Practical guidance for conducting mediation analysis with multiple mediators using inverse odds ratio weighting

Quynh C. Nguyen, Theresa L. Osypuk, Nicole M. Schmidt, M. Maria Glymour, Eric J Tchetgen Tchetgen

Research output: Contribution to journalArticle

31 Citations (Scopus)

Abstract

Despite the recent flourishing of mediation analysis techniques, many modern approaches are difficult to implement or applicable to only a restricted range of regression models. This report provides practical guidance for implementing a new technique utilizing inverse odds ratio weighting (IORW) to estimate natural direct and indirect effects for mediation analyses. IORW takes advantage of the odds ratio's invariance property and condenses information on the odds ratio for the relationship between the exposure (treatment) and multiple mediators, conditional on covariates, by regressing exposure on mediators and covariates. The inverse of the covariate-adjusted exposure-mediator odds ratio association is used to weight the primary analytical regression of the outcome on treatment. The treatment coefficient in such a weighted regression estimates the natural direct effect of treatment on the outcome, and indirect effects are identified by subtracting direct effects from total effects. Weighting renders treatment and mediators independent, thereby deactivating indirect pathways of the mediators. This new mediation technique accommodates multiple discrete or continuous mediators. IORW is easily implemented and is appropriate for any standard regression model, including quantile regression and survival analysis. An empirical example is given using data from the Moving to Opportunity (1994-2002) experiment, testing whether neighborhood context mediated the effects of a housing voucher program on obesity. Relevant Stata code (StataCorp LP, College Station, Texas) is provided.

Original languageEnglish (US)
Pages (from-to)349-356
Number of pages8
JournalAmerican journal of epidemiology
Volume181
Issue number5
DOIs
StatePublished - Mar 1 2015

Fingerprint

Odds Ratio
Survival Analysis
Obesity
Regression Analysis
Weights and Measures

Keywords

  • direct effects
  • effect decomposition
  • indirect effects
  • mediation
  • weighted regression

Cite this

Practical guidance for conducting mediation analysis with multiple mediators using inverse odds ratio weighting. / Nguyen, Quynh C.; Osypuk, Theresa L.; Schmidt, Nicole M.; Glymour, M. Maria; Tchetgen, Eric J Tchetgen.

In: American journal of epidemiology, Vol. 181, No. 5, 01.03.2015, p. 349-356.

Research output: Contribution to journalArticle

@article{f4c943fb296b4282ab162eeb7a45d0f7,
title = "Practical guidance for conducting mediation analysis with multiple mediators using inverse odds ratio weighting",
abstract = "Despite the recent flourishing of mediation analysis techniques, many modern approaches are difficult to implement or applicable to only a restricted range of regression models. This report provides practical guidance for implementing a new technique utilizing inverse odds ratio weighting (IORW) to estimate natural direct and indirect effects for mediation analyses. IORW takes advantage of the odds ratio's invariance property and condenses information on the odds ratio for the relationship between the exposure (treatment) and multiple mediators, conditional on covariates, by regressing exposure on mediators and covariates. The inverse of the covariate-adjusted exposure-mediator odds ratio association is used to weight the primary analytical regression of the outcome on treatment. The treatment coefficient in such a weighted regression estimates the natural direct effect of treatment on the outcome, and indirect effects are identified by subtracting direct effects from total effects. Weighting renders treatment and mediators independent, thereby deactivating indirect pathways of the mediators. This new mediation technique accommodates multiple discrete or continuous mediators. IORW is easily implemented and is appropriate for any standard regression model, including quantile regression and survival analysis. An empirical example is given using data from the Moving to Opportunity (1994-2002) experiment, testing whether neighborhood context mediated the effects of a housing voucher program on obesity. Relevant Stata code (StataCorp LP, College Station, Texas) is provided.",
keywords = "direct effects, effect decomposition, indirect effects, mediation, weighted regression",
author = "Nguyen, {Quynh C.} and Osypuk, {Theresa L.} and Schmidt, {Nicole M.} and Glymour, {M. Maria} and Tchetgen, {Eric J Tchetgen}",
year = "2015",
month = "3",
day = "1",
doi = "10.1093/aje/kwu278",
language = "English (US)",
volume = "181",
pages = "349--356",
journal = "American Journal of Epidemiology",
issn = "0002-9262",
publisher = "Oxford University Press",
number = "5",

}

TY - JOUR

T1 - Practical guidance for conducting mediation analysis with multiple mediators using inverse odds ratio weighting

AU - Nguyen, Quynh C.

AU - Osypuk, Theresa L.

AU - Schmidt, Nicole M.

AU - Glymour, M. Maria

AU - Tchetgen, Eric J Tchetgen

PY - 2015/3/1

Y1 - 2015/3/1

N2 - Despite the recent flourishing of mediation analysis techniques, many modern approaches are difficult to implement or applicable to only a restricted range of regression models. This report provides practical guidance for implementing a new technique utilizing inverse odds ratio weighting (IORW) to estimate natural direct and indirect effects for mediation analyses. IORW takes advantage of the odds ratio's invariance property and condenses information on the odds ratio for the relationship between the exposure (treatment) and multiple mediators, conditional on covariates, by regressing exposure on mediators and covariates. The inverse of the covariate-adjusted exposure-mediator odds ratio association is used to weight the primary analytical regression of the outcome on treatment. The treatment coefficient in such a weighted regression estimates the natural direct effect of treatment on the outcome, and indirect effects are identified by subtracting direct effects from total effects. Weighting renders treatment and mediators independent, thereby deactivating indirect pathways of the mediators. This new mediation technique accommodates multiple discrete or continuous mediators. IORW is easily implemented and is appropriate for any standard regression model, including quantile regression and survival analysis. An empirical example is given using data from the Moving to Opportunity (1994-2002) experiment, testing whether neighborhood context mediated the effects of a housing voucher program on obesity. Relevant Stata code (StataCorp LP, College Station, Texas) is provided.

AB - Despite the recent flourishing of mediation analysis techniques, many modern approaches are difficult to implement or applicable to only a restricted range of regression models. This report provides practical guidance for implementing a new technique utilizing inverse odds ratio weighting (IORW) to estimate natural direct and indirect effects for mediation analyses. IORW takes advantage of the odds ratio's invariance property and condenses information on the odds ratio for the relationship between the exposure (treatment) and multiple mediators, conditional on covariates, by regressing exposure on mediators and covariates. The inverse of the covariate-adjusted exposure-mediator odds ratio association is used to weight the primary analytical regression of the outcome on treatment. The treatment coefficient in such a weighted regression estimates the natural direct effect of treatment on the outcome, and indirect effects are identified by subtracting direct effects from total effects. Weighting renders treatment and mediators independent, thereby deactivating indirect pathways of the mediators. This new mediation technique accommodates multiple discrete or continuous mediators. IORW is easily implemented and is appropriate for any standard regression model, including quantile regression and survival analysis. An empirical example is given using data from the Moving to Opportunity (1994-2002) experiment, testing whether neighborhood context mediated the effects of a housing voucher program on obesity. Relevant Stata code (StataCorp LP, College Station, Texas) is provided.

KW - direct effects

KW - effect decomposition

KW - indirect effects

KW - mediation

KW - weighted regression

UR - http://www.scopus.com/inward/record.url?scp=84923004235&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84923004235&partnerID=8YFLogxK

U2 - 10.1093/aje/kwu278

DO - 10.1093/aje/kwu278

M3 - Article

VL - 181

SP - 349

EP - 356

JO - American Journal of Epidemiology

JF - American Journal of Epidemiology

SN - 0002-9262

IS - 5

ER -