| Background. There is epidemiological evidence supporting the role of specific dietary constituents in reducing risk for colorectal carcinogenesis. The primary objective of this project is to investigate the relationship between dietary factors, specifically micronutrient vitamins A, C, and E (project 1), folate (project 2), and vitamin D and calcium (project 3), and risk for colorectal neoplasia. Methods. Data from two separate colonoscopy-based case-control studies were used for the three projects. Diet was assessed prior to colonoscopy using a semi-quantitative Willett Food Frequency Questionnaire. Results. For Project 1: Multivariate-adjusted odds ratios comparing those with any supplemental vitamin intake and high dietary vitamin intake to those with no supplemental vitamin intake and low dietary vitamin intake across the clinic and community control groups, respectively, were 0.37 (95% CI, 0.15–0.94) and 0.64 (95% CI, 0.24–1.68) for vitamin A, and 0.67 (95% CI, 0.43–1.05) and 0.71 (95% CI, 0.45–1.11) for vitamin E. For project 2: The multivariate-adjusted odds ratio was 0.11 (95% CI, 0.03–0.39) for those who never drank alcohol and consumed high amounts of folate compared to those who consumed more than 6 drinks per day and low amounts of folate. For project 3: Adjusted odds ratios for those in the upper tertile of total calcium intake relative to those in the lower were 0.25 (0.08–0.80) among those who were Bb, 0.36 (0.15–0.85) among those who did not take NSAIDs, and 0.05 (0.01–0.41) and 0.16 (0.02–1.36) among those who were Bb and bb, respectively, and did not take NSAIDs. Conclusions. These data support the hypotheses that: higher intakes of vitamins A and E may reduce risk for colorectal adenoma (Project 1), higher levels of folate intake may reduce risk for colorectal adenoma, especially among those with low alcohol intake and perhaps among those who are homozygous for the MTHFR C677T polymorphism (Project 2) and higher calcium intakes may decrease risk for colorectal neoplasms, most readily detectable among those who do not take NSAIDs, and may be strongest among those who have at least one BSM I VDR b allele (Project 3). |
