Impulsivity and apathy in Parkinson's disease

Abstract

Impulse control disorders (ICDs) and apathy are recognized as two important neuropsychiatric syndromes associated with Parkinson's disease (PD), but as yet we understand very little about the cognitive mechanisms underlying them. Here, we review emerging findings, from both human and animal studies, that suggest that impulsivity and apathy are opposite extremes of a dopamine-dependent spectrum of motivated decision making. We first argue that there is strong support for a hypodopaminergic state in PD patients with apathy, as well as for an association between dopamine therapy and development of ICDs. However, there is little evidence for a clear dose-response relationship, and great heterogeneity of findings. We argue that dopaminergic state on its own is an insufficient explanation, and suggest instead that there is now substantial evidence that both apathy and impulsivity are in fact multi-dimensional syndromes, with separate, dissociable mechanisms underlying their 'surface' manifestations. Some of these mechanisms might be dopamine-dependent. According to this view, individuals diagnosed as impulsive or apathetic may have very different mechanisms underlying their clinical states. We propose that impulsivity and apathy can arise from dissociable deficits in option generation, option selection, action initiation or inhibition and learning. Review of the behavioural and neurobiological evidence leads us to a new conceptual framework that might help understand the variety of functional deficits seen in PD.

Figure 1

Highly schematic depiction of processes involved in motivated behaviour implicated in impulsivity and apathy. Animals and humans use, among other things, perceptual and attentional mechanisms to generate possible behavioural options. These options are then valued and compared based on features including predicted reward, punishment, effort required, time involved to outcome delivery, and probability of outcome. Following such option selection, the chosen option is then linked to the appropriate action – but critically, actions can also be cancelled by signals arriving from a change in context, for example, changes in the environment, which might mean that a behaviour is no longer optimal. Finally, the outcomes of motivated behaviours are compared with predictions made prior to making these actions. Such comparisons play a key part of learning and feeding back on option selection mechanisms. There is evidence that each of these processes might be modulated by dopamine. (Developed from Kalis et al., 2008).

Figure 2

Hypothetical relationships between distinct facets of impulsivity and apathy. This schematic illustrates how dysfunction resulting in impulsive and apathetic behaviours can occur at multiple levels of processing, often leading to superficially similar behavioural phenotypes. Optimal function along each axis will be in part determined by the behavioural context at a given point in time, for example, whether opportunities for rewards are likely to remain stable for the immediate future, or whether their availability is highly time-sensitive. This schematic is not meant to be exhaustive, but serves to illustrate the principle.

Figure 3

Classical model of basal ganglia pathophysiology in healthy, Parkinsonian, and dyskinetic states. The input to BG is from most cortical areas. Output is via GPi/SNr, which provides inhibitory input to the thalamus. The direct pathway expresses D1 receptors while the indirect pathway expresses D2 receptors. In health, GPi output is modulated by competing influences from both direct and indirect pathways. In Parkinson’s Disease, death of SNc neurons results in overactivity in the indirect pathway and underactivity in the direct pathway leading to heightened GPi output. The converse pattern is thought to account for dyskinesia induced by l-dopa therapy. Adapted from Rodriguez et al., 2009

Figure 4

Highly schematic and speculative depiction of changes that might account for age differences, behavioural findings, and occurrence of apathy and impulsivity in PD. (a) Ventral striatal responsiveness (indexed by reward sensitivity, reward learning, effortful responding, and novelty seeking) decreases with age (dashed line). (b) In late-onset PD, degeneration of dopaminergic projections to ventral striatum occurs when ventral striatal responsiveness is already at a low level causing apathy (solid black sloping line). Potentially, vulnerability mechanisms for apathy, such as deficits in option generation due to executive dysfunction, lower the threshold for the expression of apathetic behaviour. Treatment with dopaminergic medication restores ventral striatal function (solid red vertical line). (c) However, if PD occurs at a young age when ventral striatal deterioration is limited, then administration of dopaminergic medication can cause pathologically enhanced ventral striatal responsiveness (solid red vertical line). Additionally, vulnerability mechanisms, such as decreased response inhibition, reduce the threshold for expression of impulsive behaviours.