I am not sure if a good model fit is enough for accepting the causal relation beyond doubt.

Similar questions and discussions