Lifestyle risk factor modelling for complex disease prediction with respect to prostate cancer

Complex disease prediction in healthcare presents significant challenges due to the multitude of interacting risk factors and comorbidities associated with disease occurrence and prevention. Prostate cancer, as a prevalent example of a complex disease, has been shown to arise from a combination of biological, environmental, and lifestyle risk factors. While research has extensively studied individual lifestyle risk factors such as age, diet, and environmental exposures, limited attention has been given to how these factors interact when combined, particularly in the context of prostate cancer.This thesis addresses this gap by exploring the predictive impact of combined lifestyle risk factors on prostate cancer outcomes. To achieve this, a novel Predictive Algorithm Framework (PAF) is proposed, which integrates a risk factor selection and sorting algorithm adapted from the Apriori algorithm and a risk factor aggregation algorithm inspired by Opinion Geometric Pooling Algorithms. This framework facilitates the identification of significant combinations of lifestyle risk factors and estimates their influence on prostate cancer probability, offering a more nuanced understanding of how these factors interplay.Research findings demonstrate that the relationship between lifestyle risk factor combinations and prostate cancer is not linear or directly proportional. Surprisingly, certain combinations of lifestyle factors were associated with both high prevention potential and increased disease occurrence risk. This underscores the importance of targeted prevention strategies that account for the complex interplay of specific risk factor combinations. By emphasizing the nuanced impact of lifestyle risk factor modelling, this work contributes to advancing predictive methodologies for complex diseases like prostate cancer, offering actionable insights for personalized prevention strategies and improved healthcare decision-making.