Cognitive flexibility has been hypothesized to be neurochemically rooted in dopamine neurotransmission. Nonetheless, underpowered sample sizes and contradictory meta-analytic findings have obscured the role of dopamine genes in cognitive flexibility and neglected potential gene–gene interactions. In this largest neurocognitive-genetic study to date ( n = 1400), single nucleotide polymorphisms associated with elevated prefrontal dopamine levels (catechol-O-methyltransferase; rs4680) and diminished striatal dopamine (C957T; rs6277) were both implicated in Wisconsin Card Sorting Test performance. Crucially, however, these genetic effects were only evident in low-IQ participants, suggesting high intelligence compensates for, and eliminates, the effect of dispositional dopamine functioning on flexibility. This interaction between cognitive systems may explain and resolve previous empirical inconsistencies in highly educated participant samples. Moreover, compensatory gene–gene interactions were discovered between catechol-O-methyltransferase and DRD2, such that genotypes conferring either elevated prefrontal dopamine or diminished striatal dopamine—via heightened striatally concentrated D2 dopamine receptor availability—are sufficient for cognitive flexibility, but neither is necessary. The study has therefore revealed a form of epistatic redundancy or substitutability among dopamine systems in shaping adaptable thought and action, thus defining boundary conditions for dopaminergic effects on flexible behavior. These results inform theories of clinical disorders and psychopharmacological interventions and uncover complex fronto-striatal synergies in human flexible cognition.

Cognitive control has traditionally been associated with pFC based on observations of deficits in patients with frontal lesions. However, evidence from patients with Parkinson disease indicates that subcortical regions also contribute to control under certain conditions. We scanned 17 healthy volunteers while they performed a task-switching paradigm that previously dissociated performance deficits arising from frontal lesions in comparison with Parkinson disease, as a function of the abstraction of the rules that are switched. From a multivoxel pattern analysis by Gaussian Process Classification, we then estimated the forward (generative) model to infer regional patterns of activity that predict Switch/Repeat behavior between rule conditions. At 1000 permutations, Switch/Repeat classification accuracy for concrete rules was significant in the BG, but at chance in the frontal lobe. The inverse pattern was obtained for abstract rules, whereby the conditions were successfully discriminated in the frontal lobe but not in the BG. This double dissociation highlights the difference between cortical and subcortical contributions to cognitive control and demonstrates the utility of multivariate approaches in investigations of functions that rely on distributed and overlapping neural substrates.

Cognitive dysfunction in Parkinson's disease (PD) has been hypothesized to reflect a failure of cortical control. In keeping with this hypothesis, some of the cognitive deficits in PD resemble those seen in patients with lesions in the lateral pFC, which has been associated with top–down attentional control. However, there is no direct evidence for a failure of top–down control mechanisms in PD. Here we fill this gap by demonstrating disproportionate control by bottom–up attention to dimensional salience during attentional set shifting. Patients needed significantly more trials to criterion than did controls when shifting to a low-salient dimension while, remarkably, needing significantly fewer trials to criterion than did controls when shifting to a high-salient dimension. Thus, attention was captured by bottom–up attention to salient information to a greater extent in patients than in controls. The results provide a striking reinterpretation of prior set-shifting data and provide the first direct evidence for a failure of top–down attentional control, resembling that seen after catecholamine depletion in the pFC.

Executive functions are likely mediated by interconnected circuits including frontal lobe and basal ganglia structures. We assessed the executive function of task switching in patients with early-stage Huntington's disease (HD), a neurodegenerative disease affecting the basal ganglia. In two experiments, the HD patients had greater difficulty when switching than when repeating a task than matched controls, and this was true even when scaling for the overall slowing of the patients. In the first experiment, HD patients had a switching deficit even in a “pure” condition where they had to switch, predictably, and with substantial preparation time, between stimuli having only one possible response, indicating a switching deficit different from that for patients with Parkinson's disease or frontal lobe trauma, and possibly relating to inadequate activation of stimulus-response links or “response set.” In the more elaborate second experiment, we could not account for the switching deficit of the patients in terms of inadequate preparation in advance of a switch, deficient suppression of taskset processing from the preswitch trial, or impaired suppression of interference due to the presence of a competing task set. Instead, we found that part of the switching deficit was due to elevated reaction time and errors on switch trials for a repeated response (same button press as on preswitch trial) relative to an alternated response (different button press from preswitch trial). We argue that this elevated “repetition effect” for the HD patients is due to excessive inhibition of the justperformed response in advance of a switch. Alterations in the “response-setting” process alone (Experiment 1) and both the response-setting and “response inhibition” process (Experiment 2) probably arise from striatal pathology in HD, thus accounting for the task-switching deficits and showing how basal ganglia implemented response processes may underpin executive function.