Thinner prefrontal cortex in veterans with posttraumatic stress disorder

Abstract

Structural neuroimaging studies in posttraumatic stress disorder (PTSD) have focused primarily on structural alterations in the medial temporal lobe, and only a few have examined grey matter reductions in the cortex. Recent advances in computational analysis provide new opportunities to use semi-automatic techniques to determine cortical thickness, but these techniques have not yet been applied in PTSD. Twenty-five male veterans with PTSD and twenty-five male veterans without PTSD matched for age, year and region of deployment were recruited. All the subjects were scanned using MRI. Subjects' brains were aligned using cortex-based alignment in a region of interest based approach. Individual cortical thickness maps were calculated from the MR images. Regions of interest examined included the bilateral superior frontal gyri, bilateral middle frontal gyri, bilateral inferior frontal gyri, bilateral superior temporal gyri, and bilateral middle temporal gyri. In a large number of patients and controls, IQ scores and memory scores were also obtained. Individual cortical thickness maps were calculated from the MR images. Veterans with PTSD revealed reduced cortical thickness in the bilateral superior and middle frontal gyri, the left inferior frontal gyrus, and the left superior temporal gyrus. Veterans with PTSD performed significantly worse on memory measures compared to control veterans. Cortical thickness correlated with memory measures in the veterans without PTSD, but not in the veterans with PTSD. Cortical thinning in these regions may thus correspond to functional abnormalities observed in patients with PTSD.

Introduction

Neuroimaging studies in posttraumatic stress disorder (PTSD) have identified a number of functional and structural alterations that are associated with this disorder. Structural neuroimaging studies in PTSD have focused primarily on manual based hippocampal volumetry. These studies have shown that patients with PTSD are associated with bilateral smaller hippocampal volume (Geuze et al., 2005, Kitayama et al., 2005). A recent meta-analysis, confirmed the presence of significantly smaller hippocampal volumes and left amygdala volumes in patients with PTSD compared to controls with and without trauma exposure (Karl et al., 2006).

However, while considerable attention has been placed on volume changes in subcortical structures, very little attention has been directed to grey matter reductions in the cortex. Manual tracing of the anterior cingulate cortex has shown that this area is also smaller in patients with PTSD compared to controls (Woodward et al., 2006, Rauch et al., 2003). Several voxel-based morphometry studies, a whole-brain analysis technique, have been applied in PTSD. These studies have shown reduced grey matter density in the anterior cingulate gyrus, hippocampus, and insula (Chen et al., 2006, Corbo et al., 2005, Yamasue et al., 2003, Kasai et al., 2007). A recent meta-analysis of these studies confirmed that patients with PTSD revealed significantly smaller anterior cingulate cortex volume compared to trauma-exposed controls (Karl et al., 2006). In pediatric patients with PTSD, significantly smaller prefrontal lobe volume and corpus callosum volume have been found (De Bellis et al., 1999, De Bellis et al., 2002b). The findings of smaller corpus callosum volume have not been replicated in a sample of adult patients with PTSD (Kitayama et al., 2007). Functional neuroimaging studies have also identified a number of brain areas with altered activity in patients with PTSD, including the prefrontal cortex, temporal cortex, insula, amygdala, and hippocampus (Francati et al., 2007, Nemeroff et al., 2006, Hull, 2002).

Cortical thickness and volumetric analysis receive increasing interest in the imaging community because it allows relation of cognitive abilities, effects of aging, and effects of diseases to (subtle) structural changes in the brain. Practical constraints, however, related to intensive measurement technique and the use of postmortem samples, make it difficult to compare differences in cytoarchitecture across the cortex. Whole brain volumetric techniques are also compromised by large inter-individual differences in neuroanatomy making it difficult to match the same macroscopic regions between subjects. Recent advances in computational analysis, provide new opportunities to use imaging data to derive more knowledge on cortical thickness. In this study, we have applied the new cortical thickness analysis tools in Brainvoyager QX to compare veterans with PTSD to veterans without PTSD. In order to maximize the spatial correspondence mapping between cortical macrostructures, we also made use of the new technique of cortex-based alignment, which aligns reconstructed cortices using curvature information reflecting the gyral/sulcal folding pattern. It has been shown that a cortical matching approach substantially improves statistical analysis across subjects by reducing anatomical variability (Fischl et al., 1999a, Fischl et al., 1999b, Dale et al., 1999).

In this study our goal was to employ this relatively new technique to examine cortical thickness in patients with PTSD. We hypothesized that patients with PTSD would reveal reduced cortical thickness in frontal and temporal brain areas. Due to technical limitations of the cortical thickness algorithm, which performed less accurately in the inferior temporal lobe, and medial prefrontal cortex, we only examined cortical thickness in the bilateral superior, middle, and inferior frontal gyri, as well as the bilateral superior and middle temporal gyri.