:10.1002/MDS.27189
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/MDS.27189&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

19. 19.Muslimovíc D, Post B, Speelman JD, Schmand B. Cognitive profile of patients with newly diagnosed Parkinson disease. Neurology. 2005;65(8):1239–1245. [www.neurology.org](https://www.neurology.org)
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1212/01.wnl.0000180516.69442.95&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=16247051&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

20. 20.Lewis SJG, Slabosz A, Robbins TW, Barker RA, Owen AM. Dopaminergic basis for deficits in working memory but not attentional set-shifting in Parkinson’s disease. Neuropsychologia. 2005;43(6):823–832. doi:10.1016/J.NEUROPSYCHOLOGIA.2004.10.001
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.neuropsychologia.2004.10.001&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=15716155&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000227966900001&link_type=ISI) 

21. 21.Moustafa AA, Cohen MX, Sherman SJ, Frank MJ. A role for dopamine in temporal decision making and reward maximization in Parkinsonism. Journal of Neuroscience. 2008;28(47):12294–12304. doi:10.1523/JNEUROSCI.3116-08.2008
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Njoiam5ldXJvIjtzOjU6InJlc2lkIjtzOjExOiIyOC80Ny8xMjI5NCI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI0LzA5LzE2LzIwMjQuMDkuMTYuMjQzMTM3NTMuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

22. 22.Fallon SJ, Mattiesing RM, Muhammed K, Manohar S, Husain M. Fractionating the Neurocognitive Mechanisms Underlying Working Memory: Independent Effects of Dopamine and Parkinson’s Disease. Cerebral Cortex. 2017;27(12):5727–5738. doi:10.1093/CERCOR/BHX242
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/CERCOR/BHX242&link_type=DOI) 

23. 23.Simioni AC, Dagher A, Fellows LK. Effects of levodopa on corticostriatal circuits supporting working memory in Parkinson’s disease. Cortex. 2017;93:193–205. doi:10.1016/J.CORTEX.2017.05.021
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.cortex.2017.05.021&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=28675834&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

24. 24.Fournet N, Moreaud O, Roulin JL, Naegele B, Pellat J. Working memory functioning in medicated Parkinson’s disease patients and the effect of withdrawal of dopaminergic medication. Neuropsychology. 2000;14(2):247–253. doi:10.1037/0894-4105.14.2.247
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1037/0894-4105.14.2.247&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=10791864&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000087480700008&link_type=ISI) 

25. 25.Gotham A, Brown R, Marsden C. “FRONTAL” COGNITIVE FUNCTION IN PATIENTS WITH PARKINSON’S DISEASE “ON” AND “OFF” LEVODOPA. Brain. 1988;111(2):299–321. doi:10.1093/brain/111.2.299
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/111.2.299&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=3378138&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1988N657400005&link_type=ISI) 

26. 26.Bódi N, Kéri S, Nagy H, et al. Reward-learning and the novelty-seeking personality: A between-and within-subjects study of the effects of dopamine agonists on young parkinsons patients. Brain. 2009;132(9):2385–2395. doi:10.1093/brain/awp094
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/awp094&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=19416950&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000269963600010&link_type=ISI) 

27. 27.Rutledge RB, Lazzaro SC, Lau B, Myers CE, Gluck MA, Glimcher PW. Dopaminergic drugs modulate learning rates and perseveration in Parkinson’s patients in a dynamic foraging task. Journal of Neuroscience. 2009;29(48):15104–15114. doi:10.1523/JNEUROSCI.3524-09.2009
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Njoiam5ldXJvIjtzOjU6InJlc2lkIjtzOjExOiIyOS80OC8xNTEwNCI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI0LzA5LzE2LzIwMjQuMDkuMTYuMjQzMTM3NTMuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

28. 28.McGuigan S, Zhou SH, Brosnan MB, Thyagarajan D, Bellgrove MA, Chong TTJ. Dopamine restores cognitive motivation in Parkinson’s disease. Brain. 2019;142(3):719–732. doi:10.1093/brain/awy341
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/awy341&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=30689734&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

29. 29.MacDonald AA, Seergobin KN, Owen AM, et al. Differential Effects of Parkinson’s Disease and Dopamine Replacement on Memory Encoding and Retrieval. PLoS One. 2013;8(9):e74044. doi:10.1371/JOURNAL.PONE.0074044
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1371/JOURNAL.PONE.0074044&link_type=DOI) 

30. 30.Grogan JP, Tsivos D, Smith L, et al. Effects of dopamine on reinforcement learning and consolidation in parkinson’s disease. Elife. 2017;6:e26801. doi:10.7554/ELIFE.26801
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.7554/eLife.26801&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=28691905&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

31. 31.Surmeier DJ, Graves SM, Shen W. Dopaminergic modulation of striatal networks in health and Parkinson’s disease. Curr Opin Neurobiol. 2014;29:109–117. doi:10.1016/J.CONB.2014.07.008
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.conb.2014.07.008&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=25058111&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

32. 32.Chung SJ, Yoo HS, Oh JS, et al. Effect of striatal dopamine depletion on cognition in de novo Parkinson’s disease. Parkinsonism Relat Disord. 2018;51:43–48. doi:10.1016/j.parkreldis.2018.02.048
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.parkreldis.2018.02.048&link_type=DOI) 

33. 33.Nobili F, Campus C, Arnaldi D, et al. Cognitive-nigrostriatal relationships in de novo, drug-naïve Parkinson’s disease patients: A [I-123]FP-CIT SPECT Study. Movement Disorders. 2010;25(1):35–43. doi:10.1002/mds.22899
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/mds.22899&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=20058228&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

34. 34.Siepel FJ, Brønnick KS, Booij J, et al. Cognitive executive impairment and dopaminergic deficits in de novo Parkinson’s disease. Movement Disorders. 2014;29(14):1802–1808. doi:10.1002/mds.26051
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/mds.26051&link_type=DOI) 

35. 35.Liu W, Wang C, He T, et al. Substantia Nigra Integrity Correlates with Sequential Working Memory in Parkinson’s Disease. Journal of Neuroscience. 2021;41(29):6304–6313. doi:10.1523/JNEUROSCI.0242-21.2021
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Njoiam5ldXJvIjtzOjU6InJlc2lkIjtzOjEwOiI0MS8yOS82MzA0IjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjQvMDkvMTYvMjAyNC4wOS4xNi4yNDMxMzc1My5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

36. 36.Biondetti E, Gaurav R, Yahia-Cherif L, et al. Spatiotemporal changes in substantia nigra neuromelanin content in Parkinson’s disease. Brain. 2020;143(9):2757–2770. doi:10.1093/brain/awaa216
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/awaa216&link_type=DOI) 

37. 37.Vriend C, van Balkom TD, van Druningen C, et al. Processing speed is related to striatal dopamine transporter availability in Parkinson’s disease. Neuroimage Clin. 2020;26:102257. doi:10.1016/j.nicl.2020.102257
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.nicl.2020.102257&link_type=DOI) 

38. 38.Rinne JO, Portin R, Ruottinen H, et al. Cognitive Impairment and the Brain Dopaminergic System in Parkinson Disease [18 F]Fluorodopa Positron Emission Tomographic Study. Arch Neurol. 2000;57:470–475.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1001/archneur.57.4.470&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=10768619&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000086474700003&link_type=ISI) 

39. 39.Prasuhn J, Prasuhn M, Fellbrich A, et al. Association of Locus Coeruleus and Substantia Nigra Pathology With Cognitive and Motor Functions in Patients With Parkinson Disease. Neurology. 2021;97(10):e1007–e1016. doi:10.1212/WNL.0000000000012444
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1212/WNL.0000000000012444&link_type=DOI) 

40. 40.Litvan I, Goldman JG, Tröster AI, et al. Diagnostic Criteria for Mild Cognitive Impairment in Parkinson’s Disease: Movement Disorder Society Task Force Guidelines. Mov Disord. 2012;27(3):349–356. doi:10.1002/MDS.24893
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/mds.24893&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=22275317&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

41. 41.Meder D, Herz DM, Rowe JB, Lehéricy S, Siebner HR. The role of dopamine in the brain - lessons learned from Parkinson’s disease. Neuroimage. 2019;190:79–93. doi:10.1016/j.neuroimage.2018.11.021
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.neuroimage.2018.11.021&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=30465864&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

42. 42.Robbins TW, Cools R. Cognitive deficits in Parkinson’s disease: A cognitive neuroscience perspective. Movement Disorders. 2014;29(5):597–607. doi:10.1002/mds.25853
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/mds.25853&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=24757109&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

43. 43.Schultz W, Dayan P, Montague PR. A neural substrate of prediction and reward. Science (1979). 1997;275(5306):1593–1599. doi:10.1126/SCIENCE.275.5306.1593/ASSET/6CC77AD2-EA4B-4861-A4ED-A076123F94E0/ASSETS/GRAPHIC/SE1174905004.JPEG
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Mzoic2NpIjtzOjU6InJlc2lkIjtzOjEzOiIyNzUvNTMwNi8xNTkzIjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjQvMDkvMTYvMjAyNC4wOS4xNi4yNDMxMzc1My5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

44. 44.Schultz W. Dopamine reward prediction-error signalling: a two-component response. Nature Reviews *Neuroscience 2016 17:3*. 2016;17(3):183–195. doi:10.1038/nrn.2015.26
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/nrn.2015.26&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=26865020&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

45. 45.Frank MJ, Seeberger LC, O’Reilly RC. By carrot or by stick: Cognitive reinforcement learning in Parkinsonism. Science (1979). 2004;306(5703):1940–1943. doi:10.1126/science.1102941
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Mzoic2NpIjtzOjU6InJlc2lkIjtzOjEzOiIzMDYvNTcwMy8xOTQwIjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjQvMDkvMTYvMjAyNC4wOS4xNi4yNDMxMzc1My5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

46. 46.Graef S, Biele G, Kruge LK, et al. Differential influence of levodopa on reward-based learning in Parkinson’s disease. Front Hum Neurosci. 2010;4:169. doi:10.3389/FNHUM.2010.00169/BIBTEX
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.3389/FNHUM.2010.00169/BIBTEX&link_type=DOI) 

47. 47.Muhammed K, Manohar S, Ben Yehuda M, et al. Reward sensitivity deficits modulated by dopamine are associated with apathy in Parkinson’s disease. Brain. 2016;139(10):2706–2721. doi:10.1093/brain/aww188
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/aww188&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=27452600&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

48. 48.Sharp ME, Foerde K, Daw ND, Shohamy D. Dopamine selectively remediates “model-based” reward learning: A computational approach. Brain. 2016;139(2):355–364. doi:10.1093/brain/awv347
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/awv347&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=26685155&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

49. 49.McCoy B, Jahfari S, Engels G, Knapen T, Theeuwes J. Dopaminergic medication reduces striatal sensitivity to negative outcomes in Parkinson’s disease. Brain. 2019;142(11):3605–3620. doi:10.1093/brain/awz276
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/awz276&link_type=DOI) 

50. 50.Schmidt L, Braun EK, Wager TD, Shohamy D. Mind matters: Placebo enhances reward learning in Parkinson’s disease. Nat Neurosci. 2014;17(12):1793–1797. doi:10.1038/NN.3842
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/nn.3842&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=25326691&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

51. 51.Ye R, Hezemans FH, O’Callaghan C, et al. Locus coeruleus integrity is linked to response inhibition deficits in Parkinson’s disease and progressive supranuclear palsy. Journal of Neuroscience. 2023;43(42):7028–7040. doi:10.1523/JNEUROSCI.0289-22.2023
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Njoiam5ldXJvIjtzOjU6InJlc2lkIjtzOjEwOiI0My80Mi83MDI4IjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjQvMDkvMTYvMjAyNC4wOS4xNi4yNDMxMzc1My5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

52. 52.Ye R, O’Callaghan C, Rua C, et al. Locus Coeruleus Integrity from 7 T MRI Relates to Apathy and Cognition in Parkinsonian Disorders. Movement Disorders. 2022;37(8):1663–1672. doi:10.1002/mds.29072
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/mds.29072&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=35576973&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

53. 53.Sommerauer M, Fedorova TD, Hansen AK, et al. Evaluation of the noradrenergic system in Parkinson’s disease: An 11 C-MeNER PET and neuromelanin MRI study. Brain. 2018;141(2):496–504. doi:10.1093/brain/awx348
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/awx348&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=29272343&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

54. 54.Ranjbar-Slamloo Y, Fazlali Z. Dopamine and Noradrenaline in the Brain; Overlapping or Dissociate Functions? Front Mol Neurosci. 2020;12:334. doi:10.3389/FNMOL.2019.00334/BIBTEX
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.3389/FNMOL.2019.00334/BIBTEX&link_type=DOI) 

55. 55.Mather M, Harley CW. The Locus Coeruleus: Essential for Maintaining Cognitive Function and the Aging Brain. Trends Cogn Sci. 2016;20(3):214–226. doi:10.1016/J.TICS.2016.01.001
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.tics.2016.01.001&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=26895736&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

56. 56.Sharp ME, Duncan K, Foerde K, Shohamy D. Dopamine is associated with prioritization of reward-associated memories in Parkinson’s disease. Brain. 2020;143(8):2519–2531. doi:10.1093/brain/awaa182
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/awaa182&link_type=DOI) 

57. 57.Peirce J, Gray JR, Simpson S, et al. PsychoPy2: Experiments in behavior made easy. Behav Res Methods. 2019;51(1):195. doi:10.3758/S13428-018-01193-Y
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.3758/s13428-018-01193-y&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

58. 58.Blechert J, Lender A, Polk S, Busch NA, Ohla K. Food-pics_extended-an image database for experimental research on eating and appetite: Additional images, normative ratings and an updated review. Front Psychol. 2019;10:307. doi:10.3389/FPSYG.2019.00307/BIBTEX
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.3389/FPSYG.2019.00307/BIBTEX&link_type=DOI) 

59. 59.Goldman JG, Holden S, Ouyang B, Bernard B, Goetz CG, Stebbins GT. Diagnosing PD-MCI by MDS Task Force criteria: how many and which neuropsychological tests? Mov Disord. 2015;30(3):402–406. doi:10.1002/MDS.26084
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/mds.26084&link_type=DOI) 

60. 60.Hämmerer D, Callaghan MF, Hopkins A, et al. Locus coeruleus integrity in old age is selectively related to memories linked with salient negative events. Proc Natl Acad Sci U S A. 2018;115(9):2228–2233. doi:10.1073/PNAS.1712268115/-/DCSUPPLEMENTAL
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoicG5hcyI7czo1OiJyZXNpZCI7czoxMDoiMTE1LzkvMjIyOCI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI0LzA5LzE2LzIwMjQuMDkuMTYuMjQzMTM3NTMuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

61. 61.Dahl MJ, Mather M, Düzel S, et al. Higher rostral locus coeruleus integrity is associated with better memory performance in older adults. Nat Hum Behav. 2019;3(11):1203–1214. doi:10.1038/s41562-019-0715-2.Higher
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41562-019-0715-2.Higher&link_type=DOI) 

62. 62.O’Callaghan C, Hezemans FH, Ye R, et al. Locus coeruleus integrity and the effect of atomoxetine on response inhibition in Parkinson’s disease. Brain. 2021;144(8):2513–2526. doi:10.1093/BRAIN/AWAB142
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/brain/awab142&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

63. 63.Vincent RD, Neelin P, Khalili-Mahani N, et al. MINC 2.0: A flexible format for multi-modal images. Front Neuroinform. 2016;10:35. doi:10.3389/fninf.2016.00035
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.3389/fninf.2016.00035&link_type=DOI) 

64. 64.Aubert-Broche B, Fonov VS, García-Lorenzo D, et al. A new method for structural volume analysis of longitudinal brain MRI data and its application in studying the growth trajectories of anatomical brain structures in childhood. Neuroimage. 2013;82:393–402. doi:10.1016/J.NEUROIMAGE.2013.05.065
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.neuroimage.2013.05.065&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23719155&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

65. 65.Madge V, Fonov VS, Xiao Y, et al. A dataset of multi-contrast unbiased average MRI templates of a Parkinson’s disease population. Data Brief. 2023;48:109141. doi:10.1016/J.DIB.2023.109141
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/J.DIB.2023.109141&link_type=DOI) 

66. 66.Bianciardi M, Toschi N, Edlow BL, et al. Toward an In Vivo Neuroimaging Template of Human Brainstem Nuclei of the Ascending Arousal, Autonomic, and Motor Systems. Brain Connect. 2015;5(10):597–607. doi:10.1089/BRAIN.2015.0347/ASSET/IMAGES/LARGE/FIGURE7.JPEG
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1089/brain.2015.0347&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=26066023&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

67. 67.Pessiglione M, Seymour B, Flandin G, Dolan RJ, Frith CD. Dopamine-dependent prediction errors underpin reward-seeking behaviour in humans. Nature. 2006;442(7106):1042–1045. doi:10.1038/nature05051
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/nature05051&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=16929307&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000240142000050&link_type=ISI) 

68. 68.Wiecki T V., Sofer I, Frank MJ. HDDM: Hierarchical bayesian estimation of the drift-diffusion model in Python. Front Neuroinform. 2013;7(14):1–10. doi:10.3389/FNINF.2013.00014/BIBTEX
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.3389/FNINF.2013.00014/BIBTEX&link_type=DOI) 

69. 69.Pedersen ML, Frank MJ, Biele G. The drift diffusion model as the choice rule in reinforcement learning. Psychon Bull Rev. 2017;24(4):1234–1251. doi:10.3758/S13423-016-1199-Y
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.3758/s13423-016-1199-y&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=27966103&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

70. 70.Pedersen ML, Frank MJ. Simultaneous Hierarchical Bayesian Parameter Estimation for Reinforcement Learning and Drift Diffusion Models: a Tutorial and Links to Neural Data. Comput Brain Behav. 2020;3(4):458–471. doi:10.1007/S42113-020-00084-W/FIGURES/7
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/S42113-020-00084-W/FIGURES/7&link_type=DOI) 

71. 71.McDougle SD, Collins AGE. Modeling the influence of working memory, reinforcement, and action uncertainty on reaction time and choice during instrumental learning. Psychon Bull Rev. 2021;28(1):20–39. doi:10.3758/S13423-020-01774-Z/FIGURES/9
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.3758/S13423-020-01774-Z/FIGURES/9&link_type=DOI) 

72. 72.Costello H, Berry AJ, Reeves S, et al. Disrupted reward processing in Parkinson’s disease and its relationship with dopamine state and neuropsychiatric syndromes: a systematic review and meta-analysis. J Neurol Neurosurg Psychiatry. 2022;93(5):555–562. doi:10.1136/jnnp-2021-327762
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiam5ucCI7czo1OiJyZXNpZCI7czo4OiI5My81LzU1NSI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI0LzA5LzE2LzIwMjQuMDkuMTYuMjQzMTM3NTMuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

73. 73.Mitchell T, Lehéricy S, Chiu SY, Strafella AP, Stoessl AJ, Vaillancourt DE. Emerging Neuroimaging Biomarkers Across Disease Stage in Parkinson Disease: A Review. JAMA Neurol. 2021;78(10):1262–1272. doi:10.1001/JAMANEUROL.2021.1312
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1001/JAMANEUROL.2021.1312&link_type=DOI) 

74. 74.Xing Y, Sapuan AH, Martín-Bastida A, et al. Neuromelanin-MRI to Quantify and Track Nigral Depigmentation in Parkinson’s Disease: A Multicenter Longitudinal Study Using Template-Based Standardized Analysis. Movement Disorders. 2022;37(5):1028–1039. doi:10.1002/mds.28934
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/mds.28934&link_type=DOI) 

75. 75.Gaurav R, Yahia-Cherif L, Pyatigorskaya N, et al. Longitudinal Changes in Neuromelanin MRI Signal in Parkinson’s Disease: A Progression Marker. Movement Disorders. 2021;36(7):1592–1602. doi:10.1002/MDS.28531
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/MDS.28531&link_type=DOI) 

76. 76.Sulzer D, Cassidy C, Horga G, et al. Neuromelanin detection by magnetic resonance imaging (MRI) and its promise as a biomarker for Parkinson’s disease. npj Parkinson’s Disease 2018 4:1. 2018;4(1):1–13. doi:10.1038/s41531-018-0047-3
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41531-018-0047-3&link_type=DOI) 

77. 77.Cotzias GC, Papavasiliou PS, Gellene R. Modification of Parkinsonism — Chronic Treatment with L-Dopa. N Engl J Med. 1969;280(7):337–345. doi:10.1056/NEJM196902132800701
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1056/NEJM196902132800701&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=4178641&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1969C553100001&link_type=ISI) 

78. 78.Cotzias GC, Van Woert MH, Schiffer LM. Aromatic Amino Acids and Modification of Parkinsonism. N Engl J Med. 1967;276(7):374-379. doi:10.1056/NEJM196702162760703
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1056/NEJM196702162760703&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=5334614&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A19678898500003&link_type=ISI) 

79. 79.Cavanagh JF, Mueller AA, Brown DR, et al. Cognitive states influence dopamine-driven aberrant learning in Parkinson’s disease. Cortex. 2017;90:115–124. doi:10.1016/J.CORTEX.2017.02.021
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/J.CORTEX.2017.02.021&link_type=DOI) 

80. 80.Czernecki V, Pillon B, Houeto JL, Pochon JB, Levy R, Dubois B. Motivation, reward, and Parkinson’s disease: influence of dopatherapy. Neuropsychologia. 2002;40(13):2257–2267. doi:10.1016/S0028-3932(02)00108-2
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S0028-3932(02)00108-2&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=12417456&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000179716300006&link_type=ISI) 

81. 81.Grogan JP, Knight LE, Smith L, et al. Effects of Parkinson’s disease and dopamine on digit span measures of working memory. Psychopharmacology (Berl*)*. 2018;235(12):3443–3450. doi:10.1007/S00213-018-5058-6/FIGURES/2
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/S00213-018-5058-6/FIGURES/2&link_type=DOI) 

82. 82.Warden C, Hwang J, Marshall A, Fenesy M, Poston KL. The effects of dopamine on digit span in Parkinson’s disease. J Clin Mov Disord. 2016;3(5). doi:10.1186/S40734-016-0033-Z
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1186/S40734-016-0033-Z&link_type=DOI) 

83. 83.Buschman TJ, Miller EK. Top-down versus bottom-up control of attention in the prefrontal and posterior parietal cortices. Science (1979). 2007;315(5820):1860–1864. doi:10.1126/SCIENCE.1138071/SUPPL_FILE/BUSCHMAN.SOM.PDF
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Mzoic2NpIjtzOjU6InJlc2lkIjtzOjEzOiIzMTUvNTgyMC8xODYwIjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjQvMDkvMTYvMjAyNC4wOS4xNi4yNDMxMzc1My5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

84. 84.Curtis CE, D’Esposito M. Persistent activity in the prefrontal cortex during working memory. Trends Cogn Sci. 2003;7(9):415–423. doi:10.1016/S1364-6613(03)00197-9
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S1364-6613(03)00197-9&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=12963473&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000185374800010&link_type=ISI) 

85. 85.Voytek B, Knight RT. Prefrontal cortex and basal ganglia contributions to visual working memory. Proc Natl Acad Sci U S A. 2010;107(42):18167–18172. doi:10.1073/PNAS.1007277107/SUPPL_FILE/PNAS.201007277SI.PDF
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoicG5hcyI7czo1OiJyZXNpZCI7czoxMjoiMTA3LzQyLzE4MTY3IjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjQvMDkvMTYvMjAyNC4wOS4xNi4yNDMxMzc1My5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

86. 86.Gazzaley A, Nobre AC. Top-down modulation: Bridging selective attention and working memory. Trends Cogn Sci. 2012;16(2):129–135. doi:10.1016/J.TICS.2011.11.014
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.tics.2011.11.014&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=22209601&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000300803500012&link_type=ISI) 

87. 87.Friedman NP, Robbins TW. The role of prefrontal cortex in cognitive control and executive function. Neuropsychopharmacology. 2022;47(1):72–89. doi:10.1038/S41386-021-01132-0
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41386-021-01132-0&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=34408280&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 

88. 88.Chandler DJ, Gao WJ, Waterhouse BD. Heterogeneous organization of the locus coeruleus projections to prefrontal and motor cortices. Proc Natl Acad Sci U S A. 2014;111(18):6816–6821. doi:10.1073/PNAS.1320827111/-/DCSUPPLEMENTAL/PNAS.201320827SI.PDF
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoicG5hcyI7czo1OiJyZXNpZCI7czoxMToiMTExLzE4LzY4MTYiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyNC8wOS8xNi8yMDI0LjA5LjE2LjI0MzEzNzUzLmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 

89. 89.Dahl MJ, Bachman SL, Dutt S, et al. The integrity of dopaminergic and noradrenergic brain regions is associated with different aspects of late-life memory performance. Nature Aging 2023 3:9. 2023;3(9):1128–1143. doi:10.1038/s43587-023-00469-z
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s43587-023-00469-z&link_type=DOI) 

90. 90.Sara SJ. The locus coeruleus and noradrenergic modulation of cognition. Nat Rev Neurosci. 2009;10(3):211–223. doi:10.1038/nrn2573
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/nrn2573&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=19190638&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000263556500014&link_type=ISI) 

91. 91.Beyer MK, Bronnick KS, Hwang KS, et al. Verbal memory is associated with structural hippocampal changes in newly diagnosed Parkinson’s disease. J Neurol Neurosurg Psychiatry. 2013;84(1):23–28. doi:10.1136/JNNP-2012-303054
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiam5ucCI7czo1OiJyZXNpZCI7czo3OiI4NC8xLzIzIjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjQvMDkvMTYvMjAyNC4wOS4xNi4yNDMxMzc1My5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

92. 92.Citro S,  Lazzaro G Di, Cimmino AT, Giuffrè GM, Marra C, Calabresi P. A multiple hits hypothesis for memory dysfunction in Parkinson disease. Nature Reviews Neurology 2023 20:1. 2023;20(1):50–61. doi:10.1038/s41582-023-00905-z
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41582-023-00905-z&link_type=DOI) 

93. 93.Calabresi P, Picconi B, Parnetti L, Di Filippo M. A convergent model for cognitive dysfunctions in Parkinson’s disease: the critical dopamine–acetylcholine synaptic balance. Lancet Neurol. 2006;5(11):974–983. doi:10.1016/S1474-4422(06)70600-7
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S1474-4422(06)70600-7&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=17052664&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F09%2F16%2F2024.09.16.24313753.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000241591600020&link_type=ISI)

 [1]: /embed/graphic-3.gif