You are here:
Spot Sampling and Exposure Surrogate Selection as Sources of Bias in Environmental Epidemiology Studies
Sobus, J., K. Christensen, M. Phillips, T. Blessinger, M. Lorber, AND C. Tan. Spot Sampling and Exposure Surrogate Selection as Sources of Bias in Environmental Epidemiology Studies. ISES Annual Meeting, Henderson, NV, Henderson, NV, October 18 - 23, 2015.
The National Exposure Research Laboratory (NERL) Human Exposure and Atmospheric Sciences Division (HEASD) conducts research in support of EPA mission to protect human health and the environment. HEASD research program supports Goal 1 (Clean Air) and Goal 4 (Healthy People) of EPA strategic plan. More specifically, our division conducts research to characterize the movement of pollutants from the source to contact with humans. Our multidisciplinary research program produces Methods, Measurements, and Models to identify relationships between and characterize processes that link source emissions, environmental concentrations, human exposures, and target-tissue dose. The impact of these tools is improved regulatory programs and policies for EPA.
Spot measurements of chemical biomarkers are often used as quantitative exposure surrogates in environmental epidemiology studies. These measures can be expressed a number of different ways – for example, urinary biomarkers can be expressed in units of concentration (µg/ml), adjusted concentration (µg/mg creatinine), excretion rate (µg/min), etc. While based on the same initial spot concentration measure, each of these different exposure surrogates can, when used to predict a health measure of interest, lead to different conclusions about the presence and magnitude of an association. Data from the 2009-2010 NHANES were therefore used to evaluate the impact of exposure surrogate selection on observed epidemiological associations. First, measures of body size (i.e., body mass index [BMI] and waste circumference [WC]) were regressed on five urinary biomarker measures of phthalate metabolites, while controlling for influential covariates. Results varied across models, suggesting bias related to one or more of the surrogate values. To determine which surrogates may be most and least biased, a simulation experiment was performed wherein “true” exposures were known and controlled at the participant level. Specifically, a distribution of di-2-ethylhexyl phthalate (DEHP) daily intake dose was generated and individual values randomly assigned to NHANES participant IDs. Assigned doses, NHANES metadata (e.g., participant age, weight, urine flow rate), and an existing pharmacokinetic model were then used to estimate exposure surrogate values for each participant. Finally, body size measures were regressed on simulated exposure surrogate values and covariates. Results showed that while certain surrogate values were unbiased (e.g., chemical excretion rate), others introduced significant bias that led to spurious associations between DEHP exposure and body size. This presentation will review analyses and results based on the actual NHANES data and the modeling simulations, and will discuss best practices for selecting exposure surrogates based on spot urinary biomarker measurements.