A metabolic–inflammatory burden phenotype associated with urinary glucose in colorectal cancer

Background Metabolic dysregulation and chronic inflammation are frequently observed in individuals with colorectal cancer (CRC), particularly in the context of diabetes-related conditions. Identifying simple clinical indicators that reflect these combined alterations remains of interest. Urinary glucose, routinely assessed in clinical practice, may capture transient metabolic stress, but its association with integrated metabolic–inflammatory characteristics in CRC has not been systematically evaluated. Methods A hospital-based cross-sectional case–control analysis was conducted, including individuals with confirmed colorectal cancer (CRC) and non-CRC controls undergoing clinical evaluation during the same period. A composite Metabolic–Inflammatory Burden Score (MIBS) was constructed using urinary glucose status together with selected inflammatory and tumor markers. Multivariable logistic regression was performed with CRC status (1 = CRC, 0 = non-CRC) as the dependent variable, and model performance was assessed in the primary cohort. External validation was performed in an independent NHANES subset using a reduced model consistent with the variables available in that dataset. Results Individuals with higher urinary glucose categories exhibited higher levels of systemic inflammation and tumor-related markers, along with altered immune cell profiles. Urinary glucose remained associated with CRC after adjustment for demographic, metabolic, inflammatory, and molecular factors. Incorporation of urinary glucose into the composite framework was associated with improved model discrimination in the primary analysis, and similar patterns were observed in the external NHANES cohort, including differences in predicted risk distributions across urinary glucose categories. Conclusion Urinary glucose was associated with distinct metabolic–inflammatory characteristics in CRC and contributed to a composite burden framework that demonstrated consistent patterns in an independent population-based dataset.