Neuroinflammation in the Amygdala Is Associated With Recent Depressive Symptoms

Abstract

Converging evidence suggests that elevated inflammation may contribute to depression. Yet, the link between peripheral and neuro-inflammation in depression is unclear. Here using data from the UK Biobank, we estimated associations among depression, C-reactive protein as a measure of peripheral inflammation (CRP), and neuroinflammation as indexed by diffusion-basis spectral imaging-based restricted fraction (DBSI-RF). DBSI-RF was derived from diffusion-weighted imaging data (n=11,512) for whole-brain gray matter (global-RF), and regions of interest in bilateral amygdala (amygdala-RF) and hippocampus (hippocampus-RF), and CRP was estimated from blood (serum) samples. Self-reported recent depression symptoms were measured using a 4-item assessment. Linear regressions were used to estimate associations between CRP and DBSI-RFs with depression, while adjusting for the following covariates: Age, sex, body mass index, smoking, drinking, and medical conditions. Elevated CRP was associated with higher depression symptoms (β=0.04, pfdr<0.005) and reduced global-RF (β=-0.03, pfdr<0.001). Higher amygdala-RF was associated with elevated depression - an effect resilient to added covariates and CRP (β=0.02, pfdr<0.05). Interestingly, this association was stronger in individuals with a lifetime history of depression (t β=0.07, p<0.005) than in those without (β=0.03, p<0.05). Associations between global-RF or hippocampus-RF with depression were not significant, and no DBSI-RF indices indirectly linked CRP with depression (i.e., mediation effect). Peripheral inflammation and DBSI-RF neuroinflammation in the amygdala are independently associated with depression, consistent with animal studies suggesting distinct pathways of peripheral and neuro-inflammation in the pathophysiology of depression, and with investigations highlighting the role of the amygdala in stress-induced inflammation and depression.