History of the modern epidemiological concept of confounding

Abstract

The epidemiological concept of confounding has had a convoluted history. It was first expressed as an issue of group non-comparability, later as an uncontrolled fallacy, then as a controllable fallacy named confounding, and, more recently, as an issue of group noncomparability in the distribution of potential outcome types. This latest development synthesised the apparent disconnect between phases of the history of confounding. Group non-comparability is the essence of confounding, and the statistical fallacy its consequence. This essay discusses how confounding was perceived in the 18th and 19th centuries, reviews how the concept evolved across the 20th century and finally describes the modern definition of confounding.

Control of Confounding and Reporting of Results in Causal Inference Studies. Guidance for Authors from Editors of Respiratory, Sleep, and Critical Care Journals

“…The scientific, mathematical, and theoretical underpinnings of causal inference, developed by Judea Pearl, James Robins, Miguel Hernán, and others, have evolved sufficiently to permit the everyday use of causal models(9)(10)(11)(12)(13)(14)(15)(16)(17). Causal models can be represented visuallyusing directed acyclic graphs (DAGs).…”

Confounding and collinearity in regression analysis: a cautionary tale and an alternative procedure, illustrated by studies of British voting behaviour

“…This has clear implications for the size, perhaps the sign, and also the standard error of the regression coefficients associated with those collinear variables, and hence for their interpretation. The result is frequently termed confounding, the situation when the relationship between two variables is distorted because of the strength of the relationships between either one or both of them and a third variable included in the analysis (see, for example, Kish 1959 ; Morabia 2011 ; VanderWheele and Shpitser 2013 ). 1 Thus the relationship between age and abstention at an election may be confounded by the inclusion of income in the statistical modelling, if, for example, affluent young males are more likely than comparable older males to abstain but affluent young females are more likely to vote than affluent older females.…”

On the definition of a confounder

“…The literature has moved away from formal language about "confounders" and instead places the conceptual emphasis on "confounding." See Morabia (2011) for historical discussion of this point. The causal inference literature has provided a formal definition of "confounding" in terms of dependence of counterfactual outcomes and exposure, possibly conditional on covariates.…”