Increased Reward-Related Activation in the Ventral Striatum During Stress Exposure Associated With Positive Affect in the Daily Life of Young Adults With a Family History of Depression. Preliminary Findings

Abstract

Background: Being the offspring of a parent with major depression disorder (MDD) is a strong predictor for developing MDD. Blunted striatal responses to reward were identified in individuals with MDD and in asymptomatic individuals with family history of depression (FHD). Stress is a major etiological factor for MDD and was also reported to reduce the striatal responses to reward. The stress-reward interactions in FHD individuals has not been explored yet. Extending neuroimaging results into daily-life experience, self-reported ambulatory measures of positive affect (PA) were shown to be associated with striatal activation during reward processing. A reduction of self-reported PA in daily life is consistently reported in individuals with current MDD. Here, we aimed to test (1) whether increased family risk of depression is associated with blunted neural and self-reported reward responses. (2) the stress-reward interactions at the neural level. We expected a stronger reduction of reward-related striatal activation under stress in FHD individuals compared to HC. (3) the associations between fMRI and daily life self-reported data on reward and stress experiences, with a specific interest in the striatum as a crucial region for reward processing. Method: Participants were 16 asymptomatic young adults with FHD and 16 controls (HC). They performed the Fribourg Reward Task with and without stress induction, using event-related fMRI. We conducted whole-brain analyses comparing the two groups for the main effect of reward (rewarded > not-rewarded) during reward feedback in control (no-stress) and stress conditions. Beta weights extracted from significant activation in this contrast were correlated with self-reported PA and negative affect (NA) assessed over 1 week. Results: Under stress induction, the reward-related activation in the ventral striatum (VS) was higher in the FHD group than in the HC group. Unexpectedly, we did not find significant group differences in the self-reported daily life PA measures. During stress induction, VS reward-related activation correlated positively with PA in both groups and negatively with NA in the HC group. Conclusion: As expected, our results indicate that increased family risk of depression was associated with specific striatum reactivity to reward in a stress condition, and support previous findings that ventral striatal reward-related response is associated with PA. A new unexpected finding is the negative association between NA and reward-related ventral striatal activation in the HC group.

Keywords: ambulatory assessment (AA); depression; fMRI; negative affect (NA); positive affect (PA); reward; stress; striatum.

Figure 1

Fribourg reward task. Illustration of trial conditions randomly distributed in both control and stress conditions (unpredictable threat of shock). Variables used in the present study are in green [adapted from Gaillard et al., (76), p. 4].

Figure 2

Left ventral striatal (VS, i.e., putamen) region BOLD activation for comparison of FHD and HC groups during reward feedback in stress condition for contrast rewarded > not rewarded (p < 0.005 FWE). Parameter estimates (beta weight) were extracted from coordinates that showed significant activation after FWE at p < 0.05 in the ROI analyses for the main effect of reward.

Figure 3

Graphical presentation of the statistical relationships between (A) mean positive affect resp. (B) Mean negative affect and left ventral striatal (VS) region BOLD activation during reward feedback in stress condition for contrast rewarded > not rewarded. Parameter estimates (beta weight) were extracted from coordinates that showed significant activation after FWC at p < 0.05 in the ROI analyses with peak activation in the caudate. Results are presented for the entire group, the FHD group and the HC group. r, Spearmann correlation coefficient, n.s., not significant.