Semiparametric Mediation Analysis with Separately Observed Mediator and Outcome under Unmeasured Confounding

2026-06-15

AI Summary

This study addresses the challenge of mediation analysis in realistic settings where the mediator and outcome cannot be jointly observed and unmeasured confounding is present, rendering conventional methods invalid. The authors propose a data fusion framework that integrates two separate, incomplete datasets—one containing the mediator and the other the outcome—and leverages a shared instrumental variable to circumvent the need for joint observation. They develop an identification strategy that does not require pre-specifying a valid instrument, relying instead on latent variable alignment and a no-interaction assumption. Building on influence functions, they construct an estimator that achieves both multiple robustness and the semiparametric efficiency bound. Applied to assess the mediating role of immune-related gene expression in the pathway from SNP rs610932 to dementia risk, the method successfully quantifies the indirect effect, demonstrating its validity and robustness.

Abstract

Mediation analysis is widely used to disentangle causal pathways, yet in many real-world studies the mediator M and outcome Y are never jointly observed. This incompleteness breaks the standard identification strategy for natural direct and indirect effects. We introduce a novel data fusion framework that restores the identification by combining two incomplete data sources, one measuring $M$ and the other measuring Y. Our approach leverages shared instrumental variables (IVs) to circumvent the need to observe (M,Y) jointly, remains valid under unmeasured confounding via a no-interaction condition, and accommodates covariate and exposure shifts across data sources under a latent alignment condition. We establish two identification strategies, one for settings with a known set of valid IVs, and another for settings where valid IVs must be learned. We further develop semiparametric, influence-function-based estimators with multiple robustness properties, and propose an estimator that attains the semiparametric efficiency bound under appropriate conditions. We apply our framework to quantify the extent to which the effect of SNP rs610932 on dementia risk is mediated through immune-related gene-expression pathways.

Problem

• mediation analysis
• unmeasured confounding
• incomplete data
• data fusion
• causal inference

Innovation

• data fusion
• instrumental variables
• unmeasured confounding
• semiparametric estimation
• mediation analysis