Objective:

To evaluate the predictive value of non-insulin-based surrogate markers of insulin resistance (IR) for colorectal cancer (CRC) risk assessment.

Approach:

• Study Design: Prospective study analyzing data from 385,206 participants in the UK Biobank.
• Markers Evaluated: Non-insulin-based surrogate markers included TyG index, TyG-BMI, TG/HDL-C ratio, and METS-IR.
• Statistical Analysis: Cox proportional hazards models estimated hazard ratios (HRs) and 95% confidence intervals (CIs) for incident CRC.
• Follow-Up: Median follow-up of 13.8 years with 5,175 incident CRC cases documented.
• Predictive Value Assessment: Incremental predictive value and clinical utility evaluated via C-index increments and decision curve analysis (DCA).

Key Findings:

• All four IR surrogates were independently and positively associated with CRC risk (all P < 0.001).
• HRs (95% CIs) for extreme quartiles were 1.16 (1.07-1.26) for TyG index, 1.20 (1.11-1.31) for TyG-BMI, 1.09 (1.01-1.19) for TG/HDL-C ratio, and 1.21 (1.11-1.32) for METS-IR.
• METS-IR and TyG-BMI showed the most significant improvements in discrimination (C-index increments = 0.0015; P < 0.001).
• Significant interactions between IR surrogate markers and age, sex, and smoking status were observed (P-interaction ≤ 0.05).

Interpretation:

Elevated non-insulin-based IR surrogates are robust predictors of CRC risk.

Limitations:

• The study may not account for all potential confounding factors influencing CRC risk.
• Findings are based on a specific population from the UK Biobank, which may limit generalizability.

Conclusion:

The study indicates that non-insulin-based tools for assessing IR are associated with CRC risk.