Risk factors and prediction model of renal dysfunction in patients with inflammatory bowel disease: a retrospective cohort study.

BACKGROUND

Previous studies have identified the association between inflammatory bowel disease (IBD) and its renal complications, but specific risk factors and predictive models remain underexplored.

AIMS: To identify the risk factors for renal dysfunction in IBD patients and to develop a clinical model for the early detection of high-risk individuals.

METHODS

This retrospective cohort study enrolled 769 patients diagnosed with IBD from a tertiary hospital for model training and internal validation, and 298 patients from four other hospitals for external validation. A Cox regression model incorporating seven selected variables, identified through LASSO regression, was developed to determine independent risk factors, and its performance was subsequently validated in both internal and external cohorts.

RESULTS

Multivariate analysis identified seven independent predictors of renal insufficiency [estimated glomerular filtration rate (eGFR) < 90 mL/min/1.73 m2], including osteoporosis (HR: 1.77; P = 0.024), baseline eGFR (HR: 0.29; P < 0.001), baseline high-density lipoprotein (HDL) (HR: 0.52; P = 0.033), baseline C-reactive protein (CRP) (HR: 1.00; P = 0.09), biologic therapy (HR: 0.57; P = 0.011), enteral nutrition (HR: 1.86; P = 0.007), and intestinal surgery (HR: 1.64; P = 0.01). A predictive model was developed with a C-index of 0.719 for predicting renal insufficiency and 0.739 for predicting chronic kidney disease (eGFR < 60 mL/min/1.73 m2).

CONCLUSIONS

Predictors of renal dysfunction in IBD patients included osteoporosis, eGFR, HDL, CRP, biologic therapy, enteral nutrition, and intestinal surgery. Our predictive model can effectively identify IBD populations at high risk for renal dysfunction at an early stage.