Abstract
Exact formulae relating parameters in conditional and reduced generalized linear models are introduced where the reduced model omits a continuous mediator from the conditional model. For certain link functions including logit, the natural direct effect and the natural indirect effect of the counterfactual method are smaller in magnitude than, respectively, the direct effect used by the difference method and the indirect effect by the product method. Contrary to what is implicitly assumed in Jiang and VanderWeele [11] for logit link, the total effect of the counterfactual method and the total effect used for the difference method are generally not the same. They are equal to each other only under special situations. For accelerated failure time models the difference method and the product method are equivalent regardless of censoring or not, a result stated in VanderWeele [6] in the absence of censorship but proved in a misleading manner. For proportional hazards models, maximum likelihood analysis indicates that these two methods can be equivalent in the absence of censorship. In the case of logit link, one can focus on the treatment effect on the marginalized odds instead of the odds of the marginalized event so that the product method would be equivalent to the difference method. Similarly, for the proportional hazards model, one can focus on the treatment effect on the marginalized hazards instead of the hazards for the reduced model.
Acknowledgment
The author thanks the referees for their valuable comments.
-
Research funding: None declared.
-
Employment or leadership: None declared.
-
Honorarium: None declared.
-
Conflict of interest statement: The authors declare no conflicts of interest regarding this article.
References
1. VanderWeele TJ, Vansteelandt S. Odds ratios for mediation analysis for a dichotomous outcome. Am J Epidemiol 2010;172:1339–48. https://doi.org/10.1093/aje/kwq332.Search in Google Scholar PubMed PubMed Central
2. Tchetgen Tchetgen E. A note on formulae for causal mediation analysis in an odds ratio context. Epidemiol Method 2013;2:21–31. https://doi.org/10.1515/em-2012-0005.Search in Google Scholar PubMed PubMed Central
3. Gaynor SM, Schwartz J, Lin X. Mediation analysis for common binary outcomes. Stat Med 2019;38:512–29. https://doi.org/10.1002/sim.7945.Search in Google Scholar PubMed
4. Doretti M, Raggi M, Stanghellini E. Exact parametric causal mediation analysis for a binary outcome with a binary mediator, 2018. arXiv preprint arXiv:1811.00439.Search in Google Scholar
5. Stanghellini E, Doretti M. On marginal and conditional parameters in logistic regression models. Biometrika 2019;106:732–9. https://doi.org/10.1093/biomet/asz019.Search in Google Scholar
6. VanderWeele TJ. Commentary: Causal mediation analysis with survival data. Epidemiology 2011;22:582–85. https://doi.org/10.1097/EDE.0b013e31821db37e.Search in Google Scholar PubMed PubMed Central
7. Wang Z, Louis TA. Matching conditional and marginal shapes in binary random intercept models using a bridge distribution function. Biometrika 2003;90:765–75. https://doi.org/10.1093/biomet/90.4.765.Search in Google Scholar
8. Samoilenko M, Blais L, Lefebvre G. Comparing logistic and log-binomial models for causal mediation analyses of binary mediators and rare binary outcomes: evidence to support cross-checking of mediation results in practice. Obs Stud 2018;4:193–216.10.1353/obs.2018.0013Search in Google Scholar
9. Samoilenko M, Lefebvre G. Point: risk ratio equations for natural direct and indirect effects in causal mediation analysis of a binary mediator and a binary outcome—a fresh look at the formulas. Am J Epidemiol 2019;188:1201–3. https://doi.org/10.1093/aje/kwy275.Search in Google Scholar PubMed
10. VanderWeele TJ, Valeri L, Ananth CV. Counterpoint: Mediation formulas with binary mediators and outcomes and the “rare outcome assumption”. Am J Epidemiol 2019;188:1204–5. https://doi.org/10.1093/aje/kwy281.Search in Google Scholar PubMed PubMed Central
11. Jiang Z, VanderWeele TJ. When is the difference method conservative for assessing mediation? Am J Epidemiol 2015;182:105–8. https://doi.org/10.1093/aje/kwv059.Search in Google Scholar PubMed PubMed Central
12. Lange T, Hansen KW, Sørensen R, Galatius S. Applied mediation analyses: a review and tutorial. Epidemiol Health 2017;39:e2017035. https://doi.org/10.4178/epih.e2017035.Search in Google Scholar PubMed PubMed Central
13. Neuhaus JM, Jewell NP. A geometric approach to assess bias due to omitted covariates in generalized linear models. Biometrika 1993;80:807–15. https://doi.org/10.2307/2336872.Search in Google Scholar
14. Sobel ME. Asymptotic confidence intervals for indirect effects in structural equation models. Soc Methodol 1982;13:290–312. https://doi.org/10.2307/270723.Search in Google Scholar
15. Baron RM, Kenny DA. The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J Pers Soc Psychol 1986;51:1173–82. https://doi.org/10.1037/0022-3514.51.6.1173.Search in Google Scholar
16. VanderWeele T. Explanation in causal inference: methods for mediation and interaction, New York: Oxford University Press; 2015.10.1093/ije/dyw277Search in Google Scholar PubMed PubMed Central
17. Gail MH, Wieand S, Piantadosi S. Biased estimates of treatment effect in randomized experiments with nonlinear regressions and omitted covariates. Biometrika 1984;71:431–44. https://doi.org/10.1093/biomet/71.3.431.Search in Google Scholar
18. Lin DY, Psaty BM, Kronmal RA. Assessing the sensitivity of regression results to unmeasured confounders in observational studies. Biometrics 1998;54:948–63. https://doi.org/10.2307/2533848.Search in Google Scholar
19. Drake C, McQuarrie A. A note on the bias due to omitted confounders. Biometrika 1995;82:633–8. https://doi.org/10.1093/biomet/82.3.633.Search in Google Scholar
20. McCullagh P, Nelder J. Generalized linear models, 2nd ed. Chapman & Hall London; 1989.10.1007/978-1-4899-3242-6Search in Google Scholar
21. Wang K. Maximum likelihood analysis of linear mediation models with treatment–mediator interaction. Psychometrika 2019;84:719–48. https://doi.org/10.1007/s11336-019-09670-9.Search in Google Scholar PubMed
© 2020 Walter de Gruyter GmbH, Berlin/Boston