A number of epidemiologic studies have described what appear to be paradoxical associations, where an incongruous relationship is observed between a certain well-established risk factor for disease incidence and favorable clinical outcome among patients with that disease. For example, the "obesity paradox" represents the association between obesity and better survival among patients with a certain disease such as coronary heart disease. Paradoxical observations cause vexing clinical and public health problems as they raise questions on causal relationships and hinder the development of effective interventions. Compelling evidence indicates that pathogenic processes encompass molecular alterations within cells and the microenvironment, influenced by various exogenous and endogenous exposures, and that interpersonal heterogeneity in molecular pathology and pathophysiology exists among patients with any given disease. In this article, we introduce methods of the emerging integrative interdisciplinary field of molecular pathological epidemiology (MPE), which is founded on the unique disease principle and disease continuum theory. We analyze and decipher apparent paradoxical findings, utilizing the MPE approach and available literature data on tumor somatic genetic and epigenetic characteristics. Through our analyses in colorectal cancer, renal cell carcinoma, and glioblastoma (malignant brain tumor), we can readily explain paradoxical associations between disease risk factors and better prognosis among disease patients. The MPE paradigm and approach can be applied to not only neoplasms but also various non-neoplastic diseases where there exists indisputable ubiquitous heterogeneity of pathogenesis and molecular pathology. The MPE paradigm including consideration of disease heterogeneity plays an essential role in advancements of precision medicine and public health.
Keywords: Bias; Cardiovascular disease; Molecular diagnostics; Multifactorial diseases; Personalized medicine.