The Neuro-Endo-Microbio-Ome Study: A Pilot Study of Neurobiological Alterations Pre- Versus Post-Bariatric Surgery

Abstract

Background

Plausible phenotype mechanisms following bariatric surgery include changes in neural and gastrointestinal physiology. This pilot study aims to investigate individual and combined neurologic, gut microbiome, and plasma hormone changes pre- versus post-vertical sleeve gastrectomy (VSG), Roux-en-Y gastric bypass (RYGB), and medical weight loss (MWL). We hypothesized post-weight loss phenotype would be associated with changes in central reward system brain connectivity, differences in postprandial gut hormone responses, and increased gut microbiome diversity.

Methods

Subjects included participants undergoing VSG, n = 7; RYGB, n = 9; and MWL, n = 6. Ghrelin, glucagon-like peptide-1, peptide-YY, gut microbiome, and resting state functional magnetic resonance imaging (rsfMRI; using fractional amplitude of low-frequency fluctuations [fALFF]) were measured pre- and post-intervention in fasting and fed states. We explored phenotype characterization using clustering on gut hormone, microbiome, and rsfMRI datasets and a combined analysis.

Results

We observed more widespread fALFF differences post-bariatric surgery versus post-MWL. Decreased post-prandial fALFF was seen in food reward regions post-RYGB. The highest number of microbial taxa that increased post-intervention occurred in the RYGB group, followed by VSG and MWL. The combined hormone, microbiome, and MRI dataset most accurately clustered samples into pre- versus post-VSG phenotypes followed by RYGB subjects.

Conclusion

The data suggest surgical weight loss (VSG and RYGB) has a bigger impact on brain and gut function versus MWL and leads to lesser post-prandial activation of food-related neural circuits. VSG subjects had the greatest phenotype differences in interactions of microbiome, rsfMRI, and gut hormone features, followed by RYGB and MWL. These results will inform future prospective research studying gut-brain changes post-bariatric surgery.

Keywords

obesity resting state functional magnetic resonance imaging cognition gut hormones gut microbiome bariatric surgery vertical sleeve gastrectomy Roux-en-Y gastric bypass

The increased prevalence of adults with obesity has become a major health problem worldwide, having tripled over the last 40 years. In fact, 39% of adults are overweight (body mass index [BMI] ≥ 25 kg/m²) and 13% are obese (BMI ≥ 30 kg/m²) globally (World Health Organization, 2020). Bariatric surgery rapidly and sustainably decreases body mass and leads to remission of many obesity-related comorbidities, such as type 2 diabetes mellitus and metabolic syndrome (Koliaki et al., 2017; Schauer et al., 2017). The two most common procedures in current use are vertical sleeve gastrectomy (VSG), and Roux-en-Y gastric bypass (RYGB; Melton et al., 2008; Ozsoy & Demir, 2018). While the VSG removes 75 to 80% of the stomach, creating a reduction in volume, the RYGB includes both reduction in stomach capacity and re-routing of the small intestine (Mulla et al., 2018). These alterations in gastrointestinal anatomy may modulate neural reward pathways, including mesolimbic dopaminergic signaling (Ochner et al., 2010).

Recent studies using functional neuroimaging, specifically blood-oxygen-level-dependent functional magnetic resonance imaging (fMRI), have linked activation of distinct regions in the brain with changes in gut hormones and peptides (Bogdanov et al., 2020). Patients with obesity and normal weight cohorts exhibit differences in rsfMRI activity (García-García et al., 2013; Kullmann et al., 2012), and brain regions important for cognitive control, inhibition motivation, reward, and salience are involved in the neuropathology of obesity (Dong et al., 2015; Lepping et al., 2015; Zhang et al., 2015). Spontaneous low-frequency (0.01–0.08 Hz) fluctuations (LFFs) of blood-oxygen-level-dependent fMRI signals (Biswal et al., 1995) are closely related to the spontaneous neuronal activities occurring during the resting-state (Lu et al., 2007; Mantini et al., 2007), and numerous studies have used fractional amplitude of low-frequency fluctuations (fALFF) to quantify brain activity changes following VSG and RYGB (Li et al., 2018; Zeighami et al., 2021). Furthermore, shared neural activity differences post-bariatric surgery, including VSG and RYGB, also involve reward processing regions including the default mode network, salience network, and control regions (Li et al., 2018; Zeighami et al., 2021). This evidence contributes to our hypothesis that brain connectivity changes in the reward network are involved in obesity pathogenesis and reward connectivity alterations post-bariatric surgery likely contribute to sustained weight loss post-operatively. Nevertheless, brain connectivity changes post-bariatric surgery likely only contribute to a portion of patient phenotype as obesity is a complex and heterogeneous disease involving multiple body systems also including the gastrointestinal and endocrine system, among others (Lee & Mattson, 2014).

The human gut microbiome has emerged to be linked with overall health, including obesity risk. The human gut microbiome comprises those microbial populations that inhabit the gastrointestinal tract and their genes, and these microorganisms perform vital physiologic functions (Silva et al., 2020). Several studies reported that the microbiome of individuals with obesity is less diverse compared to lean individuals (Cryan & Dinan, 2012; Ley et al., 2006). The bacteria of the gut microbiome are also associated with hormones and peptides that are secreted by the gastrointestinal tract and influence feeding behavior, including ghrelin, glucagon-like peptide 1 (GLP-1), and peptide YY (PYY; Bliss & Whiteside, 2018; Covasa et al., 2019). Ghrelin, a foregut hormone, is an orexigenic peptide that stimulates food intake, while GLP-1 and PYY, produced in the hindgut, are anorexigenic and inhibit food intake (Wisser et al., 2010). Furthermore, fecal short chain fatty acids (SCFAs) may influence neurotransmitter production by modulation of precursors, suggesting that changes in the gut microbiome environment have a physiological impact that extends beyond local environment changes to influence neurologic signaling (van de Wouw et al., 2017).

Early research of the microbiota-gut-brain axis, that is, communication between the gut microbiome (and their metabolites) and the brain, set foundational knowledge for the critical role that signaling between the brain and the gastrointestinal system plays in homeostatic regulation of health and disease (Dinan & Cryan, 2017). Bidirectional communication between the brain and gut microbiome occurs through various routes including microbiome-derived metabolites, the Vagus nerve, and the enteric nervous system (Bassett et al., 2019; Wang et al., 2018). Well-described associations exist between the gut microbiome and neurologic conditions such as depression and autism (Cryan & Dinan, 2012). Still, we do not yet understand the role of brain-gut-microbiome signaling in the context of obesity and how weight loss with or without surgery may alter this connection.

While there is literature analyzing the change in the secretion of gut hormones and peptides in adjustable gastric banding compared to RYGB (Meek et al., 2016), few studies have compared patient phenotype differences after RYGB versus VSG and diet-induced weight loss (Frühbeck et al., 2004). We hypothesized that weight loss post-bariatric surgery would be reflected as improvement in the food reward mechanisms through changes in brain connectivity in central reward systems, return to physiologic post-prandial gut hormone responses, and increased diversity in gut microbiome composition. Therefore, the aims of the present pilot study were to: (i) individually investigate brain rsfMRI connectivity, gut microbiome, and gut hormone changes pre- vs. post-surgical (VSG, RYGB) and diet-induced weight loss (medical weight loss; MWL); and (ii) elucidate associations between the gut microbiome, gut hormones, and brain connectivity pre- vs. post-surgical (VSG, RYGB) and diet (MWL) weight loss interventions to understand the drivers of patient phenotype changes after weight loss. We believe this pilot study will inform if MRI, gut microbiome, and hormone measures are necessary to include in studies of the gut-brain axis in obesity and will also identify important biomarkers impacted by weight loss and/or bariatric surgery that may be further explored in future research.

Methods

This pilot study was a pre- vs. post-weight loss intervention design based on 2 bariatric surgical weight loss approaches (VSG, RYGB) and a dietary-induced weight loss or, medical weight loss (MWL) approach. As we were interested in the specific phenotype differences that occur after surgical or MWL, when we state the effect of the intervention tested, we are evaluating the impact of the prescribed surgical (RYGB or VSG) or MWL group. Selection of weight loss intervention (VSG vs. RYGB vs. MWL) was determined by the medical team at the Bariatric Surgery Program based on weight, clinical, and patient variables. The study involved two time points (study visit 1: 4–6 weeks prior to bariatric surgery or MWL and study visit 2: after treatment was initiated and the subject had lost 10% of baseline weight, 12 kg of weight or at 6 months; Figure 1). Average time from enrollment (study visit 1) to the date of intervention (bariatric surgery or MWL initiation) was 21.1 ± 22.6 days for VSG, 14.0 ± 12.6 days for RYGB, and 8.5 ± 10.2 days for MWL. Average time from the date of intervention (bariatric surgery or MWL initiation) to follow up (study visit 2) was 89.4 ± 39.7 days for VSG, 79.7 ± 18.8 days for RYGB, and 169.5 ± 22.7 days for MWL. The weight loss or time criteria endpoint for study visit 2 was selected to ensure the patients post-VSG and RYGB would not have extreme weight loss differences compared to MWL participants (Schauer et al., 2017), as weight loss post-surgical or MWL intervention was our main phenotype driver of interest (vs. days elapsed between intervention and follow-up). For expanded methods, please see Supplemental Methods.

Data Source and Collection

Subjects who qualified for bariatric surgery at the Johns Hopkins University Weight Management Center and subjects who initiated a dietary-induced weight loss plan (MWL) were enrolled from April 2016 to August 2017. A total of 30 adult subjects were enrolled in the study, but only subjects with complete datasets (MRI, microbiome, and hormone; n = 7 VSG, n = 9 RYGB, and n = 6 MWL) were included in the final analysis. Inclusion criteria included age between 18 to 55 and an initial BMI greater than 35 in subjects undergoing a weight loss treatment (bariatric surgery or dietary-induced weight loss). The University Institutional Review Board approved the study procedures and obtained informed consent from all subjects before enrollment.

Hunger Scale and Cognitive Testing

The participants rated subjective hunger and fullness using a Visual Analog Scale (VAS) during the test meal, prior to rsfMRI pre- and post-intervention (surgical or medical weight loss). Participants were administered a brief cognitive battery of 3 tests—the Letter Number Sequencing Test (LNS), Hopkins Verbal Learning Test (HVLT), Stroop Color and Word Test (SCWT). Indices from these tasks are grouped to the following domains: (1) auditory attention: LNS total score and the first learning trial (Trial 1) of the HVLT, (2) processing speed: Color Trial of the SCWT, (3) memory for auditory-verbal information: HVLT percent retained, HVLT Recognition Discrimination Index, and (4) executive functioning: Color Word Trial of the SCWT. These tests have been validated extensively (D’Souza et al., 2019; Periáñez et al., 2021), and are used in both research and clinical practice. For a description of the cognitive test methods, please see Supplemental Methods.

Test Meal and Plasma Hormone Measurements

Each subject consumed a liquid meal consisting of 237 mL Ensure Plus within a 30-minute period (test meal) at study visits one and two (pre- and post-intervention). Blood was sampled at multiple time points [2 pre-prandial fasting draws (−10 and 0 minutes)] and 15, 30, 60, 90, 120, 180, and 240 minutes post meal) for measurement of PYY, ghrelin, and GLP-1 hormones.

Stool Sample Collection and DNA extraction

Stool samples were collected at home the day before the clinic visit at baseline (4–6 weeks prior to surgical or medical weight loss intervention), and at final data collection (6 months or at 12 kg weight loss, whichever occurred first). Patients were provided with verbal instructions for self-collection in a sterile stool collection kit to be used at home. Patients then transferred 1–2 g of stool to another sterile container that was refrigerated overnight. At the clinic visit, specimens were handed off to research staff where they were transported on dry ice to a −80°C freezer and stored until deoxyribonucleic acid (DNA) extraction, sequencing, and analysis. DNA from fecal samples was isolated using the QIAGEN DNeasy PowerSoil Pro Kit, according to manufacturer’s protocol. Isolated DNA was quantified by Qubit. DNA libraries were prepared using the Illumina Nextera XT library preparation kit, with a modified protocol. Library quantity was assessed with Qubit (ThermoFisher). Libraries were sequenced on an Illumina HiSeq platform 2 × 150bp.

Gut Microbiome Sequencing and Metagenomic Analysis

Unassembled metagenomic sequencing reads were directly analyzed using the CosmosID bioinformatics software package (CosmosID Inc., Rockville, MD, United States; www.cosmosid.com), as previously described (Hourigan et al., 2018). Filtered reads were mapped to custom microbial genomic databases, and taxonomic identification was assigned with an in-house K-mer based algorithm refined against a whole genome phylogenetic tree that is described in detail in (Hourigan et al., 2018). Bacterial classification was performed to the taxonomic level of species. Microbiome BIOM files were used for downstream ordinations and statistical tests, implemented within the software packages QIIME2 (Caporaso et al., 2010), MicrobiomeAnalyst (Chong et al., 2020), and “DESeq2” packages using R (v. 3.5.1; R Foundation for Statistical Computing, Vienna, Austria). Microbial diversity of individual samples was evaluated by the alpha diversity indices Chao1 and Shannon. To evaluate the compositional difference between pre- and post-intervention (surgical or medical weight loss), beta diversity was calculated using Bray-Curtis dissimilarity, and relationships between groups pre- and post-intervention were tested using analysis of similarities (ANOSIM). Differential relative abundance (RA) of specific taxa between pre-post intervention, stratified by group (VSG, RYGB, MWL) was calculated using the parametric DESeq2 test on raw species and genus level counts (Love et al., 2014).

Resting State Functional MRI

The fMRI paradigm included 2 scans: one in a fasted state and the second after consuming the test meal (fed state). Both rsfMRIs were acquired using a 3.0 Tesla Philips Achieva Multix X-Series scanner (Phillips HealthCare, Best, Netherlands) with a multi-element 32 channel receiver head coil at the Kirby Center for Brain Imaging at the Kennedy Krieger Institute in Baltimore. 140 resting-state volumes were registered during the T2*-weighed echo-planar imaging sequence (TR = 3000 ms; TE = 30 ms; flip angle: 80°; slice thickness = 3.3 mm; slice spacing = 0 mm; slices number = 48).

fALFF Calculation

Similar to others (Li et al., 2018; Wiemerslage et al., 2017; Zeighami et al., 2021), we used a voxel-level metric fALFF to normalize ALFF, as it has been found to be more sensitive to physiological noise (Yu-Feng et al., 2007). The fALFF calculations were carried out using the CONN toolbox. The signal was linearly detrended to correct for signal drift. A fast Fourier transformation was performed and fALFF was calculated by dividing the low-frequency range (0.01–0.1 Hz) by total power (0–0.25 Hz) (Zou et al., 2008). Normalized fALFF maps were imported in SPM12, and ANOVA statistical tests were used to test for intervention effect (i.e., pre vs. post), fed condition (i.e., pre- vs. post-prandial), and their interaction. Voxels surviving an uncorrected p < 0.01 threshold were further corrected by a family wise error p < 0.05 threshold at the cluster level. Paired t-tests investigated group differences between pre- vs. post-intervention fALFF values and group differences pre- vs. post-feeding. Regions of interest were created in Marsbar as 6-mm spheres and fALFF values were extracted.

Combined Microbiome and MRI Analysis

To study the associations between gut microbiome and functional connectivity of brain regions relevant for executive functioning and interoceptive awareness, we first analyzed relationships between alpha diversity and pre-/post-prandial fALFF ratios pre vs. post weight loss. A customized script in Python was written to repeat this process with RA values of selected bacterial taxa at the genus level (25 most highly abundant taxa in the combined dataset and differentially abundant taxa from DESeq2 analysis). In this study, we sought to identify feature combinations that could characterize phenotype before and after weight loss surgery. Dimensionality reduction was performed using principal component (PC) analysis, and the MRI and microbiome datasets were reduced to five PCs. Ward hierarchical clustering was used to evaluate segmentation into pre-post intervention groups with microbiome PCs (only), MRI PCs (only), microbiome and MRI PCs, and a combined dataset with the microbiome PCs, the MRI PCs and average hormone levels at 60/90 minutes. To confirm the clustering results and test phenotype difference between pre- and post-weight loss intervention, Bray-Curtis dissimilarity was calculated and group differences tested using ANOSIM.

Statistical Analysis

Student’s t-tests were used for normally distributed variables, and Wilcoxon rank-sum calculations for non-normally distributed variables. Repeated measures ANOVAs (intervention status [pre-post] × time) were used with Sidaks’s post-hoc correction. Tests were run on raw scores for normally distributed variables and log-transformed scores when the data were not normally distributed. All statistical analyses were performed in SPSS, Python, and R software. Results are presented in the text as mean ± standard error of the mean (SEM).

Results

Weight Loss and Changes in Appetitive Ratings After Surgery

The majority of patients were female (87.5%) and their average age was 42.1 ± 8.2 years (Table 1). There was a group effect on BMI change (F_{2, 19} = 8.08, p = 0.003), weight loss percent change (F_{2, 19} = 15.48, p = 0.0001), and excess BMI loss (EBMIL; F_{2, 19} = 8.23, p = 0.002) post-surgical or medical weight loss intervention (Table 1). We found significant differences in baseline hunger ratings in RYGB and VSG, but not the MWL group (Table 1). Baseline pre-prandial hunger ratings decreased 75% in RYGB (from 42.9 ± 10.2 to 10.8 ± 5.0) and 50% in VSG (from 43 ± 9.8 to 21.3 ± 5.6) groups.

Table 1.

Patient Characteristics.

		RYGB (n = 9)		VSG (n = 7)		MWL (n = 6)
Age (years)		39.7 ± 2.75		44.3 ± 1.9		43.2 ± 3.8
Gender
Males		—		—		3 (50%)
Females		9 (100%)		7 (100%)		3 (50%)
Height (cm)		64.4 ± 0.97		64 ± 1.05		67 ± 2.05
Follow-up Time
Days post-intervention		80.9 ± 6.6		97.5 ± 16.3		201 ± 24
Days post visit 1		94.8 ± 6.5		110.7 ± 18.1		202.5 ± 27.1
Weight (kg)
Pre	Post	120.4 ± 3.85	101.4 ± 3.6*	113.2 ± 7.3	96.8 ± 5.9*	116.5 ± 7.0	112.7 ± 9.0
BMI (kg/m ² )
Pre	Post	46.4 ± 1.8	39.6 ± 1.5*	42.5 ± 1.7	36.3 ± 1.4*	40.3 ± 2.5	38.8 ± 2.6
Weight % Change		−15.8 ± 1.14		−14.4 ± 1.26		−3.8 ± 2.5
BMI Change (pre-post)		8.57 ± 1.58		5.92 ± 0.60		1.53 ± 0.83
EMBIL %		39.6 ± 4.65		35.0 ± 3.42		11.4 ± 7.23
Race/Ethnicity
Non-Hispanic Black		4 (44%)		3 (43%)		4 (67%)
Non-Hispanic White		4 (44%)		4 (57%)		2 (33%)
Hispanic		1 (11%)		—		—
Comorbidities
Type II DM		3 (33%)		2(29%)		0
Anxiety		3 (33%)		0		0
Depression		2(22%)		0		0
Hunger ratings (Visual Analog Scale)
Fasted state
Pre	Post	42.9 ± 10.2	10.8 ± 5.0*	43 ± 9.8	21.3 ± 5.6*	54.2 ± 14.5	42.3 ± 10.9
Fed state
Pre	Post	49.4 ± 11.4	2.3 ± 1.7*	31.3 ± 8.0	19.7 ± 11.3*	32.3 ± 8.9	36.7 ± 16.3

Note. *states statistical significance at p < 0.05 on *pre-intervention versus post-intervention within-group comparisons. BMI: Body mass index; DM: Diabetes mellitus; EMBIL: Percent excess BMI lost; MWL: Medical weight loss; RYGB: Roux-en-Y gastric bypass; VSG: Vertical sleeve gastrectomy.

Participants rated hunger and fullness using a visual analog scale scale of 0 to 100 mm, for example, “How hungry are you right now?” and anchored on the left by “not at all” and on the right by “extremely.” Results presented as mean ± SEM.

Resting-State fMRI Analysis

A main effect of weight loss intervention was found on fALFF in the right caudate, anterior insula, bilateral cuneus and left superior orbital gyrus, occipital regions (calcarine cortex, lingual gyrus), cerebellum, and cerebellar vermal lobules VIII-X (P_FWE-corr < 0.03). An interaction effect between prandial state and weight loss intervention was found on the right supramarginal gyrus and middle temporal gyrus (MTG; P_FWE-corr < 0.01; Supplemental Table 1). Fasted state scans revealed increased fALFF after RYGB, relative to pre, in the right posterior cingulate cortex and left middle cingulate gyrus (p < 0.05; Figure 2a, Table 2). Areas with decreased fALFF after RYGB included the right superior frontal gyrus, postcentral gyrus, precentral gyrus, left middle temporal gyrus, and inferior temporal gyrus in the fed state (p < 0.05; Figure 2b, Table 2). In the fasted state, fALFF increased prior to VSG in right cerebellum, inferior occipital gyrus, superior occipital gyrus, and in bilateral lingual gyrus (Figure 2c), and fALFF increased after VSG in left angular gyrus, supramarginal gyrus, and superior parietal lobule (Figure 2d). In the fed state, increased fALFF was seen in default mode network regions (bilateral posterior cingulate cortex), dorsal striatum (right caudate; putamen) as well as right anterior cingulate gyrus and lingual gyrus following VSG (Figure 2e). Fasted state fALFF was increased prior to intervention in right superior frontal gyrus medial segment and anterior cingulate gyrus in the MWL group (Figure 2f, Table 2).

Table 2.

fALFF Differences Pre- and Post-Interventions in Different Groups.

Groups Region	Cluster size (k)	Peak voxel (Z)	Cluster FWE p-value	MNI coordinate
				X	Y	Z
RYGB fast - post>pre
Posterior cingulate cortex_L	1394	4.25	0.01	−10	−34	26
Middle cingulate gyrus_L		4.02		−8	−22	38
Precuneus_L		3.56		−4	−42	58
RYGB fed - pre>post
Postcentral gyrus _R	1530	4.37	0.01	58	−8	38
Precentral gyrus_R		4.03		54	−24	12
Central Operculum_R		3.73		64	−14	14
Middle temporal gyrus_L	1331	3.71	0.01	−64	−38	0
Posterior Insula_L		3.63		−44	−20	−2
Angular gyrus_L		3.56		−40	−52	36
pre
VSG fast - pre>post
Occipital fusiform gyrus_R	454	3.33	0.004	30	−78	0
Inferior occipital gyrus_R		3.32		20	−82	−8
Lingual gyrus_R		3.27		30	−70	−10
VSG fast - post>pre
Precuneus_L	331	4.14	0.01	−22	−50	16
Cuneus_L		3.22		−16	−70	14
Lingual gyrus_L		3.18		−26	−66	10
MWL fast - pre>post
Anterior cingulate gyrus_L	639	3.67	0.010	−8	46	2
Superior frontal gyrus medial segment_R		3.25		18	56	−16
Groups region	Cluster size (k)	Peak voxel (Z)	Cluster FWE p -value	MNI coordinate
				X	Y	Z
RYGB fast - post>pre
Posterior cingulate cortex_R	1170	3.58	0.000	14	−42	30
Middle cingulate gyrus_L		3.42		−12	−26	38
RYGB fed - pre>post
Superior frontal gyrus_R	449	4.23	0.003	18	−10	50
Postcentral gyrus _R		4.19		18	−22	52
Precentral gyrus_R		3.34		28	−10	50
Middle temporal gyrus_L	483	3.87	0.003	−64	−22	−18
Inferior temporal gyrus_L		3.59		−52	−52	−14
VSG fast - pre>post
Cerebellum_R	546	3.23	0.003	26	−66	−18
Lingual gyrus_R		3.04		8	−72	−2
Lingual gyrus_L		2.99		−12	−72	−6
Inferior occipital gyrus_R	897	3.01	0.000	38	−84	4
Superior occipital gyrus_R		2.80		16	−88	16
VSG fast - post>pre
Angular gyrus_L	788	2.84	0.000	−40	−80	28
Superior parietal lobule_L		2.73		−36	−62	58
Supramarginal gyrus_L		2.71	0.001	−42	−56	42
VSG fed - pre>post
Caudate_R	1133	4.03	0.000	6	12	−2
Putamen_R		3.79		22	24	2
Anterior cingulate gyrus_R		3.68		16	40	6
Posterior cingulate gyrus_L	554	3.87	0.003	−6	−42	16
Posterior cingulate gyrus_R		3.33		8	−44	18
Lingual gyrus_R		3.39		14	−46	−8
MWL fast - pre>post
Superior frontal gyrus medial segment _R	464	3.60	0.010	14	50	8
Anterior cingulate gyrus _R		3.59		0	48	4

Note. MNI, coordinates referring to the standard brain of the Montreal Neurological Institute; FWE, family wise error; L, left hemisphere; R, right hemisphere. Clusters of maximally activated voxels that survived statistical thresholding (p < 0.01, uncorrected, extent threshold of >50 voxels).

Cognitive Tests

HVLT 1 results were influenced by intervention group (group F_{2, 37} = 4.96, p = 0.012) and pre- vs. post (time F_{2, 37} = 13.82, p = 0.001), but the interaction was not significant (group × time F_{2, 37} = 2.61, p = 0.087). The MWL group had a significant increase in HVLT1, but scores were not different pre- vs. post-intervention in the VSG and RYGB groups (Supplemental Table 2).

Gut Hormones

Fasting ghrelin levels were significantly lower post-VSG (Figure 3). There was an overall pre-post RYGB group but not time effect for ghrelin responses (Figure 3). Group and time responses did not differ in ghrelin levels in the MWL group (Figure 3; Supplemental Table 3). PYY levels had a significantly higher and more robust response to food intake post VSG, while overall PYY levels were influenced pre-post RYGB (Figure 3). Similar to ghrelin, there was not an overall group or time effect on PYY levels in the MWL group (Figure 3). Post-prandial GLP-1 levels were influenced by VSG and RYGB (Figure 3). There was an overall group but not time effect for GLP-1 levels pre- vs. post MWL intervention (Figure 3).

Gut Microbiome

Analyzing alpha diversity, there was an overall weight loss intervention group effect (i.e., RYGB vs. VSG vs. MWL; group F_{2, 40} = 5.79, p = 0.006), but not pre-post-intervention effect (time F_{1, 40} = 0.52, p = 0.475) for differences in the Shannon index. Interestingly, there was not a group effect (group F_{2, 40} = 1.86, p = 0.170), but there was a pre-post-intervention effect (time F_{1, 40} = 11.87, p = 0.001) for the Chao1 index (Supplemental Figure 1). Microbial beta diversity analyses did not indicate group (VSG, RYGB, MWL) effects on the observed microbial community structure pre-intervention (R = 0.034, p = 0.275). Post-intervention, there was a small but significant difference between intervention groups (R = 0.0123, p < 0.05) on overall microbial community structure (Supplemental Figure 1). At the level of genus, 30 taxa in the RYGB group, six taxa in the VSG group and three taxa in the MWL group were differentially abundant pre- vs. post-weight loss intervention (Table 3). At the taxonomic level of species, we identified 65 taxa in the RYGB group, 36 taxa in the VSG group, and 17 taxa in the MWL group as significantly different in RA pre- vs. post-weight loss intervention (Supplemental Table 4).

Table 3.

Bacterial Taxa Differences Pre- and Post-Intervention at Taxonomic Level of Genus.

Bacteria	Abundance, %
Genus	Pre-	Post-	Effect (vs. Pre-)	FDR p-value
RYGB
Alloscardovia	0.0 ± 0.0	0.00009 ± 0.0001	26.43 ± 2.39	<0.0001
Campylobacter	0.0 ± 0.0	0.00012 ± 0.0002	26.15 ± 2.73	<0.0001
Imtechella	0.00018 ± 0.0003	0.0 ± 0.0	−26.74 ± 3.01	<0.0001
Romboutsia	0.0 ± 0.0	0.00009 ± 0.0002	26.41 ± 2.99	<0.0001
Leptotrichia	0.0 ± 0.0	0.00007 ± 0.0001	26.22 ± 2.99	<0.0001
Paeniclostridium	0.0 ± 0.0	0.00006 ± 0.0001	25.84 ± 2.99	<0.0001
Eikenella	0.0 ± 0.0	0.00015 ± 0.0002	25.72 ± 2.99	<0.0001
Raoultella	0.0 ± 0.0	0.00020 ± 0.0005	23.99 ± 2.99	<0.0001
Shuttleworthia	0.0 ± 0.0	0.00002 ± 0.0000	23.07 ± 2.99	<0.0001
Neisseria	0.0 ± 0.0	0.00023 ± 0.0006	22.48 ± 2.99	<0.0001
Fretibacterium	0.0 ± 0.0	0.00010 ± 0.0002	22.26 ± 2.99	<0.0001
Negativicoccus	0.0 ± 0.0	0.00031 ± 0.0004	13.49 ± 1.86	<0.0001
Cryptobacterium	0.0 ± 0.0	0.00012 ± 0.0003	21.17 ± 2.99	<0.0001
Selenomonas	0.0 ± 0.0	0.00037 ± 0.0005	11.23 ± 2.21	<0.0001
Veillonella	0.00028 ± 0.0004	0.02325 ± 0.0351	6.14 ± 1.32	<0.0001
Fusobacterium	0.0 ± 0.0	0.00032 ± 0.0005	7.33 ± 1.89	0.0011
Cardiobacterium	0.0 ± 0.0	0.00009 ± 0.0001	11.12 ± 2.97	0.0018
Actinomyces	0.00287 ± 0.0028	0.02339 ± 0.0262	2.84 ± 0.79	0.0029
Megamonas	0.00003 ± 0.0000	0.00001 ± 0.0000	−1.94 ± 0.57	0.0059
Streptococcus	0.00354 ± 0.0055	0.03191 ± 0.0412	3.07 ± 0.92	0.0078
Granulicatella	0.00020 ± 0.0002	0.00161 ± 0.0021	2.76 ± 0.88	0.0142
Citrobacter	0.0 ± 0.0	0.00116 ± 0.0025	7.86 ± 2.53	0.0151
Escherichia	0.00002 ± 0.0000	0.0019 ± 0.0001	3.03 ±1.03	0.0242
Rothia	0.00047 ± 0.0007	0.00294 ± 0.0025	2.60 ± 0.92	0.0324
Clostridiales Family XIII. Incertae Sedis_u_g	0.00029 ± 0.0002	0.00233 ± 0.0038	2.74 ± 0.99	0.0373
Klebsiella	0.00003 ± 0.0001	0.00547 ± 0.0138	4.08 ± 1.48	0.0387
Haemophilus	0.00018 ± 0.0005	0.00867 ± 0.0128	5.50 ± 2.03	0.0427
Atopobium	0.00023 ± 0.0003	0.00461 ± 0.0097	4.09 ± 1.56	0.0497
Oribacterium	0.00001 ± 0.0000	0.00033 ± 0.0005	4.27 ± 1.61	0.0497
Staphylococcus	0.00002 ± 0.0001	0.0 ± 0.0	−5.47 ± 2.08	0.0497
VSG
Levyella	0.00006 ± 0.0002	0.00001 ± 0.0000	−22.42 ± 2.97	<0.0001
Paeniclostridium	0.00001 ± 0.0000	0.0 ± 0.0	−22.14 ± 2.97	<0.0001
Cryptobacterium	0.0 ± 0.0	0.00001 ± 0.0000	19.74 ± 2.97	<0.0001
Streptococcus	0.00090 ± 0.0008	0.00903 ± 0.0073	3.27 ± 0.63	<0.0001
Rothia	0.00006 ± 0.0001	0.00097 ± 0.0008	3.93 ± 0.85	0.0001
Veillonella	0.00005 ± 0.0001	0.00055 ± 0.0013	3.05 ± 0.89	0.0176
MWL
Aggregatibacter	0.00002 ± 0.0001	0.0 ± 0.0	−22.52 ± 3.01	<0.0001
Paeniclostridium	0.0 ± 0.0	0.00001 ± 0.0000	22.32 ± 3.00	<0.0001
Synergistes	0.0 ± 0.0	0.00001 ± 0.0000	8.14 ± 2.23	0.0130

Note. Abundance values are expressed in relative percent taxa abundance of gut microbiota sample (means ± SD). Effect is log 2-fold change ± SD. FDR p-value is false discovery rate-corrected p value. RYBG: Roux-en-Y gastric bypass; VSG: Vertical sleeve gastrectomy; MWL = medical weight loss.

Correlation between MRI and Gut Microbiome

One overarching aim of the study was to detect features that drive microbiome-MRI function connectivity relationships pre-post bariatric surgery. In subjects who underwent VSG, we found moderate negative associations between the Chao1 index and fALFF amplitudes of several brain regions involved in executive functioning and interoceptive awareness in the fed state. Chao1 index was negatively correlated with the right putamen (r = −0.594, p = 0.025) and right lingual gyrus (r = −0.597, p = 0.024). Chao1 also had moderate, non-significant correlations with the right anterior cingulate gyrus (r = −0.521, p = 0.056) and left posterior cingulate cortex (r = −0.512, p = 0.061). Subjects in the MWL group, showed a positive association between Shannon alpha diversity and fALFF amplitudes of the executive functioning regions in the fasting state, including the right superior frontal gyrus (r = 0.655, p = 0.021) and right superior frontal gyrus medial segment (r = 0.624, p = 0.030).

In the VSG group, there were associations between the RA of bacterial taxa (genus-level) and increased fALFF in occipital networks pre-VSG and that of default mode and sensorimotor networks post-VSG in the fasting state (Supplemental Table 5). Visual networks that had higher activation pre-VSG, compared to post, were negatively correlated with several anaerobic bacteria including Roseburia (r = −0.764, p = 0.001), and the RA of Veillonella genera was also positively associated with the fALFF in the superior parietal lobule (r = 0.612, p = 0.020 and r = 0.554, p = 0.040, respectively). All bacteria negatively associated with increased fALFF in angular gyrus and superior parietal lobule were anaerobic, gram-positive bacteria. Eubacterium spp. RA were positively associated with increased fALFF in brain regions related to motivation and cognitive control in the fed state. Conversely, Rothia spp. RA was negatively associated with increased fALFF in several brain regions. Five bacterial taxa had shared negative correlations between RA and activation during the fed state in both the VSG and RYGB surgery groups, while the MWL group had the smallest number of brain-microbiome correlations (Supplemental Table 5).

Phenotype Evaluation with Combined MRI, Microbiome and Hormone Data

Our final aim was to test if MRI, microbiome, hormone, or a combination of the three data types accurately differentiated pre-post weight loss phenotypes. In the VSG group, the combined microbiome, MRI and hormone dataset accurately clustered samples into pre- and post-intervention groups, and the grouped features were significantly different pre- vs. post-VSG surgery (Figure 4; R = 0.597, p = 0.002). The other datasets did not cluster pre- and post-VSG samples accurately, but the hormone only, MRI only, and MRI/microbiome datasets detected feature differences pre-post-VSG surgery (Supplemental Figure 2A). In the RYGB group, no datasets accurately clustered all subjects to pre- vs. post-RYGB intervention (Supplemental Figure 2B). Nevertheless, we found significant differences between the pre-post RYGB intervention groups in all of the datasets tested (Supplemental Figure 2C).

Discussion

In this pilot study, we evaluated the impact of pre-vs. post-VSG, RYGB, and MWL interventions on the microbiota-gut-brain axis by quantifying relationships between the gut microbiome, gut hormones, and rsfMRI. To the best of our knowledge, the present study is the first to demonstrate significant relationships between gut microbiome and rsfMRI features in subjects pre- vs. post-surgical and medical/diet induced weight loss. We observed a significant reduction in BMI and fasting hunger ratings in the surgical groups pre- vs. post-intervention that did not occur in the MWL group. A significant reduction in fasting and fed hunger ratings post-RYGB and VSG were likely associated with a reduction in the gastric volume that did not occur in the MWL group (∼30 mL and ∼120 mL in RYGB and VSG, respectively; Elder & Wolfe, 2007).

We observed a main effect as well as an interaction effect of intervention (pre- vs. post-RYGB, VSG, or MWL) and prandial states on brain resting-state activity. Several of the brain areas described have been associated with obesity or food signaling by various groups (Nota et al., 2020). For example, the fALFF decreases following RYGB intervention in the fed condition were seen in brain regions implicated in food reward (including right superior frontal gyrus, postcentral gyrus, precentral gyrus), which suggests bariatric surgery leads to lesser post-prandial activation of neural circuits underlying food approach. The cerebellum regulates feeding behavior, that is, it integrates and coordinates somatic–visceral responses to food and its decreased activation is reported in individuals with obesity in response to satiation (Gautier et al., 2000).

Pre-VSG increased sated fALFF was seen in the cingulate gyrus and striatal (caudate, putamen) and lingual gyrus. Increased fasting fALFF activation was observed in the attention and memory brain regions (left angular gyrus, superior parietal lobule, supramarginal gyrus) post-VSG. Our findings indicated that even if the sample size was similar across the groups, there were more fALFF differences in surgery versus MWL, which suggests a bigger impact of surgical intervention in brain functioning than non-surgical weight loss. Furthermore, in both surgical weight loss groups (VSG and RYGB), we observed post-prandial decreases in ghrelin and increases in PYY and GLP-1. In subjects with obesity, the post-prandial increase of circulating plasma PYY and GLP-1 and suppression of ghrelin levels is diminished (Glicksman et al., 2010), which started to reverse after both VSG and RYGB. These post-surgical hormonal changes knowingly contribute to suppression of hunger and enhancement in satiety signals in the brain, thereby improving the efficacy of these procedures (Ochner et al., 2011).

Surgical (VSG and RYGB) and dietary (MWL) weight loss significantly influenced the composition of the gut microbiome, although the global microbiome changes (alpha and beta diversity metrics) were subtle. When evaluating the Chao1 index, a marker of species richness, the number of species increased post-intervention in all groups. This increase in the Chao1 index supports the differential RA results at the species level. When we compared beta diversity across the VSG, RYGB, and MWL groups pre-weight loss, there were no significant community differences, which demonstrates global microbial communities were similar in all groups before weight loss began. Post-intervention, there was group separation observed on NMDS plots, with the RYGB group demonstrating greatest separation in 3-dimensional space. The post-intervention Chao 1 increase and subtle beta diversity changes suggest a measurable impact of weight loss on the gut microbiome, but global changes occur more slowly than changes in individual taxa.

Species-level bacteria with the greater difference pre-post intervention have functional potential relevant to weight loss and gut health. For example, in the RYGB group, Alloscardovia omnicolens was not quantified pre-RYGB surgery and had one of the highest increases post-intervention. Alloscardovia omnicolens is closely related to the Bifidobacterium genus (Mahlen & Clarridge, 2009), which are the most prevalent genera in the healthy gastrointestinal microbiota and reduction in Bifidobacterium spp. RA has been reported in overweight individuals (Eckburg et al., 2005). Other bacterial species that increased after surgical weight loss are associated with sustained weight loss or improved glucose homeostasis include Veillonella parvula (Graessler et al., 2013), Bifidobacterium bifidum (Mahadzir et al., 2017), Streptococcus thermophiles (De Lorenzo et al., 2017), and Fusobacterium spp. (De Lorenzo et al., 2017).

We found that many bacteria that significantly increased after weight loss were predominant taxa in the oral microbiome (Wilbert et al., 2020). For example, several Streptococcus, Veillonella, Prevotella, and Rothia genera were significantly increased post-weight loss (compared to pre-), with the greatest number of species classified under oral-associated genera increasing in the surgical weight loss groups. Recent research has suggested that passive transmission of oral microorganisms to the gut microbiome through saliva is more common than previously recognized (Maki et al., 2021; Schmidt et al., 2019). Our observation of increased oral-associated bacteria in the gut microbiome after surgical weight loss is consistent with other research evaluating gut microbiome changes in post bariatric surgery subjects (Ilhan et al., 2017; Palleja et al., 2016). Changes in stomach physiology and the increased gastric pH associated with bariatric surgery likely expose translocated oral bacteria to reduced levels of gastric acid and allow passage to the large intestine undisturbed (Porat et al., 2021). Streptococcus, Veillonella, and Prevotella genera are related to SCFAs (as acetate, butyrate and propionate) fermentation byproducts (Dai et al., 2011), which may contribute to weight loss by modulation of gut hormones and food intake (Lin et al., 2012). Future combined microbiome and metabolome research analyzing whether increased RA of oral-associated bacteria in the gut microbiome actively contributes to weight loss via metabolite production and signaling mechanisms will be useful.

There were several associations between fALFF amplitudes in brain regions of interest and both alpha diversity and specific bacteria that were mirrored by a clinically relevant change in RA pre- and post-bariatric surgery. Previous investigators have also reported positive association between alpha diversity of the gut microbiome and fear processing/cognitive development circuits, demonstrating a role of the microbiome in neurologic signaling (Gao et al., 2019). At the genus level, Rothia and Veillonella were positively associated with fasting fALFF values in the superior parietal lobule involved in executive functioning post-VSG. The RA of these 2 bacterial genera also increased after bariatric surgery; Rothia spp. RA increased post-surgery in VSG subjects, and Veillonella spp. RA increased after both VSG and RYGB surgery. The RA of Ruminococcus and Levyella genera were negatively associated with executive network activation in VSG patients, and both significantly decreased post VSG surgery. In other studies of the microbiome and obesity, both Ruminococcus bromii and Ruminococcus obeum species were significantly more abundant in the stool of obese subjects (Kasai et al., 2015), suggesting Ruminococcus spp. is a microbial biomarker of interest in the context of obesity.

Our final aim was to determine which features were important to distinguish pre- vs. post-weight loss intervention phenotypes in surgical and MWL groups. The clinical significance of identifying if brain, gut microbiome, and hormone data are required to differentiate patients pre- vs. post-surgical and MWL is to inform future research where larger sample sizes, inclusion of covariates, and supervised machine learning can be employed. We also believe that the physiologic and functional implications of the individual features identified in this exploratory research (i.e., brain connectivity difference in regions of interest or change of microbial taxa differentially abundant pre- vs. post-intervention) can be used to generate hypotheses for future protocols studying the microbiota-gut-brain axis in obesity. The VSG group was the only group showing completely accurate clustering into pre- and post-phenotypes. VSG patients also had the greatest reduction in BMI pre- vs. post-intervention, compared to RYGB and MWL, which may explain the strongest phenotype clustering. However, RYGB patients had the greatest overall weight loss. Importantly, based on BMI-based weight categories according to NIH guidelines (National Heart Lung and Blood Institute, nd.), all subjects remained obese, but moved from an average obesity class 3 to obesity class 2. Longer follow up periods with sustained weight loss and larger sample sizes may demonstrate greater microbiota-gut-brain axis differences with combined MRI, microbiome and hormone features, and elucidate those features changing most dramatically after weight loss.

Given that this was an exploratory study, several limitations should be considered when interpreting results. First, our sample size was small with multiple conditions considered, including pre- vs. post- VSG, RYGB, and MWL groups and pre- vs. post-prandial states. This prevented use of more sophisticated hypothesis testing. Nevertheless, despite the small sample size, it was possible to detect important microbial and neuronal features that changed post weight loss that will inform future research. Additionally, a subset of the RYGB group had diagnosed anxiety and depression, which was not present in the patients belonging to the VSG and MWL groups. As both anxiety and depression are known to impact brain connectivity and the gut microbiome through the gut-brain axis (Cryan & Dinan, 2012), the presence of these conditions will be an important covariate in future research. A large proportion of the cohort identified as females, however, this is consistent with the population as females classify about 80% of individuals who opt for bariatric surgery. Future research with balanced male and female groups will inform sex differences in microbiota-gut-brain signaling in obesity. Factors such as patient diet, stress levels, and lifestyle factors such as exercise are known to be important modulators of the gut microbiome and were not quantified during the pre- or post-intervention period. These factors should be measured and used as covariates in future studies analyzing weight loss and the gut microbiome. We included hormone assessment in our phenotype evaluation, which levels may be confounded by patient-specific factors such as menstrual cycle or diagnosis of diabetes mellitus. As this was a pilot study, we were not powered to control these factors, but future work with larger sample sizes should consider including them in the statistical models.

Conclusion

This exploratory study contributes to the current literature providing new insight into the possible mechanisms of microbiota-gut-brain axis signaling following surgical (VSG and RYGB) and medical weight loss. Utilizing a systems biology approach including rsfMRI, microbiome, and gut hormone data, and their interrelationships, our analyses suggest that surgical weight loss has a greater impact on brain and gut function pre- vs. post-intervention when compared to MWL, leading to a decrease in post-prandial activation of neural circuits underlying food reward. We demonstrated that several bacterial taxa of the gut microbiota are both associated with rsfMRI activation and differentially abundant pre- vs. post-bariatric surgery. Further, our exploratory study allowed us to analyze individual and combined datasets to determine pre- vs. post-weight loss clustering and highlighted the importance of including MRI and the microbiome in the analyses of the microbiota-gut-brain axis, as we unravel the underlying mechanisms involved in obesity. Our findings establish a strong foundation to encourage ongoing and future studies to include microbial and MRI features to improve the understanding of the microbiota-gut-brain signaling mechanisms and their alterations in subjects before and after surgical and medical weight loss interventions.

Supplemental Material

Supplemental Material - The Neuro-Endo-Microbio-Ome Study: A Pilot Study of Neurobiological Alterations Pre- Versus Post-Bariatric Surgery

Supplemental Material for The Neuro-Endo-Microbio-Ome Study: A Pilot Study of Neurobiological Alterations Pre- Versus Post-Bariatric Surgery by Khushbu Agarwal, PhD, Katherine A. Maki, PhD, Carlotta Vizioli, PhD, Susan Carnell, PhD, Ethan Goodman, PhD, Matthew Hurley, PhD, Civonnia Harris, BSc, Rita Colwell, PhD, Kimberley Steele, MD, PhD, and Paule V. Joseph, PhD in Biological Research For Nursing

Supplemental Material

Supplemental Material - The Neuro-Endo-Microbio-Ome Study: A Pilot Study of Neurobiological Alterations Pre- Versus Post-Bariatric Surgery

Supplemental Material

Supplemental Material - The Neuro-Endo-Microbio-Ome Study: A Pilot Study of Neurobiological Alterations Pre- Versus Post-Bariatric Surgery

Supplemental Material

Supplemental Material - The Neuro-Endo-Microbio-Ome Study: A Pilot Study of Neurobiological Alterations Pre- Versus Post-Bariatric Surgery

Supplemental Material

Supplemental Material - The Neuro-Endo-Microbio-Ome Study: A Pilot Study of Neurobiological Alterations Pre- Versus Post-Bariatric Surgery

Supplemental Material

Supplemental Material - The Neuro-Endo-Microbio-Ome Study: A Pilot Study of Neurobiological Alterations Pre- Versus Post-Bariatric Surgery

Supplemental Material

Supplemental Material - The Neuro-Endo-Microbio-Ome Study: A Pilot Study of Neurobiological Alterations Pre- Versus Post-Bariatric Surgery

Footnotes

Acknowledgments

The authors thank Timothy Moran, for mentorship. Larry Cheskin for assistance with MWL recruitment. Eric Stice for providing inspiration. Dorry Segev for your incredible support. Susan Oh, research nutritionists and technicians preparing “meal tests”. Research nursing for phlebotomy. James Pekar, Terri Brawner, Kathleen Kahl, Ivana Kusevic for assisting with fMRI paradigms. Brian Fanelli, Karlis Graubics for microbiome methodology and analyses. A special thank you to the ASMBS Research Committee for their compassion and understanding, for not wavering in support of the PI to see the project through to successful completion. And, importantly, our patients who participated and volunteered their time.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was supported by The American Society of Metabolic and Bariatric Surgeons, Research Grant Award, to KES (supported recruitment, gut hormone measurement, gut microbiome sequencing and fMRI imaging). PVJ is supported by the Division of Intramural Research National Institute on Alcohol Abuse and Alcoholism and Institute of Nursing Research and the Office of Workforce Diversity, National Institutes of Health Distinguished Scholar, and the Rockefeller University Heilbrunn Nurse Scholar Award. KA received Intramural Research Training Awards, National Institute of Nursing Research, National Institutes of Health, Department of Health and Human Services. KM received Intramural Research Training Awards, Clinical Center, National Institutes of Health, Department of Health and Human Services.

Authors’ Note

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Subject enrollment occurred at the Johns Hopkins Center for Bariatric Surgery, and Institutional Review Board approval was through the John Hopkins University School of Medicine.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author, [PVJ and KS], upon reasonable request.

ORCID iDs

Katherine A. Maki

Carlotta Vizioli

Supplemental Material

Supplemental material for this article is available online.

References

Bassett

S. A.

Young

Fraser

Dalziel

J. E.

Webster

Ryan

Fitzgerald

Stanton

Dinan

T. G.

Cryan

J. F.

Clarke

Hyland

Roy

N. C.

(2019). Metabolome and microbiome profiling of a stress-sensitive rat model of gut-brain axis dysfunction. Scientific Reports, 9(1), 14026. https://doi.org/10.1038/s41598-019-50593-3

Biswal

Yetkin

F. Z.

Haughton

V. M.

Hyde

J. S.

(1995). Functional connectivity in the motor cortex of resting human brain using echo-planar MRI. Magnetic Resonance in Medicine, 34(4), 537–541. https://doi.org/10.1002/mrm.1910340409

Bliss

E. S.

Whiteside

(2018). The gut-brain axis, the human gut microbiota and their integration in the development of obesity. Frontiers in Physiology, 9, 900. https://doi.org/10.3389/fphys.2018.00900

Bogdanov

V. B.

Bogdanova

O. V.

Dexpert

Delgado

Beyer

Aubert

Dilharreguy

Beau

Forestier

Ledaguenel

Magne

Aouizerate

Layé

Ferreira

Felger

Pagnoni

Capuron

(2020). Reward-related brain activity and behavior are associated with peripheral ghrelin levels in obesity. Psychoneuroendocrinology, 112, 104520. https://doi.org/10.1016/j.psyneuen.2019.104520

Caporaso

J. G.

Kuczynski

Stombaugh

Bittinger

Bushman

F. D.

Costello

E. K.

Fierer

Peña

A. G.

Goodrich

J. K.

Gordon

J. I.

Huttley

G. A.

Kelley

S. T.

Knights

Koenig

J. E.

Ley

R. E.

Lozupone

C. A.

McDonald

Muegge

B. D.

Pirrung

Knight

(2010). QIIME allows analysis of high-throughput community sequencing data. Nature Methods, 7(5), 335–336. https://doi.org/10.1038/nmeth.f.303

Chong

Liu

Zhou

Xia

(2020). Using microbiomeanalyst for comprehensive statistical, functional, and meta-analysis of microbiome data. Nature Protocols, 15(3), 799–821. https://doi.org/10.1038/s41596-019-0264-1

Covasa

Stephens

R. W.

Toderean

Cobuz

(2019). Intestinal sensing by gut microbiota: Targeting gut peptides. Frontiers in Endocrinology (Lausanne), 10, 82. https://doi.org/10.3389/fendo.2019.00082

Cryan

J. F.

Dinan

T. G.

(2012). Mind-altering microorganisms: The impact of the gut microbiota on brain and behaviour. Nature Reviews Neuroscience, 13(10), 701–712. https://doi.org/10.1038/nrn3346

Dai

Z. L.

Zhu

W. Y.

(2011). Amino acid metabolism in intestinal bacteria: links between gut ecology and host health. Frontiers in Bioscience (Landmark Ed), 16, 1768–1786. https://doi.org/10.2741/3820

10.

De Lorenzo

Costacurta

Merra

Gualtieri

Cioccoloni

Marchetti

Varvaras

Docimo

Renzo

(2017). Can psychobiotics intake modulate psychological profile and body composition of women affected by normal weight obese syndrome and obesity? A double blind randomized clinical trial [Article]. Journal of Translational Medicine, 15(1), Article 135. https://doi.org/10.1186/s12967-017-1236-2

11.

Dinan

T. G.

Cryan

J. F.

(2017). The microbiome-gut-brain axis in health and disease. Gastroenterology Clinics of North America, 46(1), 77–89. https://doi.org/10.1016/j.gtc.2016.09.007

12.

Dong

Jackson

Wang

Chen

(2015). Spontaneous regional brain activity links restrained eating to later weight gain among young women. Biological Psychology, 109, 176–183. https://doi.org/10.1016/j.biopsycho.2015.05.003

13.

D’Souza

Mollayeva

Pacheco

Javed

Colantonio

Mollayeva

(2019). Measuring change over time: A systematic review of evaluative measures of cognitive functioning in traumatic brain injury. Frontiers in Neurology, 10, 353–353. https://doi.org/10.3389/fneur.2019.00353

14.

Eckburg

P. B.

Bik

E. M.

Bernstein

C. N.

Purdom

Dethlefsen

Sargent

Gill

S. R.

Nelson

K. E.

Relman

D. A.

(2005). Diversity of the human intestinal microbial flora. Science, 308(5728), 1635–1638. https://doi.org/10.1126/science.1110591

15.

Elder

K. A.

Wolfe

B. M.

(2007). Bariatric surgery: A review of procedures and outcomes. Gastroenterology, 132(6), 2253–2271. https://doi.org/10.1053/j.gastro.2007.03.057

16.

Frühbeck

Diez-Caballero

Gil

M. J.

Montero

Gómez-Ambrosi

Salvador

Cienfuegos

J. A.

(2004). The decrease in plasma ghrelin concentrations following bariatric surgery depends on the functional integrity of the fundus. Obesity Surgery, 14(5), 606–612. https://doi.org/10.1381/096089204323093363

17.

Gao

Salzwedel

A. P.

Carlson

A. L.

Xia

Azcarate-Peril

M. A.

Styner

M. A.

Thompson

A. L.

Geng

Goldman

B. D.

Gilmore

J. H.

Knickmeyer

R. C.

(2019). Gut microbiome and brain functional connectivity in infants-a preliminary study focusing on the amygdala. Psychopharmacology, 236(5), 1641–1651. https://doi.org/10.1007/s00213-018-5161-8

18.

García-García

Jurado

Garolera

Segura

Sala-Llonch

Marqués-Iturria

Pueyo

Sender-Palacios

M. J.

Vernet-Vernet

Narberhaus

Ariza

Junqué

(2013). Alterations of the salience network in obesity: A resting-state fMRI study. Human Brain Mapping, 34(11), 2786–2797. https://doi.org/10.1002/hbm.22104

19.

Gautier

J. F.

Chen

Salbe

A. D.

Bandy

Pratley

R. E.

Heiman

Ravussin

Reiman

E. M.

Tataranni

P. A.

(2000). Differential brain responses to satiation in obese and lean men. Diabetes, 49(5), 838. https://doi.org/10.2337/diabetes.49.5.838

20.

Glicksman

Pournaras

D. J.

Wright

Roberts

Mahon

Welbourn

Sherwood

Alaghband-Zadeh

le Roux

C. W.

(2010). Postprandial plasma bile acid responses in normal weight and obese subjects. Annals of Clinical Biochemistry, 47(Pt 5), 482–484. https://doi.org/10.1258/acb.2010.010040

21.

Graessler

Qin

Zhong

Zhang

Licinio

Wong

M. L.

Chavakis

Bornstein

A. B.

Ehrhart-Bornstein

Lamounier-Zepter

Lohmann

Wolf

Bornstein

S. R.

(2013). Metagenomic sequencing of the human gut microbiome before and after bariatric surgery in obese patients with type 2 diabetes: Correlation with inflammatory and metabolic parameters. The Pharmacogenomics Journal, 13(6), 514–522. https://doi.org/10.1038/tpj.2012.43

22.

Hourigan

S. K.

Subramanian

Hasan

N. A.

Klein

Chettout

Huddleston

Deopujari

Levy

Baveja

Clemency

N. C.

Baker

R. L.

Niederhuber

J. E.

Colwell

R. R.

(2018). Comparison of infant gut and skin microbiota, resistome and virulome between neonatal intensive care unit (NICU) environments. Frontiers in Microbiology, 9, 1361. https://doi.org/10.3389/fmicb.2018.01361

23.

Ilhan

Z. E.

DiBaise

J. K.

Isern

N. G.

Hoyt

D. W.

Marcus

A. K.

Kang

D. W.

Crowell

M. D.

Rittmann

B. E.

Krajmalnik-Brown

(2017). Distinctive microbiomes and metabolites linked with weight loss after gastric bypass, but not gastric banding. The ISME Journal, 11(9), 2047–2058. https://doi.org/10.1038/ismej.2017.71

24.

Kasai

Sugimoto

Moritani

Tanaka

Oya

Inoue

Tameda

Shiraki

Ito

Takei

Takase

(2015). Comparison of the gut microbiota composition between obese and non-obese individuals in a Japanese population, as analyzed by terminal restriction fragment length polymorphism and next-generation sequencing. BMC Gastroenterology, 15, 100. https://doi.org/10.1186/s12876-015-0330-2

25.

Koliaki

Liatis

le Roux

C. W.

Kokkinos

(2017). The role of bariatric surgery to treat diabetes: Current challenges and perspectives. BMC Endocrine Disorders, 17(1), 50. https://doi.org/10.1186/s12902-017-0202-6

26.

Kullmann

Heni

Veit

Ketterer

Schick

Häring

H.-U.

Fritsche

Preissl

(2012). The obese brain: Association of body mass index and insulin sensitivity with resting state network functional connectivity. Human Brain Mapping, 33(5), 1052–1061. https://doi.org/10.1002/hbm.21268

27.

Lee

E. B.

Mattson

M. P.

(2014). The neuropathology of obesity: Insights from human disease. Acta Neuropathologica, 127(1), 3–28. https://doi.org/10.1007/s00401-013-1190-x

28.

Lepping

R. J.

Bruce

A. S.

Francisco

Yeh

H. W.

Martin

L. E.

Powell

J. N.

Hancock

Patrician

T. M.

Breslin

F. J.

Selim

Donnelly

J. E.

Brooks

W. M.

Savage

C. R.

Simmons

W. K.

Bruce

J. M.

(2015). Resting-state brain connectivity after surgical and behavioral weight loss. Obesity (Silver Spring), 23(7), 1422–1428. https://doi.org/10.1002/oby.21119

29.

Ley

R. E.

Turnbaugh

P. J.

Klein

Gordon

J. I.

(2006). Human gut microbes associated with obesity. Nature, 444(7122), 1022–1023. https://doi.org/10.1038/4441022a

30.

Lin

H. V.

Frassetto

Kowalik

E. J.

Jr. Nawrocki

A. R.

M. M.

Kosinski

J. R.

Hubert

J. A.

Szeto

Yao

Forrest

Marsh

D. J.

(2012). Butyrate and propionate protect against diet-induced obesity and regulate gut hormones via free fatty acid receptor 3-independent mechanisms. PLoS One, 7(4), Article e35240. https://doi.org/10.1371/journal.pone.0035240

31.

Shan

Liang

Nie

Liu

Duan

Huang

Zhang

Dong

Guo

Gao

Shan

(2018). Sleeve gastrectomy recovering disordered brain function in subjects with obesity: A Longitudinal fMRI Study. Obesity Surgery, 28(8), 2421–2428. https://doi.org/10.1007/s11695-018-3178-z

32.

Love

M. I.

Huber

Anders

(2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology, 15(12), 550. https://doi.org/10.1186/s13059-014-0550-8

33.

Zuo

Waltz

J. A.

Zhan

Scholl

C. A.

Rea

Yang

Stein

E. A.

(2007). Synchronized delta oscillations correlate with the resting-state functional MRI signal. Proceedings of the National Academy of Sciences of the United States of America, 104(46), 18265–18269. https://doi.org/10.1073/pnas.0705791104

34.

Mahadzir

M. D. A.

Shyam

Barua

Krishnappa

Ramamurthy

(2017). Effect of probiotic microbial cell preparation (MCP) on fasting blood glucose, body weight, waist circumference, and faecal short chain fatty acids among overweight Malaysian adults: A pilot randomised controlled trial of 4 weeks [Article]. Malaysian Journal of Nutrition, 23(3), 329–341. https://www.scopus.com/inward/record.uri?eid=2-s2.0-85040455992&partnerID=40&md5=1f45d7caef1f2342296f2596bf1acd8f

35.

Mahlen

S. D.

Clarridge

J. E.

3rd. (2009). Site and clinical significance of Alloscardovia omnicolens and Bifidobacterium species isolated in the clinical laboratory. Journal of Clinical Microbiology, 47(10), 3289–3293. https://doi.org/10.1128/jcm.00555-09

36.

Maki

K. A.

Kazmi

Barb

J. J.

Ames

(2021). The oral and gut bacterial microbiomes: Similarities, differences, and connections. Biological Research for Nursing, 23(1), 7–20. https://doi.org/10.1177/1099800420941606

37.

Mantini

Perrucci

M. G.

Del Gratta

Romani

G. L.

Corbetta

(2007). Electrophysiological signatures of resting state networks in the human brain. Proceedings of the National Academy of Sciences of the United States of America, 104(32), 13170–13175. https://doi.org/10.1073/pnas.0700668104

38.

Meek

C. L.

Lewis

H. B.

Reimann

Gribble

F. M.

Park

A. J.

(2016). The effect of bariatric surgery on gastrointestinal and pancreatic peptide hormones. Peptides, 77, 28–37. https://doi.org/https://doi.org/10.1016/j.peptides.2015.08.013

39.

Melton

G. B.

Steele

K. E.

Schweitzer

M. A.

Lidor

A. O.

Magnuson

T. H.

(2008). Suboptimal weight loss after gastric bypass surgery: Correlation of demographics, comorbidities, and insurance status with outcomes. Journal of Gastrointestinal Surgery, 12(2), 250–255. https://doi.org/10.1007/s11605-007-0427-1

40.

Mulla

C. M.

Middelbeek

R. J. W.

Patti

M. E.

(2018). Mechanisms of weight loss and improved metabolism following bariatric surgery. Annals of the New York Academy of Sciences, 1411(1), 53–64. https://doi.org/10.1111/nyas.13409

41.

National Heart Lung and Blood Institute . (n.d.). Classification of overweight and obesity by BMI, waist circumference, and associated disease risks. Retrieved May 1, 2021 from https://www.nhlbi.nih.gov/health/educational/lose_wt/BMI/bmi_dis.htm

42.

Nota

M. H. C.

Vreeken

Wiesmann

Aarts

E. O.

Hazebroek

E. J.

Kiliaan

A. J.

(2020). Obesity affects brain structure and function- rescue by bariatric surgery? Neuroscience and Biobehavioral Reviews, 108, 646–657. https://doi.org/https://doi.org/10.1016/j.neubiorev.2019.11.025

43.

Ochner

C. N.

Gibson

Carnell

Dambkowski

Geliebter

(2010). The neurohormonal regulation of energy intake in relation to bariatric surgery for obesity. Physiology & Behavior, 100(5), 549–559. https://doi.org/10.1016/j.physbeh.2010.04.032

44.

Ochner

C. N.

Gibson

Shanik

Goel

Geliebter

(2011). Changes in neurohormonal gut peptides following bariatric surgery. International Journal of Obesity (2005), 35(2), 153–166. https://doi.org/10.1038/ijo.2010.132

45.

Ozsoy

Demir

(2018). Which bariatric procedure is the most popular in the World? A bibliometric comparison. Obesity Surgery, 28(8), 2339–2352. https://doi.org/10.1007/s11695-018-3163-6

46.

Palleja

Kashani

Allin

K. H.

Nielsen

Zhang

Brach

Liang

Feng

Jørgensen

N. B.

Bojsen-Møller

K. N.

Dirksen

Burgdorf

K. S.

Holst

J. J.

Madsbad

Wang

Pedersen

Hansen

Arumugam

(2016). Roux-en-Y gastric bypass surgery of morbidly obese patients induces swift and persistent changes of the individual gut microbiota. Genome Medicine, 8(1), 67. https://doi.org/10.1186/s13073-016-0312-1

47.

Periáñez

J. A.

Lubrini

García-Gutiérrez

Ríos-Lago

(2021). Construct validity of the stroop color-word test: Influence of speed of visual search, verbal fluency, working memory, cognitive flexibility, and conflict monitoring. Archives of Clinical Neuropsychology: The Official Journal of the National Academy of Neuropsychologists, 36(1), 99–111. https://doi.org/10.1093/arclin/acaa034

48.

Porat

Vaynshtein

Gibori

Avramoff

Shaked

Dukhno

Czeiger

Sebbag

Dahan

(2021). Stomach pH before vs. after different bariatric surgery procedures: Clinical implications for drug delivery. The European Journal of Pharmaceutics and Biopharmaceutics, 160, 152–157. https://doi.org/10.1016/j.ejpb.2021.01.016

49.

Schauer

P. R.

Bhatt

D. L.

Kirwan

J. P.

Wolski

Aminian

Brethauer

S. A.

Navaneethan

S. D.

Singh

R. P.

Pothier

C. E.

Nissen

S. E.

Kashyap

S. R.

(2017). Bariatric surgery versus intensive medical therapy for diabetes - 5-year outcomes. The New England Journal of Medicine, 376(7), 641–651. https://doi.org/10.1056/NEJMoa1600869

50.

Schmidt

T. S.

Hayward

M. R.

Coelho

L. P.

S. S.

Costea

P. I.

Voigt

A. Y.

Wirbel

Maistrenko

O. M.

Alves

R. J.

Bergsten

de Beaufort

Sobhani

Heintz-Buschart

Sunagawa

Zeller

Wilmes

Bork

(2019). Extensive transmission of microbes along the gastrointestinal tract. Elife, 8, Article e42693. https://doi.org/10.7554/eLife.42693

51.

Silva

Y. P.

Bernardi

Frozza

R. L.

(2020). The role of short-chain fatty acids from gut microbiota in gut-brain communication. Frontiers in Endocrinology, 11, 25-25. https://doi.org/10.3389/fendo.2020.00025

52.

van de Wouw

Schellekens

Dinan

T. G.

Cryan

J. F.

(2017). Microbiota-gut-brain axis: Modulator of host metabolism and appetite. The Journal of Nutrition, 147(5), 727–745. https://doi.org/10.3945/jn.116.240481

53.

Wang

Harvey

Martin

van der Beek

E. M.

Knol

Cryan

J. F.

Renes

I. B.

(2018). Targeting the gut microbiota to influence brain development and function in early life. Neuroscience and Biobehavioral Reviews, 95, 191–201. https://doi.org/10.1016/j.neubiorev.2018.09.002

54.

Wiemerslage

Zhou

Olivo

Stark

Hogenkamp

P. S.

Larsson

E.-M.

Sundbom

Schiöth

H. B.

(2017). A resting-state fMRI study of obese females between pre- and postprandial states before and after bariatric surgery. European Journal of Neuroscience, 45(3), 333–341. https://doi.org/https://doi.org/10.1111/ejn.13428

55.

Wilbert

S. A.

Mark Welch

J. L.

Borisy

G. G.

(2020). Spatial ecology of the human tongue dorsum microbiome. Cell Reports, 30(12), 4003–4015.e3. https://doi.org/10.1016/j.celrep.2020.02.097

56.

Wisser

A. S.

Habbel

Wiedenmann

Klapp

B. F.

Mönnikes

Kobelt

(2010). Interactions of gastrointestinal peptides: Ghrelin and its anorexigenic antagonists. International Journal of Peptides, 2010, 817457. https://doi.org/10.1155/2010/817457

57.

World Health Organization . Obesity and Overweight; WHO: Geneva, Switzerland, 2020; Available online: https://www.who.int/ news-room/fact-sheets/detail/obesity-and-overweight (accessed on 13 October 2020)

58.

Yu-Feng

Yong

Chao-Zhe

Qing-Jiu

Man-Qiu

Meng

Li-Xia

Tian-Zi

Yu-Feng

(2007). Altered baseline brain activity in children with ADHD revealed by resting-state functional MRI. Brain & Development, 29(2), 83–91. https://doi.org/10.1016/j.braindev.2006.07.002

59.

Zeighami

Iceta

Dadar

Pelletier

Nadeau

Biertho

Lafortune

Tchernof

Fulton

Evans

Richard

Dagher

Michaud

(2021). Spontaneous neural activity changes after bariatric surgery: A resting-state fMRI study. Neuroimage, 241, 118419. https://doi.org/10.1016/j.neuroimage.2021.118419

60.

Zhang

Tian

Zang

Liu

Walter

(2015). Altered baseline brain activity differentiates regional mechanisms subserving biological and psychological alterations in obese men. Scientific Reports, 5, 11563. https://doi.org/10.1038/srep11563

61.

Zou

Q. H.

Zhu

C. Z.

Yang

Zuo

X. N.

Long

X. Y.

Cao

Q. J.

Wang

Y. F.

Zang

Y. F.

(2008). An improved approach to detection of amplitude of low-frequency fluctuation (ALFF) for resting-state fMRI: Fractional ALFF. Journal of Neuroscience Methods, 172(1), 137–141. https://doi.org/10.1016/j.jneumeth.2008.04.012

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.18 MB

0.53 MB

0.16 MB

8.56 MB

4.65 MB

3.97 MB

4.28 MB