Bounds on natural direct effects in the presence of confounded intermediate variables
- 26 November 2008
- journal article
- research article
- Published by Wiley in Statistics in Medicine
- Vol. 28 (4) , 558-571
Abstract
In epidemiological studies we often want to learn about the direct effect of an exposure on an outcome, i.e. the effect that is not relayed by a specific intermediate variable. In the literature, there are two common definitions of direct effects;controlledandnatural. When the intermediate variable and the outcome have common causes, neither the controlled nor the natural direct effect is identified. Caiet al. (Biometrics2007;64(3):695–701) derived bounds for the controlled direct effect under a set of hierarchical assumptions. In this paper we derive bounds on the natural direct effect under the same assumptions. Copyright © 2008 John Wiley & Sons, Ltd.Keywords
Funding Information
- Swedish Research Council (621-2004-3940)
- Swedish Foundation for Strategic Research (A3 02:129)
This publication has 10 references indexed in Scilit:
- Bounds on Direct Effects in the Presence of Confounded Intermediate VariablesBiometrics, 2007
- DiscussionsBiometrics, 2007
- Bounds on Causal Effects in Three-Arm Trials With Non-ComplianceJournal of the Royal Statistical Society Series B: Statistical Methodology, 2006
- Estimation of Direct Causal EffectsEpidemiology, 2006
- Improved estimation of controlled direct effects in the presence of unmeasured confounding of intermediate variablesStatistics in Medicine, 2005
- Estimation of Causal Effects via Principal Stratification When Some Outcomes are Truncated by “Death”Journal of Educational and Behavioral Statistics, 2003
- Semantics of causal DAG models and the identification of direct and indirect effects2003
- Identifiability and Exchangeability for Direct and Indirect EffectsEpidemiology, 1992
- Statistical validation of intermediate endpoints for chronic diseasesStatistics in Medicine, 1992
- Theoretical Statistics1974