The data were obtained from the case-control arm of the population-based Melbourne Colorectal Cancer Study
]. The principal alcohol findings were a significant risk for rectal cancer among habitual beer drinkers, a risk which was annulled by a high consumption of vitamin-C-containing foods [6
]. The main dietary findings were published in 1987 [12
]. The micronutrient analysis of colorectal cancer risk we reported in 2006 [13
], but this did not include the risks for rectal cancer associated with micronutrient intake among beer drinkers, therefore the data here represent unpublished material.
The study included 323 incident cases of rectal cancer and 727 age/sex frequency matched randomly chosen community controls in Melbourne, Australia, collected in 1980-1981. There was a quantitative estimate of all foods eaten during a period most representative of the previous 20 years [12
]. Whole-life beer consumption was estimated and categorised into tertiles of drinking based on all the study participants consuming beer. Alcohol data were missing for 7 cases and 3 controls, and these were excluded from the analysis. Rectal cancer risk associated with each level of beer consumption was estimated (Tables and ).
Upper levels of daily consumption of beer, number, and percentage distribution by controls and rectal cancer cases in the Melbourne Colorectal Cancer Study.
Odds ratios (ORs), 95% confidence intervals (95% CIs) and P values for categories of beer consumption for rectal cancer unadjusted and adjusted for other risk factors. Data from the Melbourne Colorectal Cancer Study.
The intake of micronutrients involved in one-carbon metabolism (folate, methionine, and vitamins B6 and B12), and four other micronutrients with antioxidant properties (vitamins C, E, lycopene, and selenium) were estimated from foods eaten using the US Department of Agriculture nutrient tables [14
], and categorised into quintiles of consumption based on all the participants in the study. Nutritional supplements were not included. Logistic regression was used to estimate associations as odds ratios (ORs) between rectal cancer and the quintiles of micronutrient intake in beer drinkers taken as a single category, to demonstrate explicitly
the effects investigated. Adjusted ORs (AORs) were obtained by adding the variables age, sex, alcohol (yes/no), body mass index (quintiles), energy intake (quintiles), family history of colorectal cancer (yes/no), oral contraceptive pill use (yes/no), cigarette pack years (none, 1–19, 20–39, ≥40), aspirin use yes/no), and NSAID use (yes/no) to the micronutrient logistic regression model, and level of beer consumption also added to the model of .
Association between quintiles of micronutrient intake and rectal cancer amongst beer consumers and nonconsumers.
To investigate the hypothesis that micronutrients may attenuate the risk of rectal cancer in beer drinkers, the data were stratified by beer consumption, and subgroup analyses were undertaken of beer consumers and nonconsumers.