Noradrenergic deficits contribute to apathy in Parkinson's disease through the precision of expected outcomes

Frank H. Hezemans, Noham Wolpe, Claire O'Callaghan, Rong Ye, Catarina Rua, P. Simon Jones, Alexander G. Murley, Negin Holland, Ralf Regenthal, Kamen A. Tsvetanov, Roger A. Barker, Caroline H. Williams-Gray, Trevor W. Robbins, Luca Passamonti, James B. RowePLOS Computational Biology 2022 · DOI

TL;DR

This study investigates the role of noradrenaline in apathy, a common symptom in Parkinson's disease. By combining a visuomotor task with atomoxetine (a noradrenaline reuptake inhibitor) and advanced MRI, the researchers found that noradrenergic function influences how precisely individuals weigh prior beliefs against sensory evidence, and that this mechanism is particularly relevant for those with more severe locus coeruleus degeneration.

What Problem Does This Paper Address?

Apathy, characterized by a lack of goal-directed behavior, is a significant and often treatment-resistant symptom in various neuropsychiatric conditions, including Parkinson's disease (PD). While dopamine has been a primary focus for understanding apathy in PD, its limitations (e.g., apathy persisting despite dopamine replacement therapy) have prompted research into other neurotransmitter systems. Noradrenaline, originating from the locus coeruleus (LC), has been implicated in signaling uncertainty and adapting to environmental changes. This paper explores the hypothesis that noradrenergic deficits contribute to apathy by altering the precision with which individuals weigh their prior beliefs about action outcomes against incoming sensory information.

Methods

The study employed a multi-faceted approach:

  • Visuomotor Task: Participants performed a task requiring them to control a ball's trajectory to hit a target. In some trials, the trajectory was hidden, and participants had to estimate the ball's final position. This task allowed for the estimation of "prior weighting," a measure reflecting the relative precision of prior beliefs versus sensory evidence.
  • Pharmacological Intervention: Participants with PD received either atomoxetine (a selective noradrenaline reuptake inhibitor) or a placebo in a double-blind, crossover design.
  • Neuroimaging: Ultra-high field (7T) magnetisation transfer imaging was used to assess the structural integrity of the locus coeruleus (LC), a key source of noradrenaline in the brain.
  • Bayesian Modeling: A hierarchical Bayesian model was used to decompose the effects of atomoxetine on prior weighting into changes in prior precision and sensory evidence precision. This allowed for a more nuanced understanding of how the drug affected the underlying computational processes.
  • Statistical Analysis: Frequentist and Bayesian statistical methods were used to analyze the data, including linear mixed-effects models and ANCOVAs, to assess relationships between apathy, LC integrity, drug effects, and prior weighting.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • Prior Weighting and Apathy Link: The study confirmed a negative correlation between prior weighting and apathy, suggesting that individuals with higher apathy rely less on their prior beliefs.
  • Atomoxetine's Baseline-Dependent Effect: Atomoxetine's impact on prior weighting was not uniform. It was most pronounced in individuals with lower locus coeruleus integrity, indicating a baseline-dependent effect.
  • Noradrenaline's Role in Belief Precision: The findings strongly suggest that noradrenaline, via the LC, plays a crucial role in modulating the precision of prior beliefs, rather than sensory evidence precision, in the context of goal-directed behavior.
  • Apathy as a Moderator: The effect of LC integrity on atomoxetine's impact on prior weighting was moderated by apathy levels, with more apathetic individuals showing a stronger response.

Key Results

  • Prior Weighting and Apathy: A significant negative association was found between prior weighting and apathy (β = -0.35, p = .028).
  • Atomoxetine and LC Integrity Interaction: A significant interaction effect between atomoxetine and locus coeruleus CNR (contrast-to-noise ratio) on prior weighting was observed (β = -0.38, p = .013), indicating that individuals with lower LC integrity showed a greater increase in prior weighting with atomoxetine compared to placebo.
  • Decomposition of Effects: The drug-induced change in prior weighting was primarily explained by changes in prior precision (β = -0.66, p = .004), not sensory evidence precision.
  • Apathy Moderation: A significant interaction between LC CNR and apathy on the drug effect on prior weighting was found (β = -0.49, p = .036).

Limitations and Future Work

  • Sample Size: The authors acknowledge that the modest sample size (N=17 for PD) may limit the power for data-driven analyses and predictive modeling, potentially leading to type II errors.
  • Apathy Severity: The PD participants were not severely apathetic, and there was no significant difference in apathy between PD and control groups. Generalizability to more severely apathetic cohorts needs further investigation.
  • Pharmacological Complexity: The potential for atomoxetine to affect dopamine levels, given its mechanism of action and the close relationship between noradrenaline and dopamine systems, cannot be entirely excluded. Future studies could use crossover designs with both noradrenergic and dopaminergic agents.
  • Future Directions: The authors suggest that locus coeruleus imaging could serve as a marker for noradrenergic function to inform stratified clinical trials of noradrenergic therapies in selected patients. Further research is needed to determine if atomoxetine can improve everyday functioning and to explore the computational underpinnings of Bayesian perceptual inference in more detail.

Summary updated

BibTeX
@article{hezemans2022noradrenergic,
  author = {Frank H. Hezemans and Noham Wolpe and Claire O'Callaghan and Rong Ye and Catarina Rua and P. Simon Jones and Alexander G. Murley and Negin Holland and Ralf Regenthal and Kamen A. Tsvetanov and Roger A. Barker and Caroline H. Williams-Gray and Trevor W. Robbins and Luca Passamonti and James B. Rowe},
  title = {{Noradrenergic deficits contribute to apathy in Parkinson's disease through the precision of expected outcomes}},
  journal = {PLOS Computational Biology},
  year = {2022},
  doi = {10.1371/journal.pcbi.1010079},
}
Figure