Identifiability and exchangeability for direct and indirect effects - PubMed (original) (raw)
Identifiability and exchangeability for direct and indirect effects
J M Robins et al. Epidemiology. 1992 Mar.
Abstract
We consider the problem of separating the direct effects of an exposure from effects relayed through an intermediate variable (indirect effects). We show that adjustment for the intermediate variable, which is the most common method of estimating direct effects, can be biased. We also show that even in a randomized crossover trial of exposure, direct and indirect effects cannot be separated without special assumptions; in other words, direct and indirect effects are not separately identifiable when only exposure is randomized. If the exposure and intermediate never interact to cause disease and if intermediate effects can be controlled, that is, blocked by a suitable intervention, then a trial randomizing both exposure and the intervention can separate direct from indirect effects. Nonetheless, the estimation must be carried out using the G-computation algorithm. Conventional adjustment methods remain biased. When exposure and the intermediate interact to cause disease, direct and indirect effects will not be separable even in a trial in which both the exposure and the intervention blocking intermediate effects are randomly assigned. Nonetheless, in such a trial, one can still estimate the fraction of exposure-induced disease that could be prevented by control of the intermediate. Even in the absence of an intervention blocking the intermediate effect, the fraction of exposure-induced disease that could be prevented by control of the intermediate can be estimated with the G-computation algorithm if data are obtained on additional confounding variables.
Similar articles
- G-computation demonstration in causal mediation analysis.
Wang A, Arah OA. Wang A, et al. Eur J Epidemiol. 2015 Oct;30(10):1119-27. doi: 10.1007/s10654-015-0100-z. Epub 2015 Nov 4. Eur J Epidemiol. 2015. PMID: 26537707 Free PMC article. - Bounds on natural direct effects in the presence of confounded intermediate variables.
Sjölander A. Sjölander A. Stat Med. 2009 Feb 15;28(4):558-71. doi: 10.1002/sim.3493. Stat Med. 2009. PMID: 19035530 - Mediation analysis in epidemiology: methods, interpretation and bias.
Richiardi L, Bellocco R, Zugna D. Richiardi L, et al. Int J Epidemiol. 2013 Oct;42(5):1511-9. doi: 10.1093/ije/dyt127. Epub 2013 Sep 9. Int J Epidemiol. 2013. PMID: 24019424 Review. - Confounding in health research.
Greenland S, Morgenstern H. Greenland S, et al. Annu Rev Public Health. 2001;22:189-212. doi: 10.1146/annurev.publhealth.22.1.189. Annu Rev Public Health. 2001. PMID: 11274518 Review.
Cited by
- Pharmacotherapy guideline concordance for major depressive disorder and its link to functioning via symptom change.
Breitzig MT, He F, Kong L, Liu G, Waschbusch DA, Yanosky JD, Liao D, Saunders EFH. Breitzig MT, et al. J Clin Transl Sci. 2024 Sep 16;8(1):e123. doi: 10.1017/cts.2024.562. eCollection 2024. J Clin Transl Sci. 2024. PMID: 39345696 Free PMC article. - MASH: MEDIATION ANALYSIS OF SURVIVAL OUTCOME AND HIGH-DIMENSIONAL OMICS MEDIATORS WITH APPLICATION TO COMPLEX DISEASES.
Chi S, Flowers CR, Li Z, Huang X, Wei P. Chi S, et al. Ann Appl Stat. 2024 Jun;18(2):1360-1377. doi: 10.1214/23-aoas1838. Epub 2024 Apr 5. Ann Appl Stat. 2024. PMID: 39328363 Free PMC article. - A Bayesian nonparametric approach for causal mediation with a post-treatment confounder.
Bae W, Daniels MJ, Perri MG. Bae W, et al. Biometrics. 2024 Jul 1;80(3):ujae099. doi: 10.1093/biomtc/ujae099. Biometrics. 2024. PMID: 39311673 - Look on the bright side: the relation between family values, positive aspects of care and caregiver burden.
Zwar L, König HH, Hajek A. Zwar L, et al. Eur J Ageing. 2024 Aug 30;21(1):23. doi: 10.1007/s10433-024-00819-9. Eur J Ageing. 2024. PMID: 39212746 Free PMC article. - Using Directed Acyclic Graphs (DAGs) to Determine if the Total Causal Effect of an Individual Randomized Physical Activity-Promoting Intervention is Identifiable.
Myers ND, Brincks AM, Lee S. Myers ND, et al. Meas Phys Educ Exerc Sci. 2024;28(3):267-282. doi: 10.1080/1091367x.2024.2318355. Epub 2024 Feb 19. Meas Phys Educ Exerc Sci. 2024. PMID: 39206182 Free PMC article.
Publication types
MeSH terms
LinkOut - more resources
Full Text Sources
Other Literature Sources