Brain Imaging and Behavior (2018) 12:1795–1803

https://doi.org/10.1007/s11682-018-9850-z

ORIGINAL RESEARCH

Difference in regional neural fluctuations and functional connectivity

in Crohn’s disease: a resting-state functional MRI study
Chunhui Bao1 · Peng Liu2 · Huirong Liu3 · Xiaoming Jin4 · Yin Shi3 · Luyi Wu1 · Xiaoqing Zeng5 · Jianye Zhang6 ·
Di Wang1 · Vince D. Calhoun7,8 · Jie Tian2 · Huangan Wu1

Published online: 20 February 2018

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Abstract
Patients with Crohn’s disease (CD) are shown to have abnormal changes in brain structures. This study aimed to further
investigate whether these patients have abnormal brain activities and network connectivity. Sixty patients with CD and 40
healthy controls (HCs) underwent resting-state functional magnetic resonance imaging (fMRI) scans. Amplitude of low-
frequency fluctuation (ALFF) and seed-based functional connectivity (FC) were used to assess differences in spontaneous
regional brain activity and functional connectivity. Compared to the HCs, patients with CD showed significantly higher
ALFF values in hippocampus and parahippocampus (HIPP/paraHIPP), anterior cingulate cortex, insula, superior frontal
cortex and precuneus. The ALFF values were significantly lower in secondary somatosensory cortex (S2), precentral gyrus,
and medial prefrontal cortex. Functional connectivities between left HIPP and left inferior temporal cortex, and right mid-
dle cingulate cortex, HIPP, and fusiform area were significantly lower. The functional connectivities between right HIPP
and right inferior orbitofrontal cortex and left HIPP were also significantly lower. Patients with CD showed higher or lower
spontaneous activity in multiple brain regions. Altered activities in these brain regions may collectively reflect abnormal
function and regulation of visceral pain and sensation, external environmental monitoring, and cognitive processing in these
patients. Lower functional connectivity of the hippocampus-limbic system was observed in these patients. These findings
may provide more information to elucidate the neurobiological mechanisms of the disease.

Keywords Crohn’s disease · Resting-state functional MRI · Amplitude of low-frequency fluctuation · Functional
connectivity

Chunhui Bao and Peng Liu contributed equally to this work.

4
* Huirong Liu Stark Neurosciences Research Institute, Indiana University
lhr_tcm@139.com School of Medicine, Indianapolis, IN 46202, USA
5
* Huangan Wu Department of Gastroenterology, Zhongshan Hospital, Fudan
wuhuangan@139.com University, Shanghai 200032, China
6
1 Department of Radiology, Shanghai Mental Health
Key Laboratory of Acupuncture and Immunological Effects,
Center, Shanghai Jiaotong University School of Medicine,
Shanghai University of Traditional Chinese Medicine,
Shanghai 200030, China
Shanghai 200030, China
7
2 The Mind Research Network, Albuquerque, NM 87131, USA
Life Sciences Research Center, School of Life Sciences
8
and Technology, Xidian University, Xi’an, Shaanxi 710071, Department of Electrical and Computer Engineering,
China University of New Mexico, Albuquerque, NM 87131, USA
3
Outpatient Department, Shanghai Research Institute
of Acupuncture and Meridian, Shanghai University
of Traditional Chinese Medicine, Shanghai 200030, China

13
Vol.:(0123456789)
1796 Brain Imaging and Behavior (2018) 12:1795–1803

Introduction activity at each voxel. Functional connectivity (FC), which

reflects synchronous spontaneous fluctuations between
Crohn’s disease (CD), a type of inflammatory bowel dis- brain regions, can provide information about functional
ease (IBD), is a chronic gastrointestinal inflammatory dis- integrity of brain networks (Fox and Raichle 2007). The
ease of unknown etiology. The major signs and symptoms combination of these two methods allows assessment of
include abdominal pain, chronic diarrhea, weight loss and the intrinsic activity of a specific brain region and its net-
fatigue. The disease typically has progressive course and work associated with a disease. Having been widely used
causes intestinal damage or even disability in severe cases in the study of functional gastrointestinal disorders (Ma
(Torres et al. 2016). Currently, there is no definitive treat- et al. 2015; Zhou et al. 2013), neurological disorders (Li
ment for CD. The condition typically requires lifelong et al. 2016; Van Hees et al. 2014), psychosis (Li et al.
medication and supportive care. Recent years have wit- 2014; Zhang et al. 2014), and other medical conditions,
nessed a steady increase in the incidence of IBD in Asia; this method shows promise in investigating functional
China has the highest incidence of IBD in Asia, which activity of the resting brain. To date, no study has been
places a heavy burden on health care resources and affects conducted on resting-state brain activity indexed by ALFF
social productivity (Ng et al. 2013). Because chronic pain, and FC in patients with CD. Therefore, these approaches
inflammation, and brain-gut interaction are known to may help elucidate the characteristics of resting-state brain
involve various brain networks (Vermeulen et al. 2014), activity in CD patients, and provide useful information on
characterizing specific changes in brain network and the pathophysiological mechanisms of CD.
activity in CD patients is important for understanding the In this study, we used ALFF and FC to explore resting-
mechanism of the disease and controlling the symptoms. state brain activity in patients with CD in remission. We
Functional abnormality in the central nervous system hypothesized that there would be differences in ALFF and
and the brain-gut axis was recently shown to play a key FC in certain brain regions between patients with CD in
role in the pathogenesis and development of CD (Al and remission and healthy controls, and that neuroimaging
Aziz 2014; Bonaz and Bernstein 2013). Studies have results would be associated with the duration of disease.
shown abnormal brain activity in CD patients in remis-
sion. Specifically, altered resting-state brain activity was
observed in brain regions involved in visceral sensation,
motility, emotional awareness, attention and cognition Methods
(Bao et al. 2016a, b). However, different brain activities
may be related to different symptoms and pathogenesis Subjects
of CD. A previous study showed that abnormal brain
activity may be related to abdominal pain sensation in This study was approved by the Ethics Committee of the
CD patients. Among patients with CD in remission, those Yueyang Hospital of Integrated Traditional Chinese and
with and without abdominal pain showed differences in Western Medicine affiliated with Shanghai University of
resting-state brain activity; further, activity profile in spe- Traditional Chinese Medicine. All subjects signed informed
cific brain areas were related to the severity of the disease consent forms.
(Bao et al. 2016a, b). Researchers have also investigated Sixty patients with CD were recruited from the outpa-
functional brain activity evoked by psychological stress tient clinics for inflammatory bowel diseases at the Shang-
in patients with CD in remission. They found altered neu- hai Research Institute of Acupuncture and Meridian, and
ral activity in hippocampus, amygdala, and insula in CD the Endoscopy Center of Zhongshan Hospital affiliated
patients (Agostini et al. 2013). However, functional brain with Fudan University. All subjects underwent systematic
activity in patients with CD is not well characterized. The and gastrointestinal screening, including colonoscopy and
key regions with abnormal brain activity in these patients intestinal mucosal biopsy. CD Endoscopic Index of Severity
have not been fully elucidated and findings from previous (CDEIS) of each subject was scored by an expert gastroen-
studies need to be further validated. terologist. To assess the level of inflammation, C-reactive
Resting-state functional magnetic resonance imaging protein, erythrocyte sedimentation rate and platelet counts
(rs-fMRI) has emerged as a useful technique to study for each subject were measured and recorded two weeks
the changes in functional activities in the resting state of prior to fMRI scans.
brain. Amplitude of low frequency fluctuations (ALFF), Inclusion criteria: right-handed (as determined by Edin-
which detect the amplitude of blood oxygen level depend- burgh handedness inventory) (Oldfield 1971); age: 18–50
ent (BOLD) signal relative to the baseline (Zang et al. years; minimum education level: 6 years of education;
2007), can be used to reflects the level of spontaneous remission lasting more than 6 months; CDAI score ≤ 150;
CDEIS score < 3.

13
Brain Imaging and Behavior (2018) 12:1795–1803 1797

Exclusion criteria: abnormal levels of inflammation 3.75 × 3.75 ­mm2, 32 sagittal slices and slice thickness: 5 mm
(C reactive protein > 10 mg/L; erythrocyte sedimentation with no gaps.
rate > 20 mm/h; platelets > 300 × 10 9/L); history of CD-
related abdominal surgery; history of treatment with gluco- Image data preprocessing
corticoids, anti TNF-α drugs, psychotropic or opioid drugs
in the past 3 months; pregnant or breastfeeding women; The imaging data processing assistant for resting-state fMRI
patients with current or previous history of neurological or (DPARSF, V3.1, http://www.restf​mri.net) was used to ana-
psychiatric conditions, head trauma or loss of conscious- lyze imaging data (Chao-Gan and Yu-Feng 2010), which is
ness; claustrophobia; and patients with metallic implants in based on SPM 12 (http://www.fil.ion.ucl.ac.uk/spm/softw​
the body. are/spm12​), and the resting state fMRI data analysis toolkit
Forty gender, age, and education matched healthy sub- (REST, V1.8, http://restf​mri.net/forum​/ REST_V1.8) (Song
jects (HCs) were recruited from the Shanghai University of et al. 2011). The first ten volumes of each functional time
Traditional Chinese Medicine via a newspaper advertise- series were discarded to allow for signals equilibrium and for
ment. They did not suffer from any gastrointestinal disorders participants’ adaptation to the scanning noise. The remain-
or pain and did not receive any medications. In addition, ing images were corrected for temporal difference in the
they had negative results on colonoscopy performed within acquisition of the different slices and then realigned to the
the preceding year as part of routine physical examination. first volume for head-motion correction. Subjects with head
All subjects were evaluated by an experienced gastroen- motion exceeded 2 mm translation or 2 degrees rotation in
terologist. To rule out psychiatric or neurological disorders, any direction were excluded. The corrected images were
psychiatric examinations were performed by an experienced further spatially normalized to the Montreal Neurological
psychiatrist, based on a structured psychiatric interview tool Institute (MNI) 152 template and resliced with isotropic
from the Diagnostic and Statistical Manual of Mental Dis- 3 mm × 3 mm × 3 mm voxel size and spatially smoothed
orders, 4th edition (DSM-IV). with a 6-mm full-width at half maximum (FWHM) Gauss-
ian kernel. After the linear trend of time courses removal,
the imaging data were then temporally band pass filtered
Symptom assessment
(0.01–0.1 Hz) to remove the effects of low-frequency drift
and high-frequency noise. The Friston 24 head-motion
The condition of CD patients was assessed by the Crohn’s
parameters, white matter signal, and cerebral spinal fluid
disease activity index (CDAI) (Best et al. 1979); the quality
signal were regressed out as nuisance covariates (Fair et al.
of life was assessed using the Inflammatory Bowel Disease
2008).
Questionnaire (IBDQ) (Irvine et al. 1994); emotional dis-
tress was assessed using the Hospital Anxiety and Depres-
Amplitude of low‑frequency fluctuation (ALFF)
sion Scale (HADS) (Zigmond and Snaith 1983).
ALFF maps were calculated using REST v1.8 software
MRI data acquisition (http://restf​mri.net/forum​/REST_V1.8) (Song et al. 2011).
First, the preprocessed time series was transformed to fre-
All MRI data were obtained from a 3T magnetic resonance quency domain using fast Fourier transform (FFT) to obtain
scanner (Siemens, TRIO, Erlangen, Germany) in the Depart- power spectrum. Then, the square root was calculated at
ment of Radiology at the Shanghai Mental Health Center. each frequency of the power spectrum and averaged across
All subjects were supine on the scanner and were told to 0.01–0.1 Hz at each voxel. This averaged square root was
relax, keep their eyes closed and not fall asleep or think. A taken as the ALFF. To reduce the global effects of variabil-
standard sponge pad was used to fill the gap between the ity across the subjects, the ALFF of each voxel was divided
head and the coil to prevent head movement and to protect by the global mean ALFF value to standardize data across
the ears from the noise of scanner. subjects (Zang et al. 2007).
A set of high-resolution 3D T1-weighted structural
images was obtained prior to functional san (TR/TE: Functional connectivity (FC)
2300 ms/2.98 ms; field of view (FOV): 256 × 256 m ­ m 2;
matrix size: 256 × 256; flip angle: 9°; in-plane resolution: CD is an autoimmune disease and hippocampus is thought
1 × 1 ­mm2; slice thickness: 1.0 mm with no gaps and 176 to play an important role in neuroimmune regulation (Lathe
slices). Functional images were acquired with a single-shot 2001). Previous studies (Agostini et al. 2013; Bao et al.
gradient–recalled echo planar imaging (EPI) sequence R/ 2015, 2016a , b) have shown differences in gray matter
TE: 2000 ms/30 ms; 180 time points; FOV: 240 × 240 ­mm2; structures and functional activity in the hippocampal cortex
matrix size: 64 × 64; flip angle: 90°; in-plane resolution: of patients with CD. This study also found abnormal ALFF

13
1798 Brain Imaging and Behavior (2018) 12:1795–1803

values in bilateral hippocampal cortex. Thus, the detected None of the subjects were smokers. 51.7% of the patients
bilateral hippocampus clusters from the ALFF analysis were presented with abdominal pain. Of these, 28 were treated
used as the regions of interest (ROIs). The MNI coordinates with mesalazine, 11 were treated with azathioprine, four
of the center of the 6-mm spherical ROI were determined by were treated with mesalazine plus azathioprine, and 17 did
the peak t-score detected at the hippocampus. For functional not receive any medication.
connectivity analysis, the mean time series was extracted
from the seed region and correlated with the time series Difference in ALFF values between the CD patients
of each voxel of the whole brain for each subject. The cor- and HCs
relation coefficients were transformed into z values using
Fisher’s r-to-z transformation to improve normality. An As showed in Table 2, compared to the HCs, the CD patients
entire brain z-score map was created for each subject (Wei showed higher ALFF values in bilateral hippocampus and
et al. 2016). In this study, the global mean timecourse was parahippocampus, left anterior cingulate cortex (ACC)
not regressed out in the model. and precuneus, and right insula and superior frontal cortex
(Fig. 1a). The CD patients also showed lower ALFF val-
Statistical analyses ues in left precentral gyrus, left medial prefrontal cortex
(MPFC), and right secondary somatosensory cortex (S2)
Clinical and demographic characteristic measures were ana- (Fig. 1b).
lyzed with Statistical Product and Service Solutions (SPSS) There was no significant correlation between the abnor-
16.0 software [SPSS Inc. Chicago, IL]. The data are pre- mal ALFF values and disease duration in the patients with
sented as mean ± standard deviation. CD (P > 0.05 for all).
In this study, we only focused on the group-difference-
related positive intrinsic connectivity, which was used to Difference in functional connectivity
identify abnormal FC connectivity within this specific brain of hippocampus between the CD patients and HCs
network. A two sample t-test was performed to assess the
differences of ALFF and FC between the two groups within As showed in Table 3, compared to HCs, the left hippocam-
the brain network, with gender, age, anxiety score of HADS pus of the CD patients had lower functional connectivity
(HADS-A) and the depression score of HADS (HADS-D) with the left inferior temporal cortex, and right MCC, hip-
as covariates. The inclusion of anxiety and depression as pocampus and fusiform (Fig. 2a). The right hippocampus
covariates was to exclude the effects of negative emotions of the CD patients had lower functional connectivity with
on the results and to focus on the impact of the disease itself the right inferior orbitofrontal cortex and left hippocampus
on brain activity. The statistical significance was set at voxel- (Fig. 2b).
wise P < 0.001 uncorrected with an extent threshold of clus-
ter-wise false discovery rate (FDR) (P < 0.05 at cluster lever,
cluster size > 45). ROI-wise analysis was used to investigate Discussion
the relationship between ALFF alteration and disease dura-
tion. Pearson’s correlation was calculated in each group at The results showed significant differences in ALFF values
a threshold of P < 0.05 with the Bonferroni correction. The in multiple brain regions in CD patients, including higher
ROIs included all nine brain regions showing different ALFF ALFF values in hippocampus and parahippocampus, ACC,
values in CD patients. The MNI coordinates of the center of precuneus, insula, and superior frontal cortex, and lower
the 6-mm spherical ROI were determined by the peak t-score ALFF values in precentral gyrus, MPFC, and S2. Functional
of all detected ROIs. connectivity in hippocampus-limbic system in CD patients
was also significantly lower than that in HCs. Together,
abnormal activities and connectivity in these brain regions
Results may suggest abnormal brain functions related to the regu-
lation of visceral sensation, pain processing, default mode
Clinical and demographic characteristics network (DMN), and neuro-immunity in these patients.

Clinical and demographic characteristics of study subjects ALFF differences between patients with CD and HCs
are summarized in Table 1. CD patients and HCs did not dif-
fer significantly with respect to demographic characteristics In this study, CD patients showed different spontaneous fluc-
including gender, age, height, and weight (P > 0.05). The tuations in brain regions including insula, ACC, MPFC, pre-
CD patients had significantly higher HADS-A and HADS-D central gyrus, S2 and hippocampal cortex, all of which are
scores than the HCs (both P < 0.01). important components of visceral sensory and pain networks

13
Brain Imaging and Behavior (2018) 12:1795–1803 1799

Table 1  Demographic and CD Patients Healthy controls P value

clinical characteristics of the (n = 60) (n = 40)
study population
Sex (male/female) 43/17 27/13 0.656
Age (years) 30.80 ± 7.62 30.68 ± 5.78 0.930
Height (cm) 169.80 ± 6.94 169.68 ± 7.03 0.930
Weight (kg) 56.48 ± 9.04 59.33 ± 6.07 0.85
HADS-A 5.85 ± 3.46 3.25 ± 1.98 0.000
HADS-D 4.57 ± 3.68 2.95 ± 1.83 0.012
Disease duration (months) 77.70 ± 50.40 — —
CDAI 76.61 ± 43.93 — —
IBDQ 173.53 ± 26.29 — —
PLT 220.93 ± 43.21 — —
ESR 10.93 ± 5.58 — —
CRP 4.48 ± 2.84 — —
CDEIS 1.14 ± 0.73 — —
Montreal classification
Age at diagnosis A1 6 — —
A2 53 — —
A3 1 — —
Location L1 14 — —
L2 12 — —
L3 34 — —
L4 0 — —
Behavior B1 20 — —
B2 4 — —
B3 17 — —
B1P 5 — —
B2P 1 — —
B3P 13 — —

Data presented as mean (± standard deviation)

CD Crohn’s disease, CDAI CD Activity Index, CDEIS CD Endoscopic Index of Severity, CRP C-reactive
protein, ESR eosinophil sedimentation rate, HADS The Hospital Anxiety and Depression Scale, HADS-A
The Anxiety score of HADS, HADS-D The Depression score of HADS, IBDQ In-flammatory Bowel Dis-
ease Questionnaire, PLT platelet levels

(Mayer et al. 2006; Van et al. 2007), and are involved in different neural network activities involved in processing
visceral sensory (pain) and movement regulation. Precuneus visceral sensory and pain in these patients. Indeed, more
and MPFC are important elements of DMN (Buckner et al. than a half of all CD patients present with abdominal pain/
2008; Mantini and Vanduffel 2013). Our previous findings abdominal distension due to long-term chronic inflammation
regarding brain areas involved in altered regional homogene- of the intestines, despite being in remission. In our previ-
ity (ReHo) in CD patients during resting-state is consistent ous study, we found different brain activities between CD
with these results (Bao et al. 2016a, b). patients with and without abdominal pain (Bao et al. 2016),
Although the inflammation is relatively stable in patients which suggests that chronic abdominal pain in patients in
with CD in remission, long-term chronic inflammation may remission may affect functional activities of brain.
have an impact on functional activities of brain owing to the In addition, the spontaneous fluctuations in DMN were
relapsing-remitting disease course. In CD patients, intesti- also different in CD patients. Precuneus and MPFC are
nal inflammation / pain signals are transmitted through the all component of DMN. Spontaneous fluctuations were
brain–gut axis to the brain, and may lead to alterations in lower in anterior DMN (MPFC), and higher in posterior
multiple functional brain networks (Bonaz and Bernstein DMN (precuneus), which suggests lower coupling between
2013). Differences in spontaneous fluctuations in insula, cin- the anterior and posterior DMN in patients with CD. The
gulate cortex, hippocampus, PFC and other brain areas asso- anterior DMN is mainly involved in monitoring one’s
ciated with visceral activity and pain regulation may suggest own state, predicting nociceptive stimulus, and emotional

13
1800 Brain Imaging and Behavior (2018) 12:1795–1803

Table 2  Brain regions with Regions Hem BA MNI t value Voxels

significant differences in ALFF
values between CD patients and X Y Z
HCs
CD > HC
Insula R 47 36 18 –9 4.76 60
ACC​ L 24 –6 21 24 3.73 111
HIPP/paraHIPP L 20 –36 –24 –6 3.92 50
HIPP/paraHIPP R 37 24 –30 –6 3.87 53
Precuneus L 19 –15 –69 39 3.96 87
Superior frontal cortex R 8 18 18 54 4.63 91
CD < HC
Precentral gyrus L 6 –48 –3 39 –3.83 52
MPFC L 10 –3 54 6 –3.89 132
S2 (Rolandic-Oper) R 48 48 –9 18 –3.99 63

The results employed age, gender, anxiety and depression as covariates. The statistical threshold was set at
P < 0.05 (false discovery rate corrected) at cluster level and the cluster size > 45
ACC​ anterior cingulate cortex, BA Brodmann area, CD Crohn’s disease, Hem hemisphere, MPFC medial
prefrontal cortex, HCs healthy controls, HIPP / paraHIPP hippocampal / parahippocampal cortex

Fig. 1  Brain regions exhibited

significant differences in ampli-
tude of low frequency fluctua-
tions (ALFF) values between
CD patients and healthy control
subjects using age, gender,
anxiety and depression as
covariates. The statistical sig-
nificance was set at voxel-wise
P < 0.001 uncorrected with an
extent threshold of cluster-wise
false discovery rate (FDR)
(P < 0.05 at cluster lever, cluster
size > 45). (a) Brain regions
exhibit significantly higher
ALFF values in CD patients;
(b) Brain regions exhibit
significantly lower ALFF values
in CD patients. ACC, anterior
cingulate cortex; CD, Crohn’s
disease; MPFC, medial prefron-
tal cortex; HC, healthy control;
HIPP / paraHIPP, hippocampal
/ parahippocampal cortex;
PCUN, precuneus; PREC,
precentral gyrus; S2, secondary
somatosensory cortex; SFC,
superior frontal cortex

13
Brain Imaging and Behavior (2018) 12:1795–1803 1801

Table 3  Brain regions showing Regions Hem BA MNI T Value Voxels

significantly lower functional
connectivity to bilateral X Y Z
hippocampus between CD
patients and HCs Left hippocampus
MCC R 23 9 –21 36 -3.83 67
Hippocampus R 36 30 –3 –30 –3.75 49
Fusiform R 20 30 –3 –39 –3.87 147
Inferior temporal cortex L 20 –42 3 –36 –4.29 104
Right hippocampus
OFC R 47 39 42 –18 –4.53 87
Hippocampus L 20 –33 –24 –15 –3.67 52

The results employed age, gender, anxiety and depression as covariates. The statistical threshold was set at
P < 0.05 (false discovery rate corrected) at cluster level and cluster size > 45
BA Brodmann area, CD Crohn’s disease, Hem hemisphere, HCs healthy controls, MCC middle cingulate
cortex, OFC orbitofrontal cortex

Fig. 2  Brain regions showed significant differences in the hippocam- ing significantly lower connectivity to the left hippocampus in CD
pus-related connectivity between CD patients and healthy control patients; (b) Brain regions showing significantly lower connectivity
subjects using age, gender, anxiety and depression as covariates. The to the right hippocampus in CD patients. CD, Crohn’s disease; HC,
statistical significance was set at voxel-wise P < 0.001 uncorrected healthy control; HIPP, hippocampus; ITC, inferior temporal cortex,
with an extent threshold of cluster-wise false discovery rate (FDR) MCC, middle cingulate cortex; OFC, orbitofrontal cortex
(P < 0.05 at cluster lever, cluster size > 45). (a) Brain regions show-

assessment or integration (Blakemore 2008; Wiech et al. Previously, we found lower gray matter volume in
2005), while the posterior DMN is mainly associated MPFC, but higher gray matter volume in precuneus in
with self-introspection and episodic memory processing patients with CD. We speculated that the different gray
(Cavanna and Trimble 2006; Margulies et al. 2009). These matter volume in CD patients may be the structural basis
results suggested that relevant DMN functions, monitoring of abnormal spontaneous activities.
of the external environment and spontaneous cognition
processing (Buckner et al. 2008; Mantini and Vanduffel
2013), were different in CD patients.

13
1802 Brain Imaging and Behavior (2018) 12:1795–1803

Differences in hippocampus‑related functional and the occurrence of disease. Second, immunosuppressive

connectivity between CD patients and HCs therapy may be a confounding factor that could not be con-
trolled in this study. This is probably a common limitation
Our study demonstrated different spontaneous functional for many fMRI studies (Mikdashi 2016).
activities in hippocampal cortex, which is consistent with In conclusion, this study found differences in spontaneous
findings of previous neuroimaging studies which showed fluctuations of activities of multiple brain regions in patients
abnormal gray matter structure and function of hippocampus with CD in remission. Since the combined functions of these
in these patients (Agostini et al. 2013; Bao et al. 2015, 2016a, brain regions involve visceral sensation, pain processing, and
b). The hippocampus plays an important role in neuroimmuno- DMN, the observed different brain activity in CD patients
logical regulation by affecting cellular and humoral immunity may reflect their abnormal functions in visceral sensation
through the hypothalamic–pituitary–adrenal (HPA) axis and and pain, external environment monitoring, and cognition
neurohumoral pathways (Lathe 2001). Further, the hippocam- processing. In addition, functional connectivity between hip-
pus can also interact with the vagus nerve to regulate the func- pocampus and visceral sensory and pain network was also
tion of immune cells in the intestinal wall through the release lower. These results may provide a better understanding of
of acetylcholine, and, thereby, be involved in neural regula- the pathophysiological mechanisms of CD.
tion of intestinal inflammation. Studies in chemically induced
models of CD intestinal inflammation have shown increased Acknowledgements Many thanks to Dr. Lili Ma at Zhongshan Hos-
pital of Fudan University for endoscopy examination and scoring.
CNS excitability and behavioral changes, which may be related This work was funded by the National Key Basic Research Program
to hippocampal microglia activation, increased expressions of of China (973 program), No. 2009CB522900, 2015CB554501; the
TNF-α, nitric oxide synthase (iNOS), and nitrite content (Hey- Program for Outstanding Medical Academic Leader, No. 80; the Pro-
darpour et al. 2016; Riazi et al. 2008), enhanced glutamatergic gram of Shanghai Academic Research Leader, No. 17XD1403400, the
National Natural Science Foundation of China, No. 81471738 and the
synaptic transmission and plasticity, and reduced neurogenesis National Institutes of Health grant (P20GM103472).
(Riazi et al. 2015; Zonis et al. 2015). Therefore, we speculate
that intestinal immune dysfunction / inflammation may affect Compliance with ethical standards
the structure and function of hippocampal cortex in patients
with CD. Conflict of interest The authors disclose no conflicts.
Whole-brain seed-based FC analysis using bilateral hip-
pocampus as seed regions showed lower functional connectiv- Research involving human participants All procedures performed in
this study were in accordance with the ethical standards of the institu-
ity between hippocampus and brain structures of the limbic tional and/or national research committee and with the 1964 Helsinki
system. These brain regions are may be mainly involved in the declaration and its later amendments or comparable ethical standards.
regulation of visceral functions (including visceral sensation
and pain). The result is consistent with the previous findings. Informed consent Informed consent was obtained from every partici-
pant included in the study.
The limbic system is considered to be a higher center for
regulation of visceral functions. In particular, visceral func-
tional activity is regulated by hippocampus and other brain
regions of the limbic system (Drevets et al. 2008; Jones et al.
2006). Previous results have confirmed different spontane- References
ous activities in networks related to visceral sensory and
Agostini, A., Benuzzi, F., Filippini, N., Bertani, A., Scarcelli, A.,
pain in CD patients. As a key brain region involved in neu- Farinelli, V., et al. (2013). New insights into the brain involve-
roimmune regulation, the hippocampus, together with other ment in patients with Crohn’s disease: a voxel-based morphometry
brain structures of the limbic systems, regulates visceral study. Neurogastroenterology & Motility, 25(2), 147-e82.
sensation and movement (Drevets et al. 2008; Jones et al. Al, O. Y., & Aziz, Q. (2014). The brain-gut axis in health and disease.
Advances in Experimental Medicine & Biology, 817, 135–153.
2006). The finding of lower functional connectivity between Bao, C. H., Liu, P., Liu, H. R., Jin, X. M., Calhoun, V. D., Wu, L. Y.,
hippocampus and other brain structures of the limbic system et al. (2016a). Different brain responses to electro-acupuncture
may suggest reduced ability of the limbic system in regulat- and moxibustion treatment in patients with Crohn’s disease. Sci-
ing visceral sensation and pain network. entific Reports, 18(6), 36636.
Bao, C. H., Liu, P., Liu, H. R., Wu, L. Y., Jin, X. M., Wang., S. Y., et al.
(2016b). Differences in regional homogeneity between patients
with Crohn’s disease with and without abdominal pain revealed
Limitations and future directions by resting-state functional magnetic resonance imaging. Pain,
157(5), 1037–1044.
Bao, C. H., Liu, P., Liu, H. R., Wu, L. Y., Shi, Y., Chen, W. F., et al.
This study has several limitations. First, this is a cross- (2015). Alterations in brain grey matter structures in patients with
sectional study, which makes it impossible to determine a Crohn’s disease and their correlation with psychological distress.
potential causal relationship between abnormal brain activity Journal of Crohn S & Colitis, 9(7), 532–540.

13
Brain Imaging and Behavior (2018) 12:1795–1803 1803

Best, W. R., Becktel, J. M., & Singleton, J. W. (1979). Rederived values Mayer, E. A., Naliboff, B. D., & Craig, A. D. B. (2006). Neuroimaging
of the eight coefficients of the Crohn’s Disease Activity Index of the brain-gut axis: from basic understanding to treatment of
(CDAI). Gastroenterology, 77(4 Pt 2), 843–846. functional GI disorders. Gastroenterology, 131(6), 1925–1942.
Blakemore, S. J. (2008). The social brain in adolescence. Nature Mikdashi, J. A. (2016). Altered functional neuronal activity in neu-
Reviews Neuroscience, 9(4), 267–277. ropsychiatric lupus: a systematic review of the fMRI investiga-
Bonaz, B. L., & Bernstein, C. N. (2013). Brain-gut interactions in tions. Seminars in Arthritis and Rheumatism, 45(4), 455–462.
inflammatory bowel disease. Gastroenterology, 144(1), 36–49. Ng, S. C., Tang, W., Ching, J. Y., Wong, M., Chow, C., Hui, M., A.
Buckner, R. L., Andrews-Hanna, J. R., & Schacter, D. L. (2008). The J. et al (2013). Incidence and phenotype of inflammatory bowel
brain’s default network: anatomy, function, and relevance to dis- disease based on results from the Asia-Pacific Crohn’s and Colitis
ease. Annals of the New York Academy of Sciences, 1124(1), 1–38. epidemiology study. Gastroenterology, 145(1), 158–165.e2.
Cavanna, A. E., & Trimble, M. R. (2006). The precuneus: a review Oldfield, R. C. (1971). The assessment and analysis of handedness: the
of its functional anatomy and behavioural correlates. Brain, Edinburgh inventory. Neuropsychologia, 9(1), 97–113.
129(Pt3), 564–583. Riazi, K., Galic, M. A., Kentner, A. C., Reid, A. Y., Sharkey, K. A.,
Chao-Gan, Y., & Yu-Feng, Z. (2010). DPARSF: a MATLAB toolbox & Pittman, Q. J. (2015). Microglia-dependent alteration of gluta-
for “pipeline” data analysis of resting-state fMRI. Frontiers in matergic synaptic transmission and plasticity in the hippocampus
Systems Neuroscience, 14(4), 13. during peripheral inflammation. Journal of Neuroscience the Offi-
Drevets, W. C., Price, J. L., & Furey, M. L. (2008). Brain structural cial Journal of the Society for Neuroscience, 35(12), 4942–4952.
and functional abnormalities in mood disorders: implications for Riazi, K., Galic, M. A., Kuzmiski, J. B., Ho, W., Sharkey, K. A., &
neurocircuitry models of depression. Brain Structure and Func- Pittman, Q. J. (2008). Microglial activation and TNFalpha produc-
tion, 213(1–2), 93–118. tion mediate altered CNS excitability following peripheral inflam-
Fair, D. A., Cohen, A. L., Dosenbach, N. U. F., Church, J. A., Miezin, mation. Proceedings of the National Academy of Sciences of the
F. M., Barch, D. M., et al. (2008). The maturing architecture of United States of America, 105(44), 17151–17156.
the brain’s default network. Proceedings of the National Academy Song, X. W., Dong, Z. Y., Long, X. Y., Li, S. F., Zuo, X. N., Zhu, C. Z.,
of Sciences of the United States of America, 105(10), 4028–4032. et al. (2011). REST: a toolkit for resting-state functional magnetic
Fox, M. D., & Raichle, M. E. (2007). Spontaneous fluctuations in brain resonance imaging data processing. Plos One, 6(9), e25031.
activity observed with functional magnetic resonance imaging. Torres, J., Mehandru, S., Colombel, J. F., & Peyrin-Biroulet, L. (2016).
Nature Reviews Neuroscience, 8(9), 700–711. Crohn’s disease. Lancet, 67(27), 822–824.
Heydarpour, P., Rahimian, R., Fakhfouri, G., Khoshkish, S., Fakhraei, Van, O. L., Coen, S. J., & Aziz, Q. (2007). Functional brain imaging
N., Salehi-Sadaghiani, M., et al. (2016). Behavioral despair asso- of gastrointestinal sensation in health and disease. World Journal
ciated with a mouse model of Crohn’s disease: Role of nitric oxide of Gastroenterology, 13(25), 3438–3445.
pathway. Progress in Neuro-Psychopharmacology & Biological Van Hees, S., Mcmahon, K., Angwin, A., De Zubicaray, G., Read, S.,
Psychiatry, 4(64), 131–141. & Copland, D. A. (2014). A functional MRI study of the relation-
Irvine, E. J., Feagan, B., Rochon, J., Archambault, A., Fedorak, R. ship between naming treatment outcomes and resting state func-
N., Groll, A., et al. (1994). Quality of life: a valid and reliable tional connectivity in post-stroke aphasia. Human Brain Mapping,
measure of therapeutic efficacy in the treatment of inflammatory 35(8), 3919–3931.
bowel disease. Canadian Crohn’s relapse prevention trial study Vermeulen, W., De Man, J. G., Pelckmans, P. A., & De Winter, B. Y.
group. Gastroenterology, 106(2), 287–296. (2014). Neuroanatomy of lower gastrointestinal pain disorders.
Jones, M. P., Dilley, J. B., Drossman, D., & Crowell, M. D. (2006). World Journal of Gastroenterology, 20(4), 1005–1020.
Brain-gut connections in functional GI disorders: anatomic and Wei, S. Y., Chao, H. T., Tu, C. H., et al. (2016). Changes in functional
physiologic relationships. Neurogastroenterology & Motility, connectivity of pain modulatory systems in women with primary
18(2), 91–103. dysmenorrhea. Pain, 157(1),92–102.
Lathe, R. (2001). Hormones and the hippocampus. Journal of Endo- Wiech, K., Seymour, B., Kalisch, R., Stephan, K. E., Koltzenburg,
crinology, 169(2), 205. M., Driver, J., et al. (2005). Modulation of pain processing in
Li, F., He, N., Li, Y., Chen, L., Huang, X., Lui, S., et al. (2014). Intrin- hyperalgesia by cognitive demand. Neuroimage, 27(1), 59–69.
sic brain abnormalities in attention deficit hyperactivity disorder: Zang, Y. F., He, Y., Zhu, C. Z., Cao, Q. J., Sui, M. Q., Liang, M., et al.
a resting-state functional MR imaging study. Radiology, 272(2), (2007). Altered baseline brain activity in children with ADHD
514–523. revealed by resting-state functional MRI. Brain & Development,
Li, F., Lui, S., Yao, L., Hu, J., Lv, P., Huang, X., et al. (2016). Longi- 29(2), 83–91.
tudinal changes in resting-state cerebral activity in patients with Zhang, X., Zhu, X., Wang, X., Zhu, X., Zhong, M., Yi, J., et al. (2014).
first-episode schizophrenia: a 1-year follow-up functional MR First-episode medication-naive major depressive disorder is asso-
imaging study. Radiology, 279(3), 867–875. ciated with altered resting brain function in the affective network.
Ma, X., Li, S., Tian, J., Jiang, G., Wen, H., Wang, T., et al. (2015). Plos One, 9(1), e85241.
Altered brain spontaneous activity and connectivity network in Zhou, G., Liu, P., Wang, J., Wen, H., Zhu, M., Zhao, R., et al. (2013).
irritable bowel syndrome patients: a resting-state fMRI study. Fractional amplitude of low-frequency fluctuation changes in
Clinical Neurophysiology, 126(6), 1190–1197. functional dyspepsia: A resting-state fMRI study. Magnetic Reso-
Mantini, D., & Vanduffel, W. (2013). Emerging roles of the brain’s nance Imaging, 31(6), 996–1000.
default network. Neuroscientist A Review Journal Bringing Neu- Zigmond, A. S., & Snaith, R. P. (1983). The hospital anxiety and
robiology Neurology & Psychiatry, 19(1), 76–87. depression scale. Acta Psychiatrica Scandinavica, 67(6), 361–370.
Margulies, D. S., Vincent, J. L., Kelly, C., Lohmann, G., Uddin, L. Q., Zonis, S., Pechnick, R. N., Ljubimov, V. A., Mahgerefteh, M., Waw-
Biswal, B. B., et al. (2009). Precuneus shares intrinsic functional rowsky, K., Michelsen, K. S., et al. (2015). Chronic intestinal
architecture in humans and monkeys. Proceedings of the National inflammation alters hippocampal neurogenesis. Journal of Neu-
Academy of Sciences of the United States of America, 106(47), roinflammation, 12, 65.
20069–20074.

13