On the Conventional Definition of Path-specific Effects: Fully Mediated Interaction With Multiple Ordered Mediators

Area of Science:

• Causal inference and mediation analysis in health research.

Background:

• Path-specific effects are crucial for understanding mediation with multiple mediators.
• Conventional definitions can lead to misinterpretation in multiple mediator analysis.
• Accurate mediation analysis is vital for understanding disease pathways.

Purpose of the Study:

• To propose a novel framework for decomposing fully mediated interaction from the average causal effect.
• To address misclassification issues in path-specific effect interpretation.
• To offer strategies for improved mediation analysis in complex health conditions.

Main Methods:

• Developed two strategies: isolating and reclassifying fully mediated interaction.
• Applied these strategies alongside conventional methods to analyze COPD patient data.
• Compared the performance of proposed strategies against the traditional decomposition approach.

Main Results:

• Misclassification of fully mediated interaction is identified as a key source of misinterpretation.
• The conventional strategy underestimates the role of dyspnea in COPD.
• Reclassifying fully mediated interaction revealed 50% of the average causal effect is mediated, particularly by dyspnea.

Conclusions:

• Proposed strategies offer superior accuracy in mediation analysis compared to conventional methods.
• Reclassifying fully mediated interaction is recommended for precise interpretation of mediation results.
• Accurate mediation analysis, especially concerning dyspnea, is critical for understanding COPD patient health status.