Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
A Simulation-Based Assessment of Alternative Explanations for Apparent Confounding in “PM Decomposition” Studies
https://orcid.org/0000-0001-6418-5188
Abstract Determining whether there is a causal link between chronic exposure to $ PM_{2.5} $ (atmospheric particulate matter with a diameter of 2.5 µm or less) and mortality is a fundamental challenge in determining appropriate air quality standards for $ PM_{2.5} $. While numerous epidemiological studies have found a statistical association between $ PM_{2.5} $ and mortality, unmeasured confounding in this relationship remains a concern. Several recent studies have examined the possibility of unmeasured confounding of the long-term association between $ PM_{2.5} $ exposure and mortality by decomposing $ PM_{2.5} $ into two orthogonal components — a “global” component that measures the national trend in pollution, and a “local” component that measures the local trend in pollution after controlling for the national trend. Generally, these “PM decomposition” studies find that while $ PM_{2.5} $ and mortality are trending downward over time at the national level, areas with steeper declines in $ PM_{2.5} $ do not have correspondingly steep declines in mortality. This finding is consistent with the long-term relationship between $ PM_{2.5} $ and mortality being confounded by some other, unmeasured, long-term trends. However, alternative explanations for these findings have been proposed under which there is still a causal link between $ PM_{2.5} $ and mortality. This study conducts simulation-based tests of four of these proposed alternative explanations — the omission of spatial variation in $ PM_{2.5} $ and mortality, confounding at the local level, measurement error, and an association between $ PM_{2.5} $ and mortality at a different time scale than those tested in the PM decomposition studies. We find that none of these alternative explanations can reproduce the results in the PM decomposition studies while simultaneously allowing for a causal link between $ PM_{2.5} $ and mortality.
A Simulation-Based Assessment of Alternative Explanations for Apparent Confounding in “PM Decomposition” Studies
https://orcid.org/0000-0001-6418-5188
Abstract Determining whether there is a causal link between chronic exposure to $ PM_{2.5} $ (atmospheric particulate matter with a diameter of 2.5 µm or less) and mortality is a fundamental challenge in determining appropriate air quality standards for $ PM_{2.5} $. While numerous epidemiological studies have found a statistical association between $ PM_{2.5} $ and mortality, unmeasured confounding in this relationship remains a concern. Several recent studies have examined the possibility of unmeasured confounding of the long-term association between $ PM_{2.5} $ exposure and mortality by decomposing $ PM_{2.5} $ into two orthogonal components — a “global” component that measures the national trend in pollution, and a “local” component that measures the local trend in pollution after controlling for the national trend. Generally, these “PM decomposition” studies find that while $ PM_{2.5} $ and mortality are trending downward over time at the national level, areas with steeper declines in $ PM_{2.5} $ do not have correspondingly steep declines in mortality. This finding is consistent with the long-term relationship between $ PM_{2.5} $ and mortality being confounded by some other, unmeasured, long-term trends. However, alternative explanations for these findings have been proposed under which there is still a causal link between $ PM_{2.5} $ and mortality. This study conducts simulation-based tests of four of these proposed alternative explanations — the omission of spatial variation in $ PM_{2.5} $ and mortality, confounding at the local level, measurement error, and an association between $ PM_{2.5} $ and mortality at a different time scale than those tested in the PM decomposition studies. We find that none of these alternative explanations can reproduce the results in the PM decomposition studies while simultaneously allowing for a causal link between $ PM_{2.5} $ and mortality.
A Simulation-Based Assessment of Alternative Explanations for Apparent Confounding in “PM Decomposition” Studies
https://orcid.org/0000-0001-6418-5188
Abstract Determining whether there is a causal link between chronic exposure to $ PM_{2.5} $ (atmospheric particulate matter with a diameter of 2.5 µm or less) and mortality is a fundamental challenge in determining appropriate air quality standards for $ PM_{2.5} $. While numerous epidemiological studies have found a statistical association between $ PM_{2.5} $ and mortality, unmeasured confounding in this relationship remains a concern. Several recent studies have examined the possibility of unmeasured confounding of the long-term association between $ PM_{2.5} $ exposure and mortality by decomposing $ PM_{2.5} $ into two orthogonal components — a “global” component that measures the national trend in pollution, and a “local” component that measures the local trend in pollution after controlling for the national trend. Generally, these “PM decomposition” studies find that while $ PM_{2.5} $ and mortality are trending downward over time at the national level, areas with steeper declines in $ PM_{2.5} $ do not have correspondingly steep declines in mortality. This finding is consistent with the long-term relationship between $ PM_{2.5} $ and mortality being confounded by some other, unmeasured, long-term trends. However, alternative explanations for these findings have been proposed under which there is still a causal link between $ PM_{2.5} $ and mortality. This study conducts simulation-based tests of four of these proposed alternative explanations — the omission of spatial variation in $ PM_{2.5} $ and mortality, confounding at the local level, measurement error, and an association between $ PM_{2.5} $ and mortality at a different time scale than those tested in the PM decomposition studies. We find that none of these alternative explanations can reproduce the results in the PM decomposition studies while simultaneously allowing for a causal link between $ PM_{2.5} $ and mortality.
A Simulation-Based Assessment of Alternative Explanations for Apparent Confounding in “PM Decomposition” Studies
Chang, Wonjun (Autor:in) / Glasgow, Garrett (Autor:in) / Ramkrishnan, Bharat (Autor:in) / Smith, Anne E. (Autor:in)
2022
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
BKL:
43.00
Umweltforschung, Umweltschutz: Allgemeines
/
43.00$jUmweltforschung$jUmweltschutz: Allgemeines
Out of Alternative Explanations
| British Library Conference Proceedings | 2010
France Confounding Lightness: Gilles Perraudin
| British Library Online Contents | 2003