Intervening on Risk Factors for Coronary Heart Disease: An Application of the Parametric GFormula
View Presentation
Miguel A Hernan, Harvard University
Murray A Mittleman, Harvard University
James M Robins, Harvard University
*Sarah Taubman, University of Pennsylvania
Keywords: causal methods, gformula, marginal structural models, coronary heart disease
Estimating the population risk of disease under hypothetical interventions—such as the population risk of coronary heart disease were everyone to quit smoking and start exercising or to start exercising if diagnosed with diabetes—can be difficult, if not impossible, using standard analytic techniques. The parametric gformula appropriately adjusts for timevarying confounders affected by prior exposures, and it is especially well suited to estimating the effect of interventions when they involve multiple risk factors (joint interventions) or depend on the value of risk factors (dynamic interventions). We use the parametric gformula to estimate the effect of various lifestyle interventions on the risk of coronary heart disease using data from the Nurses’ Health Study. We contrast this approach with the inverse probability weighting of marginal structural models.
