Many abiotic/biotic risk factors increase the chance of developing cancers, a leading cause of death across the world. These factors include pre-existing diseases (e.g., diabetes, osteoporosis and heart diseases), chronic conditions (e.g., inflammation, bronchitis and tuberculosis), chemical substances (e.g., asbestos, lead, cadmium and radon), physical conditions (e.g., age and obesity) and eating habits (e.g., low and high fat-content food). The chance of co-occurrence of these diseases (i.e., comorbid or multimorbid) is common and likely due to common perturbations in body’s functions; however, the exact mechanisms of comorbidities still remain elusive. Indeed, limited studies have been done to understand the comorbidities in the context of cancer outcomes, considering a wide range of co-occurring diseases. Here, we will study the potential risks of cancer development and the degree of its co-occurrence with other chronic diseases using the clinical data maintained in the National Cancer Institute’s (NCI) Cancer Data Access System (CDAS). In particular, we will retrieve the PLCO data [prostate, lung, colorectal, and ovarian (PLCO) cancer screening trial] from the CDAS and use the new analytical and biochemical techniques to study the developments of translational cancer biology, diagnostics, and therapeutics that have occurred over the past decades.
There are two related measures in cancer statistics: incidence rate and survival rate. We will combine these two measures (variables) in a hazard regression model framework, specifically, using the multistage proportional hazard regression model in order to handle the multiple exit modes of participants, including death, withdrawal and/or loss. We will also use a proportional hazard regression model to calculate the relative risk factor due to different comorbidities. This will allow us to rank the different comorbidities in relation to their contribution to risk factors. We will use multiple probability distributions (e.g., Weibull, Normal and Gamma distributions). Then, we will AIC (Akaike Information Criterion) to pick the best-fit model. While our interest is to understand how the risk of acquiring different types of cancer is affected by the presence of co-morbidities, we will use the following covariates in the hazard model: high blood pressure (hypertension), coronary heart disease/heart attack, stroke, emphysema, chronic bronchitis, diabetes, colorectal polyp(s), ulcerative colitis, familial polyposis, arthritis, osteoporosis, hepatitis, cirrhosis, diverticulitis, diverticulosis, gall bladder stones and inflammation. The overall cancer incident rate is also aggravated by individual’s age, physical activity and body mass, we also plan to use these parameters as control in our empirical model. Collectively, our studies will likely discover new information in cancer comorbidity and new approaches on future cancer research.

Madhusudan Dey, PhD
Associate Professor
Department of Biological Sciences
University of Wisconsin-Milwaukee
Lapham Hall 460
3209 N Maryland Ave
Milwaukee, WI-53211
Phone: 414-229-4309