Psychiatry Research: Neuroimaging ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at SciVerse ScienceDirect

Psychiatry Research: Neuroimaging

0925-49http://d

n CorrUniversfax: +39

E-m1 Th

Pleasbase

journal homepage: www.elsevier.com/locate/psychresns

Brain volumetric abnormalities in patients with anorexiaand bulimia nervosa: A Voxel-based morphometry study

Federico Amianto a,1, Paola Caroppo b,n,1, Federico D’Agata b, Angela Spalatro a, Luca Lavagnino a,Marcella Caglio b, Dorico Righi c, Mauro Bergui b, Giovanni Abbate-Daga a, Roberto Rigardetto d,Paolo Mortara b, Secondo Fassino a

a Psychiatry Section, Department of Neurosciences, University of Turin, Italyb Neurology Section, Department of Neurosciences, University of Turin, Italyc Radiology Department, University of Turin, Italyd Neuropsychiatry Section, Department of Pediatrics and Adolescence Sciences, University of Turin, Italy

a r t i c l e i n f o

Article history:Received 21 June 2012Received in revised form18 February 2013Accepted 25 March 2013

Keywords:Anorexia nervosaBulimia nervosaVoxel-based morphometry (VBM)CerebellumCaudate nucleus

27/$ - see front matter & 2013 Elsevier Irelanx.doi.org/10.1016/j.pscychresns.2013.03.010

espondence to: Neurology Section, Departmeity of Turin, Via Cherasco 15, 10126 Turin116963487.ail address: paolacaroppo@libero.it (P. Caroppe first two authors contributed equally to the

e cite this article as: Amianto, F., etd morphometry study. Psychiatry Re

a b s t r a c t

Recent studies focussing on neuroimaging features of eating disorders have observed that anorexianervosa (AN) is characterized by significant grey matter (GM) atrophy in many brain regions, especiallyin the cerebellum and anterior cingulate cortex. To date, no studies have found GM atrophy in bulimianervosa (BN) or have directly compared patients with AN and BN. We used voxel-based morphometry(VBM) to characterize brain abnormalities in AN and BN patients, comparing them with each other andwith a control group, and correlating brain volume with clinical features. We recruited 17 AN, 13 BN and14 healthy controls. All subjects underwent high-resolution magnetic resonance imaging (MRI) with aT1-weighted 3D image. VBM analysis was carried out with the FSL-VBM 4.1 tool. We found no globalatrophy, but regional GM reduction in AN with respect to controls and BN in the cerebellum, fusiformarea, supplementary motor area, and occipital cortex, and in the caudate in BN compared to AN andcontrols. Both groups of patients had a volumetric increase bilaterally in somatosensory regions withrespect to controls, in areas that are typically involved in the sensory-motor integration of body stimuliand in mental representation of the body image. Our VBM study documented, for the first time in BNpatients, the presence of volumetric alterations and replicated previous findings in AN patients. Weevidenced morphological differences between AN and BN, demonstrating in the latter atrophy of thecaudate nucleus, a region involved in reward mechanisms and processes of self-regulation, perhapsinvolved in the genesis of the binge-eating behaviors of this disorder.

& 2013 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Anorexia nervosa (AN) and bulimia nervosa (BN), the twoprincipal DSM-IV eating disorders (ED) diagnoses, are complexpsychiatric disorders in which biological, psychological, socio-cultural and environmental factors interact and coexist in differentstages of development (Sadock et al., 2009).

Recently, scientific research has focused on the study of theneurobiological basis of ED, aiming to identify biomarkers for thediagnosis and classification of these disorders, which are charac-terized by a high discrepancy between diagnostic criteria andclinical experience (Brooks et al., 2012). For this purpose, many

d Ltd. All rights reserved.

nt of Neurosciences,, Italy. Tel.: +39 116336298;

o).article.

al., Brain volumetric abnormsearch: Neuroimaging (2013

studies have used both structural and functional neuroimagingtechniques to analyze the brain regions involved in the pathophy-siology of ED (Sadock et al., 2009).

Observations of global cerebral and cerebellar atrophy in greymatter (GM) and white matter as well as ventricular enlargementin AN have led some authors to interpret these findings asrepresenting a widespread cerebral vulnerability of this disorder(Artmann et al., 1985; Dolan et al., 1988; Krieg et al., 1988; Goldenet al., 1996; Katzman et al., 1996; Lambe et al., 1997; Addoloratoet al., 1998; Swayze et al., 2003; Chui et al., 2008).

Other studies, using a manual tracing approach on brain mag-netic resonance imaging (MRI), haveshownthat somecerebral areasare particularly involved in AN, such as the thalamus, midbrain(Husain et al., 1992), paracentral lobule (Inui et al., 2002), hippocam-pus–amygdala complex (Giordano et al., 2001; Connan et al., 2006)and anterior cingulate cortex (ACC) (McCormick et al., 2008).

Voxel-based morphometry (VBM) is an unbiased automatedtechnique based on high-resolution brain MRI sequences, and is

alities in patients with anorexia and bulimia nervosa: A Voxel-), http://dx.doi.org/10.1016/j.pscychresns.2013.03.010i

F. Amianto et al. / Psychiatry Research: Neuroimaging ∎ (∎∎∎∎) ∎∎∎–∎∎∎2

widely used to measure structural differences across groups ofsubjects (Ashburner and Friston, 2000). A recent VBM study on ANpatients found a significant reduction of total WM volume andfocal GM atrophy in the cerebellum, hypothalamus, caudatenucleus, and frontal, parietal and temporal areas (Boghi et al.,2010). In another study, the authors found atrophy in the middlecingulate cortex, precuneus, and inferior and superior parietallobules in restricting-type AN patients at an early stage of thedisease, supporting the hypothesis of a regionally specific vulner-ability in the areas that are involved in mental representation ofself and body imagery (Gaudio et al., 2011). In recovered ANpatients, a significant GM decrease in the ACC has been found,correlated with the lowest lifetime body mass index (BMI)(Muhlau et al., 2007). In another study, the authors found avolume reduction in the extrastriate body area in patients withAN which has been related to the body image distortion typical ofthis disorder (Suchan et al., 2010). In a recent work, Brooks and co-workers demonstrated a volumetric reduction of the insularcortex, parahippocampal and fusiform gyrus, cerebellum andposterior cingulate cortex in AN with respect to healthy controls.Moreover, the latter displayed an age-related volumetric decline ofthe dorsal-lateral prefrontal cortex that was absent in the patients,suggesting that restraint could spare the AN group from atrophy inthis region (Brooks et al., 2011).

VBM longitudinal studies in AN patients have suggested thereversibility of structural brain abnormalities in individuals with EDafter nutritional recovery (Wagner et al., 2006), in particular GMalterations (Castro-Fornieles et al., 2010). Conversely, other studieshave shown that GM volume restitutionwas incomplete in subjectswho had been previously severely affected by AN (Joos et al., 2011).

So far, only a few studies have focused their attention on brainchanges in BN, finding that BN patients had decreased corticalmass, but that the decrease was less pronounced than in AN (Krieget al., 1989).

A recent VBM study, comparing AN and BN patients withcontrols, showed in AN a GM decrease in the ACC, frontaloperculum, temporal-parietal regions and precuneus with respectto controls, while findings in BN did not show any difference fromthose in controls (Joos et al., 2010). Another study on AN and BN,which included long-term recovered patients, did not find differ-ences with respect to controls except for an increased insularvolume in BN (Wagner et al., 2006). A further study, whichcompared BN and binge-eating disorder patients, found that bothhad increased volume in the medial orbitofrontal cortex withrespect to controls and BN displayed an enlargement of the ventralstriatum, a brain structure involved in reward processes and self-regulation (Schäfer et al., 2010). However, the previous studies didnot make direct comparisons between AN and BN, while in themajority of the current studies the authors included patients withsymptoms of long duration.

The aim of the present study was to use VBM to characterizebrain abnormalities of AN and BN patients who had received theirfirst diagnosis of an ED, comparing the groups with each other andwith healthy controls, in order to investigate those brain abnorm-alities which could be specifically linked to each ED diagnosis andto correlate them with clinical and demographic features.

2. Methods

2.1. Participants

All patients were recruited from the ED Pilot Centre of the Department ofNeuroscience, San Giovanni Battista Hospital of Turin, between November 2010 andNovember 2011.

A total of 30 patients were enrolled in this study from outpatients of the EDPilot Centre presenting a first diagnosis of ED: the group included 17 (15

Please cite this article as: Amianto, F., et al., Brain volumetric abnormbased morphometry study. Psychiatry Research: Neuroimaging (2013

restricting-type and 2 binge-purging) AN and 13 (11 purging-type and 2 notpurging) BN. All patients were diagnosed by two psychiatrists specialised in ED(A.F. and G.AD.) using the Structured Clinical Interview for Diagnosis (SCID) forDSM-IV-TR. An axis II assessment was also conducted using the SCID-II for DSM-IV-TR.A first assessment was performed by administering the SCID. After reviewing thediagnostic information, the psychiatrist made a final diagnosis of ED subtype andproposed the patient's participation in the research project.

The inclusion criteria for patients were as follows: female sex; age 16–30; right-handedness; a body mass index (BMI) of 15–18 for AN patients and 19–25 for BNpatients; no past or present psychiatric or neurological diseases, no Axis IIdisorders, no past or present pharmacological medication, no drug or alcoholabuse; no personal or family history of diabetes; no past or present psychotherapy;symptoms of less than 2 years' duration.

All patients completed the Eating Disorders Inventory-2 (Garner, 1991) toassess eating problems and the Beck Depression Inventory (BDI) to exclude majordepression symptoms. As determined by the SCID, no patients reported a previousED diagnosis.

Fourteen healthy women were recruited as controls (CN) through localadvertisements. They were interviewed by A.F. and G.AD, using the SCID, to ruleout past or present mental disorders or ED. The general inclusion criteria forcontrols were the same as for patients; the BMI was the same as for BN.

All patients and controls gave their written informed consent to the study.With juveniles, the written informed consent of their parents was obtained.The study was approved by the Ethical Committee of the San Giovanni BattistaHospital, Turin, in accordance with the Declaration of Helsinki.

2.2. MRI acquisition

MRI was acquired with a scanner at 1.5 T (Achieva, Philips). T1-Weighted 3DTurbo Gradient-Echo sequences (matrix¼256�256; voxel size 1�1�1 mm3;number of slices: 190; TR: 7 ms; TE: 3 ms; TFE shots¼89) were obtained with fullbrain coverage and isotropic voxels. Acquisition time was about 5 min.

2.3. VBM analysis

VBM was performed using the FSL-VBM 4.1 tool, part of the FSL software(FMRIB's Software Library, the University of Oxford).

For the VBM analysis, the steps below were followed (Good et al., 2001a):

1)

al), h

Preparation of T1-weighted images in the correct format (compressed NIFTI)

2) Performance of the brain extraction using the BET-extraction FSL-tool on

T1 images

3) Creation of the study-specific GM template at 2�2�2 mm3 resolution in

standard MNI space

4) Non-linear registration of all the GM images on the template. The images were

then modulated and smoothed with an isotropic Gaussian kernel of 7 mmFWHM (Full Width Half Maximum).

5)

Use of FAST-segmentation and fslstats FSL-tools (www.fmrib.ox.ac.uk/fsl) toobtain total GM and total WM for each group.

Voxel-wise GLM analysis was carried out using permutation testing. FSL-Randomise 2.8 (www.fmrib.ox.ac.uk/fsl) with 5000 permutations was used, withthe Threshold-Free Cluster Enhancement (TFCE) option. Age and BDI were used asconfounding covariates. Global GM and WM were compared among groups,regional GM in ED vs. CN, AN and BN vs. CN and AN vs. BN. Finally, GM volumein each group was correlated with BMI, age, duration of disease and BDI.

Significant results (po0.005 uncorrected for multiple comparisons) withaclusterextent(Ke)460,werereported.Clusterssurvivingaftercorrectionformultiplecomparisons were also looked at, using a corrected cluster size level of po0.05.

To obtain the anatomical localization of significant cluster peaks, the MRIcronsoftware (www.mccauslandcenter.sc.edu/mricro/mricron) with AAL (AutomatedAnatomical Labelling) and Brodmann areas (BA) templates were used. The resultswere reported in the MNI coordinates system. In the figures, the result maps werereported in accordance with neurological convention (right is right). The SPSS 17™software package (SPSS Inc., Chicago, IL, USA; www.spss.com) was used to compareglobal GM andWM and to make the statistical analysis on clinical and demographicdata: we used ANOVA to compare means among groups, using Tukey's honestlysignificant difference (HSD) test for post hoc comparison when needed (thresholdfor significant results was po0.05).

3. Results

3.1. Characteristics of participants

The demographic and clinical characteristics of patients andcontrols are summarized in Table 1.

ities in patients with anorexia and bulimia nervosa: A Voxel-ttp://dx.doi.org/10.1016/j.pscychresns.2013.03.010i

F. Amianto et al. / Psychiatry Research: Neuroimaging ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 3

The BMI of AN patients was significantly lower with respect toBN and CN groups (po0.01). Age differed significantly betweenAN and CN (AN younger than CN; po0.01).

BDI scores did not significantly differ between AN and BN; EDpatients were more depressed than CN (po0.01). All EDI-2 scoresof patients and controls are detailed in supplementary Table S1.

3.2. VBM results

3.2.1. Whole brain GM and WM differencesComparing the three groups, no significant global GM or WM

atrophy was found (supplementary Table S2).

3.2.2. GM differences between ED patients and CNED patients compared with CN had greater GM volume in the

right paracentral lobule (voxels¼1082; MNI x, y, z: 8,–34,50) and areduced GM volume in the right thalamus and left parahippo-campal gyrus (Supplementary Figs. S1–S2, Supplementary TablesS3–S4 ).

3.2.3. GM differences between AN patients and CNCompared with CN, AN patients showed an increased GM

volume in the right paracentral lobule (voxels¼659; MNI x, y, z:12,38,52), also including the precuneus (Fig. 1A, SupplementaryFig. S3, Supplementary Table S5).

Compared with CN, AN had lower GM volume in a large clusterwith a peak in the left SMA (voxels¼1229 voxels; MNI x, y, z:6,–17,64), two clusters in the bilateral cerebellar crus I (leftvoxels¼657; right voxels¼441) and other clusters in the rightprecentral, right supramarginal, right occipital inferior, left fusi-form gyri, left hippocampus and right thalamus (Fig. 1B,Supplementary Fig. S4, Supplementary Table S6).

3.2.4. GM differences between BN patients and CNBN patients showed a greater GM volume, compared with CN,

in a cluster of 1361 voxels (MNI x, y, z: 16,–34,38) in theparacentral lobule also involving the precuneus, and two clustersin the left putamen and left insula (Fig. 2A, Supplementary Fig. S5,supplementary Table S7).

BN patients showed a significantly lower GM volume, com-pared with CN, bilaterally in the caudate nucleus with a peak inthe right caudate (voxels¼355; MNI x, y, z: 14, 30, 4) and the rightthalamus (Fig. 2B, Supplementary Fig. S6, SupplementaryTable S8).

3.2.5. GM differences between AN and BN patientsBN patients showed a lower GM volume in the left caudate

(voxels¼80; MNI x, y, z: −4, 16, −6) (Fig. 3A, Supplementary Fig.S7, Supplementary Table S9).

Comparing AN and BN patients, we found a significantlyreduced GM volume in AN in a large cerebellar cluster of 3409voxels with peaks in the cerebellar crus II, in the left and rightSMA, left frontal middle, frontal inferior opercular/insula, leftpostcentral, left supramarginal, left lingual, right fusiform, left

Table 1Demographic and clinical characteristics of patients and controls.

AN (N¼17) Mean7S.D. BN (N¼13) M

Age (y) 2074 2273Duration of illness (mos) 1378 1075BMI 1671 2272BDI score 1378 1578

BMI¼body mass index; BDI¼Beck Depression Inventory; y¼years; mosn Tukey’s HSD post-hoc analysis; p significant o0.05.

Please cite this article as: Amianto, F., et al., Brain volumetric abnormbased morphometry study. Psychiatry Research: Neuroimaging (2013

occipital and temporal cortex, and bilaterally in the thalamus(Fig. 3B, Supplementary Fig. S8, Supplementary Table S10).

The fusiform atrophy also survived after correction for multiplecomparisons (po0.05 TFCE cluster corrected for multiplecomparison).

3.2.6. Correlations with age, duration of disease, BMI and BDIIn AN patients, BMI correlates directly with the right frontal

orbital cortex, right temporal middle cortex and right fusiformgyrus (Fig. 4A, Supplementary Fig. S9, Supplementary Table S11).We found no correlation between BMI and GM in BN and CN. Agecorrelates inversely with GM in AN patients in a large cluster inthe ACC (voxels¼2281; MNI x, y, z: −2,32,26) which also survivedafter correction with multiple comparisons (po0.05 TFCE clustercorrected for multiple comparison, Fig. 4B; SupplementaryFig. S10, Supplementary Table S12). We found no significantcorrelation of age and GM in BN and CN, nor any significantcorrelation of months of disease, BDI score and GM in AN and BN.

The Illustrations of axial slices (in accordance with radiologicalconvention) of all comparisons between groups, correlations andtables with coordinates in MNI space of the significant clustersareas and BA, are presented in the supplementary materials.

4. Discussion

The aim of the present study was to characterize the morpho-logical alterations of the brain in patients with AN and BN throughVBM analysis, comparing the groups with each other and controls.To our knowledge, this is the first study to find structural brainabnormalities in BN patients. Indeed, in the only VBM studycomparing AN and BN in their acute phase of disease with healthycontrols (Joos et al., 2010), the authors found only in AN patientsa GM decrease in the frontal operculum, temporal and parietalregions and precuneus compared with controls, while they failedto observe any volumetric alterations in BN. The different result ofour study could be due to greater age and longer durationof disease of the patients in the previous report (the meanduration for bulimic patients was 7 years). In another study onlong-term recovered patients with AN and BN, the authors foundonly an increased insular volume in BN (Wagner et al., 2006).

4.1. Global GM and WM differences

In the present study, no significant global GM or WM reductionwas found when comparing AN, BN and CN. This result disagreeswith other studies finding a global GM reduction in AN which hasbeen associated with the duration of the disease and with a stateof severe and prolonged malnutrion (Swayze et al., 2003;Chui et al., 2008). These alterations were thought to be relatedto the possible reversibility of the nutritional state after bodyweight restoration (Swayze et al., 2003; Castro-Fornieles et al.,2010). Also with regard to the global WM reduction, our findings donot agree with other previous findings (Castro-Fornieles et al., 2010;

ean7S.D. CN (N¼14) Mean7S.D. p Post-hocn

2473 0.01 ANoCN– 0.3 –

2172 o0.01 ANoBN, CN574 o0.01 AN, BN4CN

¼months.

alities in patients with anorexia and bulimia nervosa: A Voxel-), http://dx.doi.org/10.1016/j.pscychresns.2013.03.010i

Fig. 1. A) In red, greater regional GM in anorexia compared to controls. B) In blue, lower regional GM in anorexia compared to controls (po0.005 TFCE-uncorrected;neurological convention: right is right). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 2. A) In red, greater regional GM in bulimia compared to controls. B) In blue, lower regional GM in bulimia compared to controls. (po0.005 TFCE-uncorrected;neurological convention: right is right). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 3. A) In red, greater regional GM in anorexia compared to bulimia. B) In blue, lower regional GM in anorexia compared to bulimia. (po0.005 TFCE-uncorrected;neurological convention: right is right). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

F. Amianto et al. / Psychiatry Research: Neuroimaging ∎ (∎∎∎∎) ∎∎∎–∎∎∎4

Joos et al., 2010). Boghi and coworkers found WM alterations butsuggested that in early stages of the disease these could be theconsequences of acute alterations of eating intake (Boghi et al., 2010).The relatively high BMI (415) and the early phase of disease ofour AN group may thus explain the difference between ourfindings and those of previous findings, supporting the hypothesisthat these global brain alterations are not implicated in thepathogenesis of the disease, but possibly implicated in mainte-nance and resistance mechanisms (Kaye, 2008).

Please cite this article as: Amianto, F., et al., Brain volumetric abnormbased morphometry study. Psychiatry Research: Neuroimaging (2013

4.2. GM regional differences

In contrast, both AN and BN patients, as well as the whole EDgroup, had greater regional GM volume with respect to CN, in theparacentral lobule and precuneus and in somatosensory regionsbilaterally. This result partially disagrees with the literature, wheresome authors found atrophy in these areas, specifically in theprecuneus (Gaudio et al., 2011), which is involved in manipulationof mental images and representation of the self (Cavanna and

alities in patients with anorexia and bulimia nervosa: A Voxel-), http://dx.doi.org/10.1016/j.pscychresns.2013.03.010i

Fig. 4. A) GM and BMI direct correlation in AN subjects (po0.005 TFCE-uncorrected; neurological convention: right is right). B) GM and age inverse correlation in ANpatients. (po0.05 TFCE-cluster corrected; neurological convention: right is right).

F. Amianto et al. / Psychiatry Research: Neuroimaging ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 5

Trimble, 2006). This can be read two ways: the higher volume ofbrain areas involved in sensory-motor functions may eitherrepresent the physiological consequence of physical hyperactivitytypical of eating disorders or it may relate to the efforts of thesepatients to maintain a consistent mental image of their body inspace. Together with this second hypothesis, the contrastingevidence of Gaudio and co-workers (Gaudio et al., 2011) may beexplained by an incremented representation of these brain regionsin the early stages of the eating disorder that finally evolves,perhaps through a neurotoxicity mechanism, into hypotrophy ofthe same regions in the later stages. Further follow-up researchusing VBM is needed to confirm such a possible evolution.

In AN, we found atrophy of the cerebellum, specifically of thecerebellar crus I and II, compared with both BN and CN. This resultconfirmed the previous findings of a volumetric reduction ofcerebellar GM in AN patients, more marked in those with a longerduration of disease (Boghi et al., 2010). The hypothesis of a role ofthe cerebellum in feeding behavior is corroborated by recentexperimental evidence showing its connections to the hypothalamusthrough cerebello-hypothalamic circuits which are thought to beinvolved in the regulation of food intake and feelings of satiety(Zhu and Wang, 2008). Moreover, in recent years, many studieshave focused their attention on the role of the cerebellum incognition and emotional experience (Schmahmann and Sherman,1998; Parvizi et al., 2001; Hopyan et al., 2010). Atrophy of thecerebellar regions crus I and II, specifically involved in cerebellarcognitive and emotional functions (Stoodley and Schmahmann,2009), supported the hypothesis that cerebellar atrophy in anor-exia could be involved in the cognitive and emotional impairmentof those patients, notably including cognitive rigidity with deficitsin executive functioning and working memory, and mood altera-tions, often present since adolescence in anorexia (Abbate-Dagaet al., 2011; Brooks et al., 2011; Friederich and Herzog, 2011). Inaccord with the literature, we found many other volumetric altera-tions in AN patients, predominantly in the SMA and fusiform areawith respect to CN and BN and in the hippocampus with respect toCN. As is known, the SMA is involved in the planning and control ofmovements and in task switching (Nachev et al., 2008), therefore, itis plausible that an impairment of this region could contribute tothe cognitive-behavioral inflexibility of patients (Friederich andHerzog, 2011). SMA atrophy has already been found in patientswith acute and restored AN, supporting the idea that it is indepen-dent of the stage of disease (Friederich et al., 2012).

Fusiform area atrophy, a result that remained even aftercorrection for multiple comparisons, is of particular interest andconfirmed previous findings (Brooks et al., 2011). Atrophy of thisregion, which is involved in body shape and food processing

Please cite this article as: Amianto, F., et al., Brain volumetric abnormbased morphometry study. Psychiatry Research: Neuroimaging (2013

(Downing et al., 2001), may underlie the impaired perception ofone's own body as well as the cognitive bias in food imagingprocessing that occurs in this disorder (Brooks et al., 2011;Hummel et al., 2012). We also found atrophy in other occipitalregions also covering part of the extrastriate body areas which, asalready suggested, could be related to body image distortion, asthey are involved in the processing of static body representationwhich in patients with AN could be impaired (Suchan et al., 2010).

In AN, we also found hippocampal atrophy that confirmedprevious findings (Connan et al., 2006; Brooks et al., 2011).Atrophy in this region, which is involved in memory and learning,is not specific to AN as it has been found in other psychiatricdisorders such as depression (Caetano et al., 2004), suggesting thisregion is highly vulnerable and could be affected by exposure to anelevated cortisol level in response to chronic stress (Conrad, 2008).

So far, VBM differences between BN and CN, have been foundonly by Wagner and coworkers (Wagner et al., 2006), who showedan increased insular volume in recovered BN patients in analysescorrected for age. In our sample, BN patients had an increasedinsular volume with respect to both CN and AN. Some authorshave already demonstrated in functional imaging studies that theactivity of the insula is impaired in subjects with ED, in particularfinding they had increased activity of the anterior insula inresponse to food images (Kim et al., 2012). Connectivity analyseshave also demonstrated that the insula is hyperconnected both inBN and AN and that the patterns of connectivity are differentbetween AN and BN, with BN women displaying insular hyper-connectivity with the medial orbital cortex which may explainsome aspects of their different eating behavior with respect to ANwomen (Kim et al., 2012).

Neuroimaging studies have already suggested that insularabnormalities could be involved in the pathophysiology of EDrelated to the insula's function emotional elaboration and integra-tion of body image and coping with uncertainty (Singer et al., 2009;Nunn et al., 2011). The insula is a region involved inprocessing taste,visceral memories, regulation of appetite and which, as it receivesmany interoceptive inputs from many different neural regions,could form an integrated representation of features of the indivi-dual's internal and external environment (Craig, 2011). The insula isan important station for the exchange of information flows, andalthough it is difficult to understand its specific contributions, it hasbeen associated with the regulation of cognition and emotion, thebalance of the sympathetic and parasympathetic systems, theprocessing of pain, anxiety and empathy, and self- and body-awareness: all functions which are altered in eating disorders.

A recently introduced concept of great interest is that ofpathodysmorphia (Brooks et al., 2012), which combines the phobia

alities in patients with anorexia and bulimia nervosa: A Voxel-), http://dx.doi.org/10.1016/j.pscychresns.2013.03.010i

F. Amianto et al. / Psychiatry Research: Neuroimaging ∎ (∎∎∎∎) ∎∎∎–∎∎∎6

and anxiety in relation to how people perceive and feel theirbodies. Insula dysfunction in ED patients could explain thisdisorder, which involves exaggerated emotional arousal and anxi-ety during the processing of awareness and interoception.

The most innovative finding of the present study consists in asignificantly reduced GM volume in the caudate nucleus in BNwomen with respect to CN and AN. To our knowledge, this is thefirst VBM demonstration of the involvement of the caudate in BN.Indeed, so far functional studies have demonstrated in BN patientsabnormal activation in the caudate during functional tasks relatedto food processing (Bohon and Stice, 2012).

The caudate nucleus is a core unit within the basal ganglia,involved in reward and motivation processes (Robbins and Everitt,1996; Berridge et al., 2010). Recent studies have shown thatpatients with BN are characterized by the alteration of self-regulation processes, linked to a deficit in the recruitment offronto-striatal circuits that contributes to the genesis of binge-eating behaviors (Marsh et al., 2009). Our findings on BN patientscould support the recently outlined concept of the “impulsecontrol model” of eating disorders, in which the alterations ofmesolimbic reward response mechanisms could explain the lack ofcontrol and the impulsivity that are often present in bulimicpatients (Brooks et al., 2012).

4.3. GM correlations

The BMI of AN patients was correlated with the orbital frontalcortex, frontal pole and fusiform area, while no correlations werefound in the other groups. This result supports the idea that someof the atrophic areas such as the fusiform gyrus together with thefrontal orbital regions are particularly vulnerable to restrictingbehavior and may thus be implicated in the maintenance mechan-ism of AN. In particular, due to the role of the frontal regions indecision making (Fellows, 2007), atrophy in these regions mayaccount for the severe cognitive impairment in the more mal-nourished subjects (Abbate Daga et al., 2011).

Only in AN did age have an inverse correlation with the ACC, anarea with a role in emotion processing, assessments of motiva-tional content, and assigning emotional valence to internal andexternal stimuli. It belongs to the circuits of attachment (Swainet al., 2007) and has been considered strongly related to the patho-genesis of ED (Castro-Fornieles et al., 2010; Friederich et al., 2012).As is known, this area is particularly vulnerable to aging, togetherwith the insula to which it is closely interconnected (Good et al.,2001b). In contrast with other authors (Friederich et al., 2012), wedid not find an atrophic ACC in AN with respect to CN. Instead, in oursample we found a specific vulnerability of this region to aging in ANpatients which we did not find in either BN or CN.

4.4. Limits of the study

Of course, this study has a number of limitations. The samplesize of BN patients is smaller than that of AN patients, and thestudy requires larger numbers for each diagnostic group toconfirm our findings. Moreover, age differed between AN andCN, but this limitation is mildly mitigated using age as a covariatein the model. The patients we recruited had received a first EDdiagnosis a relatively short time after the onset of symptoms, so itwas legitimate for us to suppose that the patients were notseverely affected by brain plasticity changes secondary to thedisease or by different pharmacological treatments. However, thiscould be less true if we consider that the onset of the disorderswas probably earlier and does not correspond to the onset ofclinical symptoms in ED.

Please cite this article as: Amianto, F., et al., Brain volumetric abnormbased morphometry study. Psychiatry Research: Neuroimaging (2013

4.5. Conclusions

Our article presents new data concerning volumetric altera-tions in BN patients and differences between AN and BN. Thesedata support the existence of two different patterns of braininvolvement for each ED subtype, but also the existence of a corefunctional basis of the ED spectrum which could itself be con-sidered a psychiatric condition due to its particular brainalterations.

Acknowledgements

The present study was sponsored by the “Compagnia di SanPaolo” Bank Foundation with the “Bando Neuroscienze” grant(code: 3929IT/PF 2008.2242) assigned in 2009 to Prof. SecondoFassino and Dr. Federico Amianto. The authors thank Dr. ChiaraCaroppo for her contribution in editing the manuscript.

Appendix A. Supporting information

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.pscychresns.2013.03.010.

References

Abbate-Daga, G., Buzzichelli, S., Amianto, F., Rocca, G., Marzola, E., Shawn, M,McClintock, S.M., Fassino, S., 2011. Cognitive flexibility in verbal and nonverbaldomains and decision making in anorexia nervosa patients: a pilot study. BMCPsychiatry 11, 162.

Addolorato, G., Taranto, C., Capristo, E., Gasbarrini, G., 1998. A case of markedcerebellar atrophy in a woman with anorexia nervosa and cerebral atrophy anda review of the literature. The International Journal of Eating Disorders 24,443–447.

Artmann, H., Grau, H., Adelmann, M., Schleiffer, R., 1985. Reversible andnon-reversible enlargement of cerebrospinal fluid spaces in anorexia nervosa.Neuroradiology 27, 304–312.

Ashburner, J., Friston, K.J., 2000. Voxel-based morphometry—the methods. Neuro-image 11, 805–821.

Berridge, K.C., Ho, C.Y., Richard, J.M., DiFeliceantonio, A.G.., 2010. The tempted braineats: pleasure and desire circuits in obesity and eating disorders. BrainResearch 1350, 43–64.

Boghi, A., Sterpone, S., Sales, S., D’Agata, F., Bradac, G.B., Zullo, G., Munno, D., 2010.In vivo evidence of global and focal brain alterations in anorexia nervosa.Psychiatry Research: Neuroimaging 192, 154–159.

Bohon, C., Stice, E., 2012. Negative affect and neural response to palatable foodintake in bulimia nervosa. Appetite 58 (3), 964–970.

Brooks, S.J., Barker, G.J., O’Daly, O.G., Brammer, M., Williams, S.C., Benedict, C.,Schioth, H.B., Treasure, J., Campbell, I.C., 2011. Restraint of appetite and reducedregional brain volumes in anorexia nervosa: a voxel-based morphometricstudy. BMC Psychiatry 11, 179.

Brooks, S.J., Rask-Andersen, M., Benedict, C., Schioth, H.B., 2012. A debate on currenteating disorder diagnoses in light of neurobiological findings: is it time for aspectrum model? BMC Psychiatry 12, 76.

Caetano, S.C., Hatch, J.P., Brambilla, P., Sassi, R.B., Nicoletti, M., Mallinger, A.G., Frank, E.,Kupfer, D.J., Keshavan, M.S., Soares, J.C., 2004. Anatomical MRI study of hippo-campus and amygdala in patients with current and remitted major depression.Psychiatry Research: Neuroimaging 132, 141–147.

Castro-Fornieles, J., Caldu, X., Andres-Perpina, S., Lazaro, L., Bargallo, N., Falcon, C.,Plana, M.T., Junque, C., 2010. A cross-sectional and follow-up functional MRIstudy with a working memory task in adolescent anorexia nervosa. Neuropsy-chologia 48, 4111–4116.

Cavanna, A.E., Trimble, M.R., 2006. The precuneus: a review of its functionalanatomy and behavioural correlates. Brain 129 (Pt 3), 564–583.

Chui, H.T., Christensen, B.K., Zipursky, R.B., Richards, B.A., Hanratty, M.K., Kabani, N.J.,Mikulis, D.J., Katzman, D.K., 2008. Cognitive function and brain structure infemales with a history of adolescent-onset anorexia nervosa. Pediatrics 122,426–437.

Connan, F., Murphy, F., Connor, S.E., Rich, P., Murphy, T., Bara-Carill, N., Landau, S.,Krljes, S., Ng, V., Williams, S., Morris, R.G., Campbell, I.C., Treasure, J., 2006.Hippocampal volume and cognitive function in anorexia nervosa. PsychiatryResearch: Neuroimaging 146, 117–125.

Conrad, C.D., 2008. Chronic stress-induced hippocampal vulnerability: the gluco-corticoid vulnerability hypothesis. Reviews in the Neurosciences 19, 395–411.

alities in patients with anorexia and bulimia nervosa: A Voxel-), http://dx.doi.org/10.1016/j.pscychresns.2013.03.010i

F. Amianto et al. / Psychiatry Research: Neuroimaging ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 7

Craig, A.D., 2011. Significance of the insula for the evolution of human awareness offeelings from the body. Annals of the New York Academy of Sciences 1225,72–82.

Dolan, R.J., Mitchell, J., Wakeling, A., 1988. Structural brain changes in patients withanorexia nervosa. Psychological Medicine 18, 349–353.

Downing, P.E., Jiang, Y., Shuman, M., Kanwisher, N.G., 2001. A cortical area selectivefor visual processing of the human body. Science 293, 2470–2473.

Fellows, L.K., 2007. The role of orbitofrontal cortex in decision making: acomponent process account. Annals of the New York Academy of Sciences1121, 421–430.

Friederich, H.C., Herzog, W., 2011. Cognitive-behavioral flexibility in anorexianervosa. Current Topics in Behavioral Neurosciences 6, 111–123.

Friederich, H.C., Walther, S., Bendszus, M, Biller, A., Thomann, P., Zeigermann, S.,Katus, T., Brunner, R., Zastrow, A., Herzog, W., 2012. Grey matter abnormalitieswithin cortico-limbic-striatal circuits in acute and weight-restored anorexianervosa patients. Neuroimage 59 (2), 1106–1113.

Garner, D.M., 1991. Eating Disorder Inventory-2 Professional Manual. PsychologicalAssessment Resources, Odessa, FL.

Gaudio, S., Nocchi, F., Franchin, T., Genovese, E., Cannata, V., Longo, D., Fariello, G.,2011. Gray matter decrease distribution in the early stages of Anorexia Nervosarestrictive type in adolescents. Psychiatry Research: Neuroimaging 191, 24–30.

Giordano, G.D., Renzetti, P., Parodi, R.C., Foppiani, L., Zandrino, F., Giordano, G.,Sardanelli, F., 2001. Volume measurement with magnetic resonance imaging ofhippocampus-amygdala formation in patients with anorexia nervosa. Journal ofEndocrinological Investigation 24, 510–514.

Golden, N.H., Ashtari, M., Kohn, M.R., Patel, M., Jacobson, M.S., Fletcher, A., Shenker, I.R.,1996. Reversibility of cerebral ventricular enlargement in anorexia nervosa,demonstrated by quantitative magnetic resonance imaging. Journal of Pediatrics128, 296–301.

Good, C.D., Ashburner, J., Frackowiak, R.S., 2001a. Computational neuroanatomy:new perspectives for neuroradiology. Revue Neurologique 157, 797–806.

Good, C.D., Johnsrude, I.S., Ashburner, J., Henson, R.N., Friston, K.J., Frackowiak, R.S.,2001b. A voxel-based morphometric study of ageing in 465 normal adulthuman brains. Neuroimage 1 (Pt 1), 21–36.

Hopyan, T., Laughlin, S., Dennis, M., 2010. Emotions and their cognitive control inchildren with cerebellar tumors. Journal of the International NeuropsychologicalSociety 16, 1027–1038.

Hummel, D., Rudolf, A.K., Brandi, M.L., Untch, K.H., Grabhorn, R., Hampel, H., Mohr, H.M.,2012. Neural adaptation to thin and fat bodies in the fusiform body area and middleoccipital gyrus: an fMRI adaptation study. Human Brain Mapping , http://dx.doi.org/10.1002/hbm.22135.

Husain, M.M., Black, K.J., Doraiswamy, P.M., Shah, S.A., Rockwell, W.J., Ellinwood Jr., E.H.,Krishnan, K.R., 1992. Subcortical brain anatomy in anorexia and bulimia. BiologicalPsychiatry 31, 735–738.

Inui, A., Asakawa, A., Kasuga, M., Kamikawa, S., Uemoto, M., Watanabe, T., 2002.Paracentral cortical atrophy in patients with eating disorders. American Journalof Medicine 112, 681–683.

Joos, A., Kloppel, S., Hartmann, A., Glauche, V., Tuscher, O., Perlov, E., Saum, B.,Freyer, T., Zeeck, A., Tebartz van Elst, L., 2010. Voxel-based morphometry ineating disorders: correlation of psychopathology with grey matter volume.Psychiatry Research: Neuroimaging 182, 146–151.

Joos, A., Hartmann, A., Glauche, V., Perlov, E., Unterbrink, T., Saum, B., Tuscher, O.,Tebartz van Elst, L., Zeeck, A., 2011. Grey matter deficit in long-term recoveredanorexia nervosa patients. European Eating Disorders Review 19, 59–63.

Katzman, D.K., Lambe, E.K., Mikulis, D.J., Ridgley, J.N., Goldbloom, D.S., Zipursky, R.B.,1996. Cerebral gray matter and white matter volume deficits in adolescent girlswith anorexia nervosa. Journal of Pediatrics 129, 794–803.

Kaye, W., 2008. Neurobiology of anorexia and bulimia nervosa. Physiology andBehavior 94, 121–135.

Please cite this article as: Amianto, F., et al., Brain volumetric abnormbased morphometry study. Psychiatry Research: Neuroimaging (2013

Kim, K.R., Ku, J., Lee, J.H., Lee, H., Jung, Y.C., 2012. Functional and effectiveconnectivity of anterior insula in anorexia nervosa and bulimia nervosa.Neuroscience Letters 521 (2), 152–157.

Krieg, J.C., Pirke, K.M., Lauer, C., Backmund, H., 1988. Endocrine, metabolic, andcranial computed tomographic findings in anorexia nervosa. BiologicalPsychiatry 23, 377–387.

Krieg, J.C., Lauer, C., Pirke, K.M., 1989. Structural brain abnormalities in patientswith bulimia nervosa. Psychiatry Research 27, 39–48.

Lambe, E.K., Katzman, D.K., Mikulis, D.J., Kennedy, S.H., Zipursky, R.B., 1997.Cerebral gray matter volume deficits after weight recovery from anorexianervosa. Archives of General Psychiatry 54, 537–542.

Marsh, R., Steinglass, J.E., Gerber, A.J., Graziano O’Leary, K., Wang, Z., Murphy, D.,Walsh, B.T., Peterson, B.S., 2009. Deficient activity in the neural systems thatmediate self-regulatory control in bulimia nervosa. Archives of GeneralPsychiatry 66, 51–63.

McCormick, L.M., Keel, P.K., Brumm, M.C., Bowers, W., Swayze, V., Andersen, A.,Andreasen, N., 2008. Implications of starvation-induced change in right dorsalanterior cingulate volume in anorexia nervosa. International Journal of EatingDisorders 41, 602–610.

Muhlau, M., Gaser, C., Ilg, R., Conrad, B., Leibl, C., Cebulla, M.H., Backmund, H.,Gerlinghoff, M., Lommer, P., Schnebel, A., Wohlschlager, A.M., Zimmer, C.,Nunnemann, S., 2007. Gray matter decrease of the anterior cingulate cortexin anorexia nervosa. American Journal of Psychiatry 164, 1850–1857.

Nachev, P., Kennard, C., Husain, M., 2008. Functional role of the supplementary andpre-supplementary motor areas. Nature Reviews: Neuroscience 9, 856–869.

Nunn, K., Frampton, I., Fuglset, T.S., Törzsök-Sonnevend, M., Lask, B., 2011. Anorexianervosa and the insula. Medical Hypotheses 76 (3), 353–357.

Parvizi, J., Anderson, S.W., Martin, C.O., Damasio, H., Damasio, A.R., 2001. Patholo-gical laughter and crying: a link to the cerebellum. Brain 124, 1708–1719.

Robbins, T.W., Everitt, B.J., 1996. Neurobehavioural mechanisms of reward andmotivation. Current Opinion in Neurobiology 6 (2), 228–236.

Sadock, B.J., Sadock, V.A., Ruiz, P., 2009. Kaplan and Sadock's ComprehensiveTextbook of Psychiatry. Lippincott Williams & Wilkins, Philadelphia.

Schäfer, A., Vaitl, D., Schienle, A., 2010. Regional grey matter volume abnormalitiesin bulimia nervosa and binge-eating disorder. Neuroimage 50, 639–643.

Schmahmann, J.D., Sherman, J.C., 1998. The cerebellar cognitive affective syndrome.Brain 121, 561–579.

Singer, T., Critchley, H.D., Preuschoff, K.., 2009. A common role of insula in feelings,empathy and uncertainty. Trends in Cognitive Science 13, 334–340.

Stoodley, C.J., Schmahmann, J.D., 2009. Functional topography in the humancerebellum: a meta-analysis of neuroimaging studies. Neuroimage (2),489–501.

Suchan, B., Busch, M., Schulte, D., Gronemeyer, D., Herpertz, S., Vocks, S., 2010.Reduction of gray matter density in the extrastriate body area in women withanorexia nervosa. Behavioural Brain Research 206, 63–67.

Swain, J.E., Lorberbaum, J.P., Kose, S., Strathearn, L., 2007. Brain basis of earlyparent-infant interactions: psychology, physiology, and in vivo functionalneuroimaging studies. Journal of Child Psychology and Psychiatry 48, 262–287.

Swayze 2nd, V.W., Andersen, A.E., Andreasen, N.C., Arndt, S., Sato, Y., Ziebell, S.,2003. Brain tissue volume segmentation in patients with anorexia nervosabefore and after weight normalization. International Journal of Eating Disorders33, 33–44.

Wagner, A., Greer, P., Bailer, U.F., Frank, G.K., Henry, S.E., Putnam, K., Meltzer, C.C.,Ziolko, S.K., Hoge, J., McConaha, C., Kaye, W.H., 2006. Normal brain tissuevolumes after long-term recovery in anorexia and bulimia nervosa. BiologicalPsychiatry 59, 291–293.

Zhu, J.N., Wang, J.J., 2008. The cerebellum in feeding control: possible function andmechanism. Cellular and Molecular Neurobiology 28, 469–478.

alities in patients with anorexia and bulimia nervosa: A Voxel-), http://dx.doi.org/10.1016/j.pscychresns.2013.03.010i