Stress-induced reduction in hippocampal volume and connectivity with the ventromedial prefrontal cortex are related to maladaptive responses to stressful military service

Abstract

Previous studies have shown that people who develop psychopathology such as posttraumatic stress disorder (PTSD) following stress exposure are characterized by reduced hippocampal (HC) volume and impaired HC functional connectivity with the ventromedial prefrontal cortex (vmPFC). Nevertheless, the exact interrelationship between reduced HC volume and HC-vmPFC connectivity deficits in the context of stress has yet to be established. Furthermore, it is still not clear whether such neural abnormalities are stress induced or precursors for vulnerability. In this study, we combined measurements of MRI, functional MRI (fMRI), and diffusion tensor imaging (DTI) to prospectively study 33 a priori healthy Israeli soldiers both pre- and post-exposure to stress during their military service. Thus, we were able to assess the contributions of structural and functional features of the HC and its connectivity to the onset and progression of maladaptive response to stress (i.e., increased PTSD symptoms post-exposure). We found that soldiers with decreased HC volume following military service (i.e., post-exposure) displayed more PTSD-related symptoms post-exposure as well as reduced HC-vmPFC functional and structural connectivity post-exposure, compared to soldiers with increased HC volume following military service. In contrast, initial smaller HC volume pre-exposure did not have an effect on any of these factors. Our results therefore suggest that reduction in HC volume and connectivity with the vmPFC together mark a maladaptive response to stressful military service. As stress-induced HC volume reductions were previously shown to be reversible, these localized biological markers may carry valuable therapeutic potential.

Keywords: DTI; MRI; PTSD; fMRI; functional connectivity; structural connectivity; uncinate fasciculus; volumetry.

Figure 1

Effects of stress exposure on HC volume. (A) The plot shows the change in HC volume of each soldier postexposure vs. pre‐exposure to stress. Soldiers were divided into two groups based on the direction of change (increased HC volume, white squares and decreased HC volume, black squares). Group averages are shown by gray cross. The individual value of HC volume change was normalized to percent change by dividing it by the individual initial HC volume, thus avoiding any effects of differences in initial HC volume. (B) Bar graphs representing the groups' average severity of PTSD‐related symptom at each time point. Note that the increase in symptom severity postexposure to stress was evident only for the soldiers with decreased HC volume (P < 0.005). (C) Bar graphs representing the groups' average strength of HC‐vmPFC functional connectivity at each time point. Note the overall increase in HC‐vmPFC functional connectivity postexposure to stress, which was more prominent for the soldiers with increased HC volume (P < 0.01). (D) Bar graphs representing the groups' average FA value of the UF (i.e., structural integrity) at each time point. Note that soldiers with increased HC volume had increased UF integrity postexposure relative to pre‐exposure (P < 0.05), whereas soldiers with decreased HC volume had decreased UF integrity postexposure relative to pre‐exposure (P < 0.05). (HC, hippocampus; UF, uncinate fasciculus; FA, fractional anisotropy; N = 33; error bars ± SEM).

Figure 2

Predisposing effects of HC volume. (A) The plot shows the initial standardized HC volume pre‐exposure of each soldier. Soldiers were divided into two groups based on this measure (above the median, white squares and below the median, black squares). Group averages are shown by gray cross. Note that this measure interacted neither with the severity of PTSD‐related symptoms (B), the strength of the HC‐vmPFC functional connectivity (C) nor with the average FA value of the UF (D). There were only two main effects in that analysis, the main effect of total increase in PTSD‐related symptoms (P < 0.05) and in HC‐vmPFC functional connectivity (P < 0.001) postexposure to stress. (HC, hippocampus; UF, uncinate fasciculus; FA, fractional anisotropy; N = 33; error bars ± SEM).