Although the information derived from biological markers could conceivably be used to overcome some of the problems intrinsic to virtually all epidemiologic study designs—case definition, true exposure level, host susceptibility and resistance to factors of interest, the misclassification of study sub- 44 Rev Panam Salud Publica/Pan Am J Public Health 17(1), 2005 Opinión y análisis • Opinion and analysis jects (false positive and false negative test results), etc.—, we are still unable to resolve all such problems with the tools available at present. Biological markers seem more promising as potential indicators of the degree of susceptibility than as indicators of disease occurrence, an application requiring further technical refinement. Currently biological markers are employed in public health mainly to screen for particular diseases. Unfortunately, these markers have their limitations. For one thing, it is unlikely that they will completely eliminate the problem of false positive and false negative results, since DNA from solid tumors undergoes slight degradation due to necrosis and since genetic markers are susceptible to the effects of exposure to medication, diet, sex, ethnicity, and even the circadian cycle. And even if false positives and negatives were ultimately eliminated, it would be impossible to use many of the analytical tools based on two by two tables, such as the chi squared test, logistic regression, the Poisson regression, Cox’ proportional hazards ratio, etc., since such tools rely on comparisons of the number of false positives and negatives in the exposed and non-exposed groups. Finally, albeit no less important, certain ethical issues must be carefully considered before allowing the massive use of human genetic markers, which could lead to violations of the rights of individuals, families, and communities if carried out in an indiscriminate, unregulated fashion. Epidemiology is rapidly broadening its scope, a trend that will continue into the future; new analytical tools will be developed, and the working hypotheses to which such tools will be applied will change. At present the scientific community is paying increased attention to this field of study, but more research and discussion are needed to respond to many of the questions for which we have no satisfactory answers yet