Causal Models for Estimating the Effects of Weight Gain on Mortality - Statistics > ApplicationsReportar como inadecuado

Causal Models for Estimating the Effects of Weight Gain on Mortality - Statistics > Applications - Descarga este documento en PDF. Documentación en PDF para descargar gratis. Disponible también para leer online.

Abstract: Suppose, contrary to fact, in 1950, we had put the cohort of 18 year oldnon-smoking American men on a stringent mandatory diet that guaranteed that noone would ever weigh more than their baseline weight established at age 18. Howwould the counter-factual mortality of these 18 year olds have compared totheir actual observed mortality through 2007? We describe in detail how thiscounterfactual contrast could be estimated from longitudinal epidemiologic datasimiliar to that stored in the electronic medical records of a large HMO byapplying g-estimation to a novel structural nested model. Our analytic approachdiffers from any alternative approach in that in that, in the abscence of modelmisspecification, it can successfully adjust for i measured time-varyingconfounders such as exercise, hypertension and diabetes that are simultaneouslyintermediate variables on the causal pathway from weight gain to death anddeterminants of future weight gain, ii unmeasured confounding by undiagnosedpreclinical disease i.e reverse causation that can cause both poor weightgain and premature mortality provided an upper bound can be specified for themaximum length of time a subject may suffer from a subclinical illness severeenough to affect his weight without the illness becomes clinically manifest,and iii the prescence of particular identifiable subgroups, such as thosesuffering from serious renal, liver, pulmonary, and-or cardiac disease, in whomconfounding by unmeasured prognostic factors so severe as to render useless anyattempt at direct analytic adjustment.

Autor: James Robins


Documentos relacionados