Disparities in health status by socioeconomic factors are among the oldest and most widely corroborated epidemiologic observations. An active research program now seeks to identify the etiologic pathways through which these effects are transmitted. The standard methodologic approach for etiologic inferences concerning a social factor is to decompose effects by contrasting two adjusted effect estimates: one adjusted for potential confounders, and one adjusted for the same potential confounders plus one or more variables hypothesized to lie on the pathway through which the social factor exerts its effect. The goal is to distinguish indirect effects, through the specified intervening variables, from direct effects, transmitted via unspecified pathways. If control of hypothetical causal intermediates greatly attenuates an estimated effect, one generally infers that the effect is mediated primarily through pathways involving these quantities; minimal attenuation is interpreted as evidence that other pathways predominate. These mechanistic inferences then inform policy decisions concerning the utility of potential interventions. Although this approach is widespread in social epidemiology, it is not generally valid. We provide examples of erroneous conclusions that can arise in social epidemiology when mechanistic inference is based on this common strategy of controlling analytically for hypothetical causal intermediates as though they were confounders, and demonstrate an alternative analytic approach (Pearl's non-parametric structural equations models) which clarifies the assumptions and conditions necessary for valid effect identification and decomposition using standard methods. This thereby provides a basis for more substantial advancement in social epidemiology, and thus for efficacious policies toward successful elimination of health disparities.
Learning Objectives: At the conclusion of this presentation, attendees will be able to: 1) Define total and direct effects in social epidemiology; 2) Recognize that analytic control for hypothesized intermediates may not yield estimates of direct effects; 3) Apply the logic of nonparametric structural models to effect definition and identification in social epidemiology
Keywords: Methodology, Social Inequalities
Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: none
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.