Failure modes of the will: from goals to habits to compulsions?

Abstract

Good decisions reflect the past and improve the future. Trouble ensues when the rewards of the past conflict with the goals of the future. New goals often fail to override behaviors reinforced in the past, for instance, when patients with obsessive-compulsive disorder (OCD) try to stop the counting behavior that has been their main strategy to avoid aversive intrusions. Decision-making theories that emphasize the existence of two decision systems working in parallel provide one fruitful multidisciplinary route for understanding the origin of this conflict (1, 2): while habits reflect rewards of the past lengthily accumulated overmultiple repetitions, goal-directed decisions capture aims based on the current, more malleable understanding of the world (2). In certain disorders, such as those of impulsivity and compulsivity, an individual’s explicit aims appear to repeatedly fall prey to his or her engrained behaviors, suggesting an alteration in the adaptive tradeoff between prospective goals and retrospective habits (3, 4). In this issue of the Journal, Gillan et al. (5) add to an impressive catalog of studies on the im-balance between habitual and goal-directed decisions in OCD. Roughly speaking, these studies assess how easily we overcome habits—established through repeated pursuit of an old goal—when a novel goal arises. Gillan et al. have previously shown that patients with OCD have a tendency to establish habits faster than comparison subjects when pursuing rewards (6, 7) and, critically, also when avoiding punishments. Gillan et al. (8) first explained to participants that two visual stimuli, say L and R, would be followed by shocks to their left or right wrist, respectively. A third stimulus, S, signaled safety. Participants could avoid the shock to the wrist threatened by the stimulus by pressing a foot-key with the respective foot (e.g., the left key for stimulus L). A training period in which participants repeatedly avoided the shocks predicted for both wrists was then followed by an instructed devaluation, where the shock electrode was demonstratively disconnected from the left wrist. Participants were explicitly instructed that theywould no longer be shocked on the left wrist and that their only goal was to avoid the remaining shocks. Habitual avoidance was measured by the tendency to continue pressing the left footkeywhenseeing thenowdevalued stimulusL,despiteknowing that this effort was unnecessary because the electrode had been disconnected. OCD patients were more than four times as likely to continue avoiding, indicating that they had established stronger habits than comparison subjects over the trainingperiod.Therewasnodifference in theunderstanding of the devaluation, suggesting that avoidance was not driven by a goal-directed response to an overestimated shock probability. In their present study in this issue, Gillan et al. replicate these findings and extend them with data from functional imaging. The behavioral replication is noteworthy, since it establishes faster habitization as one feature of OCD, suggesting that compulsions in OCD originate from the rapid transformationofgoal-directed avoidance responses to obsessions into habits. Furthermore, habitization and autonomicmeasures of anxiety went hand-in-hand: patients who formed habits showed similar skin conductance responses for valued and devalued stimuli, suggesting that aspects of anxiety could be related to habit formation. And even though habits are components of everyday life, typically lacking a compelling feel, they seemed tohave a more urgent nature for OCD patients: the extent of habitization correlated with the participants’ self-reported urge to respond and also (more tentatively) with OCD symptom severity. OCD is thus one of the disorders accompanied by a shift from prospective goals toward retrospective habits, akin to addiction (7, 9). Possible origins of this shift are illuminated by the imaging results. Neurally, the goal-directed and habitual systems depend on different substrates (10), and the authors focused their analysis on these. First, they found that the establishment of habits in the subgroup of patients continuing to respond to the devalued stimulus was accompanied by hyperactivity in the caudate. Patients who did not continue to respond after devaluation showed hypoactivity. Since there was also no overall difference between OCD patients and healthy comparison subjects, it is unclear whether caudate hyperactivity merely reflects the habitization differences or OCD is ... one of the disorders accompanied by a shift from prospective goals toward retrospective habits, akin to addiction.