Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA, USA.
Department of Environmental Health, Harvard T.H. Chan School of Public Health, Boston, MA, USA.
Environ Res. 2023 Jan 1;216(Pt 4):114792. doi: 10.1016/j.envres.2022.114792. Epub 2022 Nov 11.
Previous studies on the impact of measurement error for PM were mostly simulation studies, did not control for other pollutants, or used a single regression calibration model to correct for measurement error. However, the relationship between actual and error-prone PM concentration may vary by time and region. We aim to correct the measurement error of PM predictions using stratified regression calibration and investigate how the measurement error biases the association between PM and mortality in the Medicare Cohort.
The "gold-standard" measurements of PM were defined as daily monitoring data. We regressed daily monitoring PM on modeled PM using the simple linear regression by strata of season, elevation, census division and time period. Calibrated PM was calculated with stratum-specific calibration parameters β (intercept) and β (slope) for each strata and aggregated to annual level. Associations between calibrated and error-prone annual PM and all-cause mortality among Medicare beneficiaries were estimated with Quasi-Poisson regression models.
Across 208 strata, the median of β and β were 0.62 (25% 0.0.20, 75% 1.06) and 0.93 (25% 0.87, 75% 0.99). From calibrated and error-prone PM data, we estimated that each 10 μg/m increase in PM was respectively associated with 4.9% (95%CI 4.6-5.2) and 4.6% (95%CI 4.4-4.9) increases in the mortality rate among Medicare beneficiaries, conditional on confounders.
Regression calibration parameters of PM varied by time and region. Using error-prone measures of PM underestimated the association between PM and all-cause mortality. Modern exposure models produce relatively small bias.
以往关于 PM 测量误差影响的研究大多是模拟研究，没有控制其他污染物，或使用单一回归校准模型来纠正测量误差。然而，实际和易出错的 PM 浓度之间的关系可能因时间和地区而异。我们旨在使用分层回归校准来纠正 PM 预测的测量误差，并研究 PM 测量误差如何偏倚 Medicare 队列中 PM 与死亡率之间的关系。
“金标准”的 PM 测量定义为每日监测数据。我们按季节、海拔、人口普查分区和时间段分层，使用简单线性回归将每日监测 PM 与模型化 PM 进行回归。为每个分层计算校准 PM，使用分层特定的校准参数 β（截距）和 β（斜率）。将校准 PM 汇总到年度水平。使用拟泊松回归模型估计校准和易出错的年度 PM 与 Medicare 受益人的全因死亡率之间的关联。
在 208 个分层中，β 和 β 的中位数分别为 0.62（25% 0.0.20，75% 1.06）和 0.93（25% 0.87，75% 0.99）。从校准和易出错的 PM 数据中，我们估计在调整混杂因素后，每增加 10μg/m 的 PM 分别与 Medicare 受益人的死亡率增加 4.9%（95%CI 4.6-5.2）和 4.6%（95%CI 4.4-4.9）相关。
PM 的回归校准参数因时间和地区而异。使用易出错的 PM 测量值会低估 PM 与全因死亡率之间的关联。现代暴露模型产生的偏差相对较小。
