In women, the lifetime prevalence of anorexia nervosa (AN) is 0.9%, and of bulimia nervosa (BN) 1.5% (Hudson et al., 2007), the core psychopathology consisting of drive for thinness (DT) and characteristic disturbances of eating behaviour. Structural imaging studies have shown cerebral atrophy in AN, including the anterior cingulate cortex (ACC) (Castro-Fornieles et al., 2009, Krieg et al., 1989, Joos et al., 2010, McCormick et al., 2008 and Mühlau et al., 2007). Potential pseudoatrophic changes in BN (Krieg et al., 1989) could not be verified by a recent voxel-based morphometry investigation (Joos et al., 2010). Functional magnetic resonance imaging demonstrated increased perfusion of the ACC in both, AN and BN, when confronted with disease specific stimuli, i.e. food (Schienle et al., 2009 and Uher et al., 2004). The ACC is a critical converging zone for cognitive and affective processing (Bush et al., 2000). Furthermore, the ACC is involved in processes of physiological eating behaviour (Schienle et al., 2009 and Führer et al., 2008). Decreased resting perfusion of the ACC and medial prefrontal cortex was reported in AN (Takano et al., 2001). Other brain regions implicated in the pathogenesis of eating disorders (ED) concern the parietal cortex in particular (Frank et al., 2004, Uher et al., 2004, Uher et al., 2005 and Castro-Fornieles et al., 2009)—the inferior part of which being essential for sensory integration of egocentric coordinates (Joos et al., 2010 and Medina et al., 2009). In AN, the right amygdala has received attention (Vocks et al., 2010, Miyake et al., 2010 and Seeger et al., 2002).
One tool for the non-invasive assessment of cerebral neurochemistry is magnetic resonance spectroscopy (MRS). 1H-MRS studies organic molecules of a region of interest depending on the energy absorbed by hydrogen atoms by the compound and its environment (Licata and Renshaw, 2010). Very few MRS studies have been carried out in ED patients, and they exclusively concentrated on those with AN. Investigated brain areas varied greatly. Ohrmann et al. (2004) described a reduction of the combined glutamate (Glu) + glutamine (Gln) signal (i.e. Glx) in the ACC, which was attributed to depressive symptoms, however. They reported reduced levels of myo-inositol (mI) and creatine (Cr) in the dorsolateral prefrontal cortex, in addition. Castro-Fornieles et al. (2007) reported a reduction of N-acetyl-aspartate (NAA), Glu and mI in the medial frontal cortex in 12 adolescent AN patients.
An earlier study by Roser et al. (1999) found a decrease of lipids in frontal white matter (WM) and occipital grey matter (GM) as well as a reduction of mI in the former, while another study reported increased choline (Cho) levels in the occipital WM (Mockel et al., 1999).
In summary, MRS studies have not yet yielded a clear neurochemical pattern as a hallmark of ED. Regions of interest varied, and some studies assessed metabolite ratios and not absolute signals, which make the interpretation of putative signals difficult.
Given this background, this study investigated the neurochemistry of the ACC, which is a critical region in ED. In addition to investigating restrictive AN (AN-R) patients, this is the first report to include a well defined sample of BN patients. The study aimed to examine commonalities of neurochemical deviations in ED patients, as these patients share core psychopathological features such as DT, and it also focused on subgroup differences, which are to be expected as AN is characterized by malnourishment, whereas BN by definition is not.
Based on previous results we hypothesized Glx, mI and NAA concentrations to be reduced in the ACC, in particular in AN-R.
As a second aim, we carried out an exploratory analysis to test for associations between neurochemical substrates of the ACC and core eating disorder psychopathology, i.e. DT and bulimia (Pike and Mizushima, 2005, Hartmann et al., 2009 and Penas-Lledo et al., 2009). This latter part of the study was not hypothesis-driven, as it is the first of its kind.
