MODEL-BASED ESTIMATION OF POPULATION ATTRIBUTABLE RISK UNDER CROSS-SECTIONAL SAMPLING

The covariate-adjusted population attributable risk (PAR) measures the proportionate reduced in disease prevalence in the target population when th putative risk factor is removed, after adjusting for covariate effects. This paper extends the model-based approach developed for re trospective and cohort studies to the cross-sectional sampling design. An appropriate legit linear model is utilized to estimate the covaria te-adjusted attributable risk, The asymptotic variance of this complex ratio estimate is obtained using Taylor series expansions which incor porate the sampling variation of the estimated model parameters and th e appropriate estimates of risk factor prevalence, These methods are i llustrated with cardiovascular disease risk factor data from the secon d National Health and Nutrition Examination Survey (NHANES II).