Limits...
Partitioning the population attributable fraction for a sequential chain of effects.

Mason CA, Tu S - Epidemiol Perspect Innov (2008)

Bottom Line: While the population attributable fraction (PAF) provides potentially valuable information regarding the community-level effect of risk factors, significant limitations exist with current strategies for estimating a PAF in multiple risk factor models.These strategies can result in paradoxical or ambiguous measures of effect, or require unrealistic assumptions regarding variables in the model.In addition, the approach can also be used to statistically control for confounding by other variables, while avoiding the potential pitfalls of attempting to separately differentiate direct and indirect effects.

View Article: PubMed Central - HTML - PubMed

Affiliation: College of Education and Human Development, University of Maine, Orono, ME, USA. craig.mason@umit.maine.edu

ABSTRACT

Background: While the population attributable fraction (PAF) provides potentially valuable information regarding the community-level effect of risk factors, significant limitations exist with current strategies for estimating a PAF in multiple risk factor models. These strategies can result in paradoxical or ambiguous measures of effect, or require unrealistic assumptions regarding variables in the model. A method is proposed in which an overall or total PAF across multiple risk factors is partitioned into components based upon a sequential ordering of effects. This method is applied to several hypothetical data sets in order to demonstrate its application and interpretation in diverse analytic situations.

Results: The proposed method is demonstrated to provide clear and interpretable measures of effect, even when risk factors are related/correlated and/or when risk factors interact. Furthermore, this strategy not only addresses, but also quantifies issues raised by other researchers who have noted the potential impact of population-shifts on population-level effects in multiple risk factor models.

Conclusion: Combined with simple, unadjusted PAF estimates and an aggregate PAF based on all risk factors under consideration, the sequentially partitioned PAF provides valuable additional information regarding the process through which population rates of a disorder may be impacted. In addition, the approach can also be used to statistically control for confounding by other variables, while avoiding the potential pitfalls of attempting to separately differentiate direct and indirect effects.

No MeSH data available.


Two correlated risk factors with an interaction.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2572052&req=5

Figure 6: Two correlated risk factors with an interaction.

Mentions: The final example considers the situation where there is an interaction involving risk factors that are related/correlated with each other. For example, (1) maternal smoking during pregnancy may lead to higher rates of babies born low birthweight, while low birthweight then leads to higher risk of MMR (i.e., smoking and birthweight are related) and (2) the effect of being born low birthweight on MMR may be different for those babies whose mothers also smoked, than is the effect of low birthweight for those babies whose mothers did not smoke (i.e., a smoking × birthweight interaction on MMR). Data for this example are presented in Figure 6. Applying the sequential partitioning strategy, PAFA is equal to 14.29%, the adjusted PAFB is equal to 29.12%, and the PAF for the interaction is equal to 6.59%. Applying Equation 11, the PAF for the interaction translates to 4.286 cases of the outcome.


Partitioning the population attributable fraction for a sequential chain of effects.

Mason CA, Tu S - Epidemiol Perspect Innov (2008)

Two correlated risk factors with an interaction.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2572052&req=5

Figure 6: Two correlated risk factors with an interaction.
Mentions: The final example considers the situation where there is an interaction involving risk factors that are related/correlated with each other. For example, (1) maternal smoking during pregnancy may lead to higher rates of babies born low birthweight, while low birthweight then leads to higher risk of MMR (i.e., smoking and birthweight are related) and (2) the effect of being born low birthweight on MMR may be different for those babies whose mothers also smoked, than is the effect of low birthweight for those babies whose mothers did not smoke (i.e., a smoking × birthweight interaction on MMR). Data for this example are presented in Figure 6. Applying the sequential partitioning strategy, PAFA is equal to 14.29%, the adjusted PAFB is equal to 29.12%, and the PAF for the interaction is equal to 6.59%. Applying Equation 11, the PAF for the interaction translates to 4.286 cases of the outcome.

Bottom Line: While the population attributable fraction (PAF) provides potentially valuable information regarding the community-level effect of risk factors, significant limitations exist with current strategies for estimating a PAF in multiple risk factor models.These strategies can result in paradoxical or ambiguous measures of effect, or require unrealistic assumptions regarding variables in the model.In addition, the approach can also be used to statistically control for confounding by other variables, while avoiding the potential pitfalls of attempting to separately differentiate direct and indirect effects.

View Article: PubMed Central - HTML - PubMed

Affiliation: College of Education and Human Development, University of Maine, Orono, ME, USA. craig.mason@umit.maine.edu

ABSTRACT

Background: While the population attributable fraction (PAF) provides potentially valuable information regarding the community-level effect of risk factors, significant limitations exist with current strategies for estimating a PAF in multiple risk factor models. These strategies can result in paradoxical or ambiguous measures of effect, or require unrealistic assumptions regarding variables in the model. A method is proposed in which an overall or total PAF across multiple risk factors is partitioned into components based upon a sequential ordering of effects. This method is applied to several hypothetical data sets in order to demonstrate its application and interpretation in diverse analytic situations.

Results: The proposed method is demonstrated to provide clear and interpretable measures of effect, even when risk factors are related/correlated and/or when risk factors interact. Furthermore, this strategy not only addresses, but also quantifies issues raised by other researchers who have noted the potential impact of population-shifts on population-level effects in multiple risk factor models.

Conclusion: Combined with simple, unadjusted PAF estimates and an aggregate PAF based on all risk factors under consideration, the sequentially partitioned PAF provides valuable additional information regarding the process through which population rates of a disorder may be impacted. In addition, the approach can also be used to statistically control for confounding by other variables, while avoiding the potential pitfalls of attempting to separately differentiate direct and indirect effects.

No MeSH data available.