Anorexia nervosa (AN) is an eating disorder (ED) that predominantly affects women from adolescence to adulthood (sex ratio: 10/1 = F/M); its incidence and lifetime prevalence among females are 0.5-1% and 0.5%, respectively (DSM IV-TR, American Psychiatric Association, 2000).
Clinical features of AN are body image distortion and intense fear of becoming fat, with refusal of food and consequent severe emaciation (Cash and Deagle, 1997, Epstein et al., 2001, Seeger et al., 2002 and Wagner et al., 2003). Because of self-starvation, anorexic patients could suffer from several physical consequences, such as anemia, osteoporosis, amenorrhea and other endocrine dysfunctions. Notably, cerebral alterations have been described as among these consequences. Cerebral alterations relative to controls have been demonstrated in post mortem investigations (Gagel, 1953 and Martin, 1958) and in vivo by neuroimaging studies, using computed tomography (CT) (Enzmann and Lane, 1977, Kolhmeyer et al., 1983, Artmann et al., 1985, Dolan et al., 1988, Hoffman et al., 1989 and Addolorato et al., 1998) and magnetic resonance imaging (MRI) (Golden et al., 1996, Katzman et al., 1996, Kingston et al., 1996, Swayze et al., 1996, Swayze et al., 2003, Lambe et al., 1997, Inui et al., 2002 and Miwa et al., 2004). The most frequently described macroscopic anatomical brain changes are cerebral and cerebellar gray matter (GM) and white matter (WM) atrophy and ventricular enlargement. These findings have been globally distributed over the brain, suggesting a diffuse cerebral vulnerability. However atrophy has also been reported in some discrete areas, including paracentral lobule (Inui et al., 2002), thalamus, midbrain (Husain et al., 1992), mammillary bodies (Kingston et al., 1996) and extra-striate body areas (Suchan et al., 2010). On the other hand, some authors have failed to find significant total and focal GM volumetric differences between anorexics and healthy controls (Swayze et al., 1996 and Swayze et al., 2003).
The inconsistency in findings can partly be explained by the reversibility of these lesions. Indeed, longitudinal studies have reported that both GM and WM abnormalities are partially reversible after body weight restoration (GM: Kingston et al., 1996, Katzman et al., 1997, Swayze et al., 1996 and Swayze et al., 2003; WM: Swayze et al., 2003).
Similar results have been reported for intracranial cerebrospinal fluid (CSF) volume. It has been found to be increased in the supratentorial ventricular compartments (Artmann et al., 1985 and Kingston et al., 1996) and in the subarachnoid spaces (Krieg et al., 1988, Hentschel et al., 1995 and Swayze et al., 2003), and restored by body weight gain (Golden et al., 1996 and Swayze et al., 2003). These findings have been recently confirmed in both short (Castro-Fornieles et al., 2009) and long-term recovered patients (Wagner et al., 2006), although persistent total GM and bilateral anterior cingulate cortical decrease after recovery has also been observed (Mühlau et al., 2007). Based on these abnormal brain findings, recent studies focused on the cognitive performance in patients with AN; however, correlations between cerebral morphological abnormalities and cognitive performance in AN patients have tended to be weak (Palazidou et al., 1990) or absent (Lankenau et al., 1985, Laessle et al., 1989 and Kingston et al., 1996). More interesting, a recent study suggested a morpho-functional interaction in the brain region implicated in body image processing, showing a focal alteration in the extrastriate body area in anorexic patients (Suchan et al., 2010). This result is very intriguing, suggesting a brain alteration that is likely more related to the etiopathogenesis or at least to the perpetuation of the disease than to the consequence of the disease.
Methodological differences, such as imaging techniques (i.e. CT and MRI) and morphometric methods for brain anatomy evaluation, used in the above-mentioned studies could explain some of the contradictory findings. These limitations may be overcome by voxel-based morphometry (VBM), an unbiased automated technique developed to characterize morphological brain differences across groups of subjects (Ashburner and Friston, 2000, Good et al., 2001a and Good et al., 2001b).
The aim of the present study was to characterize brain abnormalities in AN using VBM in two groups of unrecovered AN patients with different disease durations.
