Introduction

Emerging evidence suggests that poor oral health may play a role in gastrointestinal carcinogenesis through inflammatory and microbiome-mediated mechanisms. However, its association with esophageal cancer remains insufficiently explored.

Aims & Methods

We aimed to assess the association of oral health indicators and oral hygiene behaviors with esophageal cancer risk using both dental health screening and claims-based data. We conducted a population-based case-control study using the Korean National Health Insurance Service (NHIS) database and the National Health Screening (NHS) records. To enhance the reliability of our study, we examined oral hygiene from two complementary perspectives: (1) oral health assessed through dental examinations and questionnaires in the NHS database and (2) periodontal disease status determined based on a claim-based analysis of the NHIS database. Individuals newly diagnosed with esophageal cancer were identified and matched at a 1:4 ratio to cancer-free controls through rigorous propensity score (PS) matching based on sociodemographic factors, comorbidities, lifestyle behaviors, and laboratory parameters. Oral health variables were assessed using national dental screening program data (status of decayed, missing, and filled teeth; gingivitis severity; dental calculus; and self-reported oral hygiene practices) and NHIS claims data identifying periodontal disease using ICD-10 codes and procedure codes. Multivariable-adjusted conditional logistic regression analyses were performed to estimate adjusted odds ratios (aORs) with 95% confidence intervals (CIs).

Results

After PS matching, the oral health analysis included 4,238 cancer cases and 16,904 controls, while the claim-based periodontal disease analysis comprised 18,905 cancer cases and 75,604 matched controls in the final analysis. In multivariable analysis, missing teeth (adjusted OR, 1.16; 95% CI, 1.08–1.25) and inadequate oral hygiene behaviors—including brushing fewer than three times daily (aOR, 1.14; 95% CI, 1.07–1.23), infrequent toothbrushing before bedtime (aOR, 1.08; 95% CI, 1.00–1.16), infrequent use interdental cleaning tools (aOR, 1.10; 95% CI, 1.01–1.19), and dental visits in the past year (aOR, 1.09; 95% CI, 1.02–1.16)—were significantly associated with an increased risk of esophageal cancer. Additionally, claim-based periodontal disease was independently associated with elevated risk (aOR, 1.05; 95% CI, 1.01–1.09). Subgroup analyses revealed consistent associations across smoking and obesity strata, while sex-, age-, and alcohol-stratified analyses demonstrated differential patterns.

Conclusion

This study provides comprehensive, population-based evidence that both clinically assessed dental health indicators and behavioral oral hygiene practices are associated with esophageal cancer risk. The integration of screening-based and claims-based data enhances the validity of the findings and highlights oral health as a potential modifiable risk factor for esophageal cancer prevention.

Introduction

Inflammatory Bowel Disease (IBD)^1,2 requires close monitoring to detect flare ups early and guide treatment. While endoscopy is the gold standard, it is invasive, costly, and uncomfortable^1,2. Intestinal Ultrasound (IUS) is a non-invasive alternative with strong correlation to endoscopic findings—especially bowel wall thickness (BWT), a marker of disease activity³. Nevertheless, IUS remains limited due to operator dependence and a steep learning curve⁴.
AI has improved consistency in other domains, e.g. endoscopic evaluation, but its application to IUS in IBD remains underexplored. Prior models lack interpretability and require manual image cropping, or are trained on ideal, selected data, limiting clinical use^5,6.

Aims & Methods

We aim to develop a deep-learning model that automatically identifies and paints the bowel wall and measures BWT directly from raw clinical IUS images.
A training dataset of 570 images from 144 IUS videos and a testing dataset of 55 images from 55 separate exams were created. All images were extracted from previously performed IUS examinations reflecting real-world variation, including e.g. unclear boundaries and artefacts.
All images were annotated by International Bowel Ultrasound Group (IBUS) certified experts, including outline-paintings of the inner and outer bowel wall and two BWT measurements.
The AI consists of a combination of Convolutional neural networks and other image processing algorithms.
Evaluation included BWT error against the expert mean, classification accuracy using the standard IBUS 3 mm threshold³, and a leave-one-out comparison with individual doctors.

Results

The model produced predictions on 54 of 55 test images.
The model deviated from the gold standard mean by 0.98 mm (SD 1.1 mm) per image on the regression task. The average distance to expert-defined bounds was 0.44 mm (SD 0.89 mm), with 59% of predictions inside this range.
For classification (using a 3 mm threshold), the model reached an accuracy of 0.77, sensitivity of 0.69, specificity of 0.94, and Cohen’s Kappa of 0.56.
In the leave-one-out analysis, expert performance ranged from 0.89–0.93 accuracy and 0.79–0.83 Kappa. Depending on which expert was excluded from the gold standard, the model achieved 0.74–0.80 accuracy and 0.49–0.60 Kappa. Experts stayed within expert-defined bounds in 72–81% of cases, while the model did so in 46–55%.

Conclusion

Clinically, the model performed well. An error of 0.5–1 mm is negligible in practice and matches typical variation in manual measurements. Many real-world test images had BWT values near the 3 mm threshold, so small deviations led to misclassifications. In practice, a 2.5 mm reading could still raise concerns based on symptoms³.
While the system alone cannot provide expert-level BWT measurements, it could provide assistance to experts as well as non-expert and junior doctors, especially in locating the bowel, which is an essential part of IUS.
The model works on unprocessed clinical data, with the only selection criteria being that there is an identifiable bowel segment, measurable by an IBUS-certified doctor. Allowing the images to reflect realistic conditions
In summary, we developed an AI model that identifies the bowel and measures BWT with acceptable accuracy. We have already started the processes of collecting video data, to extend the AI’s functionality to a fully clinical setting.

References

1. Torres J, Mehandru S, Colombel JF, Peyrin-Biroulet L. Crohn’s disease. The Lancet. 2017;389(10080):1741-1755. doi:10.1016/S0140-6736(16)31711-1
2. Ungaro R, Mehandru S, Allen PB, Peyrin-Biroulet L, Colombel JF. Ulcerative colitis. Lancet. 2017;389(10080):1756-1770. doi:10.1016/S0140-6736(16)32126-2
3. Novak KL, Group I, Nylund K, et al. Expert Consensus on Optimal Acquisition and Development of the International Bowel Ultrasound Segmental Activity Score [IBUS-SAS]: A Reliability and Inter-rater Variability Study on Intestinal Ultrasonography in Crohn’s Disease. J Crohns Colitis. 2021;15(4):609-616. doi:10.1093/ECCO-JCC/JJAA216
4 Bove L, Meyer J, Collins M, Frampas E, Bourreille A, Le Berre C. Understanding the learning curve of intestinal ultrasound in inflammatory bowel disease: A comparative study between novice, regular, and expert. Clin Res Hepatol Gastroenterol. 2025;49(3):102548. doi:10.1016/J.CLINRE.2025.102548
5. Kumaralingam L, Le May K, Dang VB, Alavi J, Huynh HQ, Le LH. Artificial intelligence-assisted approach to assessing bowel wall thickness in pediatric inflammatory bowel disease using intestinal ultrasound images. J Crohns Colitis. 2025;19(4). doi:10.1093/ECCO-JCC/JJAF037
6. Carter D, Albshesh A, Shimon C, et al. Automatized Detection of Crohn’s Disease in Intestinal Ultrasound Using Convolutional Neural Network. Inflamm Bowel Dis. February 2023:1-6. doi:10.1093/IBD/IZAD014

Introduction

Inflammatory bowel disease (IBD), comprising Crohn’s disease (CD) and ulcerative colitis (UC), is part of a broader spectrum of immune-mediated inflammatory diseases (IMIDs). Clinical and epidemiological studies indicate a higher prevalence of IMIDs among patients with IBD and their first-degree relatives, with a significant proportion of patients with IBD also diagnosed with conditions such as asthma, psoriasis, and rheumatoid arthritis^1,2. These patterns suggest shared genetic and/or environmental risk factors. The clinical intersections call for a multidisciplinary management strategy in which decisions account for the broader immune landscape rather than focusing on each diagnosis in isolation^1,3.

Aims & Methods

This study aimed to characterize the relationship between IBD and other IMIDs and disentangle genetic and environmental contributions. This was done by analyzing the tetrachoric correlation⁴ between disease codes registered for 107,651 full sibling pairs born between 1910 and 2010 in the Danish nationwide health registries. We then estimated SNP-based correlations based on summary statistics from genome-wide association studies of each disease using LDSC⁵. Lastly, we downloaded and reanalyzed all V3-V4 16S rDNA fecal microbiota samples available from NCBI⁶ in a unified bioinformatical pipeline. We used this data to generate microbiota disease profiles and estimate their correlations across IBD and IMIDs.

Results

We show that CD and UC have distinct patterns of correlations with other IMIDs. Asthma and celiac disease were correlated with CD (asthma: ρ=0.12, 95% CI=0.10-0.13, celiac disease: ρ=0.12, 95% CI=0.08-0.17), but not UC within full siblings. Oppositely, sarcoidosis (ρ=0.16, 95% CI=0.13-0.19), multiple sclerosis (ρ=0.09, 95% CI=0.06-0.12), systemic lupus erythematosus (SLE) (ρ=0.19, 95% CI=0.13-0.25), and Sjögren’s syndrome (ρ=0.15, 95% CI=0.10-0.20) correlated with UC but not CD. The IMIDs Graves’ disease, type 1 diabetes, iridocyclitis, psoriasis, psoriatic arthritis, rheumatoid arthritis, and ankylosing spondylitis correlated with both CD and UC. Evaluating genetic and gut microbial correlations highlighted that UC exhibits microbiota-linked correlations with multiple sclerosis (ρ=0.11, 95% CI=0.01-0.21) and SLE (ρ=0.15, 95% CI=0.05-0.24) despite an estimated negative genetic overlap (multiple sclerosis: ρ=-0.18, 95% CI=-0.32--0.04, SLE: ρ=-0.21, 95% CI=-0.32--0.09), suggesting a key role for environmental factors, including microbial. In contrast, for both IBD subtypes rheumatoid arthritis showed consistent genetic (CD: ρ=0.12, 95% CI=0.02-0.21, UC: ρ=0.17, 95% CI=0.08-0.27) as well as microbial convergence (CD: ρ=0.30, 95% CI=0.20-0.38, UC: ρ=0.12, 95% CI=0.02-0.22), whereas the correlation with psoriasis appeared to be mainly genetically driven (CD SNP correlation: ρ=0.23, 95% CI=0.14-0.33, CD microbiota correlation: ρ=-0.13, 95% CI=-0.23--0.03, UC SNP correlation: ρ=0.25, 95% CI=0.15-0.35, UC microbiota correlation: ρ=0.06, 95% CI=-0.04-0.16).

Conclusion

These findings uncover heterogeneity in shared etiological pathways across immune diseases, underscoring the need for stratified approaches to diagnosis and therapeutic intervention in IBD. Importantly, our findings demonstrate that CD and UC—though often co-studied—harbor distinct immunogenetic and microbial associations, reinforcing the need for subtype-specific approaches in clinical management and research design. Further, the correlation between IBD and other IMIDs appears to be driven by different genetic and environmental mechanisms, which must be further explored.

References

1. Bezzio C, Corte C Della, Vernero M, et al. Inflammatory bowel disease and immune-mediated inflammatory diseases: looking at the less frequent associations. Therap Adv Gastroenterol 2022;15. Available at: https://journals.sagepub.com/doi/full/10.1177/17562848221115312 [Accessed March 14, 2025].
2. García MJ, Pascual M, Pozo C Del, et al. Impact of immune-mediated diseases in inflammatory bowel disease and implications in therapeutic approach. Sci Rep 2020;10:10731. Available at: https://pmc.ncbi.nlm.nih.gov/articles/PMC7330038/ [Accessed March 14, 2025].
3. Park SW, Kim TJ, Lee JY, et al. Comorbid immune-mediated diseases in inflammatory bowel disease: a nation-wide population-based study. Aliment Pharmacol Ther 2019;49:165–172. Available at: https://pubmed.ncbi.nlm.nih.gov/30506945/ [Accessed March 14, 2025].
4. Falconer DS, Mackay TFC. Introduction to Quantitative Genetics. Third Ddit. Longman Scientific & Technical; 1989.
5. Bulik-Sullivan B, Finucane HK, Anttila V, et al. An atlas of genetic correlations across human diseases and traits. Nature Genetics 2015 47:11 2015;47:1236–1241. Available at: https://www.nature.com/articles/ng.3406 [Accessed March 14, 2025].
6. Sayers EW, Bolton EE, Brister JR, et al. Database resources of the national center for biotechnology information. Nucleic Acids Res 2022;50:D20–D26. Available at: https://pubmed.ncbi.nlm.nih.gov/34850941/ [Accessed March 14, 2025].

Disclosure

TJ reports consultancy for Ferring and Pfizer. The remaining authors have no disclosures.