Confounders can make it difficult to determine the true relationship between the exposure and the outcome, because they can create the semblance of a more direct association that is not actually present.
https://www.researchgate.net/post/Moderated_mediation_with_two_independent_variables