JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Latent class analysis identifies novel coeliac disease subgroups with distinctive clinical features: a multicentric study

  • 0Department of Internal Medicine and Medical Therapeutics, University of Pavia, Pavia, Italy; First Department of Internal Medicine, San Matteo Hospital Foundation, Pavia, Italy.
European Journal of Internal Medicine +

|

June 1, 2025

|

June 1, 2025

View abstract on PubMed

Summary

This summary is machine-generated.

Latent class analysis identified four distinct subtypes of coeliac disease (CD), challenging traditional classifications. This data-driven approach refines understanding of CD phenotypes for better patient stratification.

Area Of Science

  • Gastroenterology
  • Immunology
  • Data Science

Background

  • Coeliac disease (CD) is a complex immune-mediated disorder with a broad clinical spectrum.
  • Existing classifications, such as the Oslo definitions, may not fully capture the heterogeneity of CD.
  • Novel phenotyping approaches are needed to better understand and manage CD subtypes.

Purpose Of The Study

  • To identify novel subtypes of coeliac disease (CD) based on clinical features using latent class analysis (LCA).
  • To validate these identified CD phenotypes using a predictive supervised model.
  • To compare the novel CD subtypes with the traditional Oslo classification.

Main Methods

  • A multicentric retrospective study analyzed 2478 adult CD patients from 19 Italian centers (2011-2021).
  • Latent class analysis (LCA) was applied to categorical symptom variables (gastrointestinal, hematological, neuropsychiatric, fatigue).
  • Multinomial logistic regression (MLR) validated associations between latent classes and additional clinical features.

Main Results

  • LCA identified four distinct CD classes: lower GI symptoms, upper GI manifestations, asymptomatic/nonspecific, and anemia/asthenia.
  • A partial overlap was observed between LCA-derived classes and the Oslo classification, suggesting potential misclassification.
  • Female sex, autoimmune comorbidities, and severe histological damage were significantly associated with the upper GI manifestation class (Class 2).

Conclusions

  • Latent class analysis (LCA) provides a novel, data-driven method for refining coeliac disease (CD) phenotyping.
  • The identified CD subtypes offer a more nuanced classification than traditional criteria.
  • Further validation of these LCA-derived CD phenotypes in future studies is warranted.

Keywords: