Abstract Consider a study with binary exposure, outcome, and confounder, where the confounder is nondifferentially misclassified. Epidemiologists have long accepted the unproven but oft-cited result that, if the confounder is binary, odds ratios, risk ratios, and risk differences which control for the mismeasured confounder will lie between the crude and the true measures. In this paper the authors provide an analytic proof of the result in the absence of a qualitative interaction between treatment and confounder, and demonstrate via counterexample that the result need not hold when there is a qualitative interaction between treatment and confounder. They also present an analytic proof of the result for the effect of treatment amount the treated, and describe extensions to measures conditional on or standardized over other covariates.
Ogburn, Elizabeth L. and VanderWeele, Tyler J., "On the Nondifferential Misclassification of a Binary Confounder" (September 2011). COBRA Preprint Series. Working Paper 82.