Abstract

Introduction: Micronutrients content of habitually consumed foods comprise a causal network model of cancer, but the evaluation of their effect on this pathology represents a great challenge because of the interdependence in their usual consumption. Several studies reported that nutrients can modify the urinary tract tumors (UTT) risk, although such evidence is still limited. Objective: To identify associations between dietary vitamins A, E, B6, C, phosphorus, selenium and zinc intakes and the presence of UTT in Cordoba, Argentina, considering the multicollinearity caused by the interdependence of their consumption. Methods: A case control study was carried out including 129 cases with incident histopathologically confirmed UTT and 257 controls. A food frequency questionnaire previously validated was administrated to each subject. Dietary intakes of vitamins A, E, B6, C, phosphorus, selenium and zinc were the variables of interest, each showing high correlation with each other and thus, causing collinearity. So, multiple logistic regression models were adjusted and their adaptation to the presence of correlation, Ridge regression, to obtain the odds ratio (OR). The models included terms of sex, age, body mass index (BMI), socioeconomic status, occupational exposure to carcinogens, tobacco consumption and caloric intake as covariates. Results: Vitamin E and vitamin B6 showed a slight protective effect (OR: 0.943, CI 95% 0.897-0.998 and OR: 0.730 CI 95% 0.457-1.167). Selenium was slightly promoter (OR: 1.012 CI 95% 1.001-1.023). Conclusion: When multicollinearity is considered in the model, it is possible to obtain more accurate estimates of the modulation that some micronutrients have on the risk of UTT more precisely.