Hypothalamic-pituitary-adrenal axis genetic variation and early stress moderates amygdala function

Abstract

Early life stress may precipitate psychopathology, at least in part, by influencing amygdala function. Converging evidence across species suggests that links between childhood stress and amygdala function may be dependent upon hypothalamic-pituitary-adrenal (HPA) axis function. Using data from college-attending non-Hispanic European-Americans (n=308) who completed the Duke Neurogenetics Study, we examined whether early life stress (ELS) and HPA axis genetic variation interact to predict threat-related amygdala function as well as psychopathology symptoms. A biologically-informed multilocus profile score (BIMPS) captured HPA axis genetic variation (FKBP5 rs1360780, CRHR1 rs110402; NR3C2 rs5522/rs4635799) previously associated with its function (higher BIMPS are reflective of higher HPA axis activity). BOLD fMRI data were acquired while participants completed an emotional face matching task. ELS and depression and anxiety symptoms were measured using the childhood trauma questionnaire and the mood and anxiety symptom questionnaire, respectively. The interaction between HPA axis BIMPS and ELS was associated with right amygdala reactivity to threat-related stimuli, after accounting for multiple testing (empirical-p=0.016). Among individuals with higher BIMPS (i.e., the upper 21.4%), ELS was positively coupled with threat-related amygdala reactivity, which was absent among those with average or low BIMPS. Further, higher BIMPS were associated with greater self-reported anxious arousal, though there was no evidence that amygdala function mediated this relationship. Polygenic variation linked to HPA axis function may moderate the effects of early life stress on threat-related amygdala function and confer risk for anxiety symptomatology. However, what, if any, neural mechanisms may mediate the relationship between HPA axis BIMPS and anxiety symptomatology remains unclear.

Keywords: Amygdala; Anxiety; Cortisol; Depression; Early life stress; Gene; HPA.

Figure 1

HPA axis BIMPS x ELS associated threat-related right amygdala reactivity across all participants (Y=−4; Right amygdala: max voxel: x= 28, y= −4, z= −20, t = 2.90, P_FWE-corr =0.002, 204 voxels; Left amygdala: max voxel: x= −22, y= −6, z= −18, t= 2.85, P_FWE-corr = 0.002, 179 voxels).

Figure 2

An interaction between the hypothalamic-pituitary-adrenal (HPA) axis biologically-informed multilocus profile scores (BIMPS) and early life stress (ELS) predicted right amygdala reactivity. ELS was significantly positively associated with right amygdala reactivity in individuals with high (t(308)=−2.04, p=0.04), but not low (t(308)=−1.53, p=0.13) or average (t(308)=0.55, p=0.58) BIMPS. Regions of significance from the Johnson Neyman tests are denoted in yellow (p<.05). BIMPS and ELS were analyzed as a continuous variable; or ease of visualization, we collapsed the ELS into three groupings: low (25–30), medium (31–37), and high (38–60). For alternative plotting and discussion see Supplemental Figure 1.

Figure 3

HPA Axis BIMPS predicted elevated anxious arousal (β=0.168, p=0.003). BIMPS were analyzed as a continuous measure but are collapsed into low, average and high groupings for visualization.