Structural white and gray matter differences in a large sample of patients with Posttraumatic Stress Disorder and a healthy and trauma-exposed control group: Diffusion tensor imaging and region-based morphometry

Abstract

Differences in structural white and gray matter in survivors of traumatic experiences have been related to the development and maintenance of Posttraumatic Stress Disorder (PTSD). However, there are very few studies on diffusion tensor imaging and region based morphometry comparing patients with PTSD to two control groups, namely healthy individuals with or without trauma experience. It is also unknown if differences in white and gray matter are associated. In this cross-sectional study, we examined white- and gray matter differences between 44 patients with PTSD, 49 trauma control and 61 healthy control subjects. We compared the groups applying Tract-Based Spatial Statistics (TBSS) for a whole brain white matter analysis as well as region of interest analyses for white and gray matter. First, trauma control subjects in comparison to patients with PTSD and healthy control subjects showed significantly a) higher fractional anisotropy (FA) in the left corticospinal tract and inferior fronto-occipital fasciculus than patients with PTSD, b) higher FA in the left inferior fronto-occipital-, right inferior- and right superior longitudinal fasciculi, c) higher FA in the forceps minor and d) higher volume of the left and right anterior insulae. Second, we show significant correlations between the FA in the forceps minor and the gray matter volume in the left and right anterior insulae. Third, the mean FA value in the forceps minor correlated negatively with symptom severity of PTSD and depression as well as trait anxiety, whereas the gray matter volume in the left anterior insula correlated negatively with symptom severity in PTSD. Our findings underline the importance of brain structures critically involved in emotion regulation and salience mapping. While previous studies associated these processes primarily to functional and task-based differences in brain activity, we argue that morphometrical white and gray matter differences could serve as targets in neuroscientifically-informed prevention and treatment interventions for PTSD.

Keywords: Diffusion tensor imaging; Neuroplasticity; PTSD; Trauma; White and gray matter.

Fig. 1

Diffusion Tensor Imaging. a) Results of TBSS analyses comparing the fractional anisotropy between patient with PTSD (n = 44), TC (n = 49) and HC (n = 61) subjects in an ANCOVA (yellow), between TC>HC subjects in a post-hoc t-test (red) and between TC>PTSD in a post-hoc t-test (blue). Age and sex were included as covariates in the analyses. All results are FWE-corrected (α <.05). b) Boxplots with significant (α_{bonferroni_cor} =.05/20 =.0025) differences in mean FA value of the forceps minor between patients with PTSD (n = 44), TC (n = 49) and HC (n = 61) subjects (n = 154). c) Significant correlation (α_{bonferroni_cor} =.05/4 =.0125) between mean FA value in the forceps minor and the mean CAPS score for TC subjects (n = 49) and patients with PTSD (n = 44). d) Anatomical images with mean FA skeleton used for the TBSS analysis (in green). The contrast between TC subjects and patients with PTSD is marked in yellow to red. The forceps minor is marked in blue as a region of interest for clarification. [Abbreviations: ANCOVA - Analysis of Covariance; CAPS - Clinician-Administered PTSD Scale; FA - Fractional anisotropy; FM - Forceps minor; FWE - Family-wise error correction; HC - Healthy control subjects; I - Inferior; L - Left; P - Posterior; PTSD - Patients with posttraumatic stress disorder; R - Right; S - Superior; TBSS - Tract-based spatial statistics; TC - Trauma control subjects; * α <.05; ** α <.01; *** α <.001].

Fig. 2

Region Based Morphometry. a) Boxplots with mean volume of left anterior insula (in cm³) in all three groups. The results show a significant ( α_{bonferroni_cor}=.05/14 =.0036) difference in volume between patients with PTSD (n=42), TC (n=47) and HC (n=58) subjects. Post-hoc t-tests revealed significant differences in volume for the contrasts TC>HC subjects and TC>PTSD. b) Significant negative correlation ( α_{bonferroni_cor}=.05/5 =.0125) between mean volume of left anterior insula (in cm³) and the mean CAPS score for TC subjects (n = 47) and patients with PTSD (n = 42). c) Outline of the left and right anterior insulae. d) Boxplots with mean volume of right anterior insula (in cm³) in all three groups. The results show a significant ( α_{bonferroni_cor}=.05/14=.0036) difference in volume between patients with PTSD (n = 42), TC (n = 47) and HC (n = 58) subjects. Post-hoc t-tests revealed significant differences in volume for the contrasts TC>HC subjects and TC>PTSD. [Abbreviations: CAPS - Clinician-Administered PTSD Scale; HC - Healthy control subjects; lAntIns - Left anterior insula; PTSD - Patients with posttraumatic stress disorder; rAntIns - Right anterior insula; TC - Trauma control subjects; * α <.05; ** α <.01; *** α <.001].

Fig. 3

White and gray matter coupling. Significant positive correlation ( α_{bonferroni_cor} =.05/4 =.0125) between mean FA value in forceps minor and a) volume of the lAntIns (PTSD, n = 42; TC, n = 47; HC, n = 58) and b) volume of the rAntIns (PTSD, n = 42; TC, n = 47; HC, n = 58). [Abbreviations: HC - Healthy control subjects; lAntIns - Left anterior insula; PTSD - Patients with posttraumatic stress disorder; rAntIns - Right anterior insula; TC - Trauma control subjects; * α <.05; ** α <.01; *** α <.001].