A longitudinal study of prediagnostic metabolic biomarkers and the risk of molecular subtypes of colorectal cancer

• Published in: Scientific Reports
• Main Authors: Myte, Robin, Harlid, Sophia, Sundkvist, Anneli, Gylling, Björn, Häggström, Jenny, Zingmark, Carl, Burström, Anna Löfgren, Palmqvist, Richard, Guelpen, Bethany Van
• Other Authors: Cancer Research Foundation in Northern Sweden, Cancerfonden
• Format: Article in Journal/Newspaper
• Language: English
• Published: Springer Science and Business Media LLC 2020
• Subjects: Multidisciplinary; Northern Sweden
• Online Access: http://dx.doi.org/10.1038/s41598-020-62129-1; http://www.nature.com/articles/s41598-020-62129-1.pdf; http://www.nature.com/articles/s41598-020-62129-1

Abstract Body fatness increases the risk of colorectal cancer (CRC). Insulin resistance and altered adipokines are potential mechanisms, but previous biomarker studies have been inconsistent. Intertumoral heterogeneity might provide an explanation. We investigated insulin, C-peptide, adiponectin, and leptin in relation to CRC molecular subtypes using a nested case-control design (1010 cases, 1010 matched controls, median 12.3 years from baseline to CRC diagnosis) from the population-based Northern Sweden Health and Disease Study. Repeated samples were available from 518 participants. Risks of CRC and subtypes, defined by tumor BRAF and KRAS mutations and microsatellite instability (MSI) status, were estimated using conditional logistic regression and linear mixed models. Higher C-peptide and lower adiponectin were associated with increased CRC risk (odds ratios per standard deviation increase (95% CI): 1.11 (1.01, 1.23) and 0.91 (0.83, 1.00), respectively), though weakened when adjusted for body mass index. Insulin and leptin were not associated with CRC risk. Within-individual time trajectories were similar in cases and controls, and no subtype-specific relationships were identified (all P heterogeneity > 0.1). Adiponectin was weakly inversely associated with the risk of KRAS -mutated (P = 0.08) but not BRAF -mutated or KRAS/BRAF -wildtype CRC, consistent with the one previous study. These findings contribute to an increased understanding of the complex role of body size in CRC.