confounding

Syllable breakdown

Dictionary meaning

a distortion in the estimated association between an exposure and an outcome caused by a third variable — the confounder — that is associated with both the exposure and the outcome independently, mixing its effect with the exposure-outcome association being studied; confounding is a fundamental challenge in observational epidemiology because, unlike randomised controlled trials where random assignment balances known and unknown confounders between groups, observational studies cannot eliminate the possibility that observed associations are due to pre-existing differences between exposed and unexposed groups rather than to the exposure itself; classic examples of confounding include the association between coffee drinking and lung cancer — which is confounded by smoking, because coffee drinkers are more likely to smoke — and the association between carrying a lighter and lung cancer — confounded by both smoking and coffee drinking; confounding can be controlled in study design by restriction, matching, or randomisation, and in analysis by stratification, standardisation, or multivariable regression

Examples

• Age is a powerful confounder in studies of physical activity and mortality — older people exercise less and die more, but age rather than inactivity may explain much of the association.
• Propensity score matching was used to control for confounding by treatment indication in the observational study of statin effects.