Background:
The placebo literature suggests that both conceptual (i.e. socially instructed beliefs) and associative learning processes are critical for the genesis of placebo effects. Several studies have performed placebo 'conditioning' (associating a sham treatment with reduced pain through repeated experience) but interfered with conceptual processing by informing subjects that the intensity of the stimulus was being lowered. This manipulation prevented the attribution of pain reductions to the placebo treatment during the learning process. These studies showed no conditioned analgesia. However, when the same 'conditioning' was performed without the verbal explanation for why the treatment seemed to work, robust placebo effects were created. Conceptual processes appear to be critical. Conversely, several studies have manipulated conceptual expectations alone, by manipulating verbal instructions, and have found markedly reduced or absent placebo analgesia on both pain report and brain event-related potentials. Learning by experience also seems to be critical.
These studies separately have led to the conclusions that 'expectancy' and 'conditioning' are each critical processes, and debates have focused on which one is the driver of placebo effects. The investigators propose another view: Both processes are critical, and they interact. Experience drives changes in value learning systems, but in any type of value learning, there is a 'credit assignment' problem, and the brain must decide which cue-outcome associations to update as a result of experience: Is the pain reduced because the treatment was effective or because the cause of pain changed? Conceptual processes fill the gap, drawing on explicit memory and generalization from similar past experiences to solve the credit assignment problem, creating analgesia if experienced relief is attributed to the treatment. This view is compatible with older information based theories of conditioning and new evidence that rats and humans alike maintain expectancies about specific outcomes and mental models of contingencies that are distinct from associative learning. In spite of dozens of published studies demonstrating effective placebo analgesia with the established paradigm the investigators use, the precise nature of the learning that occurs is unknown, because placebo analgesia has not typically been studied from a learning-systems perspective. The aim of the present study is to assess the influence of social information and associative learning on placebo analgesia.
In addition, interactions between neurochemical systems and placebo analgesia have hardly been explored, and this proposal represents a significant effort in that regard. For example, oxytocin interacts synergistically with opioids in the PAG (periaqueductal gray) (and CCK (Cholecystokinin)) to relieve pain, and in a separate literature reliably increases trust and reduces anxiety in interpersonal situations. In spite of the fact that oxytocin has been proposed as central to the placebo effect and can be administered to humans with no known subjective effects or side effects, its role in placebo analgesia has not been explored extensively. The experiment proposed here will clarify the roles of the oxytocin system and its contributions to social facilitation of analgesia, and will be instrumental in developing a systems-based model of placebo effects.
Experimental Design:
Participants will perform two tasks in each experimental session. First, they will perform a social-influence and learning task to investigate the effects of oxytocin on social instruction effects and learning on pain. Second, they will perform a social-support during pain, to test the effects of oxytocin on the pain-alleviating effects of social support during pain.
