Fig. 1From: Identification of confounder in epidemiologic data contaminated by measurement error in covariatesDirect acyclic graph of the causal effect between blood mercury and depressionBack to article page