Sex-specific neural responses to acute psychosocial stress in depression

Abstract

Major Depressive Disorder (MDD) is characterized by increased stress sensitivity. Emerging findings in healthy adults suggest that stress responses within limbic/striatal-prefrontal regions are moderated by sex and unfold over time. Thus, we hypothesized that stress response abnormalities in MDD might be affected by sex and stress exposure time. The Montreal Imaging Stress Task was administered to 124 unmedicated patients with first-episode MDD (76 females) and 243 healthy controls (HC; 137 females) during functional magnetic resonance imaging (fMRI). Based on prior studies, amygdala, hippocampus, medial orbitofrontal cortex (mOFC), nucleus accumbens (NAc) and dorsolateral prefrontal cortex (dlPFC) were selected as a priori regions of interest. In a complementary approach, we probed the effects of stress on the frontoparietal network (FPN) and a network including the amygdala, NAc and anterior cingulate cortex (ACC). Across groups, males exhibited higher dlPFC activity and right FPN amplitude than females. Relative to female HCs, the female MDD group had less deactivation in limbic/striatal regions (amygdala, NAc, hippocampus, Amygdala-NAc-ACC network). Furthermore, unlike female HCs, the female MDD group failed to show a significant increase of deactivation over stress exposure time in the amygdala, mOFC and NAc. Our findings confirm the importance of considering sex differences when investigating neural stress responses. Case-control differences in neural stress responses observed in females (but not males) provide insights into sex differences in the etiology and pathophysiology of depression. The failure to deactivate limbic/NAc regions in depressed females point to dysfunction of adaptive stress responses over stress exposure time.

Fig. 1. Neural responses to acute psychosocial stress.

A Location of medial orbitofrontal cortex, hippocampus, amygdala, nucleus accumbens, dorsolateral prefrontal cortex. B Sex differences in dorsolateral prefrontal cortex. C Sex × Diagnosis interaction effect in amygdala. D Sex × Diagnosis interaction effect in hippocampus. E Sex × Diagnosis interaction effect in nucleus accumbens. Estimated-mean are plotted, and error bars represents standard error (SE). mOFC medial orbitofrontal cortex, dlPFC dorsolateral prefrontal cortex, HC healthy controls, MDD major depressive disorder. *p _Bonferroni < 0.05, **p _Bonferroni < 0.01.

Fig. 2. Subjective and cortisol stress responses.

A All four groups exhibited higher post-MIST subjective stress evaluation level in comparison to the pre-MIST subjective stress level. Estimated-mean are plotted, and the error bar represent SE. B Significant main effect of time in cortisol concentration over the stress exposure. ***p _Bonferroni < 0.001. HC healthy controls, MDD major depressive disorder, MIST Montreal Imaging Stress Task.

Fig. 3. Neural stress responses in different runs.

A significant Time × Diagnosis × Sex interaction effect emerged in the (A) amygdala, (B) medial orbitofrontal cortex, and (C) nucleus accumbens. Estimated-mean are plotted, and error bar represent SE. HC healthy controls, MDD major depressive disorder. *p _Bonferroni < 0.05, **p _Bonferroni < 0.01, ***p _Bonferroni < 0.001.

Fig. 4. Neural network responses to acute psychosocial stress.

Group spatial maps of the (A) right frontoparietal network, (B) the left frontoparietal network, and (C) amygdala-NAc-ACC network. D Sex differences in the right frontoparietal network. E Sex × Diagnosis interaction effect in the amygdala-NAc-ACC network. F Diagnosis × Time interaction effect in the amygdala-NAc-ACC network. Estimated-mean are plotted, and error bars represent SE. HC healthy controls, MDD major depressive disorder, FPN frontoparietal network, NAc nucleus accumbens, ACC anterior cingulate cortex. *p _Bonferroni < 0.05, **p _Bonferroni < 0.01, ***p _Bonferroni < 0.001.