Restless 'rest': intrinsic sensory hyperactivity and disinhibition in post-traumatic stress disorder

Abstract

Post-traumatic stress disorder is characterized by exaggerated threat response, and theoretical accounts to date have focused on impaired threat processing and dysregulated prefrontal-cortex-amygdala circuitry. Nevertheless, evidence is accruing for broad, threat-neutral sensory hyperactivity in post-traumatic stress disorder. As low-level, sensory processing impacts higher-order operations, such sensory anomalies can contribute to widespread dysfunctions, presenting an additional aetiological mechanism for post-traumatic stress disorder. To elucidate a sensory pathology of post-traumatic stress disorder, we examined intrinsic visual cortical activity (based on posterior alpha oscillations) and bottom-up sensory-driven causal connectivity (Granger causality in the alpha band) during a resting state (eyes open) and a passive, serial picture viewing state. Compared to patients with generalized anxiety disorder (n = 24) and healthy control subjects (n = 20), patients with post-traumatic stress disorder (n = 25) demonstrated intrinsic sensory hyperactivity (suppressed posterior alpha power, source-localized to the visual cortex-cuneus and precuneus) and bottom-up inhibition deficits (reduced posterior→frontal Granger causality). As sensory input increased from resting to passive picture viewing, patients with post-traumatic stress disorder failed to demonstrate alpha adaptation, highlighting a rigid, set mode of sensory hyperactivity. Interestingly, patients with post-traumatic stress disorder also showed heightened frontal processing (augmented frontal gamma power, source-localized to the superior frontal gyrus and dorsal cingulate cortex), accompanied by attenuated top-down inhibition (reduced frontal→posterior causality). Importantly, not only did suppressed alpha power and bottom-up causality correlate with heightened frontal gamma power, they also correlated with increased severity of sensory and executive dysfunctions (i.e. hypervigilance and impulse control deficits, respectively). Therefore, sensory aberrations help construct a vicious cycle in post-traumatic stress disorder that is in action even at rest, implicating dysregulated triangular sensory-prefrontal-cortex-amygdala circuitry: intrinsic sensory hyperactivity and disinhibition give rise to frontal overload and disrupt executive control, fuelling and perpetuating post-traumatic stress disorder symptoms. Absent in generalized anxiety disorder, these aberrations highlight a unique sensory pathology of post-traumatic stress disorder (ruling out effects merely reflecting anxious hyperarousal), motivating new interventions targeting sensory processing and the sensory brain in these patients.

Keywords: Granger causality; alpha/gamma oscillations; post-traumatic stress disorder (PTSD); resting state; sensory hyperactivity.

Figure 1

Alpha and gamma power. (A) Power spectra averaged over occipitoparietal electrodes for each group at each state. Shaded ribbons = standard error of the mean (SEM). (B) Alpha power magnitudes at left, middle and right posterior sites. Error bars = SEM. (C) Scalp topographical maps of alpha powers, with electrodes included in posterior (occipitoparietal) sites bolded and circled. (D) eLORETA localized the source of alpha activity to the bilateral cuneus/precuneus. (E) Power spectra averaged over frontal electrodes for each group at each state. Shaded ribbons = SEM. (F) Gamma power averages at left and right frontal sites. Error bars = SEM. (G) Scalp topographical maps of gamma powers, with frontal electrodes bolded and circled. (H) eLORETA localized the source of gamma activity to the bilateral prefrontal cortex, including superior frontal gyrus and dorsal anterior cingulate cortex. ^*P < 0.05; ^**P < 0.01. HC = healthy controls.

Figure 2

Granger causality. (A) S-RS/M-RS: Reduction in right hemisphere posterior→frontal and frontal→posterior alpha causality in PTSD (versus GAD/healthy controls). (B) Average Granger causality values for both directions in both hemispheres. BU = bottom-up; TD = top-down. Error bars = SEM. ^*P < 0.05; ^**P < 0.01; ^***P < 0.005; ^†P < 0.1.

Figure 3

Correlation analyses. (A) Correlations across alpha/gamma powers and Granger causality. Pearson r’s for pairwise correlations in both states (S-RS/M-RS). Solid black lines indicate significant correlations in both states; the solid grey line indicates correlation in M-RS only (r = 0.35). Dashed grey lines indicate non-significant correlations in either state. (B) Correlations between bottom-up Granger causality and severity in hypervigilance and difficulty in impulse control (correlations remained significant after co-varying out substance use). BU = bottom-up; TD = top-down. ^*P < 0.05; ^**P < 0.01; ^***P < 0.005.

Figure 4

Schematic presentation of a sensory hypothesis of PTSD. A vicious cycle in action during resting state, entailing (i) sensory hyperactivity; (ii) deficient bottom-up (BU), sensory inhibition; (iii) frontal overload (due to sensory overflow); and (iv) deficient top-down (TD), executive inhibition and regulation.