Methodological issues in epidemiologic studies of dietary fat and cancer

Every scientific discipline has unique methodological constraints. Each scientific discipline also has differing degrees of power to resolve different questions. The methods of epidemiology are not exempt from these generalizations. In this brief review, several categories of methodological problems in epidemiological studies of dietary fat and cancer have been cited. Other problems could have been discussed and, several more examples illustrating the occurrence of the problems in research could have been presented. These problems notwithstanding, epidemiologic study of the association of dietary fat to cancer is an essential investigative tool in moving toward a better understanding of the public health impact of dietary fat. There are no known biological universals, and, as persuasive as data from animal models may be, extrapolation from the laboratory environment to human experience always is tentative. Laboratory methods require human validation before application to populations at risk of disease. Given the mutual dependency of epidemiologic studies and laboratory approaches, it is useful that the evidence from both realms be integrated and evaluated together as they are in this volume. Finally, given the limitations of any single epidemiologic method applied to the topic, future progress in understanding the role of dietary fat in cancer cause and prevention must rely on the application of multiple methodologies. By this means, it may be possible to compare and contrast the findings from different approaches to yield insights which would not otherwise be achievable.