Age-Period-Cohort Analyses in Epidemiological Studies: Clarifying Assumptions and Validating Results

Ryan K. Masters , University of Colorado Boulder
Daniel A. Powers, University of Texas at Austin

Age-Period-Cohort (APC) models are often used to decompose health trends into period- and cohort-based sources, but their use remains contentious. Central to the contention are researchers’ failures to 1) clearly state their analytic assumptions and/or 2) thoroughly evaluate model results. These failures generate confusion about the merits of APC methods, and different APC approaches are often treated and discussed as one and the same. We propose three simple guidelines to help practitioners of APC methods articulate their assumptions and validate their results. To demonstrate the usefulness of the guidelines, we apply them to results recently published in American Journal of Epidemiology about black-white differences in U.S. heart disease mortality. The application of the guidelines reveals that some APC methods produce inconsistent results that are highly sensitive to researcher manipulation while others estimate results that are robust to researcher manipulation and consistent across APC models.

See

 Presented in Session 167. Modeling Mortality