An fMRI task for exploring memory effects of significant-other activation in social cognition

Abstract

A novel functional magnetic resonance imaging (fMRI) paradigm for exploring the neurobiologic basis of a key aspect of social cognition was developed, adapted from Andersen’s transference paradigm (Andersen & Chen, 2002). This paradigm captures the extent to which current memory in relation to a stranger is impacted by activated knowledge of a significant other (SO). Forty healthy young adult participants were seen for two separate sessions and misled to believe that the sessions were unrelated to one another. In the first, participants listed descriptive sentences for each of four different important people in their lives (SOs), chose 20 irrelevant adjectives about each from a list, and provided descriptive sentences about 80 famous individuals. In the second session, at least two months later, participants were shown 20 sentences about each of six different supposed strangers (“targets”). Unbeknownst to participants, two of the targets resembled SOs described by the participant in the initial session (“the experimental targets”), two resembled SOs described by a different participant (the yoked-control), and two were constructed from a random assortment of sentences the participant listed about famous individuals (the no-representation control). The participants were then shown 30 sentences about each target and asked to rate how likely it was that they were just exposed to each about that particular target. For targets resembling SOs, 10 of the 30 sentences (the “lures”) had, in fact, not been seen previously about the target, but were from the participant’s own description of the pertinent SO. Therefore, participants should incorrectly rate these lures higher when the target resembled their own SO, as compared to the comparable lures in either control condition. Functional MRI images were collected during the entire second session using a GE Signa 3T whole body scanner. Individual and group level fMRI analyses were conducted using SPM8. Replicating Andersen’s findings, the average memory recognition score was higher for the “lures” in the SO condition (M =2.69; SD = .61) than for those in the yoked-control condition (M = 2.55; SD = .65), paired t (39) = 2.01, p = .05). In fMRI analyses, betas reflected the tendency of neural activity to increase when SO-relevant targets were presented during the learning phase and these were used to produce a second-level contrast map which showed significant differences in activation during the presentation of SO-resembling targets, in comparison with the yoked-control condition. Findings were significant at the level of p < .05 after correction for multiple comparisons using the false discovery rate method. Neural activity in bilateral fusiform cortex and the superior frontal gyrus increased during SO activation, while neural activity in the subgenual ACC (sgACC) decreased. This paradigm is a powerful tool for exploring the neural basis of social cognition and may well prove useful in characterizing psychiatric disorders in a developmentally sophisticated way. METHOD ● Participants 40 Healthy Young Adults Ages 18-30, mean=21.9 15 male, 25 female ● Initial Session ● Experimental Session ● Learning Task ● Memory Recognition Task RESULTS Fig 1. Statistically significant recognition memory effect, n=40, p=.05 Fig 2. Healthy control schema-relevant activity (red= increasing, blue=decreasing), N=40, corrected p<.05.  Selects 4 Significant Others and provides 10 positive and 10 negative short sentences about each  Identifies 25 “irrelevant” adjectives about each Significant Other  Provides 40 positive and 40 negative short sentences about 80 well-known individuals