Institute for Risk Assessment Sciences, Utrecht University, Utrecht, Netherlands.
Department of Hygiene, Epidemiology and Medical Statistics, School of Medicine, National and Kapodistrian University of Athens, Athens, Greece.
Environ Health Perspect. 2021 Apr;129(4):47009. doi: 10.1289/EHP8368. Epub 2021 Apr 12.
Inconsistent associations between long-term exposure to particles with an aerodynamic diameter [fine particulate matter ()] components and mortality have been reported, partly related to challenges in exposure assessment.
We investigated the associations between long-term exposure to elemental components and mortality in a large pooled European cohort; to compare health effects of components estimated with two exposure modeling approaches, namely, supervised linear regression (SLR) and random forest (RF) algorithms.
We pooled data from eight European cohorts with 323,782 participants, average age 49 y at baseline (1985-2005). Residential exposure to 2010 annual average concentration of eight components [copper (Cu), iron (Fe), potassium (K), nickel (Ni), sulfur (S), silicon (Si), vanadium (V), and zinc (Zn)] was estimated with Europe-wide SLR and RF models at a scale. We applied Cox proportional hazards models to investigate the associations between components and natural and cause-specific mortality. In addition, two-pollutant analyses were conducted by adjusting each component for mass and nitrogen dioxide () separately.
We observed 46,640 natural-cause deaths with 6,317,235 person-years and an average follow-up of 19.5 y. All SLR-modeled components were statistically significantly associated with natural-cause mortality in single-pollutant models with hazard ratios (HRs) from 1.05 to 1.27. Similar HRs were observed for RF-modeled Cu, Fe, K, S, V, and Zn with wider confidence intervals (CIs). HRs for SLR-modeled Ni, S, Si, V, and Zn remained above unity and (almost) significant after adjustment for both and . HRs only remained (almost) significant for RF-modeled K and V in two-pollutant models. The HRs for V were 1.03 (95% CI: 1.02, 1.05) and 1.06 (95% CI: 1.02, 1.10) for SLR- and RF-modeled exposures, respectively, per , adjusting for mass. Associations with cause-specific mortality were less consistent in two-pollutant models.
Long-term exposure to V in was most consistently associated with increased mortality. Associations for the other components were weaker for exposure modeled with RF than SLR in two-pollutant models. https://doi.org/10.1289/EHP8368.
长期暴露于空气动力学直径为 [细颗粒物 ()] 的颗粒及其成分与死亡率之间的关联一直存在争议，部分原因与暴露评估方面的挑战有关。
我们在一个大型欧洲队列中研究了长期暴露于元素成分与死亡率之间的关系，并比较了两种暴露建模方法（即监督线性回归 [SLR] 和随机森林 [RF] 算法）估计的元素成分的健康效应。
我们汇集了来自八个欧洲队列的数据，共有 323782 名参与者，基线时平均年龄为 49 岁（1985-2005 年）。采用欧洲范围内的 SLR 和 RF 模型，以 为尺度，估算了 2010 年 8 种 元素 [铜 (Cu)、铁 (Fe)、钾 (K)、镍 (Ni)、硫 (S)、硅 (Si)、钒 (V) 和锌 (Zn)] 的年平均浓度的居住暴露量。我们应用 Cox 比例风险模型来研究各成分与自然和特定原因死亡率之间的关联。此外，还通过分别为每个成分调整 质量和二氧化氮 () 进行了双污染物分析。
我们观察到 46640 例自然原因死亡，有 6317235 人年随访，平均随访时间为 19.5 年。在单污染物模型中，所有基于 SLR 模型的成分与自然原因死亡率均具有统计学意义，风险比 (HR) 为 1.05 至 1.27。RF 模型估算的 Cu、Fe、K、S、V 和 Zn 的 HR 也观察到相似的结果，但置信区间 (CI) 更宽。基于 SLR 模型估算的 Ni、S、Si、V 和 Zn 的 HR 在调整 和 后仍高于 1 且（几乎）有统计学意义。仅在双污染物模型中，RF 模型估算的 K 和 V 的 HR （几乎）具有统计学意义。对于暴露于 SLR 和 RF 模型的每 ，V 的 HR 分别为 1.03（95%CI：1.02，1.05）和 1.06（95%CI：1.02，1.10），同时调整了 质量。在双污染物模型中，与特定原因死亡率的关联不太一致。
长期暴露于 中的 V 与死亡率升高最密切相关。在双污染物模型中，基于 RF 模型估算的其他成分的关联比基于 SLR 模型的弱。https://doi.org/10.1289/EHP8368.
Zotero 插件