Cancer
Understanding Colorectal Cancer Etiology: The Paradigm III Conceptual Model Tanvi Srivastava* Tanvi Srivastava Srivastava Srivastava Srivastava Srivastava Srivastava Srivastava Srivastava Srivastava UCSF
Colorectal cancer (CRC) is the second leading cause of cancer-related death in the US. By 2030, ~25% of CRC cases are projected to occur in individuals under age 50. To better understand the multifactorial etiology of CRC and early-onset CRC (EO-CRC), we developed a dynamic conceptual model to guide research efforts and improve risk communication to the public.
Building on the Paradigm II breast cancer etiology model, we conducted a literature review to identify risk factors for CRC and EO-CRC. Each factor was assessed for data quality, strength of evidence, and potential interactions, and classified within four domains: biologic, behavioral, social, and environmental. We graded the strength of association as strong (RR > 2.0), modest (RR = 1.5–2.0), or weak (RR < 1.5), and rated overall evidence quality as high, medium, or low according to US Preventive Services Task Force criteria, considering study design, internal validity, consistency, precision, directness and applicability to the target population.
We identified 31 factors of potential etiologic importance across the four domains associated with CRC and EO-CRC. High-quality evidence shows higher EO-CRC incidence in Hispanic populations. Alcohol intake and obesity modestly raise overall CRC risk. Enterotoxigenic Bacteroides fragilis and inflammatory bowel disease (IBD) strongly increase overall CRC incidence. IBD also modestly raises EO-CRC risk. Medium-quality evidence shows that higher vegetable consumption modestly reduces CRC incidence. Low-quality evidence suggests that lower area-level income may weakly increase CRC risk. The Paradigm III framework will be publicly available as an interactive, evidence-based online tool.
This conceptual model provides a novel framework for examining multiple etiologic factors that affect CRC risk and supports ongoing mathematical modeling of long-term benefits of interventions targeting modifiable biologic, behavioral, social, and environmental factors.