Prediction Models for Celiac Disease Development in Children From High-Risk Families: Data From the PreventCD Cohort

Area of Science:

• Pediatric Gastroenterology
• Genetics
• Immunology

Background:

• Celiac disease (CD) screening is recommended for children with affected first-degree relatives (FDR).
• Optimal screening frequency and age remain undetermined.
• This study aimed to identify CD risk factors and develop predictive models for personalized screening strategies.

Purpose of the Study:

• To identify variables influencing celiac disease (CD) risk in genetically predisposed children.
• To develop and validate clinical prediction models for individualized CD screening advice.
• To create a risk score and prediction tool for personalized CD risk assessment.

Main Methods:

• Prospective analysis of the PreventCD-birth cohort (944 children) over 10 years.
• Development of prediction models using multivariable Cox regression and backward elimination.
• Validation of models using the independent NeoCel cohort, assessing discrimination with Harrell's c-index.

Main Results:

• 135/944 children developed CD, with a mean age of 4.3 years.
• CD risk was significantly higher in girls and Human Leukocyte Antigen (HLA)-DQ2 homozygous individuals.
• Prediction models demonstrated good fit in the validation cohort (Cox regression 0.81).

Conclusions:

• Celiac disease (CD) develops early in genetically predisposed children.
• Risk is influenced by gender, age, and HLA-DQ type, crucial for screening advice.
• An online tool, Prediction, is available for personalized screening recommendations.