Diabetes as an independent risk factor for severe inflammatory bowel disease: evidence from an inflammation–metabolism–liver coupling framework

Background: The clinical course of inflammatory bowel disease (IBD) varies widely, and identifying factors associated with severe disease is essential for risk stratification. Diabetes has been proposed as a potential determinant of adverse outcomes, yet its independent role in disease progression remains unclear.

Methods: We retrospectively analyzed clinical data from patients with mild and severe IBD. Demographic, inflammatory, hepatic, and metabolic parameters were compared between groups. Logistic regression was used to identify independent predictors of severe IBD. Model performance was assessed using receiver operating characteristic curves, calibration analysis, and cross-validation. An inflammation–metabolism–liver coupling index (IMLCI) was constructed to integrate key predictors.

Results: Patients with severe IBD exhibited significantly higher levels of inflammatory markers (WBC, neutrophil percentage, CRP), impaired hepatic function indices (ALT, AST, bilirubin), and adverse metabolic profiles (elevated TG and LDL, reduced HDL and vitamin B12). Diabetes was strongly associated with severe IBD (odds ratio = 3.81, P < 0.001), confirming its independent effect beyond traditional risk factors. The inflammation–metabolism–liver coupling index (IMLCI) demonstrated excellent discrimination (AUC = 0.900 in the training cohort and 0.891 in the testing cohort), good calibration (Hosmer–Lemeshow P = 0.83), and robust internal validation, outperforming single laboratory or metabolic biomarkers.

Conclusion: Diabetes represents a strong independent risk factor for severe IBD. The proposed IMLCI framework, integrating inflammation, metabolism, and liver function, demonstrates high predictive accuracy and may provide a practical tool for early identification and risk management of patients with severe IBD.

1 Introduction

Inflammatory bowel disease (IBD), including Crohn’s disease and ulcerative colitis, is a chronic relapsing disorder of the gastrointestinal tract characterized by immune dysregulation and persistent inflammation (1, 2). Over the past decade, the global prevalence of IBD has increased markedly (3), with rising incidence observed in both Western countries and newly industrialized regions (4, 5). This epidemiological trend is closely associated with lifestyle transitions, particularly the increasing burden of obesity, insulin resistance, and metabolic abnormalities in IBD populations (6, 7). Accumulating evidence suggests that metabolic dysfunction not only promotes disease activity and recurrence but also contributes to complications (8), underscoring the critical role of systemic metabolism in determining IBD severity.

Diabetes mellitus, as a prototypical metabolic disorder (9), has recently attracted considerable attention for its impact on IBD outcomes. Clinical observations have shown that IBD patients with comorbid diabetes often present with more severe disease courses (10), including increased hospitalization, higher susceptibility to infections, and reduced therapeutic response. Several cohort studies have further indicated that diabetes may act as an independent risk factor for the development of severe IBD, beyond traditional inflammatory predictors (11, 12). However, the underlying mechanisms linking diabetes to IBD progression remain incompletely understood, and robust evidence from comprehensive clinical analyses is still lacking.

The interplay among inflammation, metabolism, and liver function represents a key biological framework for understanding the heterogeneity of IBD (13). The liver serves as a central hub of immunometabolic regulation, and hepatic dysfunction can amplify systemic inflammatory responses. Concurrently, metabolic disturbances such as dyslipidemia and micronutrient deficiencies may exacerbate immune imbalance, while elevated inflammatory markers reflect disease burden (14–17). Although these associations have been increasingly recognized, few studies have systematically integrated these domains to establish a predictive model. To address this gap, we developed an inflammation–metabolism–liver coupling index (IMLCI) to evaluate the independent contribution of diabetes to severe IBD and to determine its predictive utility in clinical practice.

2 Materials and methods

2.1 Study design and population

This retrospective observational study was conducted using a clinical cohort of patients diagnosed with inflammatory bowel disease (IBD) in China.

Eligible participants were classified into mild and severe groups based on standardized clinical and laboratory criteria, consistent with the ECCO 2023 guidelines and the Chinese IBD Consensus 2024.

Mild IBD was defined as a low inflammatory state, characterized by C-reactive protein (CRP) < 10 mg/L, white blood cell (WBC) count < 8 × 10⁹/L, and normal hepatic enzyme levels (alanine aminotransferase [ALT] < 40 U/L and aspartate aminotransferase [AST] < 40 U/L), without systemic corticosteroid therapy or hospitalization.

Severe IBD was defined based solely on clinical criteria to avoid incorporation bias. Specifically, patients were classified as having severe IBD if they required systemic corticosteroid therapy, biologic treatment, or hospitalization due to active disease, consistent with ECCO 2023 guidelines. Laboratory parameters (CRP, WBC, ALT/AST, albumin) were excluded from this definition to ensure independence from model predictors.

A total of 7,869 patients were included in the final analysis after excluding those with incomplete demographic, biochemical, or clinical data. The overall workflow, including data collection, variable preprocessing, model development, and internal validation, is illustrated in Figure 1.

Figure 1

Figure 1. Study workflow for patient selection, variable processing, model development, and validation. The flowchart illustrates the inclusion of patients with inflammatory bowel disease (IBD) and classification into mild and severe groups based on lab-free clinical criteria (to avoid incorporation bias), followed by data collection of demographic, inflammatory, hepatic, and metabolic variables, construction of the inflammation–metabolism–liver coupling index (IMLCI), and subsequent statistical analyses.

2.2 Data collection and variables

Comprehensive clinical data were extracted from patient records. Demographic variables included age and sex. Inflammatory markers comprised white blood cell (WBC) count, neutrophil percentage, lymphocyte percentage, monocyte percentage, C-reactive protein (CRP), neutrophil-to-lymphocyte ratio (NLR), and monocyte-to-lymphocyte ratio (MLR). Liver function indicators included alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), lactate dehydrogenase (LDH), total bilirubin, direct bilirubin, albumin, globulin, and the albumin-to-globulin (A/G) ratio. Metabolic and nutritional indices included body mass index (BMI), fasting plasma glucose (FPG), glycated hemoglobin (HbA1c), triglycerides (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), and vitamin B12. Comorbidity status of diabetes (yes or no) was also recorded. Information on IBD subtype, disease duration, surgical history, and medication exposures was partially available but not consistently recorded and thus was not included in the multivariable model. However, laboratory and metabolic parameters incorporated into the IMLCI partially capture systemic inflammation and disease activity.

2.3 Variable processing and composite index construction

Continuous variables were standardized before analysis. An inflammation–metabolism–liver coupling index (IMLCI) was derived by integrating significant predictors from multiple domains. Variable selection was performed using logistic regression and machine learning techniques, specifically extreme gradient boosting (XGBoost) combined with SHapley Additive exPlanations (SHAP) for feature importance ranking.

2.4 Statistical analysis

Baseline differences between mild and severe IBD were compared using standard statistical tests. Logistic regression identified independent predictors of severe IBD. Model discrimination was evaluated by receiver operating characteristic (ROC) curves and the area under the curve (AUC), and calibration by calibration curves. Clinical utility was assessed through decision curve analysis (DCA). Subgroup and sensitivity analyses were conducted to confirm model stability.

Continuous variables were z-score–standardized before modeling. To prevent data leakage, all imputation, scaling, and feature-selection steps were performed within the training set and applied to the testing set. Categorical variables were dummy-encoded; missing values (<5%) were imputed via predictive mean matching. Highly collinear features (|r| > 0.8) were removed. The IMLCI model was built with XGBoost, and predictor contributions were interpreted using SHAP analysis.

All modeling procedures followed a leakage-resistant framework. Laboratory variables were excluded from the outcome definition to avoid predictor–outcome overlap, and a sensitivity analysis with the laboratory-inclusive outcome was used to verify robustness.

2.5 Internal validation and model interpretation

To evaluate the robustness and interpretability of the inflammation–metabolism–liver coupling index (IMLCI), the dataset was randomly divided into a training cohort (70%) and a testing cohort (30%) using stratified sampling. Model discrimination was assessed by the area under the receiver operating characteristic curve (AUC) in both cohorts. Internal stability was verified through 500 bootstrap resampling iterations performed on the training cohort, and the mean AUC was used to obtain optimism-corrected estimates.

All preprocessing, scaling, and model fitting were conducted exclusively within the training cohort, followed by independent testing to avoid information leakage and overfitting. Multicollinearity was examined using variance inflation factors (VIF), all below 2.5. Predictor contributions were quantified using SHAP analysis within the XGBoost framework. All analyses were performed in R (version 4.4.0) with the pROC, patchwork, and iml packages.

3 Results

3.1 Baseline characteristics

A total of 7,869 patients diagnosed with inflammatory bowel disease (IBD) were included in the final analysis, comprising 3,842 males (48.8%) and 4,027 females (51.2%), with a mean age of 47.6 ± 12.4 years.

The baseline characteristics of patients with mild and severe IBD are summarized in Table 1.

Table 1

Table 1. Baseline characteristics of patients with mild and severe inflammatory bowel disease (IBD).

Patients with severe IBD were older (52.18 ± 12.01 vs. 45.03 ± 11.95 years, P<0.001) and exhibited significantly higher levels of inflammatory markers, including WBC (8.94 ± 2.00 vs. 6.99 ± 1.98 ×10⁹/L, P<0.001), neutrophil percentage (64.99 ± 7.90% vs. 58.10 ± 7.82%, P<0.001), and CRP (14.93 ± 5.01 vs. 5.89 ± 5.01 mg/L, P<0.001).

Lymphocyte percentage was lower in the severe group (24.94 ± 5.95% vs. 30.10 ± 5.96%, P<0.001), while both NLR and MLR were elevated (2.79 ± 0.88 vs. 2.02 ± 0.55 and 0.30 ± 0.13 vs. 0.21 ± 0.09, respectively; both P<0.001).

Liver function indices were significantly impaired, with higher ALT (54.73 ± 14.86 vs. 32.03 ± 14.95 U/L, P<0.001), AST (47.90 ± 12.24 vs. 30.07 ± 12.02 U/L, P<0.001), and bilirubin levels, while albumin was reduced (33.81 ± 3.91 vs. 40.02 ± 3.96 g/L, P<0.001).

Metabolic and nutritional disturbances included higher BMI, FPG, HbA1c, TG, and LDL, along with lower HDL and vitamin B12 (all P<0.001). Diabetes was more prevalent in patients with severe IBD (39.5% vs. 25.1%, P<0.001).

3.2 Independent predictors of severe IBD

Multivariable logistic regression identified several independent predictors of severe IBD, as shown in Table 2. Age was associated with increased risk (OR = 1.58, 95% CI: 1.29–1.93, P<0.001). Among inflammatory markers, WBC (OR = 3.17, 95% CI: 2.53–4.02), neutrophil percentage (OR = 2.56, 95% CI: 2.07–3.20), lymphocyte percentage (OR = 0.36, 95% CI: 0.28–0.44), and monocyte percentage (OR = 1.67, 95% CI: 1.37–2.04) were significant predictors (all P<0.001). Liver function indicators, including ALT (OR = 8.06, 95% CI: 6.10–10.87), AST (OR = 5.46, 95% CI: 4.25–7.12), total bilirubin (OR = 3.90, 95% CI: 3.09–4.98), direct bilirubin (OR = 7.18, 95% CI: 5.46–9.64), and A/G ratio (OR = 0.25, 95% CI: 0.19–0.31), were also independently associated with severe IBD. Metabolic markers such as TG (OR = 2.52, 95% CI: 2.03–3.17) and LDL (OR = 3.33, 95% CI: 2.65–4.26) were risk factors, while HDL (OR = 0.37, 95% CI: 0.30–0.46) and vitamin B12 (OR = 0.16, 95% CI: 0.12–0.21) were protective. Diabetes emerged as a strong independent predictor (OR = 3.81, 95% CI: 1.85–7.98, P<0.001).

Table 2

Table 2. Multivariable logistic regression analysis of predictors for severe inflammatory bowel disease (IBD).

3.3 Model performance

The diagnostic performance of predictive models is summarized in Table 3. The IMLCI model achieved an AUC of 0.900 (95% CI: 0.885–0.915) in the training set and 0.889 (95% CI: 0.873–0.904) in the testing set, indicating excellent discrimination. Calibration was good (Hosmer–Lemeshow P = 0.876 for training, 0.862 for testing), and fivefold cross-validation yielded a consistent mean AUC of 0.89. The IMLCI clearly outperformed individual predictors such as WBC (AUC = 0.757), ALT (AUC = 0.860), and BMI (AUC = 0.540), showing superior diagnostic accuracy and potential clinical utility.

Table 3

Table 3. Diagnostic performance and internal validation of the inflammation–metabolism–liver coupling index (IMLCI).

Prior to model fitting, multicollinearity diagnostics were performed for all candidate predictors. All variables exhibited variance inflation factors (VIF) below 2.5, indicating no evidence of multicollinearity. The complete regression coefficients and VIF values are provided in Supplementary Table S1.

3.4 Receiver operating characteristic analysis

As illustrated in Figure 2, the IMLCI exhibited outstanding discrimination for severe IBD, with its ROC curve almost reaching the upper-left corner, indicating markedly higher predictive accuracy compared with single-variable models.

Figure 2

Figure 2. (A) Heatmap showing hierarchical clustering of inflammatory, hepatic, and metabolic variables in mild versus severe inflammatory bowel disease (IBD). Samples are grouped based on similarity of biochemical profiles, with stage annotated as mild (cyan) or severe (red). (B) Box plots comparing representative variables (age, WBC, ALT, BMI, and vitamin B12) between mild and severe IBD groups, illustrating significant differences across key inflammatory, hepatic, and metabolic indicators.

3.5 Calibration analysis

Calibration plots are presented in Figure 3. The IMLCI curve was closely aligned with the ideal diagonal line, indicating excellent agreement between predicted and observed probabilities across risk strata.

Figure 3

Figure 3. (A) Forest plot displaying odds ratios (ORs) with 95% confidence intervals for independent predictors of severe inflammatory bowel disease (IBD), including inflammatory markers, hepatic indicators, metabolic variables, and diabetes. (B) Receiver operating characteristic (ROC) curves comparing the predictive performance of the inflammation–metabolism–liver coupling index (IMLCI) with single predictors such as WBC, ALT, and BMI. (C) Calibration curve showing the agreement between predicted and observed probabilities for the IMLCI model, demonstrating excellent calibration accuracy.

3.6 Internal validation of the IMLCI model

The internal validation demonstrated that the inflammation–metabolism–liver coupling index (IMLCI) maintained excellent and stable discriminatory performance. In the training cohort, the model achieved an AUC of 0.90, and in the independent testing cohort, an AUC of 0.89 (Figure 4A). Bootstrap validation with 500 resampling iterations on the training cohort yielded a mean AUC of 0.90 (Figure 4B), confirming the model’s internal stability and minimal overfitting.

Figure 4

Figure 4. Internal validation and clinical utility of the inflammation–metabolism–liver coupling index (IMLCI) model. (A) Receiver operating characteristic (ROC) curves of the IMLCI in the training and independent testing cohorts, showing excellent and stable discrimination (AUC = 0.90 and 0.89, respectively). (B) Bootstrap validation with 500 resampling iterations on the training cohort demonstrated consistent performance (mean AUC = 0.90), confirming model stability and minimal overfitting. (C) Decision curve analysis (DCA) revealed that the IMLCI provided higher net clinical benefit than the “treat-all” and “treat-none” strategies across threshold probabilities of 0.1–0.8, supporting its clinical applicability for individualized risk prediction.

To further assess its clinical applicability, decision curve analysis (DCA) was performed to evaluate the net benefit of using the IMLCI for severe IBD prediction across a range of threshold probabilities (Figure 4C). The IMLCI demonstrated consistently higher net benefit than the “treat-all” and “treat-none” strategies between threshold probabilities of 0.1–0.8, indicating tangible clinical utility in guiding individualized risk stratification. For illustrative risk stratification, IMLCI scores were divided into tertiles (low, moderate, and high risk) based on their distribution in the training cohort. Higher IMLCI tertiles were associated with an increased probability of severe IBD, further demonstrating the index’s potential for clinical decision support.

In addition, the IMLCI’s performance was contextualized relative to established disease activity indices such as the Mayo score and Crohn’s Disease Activity Index (CDAI). While these conventional tools primarily rely on clinical symptoms and endoscopic findings, the IMLCI integrates objective biochemical indicators reflecting systemic inflammation, hepatic metabolism, and immune dysregulation. Within this cohort, the IMLCI demonstrated a discrimination ability (AUC ≈ 0.90) comparable to or exceeding that of traditional indices typically ranging from 0.75 to 0.85.

3.7 Sensitivity analysis using the laboratory-inclusive definition

To assess the robustness of the IMLCI under alternative outcome specifications, a sensitivity analysis was performed using the original laboratory-inclusive definition of severe IBD, which incorporated CRP, WBC, ALT/AST, and albumin thresholds.

The IMLCI maintained similar discrimination (AUC = 0.91; HL P = 0.83) and calibration to the primary lab-free model, suggesting that model performance was not driven by circularity.

However, as expected, the lab-free definition produced slightly lower but more realistic discrimination (AUC = 0.89), confirming that excluding laboratory components from the outcome effectively prevented incorporation bias while preserving predictive validity. Subgroup comparisons based on available clinical records showed no significant difference in diabetes prevalence or laboratory markers between ulcerative colitis and Crohn’s disease, suggesting that the IMLCI may generalize across subtypes.

4 Discussion

This study confirmed that diabetes is a robust and independent risk factor for severe IBD and established the inflammation–metabolism–liver coupling index (IMLCI), which demonstrated excellent predictive performance. Patients with severe IBD exhibited pronounced systemic inflammation, impaired hepatic function, and metabolic dysregulation, all of which were integrated into the IMLCI framework. Compared with single biomarkers, the IMLCI achieved superior discrimination, calibration, and internal validation, underscoring its clinical potential for risk stratification in IBD.

Our findings are consistent with recent epidemiological evidence indicating that metabolic abnormalities increase the risk of adverse outcomes in IBD (18, 19). Over the past two decades, the global prevalence of diabetes has risen substantially, and the proportion of IBD patients with concomitant diabetes has also increased (20, 21). Previous studies have shown that diabetes is associated with higher hospitalization rates, increased need for surgical interventions, and greater risk of treatment failure among patients with IBD. The systemic proinflammatory state, endothelial dysfunction, and gut microbiota alterations characteristic of diabetes may synergistically accelerate IBD progression, corroborating the independent association observed in our cohort (20, 22). To minimize the potential incorporation bias inherent in earlier designs, we redefined severe IBD using only clinical endpoints independent of laboratory variables. This re-specification ensured that predictors such as CRP, WBC, and liver enzymes were not involved in outcome determination, thereby preventing circular associations. The model maintained strong discrimination (AUC ≈ 0.89) under this stricter framework, demonstrating genuine predictive validity.

In addition to diabetes, several inflammatory and hepatic biomarkers emerged as important predictors of severe IBD (23–25). Elevated white blood cell counts, neutrophil percentage, and CRP levels reflected heightened systemic inflammation, while reduced lymphocyte counts indicated compromised adaptive immunity (26). Similarly, abnormal liver enzymes and bilirubin levels suggested hepatocellular injury and bile metabolism dysfunction, both of which have been associated with increased disease activity and poor prognosis (27, 28).

Our findings provide a potential biological explanation for the observed association between diabetes and IBD severity. Chronic hyperglycemia and insulin resistance can amplify oxidative stress and promote the release of pro-inflammatory cytokines such as IL-6 and TNF-α, leading to sustained immune activation. Meanwhile, hepatic metabolic stress—reflected by elevated ALT, AST, and bilirubin—may disrupt bile acid signaling and lipid metabolism, impairing intestinal barrier function and promoting bacterial translocation.

Integrating inflammation, metabolism, and liver function into a composite index substantially enhanced the model’s predictive accuracy (29, 30). Traditional clinical scoring systems often depend on symptoms and endoscopic findings (31–35), which typically lag behind biochemical alterations. In contrast, the IMLCI is derived from objective laboratory markers that are routinely available, offering a pragmatic and reproducible tool for early risk identification in patients with IBD. In the revised analysis, the IMLCI achieved an AUC of 0.90 in the training cohort and 0.891 in the testing cohort, indicating excellent but realistic discrimination. This multidimensional approach is consistent with recent advances in machine learning–based models for chronic inflammatory diseases and underscores the value of integrating metabolic and hepatic biomarkers into inflammation-oriented prediction frameworks. Clinically, the IMLCI provides an objective, laboratory-based complement to symptom-driven indices such as the Mayo score or CDAI. By integrating inflammatory, metabolic, and hepatic biomarkers, it enables early recognition of patients at high risk of severe IBD who may benefit from closer monitoring or intensified treatment. This laboratory-based approach also offers a reproducible framework for tracking disease activity and treatment response over time.

5 Limitations and future perspectives

This study has several limitations. Its single-center, retrospective design may introduce selection bias and restrict generalizability. To minimize bias, standardized diagnostic criteria, uniform laboratory procedures, and random stratified sampling were applied. Internal validation using stratified splitting and 500 bootstrap iterations confirmed the stability of the IMLCI and limited overfitting. External multicenter validation remains essential, and future studies will prospectively validate and recalibrate the IMLCI for clinical application. Some clinically relevant variables—such as IBD subtype, disease duration, surgical history, and medication exposures—were inconsistently recorded and therefore excluded from the multivariable model. Nevertheless, the laboratory and metabolic parameters integrated in the IMLCI likely capture systemic disease burden and treatment effects. Future multicenter cohorts incorporating these clinical details will help refine and externally validate the model.

6 Conclusion

This study demonstrated that diabetes is a strong independent risk factor for severe IBD. By integrating inflammatory, metabolic, and hepatic biomarkers, the proposed inflammation–metabolism–liver coupling index (IMLCI) achieved superior predictive accuracy compared with conventional single indicators. The IMLCI provides a pragmatic and objective approach for early identification of high-risk patients, with potential value in guiding timely intervention and personalized disease management.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

Ethics statement

This study was conducted in accordance with the principles of the Declaration of Helsinki. The protocol was reviewed and approved by the Ethics Committee of Shanghai Sixth People’s Hospital, Shanghai Jiao Tong University School of Medicine (Approval No. 2024-KY-172(K), Date: November 15, 2024). Given the retrospective design, the requirement for informed consent was waived by the ethics committee. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

Author contributions

BH: Funding acquisition, Resources, Visualization, Writing – original draft, Writing – review & editing. HA: Funding acquisition, Resources, Visualization, Writing – original draft, Writing – review & editing. LC: Conceptualization, Methodology, Software, Writing – original draft. YQ: Data curation, Investigation, Methodology, Writing – original draft. YS: Data curation, Methodology, Software, Writing – original draft. SS: Data curation, Methodology, Software, Writing – original draft. HW: Data curation, Investigation, Software, Validation, Writing – original draft. ML: Conceptualization, Formal analysis, Project administration, Validation, Writing – original draft. LL: Formal analysis, Investigation, Validation, Writing – original draft. RW: Conceptualization, Formal analysis, Methodology, Validation, Writing – review & editing. HH: Funding acquisition, Resources, Software, Visualization, Writing – review & editing. WC: Funding acquisition, Resources, Validation, Visualization, Writing – review & editing.

Funding

The author(s) declare financial support was received for the research and/or publication of this article. This work was supported by the National Natural Science Foundation of China (No. 82204925), the Financial Project of Fujian Province (No. X2024004), the Natural Science Foundation of Fujian Province, China (No. 2024J01782, 2025J010039), and the Scientific and Technological Innovation Joint Funding Program (2024Y9496).

Acknowledgments

The authors thank all clinicians, nurses, and staff members involved in patient care and data collection at Shanghai Sixth People’s Hospital.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fendo.2025.1707507/full#supplementary-material

References

1. Bisgaard TH, Allin KH, Keefer L, Ananthakrishnan AN, and Jess T. Depression and anxiety in inflammatory bowel disease: epidemiology, mechanisms and treatment. Nat Rev Gastroenterol Hepatol. (2022) 19:717–26. doi: 10.1038/s41575-022-00634-6

PubMed Abstract | Crossref Full Text | Google Scholar

2. Bruner LP, White AM, and Proksell S. Inflammatory bowel disease. Prim Care. (2023) 50:411–27. doi: 10.1016/j.pop.2023.03.009

PubMed Abstract | Crossref Full Text | Google Scholar

3. Khanna R and Kuenzig ME. Global incidence and prevalence of inflammatory bowel disease over time. Gastroenterology. (2025). doi: 10.1053/j.gastro.2025.08.003

PubMed Abstract | Crossref Full Text | Google Scholar

4. Wang R, Li Z, Liu S, and Zhang D. Global, regional and national burden of inflammatory bowel disease in 204 countries and territories from 1990 to 2019: a systematic analysis based on the Global Burden of Disease Study 2019. BMJ Open. (2023) 13:e065186. doi: 10.1136/bmjopen-2022-065186

PubMed Abstract | Crossref Full Text | Google Scholar

5. Rodrigues VF, Elias-Oliveira J, Pereira ÍS, Pereira JA, Barbosa SC, Machado MSG, et al. Akkermansia muciniphila and gut immune system: A good friendship that attenuates inflammatory bowel disease, obesity, and diabetes. Front Immunol. (2022) 13:934695. doi: 10.3389/fimmu.2022.934695

PubMed Abstract | Crossref Full Text | Google Scholar

6. Zhang Y, Kang Q, He L, Chan KI, Gu H, Xue W, et al. Exploring the immunometabolic potential of Danggui Buxue Decoction for the treatment of IBD-related colorectal cancer. Chin Med. (2024) 19:117. doi: 10.1186/s13020-024-00978-y

PubMed Abstract | Crossref Full Text | Google Scholar

7. Wangchuk P, Yeshi K, and Loukas A. Ulcerative colitis: clinical biomarkers, therapeutic targets, and emerging treatments. Trends Pharmacol Sci. (2024) 45:892–903. doi: 10.1016/j.tips.2024.08.003

PubMed Abstract | Crossref Full Text | Google Scholar

8. Xu L, Shao Z, Fang X, Xin Z, Zhao S, Zhang H, et al. Exploring precision treatments in immune-mediated inflammatory diseases: Harnessing the infinite potential of nucleic acid delivery. Explor (Beijing). (2024) 5:20230165. doi: 10.1002/EXP.20230165

PubMed Abstract | Crossref Full Text | Google Scholar

9. Xu X, Guo Y, Chen M, Li N, Sun Y, Ren S, et al. Hypoglycemic activities of flowers of Xanthoceras sorbifolia and identification of anti-oxidant components by off-line UPLC-QTOF-MS/MS-free radical scavenging detection. Chin Herbal Med. (2024) 16:151–61. doi: 10.1016/j.chmed.2022.11.009

PubMed Abstract | Crossref Full Text | Google Scholar

10. Frumento D and Țălu Ș. Immunological linkages between inflammatory bowel diseases and type 2 diabetes. Biomedicines. (2025) 13:2224. doi: 10.3390/biomedicines13092224

PubMed Abstract | Crossref Full Text | Google Scholar

11. Bower JAJ, O’Flynn L, Kakad R, and Aldulaimi D. Effect of inflammatory bowel disease treatments on patients with diabetes mellitus. World J Diabet. (2021) 12:1248–54. doi: 10.4239/wjd.v12.i8.1248

PubMed Abstract | Crossref Full Text | Google Scholar

12. Bezzio C, Della Corte C, Vernero M, Di Luna I, Manes G, and Saibeni S. Inflammatory bowel disease and immune-mediated inflammatory diseases: looking at the less frequent associations. Therap Adv Gastroenterol. (2022) 15:17562848221115312. doi: 10.1177/17562848221115312

PubMed Abstract | Crossref Full Text | Google Scholar

13. Fuschillo G, Celentano V, Rottoli M, Sciaudone G, Gravina AG, Pellegrino R, et al. Influence of diabetes mellitus on inflammatory bowel disease course and treatment outcomes. A Syst Rev meta-anal Dig Liver Dis. (2023) 55:580–6. doi: 10.1016/j.dld.2022.08.017

PubMed Abstract | Crossref Full Text | Google Scholar

14. Lu J, Shen X, Li H, and Du J. Recent advances in bacteria-based platforms for inflammatory bowel diseases treatment. Explor (Beijing). (2024) 4:20230142. doi: 10.1002/EXP.20230142

PubMed Abstract | Crossref Full Text | Google Scholar

15. Meng Z, Li M, Wang X, Zhang K, Wu C, and Zhang X. Inula britannica ameliorates alcohol-induced liver injury by modulating SIRT1-AMPK/Nrf2/NF-κB signaling pathway. Chin Herb Med. (2024) 16:667–78. doi: 10.1016/j.chmed.2023.12.006

PubMed Abstract | Crossref Full Text | Google Scholar

16. Song Y, Chen C, and Li W. Ginsenoside Rb1 in cardiovascular and cerebrovascular diseases: A review of therapeutic potentials and molecular mechanisms. Chin Herb Med. (2024) 16:489–504. doi: 10.1016/j.chmed.2024.09.006

PubMed Abstract | Crossref Full Text | Google Scholar

17. Jiang X, Wang H, Nie K, Gao Y, Chen S, Tang Y, et al. Targeting lipid droplets and lipid droplet-associated proteins: a new perspective on natural compounds against metabolic diseases. Chin Med. (2024) 19:120. doi: 10.1186/s13020-024-00988-w

PubMed Abstract | Crossref Full Text | Google Scholar

18. Zhu W, Yan M, Cao H, Zhou J, and Xu Z. Effects of clostridium butyricum capsules combined with rosuvastatin on intestinal flora, lipid metabolism, liver function and inflammation in NAFLD patients. Cell Mol Biol (Noisy-le-grand). (2022) 68:64–9. doi: 10.14715/cmb/2022.68.2.10

PubMed Abstract | Crossref Full Text | Google Scholar

19. Saez A, Gomez-Bris R, Herrero-Fernandez B, Mingorance C, Rius C, and Gonzalez-Granado JM. Innate lymphoid cells in intestinal homeostasis and inflammatory bowel disease. Int J Mol Sci. (2021) 22:7618. doi: 10.3390/ijms22147618

PubMed Abstract | Crossref Full Text | Google Scholar

20. Dubey RK and Shukla S. Exploring novel herbal compounds and formulations for inflammatory bowel disease (IBD) management. Zhongguo Ying Yong Sheng Li Xue Za Zhi. (2023) 39:e20230003. doi: 10.62958/j.cjap.2023.003

PubMed Abstract | Crossref Full Text | Google Scholar

21. Francis KL, Alonge KM, Pacheco MC, Hu SJ, Krutzsch CA, Morton GJ, et al. Diabetes exacerbates inflammatory bowel disease in mice with diet-induced obesity. World J Gastroenterol. (2023) 29:4991–5004. doi: 10.3748/wjg.v29.i33.4991

PubMed Abstract | Crossref Full Text | Google Scholar

22. Su W, Yang Y, Zhao X, Cheng J, Li Y, Wu S, et al. Potential efficacy and mechanism of eight mild-natured and bitter-flavored TCMs based on gut microbiota: A review. Chin Herbal Med. (2024) 16:42–55. doi: 10.1016/j.chmed.2023.08.001

PubMed Abstract | Crossref Full Text | Google Scholar

23. Jia R, Zheng H, Li S, Chen W, Yang Y, Wu H, et al. QingChang-XiaoPi decoction ameliorates intestinal inflammation of ulcerative colitis by regulating the pathogenicity of Th17 cells. Phytomedicine. (2024) 132:155779. doi: 10.1016/j.phymed.2024.155779

PubMed Abstract | Crossref Full Text | Google Scholar

24. Gong XX, Cao LH, Ni HX, Zang ZY, Chang H, et al. Chinese herbal medicine for the treatment of diabetic nephropathy: From clinical evidence to potential mechanisms. J Ethnopharmacol. (2024) 330:118179. doi: 10.1016/j.jep.2024.118179

PubMed Abstract | Crossref Full Text | Google Scholar

25. Xu T, He P, namWangdu S, et al. Revealing the improvement of diabetes by Si Wei Jiang Huang Tang San through ERK/HIF1α signaling pathway via network pharmacology. J Ethnopharmacol. (2024) 319:117254. doi: 10.1016/j.jep.2023.117254

PubMed Abstract | Crossref Full Text | Google Scholar

26. Abulaiti K, Aikepa M, Ainaidu M, Wang J, Yizibula M, and Aikemu M. Metabolomics combined with network pharmacology reveals anti-asthmatic effects of Nepeta bracteata on allergic asthma rats. Chin Herb Med. (2024) 16:599–611. doi: 10.1016/j.chmed.2024.02.001

PubMed Abstract | Crossref Full Text | Google Scholar

27. Massironi S, Viganò C, Palermo A, Pirola L, Mulinacci G, Allocca M, et al. Inflammation and malnutrition in inflammatory bowel disease. Lancet Gastroenterol Hepatol. (2023) 8:579–90. doi: 10.1016/S2468-1253(23)00011-0

PubMed Abstract | Crossref Full Text | Google Scholar

28. Li Y, Li F, Ding M, et al. Chuanxiong Rhizoma extracts prevent liver fibrosis via targeting CTCF-c-MYC-H19 pathway. Chin Herb Med. (2023) 16:82–93. doi: 10.1016/j.chmed.2023.07.003

PubMed Abstract | Crossref Full Text | Google Scholar

29. Saez A, Herrero-Fernandez B, Gomez-Bris R, Sánchez-Martinez H, and Gonzalez-Granado JM. Pathophysiology of inflammatory bowel disease: innate immune system. Int J Mol Sci. (2023) 24:1526. doi: 10.3390/ijms24021526

PubMed Abstract | Crossref Full Text | Google Scholar

30. Dorner H, Stolzer I, Mattner J, et al. Gut pathobiont-derived outer membrane vesicles drive liver inflammation and fibrosis in primary sclerosing cholangitis-associated inflammatory bowel disease. Gastroenterology. (2024) 167:1183–1197.e16. doi: 10.1053/j.gastro.2024.06.032

PubMed Abstract | Crossref Full Text | Google Scholar

31. Zhang H, Zhao S, and Cao Z. Impact of Epstein-Barr virus infection in patients with inflammatory bowel disease. Front Immunol. (2022) 13:1001055. doi: 10.3389/fimmu.2022.1001055

PubMed Abstract | Crossref Full Text | Google Scholar

32. Ben Nasr M, Frumento D, and Fiorina P. Adipose stem cell therapy for chronic pancreatitis. Mol Ther. (2017) 25:2438–9. doi: 10.1016/j.ymthe.2017.10.007

PubMed Abstract | Crossref Full Text | Google Scholar

33. Frumento D, Ben Nasr M, El Essawy B, D’Addio F, Zuccotti GV, and Fiorina P. Immunotherapy for type 1 diabetes. J Endocrinol Invest. (2017) 40:803–14. doi: 10.1007/s40618-017-0641-y

PubMed Abstract | Crossref Full Text | Google Scholar

34. Li A, Niu Y, Xu W, Liang B, Zhang R, Mei R, et al. Sophora Flos extract ameliorates inflammatory bowel disease via inhibiting the PI3K/AKT/mTOR signaling pathway activation and modulating M1/M2 polarization in macrophages. Phytomedicine. (2025) 146:157095. doi: 10.1016/j.phymed.2025.157095

PubMed Abstract | Crossref Full Text | Google Scholar

35. Feng Z, Jiao L, Wu Z, et al. A novel nanomedicine ameliorates acute inflammatory bowel disease by regulating macrophages and T-cells. Mol Pharm. (2021) 18:3484–95. doi: 10.1021/acs.molpharmaceut.1c00415

PubMed Abstract | Crossref Full Text | Google Scholar