Associations between triglyceride-glucose related indices and the risk of incident pancreatic cancer: a large-scale prospective cohort study in the UK Biobank.

Abstract

Background: Triglyceride-Glucose (TyG) index, a novel surrogate marker for insulin resistance, has been linked to the risk of various cancers, such as breast and colorectal cancers. However, the relationship between the TyG index and its related indices with the risk of pancreatic cancer(PC) remains unclear.

Methods: This large-scale, prospective cohort study utilized data from the UK Biobank, involving 428,152 participants who were free of PC at baseline. The primary outcome was incident PC. To evaluate the relationships between TyG-related indices and PC onset, covariate-adjusted Cox proportional hazards regression models and restricted cubic spline (RCS) analyses were employed. Additionally, we explored the relationships between TyG-related indices and incident PC within subgroups defined by age, gender, diabetes, hypertension, pancreatitis history, smoking status, and drinking status. Sensitivity analyses were conducted to confirm the robustness of our findings.

Results: Over an average follow-up period of 13.8 years, a total of 1,759 PC cases occurred. The risk of developing PC increased with higher levels of TyG-related indices. RCS curves illustrated a linear positive relationship between TyG-related indices and incident PC. Subgroup analyses indicated that the relationships between the TyG index, TyG-waist circumference, and TyG-waist-to-hip ratio and PC risk were more pronounced in participants without hypertension, without pancreatitis history, and those with drinking history (P for interaction < 0.05). Sensitivity analyses further validated the robustness of these findings.

Conclusions: TyG-related indices were independently related with a higher risk of incident PC, highlighting the importance of incorporating these indices into PC risk assessment tools and provides strong support for constructing a more comprehensive PC risk assessment framework.