Neurobiological Mechanisms of Self-Control in Value-Based Choices

Saturday, February 16, 2013
Room 309 (Hynes Convention Center)
Todd Hare , University of Zürich, Zürich, Switzerland
Self-control is an important component of dietary choices and for some individuals it is a critical factor in maintaining a healthy body weight. I will present data that are designed to test a two-part hypothesis: 1) the overall value of a reward predicting stimulus is represented in ventral medial prefrontal cortex (vmPFC) regardless of self-control level during decision making, and 2) self-control is implemented by means of interactions between the dorsolateral prefrontal cortex (dlPFC) and vmPFC.

We used a combination of functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) to compare neural activity during successful self-control (SSC) and non-self-control (NSC) choices over food and monetary rewards. In both cases, the overall value of the reward was reflected in the activity of vmPFC and the timing was similar for both SSC and NSC decisions. In other words, the vmPFC represented the overall value the rewards in a manner consistent part one of the hypothesis.

To determine if dlPFC played a role in self-control choices, we compared SSC and NSC decisions. There was greater activity in dlPFC for SSC compared to NSC. Furthermore, we found using dynamic causal modeling (DCM) of the fMRI data that this region of dlPFC showed increased signaling to the vmPFC during self-control choices. Moreover, these DCM parameters could be used to explain individual differences in self-control behavior across subjects. We also used EEG to confirm that the timing of interactions between dlPFC and vmPFC at the millisecond level is consistent with dlPFC signaling to vmPFC at the time of choice.

In summary, our data support both aspects of the hypothesis. First, regardless of self-control level, the overall value of a reward predicting stimulus is represented in the activity levels of vmPFC. Second, dlPFC can modulate of value computations in the vmPFC.