It should be noted that relative figures are the standard method of reporting effects in epidemiology, for good reasons and bad. But among them is that the regr

It should be noted that relative figures are the standard method of reporting effects in epidemiology, for good reasons and bad. But among them is that the regression models needed for absolute differences under more sophisticated study designs, controlling for confounding variables, etc. may not be available. http://journals.lww.com/epidem/Fulltext/2010/01000/On_the_Origin_of_Risk_Relativism.2.aspx http://journals.lww.com/epidem/Fulltext/2010/01000/On_the_Or...