Abstract
Objectives
Pediatric acute pancreatitis (AP), Crohn's disease (CD), and irritable bowel syndrome (IBS) are associated with gut dysbiosis, but differences and similarities between conditions are unknown. We hypothesized that gut microbial ecology would differ across these disorders.
Methods
Stool was collected from 120 subjects (AP [n = 30], CD [n = 29], IBS Rome IV [n = 27], and healthy controls [HC, n = 34]). Shotgun metagenomic sequencing was performed on extracted DNA and taxonomic and functional profiles obtained using sylph and HUMAnN3 with default parameters.
Results
Age interquartile range for all participants was 8.1–17.7 years. Shannon diversity was decreased in AP compared to IBS or HC (p < 0.0001) and similar to CD (p = 0.97). CD differed from IBS (p = 0.001) and HC (p < 0.0001) while IBS and HC were similar (p = 0.61). Ordination of the first two principal coordinate analyses axes showed sample clustering by condition (R 2 = 0.12, p < 0.001), and differences between all conditions in pairwise comparisons (p < 0.001). Escherichia coli, Ruminococcus gnavus, Staphylococcus aureus, and Thomasciavelia ramosa remained enriched when all conditions (AP, CD, and IBS) were compared as a single group to HC. Using a random forest machine learning algorithm for species relative abundance, the ability to classify a sample to each condition versus all others was highest for CD (area under the receiver operative characteristic curve, AUC = 0.97), followed by AP (AUC = 0.92), HC (AUC = 0.88), and IBS (AUC = 0.83).
Conclusion
Organic disorders (AP and CD) are associated with significant gut dysbiosis than IBS which appears more like HC. Interventions targeting shifts in commensals in AP and CD may be helpful in improving outcomes in both disorders.
Keywords: disorders of gut–brain interaction, dysbiosis, metagenomic sequencing, organic disorders, shannon diversity
What is Known
Gut microbial dysbiosis has been implicated in the pathophysiology of several pediatric gastrointestinal disorders.
Dysbiosis has been shown to affect outcomes in these disorders.
What is New
Children with organic gastrointestinal (GI) disorders particularly acute pancreatitis (AP) and Crohn's disease (CD) show gut taxonomy and functional profiles that are different from those with irritable bowel syndrome and healthy children.
This study highlights influence of the gut microbiota profile in pediatric GI disorders particularly CD and AP. It supports developing microbiome‐targeted therapies to improve outcomes in these GI disorders.
1. INTRODUCTION
Intestinal microbiota are involved in maintaining homeostasis and the regulation of various gastrointestinal (GI) physiologic processes. 1 Microbial dysbiosis and altered metabolic function have been observed in intestinal and extra‐intestinal disorders, 2 some of which involve gut inflammation and others disruption of the gut‐brain axis. 3 Decreased gut microbial diversity, such as 4 , 5 , 6 , 7 lessened abundance of Bacteroides and Ruminococcus species, and fewer butyrate‐producing organisms such as Faecalbacterium prausnitzii and Roseburia hominis has been reported in patients with inflammatory bowel disease (IBD). 8 , 9 , 10 Similarly, pathobionts like Escherichia and Klebsiella are increased in active IBD. 8 , 11 , 12
Children with irritable bowel syndrome (IBS) have been shown to have elevated Firmicutes, including Clostridial species, and increased long chain fatty acid (LCFA) synthesis pathways suggesting pain‐potentiating effects. 4 Recent studies suggest that microbial dysbiosis may also contribute to acute pancreatitis (AP). 5 Severe AP is associated with increased Enterococcus faecalis and Clostridium innocuum in children, for example. 5 Our data and those of others have shown gut dysbiosis in different GI conditions, Crohn's disease (CD), AP, and IBS being the most common. In our institution, using the same microbiome sequencing and analytic pipeline, we previously demonstrated that gut microbial profiles differ in children with CD, AP, and IBS when compared to healthy children. There is limited data about the relationship of microbial variations across these conditions and their role in pathogenesis. It is unclear if altered intestinal flora plays a role in shared pathogenesis or provides a spectrum of change between organic GI conditions and IBS. Thus, we aimed to compare microbial composition and function in pediatric patients with IBD‐CD, IBS, and AP as well as with healthy controls (HCds). We utilized previously sequenced stool samples and existing retrospective clinical data in each cohort and reprocessed all data for uniformity. Cross comparison across these conditions and with HC has not been explored prior, and is thus novel. 4 , 6 , 13
2. METHODS
2.1. Ethics statement
All studies were approved by the Institutional Review Board at Cincinnati Children Hospital Medical Center (CCHMC). Informed consent was obtained for collecting questionnaires and microbiome data before participants were included in the study.
2.2. Study population
Cohorts were obtained from three prior observational studies enrolling at the same tertiary institution from October 1, 2023, to June 1, 2024. Demographic data (age, gender, race, and anthropometrics) and medical history were obtained for each cohort and entered into REDCap data registries (©Vanderbilt University, Nashville, TN). Stool samples were collected similarly for each cohort and DNA extraction and metagenomic sequencing by the Microbial Genomics and Metagenomics Laboratory (MGML) at CCHMC followed a standardized protocol (Figure S1). Analyses of these patient cohorts (or subsets of the cohort) (with different aims) have been published previously. 5 , 6 , 7 A brief description of each cohort is provided below.
2.2.1. AP cohort
This is a sub‐analysis of a prospective AP registry enrolling patients age <21 years with first‐time AP. Stool samples were collected within 7 days of onset of AP symptoms, and participants were prospectively followed for up to 1 year. Clinical data and stool samples were collected at baseline as well as at multiple time points (3 months and 12 months) post AP episode. Most participants had mild AP (n = 23), without pancreatic or systemic complications. Only three patients received antibiotics at the time of stool collection. 5 Only baseline stool samples were included in this sub‐analysis.
2.2.2. CD cohort
Participants <20 years with a diagnosis of CD were recruited as previously described. 6 This was a sub‐analysis of an existing IBD biorepository where patients, regardless of time since CD diagnosis, submitted stool samples and brief symptom questionnaires (patient‐reported outcomes or PRO). Stool collections were done in the outpatient setting. Prior research on this cohort examined longitudinal shifts associated with Clostridioides difficile infection and time to surgery as well as PRO and microbial shifts. 6 , 13 Only the baseline samples for subjects with a fecal calprotectin >250 µg/g (indicating active mucosal inflammation) 14 were included in this sub‐analysis. Patients ranged in severity from active to quiescent disease and one patient received antibiotics just prior to stool collection.
2.2.3. IBS cohort
This was a sub‐analysis of a prospective study assessing microbiome changes in adolescents 11–18 years who met the Rome IV criteria for IBS and underwent auricular vagal nerve stimulation. 14 Participants did not have organic GI disorders and those on probiotics, antibiotics, or complete enteral nutrition were excluded as described prior. Participant symptoms ranged in severity from mild to moderate based on IBS‐symptom severity scoring. 4 Stool samples were collected at baseline, immediately following treatment completion and at the 3‐month follow‐up. Only baseline stool samples were included in this sub‐analysis.
2.2.4. HC cohort
Data from HC recruited in the Precision Crohn's Disease Management Utilizing Predictive Protein Panels (ENvISION study: NCT04131504) were used. Participants had no known organic GI disorder or Rome IV criteria for IBS and did not receive any antibiotics within 30 days of enrollment. 14
2.3. Procedures
2.3.1. Stool collection
Stool samples were collected using 58‐EZ‐SAMPLER kit with sterile self‐sealing containers with three collection tubes for both the IBS and CD cohorts. 6 For the AP cohort, only one collection tube was used for metagenomic sequencing. 5 Stool samples were frozen and stored at −80°C until analysis, and sequencing occurred within 1 year of collection.
2.3.2. Stool DNA extraction, sequencing, and bioinformatic processing
Whole‐genome shotgun metagenomic sequencing of microbial DNA was used to identify taxonomic and functional features of the stool microbiome. Please see Table S1 for details.
2.3.3. Intestinal inflammation assessment
Fecal calprotectin levels were obtained on IBS, IBD, AP cohorts, and HC to evaluate for intestinal inflammation via the. Inova Quanta Lite® Calprotectin ELISA assay (Bühlmann Laboratories AG, Schönenbuch, Switzerland). 15
2.4. Outcomes
2.4.1. Primary
We compared similarities and differences of gut taxonomic profiles, including alpha diversity (measured via Observed species and Shannon index), beta diversity, and species relative abundance across different GI disorders in children and adolescents with AP, CD, IBS, and HC.
2.4.2. Secondary
We compared functional profiles assessed by pathway relative abundance.
2.5. Statistical analysis
Demographic and clinical characteristics of participants in each cohort were described using descriptive statistics. Analysis of variance (ANOVA) and chi‐square/Fisher's exact tests were used to test differences in demographic and clinical characteristics. Multiple linear regression was used to test for differences in species richness and Shannon diversity adjusted for age, sex, and body mass index (BMI) z‐score after subsampling to the lowest observed read depth (1.3 million reads). Pairwise comparisons were carried out for each cohort and Tukey's correction applied to account for multiple testing. Ordinations of the first two principal coordinate axes obtained from the Bray–Curtis dissimilarity matrix after subsampling were used to assess the similarity in the microbial community composition of samples from each cohort. Permutational analysis of variance (PERMANOVA) was performed using adonis2 function in vegan (version 2.6.4) 16 was used to partition the sums of squares for each cohort centroid under the default settings adjusted for age, sex, and BMI z‐score. Pairwise comparisons were carried out for each cohort and Tukey's correction was applied to account for the multiple testing. Microbiome Multivariable Associations with Linear Models (MaAsLin2) (version 1.12.0) 16 was used to identify differentially abundant microbial species, MetaCyc pathways, and enzyme‐catalyzed reactions between cohorts and HC, as well as for specific cohort contrasts of interest (e.g., AP vs. CD and AP/CD vs. IBS). Please see further details in Table S1 and Figure S1.
3. RESULTS
3.1. Cohort demographics (Table 1)
Table 1.
Demographics and baseline characteristics.
| Total (N = 120) | AP (n = 30) | IBS (n = 27) | CD (n = 29) | HC (n = 34) | p‐value | |
|---|---|---|---|---|---|---|
| Demographics | ||||||
| Age years, median (IQR) | 15.2 (12.4, 17.1) | 12.2 (8.1, 17.6) | 15.8 (14.7, 17.1) | 15.5 (12.8, 17.4) | 14.1 (11.7, 16.6) | 0.064 |
| Male, n (%) | 57 (48%) | 15 (50%) | 5 (19%) | 17 (59%) | 20 (59%) | 0.006 |
| Non‐white, n (%) | 18 (15%) | 5 (17%) | 3 (11%) | 3 (10%) | 7 (21%) | 0.682 |
| BMI, median (IQR), n = 113 | 19.7 (17.6, 22.9) | 18.3 (16.2, 21.2) | 21.2 (19.5, 25.6) | 19.1 (17.3, 21.8) | 19.6 (18.0, 22.1) | 0.006 |
| BMI, z‐score n = 112 | 0.01 (1.21) | −0.15 (1.56) | 0.42 (1.16) | −0.21 (0.97) | 0.01 (1.01) | 0.213 |
| Clinical characteristics | ||||||
| Fecal calprotectin, median [IQR] | 340 [50, 841] | 75 [74, 78] | 599 [367, 1775] | 29 [20, 35] | <0.001 | |
| Hospitalization status | All hospitalized | Outpatient | Outpatient | Outpatient | ||
| Disease severity n (%) | Mild: 23 (76%) Moderate to severe: 7 (24%) | Mild: 14 (25%) Moderate to severe: 27 (47.36%) | Inactive: 23 (79%) Mild: 2 (7%) Moderate to severe: 4 (14%) | NA | ||
| Antibiotic use | 3 (10%) | 0 | 1 (3%) | 0 | ||
Note: Age, BMI, and fecal calprotectin presented in median (IQR); gender, race, and disease severity reported in n (%); *IBS severity based on functional disability. Abbreviations: AP, acute pancreatitis; BMI, body mass index; CD, Crohn's disease; HC, healthy controls; IBS, irritable bowel syndrome; IQR, interquartile range.
In our total sample (n = 120), the median age interquartile range (IQR) was 15.2 years (12.4–17.1 years), mostly female and white (Table 1). Most CD and HC participants were male (59% each). The IBS cohort was predominantly female (81%), while the AP cohort was equally distributed between males and females (50% each) p = 0.006. Median BMI was higher in IBS 21.2 (19.5–25.6) than AP 18.3 (16.2–21.2), CD 19.1 (17.3–21.8), and HC cohorts 19.6 (18.0–22.1), (p = 0.025). In the CD cohort, 22 patients had ileocolonic disease, four patients had colon‐only disease, and three patients had ileal CD. Fecal calprotectin was significantly different between the four groups, CD, AP, IBS, and HC cohorts p < 0.001, as expected.
3.2. Alpha diversity
The number of observed species significantly differed across cohorts after controlling for age, gender, and BMI z‐scores. Differences were seen between the AP and IBS (p < 0.0001), AP and CD (p = 0.02), IBS and CD (p = 0.0014) cohorts, HC and AP (p < 0.0001) and HC and CD (p < 0.0001) cohorts (Figure 1A). The IBS and HC cohorts were similar (p = 0.78).
Figure 1.
Comparison of alpha and beta diversity between conditions. (A) Differences in observed species, (B) Shannon (alpha) diversity, (C) principal component analysis ordination plot (beta‐diversity). AP, acute pancreatitis; CD, Crohn's disease; HC, healthy controls; IBS, irritable bowel syndrome.
Shannon diversity differed between AP and IBS (p < 0.0001), IBS and CD (p = 0.0001), HC and AP (p ≤ 0.0001) and HC and CD cohorts (p < 0.0001, Figure 1B) with no differences between AP and CD cohorts (p = 0.97).
3.3. Beta diversity
The ordination of first two principal coordinate analyses (PCoA) axes using the Bray‐Curtis dissimilarity matrix showed clustering of participant samples by GI condition (R 2 = 0.12, p < 0.001), and significant differences between all conditions for pairwise comparisons (p < 0.002) and also when AP and CD samples were combined as a single group and compared to HC (p = 0.001, Figure 1C).
3.4. Differential species abundance
When compared to the HC cohort, species enriched in the AP cohort included Blautia hydrogenotrophica, Clostridium innocuum, Eggerthella lenta, Ruminococcus gnavus, and Thomasciavelia ramosa. In the CD cohort, enriched species included Clostridium symbosium, Escherichia coli, Staphylococcus aureus. The IBS cohort appeared similar to HC cohort with no species identified as differentially abundant based on a false discovery rate (FDR) p‐value < 0.05 and absolute log2 fold‐change ≥2. When all three cohorts (AP, CD, and IBS) together were compared to the HC cohort, E coli, R. gnavus, S. aureus, and T. ramosa remained enriched (Figure 2A). B. hydrogenotrophica, C. innocuum, E. lenta, and Lacticaseibacillus rhamnosus remained enriched in the AP cohort when compared to the CD cohort (Figure 2B). There were no species enriched in the AP and CD cohorts when combined as a single group and compared to the IBS cohort (Figure 2C,D). The details of all species abundance are provided in Table S1.
Figure 2.
Comparison of differentially abundant species between conditions. (A) Heatmap showing differentially abundant species plotted by the Log2 Fold‐Change; (B) Species abundance analysis plots (patients vs. healthy controls) for species identified in any comparisons with FDR‐adjusted p < 0.05 and absolute log2 fold change ≥ 3; (C) Species abundance analysis plots (AP vs. CP) for species with FDR‐adjusted p < 0.05 and absolute log2 fold change ≥ 2; (D) Species abundance analysis plots (AP+CP vs. IBS) for species with FDR‐adjusted p < 0.05 and absolute log2 fold change ≥ 2. AP, acute pancreatitis, CD, Crohn's disease, HC, healthy controls; IBS, irritable bowel syndrome. Species with an FDR p‐value of < 0.05 are plotted and are represented by a blue bar, while the instances with a gray bar had a higher FDR p‐value.
3.5. Differential pathway analysis
There were significant differences in the Metacyc pathways seen by GI condition (R 2 = 0.2, p = 0.001) even after controlling age, gender, and BMI z‐scores. Pairwise comparisons showed that Metacyc pathways were different for each GI organic condition; AP, CD versus HC (p = 0.001), and IBS versus HC (p = 0.045). This significant difference was also seen in CD versus AP (p = 0.001), CD versus IBS (p = 0.001) and IBS versus AP (p = 0.001). When AP and CD cohorts were combined as a single group and compared with HC (p = 0.002). Several pathways that were enriched in each of the organic GI condition cohorts; AP and CD cohorts compared to HC cohorts were not enriched in the IBS cohort when it was compared to the HC cohort except Nicotinamide adenine dinucleotide (NAD). Biosynthesis II NAD salvage pathway III to nicotinamide. Riboside which is involved in NAD synthesis. We also found a pathway involved in L‐arginine degradation in the AP and CD cohorts compared to HC that was not present in the IBS cohort (log two‐fold change of ≥2 p < 0.05).
When compared to HC, the IBS cohort was enriched in pathways utilizing mevalonate for biosynthetic pathways of terpenes/terpenoids like isoprenoid. Pathway involved in pyridoxal 5 phosphate salvage was enriched in the combined group of organic GI disorders (AP + CD) when compared to IBS alone and a pathway involved in carbohydrate storage and metabolism was depleted in the combined AP and CD group when compared to the IBS cohort.
Further, PCoA axes ordination obtained from Bray Curtis dissimilarity matrix also showed clustering of pathways by condition; AP versus HC (p = 0.001), IBS versus HC (p = 0.033), CD versus HC (p = 0.001), IBS versus AP (p = 0.001), CD versus AP (p = 0.001), CD versus IBS (p = 0.001). Differential pathway analysis for enzyme‐catalyzed reactions revealed several pathways involved in amino acid metabolism were enriched in the cohort with organic GI orders (AP + CD) combined together and compared to the IBS cohort while pathways involved in carbohydrate metabolism were depleted in the GI organic disorders group compared to the IBS cohort (Figure S2). Details of all pathways and enzymes are provided in Tables S2 and S3 respectively.
3.6. Model assessment
We built a random forest model to assess the ability to classify each sample to its cohort of origin based only on the species relative abundance. The out‐of‐sample multiclass area under the receiver operating curve (AUC) was 0.90 with specificity of 90% highlighting the ability to correctly identify the cohort of origin based on the relative species abundance. The AUC for classifying each sample the correct cohort versus any other as follows: CD (AUC = 0.97, 95% confidence interval [CI] [96; 98]), AP (AUC = 0.92, 95% CI [89; 95]), HC (AUC = 0.88, 95% CI [86; 91]), and IBS (AUC = 0.83, 95% CI [79; 87]) (Figure 3). Figure S2 shows the top 10 species based on their variable importance as determined by the random forest model.
Figure 3.
Receiver operating characteristic curves of random forest machine learning plot. (A) Out‐of‐sample multiclass AUC using random forest machine learning algorithm. (B) Top 10 species determining random forest variable importance. AP, acute pancreatitis; AUC, area under the receiver operating characteristic curve. CD, Crohn's disease; FDR, false discovery rate; HC, healthy controls; IBS, irritable bowel syndrome.
4. DISCUSSION
Our study showed that while both alpha and beta diversity were similar between HC and IBS cohorts, the IBS cohort differed in alpha diversity from both organic disease cohorts. Prior studies have shown that alpha diversity does not differ with indices of severity for neither CD 6 nor AP. 5 This difference then is unlikely to be a consequence of disease activity and may suggest a role for intestinal flora in the pathophysiology of these inflammation‐driven diseases.
The IBS and HC cohorts were similar in overall species relative abundance, and no significant differences were seen. However, other studies have previously reported microbial shifts between IBS and HC 17 including enrichment of inflammatory pathways in IBS. 18 The HC‐IBS similarities in this study suggest a spectrum of deviance in the intestinal flora from a healthy gut to that associated with disease. This may explain the limitation of using microbial therapies to achieve optimal outcomes for IBS. 19 , 20 , 21 , 22
The association of intestinal dysbiosis and CD has been studied, though an exact and consistent microbial signature across disease activity, phenotype, age, and geography, has not been identified. 23 While it is thought that an altered microbiome contributes to CD pathogenesis and therapeutic response, this hypothesis remains under investigation. 24 Further, both human and animal studies have shown that AP is associated with gut microbial dysbiosis, and certain species are relatively abundant in patients with severe AP when compared to mild AP. 5 , 25 , 26 We identified previously reported microbial shifts associated with active CD and revealed unexpected microbial commonalities between those with CD and AP.
Consistent with prior studies, E. coli was relatively abundant in CD and AP 5 compared to HC, while Ruminococcus strains were relatively reduced. 8 , 11 , 12 R. gnavus, which has been previously reported in association with CD, AP, and IBS, and produces pro‐inflammatory compounds, remained differentially abundant in our AP cohort in this study relative to HC but not in CD. 27 In both CD and AP relative to HC, we found a depletion of Bifidobacterium species. These species are short‐chain‐fatty‐acid‐producing commensals known to have a positive effect on IBD and human health. 28 When AP and CD were directly compared, Bifidobacterium longum, the most common microbe in the infant gut was differentially abundant in the AP cohort. 29 However, other butyrate‐producing commensals such as Roseburia intestinalis, were depleted in AP relative to CD. It is possible that these shifts represent a disease state in general, but also plausible that commensals have a role not only in intestinal mucosal disease but in pancreatitis as well.
No change in differential abundance was found in Faecalbacterium prauznitzii in our CD cohort, but the AP cohort showed a decreased relative abundance compared to HC. Reduction in this commensal may have disease implications that extend beyond intestinal mucosal inflammation. Expansion of beneficial commensals such as Bifidobacteria via ingestion of human milk oligosaccharides including 2’ fucosyllactose, difucosyllactose, and 3’ sialyllactose, has been reported. 30 If this is effective in altering commensal populations, this could prove an additional therapeutic pathway for organic GI disorders associated with dysbiosis such as AP.
MetaCyc pathway analysis showed that many pathways enriched in AP and CD compared to the HC were not enriched in the IBS cohort. The AP and CD cohorts had enrichment of a pathway involved in arginine degradation and fatty acid oxidation compared to HC cohort that was not seen in the IBS cohort. While the significance of these pathway variations is unknown, these differences may highlight opportunities for additional investigation.
A random forest model prediction model based on relative species abundance showed higher sensitivity for predicting CD and AP. If similar performance can be obtained upon external validation, then fecal sampling may ultimately serve as additional non‐invasive diagnostic tool.
To our knowledge, this is the first and largest study of children and young adults comparing microbial signatures across different GI conditions. Limitations include stools collected over a broad timepoint in a small, heterozygous CD cohort and heterogeneity in prior antibiotic exposure between groups. We did not correlate the microbiome with endoscopic activity or radiographic inflammation in the organic GI conditions, nor control for disease location, duration, severity, medications, or antibiotic and probiotic exposure. We only utilized fecal calprotectin as a marker of active inflammation and did not include endoscopic or radiographic information. We also did not compare microbiome by specific clinical symptoms in the CD cohort (abdominal pain or increased stool frequency, for example) as the cohort was too small for this type of subgroup analysis. Although we selected the GI conditions which were already studied in our institution, access the same fecal collection and storage processes and the same metagenomic processing pipeline gives uniformity to our results, which is a strength. There was no overlap with patient enrollment across disease cohorts. This study contributes to the body of knowledge that gut dysbiosis contributes to GI pathology and microbial signatures may 1 day be a non‐invasive differentiating marker between inflammatory versus functional disease or, better yet, a target for personalized therapeutic intervention.
5. CONCLUSION
In conclusion, our study compared the fecal microbiome across GI conditions in children and adolescents. IBS resembled HC, while AP and CD were similar. We uncovered novel microbial signatures and pathways differentiating these conditions and a model to identify each disease type based on their microbial patterns. This will pave way for future multi‐center studies to better understand the relationship between dysbiosis and gut inflammation and to develop targeted therapies.
CONFLICT OF INTEREST STATEMENT
Dr. Neha R. Santucci is a key opinion leader for Kate Farms, Gi Health Foundation and consultant for Ironwood Pharmaceuticals. The remaining authors dec;are no conflicts of interest.
Supporting information
Supporting Figure 1: Methodology, CD: Crohn's disease, AP: acute pancreatitis, IBS: irritable bowel syndrome.
Supporting Figure 2: Comparison of differential pathways and enzymes between conditions, A: MetaCyc pathway analysis plots showing differentially abundant pathways. Pathways with an FDR p‐value of < 0.1 are plotted and are represented by a green bar, while the instances with a gray bar had a higher FDR p‐value; B. Differentially abundant enzyme plots. Pathways with an FDR p‐value of <replace0.1 are plotted and are represented by a red bar, while the instances with a gray bar had a higher FDR p‐value. Features plotted by the Log2 Fold‐Change; CD: Crohn's disease, AP: acute pancreatitis, IBS: irritable bowel syndrome, HC: healthy controls.
Supporting File S1.
Supporting File S2.
Supporting File S3.
ACKNOWLEDGMENTS
Adalyn Heins and Priyanshi Shah for assistance with Biorender figures. This was supported by the National Institutes of Health ‐ National Institute of Diabetes and Digestive and Kidney Disease K23DK135797 (N.S.), 23DK118190 (M.A.H.), R03 DK131156 (M.A.H.). This project was supported in part by NIH P30 DK078392, the Digestive Diseases Research Core Center in Cincinnati (L.D.), and the Leona M. and Harry B. Helmsley Charitable Trust for the ENvISION study: NCT04131504 (P.M. and L.D.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Santucci NR, Dike CR, Hellmann J, et al. Gut microbiome in pediatric acute pancreatitis and Crohn's disease versus irritable bowel syndrome and healthy controls. J Pediatr Gastroenterol Nutr. 2026;83:728‐736. 10.1002/jpn3.70504
Neha R. Santucci and Chinenye R. Dike shared co‐first authorship.
DATA AVAILABILITY STATEMENT
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
REFERENCES
- 1. Heintz‐Buschart A, Wilmes P. Human gut microbiome: function matters. Trends Microbiol. 2018;26(7):563‐574. 10.1016/j.tim.2017.11.002 [DOI] [PubMed] [Google Scholar]
- 2. Davenport ER, Sanders JG, Song SJ, Amato KR, Clark AG, Knight R. The human microbiome in evolution. BMC Biol. 2017;15(1):127. 10.1186/s12915-017-0454-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Martin CR, Osadchiy V, Kalani A, Mayer EA. The brain‐gut‐microbiome axis. Cell Mol Gastroenterol Hepatol. 2018;6(2):133‐148. 10.1016/j.jcmgh.2018.04.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Castillo DF, Denson LA, Haslam DB, et al. The microbiome in adolescents with irritable bowel syndrome and changes with percutaneous electrical nerve field stimulation. Neurogastroenterol Motil. 2023;35(7):e14573. 10.1111/nmo.14573 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Dike CR, Ollberding NJ, Thompson T, et al. Acute pancreatitis is associated with gut dysbiosis in children. Dig Liver Dis. 2024;56(3):444‐450. 10.1016/j.dld.2023.10.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Hellmann J, Ta A, Ollberding NJ, et al. Patient‐reported outcomes correlate with microbial community composition independent of mucosal inflammation in pediatric inflammatory bowel disease. Inflamm Bowel Dis. 2023;29(2):286‐296. 10.1093/ibd/izac175 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Thapar N, Benninga MA, Crowell MD, et al. Paediatric functional abdominal pain disorders. Nat Rev Dis Primers. 2020;6(1):89. 10.1038/s41572-020-00222-5 [DOI] [PubMed] [Google Scholar]
- 8. Gevers D, Kugathasan S, Denson LA, et al. The treatment‐naive microbiome in new‐onset Crohn's disease. Cell Host Microbe. 2014;15(3):382‐392. 10.1016/j.chom.2014.02.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Lloyd‐Price J, Arze C, Ananthakrishnan AN, et al. Multi‐omics of the gut microbial ecosystem in inflammatory bowel diseases. Nature. 2019;569(7758):655‐662. 10.1038/s41586-019-1237-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Machiels K, Joossens M, Sabino J, et al. A decrease of the butyrate‐producing species Roseburia hominis and Faecalibacterium prausnitzii defines dysbiosis in patients with ulcerative colitis. Gut. 2014;63(8):1275‐1283. 10.1136/gutjnl-2013-304833 [DOI] [PubMed] [Google Scholar]
- 11. Kugathasan S, Denson LA, Walters TD, et al. Prediction of complicated disease course for children newly diagnosed with Crohn's disease: a multicentre inception cohort study. Lancet (London, England). 2017;389(10080):1710‐1718. 10.1016/s0140-6736(17)30317-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Lewis JD, Chen EZ, Baldassano RN, et al. Inflammation, antibiotics, and diet as environmental stressors of the gut microbiome in pediatric Crohn's disease. Cell Host Microbe. 2015;18(4):489‐500. 10.1016/j.chom.2015.09.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Hellmann J, Andersen H, Fei L, et al. Microbial shifts and shorter time to bowel resection surgery associated with C. difficile in pediatric Crohn's disease. Inflamm Bowel Dis. 2020;26(8):1212‐1221. 10.1093/ibd/izz263 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Hyams JS, Di Lorenzo C, Saps M, Shulman RJ, Staiano A, van Tilburg M. Childhood functional gastrointestinal disorders: child/adolescent. Gastroenterology. 2016;150:1456‐1468.e2. 10.1053/j.gastro.2016.02.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Konikoff MR, Denson LA. Role of fecal calprotectin as a biomarker of intestinal inflammation in inflammatory bowel disease. Inflamm Bowel Dis. 2006;12(6):524‐534. 10.1097/00054725-200606000-00013 [DOI] [PubMed] [Google Scholar]
- 16. Mallick H, Rahnavard A, McIver LJ, et al. Multivariable association discovery in population‐scale meta‐omics studies. PLoS Comput Biol. 2021;17(11):e1009442. 10.1371/journal.pcbi.1009442 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Hollister EB, Oezguen N, Chumpitazi BP, et al. Leveraging human microbiome features to diagnose and stratify children with irritable bowel syndrome. J Mol Diagn. 2019;21(3):449‐461. 10.1016/j.jmoldx.2019.01.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Mars RAT, Yang Y, Ward T, et al. Longitudinal multi‐omics reveals subset‐specific mechanisms underlying irritable bowel syndrome. Cell. 2020;182(6):1460‐1473.e17. 10.1016/j.cell.2020.08.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Chumpitazi BP, Cope JL, Hollister EB, et al. Randomised clinical trial: gut microbiome biomarkers are associated with clinical response to a low FODMAP diet in children with the irritable bowel syndrome. Aliment Pharmacol Ther. 2015;42(4):418‐427. 10.1111/apt.13286 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Sokol H, Landman C, Seksik P, et al. Fecal microbiota transplantation to maintain remission in Crohn's disease: a pilot randomized controlled study. Microbiome. 2020;8(1):12. 10.1186/s40168-020-0792-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Suskind DL, Lee D, Kim YM, et al. The specific carbohydrate diet and diet modification as induction therapy for pediatric crohn's disease: a randomized diet controlled trial. Nutrients. 2020;12(12):3749. 10.3390/nu12123749 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Wu J, Lv L, Wang C. Efficacy of fecal microbiota transplantation in irritable bowel syndrome: a meta‐analysis of randomized controlled trials. Front Cell Infect Microbiol. 2022;12:827395. 10.3389/fcimb.2022.827395 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Pittayanon R, Lau JT, Leontiadis GI, et al. Differences in gut microbiota in patients with vs without inflammatory bowel diseases: a systematic review. Gastroenterology. 2020;158(4):930‐946.e1. 10.1053/j.gastro.2019.11.294 [DOI] [PubMed] [Google Scholar]
- 24. Caenepeel C, Falony G, Machiels K, et al. Dysbiosis and associated stool features improve prediction of response to biological therapy in inflammatory bowel disease. Gastroenterology. 2024;166(3):483‐495. 10.1053/j.gastro.2023.11.304 [DOI] [PubMed] [Google Scholar]
- 25. Li X, He C, Li N, et al. The interplay between the gut microbiota and NLRP3 activation affects the severity of acute pancreatitis in mice. Gut Microbes. 2020;11(6):1774‐1789. 10.1080/19490976.2020.1770042 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. van den Berg FF, Hugenholtz F, Boermeester MA, Zaborina O, Alverdy JC. Spatioregional assessment of the gut microbiota in experimental necrotizing pancreatitis. BJS Open. 2021;5(5):zrab061. 10.1093/bjsopen/zrab061 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Henke MT, Kenny DJ, Cassilly CD, Vlamakis H, Xavier RJ, Clardy J. Ruminococcus gnavus, a member of the human gut microbiome associated with Crohn's disease, produces an inflammatory polysaccharide. Proc Natl Acad Sci USA. 2019;116(26):12672‐12677. 10.1073/pnas.1904099116 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Yao S, Zhao Z, Wang W, Liu X. Bifidobacterium longum: protection against inflammatory bowel disease. J Immunol Res. 2021;2021:8030297. 10.1155/2021/8030297 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Turroni F, Peano C, Pass DA, et al. Diversity of bifidobacteria within the infant gut microbiota. PLoS One. 2012;7(5):e36957. 10.1371/journal.pone.0036957 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Otaru N, Bajic D, Van den Abbeele P, et al. Bifidogenic effect of human milk oligosaccharides on pediatric IBD fecal microbiota. Microorganisms. 2024;12(10):1977. 10.3390/microorganisms12101977 [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
Supporting Figure 1: Methodology, CD: Crohn's disease, AP: acute pancreatitis, IBS: irritable bowel syndrome.
Supporting Figure 2: Comparison of differential pathways and enzymes between conditions, A: MetaCyc pathway analysis plots showing differentially abundant pathways. Pathways with an FDR p‐value of < 0.1 are plotted and are represented by a green bar, while the instances with a gray bar had a higher FDR p‐value; B. Differentially abundant enzyme plots. Pathways with an FDR p‐value of <replace0.1 are plotted and are represented by a red bar, while the instances with a gray bar had a higher FDR p‐value. Features plotted by the Log2 Fold‐Change; CD: Crohn's disease, AP: acute pancreatitis, IBS: irritable bowel syndrome, HC: healthy controls.
Supporting File S1.
Supporting File S2.
Supporting File S3.
Data Availability Statement
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.