Complete effect decomposition for an arbitrary number of multiple ordered mediators with time-varying confounders: A method for generalized causal multi-mediation analysis
Restricted accessResearch articleFirst published online January, 2023
Complete effect decomposition for an arbitrary number of multiple ordered mediators with time-varying confounders: A method for generalized causal multi-mediation analysis
Causal mediation analysis is advantageous for mechanism investigation. In settings with multiple causally ordered mediators, path-specific effects have been introduced to specify the effects of certain combinations of mediators. However, most path-specific effects are unidentifiable. An interventional analog of path-specific effects is adapted to address the non-identifiability problem. Moreover, previous studies only focused on cases with two or three mediators due to the complexity of the mediation formula in a large number of mediators. In this study, we provide a generalized definition of traditional path-specific effects and interventional path-specific effects with a recursive formula, along with the required assumptions for nonparametric identification. Subsequently, a general approach is developed with an arbitrary number of multiple ordered mediators and with time-varying confounders. All methods and software proposed in this study contribute to comprehensively decomposing a causal effect confirmed by data science and help disentangling causal mechanisms in the presence of complicated causal structures among multiple mediators.
MacKinnonDP. Introduction to statistical mediation analysis. New York: Routledge, 2008.
2.
KlineRB. Principles and practice of structural equation modeling. New York: Guilford Press, 2015.
3.
RobinsJMGreenlandS. Identifiability and exchangeability for direct and indirect effects. Epidemiology1992; 3: 143–155.
4.
PearlJ. Direct and indirect effects. In: Proceedings of the seventeenth conference on uncertainty in artificial intelligenceSan Francisco, CA, USA, 2001, pp.411 − 4420. Morgan Kaufmann Publishers Inc.
5.
VanderWeeleTVansteelandtS. Conceptual issues concerning mediation, interventions and composition. Stat Interface2009; 2: 457–468.
6.
van der LaanMJPetersenML. Direct effect models. Int J Biostat2008; 4: article 23.
7.
VanderWeeleTJ. Marginal structural models for the estimation of direct and indirect effects. Epidemiology2009; 20: 18–26.
8.
ImaiKKeeleLTingleyD. A general approach to causal mediation analysis. Psychol Methods2010; 15: 309.
9.
VanderWeeleTJ. Bias formulas for sensitivity analysis for direct and indirect effects. Epidemiology (Cambridge, Mass)2010; 21: 540.
10.
VanderWeeleTJVansteelandtS. Odds ratios for mediation analysis for a dichotomous outcome. Am J Epidemiol2010; 172: 1339–1348.
11.
DidelezVDawidPGenelettiS. Direct and indirect effects of sequential treatments. arXiv preprint arXiv:120668402012.
12.
Tchetgen TchetgenEJShpitserI. Semiparametric theory for causal mediation analysis: efficiency bounds, multiple robustness and sensitivity analysis. Ann Stat2012; 40: 1816–1845.
13.
ValeriLVanderWeeleTJ. Mediation analysis allowing for exposure–mediator interactions and causal interpretation: theoretical assumptions and implementation with SAS and SPSS macros. Psychol Methods2013; 18: 137.
14.
VanderWeeleTJVansteelandtS. Mediation analysis with multiple mediators. Epidemiol Method2014; 2: 95–115.
15.
VanderweeleTJVansteelandtSRobinsJM. Effect decomposition in the presence of an exposure-induced mediator-outcome confounder. Epidemiology2014; 25: 300–306.
16.
FasanelliFGiraudoMTRicceriF, et al. Marginal time-dependent causal effects in mediation analysis with survival data. Am J Epidemiol2019; 188: 967–974.
17.
AvinCShpitserIPearlJ. Identifiability of path-specific effects. In: Proceedings of the 19th International Joint Conference on Artificial Intelligence, 2005.
TaguriMFeatherstoneJChengJ. Causal mediation analysis with multiple causally non-ordered mediators. Stat Methods Med Res2015; 27: 3–19.
20.
WangWNelsonSAlbertJM. Estimation of causal mediation effects for a dichotomous outcome in multiple-mediator models using the mediation formula. Stat Med2013; 32: 4211–4228.
21.
HuangY-TYangH-I. Causal mediation analysis of survival outcome with multiple mediators. Epidemiology2017; 28: 370–378.
22.
SteenJLoeysTMoerkerkeB, et al. Flexible mediation analysis with multiple mediators. Am J Epidemiol2017; 186: 184–193.
23.
LohWWMoerkerkeBLoeysT, et al. Interventional effect models for multiple mediators. arXiv preprint arXiv:1907084152019.
24.
VansteelandtSDanielRM. Interventional effects for mediation analysis with multiple mediators. Epidemiology (Cambridge, Mass) 2017; 28: 258.
25.
Moreno-BetancurMMoranPBeckerD, et al. Mediation effects that emulate a target randomised trial: simulation-based evaluation of ill-defined interventions on multiple mediators. Stat Methods Med Res2021; 30: 1395–1412.
26.
PreacherKJHayesAF. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behav Res Methods2008; 40: 879–891.
27.
TaiASLinSH. Integrated multiple mediation analysis: a robustness-specificity trade-off in causal structure. Stat Med2021; 40: 4541–4567.
28.
DanielRDe StavolaBCousensS, et al. Causal mediation analysis with multiple mediators. Biometrics2015; 71: 1–14.
29.
AlbertJMChoJILiuY, et al. Generalized causal mediation and path analysis: extensions and practical considerations. Stat Methods Med Res2019; 28: 1793–1807.
30.
LinS-HVanderWeeleT. Interventional approach for path-specific effects. J Causal Inference2017; 5.
31.
GenelettiS. Identifying direct and indirect effects in a non-counterfactual framework. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 2007; 69: 199–215.
32.
ZhengWvan der LaanMJ. Causal mediation in a survival setting with time-dependent mediators. 2012.
33.
LinSHYoungJGLoganR, et al. Mediation analysis for a survival outcome with time-varying exposures, mediators, and confounders. Stat Med2017; 36: 4153–4166.
34.
VanderWeeleTJTchetgen TchetgenE. Mediation analysis with time-varying exposures and mediators. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 2017; 79(3): 917–938.
35.
ZhengWvan der LaanM. Longitudinal mediation analysis with time-varying mediators and exposures, with application to survival outcomes. J Causal Inference2017; 5: 20160006.
36.
VanderWeeleTJTchetgen TchetgenEJ. Mediation analysis with time-varying exposures and mediators. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 2017; 79: 917–938.
37.
TaiA-SLinS-HChuY-C, et al. Causal mediation analysis with multiple time-varying mediators. Epidemiology2022(In press).
38.
YamamuroSShinozakiTIimuroS, et al. Mediational g-formula for time-varying treatment and repeated-measured multiple mediators: application to atorvastatin’s effect on cardiovascular disease via cholesterol lowering and anti-inflammatory actions in elderly type 2 diabetics. Stat Methods Med Res2021; 30: 1782–1799.
39.
RobinsJ. A new approach to causal inference in mortality studies with a sustained exposure period—application to control of the healthy worker survivor effect. Math Model1986; 7: 1393–1512.
40.
PearlJ. Causal inference in statistics: an overview. Stat Surv2009; 3: 96–146.
41.
VanderWeeleTJ. Concerning the consistency assumption in causal inference. Epidemiology2009; 20: 880–883.
42.
VanderWeeleTJ. A unification of mediation and interaction: a four-way decomposition. Epidemiology2014; 25: 749–761.
43.
VanderWeeleTJShrierI. Sufficient cause representation of the four-way decomposition for mediation and interaction. Epidemiology2016; 27: e32–e33.
44.
BollenKA. Structural equation models that are nonlinear in latent variables: a least-squares estimator. Sociol Methodol1995; 25: 223–252.
45.
MarcoulidesGASchumackerRE. Advanced structural equation modeling: Issues and techniques. New York: Psychology Press, 2013.
46.
SobelME. Some new results on indirect effects and their standard errors in covariance structure models. Sociol Methodol1986; 16: 159–186.
47.
KingGTomzMWittenbergJ. Making the most of statistical analyses: Improving interpretation and presentation. Available at SSRN 10837382000.
48.
HuangBSivaganesanSSuccopP, et al. Statistical assessment of mediational effects for logistic mediational models. Stat Med2004; 23: 2713–2728.
49.
GlynnAN. Estimating and bounding mechanism specific causal effects. In: 25th annual summer meeting of the Society for Political Methodology, Ann Arbor, Michigan2008.
50.
WangYTaylorJM. A measure of the proportion of treatment effect explained by a surrogate marker. Biometrics2002; 58: 803–812.
51.
Tchetgen TchetgenEJ. On causal mediation analysis with a survival outcome. Int J Biostat2011; 7: 1–38.
52.
Tchetgen TchetgenEJ. Identification and estimation of survivor average causal effects. Stat Med2014; 33: 3601–3628.
53.
ChenC-JYangH-ISuJ, et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA2006; 295: 65–73.
54.
ChenCLYangHIYangWS, et al. Metabolic factors and risk of hepatocellular carcinoma by chronic hepatitis B/C infection: a follow-up study in Taiwan. Gastroenterology2008; 135: 111–121.
55.
IloejeUHYangHIJenCL, et al. Risk and predictors of mortality associated with chronic hepatitis B infection. Clin Gastroenterol Hepatol2007; 5: 921–931.
56.
HuangY-TJenC-LYangH-I, et al. Lifetime risk and sex difference of hepatocellular carcinoma among patients with chronic hepatitis B and C. J Clin Oncol2011; 29: 3643.
57.
HuangY-TFreemanJRYangH-I, et al. Mediation effect of hepatitis B and C on mortality. Eur J Epidemiol2016; 31: 625–633.
58.
HuangY-TYangH-ILiuJ, et al. Mediation analysis of hepatitis B and C in relation to hepatocellular carcinoma risk. Epidemiology2016; 27: 14–20.
59.
YangYHChiangHJYipHK, et al. Risk of new-onset atrial fibrillation among Asian chronic hepatitis C virus carriers: a nationwide population-based cohort study. J Am Heart Assoc2019; 8: e012914.
60.
Van Der LaanMJRubinD. Targeted maximum likelihood learning. Int J Biostat2006; 2: article 11.
61.
ZhengWvan der LaanMJ. Targeted maximum likelihood estimation of natural direct effects. Int J Biostat2012; 8: article 3.
62.
LinS-HYoungJLoganR, et al. Mediation analysis with survival outcome and time-varying exposures, mediators, and confounders. Stat Med2017; 36: 4153–4166.
63.
TaiA-STsaiC-ALinS-H. Survival mediation analysis with the death-truncated mediator: the completeness of the survival mediation parameter. Stat Med2021; 40: 3953–3974.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
