Evidence-appraisal glossary
Mediation Analysis
A set of methods that split a total effect into the part carried through an intermediate variable (the indirect effect) and the part acting by other routes (the direct effect). It answers how, or through what, an exposure changes an outcome.
Also called: causal mediation, direct and indirect effects.
Mediation analysis asks whether a treatment works because it changes some intermediate step, for instance whether a drug prevents heart attacks by lowering blood pressure or through other pathways. Valid estimates require no unmeasured confounding of the exposure-mediator, exposure-outcome, and mediator-outcome relationships, which is a demanding set of conditions. Adjusting for a mediator incorrectly can itself introduce collider bias, so these analyses need to be planned with an explicit causal diagram.
This is a plain-language methodology definition for reading research. It is general education, not medical advice.