Anterior cingulate activity to salient stimuli is modulated by autonomic arousal in Posttraumatic Stress Disorder

Abstract

Reduced ventral anterior cingulate (vACC) activity to threat is thought to reflect an impairment in regulating arousal networks in posttraumatic stress disorder (PTSD). Concurrent functional magnetic resonance imaging (fMRI) and skin conductance response (SCR) recording were used to examine neural functioning when arousal networks are engaged. Eleven participants with PTSD and 11 age- and sex-matched non-traumatized controls performed an oddball task that required responding to salient, non-trauma-related auditory target tones embedded in lower frequency background tones. Averaged target-background analyses revealed significantly greater dorsal ACC, supramarginal gyrus, and hippocampal activity in PTSD relative to control participants.With-SCR target responses resulted in increased vACC activity in controls, and dorsal ACC activity in PTSD. PTSD participants had reduced vACC activity relative to controls to target tones when SCR responses were present. This reduction in vACC in PTSD relative to controls was not apparent in without-SCR responses. These findings suggest that a reduction in vACC in PTSD occurs specifically when arousal networks are engaged.

Introduction

A disturbance in the regulation of arousal is a core pathology implicated in posttraumatic stress disorder (PTSD: Frewen and Lanius, 2006). Although many neuroimaging studies examine limbic activity in response to threat in PTSD, no imaging studies have directly examined neural activity in relation to autonomic arousal.

Neuroimaging studies in PTSD have reported impairments in ventral and rostral anterior cingulate activity (ACC: Bremner et al., 1999, Shin et al., 2005, Williams et al., 2006a) and ventral medial prefrontal regions (Lanius et al., 2001) in response to threatening stimuli. Some neuroimaging studies in PTSD have reported concurrent increases in amygdala activity (Shin et al., 2005, Williams et al., 2006a). The ventral portion of the ACC is thought to regulate affective arousal and limbic networks (Devinsky et al., 1995, Bush et al., 2000), and ventro-medial prefrontal regions have been implicated in fear-extinction processes (Phelps et al., 2004). These findings have led to a neurobiological model of PTSD in which reductions in ventromedial PFC inhibitory activity are thought to lead to heightened amygdala reactivity to threat (Frewen and Lanius, 2006, Rauch et al., 2006.

Reductions in anterior cingulate activity in PTSD have not always been found. In a recent study, we examined fMRI responses to salient, non-threatening tones in a selective attention oddball paradigm and found increased dorsal and rostral ACC activity coupled with increased left amygdala reactivity in PTSD relative to control participants (Bryant et al., 2005). We concluded that the reductions in rostral ACC activity may be specific to stimuli that engage arousal networks (such as threatening stimuli).

The present study aims to test this hypothesis by examining neural activity associated with autonomic arousal in PTSD in response to salient stimuli in a cognitive task. The selective attention oddball task reliably engages the ACC (Bryant et al., 2005, Yoshiura et al., 1999) and will allow us to examine the impact of arousal, as indexed by the skin conductance response (SCR) without the potential confound of emotional stimuli. In a recent functional magnetic resonance imaging (fMRI) oddball study in healthy controls, we found that responses to with-SCR target tones were associated with greater activity in amygdala, ventromedial and lateral frontal regions, whereas responses to without-SCR targets were associated with activity in dorsolateral frontal cortex and supramarginal gyrus (Williams et al., 2007).

In the current study, we compare PTSD participants with age- and sex-matched healthy controls on their responses to targets that evoke autonomic arousal in an auditory oddball paradigm. If reductions in vACC activity in PTSD require the engagement of arousal networks, vACC activity should be reduced in PTSD relative to control participants in targets with SCR, and this effect should not be apparent for targets without SCR. Enhanced amygdala activity should be expected to occur in the PTSD group for targets with SCR relative to controls.