Comparison of methods for the analysis of relatively simple mediation models

Judith J.M. Rijnhart, Jos W.R. Twisk, Mai J.M. Chinapaw, Michiel R. de Boer, Martijn W. Heymans

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

Background/aims Statistical mediation analysis is an often used method in trials, to unravel the pathways underlying the effect of an intervention on a particular outcome variable. Throughout the years, several methods have been proposed, such as ordinary least square (OLS) regression, structural equation modeling (SEM), and the potential outcomes framework. Most applied researchers do not know that these methods are mathematically equivalent when applied to mediation models with a continuous mediator and outcome variable. Therefore, the aim of this paper was to demonstrate the similarities between OLS regression, SEM, and the potential outcomes framework in three mediation models: 1) a crude model, 2) a confounder-adjusted model, and 3) a model with an interaction term for exposure-mediator interaction. Methods Secondary data analysis of a randomized controlled trial that included 546 schoolchildren. In our data example, the mediator and outcome variable were both continuous. We compared the estimates of the total, direct and indirect effects, proportion mediated, and 95% confidence intervals (CIs) for the indirect effect across OLS regression, SEM, and the potential outcomes framework. Results OLS regression, SEM, and the potential outcomes framework yielded the same effect estimates in the crude mediation model, the confounder-adjusted mediation model, and the mediation model with an interaction term for exposure-mediator interaction. Conclusions Since OLS regression, SEM, and the potential outcomes framework yield the same results in three mediation models with a continuous mediator and outcome variable, researchers can continue using the method that is most convenient to them.

Original languageEnglish
Pages (from-to)130-135
Number of pages6
JournalContemporary Clinical Trials Communications
Volume7
DOIs
Publication statusPublished - 1 Sep 2017

Fingerprint

Least-Squares Analysis
Research Personnel
Randomized Controlled Trials
Confidence Intervals

Keywords

  • Cross-sectional data
  • Indirect effect
  • Mediation analysis
  • Ordinary least square regression
  • Potential outcomes framework
  • Structural equation modeling

Cite this

Rijnhart, Judith J.M. ; Twisk, Jos W.R. ; Chinapaw, Mai J.M. ; de Boer, Michiel R. ; Heymans, Martijn W. / Comparison of methods for the analysis of relatively simple mediation models. In: Contemporary Clinical Trials Communications. 2017 ; Vol. 7. pp. 130-135.
@article{11688ab383bf425f89b19953b0ed0f7e,
title = "Comparison of methods for the analysis of relatively simple mediation models",
abstract = "Background/aims Statistical mediation analysis is an often used method in trials, to unravel the pathways underlying the effect of an intervention on a particular outcome variable. Throughout the years, several methods have been proposed, such as ordinary least square (OLS) regression, structural equation modeling (SEM), and the potential outcomes framework. Most applied researchers do not know that these methods are mathematically equivalent when applied to mediation models with a continuous mediator and outcome variable. Therefore, the aim of this paper was to demonstrate the similarities between OLS regression, SEM, and the potential outcomes framework in three mediation models: 1) a crude model, 2) a confounder-adjusted model, and 3) a model with an interaction term for exposure-mediator interaction. Methods Secondary data analysis of a randomized controlled trial that included 546 schoolchildren. In our data example, the mediator and outcome variable were both continuous. We compared the estimates of the total, direct and indirect effects, proportion mediated, and 95{\%} confidence intervals (CIs) for the indirect effect across OLS regression, SEM, and the potential outcomes framework. Results OLS regression, SEM, and the potential outcomes framework yielded the same effect estimates in the crude mediation model, the confounder-adjusted mediation model, and the mediation model with an interaction term for exposure-mediator interaction. Conclusions Since OLS regression, SEM, and the potential outcomes framework yield the same results in three mediation models with a continuous mediator and outcome variable, researchers can continue using the method that is most convenient to them.",
keywords = "Cross-sectional data, Indirect effect, Mediation analysis, Ordinary least square regression, Potential outcomes framework, Structural equation modeling",
author = "Rijnhart, {Judith J.M.} and Twisk, {Jos W.R.} and Chinapaw, {Mai J.M.} and {de Boer}, {Michiel R.} and Heymans, {Martijn W.}",
year = "2017",
month = "9",
day = "1",
doi = "10.1016/j.conctc.2017.06.005",
language = "English",
volume = "7",
pages = "130--135",
journal = "Contemporary Clinical Trials Communications",
issn = "2451-8654",
publisher = "Elsevier Inc.",

}

Comparison of methods for the analysis of relatively simple mediation models. / Rijnhart, Judith J.M.; Twisk, Jos W.R.; Chinapaw, Mai J.M.; de Boer, Michiel R.; Heymans, Martijn W.

In: Contemporary Clinical Trials Communications, Vol. 7, 01.09.2017, p. 130-135.

Research output: Contribution to JournalArticleAcademicpeer-review

TY - JOUR

T1 - Comparison of methods for the analysis of relatively simple mediation models

AU - Rijnhart, Judith J.M.

AU - Twisk, Jos W.R.

AU - Chinapaw, Mai J.M.

AU - de Boer, Michiel R.

AU - Heymans, Martijn W.

PY - 2017/9/1

Y1 - 2017/9/1

N2 - Background/aims Statistical mediation analysis is an often used method in trials, to unravel the pathways underlying the effect of an intervention on a particular outcome variable. Throughout the years, several methods have been proposed, such as ordinary least square (OLS) regression, structural equation modeling (SEM), and the potential outcomes framework. Most applied researchers do not know that these methods are mathematically equivalent when applied to mediation models with a continuous mediator and outcome variable. Therefore, the aim of this paper was to demonstrate the similarities between OLS regression, SEM, and the potential outcomes framework in three mediation models: 1) a crude model, 2) a confounder-adjusted model, and 3) a model with an interaction term for exposure-mediator interaction. Methods Secondary data analysis of a randomized controlled trial that included 546 schoolchildren. In our data example, the mediator and outcome variable were both continuous. We compared the estimates of the total, direct and indirect effects, proportion mediated, and 95% confidence intervals (CIs) for the indirect effect across OLS regression, SEM, and the potential outcomes framework. Results OLS regression, SEM, and the potential outcomes framework yielded the same effect estimates in the crude mediation model, the confounder-adjusted mediation model, and the mediation model with an interaction term for exposure-mediator interaction. Conclusions Since OLS regression, SEM, and the potential outcomes framework yield the same results in three mediation models with a continuous mediator and outcome variable, researchers can continue using the method that is most convenient to them.

AB - Background/aims Statistical mediation analysis is an often used method in trials, to unravel the pathways underlying the effect of an intervention on a particular outcome variable. Throughout the years, several methods have been proposed, such as ordinary least square (OLS) regression, structural equation modeling (SEM), and the potential outcomes framework. Most applied researchers do not know that these methods are mathematically equivalent when applied to mediation models with a continuous mediator and outcome variable. Therefore, the aim of this paper was to demonstrate the similarities between OLS regression, SEM, and the potential outcomes framework in three mediation models: 1) a crude model, 2) a confounder-adjusted model, and 3) a model with an interaction term for exposure-mediator interaction. Methods Secondary data analysis of a randomized controlled trial that included 546 schoolchildren. In our data example, the mediator and outcome variable were both continuous. We compared the estimates of the total, direct and indirect effects, proportion mediated, and 95% confidence intervals (CIs) for the indirect effect across OLS regression, SEM, and the potential outcomes framework. Results OLS regression, SEM, and the potential outcomes framework yielded the same effect estimates in the crude mediation model, the confounder-adjusted mediation model, and the mediation model with an interaction term for exposure-mediator interaction. Conclusions Since OLS regression, SEM, and the potential outcomes framework yield the same results in three mediation models with a continuous mediator and outcome variable, researchers can continue using the method that is most convenient to them.

KW - Cross-sectional data

KW - Indirect effect

KW - Mediation analysis

KW - Ordinary least square regression

KW - Potential outcomes framework

KW - Structural equation modeling

UR - http://www.scopus.com/inward/record.url?scp=85021330774&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85021330774&partnerID=8YFLogxK

U2 - 10.1016/j.conctc.2017.06.005

DO - 10.1016/j.conctc.2017.06.005

M3 - Article

VL - 7

SP - 130

EP - 135

JO - Contemporary Clinical Trials Communications

JF - Contemporary Clinical Trials Communications

SN - 2451-8654

ER -