Purpose: To explore the hypothesis that a combination of several risk factors acting through the same pathway may produce an overall large increase in risk even in the presence of weak associations with each individual factor.
Methods: Using oxidative stress pathway as an example, we propose an oxidative stress score (OSS), where high and low pro-oxidant exposures expressed as continuous variables are assigned values of 0 and 1, while high and low antioxidant exposures are assigned values of 1 and 0, respectively. Dichotomous variables for pro-oxidant and antioxidant exposures are scored in a similar fashion. All individual scores are then summed to calculate the overall OSS, where higher and lower values indicate a shift toward antioxidant and pro-oxidant exposures, respectively.
Results: We illustrate this approach by using data from two previously-conducted case-control studies: a colonoscopy-based colorectal adenoma study, and a population-based prostate cancer study. In this pilot illustration we found a substantial decrease in risk associated with a high OSS for both prostate cancer and colorectal adenoma. By contrast, analyses for individual OSS components demonstrated no discernible pattern.
Conclusions: Our exploratory analyses serve as a demonstration of a method and warrant further confirmation on a larger scale.