Interaction between fatty pancreas disease and genetically predicted glucose-dependent insulinotropic polypeptide on incident type 2 diabetes: evidence from the UK Biobank.

PURPOSE: Glucose-dependent insulinotropic polypeptide (GIP) plays a crucial role in lipid metabolism. The effect of GIP on pancreatic lipid and whether it modulates risk of type 2 diabetes (T2D) associated with fatty pancreas remain unknown. The aim of this study was to investigate the interaction between genetically predicted GIP levels and fatty pancreas in the development of T2D.

METHODS

This is a large-scale cohort study using data from the UK Biobank. Participants of White ethnicity without diabetes at the imaging visit were included in the analysis. The loss-of-function GIPR variant E354Q was used for the prediction of fasting GIP levels, and a polygenic risk score (PRS) of postprandial GIP was used for the prediction of postprandial GIP levels. The presence of fatty pancreas disease (FPD) was determined with magnetic resonance imaging (MRI). Diagnosis of T2D was ascertained based on ICD10-CM diagnosis code E11. During a median follow-up of 51 months, 276 cases of incident T2D were identified.

RESULTS

A significant interaction between the carrying status of E354Q and FPD (p for interaction = 0.018) and between 2hGIP PRS and FPD (p for interaction = 0.015) in the development of T2D was found. FPD is associated with a greater increase of risk of T2D in individuals without E354Q [hazard ratio (HR) 2.44, 95% confidence interval (CI) 1.78-3.34] or with higher levels of genetically predicted postprandial GIP (HR 2.64, 95% CI 1.86-3.76).

CONCLUSION

Our findings show that genetically predicted GIP modifies risk of T2D associated with FPD, suggesting that GIP may play a role in linking pancreatic fat accumulation to metabolic dysfunction. These findings are derived from genetically predicted rather than measured GIP.