Multipollutant Measurement Error in Air Pollution Epidemiology Studies Arising from Predicting Exposures with Penalized Regression Splines

Bergen, Silas; Sheppard, Lianne; Kaufman, Joel D.; Szpiro, Adam A.

doi:10.1111/rssc.12144

Cited by 19 publications

(11 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…However, only a single pollutant measure is collected and multipollutant analyses, like in Bergen et al. (), are not feasible.…”

Section: Discussionmentioning

confidence: 99%

“…Epidemiological longitudinal studies of acute health effects of outdoor air pollution often use estimated exposure of the study participants by one or several fixed outdoor monitoring sites located in an urban background of the study area (e.g. Alexeeff, Carroll, & Coull, ; Bergen, Sheppard, Kaufman, & Szpiro, ). The ambient air pollutant concentrations measured at the fixed sites are assumed to represent the population–averaged exposure and this value usually enters the analysis.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Mixtures of Berkson and classical covariate measurement error in the linear mixed model: Bias analysis and application to a study on ultrafine particles

Deffner¹,

Küchenhoff²,

Breitner

et al. 2018

Biometrical J

View full text Add to dashboard Cite

The ultrafine particle measurements in the Augsburger Umweltstudie, a panel study conducted in Augsburg, Germany, exhibit measurement error from various sources. Measurements of mobile devices show classical possibly individual-specific measurement error; Berkson-type error, which may also vary individually, occurs, if measurements of fixed monitoring stations are used. The combination of fixed site and individual exposure measurements results in a mixture of the two error types. We extended existing bias analysis approaches to linear mixed models with a complex error structure including individual-specific error components, autocorrelated errors, and a mixture of classical and Berkson error. Theoretical considerations and simulation results show, that autocorrelation may severely change the attenuation of the effect estimations. Furthermore, unbalanced designs and the inclusion of confounding variables influence the degree of attenuation. Bias correction with the method of moments using data with mixture measurement error partially yielded better results compared to the usage of incomplete data with classical error. Confidence intervals (CIs) based on the delta method achieved better coverage probabilities than those based on Bootstrap samples. Moreover, we present the application of these new methods to heart rate measurements within the Augsburger Umweltstudie: the corrected effect estimates were slightly higher than their naive equivalents. The substantial measurement error of ultrafine particle measurements has little impact on the results. The developed methodology is generally applicable to longitudinal data with measurement error.

show abstract

“…However, only a single pollutant measure is collected and multipollutant analyses, like in Bergen et al. (), are not feasible.…”

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Mixtures of Berkson and classical covariate measurement error in the linear mixed model: Bias analysis and application to a study on ultrafine particles

Deffner¹,

Küchenhoff²,

Breitner

et al. 2018

Biometrical J

View full text Add to dashboard Cite

show abstract

“…Denote the collection of all monitor observations

{boldx}^{*} = {false(x false({bolds}_{1}^{*} false), …, x false({bolds}_{n^{*}}^{*} false) false)}^{sans-serifT}

and their corresponding covariates

{boldR}^{*} = {false(boldr false({bolds}_{1}^{*} false), …, boldr false({bolds}_{n^{*}}^{*} false) false)}^{sans-serifT}

. Following Yu and Ruppert () and Bergen, Sheppard, Kaufman, and Szpiro (), we can write the model coefficients as

\begin{matrix} alignleft \\ align-1 {\hat{γ}}_{λ, n^{*}} & align-2 = \underset{θ}{argmin} \frac{1}{n^{*}} \sum_{i = 1}^{n^{*}} {(x (s_{i}^{*}) - r {(s_{i}^{*})}^{T} θ)}^{2} + λ θ^{T} D θ \\ align-1 & align-2 = {({R^{*}}^{T} R^{*} + λ n^{*} D)}^{- 1} {R^{*}}^{T} x, \end{matrix}

where λ is a fixed penalty parameter and

D \in {double-struckR}^{p \times p}

is a weighting matrix for the...…”

Section: Error In the Area‐level Average Estimatesmentioning

confidence: 99%

“…Bergen and Szpiro () develop a method for bias‐correcting

{\overset{β}{true}}_{1}

that accounts for bias in predictions from a penalized‐regression exposure model. However, that approach, like most other measurement error methodology (Alexeeff, Carroll, & Coull, ; Bergen et al, ; Gryparis, Paciorek, Zeka, Schwartz, & Coull, ; Szpiro & Paciorek, ), was developed under a linear health model, and there currently is no analogue for a generalized linear model setting. However, a nonparametric bootstrap procedure can be used to correct the standard errors of

{\overset{β}{true}}_{1}

(Bergen & Szpiro, ; Keller et al, ; Szpiro & Paciorek, ).…”

Section: Impact On Health Model Inferencementioning

confidence: 99%

Error in estimating area‐level air pollution exposures for epidemiology

Keller

Peng

2019

Environmetrics

View full text Add to dashboard Cite

In air pollution epidemiology, measurements of relevant exposure concentrations are typically made at point locations, resulting in spatial misalignment between the exposure data and health outcomes aggregated at the area level. To obtain values that match the spatial units of the health data, observations can be averaged directly or prediction models developed. We present a framework for evaluating the error in aggregating exposure concentrations to the area unit. We present estimators of mean squared error that can be used for model selection. We find that exposure prediction models, even when misspecified, outperform monitor averages in settings with realistic numbers of monitors and that important reductions in error of the health effect estimate can be obtained when restricting to areas with a monitor. In an analysis of long‐term particulate matter concentrations across the United States, we estimated the error of the prediction model approach to be less than that of the monitor averaging approach on average across counties. We present health effect estimates about particulate matter exposure and pediatric asthma morbidity in the Medicaid population using each approach. Our findings support the use of a prediction model for estimating area‐wide averages, even when restricting to areas that contain a monitor.

show abstract

“…It should be noted that multipollutant models have their own issues with interpretation (Bergen et al 2016;Dionisio, Baxter, and Chang 2014). One issue is that of exposure error, for which there is a large literature related to air pollution epidemiology studies.…”

Section: Co-pollutants In Panel Studiesmentioning

confidence: 99%

Ozone exposure and pulmonary effects in panel and human clinical studies: Considerations for design and interpretation

Rohr

2018

Journal of the Air & Waste Management Association

View full text Add to dashboard Cite

The pulmonary health impacts of ozone exposure have been well documented using both epidemiological and chamber study designs. However, there are a number of specific methodological and related issues that should be considered when interpreting the results of these studies and planning additional research, including the standardization of exposure and health metrics to facilitate comparisons among studies.

show abstract

Multipollutant Measurement Error in Air Pollution Epidemiology Studies Arising from Predicting Exposures with Penalized Regression Splines

Cited by 19 publications

References 35 publications

Mixtures of Berkson and classical covariate measurement error in the linear mixed model: Bias analysis and application to a study on ultrafine particles

Mixtures of Berkson and classical covariate measurement error in the linear mixed model: Bias analysis and application to a study on ultrafine particles

Error in estimating area‐level air pollution exposures for epidemiology

Ozone exposure and pulmonary effects in panel and human clinical studies: Considerations for design and interpretation

Contact Info

Product

Resources

About