Perceptual decisions based on previously learned information are independent of dopaminergic tone

Abstract

Both cognitive and motor symptoms in people with Parkinson's disease (PD) arise from either too little or too much dopamine (DA). Akinesia stems from DA neuronal cell loss, and dyskinesia often stems from an overdose of DA medication. Cognitive behaviors typically associated with frontal cortical function, such as working memory and task switching, are also affected by too little or too much DA in PD. Whether motor and cognitive circuits overlap in PD is unknown. In this article, we show that whereas motor performance improves in people with PD when on dopaminergic medication compared with off medication, perceptual decision-making based on previously learned information (priors) remains impaired whether on or off medications. To rule out effects of long-term DA treatment and dopaminergic neuronal loss such as occur in PD, we also tested a group of people with dopa-unresponsive focal dystonia, a disease that involves the basal ganglia, like PD, but has motor symptoms that are insensitive to dopamine treatment and is not thought to involve frontal cortical DA circuits, unlike PD. We found that people with focal dystonia showed intact perceptual decision-making performance but impaired use of priors in perceptual decision-making, similar to people with PD. Together, the results show a dissociation between motor and cognitive performance in people with PD and reveal a novel cognitive impairment, independent of sensory and motor impairment, in people with focal dystonia. The combined results from people with PD and people with focal dystonia provide mechanistic insights into the role of basal ganglia non-dopaminergic circuits in perceptual decision-making based on priors.

Keywords: Parkinson’s disease; basal ganglia; bias; cognition; dopa-unresponsive dystonia, dopamine; dystonia; focal dystonia; glass patterns; implicit learning; movement disorders; perception; priors.

Fig. 1.

Perceptual decision-making task. A: example images of Glass patterns. The strength of the sensory information contained in the Glass pattern varies with the number of correlated dots (coherence). Examples shown are 0% (no information about orientation), 13%, 35%, and 100% (all pairs sharing same orientation). B: example red and green rightward Glass patterns are shown, each with 100% coherence. This illustrates one example session in which the red Glass pattern is associated with equal orientation priors (50:50) and the green Glass pattern is associated with unequal orientation priors (75:25). In this example, the rightward orientation occurred more often. The direction of the orientation priors and the color with which they were associated were counterbalanced across participants. C: spatial and temporal arrangement of the task. The boxes indicate the visual display. The black spots show the fixation spot at the center, and the 2 choice targets appear peripherally, indicating the 2 possible orientations. The Glass pattern appears at the central location. Participants indicated what direction they perceived by pressing the “O” (leftward) or the “P” (rightward) key on a computer keyboard using one hand, illustrated by the finger. Participants heard a low pitch tone (a beep) for every correct trial and received no feedback for incorrect trials, illustrated by the audio symbol. No explicit reward was provided. D: proportion of positive (leftward) choices is plotted against orientation strength for 12 healthy people (HC; 5 men, 7 women, mean age 63 yr). These data are fitted with a logistic function and provide a measure of the response bias (α) and the slope or sensitivity of the psychometric function (β). The parameters of the fits were used to compare performance between the equal and unequal prior conditions within groups. The gray diamond and line show the data and the logistic fit in the equal prior trials (50:50). Age- and sex- matched HC make accurate decisions in conditions of sensory certainty (100%). Their performance decreases as sensory information becomes less clear (35% and 13%) and reaches chance level when sensory information is ambiguous (0%) and participants guess. E: comparisons between the equal and unequal prior conditions in participants presented with more frequent positive orientation during the unequal prior trials. The gray lines and upward triangles show the data in the equal prior trials (50:50), whereas the black lines and upward triangles show the data for unequal positive prior trials (25:75). F: same as in E in participants presented with more frequent negative orientation during the unequal prior trials. The gray lines and downward triangles show the data in the equal prior trials (50:50), whereas the black lines and downward triangles show the data for unequal negative prior trials (75:25). G: logistic fits, α and β parameters, for the fits in the negative, equal (for negative), equal (for positive), and positive prior trials (*P = 0.04, paired t-test). Error bars are SE. Data in this group are from Perugini et al. (2016).

Fig. 3.

Dopamine (DA) medications improve motor symptoms. A: plot shows the UPDRS III scores for PDon (15 ± 1.66) and PDoff (18 ± 1.7). Significantly higher scores during the off session indicate that abstinence from DA medications impaired movement in people with PD [UPDRS III on vs. off, t(14) = 2.27, P = 0.039, paired t-test]. B: plot shows the “difference” UPDRS III scores between off and on session for each patient. Each square represents a value obtained by subtracting the UPDRS score when patients were on medications from the UPDRS score when patients were off medications. A positive value means that medications improve patients’ motor symptoms (UPDRS scores off > on, n = 11). A negative value means that medications worsen patients’ motor symptoms (UPDRS OFF < on, n = 1). A value equal to 0 means that medications did not affect patients’ motor symptoms during assessment (UPDRS off = on, n = 3). C: “difference” use of prior (proportion of more frequent choices for PDon minus proportion of more frequent choices for PDoff) against UPDRS III scores for PDoff minus UPDRS III scores for PDon. We plotted data only from people with PD who had improved motor behavior with DA (n = 11). Positive values on the x-axis mean that DA improved motor behavior, and positive values on the y-axis mean that DA improved use of the prior at the decision-making task. There is no correlation between the effects of DA on the use of prior and on motor behavior (r = 0.12, P = 0.7). D: plot shows the distribution of the reaction times (RT) in milliseconds (msec) across orientation strength for correct trials in PDon. Black circles show the unequal prior trials (75:25/25:75), and gray circles show the equal prior trials (50:50). It is important to note that the RT shown include sensory, motor, and decision times. E: same as in D for PDoff. Error bars are SE.

Fig. 4.

People with dopa-unresponsive focal dystonia are impaired at using priors to guide perceptual decision-making. A: proportion of positive choices are plotted as a function of orientation strength for a group of 16 people with dopa-unresponsive focal dystonia (DY; 2 men, 14 women, mean age 61 yr). Data are plotted as in Fig. 1D. People with dystonia are able to perform the task well based on the amount of sensory information provided, and they guess when sensory information is ambiguous. This rules out interpretations based on motor or perceptual impairments. B: comparisons between the equal and unequal prior conditions for participants presented with more frequent positive orientation in the unequal prior trials. Data are plotted as in Fig. 1E. People with dystonia are not able to bias their choices based on prior information. C: same as in B for participants presented with more frequent negative orientation in the unequal prior trials. Data are plotted as in Fig. 1F. Error bars are SE. D: logistic fits, α and β parameters, for dystonia. Data are plotted as in Fig. 1G. E: use of prior for dystonia against the scores on the Fahn-Marsden scale. We excluded one participant with musician’s dystonia from this plot, because the participant had a score of 0 on the Fahn-Marsden (n = 15). There is no correlation between the use of prior and motor impairment in people with dystonia (r = 0.03, P = 0.9).

Fig. 5.

People with PD and dystonia are not impaired at learning from positive or negative feedback. Graphs show the mean proportion of win-stay (A) and lose-shift (B) in each group, separated in trials with the equal prior stimuli and trials with the unequal prior stimuli with 0% coherence stimuli. Win-stay refers to the proportion of repeating the same response (leftward or rightward) after a correct trial (feedback). Lose-shift refers to the proportion of changing response after an error trial (no feedback). The win-stay and loose-shift trials only represent half of the total trials present in the experiment, for each stimulus color (equal and unequal). Indeed, the total number of trials for each stimulus color is given by summing the win-stay trials, lose-shift trials, win-shift trials (where participants changed response after correct trials), and lose-stay trials (where participants repeated the same response after an incorrect trial). We found a significant difference in the proportions of win-stay between equal and unequal prior trials in all groups, using a 2 × 2 mixed ANOVA [F(1,54) = 126, ***P < 0.0001), no difference between groups [F(3,54) = 1.46, P = 0.24], and no significant interaction between groups and prior condition [F(3,54) = 1.76, P = 0.166]. Similarly, we found a significant difference in the proportions of lose-shift between equal and unequal prior trials in all groups [F(1,54) = 43.2, ***P < 0.0001], no difference between groups [F(3,54) = 1.45, P = 0.24], and no significant interaction between groups and prior condition [F(3,54) = 1.88, P = 0.144]. The data show that participants repeated responses that led to a positive outcome more often for the unequal prior stimuli compared with the equal prior stimuli. Conversely, they adopted a lose-shift strategy more for the equal prior trials compared with the unequal prior trials. Error bars are SE.

Fig. 2.

Dopamine medications do not affect the ability to use priors in perceptual decision-making performance in PD. A: proportion of positive choices is plotted as a function of orientation strength for a group of 15 people with PD (6 men, 9 women, mean age 64 yr) after taking dopaminergic medications. Data are plotted as in Fig. 1D. PDon are able to perform the task well based on the amount of sensory information provided, and they guess when sensory information is ambiguous. This rules out interpretations based on motor or perceptual impairments. B: comparisons between the equal and unequal prior conditions for participants presented with more frequent positive orientation during the unequal prior trials. Data are plotted as in Fig. 1E. PDon are not able to bias their choices based on prior information. C: same as in B for participants presented with more frequent negative orientation during the unequal prior trials. Data are plotted as in Fig. 1F. D: logistic fits, α and β parameters, for PDon. Data are plotted as in Fig. 1G. E: same as in A for the same 15 people with PD, but after at least 16 h of withdrawal and abstinence from dopaminergic medications. F: same as in B for PDoff. G: same as in C for PDoff. H: same as in D for PDoff. Error bars are SE.