Few studies have used Voxel-Based Morphometry (VBM) to examine brain structure in Anorexia Nervosa patients. The purpose of the present study was to investigate a sample of Anorexia Nervosa restrictive type (AN-r) adolescent patients in the early stages of the illness, using VBM in order to characterize morphometric gray matter (GM) changes. Participants were 16 AN-r female patients (with no other psychiatric disorders) whose AN-r had been in progress for less than 12 months and 16 age-matched healthy female subjects. High-resolution T1-weighted magnetic resonance images were preprocessed according to the optimized VBM method, and statistically analyzed. The analyses revealed a significant global GM decrease in the AN-r patients; furthermore, a significant region-specific decrease in GM volume was found bilaterally in the middle cingulate cortex, the precuneus, and the inferior and superior parietal lobules. The significant early GM decrease in the aforementioned regions in AN-r adolescent patients suggests that there might be a region-specific GM vulnerability that could play a role in the pathophysiology of the disease. Given that these regions are also involved in the manipulation of mental images and the mental representation of the self, this might explain the presence of a distorted body image in these patients.
Anorexia Nervosa (AN) is a disorder that generally begins in adolescence and primarily affects females. Its relevance is not only related to the increased incidence of AN over the last few decades (Lucas et al., 1999), but also to the high occurrence of medical complications, which can occasionally prove fatal (Harris and Barraclough, 1998).
The DSM-IV-TR classification of AN recognizes two subtypes: AN restrictive type (AN-r) and AN binge/purging type (AN-b/p).
The use of advanced brain-imaging techniques (functional magnetic resonance imaging, positron emission tomography, and single photon emission computed tomography) allowed the identification of several cerebral regions involved in AN. However, the direct involvement of these regions in the onset of AN still remains unclear and under debate (Chowdhury et al., 2003).
In previously published neuroimaging structural studies, most authors (Dolan et al., 1988, Krieg et al., 1988, Golden et al., 1996 and Kohn et al., 1997) showed a global gray matter (GM) and white matter (WM) decrease with a cerebral spinal fluid (CSF) increase, while Swayze et al. (2003) demonstrated a WM decrease and a CSF increase, with no significant GM decrease. Some authors indicated a relationship between brain mass loss and severity of AN (Kohlmeyer et al., 1983, Golden et al., 1996 and Katzman et al., 1996), a full reversibility of brain mass loss (Golden et al., 1996, Swayze et al., 1996 and Chowdhury et al., 2003), and persistence of GM decrease (Katzman et al., 1997 and Lambe et al., 1997).
A few authors used magnetic resonance imaging (MRI) and manual volume measurement in AN patients demonstrating a significant volume reduction in the hypophysis (Doraiswamy et al., 1990) and the amygdala–hippocampus complex (Giordano et al., 2001).
In recent years, the availability of the automated Voxel-Based Morphometry (VBM) technique has enabled us to make a more reliable global and regional evaluation of GM, WM and CSF volume and concentration alterations. VBM allows for a voxel-by-voxel assessment of differences across the whole brain, thus enabling us to evaluate regional volumetric alterations without a priori hypotheses about their localization ( Ashburner and Friston, 2000). This approach is also minimally operator-dependent, in that it avoids the potential confounds and challenges associated with Regions of Interest hand-tracing methods ( Gilbert et al., 2008). VBM is a group analysis technique, which enables operators to reveal statistically significant local morphological differences between samples: the experimental question it seeks to answer is whether the clinical sample shows specific structural features that could be related to the pathology. Thus a reliable analysis must rely on homogeneous samples of appropriate size. In AN patients this technique produced differing results: Wagner et al. (2006) found no GM, WM and CSF differences between 40 recovered patients with eating disorders (ED) and a healthy control group; Mühlau et al. (2007) demonstrated, in a sample of 22 recovered AN patients, a global GM loss of 1% and a region-specific GM loss of 5% in the anterior cingulate cortex; Castro-Fornieles et al. (2009) in 12 AN patients (9 AN-r and 3 AN-b/p) showed a global GM decrease which normalized at follow-up (after 7 months and weight recovery) and several regions particularly affected (temporal and parietal areas), although only left and right supplementary motor areas and middle cingulate cortex remained significantly altered at follow-up.
The purpose of the present study is to perform, via VBM, a global and local GM analysis in a sample of adolescent patients whose AN-r had been in progress for less than 12 months at the time of scanning.
