S. de Wit

Corticostriatal Connectivity Underlies Individual Differences in the Balance between Habitual and Goal-Directed Action Control

Why are some individuals more susceptible to the formation of inflexible habits than others? In the present study, we used diffusion tensor imaging to demonstrate that brain connectivity predicts individual differences in relative goal-directed and habitual behavioral control in humans. Specifically, vulnerability to habitual “slips of action” toward no-longer-rewarding outcomes was predicted by estimated white matter tract strength in the premotor cortex seeded from the posterior putamen (as well as by gray matter density in the posterior putamen as determined with voxel-based morphometry). In contrast, flexible goal-directed action was predicted by estimated tract strength in the ventromedial prefrontal cortex seeded from the caudate. These findings suggest that integrity of dissociable corticostriatal pathways underlies individual differences in action control in the healthy population, which may ultimately mediate vulnerability to impulse control disorders.

Working for food you don't desire. Cues interfere with goal-directed food-seeking

Why do we indulge in food-seeking and eating behaviors at times when we are already fully sated? In the present study we investigated the hypothesis that food-associated cues in the environment can interfere with goal-directed action by eliciting food-seeking that is independent of the current desirability of the outcome. To this end, we used a computerized task in which participants learned to press keys for chocolate and popcorn rewards. Subsequently, we investigated whether satiation on one of these rewards would bias choice toward the other, still desirable, food reward. We found that satiation did indeed selectively reduce responding on the associated key in the absence of food-associated cues. In contrast, in a Pavlovian-instrumental transfer (PIT) test, satiation failed to reduce cue-elicited food-seeking: in line with our hypothesis, cues that had previously been paired with chocolate and popcorn led to increased responding for the signaled food reward, independent of satiation. Furthermore, we show that food-associated cues will not only bias choice toward the signaled food (outcome-specific transfer), but also enhance the vigor of responding generally (general transfer). These findings point to a mechanism that may underlie the powerful control that cues in our obesogenic environment exert over our behavior.

Motivational mechanisms and outcome expectancies underlying the approach bias toward addictive substances

Human behavior can be paradoxical, in that actions can be initiated that are seemingly incongruent with an individual’s explicit desires. This is most commonly observed in drug addiction, where maladaptive behavior (i.e., drug seeking) appears to be compulsive, continuing at great personal cost. Approach biases toward addictive substances have been correlated with actual drug-use in a number of studies, suggesting that this measure can, in some cases, index everyday maladaptive tendencies. At present it is unclear whether this bias to drug cues is a Pavlovian conditioned approach response, a habitual response, the result of a Pavlovian-instrumental transfer process, or a goal-directed action in the sense that expectancy of the rewarding effects of drugs controls approach. We consider this question by combining the theoretical framework of associative learning with the available evidence from approach bias research. Although research investigating the relative contributions of these mechanisms to the approach bias is to date relatively limited, we review existing studies and also outline avenues for future research.

Motivational mechanisms underlying the approach bias to cigarettes

Approach Avoidance tasks measure approach bias, a behavioral tendency to be faster at approaching rather than avoiding drug cues. Approach bias has been measured in a number of different drug-using populations and there is evidence to suggest that approach bias measurements correlate with drug use. Little is known, however, about the motivational mechanisms underlying the approach bias. In the current study we assessed whether the approach bias to cigarettes was immediately sensitive to changes in the incentive value of smoking. We examined the change from baseline in a participant group, after half the group had been given the opportunity to smoke. Specifically, we examined whether the approach bias has the characteristics of a cue-elicited behavior or is flexibly modulated by current desire. Results showed that while the baseline approach-bias score in deprived cigarette smokers correlated with craving, smoking a cigarette led to reduced craving but an increased approach bias score. We discuss a possible account of these findings in terms of an ideomotor outcome-response priming mechanism.

Current limits of experimental research into habits and future directions

The outcome devaluation paradigm allows for the differentiation between goal-directed actions and habits and has been used extensively in both animal and human studies. This has been a fruitful avenue of research with translational human research reporting reduced sensitivity to devaluation in a number of clinical populations using various different paradigms. However, these paradigms do not provide a direct window on the extent to which impaired performance is due to strong habit formation or to weak goal-directed control. We review the indirect nature of the existing evidence from both animal and human studies and suggest avenues for future research including combining behavioral repetition with manipulations to reduce goal-directed control processes as well as alternative paradigms such as the Pavlovian-to-instrumental transfer task.

Age-related changes in deterministic learning from positive versus negative performance feedback

Feedback-based learning declines with age. Because older adults are generally biased toward positive information (“positivity effect”), learning from positive feedback may be less impaired than learning from negative outcomes. The literature documents mixed results, due possibly to variability between studies in task design. In the current series of studies, we investigated the influence of feedback valence on reinforcement learning in young and older adults. We used nonprobabilistic learning tasks, to more systematically study the effects of feedback magnitude, learning of stimulus–response (S–R) versus stimulus–outcome (S–O) associations, and working-memory capacity. In most experiments, older adults benefitted more from positive than negative feedback, but only with large feedback magnitudes. Positivity effects were pronounced for S–O learning, whereas S–R learning correlated with working-memory capacity in both age groups. These results underline the context dependence of positivity effects in learning and suggest that older adults focus on high gains when these are informative for behavior.