Cargando…

Adjustment for time-invariant and time-varying confounders in ‘unexplained residuals’ models for longitudinal data within a causal framework and associated challenges

‘Unexplained residuals’ models have been used within lifecourse epidemiology to model an exposure measured longitudinally at several time points in relation to a distal outcome. It has been claimed that these models have several advantages, including: the ability to estimate multiple total causal ef...

Descripción completa

Detalles Bibliográficos
Autores principales: Arnold, KF, Ellison, GTH, Gadd, SC, Textor, J, Tennant, PWG, Heppenstall, A, Gilthorpe, MS
Formato: Online Artículo Texto
Lenguaje:English
Publicado: SAGE Publications 2018
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6484949/
https://www.ncbi.nlm.nih.gov/pubmed/29451093
http://dx.doi.org/10.1177/0962280218756158
_version_ 1783414193666392064
author Arnold, KF
Ellison, GTH
Gadd, SC
Textor, J
Tennant, PWG
Heppenstall, A
Gilthorpe, MS
author_facet Arnold, KF
Ellison, GTH
Gadd, SC
Textor, J
Tennant, PWG
Heppenstall, A
Gilthorpe, MS
author_sort Arnold, KF
collection PubMed
description ‘Unexplained residuals’ models have been used within lifecourse epidemiology to model an exposure measured longitudinally at several time points in relation to a distal outcome. It has been claimed that these models have several advantages, including: the ability to estimate multiple total causal effects in a single model, and additional insight into the effect on the outcome of greater-than-expected increases in the exposure compared to traditional regression methods. We evaluate these properties and prove mathematically how adjustment for confounding variables must be made within this modelling framework. Importantly, we explicitly place unexplained residual models in a causal framework using directed acyclic graphs. This allows for theoretical justification of appropriate confounder adjustment and provides a framework for extending our results to more complex scenarios than those examined in this paper. We also discuss several interpretational issues relating to unexplained residual models within a causal framework. We argue that unexplained residual models offer no additional insights compared to traditional regression methods, and, in fact, are more challenging to implement; moreover, they artificially reduce estimated standard errors. Consequently, we conclude that unexplained residual models, if used, must be implemented with great care.
format Online
Article
Text
id pubmed-6484949
institution National Center for Biotechnology Information
language English
publishDate 2018
publisher SAGE Publications
record_format MEDLINE/PubMed
spelling pubmed-64849492019-06-03 Adjustment for time-invariant and time-varying confounders in ‘unexplained residuals’ models for longitudinal data within a causal framework and associated challenges Arnold, KF Ellison, GTH Gadd, SC Textor, J Tennant, PWG Heppenstall, A Gilthorpe, MS Stat Methods Med Res Articles ‘Unexplained residuals’ models have been used within lifecourse epidemiology to model an exposure measured longitudinally at several time points in relation to a distal outcome. It has been claimed that these models have several advantages, including: the ability to estimate multiple total causal effects in a single model, and additional insight into the effect on the outcome of greater-than-expected increases in the exposure compared to traditional regression methods. We evaluate these properties and prove mathematically how adjustment for confounding variables must be made within this modelling framework. Importantly, we explicitly place unexplained residual models in a causal framework using directed acyclic graphs. This allows for theoretical justification of appropriate confounder adjustment and provides a framework for extending our results to more complex scenarios than those examined in this paper. We also discuss several interpretational issues relating to unexplained residual models within a causal framework. We argue that unexplained residual models offer no additional insights compared to traditional regression methods, and, in fact, are more challenging to implement; moreover, they artificially reduce estimated standard errors. Consequently, we conclude that unexplained residual models, if used, must be implemented with great care. SAGE Publications 2018-02-16 2019-05 /pmc/articles/PMC6484949/ /pubmed/29451093 http://dx.doi.org/10.1177/0962280218756158 Text en © The Author(s) 2018 http://creativecommons.org/licenses/by/4.0/ This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage).
spellingShingle Articles
Arnold, KF
Ellison, GTH
Gadd, SC
Textor, J
Tennant, PWG
Heppenstall, A
Gilthorpe, MS
Adjustment for time-invariant and time-varying confounders in ‘unexplained residuals’ models for longitudinal data within a causal framework and associated challenges
title Adjustment for time-invariant and time-varying confounders in ‘unexplained residuals’ models for longitudinal data within a causal framework and associated challenges
title_full Adjustment for time-invariant and time-varying confounders in ‘unexplained residuals’ models for longitudinal data within a causal framework and associated challenges
title_fullStr Adjustment for time-invariant and time-varying confounders in ‘unexplained residuals’ models for longitudinal data within a causal framework and associated challenges
title_full_unstemmed Adjustment for time-invariant and time-varying confounders in ‘unexplained residuals’ models for longitudinal data within a causal framework and associated challenges
title_short Adjustment for time-invariant and time-varying confounders in ‘unexplained residuals’ models for longitudinal data within a causal framework and associated challenges
title_sort adjustment for time-invariant and time-varying confounders in ‘unexplained residuals’ models for longitudinal data within a causal framework and associated challenges
topic Articles
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6484949/
https://www.ncbi.nlm.nih.gov/pubmed/29451093
http://dx.doi.org/10.1177/0962280218756158
work_keys_str_mv AT arnoldkf adjustmentfortimeinvariantandtimevaryingconfoundersinunexplainedresidualsmodelsforlongitudinaldatawithinacausalframeworkandassociatedchallenges
AT ellisongth adjustmentfortimeinvariantandtimevaryingconfoundersinunexplainedresidualsmodelsforlongitudinaldatawithinacausalframeworkandassociatedchallenges
AT gaddsc adjustmentfortimeinvariantandtimevaryingconfoundersinunexplainedresidualsmodelsforlongitudinaldatawithinacausalframeworkandassociatedchallenges
AT textorj adjustmentfortimeinvariantandtimevaryingconfoundersinunexplainedresidualsmodelsforlongitudinaldatawithinacausalframeworkandassociatedchallenges
AT tennantpwg adjustmentfortimeinvariantandtimevaryingconfoundersinunexplainedresidualsmodelsforlongitudinaldatawithinacausalframeworkandassociatedchallenges
AT heppenstalla adjustmentfortimeinvariantandtimevaryingconfoundersinunexplainedresidualsmodelsforlongitudinaldatawithinacausalframeworkandassociatedchallenges
AT gilthorpems adjustmentfortimeinvariantandtimevaryingconfoundersinunexplainedresidualsmodelsforlongitudinaldatawithinacausalframeworkandassociatedchallenges