Limits...
Challenges and novel approaches for investigating molecular mediation

View Article: PubMed Central - PubMed

ABSTRACT

Understanding mediation is useful for identifying intermediates lying between an exposure and an outcome which, when intervened upon, will block (some or all of) the causal pathway between the exposure and outcome. Mediation approaches used in conventional epidemiology have been adapted to understanding the role of molecular intermediates in situations of high-dimensional omics data with varying degrees of success. In particular, the limitations of observational epidemiological study including confounding, reverse causation and measurement error can afflict conventional mediation approaches and may lead to incorrect conclusions regarding causal effects. Solutions to analysing mediation which overcome these problems include the use of instrumental variable methods such as Mendelian randomization, which may be applied to evaluate causality in increasingly complex networks of omics data.

No MeSH data available.


A simplistic representation of mediation. (A) Complete mediation - M is the only mechanism by which E can change Y. (B) Partial mediation - In practice, it is more likely that E has an effect on Y other than those operating by changing M. Mediation aims to partition the total (causal) effect of E on Y into mediated effects (effects that operate by changing the mediator, M) and non-mediated effects.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5036871&req=5

ddw197-F2: A simplistic representation of mediation. (A) Complete mediation - M is the only mechanism by which E can change Y. (B) Partial mediation - In practice, it is more likely that E has an effect on Y other than those operating by changing M. Mediation aims to partition the total (causal) effect of E on Y into mediated effects (effects that operate by changing the mediator, M) and non-mediated effects.

Mentions: A mediator (M) is a variable that is on the causal path from an exposure (E) to an outcome variable (Y). It causes the outcome and is itself caused by the exposure. There are a variety of statistical methods that have been introduced for analysing mediation, from simple regression-based systems and structural equation models to more novel parametric and semi-parametric methods (8), and these have been widely implemented (Fig. 2).Figure 2.


Challenges and novel approaches for investigating molecular mediation
A simplistic representation of mediation. (A) Complete mediation - M is the only mechanism by which E can change Y. (B) Partial mediation - In practice, it is more likely that E has an effect on Y other than those operating by changing M. Mediation aims to partition the total (causal) effect of E on Y into mediated effects (effects that operate by changing the mediator, M) and non-mediated effects.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5036871&req=5

ddw197-F2: A simplistic representation of mediation. (A) Complete mediation - M is the only mechanism by which E can change Y. (B) Partial mediation - In practice, it is more likely that E has an effect on Y other than those operating by changing M. Mediation aims to partition the total (causal) effect of E on Y into mediated effects (effects that operate by changing the mediator, M) and non-mediated effects.
Mentions: A mediator (M) is a variable that is on the causal path from an exposure (E) to an outcome variable (Y). It causes the outcome and is itself caused by the exposure. There are a variety of statistical methods that have been introduced for analysing mediation, from simple regression-based systems and structural equation models to more novel parametric and semi-parametric methods (8), and these have been widely implemented (Fig. 2).Figure 2.

View Article: PubMed Central - PubMed

ABSTRACT

Understanding mediation is useful for identifying intermediates lying between an exposure and an outcome which, when intervened upon, will block (some or all of) the causal pathway between the exposure and outcome. Mediation approaches used in conventional epidemiology have been adapted to understanding the role of molecular intermediates in situations of high-dimensional omics data with varying degrees of success. In particular, the limitations of observational epidemiological study including confounding, reverse causation and measurement error can afflict conventional mediation approaches and may lead to incorrect conclusions regarding causal effects. Solutions to analysing mediation which overcome these problems include the use of instrumental variable methods such as Mendelian randomization, which may be applied to evaluate causality in increasingly complex networks of omics data.

No MeSH data available.