Limits...
Causality, mediation and time: a dynamic viewpoint.

Aalen OO, R√łysland K, Gran JM, Ledergerber B - J R Stat Soc Ser A Stat Soc (2012)

Bottom Line: Graphical representations based on local independence graphs and dynamic path analysis are used to facilitate communication as well as providing an overview of the dynamic relations 'at a glance'.The relationship between causality as understood in a mechanistic and in an interventionist sense is discussed.An example using data from the Swiss HIV Cohort Study is presented.

View Article: PubMed Central - PubMed

Affiliation: University of Oslo Norway.

ABSTRACT
Summary. Time dynamics are often ignored in causal modelling. Clearly, causality must operate in time and we show how this corresponds to a mechanistic, or system, understanding of causality. The established counterfactual definitions of direct and indirect effects depend on an ability to manipulate the mediator which may not hold in practice, and we argue that a mechanistic view may be better. Graphical representations based on local independence graphs and dynamic path analysis are used to facilitate communication as well as providing an overview of the dynamic relations 'at a glance'. The relationship between causality as understood in a mechanistic and in an interventionist sense is discussed. An example using data from the Swiss HIV Cohort Study is presented.

No MeSH data available.


Patients not on HAART: two regression analyses are presented, with (a), (b) increments of HIV-1 RNA and (c), (d) increments of CD4 cell values as dependent variables, and lagged values of HIV-1 RNA and CD4 cell values as independent values; the analysis is performed for each month
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3500875&req=5

fig010: Patients not on HAART: two regression analyses are presented, with (a), (b) increments of HIV-1 RNA and (c), (d) increments of CD4 cell values as dependent variables, and lagged values of HIV-1 RNA and CD4 cell values as independent values; the analysis is performed for each month

Mentions: As an illustration we shall analyse data on CD4 cell count and HIV-1 RNA for patients on treatment and patients not on treatment. In the first group we would expect that the primary process is the reduction of HIV-1 RNA, and that this then drives the increase in the CD4 cell count, i.e. we would expect that CD4 is locally dependent on HIV-1 RNA, but not the other way round. In Fig. 9 we present results of regression analyses performed at every month where dependent variables are increments of HIV-1 RNA and CD4 cell counts respectively, and where the independent variables are in both cases lagged values of HIV-1 RNA and CD4 cell counts. The analyses show (Fig. 9(c)) that lagged values of HIV-1 RNA have a significant effect on the CD4 cell count increments, whereas lagged values of CD4 do not have an effect on HIV-1 RNA increments (Fig. 9(b)). This indicates that the CD4 process is locally dependent on the HIV-1 RNA process, whereas in the opposite direction there is local independence. This fits with the biological understanding of how the treatment works: that it decreases the viral load which then drives an improvement of the immune system. Fig. 10 presents the same analysis for the untreated group and we see similar results. This is not surprising: although there is no treatment to lower the viral load, still the amount of virus may change over time and be a driving force for changes in the CD4 cell level.


Causality, mediation and time: a dynamic viewpoint.

Aalen OO, R√łysland K, Gran JM, Ledergerber B - J R Stat Soc Ser A Stat Soc (2012)

Patients not on HAART: two regression analyses are presented, with (a), (b) increments of HIV-1 RNA and (c), (d) increments of CD4 cell values as dependent variables, and lagged values of HIV-1 RNA and CD4 cell values as independent values; the analysis is performed for each month
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3500875&req=5

fig010: Patients not on HAART: two regression analyses are presented, with (a), (b) increments of HIV-1 RNA and (c), (d) increments of CD4 cell values as dependent variables, and lagged values of HIV-1 RNA and CD4 cell values as independent values; the analysis is performed for each month
Mentions: As an illustration we shall analyse data on CD4 cell count and HIV-1 RNA for patients on treatment and patients not on treatment. In the first group we would expect that the primary process is the reduction of HIV-1 RNA, and that this then drives the increase in the CD4 cell count, i.e. we would expect that CD4 is locally dependent on HIV-1 RNA, but not the other way round. In Fig. 9 we present results of regression analyses performed at every month where dependent variables are increments of HIV-1 RNA and CD4 cell counts respectively, and where the independent variables are in both cases lagged values of HIV-1 RNA and CD4 cell counts. The analyses show (Fig. 9(c)) that lagged values of HIV-1 RNA have a significant effect on the CD4 cell count increments, whereas lagged values of CD4 do not have an effect on HIV-1 RNA increments (Fig. 9(b)). This indicates that the CD4 process is locally dependent on the HIV-1 RNA process, whereas in the opposite direction there is local independence. This fits with the biological understanding of how the treatment works: that it decreases the viral load which then drives an improvement of the immune system. Fig. 10 presents the same analysis for the untreated group and we see similar results. This is not surprising: although there is no treatment to lower the viral load, still the amount of virus may change over time and be a driving force for changes in the CD4 cell level.

Bottom Line: Graphical representations based on local independence graphs and dynamic path analysis are used to facilitate communication as well as providing an overview of the dynamic relations 'at a glance'.The relationship between causality as understood in a mechanistic and in an interventionist sense is discussed.An example using data from the Swiss HIV Cohort Study is presented.

View Article: PubMed Central - PubMed

Affiliation: University of Oslo Norway.

ABSTRACT
Summary. Time dynamics are often ignored in causal modelling. Clearly, causality must operate in time and we show how this corresponds to a mechanistic, or system, understanding of causality. The established counterfactual definitions of direct and indirect effects depend on an ability to manipulate the mediator which may not hold in practice, and we argue that a mechanistic view may be better. Graphical representations based on local independence graphs and dynamic path analysis are used to facilitate communication as well as providing an overview of the dynamic relations 'at a glance'. The relationship between causality as understood in a mechanistic and in an interventionist sense is discussed. An example using data from the Swiss HIV Cohort Study is presented.

No MeSH data available.