Sexually dimorphic gray matter volume reduction in patients with panic disorder

Abstract

While clinical features of panic disorder show significant sexual dimorphism, previous structural MRI studies have not sufficiently controlled for sex when looking at regional brain abnormalities in panic disorder. Using optimized voxel-based morphometry (VBM), regional gray matter volume was compared between 24 patients (male/female: 9/15) with panic disorder and 24 healthy subjects matched for age and sex. Significant gray matter volume reductions were found in the bilateral dorsomedial and right ventromedial prefrontal cortices, right amygdala, anterior cingulate cortex, bilateral insular cortex, occipitotemporal gyrus and left cerebellar vermis in the patients compared with the controls. Among these regions, the VBM revealed significant sexual dimorphism: volume reduction in the right amygdala and the bilateral insular cortex was significantly greater in the males, while reduction in the right superior temporal gyrus was greater in females. Furthermore, a significant reduction in the dorsolateral and ventrolateral prefrontal cortices, thalamus, and parietal cortex was specific to the female patients. The present study demonstrated the morphological changes in extensive brain regions of patients with panic disorder compared with the sex-matched controls. The current results further suggested that the sexually dimorphic clinical phenotypes of panic disorder might have a neurobiological background even at the structural level of the brain.

Introduction

Many functional neuroimaging studies have indicated brain abnormalities in patients with panic disorder. Pillay et al. (2006), for example, demonstrated an association between activation in the amygdala and cingulate cortex and fearful-neutral face recognition in patients with panic disorder using functional MRI. Schunck et al. (2006) reported that the cerebellar vermis, temporal pole, insular cortex and thalamus were activated when panic attacks were induced in healthy subjects by injecting them with cholecystokinin tetrapeptide (CCK-4). Furthermore, Fischer et al. (1998) used positron emission tomography (PET) to show that regional cerebral blood flow (rCBF) decreased in the right orbitofrontal cortex, anterior cingulate cortex (ACC), and anterior temporal cortex when a patient unexpectedly experienced a panic attack,. These findings suggest that extensive cerebrocortical regions – especially the prefrontal cortex, limbic structures and temporal cortex – were associated with the pathophysiology of panic disorder. Coplan and Lydiard (1998) suggested that dysfunction of the prefrontal cortex and excessive activation of the limbic structures, which include the amygdala, play an important role in panic disorder. Building upon that idea, Gorman et al. (2000) suggested that, in addition to the amygdala and prefrontal cortex, several other regions, including the hippocampus, thalamus, insular cortex and cingulate cortex, also contributed to the pathogenesis of panic disorder.

Nevertheless, structural imaging studies could show only four significant abnormal regions separately in patients with panic disorder, volume reduction of the amygdala (Massana et al., 2003b), parahippocampus (Massana et al., 2003a), putamen (Yoo et al., 2005) and temporal lobe (Fontaine et al., 1990, Vythilingam et al., 2000, Uchida et al., 2003). Other studies have reported a less significant volume reduction of the right inferior frontal gyrus, left superior frontal gyrus and left superior temporal gyrus (STG) (Yoo et al., 2005), as well as a relatively insignificant volume increase of the brainstem (Protopopescu et al., 2006), and no change in the hippocampus (Massana et al., 2003b). Therefore, to our knowledge, no previous studies have shown a significant concurrent volume reduction in cerebrocortical regions considered to be interconnected and implicated in panic disorder.

Several epidemiological studies have described sex differences in panic disorder; for example, the prevalence is slightly more than twice as high among women as among men (Von Korff et al., 1985). While the highest rate for panic attacks among men is in the age range of 15–24 years, for women the peak is from 35–44 years (Eaton et al., 1994). These epidemiological findings indicate the existence of sex differences with a neural background in panic disorder. There are some reports that show sex differences in brain response to fear/anxiety stress. According to these previous reports, sex differences in regional brain response to fearful/anxious stimuli were found in the amygdala, ACC and insula although the results were various (e.g., Butler et al., 2005, Hamann, 2005). Sex differences in brain anatomy have also been reported in previous studies. For example, Im et al. (2006) described significantly greater cortical thickening in the right STG, generally identified as an anxiety-related region in panic disorder (e.g. Bisaga et al., 1998, Yoo et al., 2005), in healthy female subjects than male subjects. Recent studies have further reported sexual dimorphism in neuroanatomical correlates, including the prefrontal cortex and limbic structures, of anxiety-related personality traits in healthy adults (Yamasue et al., 2007). However, no previous study has examined sexual dimorphism in the neuroanatomical background of panic disorder.

The present study hypothesized that the whole picture of gray matter volume reduction in the prefrontal cortex, limbic structures such as the amygdala, ACC, and insular cortex, temporal cortex and cerebellar vermis in patients with panic disorder would be clarified by carefully controlling for sex and comparing the patients with matched healthy controls. Furthermore, the sexual dimorphism of gray matter volume reduction in patients was examined using sex-by-diagnosis interaction analysis.