CANCER, ARSENIC, MORTALITY AND THEIR RELATIONSHIP

Arsenic is a natural element found in the environment in organic and inorganic form. The inorganic form is much more toxic and is found in ground water, surface water and many foods. This form is responsible for many adverse health effects like cancer and cardiovascular and neurological effects. Present work is based on the available mortality data on lung, bladder and liver cancer. The purpose of the study is to see the effect of various predictors like Gender, Cancer Type (Lung, Liver and Bladder), Arsenic Concentration (High, Medium and low), Age and Person Years on Mortality. We take a model-based approach to analyze this data. Three regression techniques: Logistic, Poisson and Negative Binomial Models are used to assess the effect of these explanatory variables on the response or dependent variable. The major finding of the study is that either all or most of the categories of the predictors are significantly associated with the mortality. The logistic model finds Age as the only significant predictor (p < .0001). The models for the actual frequency counts like Poisson Regression and Negative Binomial find all the five predictors significantly associated with the mortality.