Risk Factors Analysis and Predictive Model Construction for Autoimmune Gastritis: A Nationwide Multicenter Case–Control Study in China

Objective

Here, we ascertained the clinical characteristics of Chinese patients with autoimmune gastritis (AIG) and determined the correlation of dietary and lifestyle factors with AIG occurrence and development to establish a noninvasive predictive model for AIG.

Methods

In this case–control study, we enrolled 479 patients from seven independent centers nationwide in China; of them, 279 had AIG, 112 had chronic atrophic gastritis mostly in the antrum, and 88 had chronic nonatrophic gastritis. Their clinical and lifestyle data were systematically collected and analyzed. Finally, a multivariate logistic regression disease prediction model was then established and validated.

Results

Most of the 279 patients with AIG were middle-aged, older, and female. In the predictive model of AIG, the larger amount of cooking oil used per meal and comorbid autoimmune thyroid disease was considered risk factors, and a diet rich in vitamin B12 was considered a protective factor. We plotted a receiver operating characteristic (ROC) curve of the model in the discovery and validation cohorts, and the areas under the ROC curves were 0.72 and 0.74, respectively. In addition, dietary structure, eating habits, sleep quality, and smoking status were noted to be correlated with the occurrence of gastrointestinal symptoms and complications, as well as histopathological grades of AIG.

Conclusion

Dietary and lifestyle factors may predict AIG risk in Chinese populations and were related to AIG prognosis.