Distinct effects of apathy and dopamine on effort-based decision-making in Parkinson's disease

Abstract

Effort-based decision-making is a cognitive process crucial to normal motivated behaviour. Apathy is a common and disabling complication of Parkinson's disease, but its aetiology remains unclear. Intriguingly, the neural substrates associated with apathy also subserve effort-based decision-making in animal models and humans. Furthermore, the dopaminergic system plays a core role in motivating effortful behaviour for reward, and its dysfunction has been proposed to play a crucial role in the aetiology of apathy in Parkinson's disease. We hypothesized that disrupted effort-based decision-making underlies the syndrome of apathy in Parkinson's disease, and that this disruption may be modulated by the dopaminergic system. An effort-based decision-making task was administered to 39 patients with Parkinson's disease, with and without clinical apathy, ON and OFF their normal dopaminergic medications across two separate sessions, as well as 32 healthy age- and gender-matched controls. On a trial-by-trial basis, participants decided whether to accept or reject offers of monetary reward in return for exerting different levels of physical effort via handheld, individually calibrated dynamometers. Effort and reward were manipulated independently, such that offers spanned the full range of effort/reward combinations. Apathy was assessed using the Lille apathy rating scale. Motor effects of the dopamine manipulation were assessed using the Unified Parkinson's Disease Rating Scale part three motor score. The primary outcome variable was choice (accept/decline offer) analysed using a hierarchical generalized linear mixed effects model, and the vigour of squeeze (Newtons exerted above required force). Both apathy and dopamine depletion were associated with reduced acceptance of offers. However, these effects were driven by dissociable patterns of responding. While apathy was characterized by increased rejection of predominantly low reward offers, dopamine increased responding to high effort, high reward offers, irrespective of underlying motivational state. Dopamine also exerted a main effect on motor vigour, increasing force production independently of reward offered, while apathy did not affect this measure. The findings demonstrate that disrupted effort-based decision-making underlies Parkinson's disease apathy, but in a manner distinct to that caused by dopamine depletion. Apathy is associated with reduced incentivization by the rewarding outcomes of actions. In contrast, dopamine has a general effect in motivating behaviour for high effort, high reward options without altering the response pattern that characterizes the apathetic state. Thus, the motivational deficit observed in Parkinson's disease appears not to be simply secondary to dopaminergic depletion of mesocorticolimbic pathways, suggesting non-dopaminergic therapeutic strategies for apathy may be important future targets.

Figures

Figure 1

Effort-based decision-making task. On a trial-by-trial basis, participants were presented with offers of a certain amount of reward (apples on an apple tree, with each apple worth 1p) in return for a level of physical effort [ranging between 10% and 80% of a subjects’ previously determined maximal voluntary contraction (MVC), held for 1 s] (A and B). They were instructed to weigh up each offer, deciding ‘whether it was worth it’. If they accepted an offer (by squeezing the left-hand grip) the tree moved to the left or right of the screen, indicating which hand they had to perform the force squeeze with. They had a 5-s window within which to achieve the required force level. If they rejected the offer (by squeezing the right-hand grip) they waited the same 5-s period. After a feedback phase they then moved onto the next trial. After a practice session, participants worked through 180 trials, which evenly sampled the 6 × 6 ‘decision space’ over five blocks (C). Example force trace from a single trial (D, left). Participants parametrically modulated force output to task requirements. Superimposed force traces for all patients, ON and OFF their dopaminergic medications. Each required force level is a different colour, and the solid lines show the minimum required level (D, middle). Both patients and healthy controls modulated squeeze force appropriately (D, right). PD = Parkinson’s disease.

Figure 2

Raw acceptance results as apathy status, and dopamine state, varied. Mean acceptance rates for all patients, ON and OFF their normal medications. As reward on offer increased, and effort required decreased, the probability of a participant accepting an offer increased, raw data (A) and modelled data (B). Apathetic Parkinson’s disease patients accepted significantly fewer offers than healthy controls or non-apathetic Parkinson’s disease patients (C, ON state shown, error bars are ± SEM). Apathy level (action initiation subscale of LARS) strongly correlated with acceptance rate (D, r = 0.37, P = 0.019; ON state shown). Dopamine depletion (OFF state) was associated with significantly reduced acceptance rates, irrespective of apathy status (E, dotted line is mean effect ± SEM). PD = Parkinson’s disease.

Figure 3

Distinct effects of apathy and dopamine depletion on response pattern. Change in acceptance in apathetic compared to non-apathetic patients, as both reward and effort increase. The panels show difference in responding (% accepted in no apathy group − % accepted in apathy group) at each reward level, collapsed across effort (A, left) and effort level, collapsed across reward (A, right). The reduced acceptance rate seen in apathetic patients was driven by reduced responding to low reward offers, particularly when effort costs were lower (Apathy × Reward × Effort interaction P = 0.004). A different behavioural pattern characterized the effects of dopamine depletion (OFF state) on choice. Panels show the mean change in acceptance (ON minus OFF) across patients, as reward (B, left) and effort (B, right) increases. The difference between ON and OFF states was driven by increased responding to predominantly high effort, higher reward offers (Dopamine × Reward × Effort interaction P = 0.003). These differences in acceptance rates associated with apathy (C, left) and dopamine depletion (C, right) manifest within distinct regions of the 6 × 6 decision space grid. The grey shaded plane (z-axis = 0) represents no difference in acceptance. All error bars are ± SEM. PD = Parkinson’s disease.

Figure 4

Dopamine depletion, but not apathy, reduces motor vigour following decision to engage. Motor vigour was indexed by excess force generated above the required effort level, i.e. how much more than was required that a participant squeezed (A). Dopamine (ON state) significantly increased response vigour (indexed by OFF state), while the presence of apathy was not associated with a significant change in the vigour of response (B). Despite greater force production at each effort level in the ON state, there was no change in perceived physical exertion between ON and OFF state, nor between patients and controls (C). Maximal voluntary contraction did not differ between ON and OFF states (D). PD = Parkinson’s disease.