Plasma Choline Metabolites and Colorectal Cancer Risk in the Women's Health Initiative Observational Study

Few studies have examined associations between plasma choline metabolites and risk of colorectal cancer. Therefore, we investigated associations between plasma biomarkers of choline metabolism choline, betaine, dimethylglycine, and trimethylamine N-oxide (TMAO) and colorectal cancer risk among postmenopausal women in a case-control study nested within the Women's Health Initiative Observational Study. We selected 835 matched case-control pairs, and cases were further stratified by tumor site (proximal, distal, or rectal) and stage (local/regional or metastatic). Colorectal cancer was assessed by self-report and confirmed by medical records over the mean of 5.2 years of follow-up. Baseline plasma choline metabolites were measured by LC/MS-MS. In multivariable-adjusted conditional logistic regression models, plasma choline tended to be positively associated with rectal cancer risk OR (95% confidence interval, CI)(highest vs. lowest quartile) = 2.44 (0.93-6.40); P trend = 0.08, whereas plasma betaine was inversely associated with colorectal cancer overall 0.68 (0.47-0.99); P trend = 0.01 and with local/regional tumors 0.64 (0.42-0.99); P trend = 0.009. Notably, the plasma betaine:choline ratio was inversely associated with colorectal cancer overall 0.56 (0.39-0.82); P trend = 0.004 as well as with proximal 0.66 (0.41-1.06); P trend = 0.049, rectal 0.27 (0.10-0.78); P trend = 0.02, and local/regional 0.50 (0.33-0.76); P trend = 0.001 tumors. Finally, plasma TMAO, an oxidative derivative of choline produced by intestinal bacteria, was positively associated with rectal cancer 3.38 (1.25-9.16); P trend = 0.02 and with overall colorectal cancer risk among women with lower (vs. higher) plasma vitamin B12 levels (P interaction = 0.003). Collectively, these data suggest that alterations in choline metabolism, which may arise early in disease development, may be associated with higher risk of colorectal cancer. The positive association between plasma TMAO and colorectal cancer risk is consistent with an involvement of the gut microbiome in colorectal cancer pathogenesis.