Metabolic and Laboratory Biomarkers in Early-Onset Versus Late-Onset Colorectal Cancer: A Case-Control Study

Background: The incidence of early-onset colorectal cancer (EOCRC) is rising, yet the relative contribution of metabolic, inflammatory, and laboratory abnormalities remains incompletely defined. Objectives: We compared these associations between EOCRC and late-onset colorectal cancer (LOCRC) while addressing the possibility that some laboratory abnormalities may reflect occult cancer rather than antecedent risk. Methods: We conducted a matched case-control study using the TriNetX US Network. Adults diagnosed with CRC between 2010 and 2023 were identified as EOCRC (18-49 years) or LOCRC (50-75 years). Patients with prior malignancy, inflammatory bowel disease, hereditary or familial CRC risk, or prior colectomy were excluded. Three separate analyses were performed. First, a direct EOCRC-versus-LOCRC comparison evaluated gastrointestinal symptoms during the 6 months preceding diagnosis. Second, EOCRC and LOCRC were each compared with their respective matched cancer-free controls to assess clinical, metabolic, and laboratory features during the 24 months preceding diagnosis. When multiple laboratory values were available, the most recent value preceding the index date was used. Conditional logistic regression estimated adjusted odds ratios with 95% confidence intervals, with Bonferroni correction applied for multiple comparisons. Results: The direct matched EOCRC-versus-LOCRC comparison included 7752 patients with CRC, comprising 2584 with EOCRC and 5168 with LOCRC. EOCRC more frequently presented with rectal bleeding, abdominal pain, diarrhea, iron-deficiency anemia, and weight loss. Rectal tumors were more common in EOCRC, whereas proximal tumors were more common in LOCRC. In separate control-based analyses, 3217 patients with EOCRC and 12,112 patients with LOCRC were compared with 6434 and 24,336 matched cancer-free controls, respectively. The strongest independent features associated with EOCRC were severe obesity (aOR 2.61), microcytosis (aOR 2.29), low ferritin (aOR 2.11), and elevated C-reactive protein (aOR 1.87). Similar but generally attenuated associations were observed in LOCRC. In adjusted EOCRC-versus-LOCRC analyses, obesity (aOR 1.38), metabolic syndrome (aOR 1.41), and MASH (aOR 1.22) remained more closely associated with EOCRC. Conclusions: EOCRC is associated with a distinct clinical-metabolic phenotype, with more pronounced metabolic, inflammatory, and hematologic abnormalities than LOCRC. These findings should be interpreted as hypothesis-generating prediagnostic associations, not as validated predictors or causal risk factors.