SEARCH

SEARCH BY CITATION

Keywords:

  • brain neuroplastic changes;
  • cingulate cortex;
  • Crohn’s disease;
  • frontal lobe;
  • pain matrix;
  • voxel-based morphometry

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Funding
  8. Disclosure
  9. Author contributions
  10. References

Background  Crohn’s disease (CD) is a chronic intestinal disorder characterized by overproduction of inflammatory cytokines and recurrent abdominal pain. Recently, brain morphological abnormalities in the pain matrix were found in patients with chronic pain disorders including irritable bowel syndrome. To investigate potential structural brain changes associated with CD, we used magnetic resonance imaging (MRI). Furthermore, we tested whether in patients gray matter (GM) volumes correlated with disease duration.

Methods  Eighteen CD patients in remission and 18 healthy controls underwent structural MRI. Voxel-based morphometry (VBM) is a fully automated technique allowing identification of regional differences in the amount of GM enabling an objective analysis of the whole brain between groups of subjects. VBM was used for comparisons and correlation analysis.

Key Results  With respect to controls, CD patients exhibited decreased GM volumes in portion of the frontal cortex and in the anterior midcingulate cortex. Disease duration was negatively correlated with GM volumes of several brain regions including neocortical and limbic areas.

Conclusions & Inferences  Crohn’s disease is associated with brain morphological changes in cortical and subcortical structures involved in nociception, emotional, and cognitive processes. Our findings provide new insight into the brain involvement in chronic inflammatory bowel disorders.


Abbreviations:
ACC

anterior cingulate cortex

aMCC

anterior midcingulate cortex

BA

Brodmann area

BBB

blood–brain barrier

BGA

brain gut axis

BMI

body mass index

CC

cingulate cortex

CD

Crohn’s disease

CDAI

Crohn’s disease activity index

CRP

C reactive protein

DLPFC

dorso-lateral prefrontal cortex

dPCC

dorsal posterior cingulate cortex

FWE

family wise error

GM

gray matter

IBD

inflammatory bowel disease

IBS

irritable bowel syndrome

MOFC

medio-orbital frontal cortex

pACC

pregenual anterior cingulate cortex

PCC

posterior cingulate cortex

pMCC

posterior midcingulate cortex

RCS

retro splenial cortex

ROI

region of interest

SVC

small volume correction

sACC

subgenual anterior cingulate cortex

TE

echo time

TR

repetition time

VBM

voxel-based morphometry

vPCC

ventral posterior cingulate cortex

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Funding
  8. Disclosure
  9. Author contributions
  10. References

Crohn’s disease (CD), a type of inflammatory bowel disease (IBD), is a chronic, relapsing, and remitting disorder of the bowel. The main symptoms of CD include bloody diarrhea, abdominal pain, fever, and weight loss. As for many chronic disorders, IBD is accompanied with stress,1,2 emotional disturbances,3–5 and impaired cognitive functioning.6 CD patients report significant pain symptoms during acute flares, and a substantial portion of patients with quiescent CD continue to complain about pain symptoms.7

Morphological abnormalities were found in the pain matrix, the distributed neural circuit involved in nociception, in chronic pain syndromes including fibromyalgia,8 low back pain,9 migraine,10 and irritable bowel syndrome (IBS),11–13 a functional bowel disorder characterized by abnormal stool frequency and abdominal pain.14,15 These alterations involved different brain regions for each pain syndrome (i.e. thalamus, insula, and brain stem) but overlapped in the frontal lobe and in the cingulate cortex (CC), a limbic structure belonging to the medial pain system, a fundamental part of the pain matrix.14 Although recurrent abdominal pain is a cardinal symptom of CD, the potential morphological changes occurring in the pain matrix in CD patients have not been investigated.

In CD patients, the chronic exposure of the brain to inflammatory mediators might promote neuroplastic events involving gray matter (GM) loss. Neurons within discrete brain regions including the CC, the amygdala, and hippocampus express cytokine receptors.16,17 Moreover, endothelial cells forming the blood–brain barrier (BBB), perivascular macrophages, and glial cells are capable of both constitutive and inducible cytokine secretion.17 Indeed, research in animal models observed increased cytokine levels within the brain after peripheral inflammatory challenges, suggesting a local synthesis of cytokines as part of a systemic immune response.16,18 The immune-to-brain communication, connecting peripheral inflammation and the brain, exists via humoral and neural pathways.17 Circulating cytokines can cross the BBB and propagate inflammatory signals within the brain through the activation of the endothelial cells, perivascular macrophages, and glial cells.19 Furthermore, intestinal inflammatory signals may reach the brain via the brain gut axis (BGA), the anatomo-functional substrate for bidirectional communications between the brain and the gastrointestinal tract.20

We hypothesized that, compared with healthy controls, CD patients may exhibit differences in GM volume. In light of the possibility that chronic pain and inflammatory mediators have elicited GM loss within the brains of patients, we expected to find decreased GM volumes in the pain matrix and in the brain structures associated with the BGA. Voxel-based morphometry (VBM) is a fully automated technique allowing identification of regional differences in the amount of GM enabling an objective analysis of the whole brain between groups of subjects.21 We conducted an exploratory VBM study across 18 CD patients in remission and 18 healthy controls to compare the brain morphology between the two groups. In addition, a region of interest (ROI) analysis was performed on the anterior cingulate. Finally, to evaluate the possibility that the chronic bowel disorder was associated with progressive GM changes, we correlated the GM volume with the disease duration. Further knowledge of brain morphological modifications in CD may provide new insights into the potential relationships between GM changes, cognitive impairment, emotional disturbances, and pain symptoms in affected patients.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Funding
  8. Disclosure
  9. Author contributions
  10. References

Participant’s recruitment

This study was carried out with the approval of the local Ethics Committee and all participants signed written informed consent. Physicians attending two collaborating IBD Units (S. Orsola-Malpighi university hospital in Bologna and Policlinico university hospital in Modena) examined all participants and enrolled CD patients. In addition, all participants were screened for neurological and psychiatric disorders using a structured interview for the diagnostic and statistical manual of mental disorder IV edition (DSM IV). Handedness was assessed using the Edinburgh handedness inventory.22

CD patients  Eligible CD patients were consecutively recruited when they met the inclusion and exclusion criteria. Hematological and endoscopic data were obtained from routinely conducted blood tests and ileo-colonscopies. We excluded patients with recent flares to avoid potential confounders. Other studies indicate that the medications used during inflammatory exacerbations such as steroids 23 and opioids 24 may induce acute GM changes. Moreover, given the intrinsic heterogeneity of disease characteristics in CD, we attempted to recruit patients homogeneous at least for disease activity.

Inclusion criteria were as follows: (1) range of age between 18 and 50, (2) clinical remission for at least 12 months, (3) CD diagnosed at least 3 years prior to enrollment, (4) right-handed.

Exclusion criteria were as follows: (1) clinical [Crohn’s disease activity index (CDAI) 25 >150], (2) endoscopic and (3) biological [C reactive protein (CRP) level in mg L−1 > 5] signs of disease activity, (4) use of corticosteroids, biologics, and psychotropic medications in the previous 6 months, (5) perianal disease, (6) presence of extraintestinal manifestations in the past 12 months, (7) current or prior history of neurological or psychiatric disease and pain syndromes, (8) claustrophobia, (9) presence of metallic implants in the body.

At recruitment, were collected as follows: (1) date of birth, (2) gender, (3) body mass index (BMI), (4) educational level. Moreover, according to the Montreal classification,26 were collected (5) age at diagnosis (A), (6) disease location (L), (7) disease behavior (B). Furthermore, we collected: (8) information about current and past pain symptoms. A questionnaire with the following items was used. From CD onset (i) did you suffer from recurrent abdominal pain for at least 6 months? (ii) Was the pain influenced by stenosis or fistula? (iii) Did you suffer from non-abdominal pain, for example arthralgia? In the past 6 months, (iv) did you suffer from abdominal pain? If yes, (v) how many times? At last twice, two to ten times or more than ten times? (vi) Was it mild, moderate, or severe? (vii) Was it characterized by discomfort, bloating, or cramps? Moreover, we collected: (9) perianal disease, (10) extraintestinal manifestations, (11) smoking history, (12) previous history of intestinal surgery, (13) clinical (CDAI) and (14) biological (CRP) parameters of disease activity, (15) past treatment with biologics, (16) possible maintenance treatments.

Control group  A group of 18 healthy subjects (10 females), were recruited with advertisements among staff from Bologna and Modena Universities. They underwent the same screening procedures as the CD patients and took part in the study as members of the control group.

Nomenclature

For the nomenclature of the CC, we used the four-region model proposed by Vogt et al.27–29 According to this model, the CC was subdivided in anterior cingulate cortex (ACC), midcingulate (MCC), posterior cingulate (PCC), and retrosplenial cortex (RCS). The ACC was further subdivided into subgenual (sACC) and pregenual (pACC) subregions. The MCC was further subdivided into anterior midcingulate cortex (aMCC) and posterior midcingulate cortex (pMCC) subregions. The PCC was further subdivided into the dorsal (dPCC) and ventral (vPCC) subregions. Brodmann’s area (BA) 24 was subdivided into areas 24 ACC, a24′ (aMCC), p24′ (pMCC), whereas BA 23 was subdivided into areas 23d and d23 (dPCC) and v23 (vPCC).

Neuroimaging protocol

Scanning was performed at the Department of Neuroscience-Baggiovara hospital of Modena university. T1-weighted images were acquired using a 3 Tesla Philips Intera system. A SPGR pulse sequence [echo time (TE) = 4.6, repetition time (TR) = 9.9 ms] was used. One hundred seventy contiguous slices were acquired (voxel size = 1 × 1 × 1 mm). VBM was used to identify GM abnormalities in cortical and subcortical structures as well as the correlation with disease duration. The Matlab 7.1 and SPM5 (Wellcome Trust Centre for Neuroimaging, London, UK) software were used for data analysis.

VBM procedure  Preprocessing: T1 volumes were preprocessed and segmented using vbm5 toolbox (http://dbm.neuro.uni-jena.de/vbm/vbm5-for-spm5/) as implemented in SPM5. This is an improved version of the standard VBM procedure described in detail elsewhere 30 and already used in previous studies.30–35 In brief, both the optimized method for VBM and VBM5 comprised several processes: normalization, segmentation, modulation, and smoothing. The normalization step put the individual images in a common space: images were spatially normalized to a widely used T1-weighted MRI template in stereotaxic space, the Montreal Neurological Institute/International Consortium for Brain Mapping (MNI/ICBM) 152 standard using linear transformations. Then, images were segmented in GM, white matter (WM), and cerebrospinal fluid (CSF) volumes. With respect to this preprocessing step, VBM5 differs from the standard VBM procedure. Indeed, segmentation is performed without tissue priors, thus increasing classification accuracy. Volumes were then modulated with Jacobian determinants. Modulation involves scaling by the amount of contraction, so that the total amount of GM in the modulated GM volumes remains the same, as it would be in the original images.

Whole brain analysis: Total GM volumes of normalized-modulated images were compared between groups using a two-sample T-test. In addition, a linear correlation analysis was used for correlating GM volume and disease duration. To remove the confounding effects of different brain sizes, we used modulated images correcting for non-linear warping only. In addition, we modeled the total amount of GM, WM, and CSF, as nuisance parameters (additive effects) in both analyses. Modulation is recommended if interested in volume changes rather than differences in concentration (or density). Finally, the GM volumes were smoothed with a Gaussian kernel of 12 × 12 × 12 mm width.

For the T-test and the correlation analysis, the threshold at the voxel level was set at P < 0.001 uncorrected for multiple comparisons. In addition, a cluster threshold of P < 0.05 uncorrected was used. This threshold corresponded to k > 280 voxels for the T-test and to k > 175 voxels for the correlation.

ROI analysis: We performed a ROI analysis on the CC. To avoid any drawing bias, we used a standard anatomical ROI (http://marsbar.sourceforge.net/index.html).36,37 Therefore, instead of drawing individual ROI on the single subject T1 images, the standard anatomical ROI has been applied on the individual normalized GM volume. Then, a T-test has been conducted on the resulting images to evaluate differences between CD patients and controls. At the voxel level, the threshold was set at P < 0.001 uncorrected. At the cluster level, correction was made with the small volume correction (SVC). The threshold was set at P < 0.005 with family wise error (FWE) correction.35

Statistical analysis for non-imaging variables

Statistics were carried out using SPSS software (Chicago, IL, USA). T-test was used for comparisons between the two groups and Chi-squared test (χ2) was used for categorical variables.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Funding
  8. Disclosure
  9. Author contributions
  10. References

Participants

All participants were right-handed. None of the participants were smokers at recruitment. No differences were observed between controls (N = 18) and patients (N = 18) for age, gender, educational level, and BMI (Table 1).

Table 1.   Sociodemographic and clinical characteristics of the two study groups. Values denote mean (± Standard Deviation) or numbers of subjects. NA, not applicable
 CD PatientsControls P
N = 18 N = 18
Sociodemographics
 Age (years)32.8 (±7.22)29.3 (±5.63)0.12
 Sex (Male/Female)8/108/101
 Education (years)14.1 (±3.13)15.8 (±2.41)0.1
Clinical characteristics
 BMI22.50 (±3.69)22.53 (±2.54)0.98
 Duration of illness (years)10.33 (±5.94)NA 
 Duration of remission (years)4.30 (±3.93)NA 
Montreal classification
 Age at diagnosisA12  
A216  
A30  
 LocationL110  
L20  
L38  
L40  
 BehaviorB118  
B20  
B30  
 Previous extraintestinal manifestationsArthralgia4  
Erithema Nodosum1  
None14  
 Previous surgeryIleal resection8  
Never treated10  
 Use of biologicsPreviously treated6  
Never treated12  

Patients’ characteristics

The CD group consisted of 18 patients (10 females) in remission. Endoscopic and clinical data were available in all patients. The CDAI was <150 in all patients whereas the mean level of CRP expressed in mg L−1 was 2.37 (1.05) ranging from 0.5 to 3.8. CD duration varied from 3 to 28 years with an average duration of 10.33 (5.94) years, whereas the duration of remission varied from 1 to 14 years with an average duration of 4.30 (3.93) years.

All patients suffered from abdominal pain for at least 6 months. For eight patients, pain was influenced by stenosis and fistulas that were subsequently treated with surgery. Among these eight patients, two also had arthralgia. Two other patients (classified B1) suffered from joint pain and one had erythema nodosum. In the 6 months prior to enrollment (during remission), 10 patients had symptoms of mild abdominal pain at least two times characterized by symptoms such as discomfort, bloating, and cramps. Finally, as revealed by the CDAI, one patient suffered from mild abdominal pain in the week prior to recruitment.

According to the Montreal classification, two patients fell into the A1 category and 16 patients into A2. The mean age at diagnosis was 22.44 (5.47) years and ranged from 13 to 33 years. In terms of disease location, 10 patients reported localization ileum (L1), whereas eight patients ileo-colon (L3). At recruitment, the disease behavior was classified not-stricturing and not-penetrating (B1) in all patients. No patients suffered in the past 12 months from extraintestinal manifestations. Moreover, none of patients had surgery in the past 12 months, but eight patients previously underwent an ileal resection.

Six patients were previously treated with biologics (infliximab or adalimumab). Finally, 14 patients had maintenance treatment with 5-aminosalicylic acid (ASA) agents, whereas four patients had no maintenance treatments (Table 1).

VBM

Whole brain analysis  We first tested the GM differences between CD patients and controls using the total intracranial volume as covariate of non-interest (Table 2 and Fig. 1). The CD group showed decreased GM volume bilaterally in the superior and medial frontal gyrus (BA 9) and in the middle and superior frontal gyrus (BA 6, 8 on the left side and BA 10 on the right hemisphere). No significant increased GM volume in CD patients relative to controls was found.

Table 2.   Gray matter differences between CD patients and controls
 ClusterVoxelLevelMNI coord.
kZ P x,y,z (mm)
  1. P < 0.001 uncorrected (voxel level), cluster corrected with SVC P < 0.005 (FWE).

  2. L, left; R, right; BA, Brodmann area.

Whole brain analysis
 R, L superior and medial frontal gyrus (BA 9)2843.33<0.00144535
 3.15<0.001−44433
 L middle and superior frontal gyrus (BA 6, 8)3323.79<0.001−222558
 R middle and superior frontal gyrus (BA 10)3693.59<0.001275312
P < 0.001 uncorrected for multiple comparisons (voxel level).
P < 0.05 uncorrected (cluster level) k > 280 voxels.
ROI analysis
 R anterior midcingulate (area a24′, BA 24)       98    3.22  <0.001   4 2820
image

Figure 1.  Surface rendering (A) and sagittal sections (B) showing the gray matter reduction in CD patients with respect to controls. Numbers below each sagittal slice represent the x coordinate in MNI space. Clusters are superimposed on the MNI template implemented in SPM5. a = R Middle and Superiore Frontal Gyrus (BA 10), MNI coordinates 27, 53, 12; b = R Anterior Midcingulate (area a24′, BA 24), MNI coordinates: 4, 28, 20; c = R Superior and Medial Frontal Gyrus (BA 9), MNI coordinates 4, 45, 35; d = L Middle and Superior Frontal Gyrus (BA 6, 8), MNI coordinates −22, 25, 58. a, c, d: whole brain analysis; b: ROI analysis.

Download figure to PowerPoint

ROI analysis  The ROI analysis on the CC revealed a decreased GM volume in the aMCC (area a24′, BA 24).

Correlation analysis  By means of a correlation analysis, we tested whether differences in the GM volume were associated with disease duration in CD patients. To this purpose, we correlated GM volume and years of disease duration using the total intracranial volume, and the age as covariates of non-interest. We found a wide network correlating negatively with disease duration in both hemispheres (Table 3 and Fig. 2). Clusters comprised the left precentral gyrus (BA 6), the right superior and middle frontal gyrus (BA 8, 10), the bilateral middle and superior frontal gyrus (BA 28, 6). Other clusters were found in the sACC (BA 25), in the pMCC (area p24′, BA 24) and in the PCC (areas d23 and v23, BA 23). Other clusters were found bilaterally in the parahippocampal gyrus (BA 28, 35, 36), as well as in the right superior temporal gyrus (BA 39), and in the left inferior temporal gyrus (BA 20, 36, 38).

Table 3.   Gray matter negative correlation with disease duration (years of disease)
 ClusterVoxelLevelMNI coord.
KZ P x,y,z (mm)
  1. P < 0.001 uncorrected for multiple comparisons (voxel level).

  2. P < 0.05 uncorrected (cluster level); k > 175 voxels.

  3. L, left; R, right; BA, Brodmann area.

Whole brain analysis
 L precentral gyrus (BA 6)4834.14<0.001−47−825
 R superior frontal gyrus (BA 8)2693.52<0.001173249
 3.38<0.001122551
 R middle frontal gyrus2403.47<0.00133479
 3.23<0.001384316
 R middle and superior frontal gyrus3163.67<0.001231257
 L middle and superior frontal gyrus3853.64<0.001−24860
 3.13<0.001−301053
 L subgenual anterior cingulate (BA 25)1883.51<0.001−313−10
 R posterior Midcingulate2373.35<0.0019−541
 L posterior cingulate (areas d23, v23, BA 23)2143.47<0.001−5−4323
 R parahippocampal gyrus (BA 28, 35, 36)4174.13<0.00123−31−10
 L parahippocampal gyrus (BA 28, 35, 36)1093.28<0.001−22−8−23
 R superior temporal gyrus (BA 39)3013.67<0.00163−6829
 L inferior temporal gyrus, Uncus (BA 20, 36, 38)5783.50<0.001−251−42
 L inferior temporal gyrus (BA 20)3423.40<0.001−50−6−39
image

Figure 2.  Surface rendering (A) and sagittal sections (B) showing the gray matter negative correlation with disease duration (years of disease) in CD patients. Numbers below each sagittal slice represent the x coordinate in MNI space. Same overlay procedures as in Fig. 1. a = R Posterior Midcingulate (area p24′ BA 24), MNI coordinates 9, −5, 41; b = R Superior Frontal Gyrus (BA 8), MNI coordinates 12, 25, 51; c = L Posterior Cingulate (area d23 and v23, BA 23), MNI coordinates −5, −43, 23; d = L Subgenual Anterior Cingulate (BA 25), MNI coordinates −3, 13, −10; e = R Parahippocampal Gyrus (BA 28, 35, 36), MNI coordinates 23, −31, −10; f = R Middle and Superior Frontal Gyrus (BA 8), MNI coordinates 23, 12, 57.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Funding
  8. Disclosure
  9. Author contributions
  10. References

In this study, brain morphological differences were found between a group of CD patients in remission and healthy controls. Compared with controls, CD patients exhibited a decreased GM volume in a portion of the dorso-lateral prefrontal cortex (DLPFC) as well as in the aMCC. Furthermore, in CD patients the duration of the disease was negatively correlated with GM volumes of a network of brain structures including the PFC and the cingulate gyrus. The CC is consistently activated in imaging studies during acute pain,15,27 playing an important role in predicting and avoiding noxious stimuli.28,29 Overall, these brain areas are involved in cognitive and emotional functions and nociception.14,15,27

VBM detects changes in GM concentration voxel by voxel as well as changes in the classification of individual voxels, for instance from WM to GM30 and probably a combination of both. In general, a decrease in GM could be due to a decrease in cell size, neural or glial cell apoptosis, a decrease in spine density, dendritic atrophy, or even changes in blood flow or interstitial fluid.

Morphological alterations in the pain matrix have been identified in patients suffering from persistent pain disorders including fibromyalgia,8 low back pain,9 headache/migraine,10 vulvodynia,38 and IBS.11,12 Although the structural abnormalities found in these studies were different for each pain syndrome, the most common findings were a decrease in GM in the frontal and cingulate cortices.39 Our findings show a remarkable overlap between GM abnormalities in CD and in chronic pain syndromes, suggesting a common basis. The mechanisms underlying these neuroplastic changes have not been completely elucidated, but excito-toxicity and apoptosis related to increased cytokine release, have been implicated.40 In chronic pain syndromes, the excessive pain inputs are considered the consequence of abnormal nociceptive/antinociceptive functions on different levels of the neuraxis.39,41 According to this view, the decreased GM volumes in the pain matrix are believed the consequence of neural loss induced by excito-toxicity.39 Likewise, the decreased GM volumes which we have found in CD patients might reflect atrophy secondary to excito-toxicity. In the context of CD, the recurrent intestinal inflammatory exacerbations determine abundant pain inputs. Furthermore, occult inflammation and inflammatory-induced changes in the efferent neurons are believed to be involved in the development of pain symptoms in CD patients in remission. These symptoms (usually defined IBS-like symptoms) are common in CD.7 The persistency of occult inflammation and pain might reflect the persistency of the GM changes in the pain matrix that we have found in patients in remission.42 Further studies involving a larger sample of CD patients with IBS-like symptoms might shed more light on this issue.

The overproduction of inflammatory mediators in CD provides other plausible mechanisms underlying the brain structural changes. In CD patients, inflammatory cytokines may promote GM changes by inducing apoptosis in astrocytes and oligodendrocytes.43,44 In addition, inflammatory mediators are believed to induce decreased neurogenesis and to exert cytotoxic effects via increased glutamatergic activation and increased oxidative stress.43 Despite the fact that these mediators are relatively large polypeptides, virtually unable to cross the BBB, circulating cytokines could reach the brain through leaky regions in the BBB and active transport.17–19

Furthermore, circulating cytokines are believed to propagate their message within the brain via activation of endothelial cells, glial cells, and cerebral perivascular macrophages.17–19,44,45 Moreover, inflammatory signals arising from the gut may reach the brain via the BGA. Cytokines may relay their signals to the vagal nucleus of the solitary tract (NST) in the brainstem by binding the receptors of the peripheral afferent fibers of the vagus nerve. Thereafter, these signals may be projected to several cortical and subcortical areas including the CC and DLPFC.20,45

The brain morphological changes occurring in CD patients might have clinical implications. The GM changes in the pain matrix in CD might be involved in IBS-like symptoms, or hyperalgesia, occurring in patients. Furthermore, GM atrophy in frontal and limbic areas might represent the anatomical substrate for the development of cognitive 6 and emotional disturbances 2–5 that have already been observed in IBD patients. In our preliminary study, we did not evaluate in participants the emotional and the cognitive functioning. This represents a limit of the study that need to be addressed in further studies. Another limit regards the cross-sectional design of our study that does not allow us to infer causality direction. The negative correlation we found between GM volumes of several neural structures and the disease duration suggests that the morphological changes might be at least in part secondary to the chronic intestinal disorder. Nevertheless, we cannot exclude the possibility that the reduced GM volume we found in CD patients could be a preexisting condition, rather than a consequence of the bowel disorder. Longitudinal studies are needed to both validate and untangle the nature of the relationship. For correlation analysis, we adopted the duration of illness. This parameter does not take into account other important factors such as the number and severity of relapses, the kind and the duration of treatments.

In conclusion, our exploratory study has shown brain morphological abnormalities in CD patients in the pain matrix and in cognitive and emotional neural structures. Our findings represent a further step in understanding the brain involvement in IBD and could pave the way to further research aimed at investigating the plausible implications of neural changes in the development of disease manifestations such as hyperalgesia, emotional disturbances, and cognitive decline.

Funding

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Funding
  8. Disclosure
  9. Author contributions
  10. References

NF is funded by the Gordon Edward Small’s Charitable Trust (Scottish Charity Register: SC008962).

The work was independent of supporters.

Author contributions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Funding
  8. Disclosure
  9. Author contributions
  10. References

Study concept and design: Alessandro Agostini.

Acquisition and analysis of imaging data: Francesca Benuzzi.

Tecnical support for MRI: Valentina Farinelli.

Interpretation of imaging data: Francesca Benuzzi, Nicola Filippini.

Recruitment, clinical and endoscopic evaluations of participants: Angela Bertani, Antonella Scarcelli, Carlo Calabrese, Fernando Rizzello, Paolo Gionchetti.

Language assistance: Catherine Marchetta.

Study supervisor: Mauro Ercolani, Massimo Campieri, Paolo Nichelli.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Funding
  8. Disclosure
  9. Author contributions
  10. References