Patients with posttraumatic stress disorder (PTSD) exhibit smaller regional brain volumes in commonly reported regions including the amygdala and hippocampus, regions associated with fear and memory processing. In the current study, we have conducted a voxel-based morphometry (VBM) meta-analysis using whole-brain statistical maps with neuroimaging data from the ENIGMA-PGC PTSD working group.

T1-weighted structural neuroimaging scans from 36 cohorts (PTSD n = 1309; controls n = 2198) were processed using a standardized VBM pipeline (ENIGMA-VBM tool). We meta-analyzed the resulting statistical maps for voxel-wise differences in gray matter (GM) and white matter (WM) volumes between PTSD patients and controls, performed subgroup analyses considering the trauma exposure of the controls, and examined associations between regional brain volumes and clinical variables including PTSD (CAPS-4/5, PCL-5) and depression severity (BDI-II, PHQ-9).

PTSD patients exhibited smaller GM volumes across the frontal and temporal lobes, and cerebellum, with the most significant effect in the left cerebellum (Hedges’ g = 0.22, pcorrected = .001), and smaller cerebellar WM volume (peak Hedges’ g = 0.14, pcorrected = .008). We observed similar regional differences when comparing patients to trauma-exposed controls, suggesting these structural abnormalities may be specific to PTSD. Regression analyses revealed PTSD severity was negatively associated with GM volumes within the cerebellum (pcorrected = .003), while depression severity was negatively associated with GM volumes within the cerebellum and superior frontal gyrus in patients (pcorrected = .001).

PTSD patients exhibited widespread, regional differences in brain volumes where greater regional deficits appeared to reflect more severe symptoms. Our findings add to the growing literature implicating the cerebellum in PTSD psychopathology.

Recent stressful life events (SLEs) are an established risk factor for a range of psychiatric disorders. Animal studies have shown evidence of gray matter (GM) reductions associated with stress, and previous work has found similar associations in humans. However longitudinal studies investigating the association between stress and changes in brain structure are limited.

The current study uses longitudinal data from the UK Biobank and comprises 4,543 participants with structural neuroimaging and recent SLE data (mean age = 61.5 years). We analyzed the association between recent SLEs and changes in brain structure, determined using the longitudinal FreeSurfer pipeline, focusing on total GM volume and five a priori brain regions: the hippocampus, amygdala, anterior cingulate cortex, orbitofrontal cortex, and insula. We also examined if depression and childhood adversity moderated the relationship between SLEs and brain structure.

Individuals who had experienced recent SLEs exhibited a slower rate of hippocampal decrease over time compared to individuals who did not report any SLEs. Individuals with depression exhibited smaller GM volumes when exposed to recent SLEs. There was no effect of childhood adversity on the relationship between SLEs and brain structure.

Our findings suggest recent SLEs are not directly associated with an accelerated decline in brain volumes in a population sample of older adults, but instead may alter brain structure via affective disorder psychopathology. Further work is needed to investigate the effects of stress in younger populations who may be more vulnerable to stress-induced changes, and may yet pinpoint brain regions linked to stress-related disorders.