Dataset Information

Temporal aspects of air pollutant measures in epidemiologic analysis: a simulation study.

ABSTRACT: Numerous observational studies have assessed the association between ambient air pollution and chronic disease incidence, but there is no uniform approach to create an exposure metric that captures the variability in air pollution through time and determines the most relevant exposure window. The purpose of the present study was to assess ways of modeling exposure to air pollution in relation to incident hypertension. We simulated data on incident hypertension to assess the performance of six air pollution exposure metrics, using characteristics from the Black Women's Health Study. Each metric made different assumptions about how to incorporate time trends in pollutant data, and the most relevant window of exposure. We use observed values for particulate matter ? 2.5 microns (PM2.5) for this cohort to create the six exposure metrics and fit Cox proportional hazards models to the simulated data using the six metrics. The optimal exposure metric depends on the underlying association between PM2.5 and disease, which is unknown. Metrics that incorporate exposure information from multiple years tend to be more robust and suffer from less bias. This study provides insight into factors that influence the metric used to quantifying exposure to PM2.5 and suggests the need for careful sensitivity analyses.

SUBMITTER: White LF

PROVIDER: S-EPMC4726372 | biostudies-literature | 2016 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Temporal aspects of air pollutant measures in epidemiologic analysis: a simulation study.

White Laura F LF Yu Jeffrey J Jerrett Michael M Coogan Patricia P

Scientific reports 20160121

Numerous observational studies have assessed the association between ambient air pollution and chronic disease incidence, but there is no uniform approach to create an exposure metric that captures the variability in air pollution through time and determines the most relevant exposure window. The purpose of the present study was to assess ways of modeling exposure to air pollution in relation to incident hypertension. We simulated data on incident hypertension to assess the performance of six ai ...[more]

PMID: 26791428

Similar Datasets

Project description:BACKGROUND:Cohort studies have documented associations between fine particulate matter air pollution (PM2.5) and mortality risk. However, there remains uncertainty regarding the contribution of co-pollutants and the stability of pollution-mortality associations in models that include multiple air pollutants. Furthermore, it is unclear whether the PM2.5-mortality relationship varies spatially, when exposures are decomposed according to scale of spatial variability, or temporally, when effect estimates are allowed to change between years. METHODS:A cohort of 635,539 individuals was compiled using public National Health Interview Survey (NHIS) data from 1987 to 2014 and linked with mortality follow-up through 2015. Modelled air pollution exposure estimates for PM2.5, other criteria air pollutants, and spatial decompositions (<?1?km, 1-10?km, 10-100?km, >?100?km) of PM2.5 were assigned at the census-tract level. The NHIS samples were also divided into yearly cohorts for temporally-decomposed analyses. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) in regression models that included up to six criteria pollutants; four spatial decompositions of PM2.5; and two- and five-year lagged mean PM2.5 exposures in the temporally-decomposed cohorts. Meta-analytic fixed-effect estimates were calculated using results from temporally-decomposed analyses and compared with time-independent results using 17- and 28-year exposure windows. RESULTS:In multiple-pollutant analyses, PM2.5 demonstrated the most robust pollutant-mortality association. Coarse fraction particulate matter (PM2.5-10) and sulfur dioxide (SO2) were also associated with excess mortality risk. The PM2.5-mortality association was observed across all four spatial scales of PM2.5, with higher but less precisely estimated HRs observed for local (<?1?km) and neighborhood (1-10?km) variations. In temporally-decomposed analyses, the PM2.5-mortality HRs were stable across yearly cohorts. The meta-analytic HR using two-year lagged PM2.5 equaled 1.10 (95% CI 1.07, 1.13) per 10??g/m3. Comparable results were observed in time-independent analyses using a 17-year (HR 1.13, CI 1.09, 1.16) or 28-year (HR 1.09, CI 1.07, 1.12) exposure window. CONCLUSIONS:Long-term exposures to PM2.5, PM2.5-10, and SO2 were associated with increased risk of all-cause and cardiopulmonary mortality. Each spatial decomposition of PM2.5 was associated with mortality risk, and PM2.5-mortality associations were consistent over time.

Project description:BackgroundMost epidemiological studies estimate associations without considering exposure measurement error. While some studies have estimated the impact of error in single-exposure models we aimed to quantify the effect of measurement error in multi-exposure models, specifically in time-series analysis of PM2.5, NO2, and mortality using simulations, under various plausible scenarios for exposure errors. Measurement error in multi-exposure models can lead to effect transfer where the effect estimate is overestimated for the pollutant estimated with more error to the one estimated with less error. This complicates interpretation of the independent effects of different pollutants and thus the relative importance of reducing their concentrations in air pollution policy.MethodsMeasurement error was defined as the difference between ambient concentrations and personal exposure from outdoor sources. Simulation inputs for error magnitude and variability were informed by the literature. Error-free exposures with their consequent health outcome and error-prone exposures of various error types (classical/Berkson) were generated. Bias was quantified as the relative difference in effect estimates of the error-free and error-prone exposures.ResultsMortality effect estimates were generally underestimated with greater bias observed when low ratios of the true exposure variance over the error variance were assumed (27.4% underestimation for NO2). Higher ratios resulted in smaller, but still substantial bias (up to 19% for both pollutants). Effect transfer was observed indicating that less precise measurements for one pollutant (NO2) yield more bias, while the co-pollutant (PM2.5) associations were found closer to the true. Interestingly, the sum of single-pollutant model effect estimates was found closer to the summed true associations than those from multi-pollutant models, due to cancelling out of confounding and measurement error bias.ConclusionsOur simulation study indicated an underestimation of true independent health effects of multiple exposures due to measurement error. Using error parameter information in future epidemiological studies should provide more accurate concentration-response functions.

Dataset Information

Temporal aspects of air pollutant measures in epidemiologic analysis: a simulation study.

Publications

Temporal aspects of air pollutant measures in epidemiologic analysis: a simulation study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets