Model-Robust Direct Effect Under Confounder-Mediator Ambiguity

TL;DR

This paper introduces a model-robust approach to estimating direct effects in observational studies, addressing confounder-mediator ambiguity and providing estimands valid under both roles.

Contribution

It develops a new estimand that remains interpretable whether the focal variable is a confounder or mediator, with a doubly robust estimator and sensitivity bounds.

Findings

The proposed estimand equals the ATE or interventional direct effect depending on the variable's role.

Simulation studies demonstrate the robustness of the estimator across scenarios.

Application to NHANES data shows differences from traditional mediation analyses.

Abstract

Direct effect analyses usually require deciding whether a focal variable is a pre-exposure confounder or a post-exposure mediator. In observational studies, that distinction may be unclear because timing is measured coarsely or the variable reflects an evolving process. Considering the average treatment effect (ATE) and the natural direct effect (NDE) as a common notion of the direct effect when the focal variable is a confounder and a mediator, respectively, we show that, in general, no single observed-data estimand recovers both the ATE when the focal variable is a confounder and the NDE when it is a mediator. Consequently, if a practitioner applies an NDE estimator when the variable is actually pre-exposure, the resulting estimate may have no clear causal interpretation. We identify a no-additive-interaction condition under which these quantities coincide, develop sensitivity bounds for departures from that condition, and propose an alternative model-robust estimand. This estimand equals the ATE when the variable is pre-exposure and an interventional direct effect when it is post-exposure. Moreover, within a natural class of outcome-free stochastic direct effects, it is the unique observed-data functional that remains causally interpretable under both structural roles of the focal variable. We derive an efficient influence function and a doubly robust estimator, yielding robustness at two levels: the estimand is model-robust across the two causal scenarios, and the estimator is doubly robust with respect to nuisance estimation. In simulations and in an NHANES application on elevated PFAS burden, kidney function, and uric acid, mediation-based analyses yielded materially different reported estimates.