Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Oct 1.
Published in final edited form as: Neurogastroenterol Motil. 2022 May 13;34(10):e14396. doi: 10.1111/nmo.14396

Cine gastric MRI reveals altered Gut-Brain Axis in Functional Dyspepsia: gastric motility is linked with brainstem-cortical fMRI connectivity

Roberta Sclocco 1,2,†, Harrison Fisher 1,†, Rowan Staley 1,3, Kyungsun Han 1,4, April Mendez 3, Andrew Bolender 1,3, Jaume Coll-Font 5,6, Norman W Kettner 2, Christopher Nguyen 5,6,7, Braden Kuo 3,†, Vitaly Napadow 1,2,†
PMCID: PMC9529794  NIHMSID: NIHMS1808604  PMID: 35560690

Abstract

Background:

Functional dyspepsia (FD) is a disorder of gut-brain interaction, and its putative pathophysiology involves dysregulation of gastric motility and central processing of gastric afference. The vagus nerve modulates gastric peristalsis and carries afferent sensory information to brainstem nuclei, specifically the nucleus tractus solitarii (NTS). Here, we combine MRI assessment of gastric kinematics with measures of NTS functional connectivity to the brain in patients with FD and healthy controls (HC), in order to elucidate how gut-brain axis communication is associated with FD pathophysiology.

Methods:

FD and HC subjects experienced serial gastric MRI and brain fMRI following ingestion of a food-based contrast meal. Gastric function indices estimated from 4D cine MRI data were compared between FD and HC groups using repeated measure ANOVA models, controlling for ingested volume. Brain connectivity of the NTS was contrasted between groups and associated with gastric function indices.

Key Results:

Propagation velocity of antral peristalsis was significantly lower in FD compared to HC. The brain network defined by NTS connectivity loaded most strongly onto the Default Mode Network, and more strongly onto the Frontoparietal Network in FD. FD also demonstrated higher NTS connectivity to insula, anterior cingulate and prefrontal cortices, and pre-supplementary motor area. NTS connectivity was linked to propagation velocity in HC, but not FD, whereas peristalsis frequency was linked with NTS connectivity in patients with FD.

Conclusions & Inferences:

Our multi-modal MRI approach revealed lower peristaltic propagation velocity linked to altered brainstem-cortical functional connectivity in patients suffering from FD suggesting specific plasticity in gut-brain communication.

Keywords: functional dyspepsia, stomach, MRI, interoception, vagus nerve

Graphical Abstract

graphic file with name nihms-1808604-f0001.jpg

1. Introduction:

Functional dyspepsia (FD) is a disorder of gut-brain interaction (DGBI) characterized by a wide range of bothersome upper-gastric symptoms in the absence of any organic, systemic, or metabolic disease (Tack et al., 2006). The etiology of this gastrointestinal (GI) disorder is likely heterogeneous, but may include visceral hypersensitivity, altered gastric motility, Helicobacter pylori infection, autonomic or central nervous system dysfunction, psychosocial factors, and genetic susceptibility (Tack et al., 2004a) (Tack et al., 2004b). FD pathophysiology is not well understood, but may include altered gastric function, such as emptying (Kim et al., 2001)(Lorena et al., 2004) and accommodation (Lunding et al., 2006)(Sarnelli et al., 2003), and gastric myoelectric activity (Lin et al., 1999)(Leahy et al., 1999), as well as alterations in brain processing that may be related to gastric afference (Lee et al., 2016). It is also unknown if gastric and/or brain function may differentiate FD subtypes, i.e. epigastric pain syndrome (EPS) and postprandial distress syndrome (PDS). Thus, greater understanding of FD etiology and pathophysiology may require not only more advanced characterization of gastric mechanical properties, but also deeper investigation into gut-brain axis communication.

The vagus nerve acts as the major link in communication between the stomach and the brain, carrying both descending motor signals (Teckentrup et al., 2020) (Lu et al., 2018) and ascending sensory information (Browning et al., 2017). Specifically, viscerosensory afference from the GI system is carried by the vagus nerve to the nucleus tractus solitarii (NTS) in the brainstem, which then projects to higher brainstem and cortical/subcortical regions including those involved in interoception and central autonomic regulation (Chen et al., 2021, Berntson and Khalsa, 2021). Importantly, a recent study found that mechanosensory stretch receptors in the stomach and duodenum that respond to wall distention project directly to vagal neurons (Bai et al., 2019). Hence, gastric kinematics are likely an excellent proxy for vagal afference to the brain. Recent advances in Magnetic Resonance Imaging (MRI) technology have allowed for 4D cine assessments of motility and kinematic metrics (Sclocco et al., 2021), in addition to comprehensive measures of volume and emptying, as previously reported with conventional MRI (de Zwart and de Roos, 2010) (Marciani, 2011). Combining assessment of gastric kinematics with measures of brain processing of vagal afference may help elucidate how gut-brain axis communication is associated with FD pathophysiology.

Previous research has noted altered brain activity and connectivity in patients with FD compared to healthy controls, which may be linked with GI symptoms (for reviews see (Skrobisz et al., 2019) and (Kano et al., 2018)). Recent studies have begun to link brain functional MRI (fMRI) assessments with gastric physiology, such as gastric electrical rhythm as measured by cutaneous electrogastrogram (EGG) (Rebollo et al., 2018) (Rebollo and Tallon-Baudry, 2021) (Choe et al., 2021), in healthy controls. However, no study to date has directly linked mechanoreceptive gastric afference, approximated by MRI metrics of gastric kinematics, to brainstem viscerosensory connectivity in patients with FD. In this study, we combined our previously validated 4D cine MRI gastric imaging technique (Sclocco et al., 2021) with brain fMRI in order to examine stomach function and the central processing of gastric afference in FD. We hypothesized that compared to healthy adults, patients with FD would demonstrate altered antral motility and NTS connectivity to higher cortical regions known for executive and cognitive processing of sensory signaling. Furthermore, we hypothesized that gastric dysmotility in FD would be directly linked with NTS connectivity to these cortical processing brain regions.

2. Methods

This study was conducted as a single-center neuroimaging study, pre-registered with ClinicalTrials.gov (NCT03603730). The study took place at the Center for Neurointestinal Health, Department of Gastroenterology, Massachusetts General Hospital (MGH) and the Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, MGH, in Boston, MA from October, 2018 to March, 2021. All study protocols were approved by MGH and Partners Human Research Committee and all subjects provided written informed consent.

2.1. Participants

Participants aged 18 to 65 years old with history of epigastric pain and/or burning, early satiation, or postprandial fullness meeting ROME IV criteria (Drossman, 2016) for FD for at least six months were recruited to this study. Additionally, patients were required to have had an upper endoscopy within the last 2 years which showed no structural or organic cause for their upper GI symptoms or carry a diagnosis of FD with a previous normal upper GI endoscopy. Patients were excluded if they had a previous gastrointestinal surgery, electrolyte disturbances, renal insufficiency, iron overload disorders, or estimated Glomerular Filtration Rate (eGFR) less than 60 mL/min/1.73m2. Patients were also excluded if they had diabetes, mitochondrial disease, severe autonomic dysfunction, small fiber polyneuropathy, history of arrhythmias, or other conditions that could interfere with the study. Patients were required to stop taking medications that affect gastrointestinal motility or containing tetrahydrocannabinol (THC) at least seven days prior to the start of the study until completion of the last visit. None were on chronic daily use of narcotics. Age- and sex-matched healthy controls without gastrointestinal issues were enrolled for comparison. All subjects were required to be free of contraindications for MRI including metallic implants, cardiac pacemakers, pregnancy, and any other characteristic that increases the risk of being in a high magnetic field. Subjects were also excluded if they had a documented allergy to pineapple, which was used in the test meal during the imaging visit.

2.2. Study protocol

Subjects meeting enrollment criteria attended a screening visit, a behavioral assessment session involving a nutrient drink test, and an imaging visit. At the behavioral visit, participants also completed clinical questionnaires—Patient Assessment of Gastrointestinal Disorders Symptom Severity Index (PAGI-SYM) (Rentz et al., 2004), short form-Nepean Dyspepsia Index (SF-NDI) (Tack et al., 2006), and Patient-Reported Outcomes Measurement Information System (PROMIS®) Anxiety and Depression short forms (Cella et al., 2010).

In preparation for the imaging visit, participants were instructed to stop eating by 10pm of the day before the MRI scan, which occurred between 7am and 11am. A lactose-free, high caloric meal (470 ml semi-solid pudding; 0.44 g of fat, 154 g of carbohydrates, of which 133 g sugar), prepared by incorporating 400 g of fresh, blended pineapple (200 kcal) into 150 g of instant vanilla pudding (Jell-O, Kraft Foods, 360 kcal), was used as a food-based contrast meal, given the known high manganese content for pineapple (Riordan et al., 2004) (Arthurs et al., 2014). Subjects were instructed to consume as much of the 470 ml pudding as they could, up to their maximum tolerable limit, and the amount consumed was recorded along with the duration of consumption. Post meal, subjects were sequentially scanned in a supine position with abdominal (gastric) MRI (+15, +45, +70 minutes) and brain fMRI (+30, +60, +80 minutes) at 3T (Siemens Skyra) (Figure 1). The measurements were acquired as soon as the subject could be placed into the MRI bore after they finished the meal intake, following the nuclear medicine standard for clinical evaluations. Before each scan, subjects were asked to rate their current level of abdominal discomfort using a numerical rating scale of 0 to 100, with anchors of unnoticeable and unbearable, respectively.

Figure 1.

Figure 1.

Open in a new tab

Gut and brain MR imaging protocol. Following ingestion of the contrast meal, subjects were positioned supine on the MRI table which was translated to allow for sequential gut (G0–2) and brain (B0–2) MRI, three scans each per body region.

2.3. Gut and Brain MRI Data Acquisition

Stomach images were acquired with an eight-channel spine coil and a twelve-channel flexible body coil, positioned over the abdomen, as previously described (Sclocco et al., 2021). With the field of view positioned over the stomach, forty volumes over five minutes of 4D cine-MRI gastric scans were collected at each stomach imaging timepoint. This real time 3D cine sequence (3D stack of stars Gradient Echo Fast Low Angle Shot (GRE FLASH), TR = 2.7 ms, TE = 1.14 ms, FA = 3 deg, real-time temporal resolution = 7 s, 40 body coronal slices, native 2.1 × 2.1 mm2 in-plane resolution, 3.5 mm slice thickness) does not require breath holding.

Blood-oxygenation level-dependent (BOLD) fMRI data were acquired using a sixty-four-channel head/neck coil during the brain scans. Whole-brain fMRI data were acquired with gradient-echo echo-planar imaging (EPI) and a Simultaneous Multi-Slice acquisition with multiband factor 5 with the following parameters: 2.04 × 2.04 × 2.00 mm voxel size, 75 axial slices, repetition time (TR) 1270 ms, echo time (TE) 33 ms, flip angle 65°, 288 time-series measurements. After the first fMRI scan, an additional volume was collected having opposite phase encoding, in order to estimate and correct susceptibility-induced distortion during pre-processing (topup, FSL, (Andersson et al., 2003)). To aid brain to template co-registration, a high-resolution (1 mm isotropic voxel size) T1-weighted structural MRI scan was collected for each subject with a multi-echo MP-RAGE pulse sequence (TR = 2530 ms, TE1/TE2/TE3/TE4 = 1.69/3.55/5.41/7.27 ms, flip angle = 7°, field of view = 256 256 mm2, 176 axial slices.

2.4. 4D Cine Stomach Image Analysis

Full details of the stomach imaging analysis pipeline using custom Matlab scripts (R2020a, The MathWorks, Natick, MA, USA) were previously described (Sclocco et al., 2021). Briefly, a manual segmentation was performed on the first volume of each abdominal scan to generate a 3D mask of the stomach. This initial segmentation was then propagated to the remaining volumes with a nonlinear geometric transformation field (imregdemons, Matlab), yielding a 4D gastric mask. The calculated transformation matrices were inverted and applied to the initial mask. This 4D gastric mask was automatically segmented into air and meal volumes with histogram-based thresholding. Volume change was computed as the percentage change in total and meal volumes at G1 and G2 timepoints relative to G0.

Antral motility measures were calculated by first cropping the stomach volumes to a cuboid oriented along the longitudinal axis of the antrum (Figure 2). The number of voxels in each cross-sectional slice of the antrum was calculated for each volume. These time-varying cross-sectional area values at several points along the antrum allowed for quantification of kinematic patterns of antral contractions. At each antral cross-sectional slice, contraction frequency and percent lumen occlusion were computed. Values were averaged over the antral slices for statistical analyses. Peristaltic propagation velocity was calculated by tracking contractile wave peaks across the slice timeseries, comparing the middle antral slice and those 10 slices proximal and distal to the middle. Antral deformation was calculated from the 4D deformation field of the outer layer of the cropped antral volume. Cumulative deformation was computed for each location on the antral wall and the average value was taken for statistical analyses.

Figure 2.

Figure 2.

Open in a new tab

Calculation of motility indices from cross-sectional areas of antral slices. A cuboid was positioned along the parallel axis of the antrum, and cross-sectional areas were computed at each slice for every acquired imaging timepoint. Peristaltic propagation velocity and frequency were calculated in addition to antral occlusion by tracking the contraction cycles over time and distance along the antrum.

2.5. Brain fMRI Data Preprocessing

Full details of the BOLD fMRI image preprocessing are available in the Supplementary Materials. Briefly, BOLD data were first corrected for cardiac and respiratory pulsatility noise (3dretroicor, AFNI, (Glover et al., 2000)) before preprocessing the functional and structural images with FMRIPREP (version 20.1.1 (Esteban et al., 2019)). This preprocessing pipeline yields a segmentation of the structural image into constituent tissues, motion corrected functional images, co-registration of images to standard space, and noise regressors generated from motion realignment parameters and white matter and cerebrospinal fluid (CSF) tissue signals. Following FMRIPREP, BOLD fMRI data were skull stripped (3dSkullStrip, AFNI) and denoised by fitting a general linear model (GLM) containing the noise regressors and a censoring confound matrix of high head motion time points (framewise displacement greater than 0.5mm) (Power et al., 2014). This GLM was combined with a temporal highpass filter (0.008 Hz) and applied as a single step (3dTproject, AFNI) to prevent reintroduction of noise from a modular preprocessing pipeline (Lindquist et al., 2019). Spatial smoothing of the residual signal image after denoising was performed with a custom procedure that applied smoothing separately to the brainstem with a specific brainstem-only mask which removed surface vessel noise sources (Gaussian kernel with a full width half maximum, FWHM=3mm), and to the rest of the brain (FWHM=5mm). The smoothed, denoised image was then registered to standard space with the nonlinear warp computed by FMRIPREP and used for functional connectivity analyses. BOLD data with excessive estimated head motion (greater than 1mm mean framewise displacement) were dropped from analyses.

2.6. Nucleus of the solitary tract seed-voxel functional connectivity analysis

Following preprocessing, average BOLD timeseries extracted from a region previously localized as the NTS (Sclocco et al., 2019) (Figure 4), were used to generate seed-to-voxel whole-brain functional connectivity maps. Each preprocessed BOLD image was temporarily resampled to 1mm isotropic resolution to match that of the left NTS seed map when extracting the timeseries. The seed-voxel maps were generated using 2mm isotropic data. These maps were combined across scans for each subject in a fixed effects model (FEAT, FSL) to generate individual connectivity maps. In order to assess network properties of NTS connectivity, similarity between these NTS functional connectivity maps and the seven canonical resting state networks (Yeo et al., 2011) was assessed using point-biserial correlations and compared between FD and healthy control groups in a mixed-effects ANOVA (JASP Team, 2020) with between subjects effect of Group and fixed effect of Network. Post-hoc pairwise t-tests were performed to assess group differences for each network. The subject-level NTS connectivity maps were then contrasted between groups (mixed effects, FLAME 1+2, FSL), correcting for multiple comparisons with a cluster forming threshold of z > 2.3 and a cluster extent threshold of p < 0.05 to control the family wise error rate.

Figure 4.

Figure 4.

Open in a new tab

Altered patterns of NTS functional connectivity in patients with FD compared to healthy controls. A) The NTS functional connectivity pattern loaded most strongly onto the default mode network in both groups. Patients with FD displayed higher loading of NTS functional connectivity pattern onto the frontoparietal network compared to healthy controls. B) Seed-voxel whole brain functional connectivity analysis demonstrated greater connectivity between the NTS and the anterior cingulate cortex, pre-supplementary motor area, ventrolateral prefrontal cortex, and anterior insula in patients with FD relative to healthy controls. FD – functional dyspepsia, HC – healthy control, NTS – nucleus tractus solitarii.

2.7. Statistical Analyses

Differences in demographic measures between patients with FD and healthy controls were assessed with unpaired t-tests while clinical questionnaire scores were compared using non-parametric Mann-Whitney tests given the non-gaussian distributions of the PAGI-SYM, SF-NPD, and PROMIS measures.

Volumetric and antral motility indices were evaluated using mixed effect models for repeated measures (LmerTest package, R) with fixed effects of Scan (G0, G1, G2) and Group (FD, healthy controls), as well as Group*Scan interaction, and random effect of Subject. Ingested Volume was also included as a covariate. When the effect of Scan was not statistically significant, indices were averaged across available stomach scans for each session for correlations with brain functional connectivity measures. The range of peristaltic contraction frequencies for each subject were computed as the difference between the maximum and minimum frequency, in order to assess the stability of this measure.

Z-scored values from clusters showing significant group differences were extracted from each subject’s mean NTS connectivity map using spherical masks (3mm radius) with centers located at the cluster peak z-value (NiftiSpheresMasker, Nilearn). These connectivity values were correlated with the averaged antral motility indices. Multiple comparisons were FDR-corrected through the Benjamin-Hochberg procedure (pFDR= 0.005). However, since this study is exploratory in nature, we also report correlations that passed the uncorrected threshold of p < 0.05. Statistical tests were performed in R (R Core Team, 2017) unless otherwise specified, and plotting was performed with ggplot2 (Wickham, Hadley, 2016) and raincloud plots packages (Allen et al., 2021). All statistical results are reported as mean +/− standard deviation unless otherwise specified.

3. Results

3.1. Demographic and Clinical Characterization

FD (N=15) and healthy control (N=14) groups did not differ in age (FD: 29.1 +/− 13.2 yrs; HC: 31.4 +/− 8.2 yrs; t = −0.57, p = 0.58), sex (FD: 13 females, 2 males; HC: 9 females, 5 males; proportion test z = 1.41, p = 0.16) or body mass index (FD: 24.0 +/− 3.9 kg/m2; HC: 24.8 +/− 3.6 kg/m2; t = −0.56, p = 0.58). As expected, the groups differed on clinical outcomes (Table 1). Specifically, patients with FD reported more severe gastrointestinal symptoms on the PAGI-SYM and disrupted quality of life on the SF-NDI. Patients with FD also reported greater anxiety and depression severity on the PROMIS short forms. Gastric emptying scores were normal (94.01 +/− 7.79 %), for patients with FD with available scores (N=8). This group of patients with FD was composed of eleven patients that were predominantly PPD subtype, three patients with FD that were predominantly EPS subtype, and one subject that equally overlapped with both subtypes.

Table 1.

Clinical Data

Measure FD HC Test Statistic p-value
Short Nepean 26.9 +/− 6.80 (N=15) 10.0 +/− 0.76 (N=15) W = 225.0 < 0.001
PAGI-SYM 43.6 +/− 18.6 (N=14) 3.53 +/− 4.37 (N=15) W = 208.0 < 0.001
PROMIS Anxiety 53.5 +/− 7.04 (N=15) 47.2 +/− 7.76 (N=14) W = 150.5 0.049
PROMIS Depression 49.1 +/− 7.71 (N=15) 42.9 +/− 5.55 (N=14) W = 155.0 0.027
Open in a new tab

During the MRI visits, patients with FD consumed significantly less of the contrast agent meal compared to healthy controls (FD: 337 +/− 141mL, HC: 456 +/− 48.7mL; Mann Whitney W = 13, p < 0.001); however, both groups ate at comparable rates (FD: 54.5 +/− 29.5mL/min; HC: 61.5 +/− 37.1; t = −0.40, p = 0.69). After consuming the meal, patients with FD experienced greater discomfort compared to healthy controls, both when averaged across all post-meal measurements (FD: 16.5 +/− 16.0; HC: 2.73 +/− 5.16; W = 155.5, p = 0.006) and when considering only the maximum post-meal discomfort (FD: 26.4 +/− 20.3; HC: 5.7 +/−10.0; W = 156.5, p = 0.005) as rated at the MRI visit. Within the FD group, the volume of meal ingested was negatively correlated with both maximum discomfort (spearman rho = −0.64, p = 0.025) and with PAGI-SYM score (spearman rho = −0.72, p = 0.003). Furthermore, maximum discomfort was significantly correlated with PAGI-SYM score within the FD group (spearman rho = 0.65, p = 0.009). Following meal consumption, subjects underwent three abdominal scans and three brain scans, alternating between the two. There were no significant differences between groups in the times that each scan was acquired relative to the end of meal consumption (Supplementary Table S1).

3.2. Volumetric and Antral Motility Indices

Patients with FD and healthy controls were compared on each volumetric and antral motility measure using mixed effects models, controlling for the effect of Scan and Ingested Volume. There was a significant main effect of Group on peristaltic propagation velocity (F = 8.02, p = 0.006) (Figure 3A). Post-hoc testing revealed that, compared to healthy controls, patients with FD displayed lower mean peristaltic propagation velocity (post-hoc t-test: FD: 3.8 +/− 0.8 mm/s; HC: 5.6 +/− 1.6 mm/s; t = −3.78, p = 0.001) (Figure 3). This difference was primarily driven by higher peristaltic propagation velocity in healthy controls at stomach scan timepoints G0 and G1, but not G2 (Supplementary Table S2). As a follow up, we assessed the possible influence of the antrum size on the propagation velocity, and found no significant correlation between antral volume of interest and estimated velocity, both in HC (r = 0.05, p = 0.752) and FD (r = 0.17, p = 0.329). Peristaltic contraction frequencies were consistently estimated to be in the normal gastric rhythm range for both groups (FD = 2.9 +/− 0.1 cpm; HC = 2.9 +/− 0.1 cpm), and were relatively stable across timepoints for each subject (FD max – min frequency = 0.2 +/− 0.1 cpm; HC = 0.2 +/− 0.1 cpm), but not different between groups (Main effect of Group: F = 2.55, p = 0.12). A significant Group*Scan interaction in peristaltic frequency was also identified (F = 4.42, p = 0.017). For antral occlusion, there were trends toward significance for a main effect of Group (FD = 11.9 +/− 4.84 %; HC = 13.7 +/− 3.91%; F = 3.89, p = 0.061). No main effect of Group was found for mean antral wall deformation (F = 0.43, p = 0.52), which showed however a significant effect of Scan (F = 4.36, p = 0.018). This last observation aligns with the fact that mean deformation is related to Meal and Total Volume, which show strong Scan effects as they measure gastric emptying over time.

Figure 3.

Figure 3.

Open in a new tab

Patients with FD showed comparable gastric volumetric measurements, but altered antral motility indices relative to healthy controls. A) Patients with FD displayed lower peristaltic propagation velocity and lower antral occlusion compared to healthy controls. No differences were observed in frequency or mean deformation of the antral contractions. B) Total, meal, and air volumes were assessed at each gut imaging timepoint. Patients and controls did not differ in rates of volume change (emptying was consistent across groups), and absolute volumetric differences were driven by differences in ingested volumes. P-values shown represent the main effect of Group in mixed effects models. FD – functional dyspepsia, HC – healthy controls.

Compared to healthy controls, patients with FD showed lower total, meal, and air volumes (Figure 3B), but the main effect of Group was not significant (Total Volume: F = 2.90, p = 0.10; Meal Volume: F = 1.21, p =0.28; Air Volume: F = 1.83, p = 0.19) when Ingested Volume was included the model. Two measures of gastric emptying rate, the percent change in total volume and meal volume from G0 to G1 and from G0 to G2, did not show a significant main effects of Group (% change total volume: F = 0.89, p = 0.35; % change meal volume: F = 0.20, p = 0.66).

3.3. NTS Functional Connectivity

In order to compare the whole-brain functional connectivity patterns quantified by NTS seed connectivity, to the seven major resting state networks (Yeo et al., 2011), we computed point-biserial correlations for each subject’s NTS connectivity map using binary masks for each network (Figure 4a). These correlations for each subject were then z-scored and passed up to a group analysis using a mixed effects ANOVA model (JASP), which yielded a significant effect of Network (F = 9.58, p < 0.001). The NTS connectivity pattern loaded most strongly onto the Default Mode Network (DMN) in both groups. Adjusting for multiple comparison with Bonferroni-Holm p-value adjustment, t-tests of NTS loading values between pairs of the seven networks revealed that the loading on the DMN (0.16 +/− 0.18) was significantly higher than that of the Somatomotor (−0.07 +/− 0.14, t = −4.67, p.adj = 0.001), Dorsal Attention (−0.03 +/− 0.11, t = 3.61, p.adj = 0.025), Ventral Attention (−0.08 0.11, t = 4.96, p.adj < 0.0001), Limbic (−0.04 +/− 0.10, t = 6.26, p.adj < 0.0001), and Frontoparietal (−0.02 +/− 0.11, t = 4.23, p.adj = 0.005) networks. We also ran exploratory post-hoc t-tests comparing NTS network loading values between FD and HC groups for each network. The NTS loading onto Frontoparietal Network (FP) was stronger in patients with FD compared to HCs (FD: 0.03 +/− 0.10, HC: −0.07 +/− 0.10, t = 2.58, p = 0.016).

In a whole-brain voxel-wise analysis, patients with FD demonstrated higher NTS connectivity to anterior cingulate cortex (ACC), pre supplementary motor area (preSMA), bilateral ventrolateral prefrontal cortices (vlPFC), and the left anterior insula (aIns) relative to healthy controls (Figure 4b, Supplementary Table S2). No brain regions showed greater NTS connectivity in healthy controls relative to patients with FD.

3.4. Correlation between FC and Antral Motility Indices

NTS functional connectivity to five regions of interest (ROIs) (Supplementary Table S2) noted above from the FD vs healthy control group contrast map was correlated with antral motility indices (Figure 5). In a combined group (FD and HC) sample, average NTS connectivity to preSMA and left vlPFC ROIs was negatively correlated to average peristaltic propagation velocity (preSMA r = −0.51, p = 0.009, uncorrected; left vlPFC r = −0.63, p = 0.007, FDR-corrected). Peristaltic frequency in the combined group was significantly correlated to the connectivity between NTS and preSMA (r = −0.40, p = 0.046, uncorrected), left vlPFC (r = −0.43, p = 0.033, uncorrected), and the left aIns (r = −0.58, p = 0.002, FDR-corrected).

Figure 5.

Figure 5.

Open in a new tab

Antral motility indices are correlated with functional connectivity between the NTS and motor and processing regions. In the combined group (FD and HC) sample, peristaltic frequency was correlated to functional connectivity between the NTS and the left vlPFC, preSMA, and left anterior insula. These correlations were significant for a non-combined, FD-alone, group as well (Table 2). Peristaltic propagation velocity was correlated to functional connectivity between the NTS and the left vlPFC and preSMA. The correlation between velocity and NTS / vlPFC connectivity was also significant for a non-combined, HC-alone, group as well. vlPFC – ventrolateral prefrontal cortex, preSMA – pre-supplementary motor area, LaIns – left anterior insula, NTS – nucleus tractus solitarii.

Correlations between NTS functional connectivity to regions of interest and motility indices were also assessed separately for each group. Within the healthy control group alone, there was a significant correlation between average propagation velocity and NTS / left vlPFC connectivity remained significant (r = −0.63, p = 0.021, uncorrected). These correlation between brain connectivity and peristaltic frequency were also significant within the FD group alone (preSMA: r = −0.59, p = 0.041, uncorrected; left vlPFC: r = −0.70, p = 0.012, uncorrected; left aIns: r = −0.69, p = 0.014, uncorrected), but not within the healthy control group alone (Table 2).

Table 2.

Gut-Brain Correlations within Groups

Gut Metric Brain Metric FD HC
velocity NTS/preSMA FC r = 0.22 p = 0.50 r = −0.34 p = 0.26
velocity NTS/L vlPFC FC r = −0.16 p = 0.62 r = −0.63 p = 0.021 †
frequency NTS/preSMA FC r = −0.59 p = 0.041 † r = −0.15 p = 0.73
frequency NTS/L vlPFC FC r = −0.70 p = 0.012 † r = 0.13 p = 0.68
frequency NTS/LaIns FC r = −0.69 p = 0.014 † r = −0.41 p = 0.46
Open in a new tab

(†p < 0.05, uncorrected for multiple comparisons)

4. Discussion

Functional dyspepsia has been hypothesized to be a disorder of gut-brain axis signaling. Our multi-modal MRI approach evaluated gut-brain axis function, demonstrating the ability to assess both gastric kinematics and brain processing of gastric afference, as well as the links between gut and brain function in a post-meal state. Compared to healthy controls, patients with FD demonstrated slower propagation of gastric peristaltic waves and altered brain connectivity – i.e. shifted NTS connectivity from the self-referential processing DMN to the executive control FP network, suggesting altered cognitive processing of interoceptive (gastric) signaling in FD. Importantly, NTS connectivity was directly linked to the velocity of antral contraction wave propagation in healthy controls, but not in patients with FD. On the other hand, the frequency of gastric peristalsis was more tightly linked with NTS connectivity in patients with FD rather than healthy controls. These associations suggest that patients suffering from FD show specific plasticity in gut-brain communication, which can be assessed with our novel multi-modal MRI approach.

This study linked gastric kinematics with brainstem-cortical functional connectivity. Velocity and frequency of antral peristaltic waves were linked to connectivity between the NTS and vlPFC, preSMA, and anterior insula. In particular, the insular cortex has been strongly implicated in interoceptive processing, especially with regard to receiving signals from the NTS (Berntson and Khalsa, 2021). The anterior division of the insula may mediate interoceptive attention (Wang et al., 2019), and higher connectivity to the NTS in patients with FD may reflect greater attention to bothersome gastric sensations. The pre-supplementary motor area is involved in motor planning and execution, particularly internally motivated actions, along with the supplementary motor area and motor cortex (Nachev et al., 2008). However, the preSMA may specialize more in the higher order aspects of motor coordination such as procedural learning and perception (Ruan et al., 2018). Communication between the NTS and preSMA could reflect higher order coordination and processing of visceral afference and abdominal muscle activity in the presence of bothersome sensations such as nausea (Babic and Browning, 2014). Patients with FD also displayed higher connectivity between the NTS and bilateral vlPFC, a region implicated in motor inhibition (Levy and Wagner, 2011) and in cognitive control, such as in retrieval and storage of working memory (Badre and Wagner, 2007). Taken together, our results suggest that communication between the major visceral afference integration center in the brainstem (NTS) and cortical areas involved in interoception, motor planning, and cognitive control may be relevant for FD pathophysiology, including the processing or control of gastric kinematics.

The link between gastric function and brainstem-cortical connectivity highlighted in the present study complements prior explorations of gut-brain communication. For instance, brain fMRI activity in somatomotor and DMN (e.g. posterior cingulate cortex, precuneus) regions was shown to correlate with basal EGG rhythm in healthy subjects (Rebollo et al., 2018). Another study confirmed synchrony between EGG and resting fMRI signal in the DMN and a dorsal somatomotor network (Choe et al., 2021) in a highly sampled healthy adult. Our study found that NTS connectivity patterns loaded most strongly onto the DMN in both patients with FD and healthy controls, corroborating involvement of DMN in cortical processing of gastric afference. Interestingly, increased functional connectivity between the hippocampus and the frontal regions of the DMN has been linked to central insulin release (Kullmann et al., 2017). Differential activation with administration of the motilin-receptor agonist erythromycin has been shown in right anterior insular cortex, and also linked to insulin release (Zhao et al., 2018). Assessing insulin levels and blood glucose in future studies would provide additional insight into the underlying metabolic factors that interact with gut-brain communication. Additionally, we observed that NTS connectivity to preSMA was coupled to peristaltic frequency and velocity. While EGG assessments may reflect peristaltic frequency, our MRI-based gastric kinematic metrics can provide a more complete evaluation of gastric physiology. Firstly, our study demonstrated that FD and healthy controls have similar frequencies of gastric contractions, corroborating a recent study which found that EGG measures in patients with FD were far more similar to controls than nausea and vomiting syndromes (Carson et al., 2021). Thus, peristaltic frequency, whether measured by EGG or MRI, may not be the most relevant measure for understanding FD pathophysiology. Secondly, studies have reported mechano-electrical dissociation between the EGG signal, which measures gastric slow waves generated by the interstitial cells of Cajal (ICCs) (Koch and Stern, 2004), and gastric peristaltic contractions. For example, some patients with obstructive post-surgical ileus can show normal gastric rhythms (Frasko et al., 2008). In fact, afference related to mechanical action of the stomach (and intestine) is transduced by several distinct populations of vagal sensory neurons. For instance, the pattern generating ICCs form multi-contact complexes with intramuscular arrays (IMAs) (Powley and Phillips, 2011), vagal mechanosensory neurons embedded in gastric smooth muscle. Another type of vagal mechanosensory neurons, intraganglionic laminar endings (IGLEs), innervate the antrum, fundus, and corpus as well as the intestine, but are most densely distributed on either side of the pylorus (Bai et al., 2019). While IMAs may be more directly coupled to ICC activity, the dense IGLE populations near the pylorus might capture additional aspects of gastric distension and contractions. Therefore, measurements of gastric kinematics may provide additional insights into mechanosensation in the gut beyond the dominant gastric contraction rhythm, commonly assessed by EGG.

Using our novel MRI assessment of gastric kinematics, recently validated in healthy adults (Sclocco et al., 2021), we found that both healthy controls and patients with FD demonstrated decreasing total and meal volumes across timepoints, with no significant differences between groups, when controlling for ingested volume. While delayed gastric emptying may be present in FD, it is not a consistent hallmark of the disorder (Talley et al., 2006), particularly when compared to gastroparesis (Stein et al., 2014). Similarly, our method identified contraction frequencies in the normogastric range (2–4 cycles/minute) for both groups, consistent within each subject. However, antral peristaltic waves propagated with lower velocity and lower luminal occlusion in FD compared to healthy controls. These alterations may indicate less efficient or effective peristalsis that could impair gastric function, even if global measures of gastric emptying were normal. Thus, reduced velocity and occlusion suggest lower contractility and reduced antral efficiency in translating stomach contents toward the pylorus, potentially contributing to bothersome clinical symptoms in FD.

While we report novel findings in this study, it is not without limitations. For instance, our MRI assessments concluded after approximately 80 minutes post meal, whereas traditional methods assess gastric activity up to 4 hours after ingestion (Farrell, 2019). We instructed subjects to consume their maximum tolerable amount of contrast meal, and found lower volumes ingested by patients with FD, on average. For this reason, Ingested Volume was controlled for in ANOVA models, and this variable explained apparent group differences in volumetric measures. Furthermore, lower meal consumption was correlated with higher PAGI-SYM scores, and discomfort experienced during the scanning visit. These associations support the hypothesis that lower volume for the FD group was explained by patients with FD stopping meal consumption earlier as existing symptoms were exacerbated. Future studies could try to assess FD versus healthy control differences in gastric kinematics with comparable stomach volumes. Additionally, the MRI pulse sequence used in the study did not provide sufficient tissue contrast to allow for a baseline preprandial scan. We have attempted to address this limitation by keeping all participants NPO for at least 12 hours prior to the scan. Furthermore, we have since developed a new pulse sequence that successfully captures gastric volumetric and kinematic information in a fasted (preprandial) state, and are currently working to apply this protocol in future studies. Lastly, no measure except for Ingested Volume was directly correlated to the discomfort experienced by patients with FD during the imaging visit. Our study sample included predominantly PDS FD subtypes, but also several EPS subtype patients, potentially contributing to the broad range of discomfort ratings in the FD group. It may be possible that PDS patients may exhibit a tighter link between a gut or brain metric and discomfort, which should be explored in a future large cohort study focused on this question.

In conclusion, our multi-modal MRI study demonstrated the ability to assess gut-brain axis plasticity in FD, linking gastric kinematics, and NTS-cortical functional connectivity providing novel insights into FD pathophysiology.

Supplementary Material

Supinfo

Acknowledgements

The present work was supported by the following organizations: US National Instituted of Health (NIH), Office Of The Director (OT2-OD023867); Center for Functional Neuroimaging Technologies (P41-EB015896); National Center for Complementary and Integrative Health (P01-AT006663); National Institute of Diabetes and Digestive and Kidney Diseases (R21-DK116029, U01DK112193); NIDDK Diabetic Complications Consortium (DK076169, DK115255). This work also involved the use of instrumentation supported by the NIH Shared Instrumentation Grant Program and/or High-End Instrumentation Grant Program; specifically, Grant no. S1-0RR023043.

Footnotes

Conflicts of interest: none.

Data Availability Statement:

The datasets generated during and analyzed during the current study are available from the corresponding author on reasonable request.

5. References

  1. Allen M, Poggiali D, Whitaker K, Marshall TR, van Langen J, Kievit RA (2021) Raincloud plots: a multi-platform tool for robust data visualization. Wellcome Open Res 4:63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Andersson JLR, Skare S, Ashburner J (2003) How to correct susceptibility distortions in spin-echo echo-planar images: application to diffusion tensor imaging. NeuroImage 20:870–888. [DOI] [PubMed] [Google Scholar]
  3. Arthurs OJ, Graves MJ, Edwards AD, Joubert I, Set PAK, Lomas DJ (2014) Interactive neonatal gastrointestinal magnetic resonance imaging using fruit juice as an oral contrast media. BMC Med Imaging 14:33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Babic T, Browning KN (2014) The role of vagal neurocircuits in the regulation of nausea and vomiting. Eur J Pharmacol 722:38–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Badre D, Wagner AD (2007) Left ventrolateral prefrontal cortex and the cognitive control of memory. Neuropsychologia 45:2883–2901. [DOI] [PubMed] [Google Scholar]
  6. Bai L, Mesgarzadeh S, Ramesh KS, Huey EL, Liu Y, Gray LA, Aitken TJ, Chen Y, Beutler LR, Ahn JS, Madisen L, Zeng H, Krasnow MA, Knight ZA (2019) Genetic Identification of Vagal Sensory Neurons That Control Feeding. Cell 179:1129–1143.e23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Berntson GG, Khalsa SS (2021) Neural Circuits of Interoception. Trends in Neurosciences 44:17–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Browning KN, Verheijden S, Boeckxstaens GE (2017) The Vagus Nerve in Appetite Regulation, Mood, and Intestinal Inflammation. Gastroenterology 152:730–744. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Carson DA, Bhat S, Hayes TCL, Gharibans AA, Andrews CN, O’Grady G, Varghese C (2021) Abnormalities on Electrogastrography in Nausea and Vomiting Syndromes: A Systematic Review, Meta-Analysis, and Comparison to Other Gastric Disorders. Dig Dis Sci Available at: 10.1007/s10620-021-07026-x [Accessed August 3, 2021]. [DOI] [PubMed] [Google Scholar]
  10. Cella D et al. (2010) The Patient-Reported Outcomes Measurement Information System (PROMIS) developed and tested its first wave of adult self-reported health outcome item banks: 2005–2008. J Clin Epidemiol 63:1179–1194. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chen WG, Schloesser D, Arensdorf AM, Simmons JM, Cui C, Valentino R, Gnadt JW, Nielsen L, Hillaire-Clarke CS, Spruance V, Horowitz TS, Vallejo YF, Langevin HM (2021) The Emerging Science of Interoception: Sensing, Integrating, Interpreting, and Regulating Signals within the Self. Trends in Neurosciences 44:3–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Choe AS, Tang B, Smith KR, Honari H, Lindquist MA, Caffo BS, Pekar JJ (2021) Phase-locking of resting-state brain networks with the gastric basal electrical rhythm. PLOS ONE 16:e0244756. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. de Zwart IM, de Roos A (2010) MRI for the evaluation of gastric physiology. Eur Radiol 20:2609–2616. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Drossman DA (2016) Functional Gastrointestinal Disorders: History, Pathophysiology, Clinical Features and Rome IV. Gastroenterology:S0016-5085(16)00223-7. [DOI] [PubMed] [Google Scholar]
  15. Esteban O, Markiewicz CJ, Blair RW, Moodie CA, Isik AI, Erramuzpe A, Kent JD, Goncalves M, DuPre E, Snyder M, Oya H, Ghosh SS, Wright J, Durnez J, Poldrack RA, Gorgolewski KJ (2019) fMRIPrep: a robust preprocessing pipeline for functional MRI. Nat Methods 16:111–116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Farrell MB (2019) Gastric Emptying Scintigraphy. Journal of Nuclear Medicine Technology 47:111–119. [DOI] [PubMed] [Google Scholar]
  17. Frasko R, Maruna P, Gurlich R, Trca S (2008) Transcutaneous electrogastrography in patients with ileus. Relations to interleukin-1beta, interleukin-6, procalcitonin and C-reactive protein. Eur Surg Res 41:197–202. [DOI] [PubMed] [Google Scholar]
  18. Glover GH, Li TQ, Ress D (2000) Image-based method for retrospective correction of physiological motion effects in fMRI: RETROICOR. Magn Reson Med 44:162–167. [DOI] [PubMed] [Google Scholar]
  19. Kano M, Dupont P, Aziz Q, Fukudo S (2018) Understanding Neurogastroenterology From Neuroimaging Perspective: A Comprehensive Review of Functional and Structural Brain Imaging in Functional Gastrointestinal Disorders. Journal of Neurogastroenterology and Motility 24:512–527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kim DY, Delgado-Aros S, Camilleri M, Samsom M, Murray JA, O’Connor MK, Brinkmann BH, Stephens DA, Lighvani SS, Burton DD (2001) Noninvasive measurement of gastric accommodation in patients with idiopathic nonulcer dyspepsia. Am J Gastroenterol 96:3099–3105. [DOI] [PubMed] [Google Scholar]
  21. Koch KL, Stern RM (2004) Handbook of Electrogastrography. Oxford University Press. [Google Scholar]
  22. Kullmann S, Heni M, Veit R, Scheffler K, Machann J, Häring HU, Fritsche A, Preissl H (2017). Intranasal insulin enhances brain functional connectivity mediating the relationship between adiposity and subjective feeling of hunger. Scientific reports 7(1):1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Leahy A, Besherdas K, Clayman C, Mason I, Epstein O (1999) Abnormalities of the electrogastrogram in functional gastrointestinal disorders - PubMed. American Journal of Gastroenterology 94:1023–1028. [DOI] [PubMed] [Google Scholar]
  24. Lee I-S, Wang H, Chae Y, Preissl H, Enck P (2016) Functional neuroimaging studies in functional dyspepsia patients: a systematic review. Neurogastroenterol Motil 28:793–805. [DOI] [PubMed] [Google Scholar]
  25. Levy BJ, Wagner AD (2011) Cognitive control and right ventrolateral prefrontal cortex: reflexive reorienting, motor inhibition, and action updating. Ann N Y Acad Sci 1224:40–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lin Z, Eaker EY, Sarosiek I, McCallum RW (1999) Gastric myoelectrical activity and gastric emptying in patients with functional dyspepsia. Am J Gastroenterol 94:2384–2389. [DOI] [PubMed] [Google Scholar]
  27. Lindquist MA, Geuter S, Wager TD, Caffo BS (2019) Modular preprocessing pipelines can reintroduce artifacts into fMRI data. Human Brain Mapping 40:2358–2376. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lorena SLS, Tinois E, Brunetto SQ, Camargo EE, Mesquita MA (2004) Gastric emptying and intragastric distribution of a solid meal in functional dyspepsia: influence of gender and anxiety. J Clin Gastroenterol 38:230–236. [DOI] [PubMed] [Google Scholar]
  29. Lu K-H, Cao J, Oleson S, Ward MP, Phillips RJ, Powley TL, Liu Z (2018) Vagus nerve stimulation promotes gastric emptying by increasing pyloric opening measured with magnetic resonance imaging. Neurogastroenterol Motil 30:e13380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lunding JA, Tefera S, Bayati A, Gilja OH, Mattsson H, Hausken T, Berstad A (2006) Pressure-induced gastric accommodation studied with a new distension paradigm. Abnormally low accommodation rate in patients with functional dyspepsia. Scand J Gastroenterol 41:544–552. [DOI] [PubMed] [Google Scholar]
  31. Marciani L (2011) Assessment of gastrointestinal motor functions by MRI: a comprehensive review. Neurogastroenterol Motil 23:399–407. [DOI] [PubMed] [Google Scholar]
  32. Nachev P, Kennard C, Husain M (2008) Functional role of the supplementary and pre-supplementary motor areas. Nat Rev Neurosci 9:856–869. [DOI] [PubMed] [Google Scholar]
  33. Power JD, Mitra A, Laumann TO, Snyder AZ, Schlaggar BL, Petersen SE (2014) Methods to detect, characterize, and remove motion artifact in resting state fMRI. Neuroimage 84:320–341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Powley TL, Phillips RJ (2011) Vagal intramuscular array afferents form complexes with interstitial cells of Cajal in gastrointestinal smooth muscle: Analogues of muscle spindle organs? Neuroscience 186:188–200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. R Core Team (2017) R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Available at: https://www.r-project.org/. [Google Scholar]
  36. Rebollo I, Devauchelle A-D, Béranger B, Tallon-Baudry C (2018) Stomach-brain synchrony reveals a novel, delayed-connectivity resting-state network in humans. eLife 7 Available at: 10.7554/eLife.33321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Rebollo I, Tallon-Baudry C (2021) The sensory and motor components of the cortical hierarchy are coupled to the rhythm of the stomach during rest. bioRxiv:2021.05.26.445829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Rentz AM, Kahrilas P, Stanghellini V, Tack J, Talley NJ, de la Loge C, Trudeau E, Dubois D, Revicki DA (2004) Development and psychometric evaluation of the patient assessment of upper gastrointestinal symptom severity index (PAGI-SYM) in patients with upper gastrointestinal disorders. Qual Life Res 13:1737–1749. [DOI] [PubMed] [Google Scholar]
  39. Riordan RD, Khonsari M, Jeffries J, Maskell GF, Cook PG (2004) Pineapple juice as a negative oral contrast agent in magnetic resonance cholangiopancreatography: a preliminary evaluation. Br J Radiol 77:991–999. [DOI] [PubMed] [Google Scholar]
  40. Ruan J, Bludau S, Palomero-Gallagher N, Caspers S, Mohlberg H, Eickhoff SB, Seitz RJ, Amunts K (2018) Cytoarchitecture, probability maps, and functions of the human supplementary and pre-supplementary motor areas. Brain Struct Funct 223:4169–4186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sarnelli G, Caenepeel P, Geypens B, Janssens J, Tack J (2003) Symptoms associated with impaired gastric emptying of solids and liquids in functional dyspepsia. Am J Gastroenterol 98:783–788. [DOI] [PubMed] [Google Scholar]
  42. Sclocco R, Garcia RG, Kettner NW, Isenburg K, Fisher HP, Hubbard CS, Ay I, Polimeni JR, Goldstein J, Makris N, Toschi N, Barbieri R, Napadow V (2019) The influence of respiration on brainstem and cardiovagal response to auricular vagus nerve stimulation: A multimodal ultrahigh-field (7T) fMRI study. Brain Stimul 12:911–921. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Sclocco R, Nguyen C, Staley R, Fisher H, Mendez A, Velez C, Kettner NW, Kuo B, Napadow V (2021) Non-uniform gastric wall kinematics revealed by 4D Cine magnetic resonance imaging in humans. Neurogastroenterology & Motility:e14146. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Skrobisz K, Piotrowicz G, Drozdowska A, Markiet K, Sabisz A, Naumczyk P, Rydzewska G, Szurowska E (2019) Use of functional magnetic resonance imaging in patients with irritable bowel syndrome and functional dyspepsia. Prz Gastroenterol 14:163–167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Stein B, Everhart KK, Lacy BE (2014) Treatment of functional dyspepsia and gastroparesis. Curr Treat Options Gastroenterol 12:385–397. [DOI] [PubMed] [Google Scholar]
  46. Tack J, Bisschops R, Sarnelli G (2004a) Pathophysiology and treatment of functional dyspepsia. Gastroenterology 127:1239–1255. [DOI] [PubMed] [Google Scholar]
  47. Tack J, Caenepeel P, Corsetti M, Janssens J (2004b) Role of tension receptors in dyspeptic patients with hypersensitivity to gastric distention - Gastroenterology. Gastroenterology 127:1058–1066. [DOI] [PubMed] [Google Scholar]
  48. Tack J, Talley NJ, Camilleri M, Holtmann G, Hu P, Malagelada J-R, Stanghellini V (2006) Functional Gastroduodenal Disorders. Gastroenterology 130:1466–1479. [DOI] [PubMed] [Google Scholar]
  49. Talley NJ, Locke GR, Lahr BD, Zinsmeister AR, Tougas G, Ligozio G, Rojavin MA, Tack J (2006) Functional dyspepsia, delayed gastric emptying, and impaired quality of life. Gut 55:933–939. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Teckentrup V, Neubert S, Santiago JCP, Hallschmid M, Walter M, Kroemer NB (2020) Non-invasive stimulation of vagal afferents reduces gastric frequency. Brain Stimulation: Basic, Translational, and Clinical Research in Neuromodulation 13:470–473. [DOI] [PubMed] [Google Scholar]
  51. Wang X, Wu Q, Egan L, Gu X, Liu P, Gu H, Yang Y, Luo J, Wu Y, Gao Z, Fan J (2019) Anterior insular cortex plays a critical role in interoceptive attention Stephan KE, Frank MJ, Stephan KE, Faull OK, Allen M, eds. eLife 8:e42265. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Wickham Hadley (2016) ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag; New York. Available at: https://ggplot2.tidyverse.org. [Google Scholar]
  53. Yeo BTT, Krienen FM, Sepulcre J, Sabuncu MR, Lashkari D, Hollinshead M, Roffman JL, Smoller JW, Zöllei L, Polimeni JR, Fischl B, Liu H, Buckner RL (2011) The organization of the human cerebral cortex estimated by intrinsic functional connectivity. J Neurophysiol 106:1125–1165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Zhao D, Meyer-Gerspach AC, Deloose E, Iven J, Weltens N, Depoortere I, O’daly O, Tack J, Van Oudenhove L (2018) The motilin agonist erythromycin increases hunger by modulating homeostatic and hedonic brain circuits in healthy women: a randomized, placebo-controlled study. Scientific reports 8(1):1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supinfo
NIHMS1808604-supplement-Supinfo.docx (30.1KB, docx)

Data Availability Statement

The datasets generated during and analyzed during the current study are available from the corresponding author on reasonable request.

RESOURCES