Long-Term Exposure to Fine Particle Elemental Components and Natural and Cause-Specific Mortality-a Pooled Analysis of Eight European Cohorts within the ELAPSE Project

Abstract

Background: Inconsistent associations between long-term exposure to particles with an aerodynamic diameter $\le 2.5\mu m$ [fine particulate matter (${\text{PM}}_{2.5}$)] components and mortality have been reported, partly related to challenges in exposure assessment.

Objectives: We investigated the associations between long-term exposure to ${\text{PM}}_{2.5}$ elemental components and mortality in a large pooled European cohort; to compare health effects of ${\text{PM}}_{2.5}$ components estimated with two exposure modeling approaches, namely, supervised linear regression (SLR) and random forest (RF) algorithms.

Methods: 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 ${\text{PM}}_{2.5}$ 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 $100\times 100m$ 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 ${\text{PM}}_{2.5}$ mass and nitrogen dioxide (${\text{NO}}_2$) separately.

Results: 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 ${\text{PM}}_{2.5}$ and ${\text{NO}}_2$. 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 $2{\text{ng}/m}^3$, adjusting for ${\text{PM}}_{2.5}$ mass. Associations with cause-specific mortality were less consistent in two-pollutant models.

Conclusion: Long-term exposure to V in ${\text{PM}}_{2.5}$ 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.

Figure 1.

Distribution of component exposure at participant addresses estimated from supervised linear regression and random forest models. (A) ${\text{PM}}_{2.5}$ copper and ${\text{PM}}_{2.5}$ iron; (B) ${\text{PM}}_{2.5}$ potassium and ${\text{PM}}_{2.5}$ nickel; (C) ${\text{PM}}_{2.5}$ sulfur and ${\text{PM}}_{2.5}$ silicon; and (D) ${\text{PM}}_{2.5}$ vanadium and ${\text{PM}}_{2.5}$ zinc. The boundary of the box closest to zero indicates P25; the boundary of the box furthest from zero, P75; the bold vertical line inside the box, P50; and the whiskers, P5 and P95. (See Table S11 for exposure distribution of components for the pooled cohort.) Subcohorts are shown from North to South. Note: P, percentile; ${\text{PM}}_{2.5}$, fine particulate matter.

Figure 2.

Associations of ${\text{PM}}_{2.5}$ composition with natural mortality in single-pollutant and two-pollutant models in SLR and RF analyses. Total number of $\text{observations}=323,782$; person-years at $\text{risk}=6,317,235$; number of deaths from natural $\text{mortality}=46,640$. HRs (95% CIs) are presented for the following increments: ${\text{PM}}_{2.5}$ Cu, $5{\hspace{0.17em}\text{ng}/\mathrm{m}}^3$; ${\text{PM}}_{2.5}$ Fe, $100{\hspace{0.17em}\text{ng}/\mathrm{m}}^3$; ${\text{PM}}_{2.5}$ K, $50{\hspace{0.17em}\text{ng}/\mathrm{m}}^3$; ${\text{PM}}_{2.5}$ Ni, $1{\hspace{0.17em}\text{ng}/\mathrm{m}}^3$; ${\text{PM}}_{2.5}$ S, $200{\hspace{0.17em}\text{ng}/\mathrm{m}}^3$; ${\text{PM}}_{2.5}$ Si, $100{\hspace{0.17em}\text{ng}/\mathrm{m}}^3$; ${\text{PM}}_{2.5}$ V, $2{\hspace{0.17em}\text{ng}/\mathrm{m}}^3$; ${\text{PM}}_{2.5}$ Zn, $10{\hspace{0.17em}\text{ng}/\mathrm{m}}^3$. (See Table S14 for corresponding numeric data.) The main model was adjusted for subcohort identification, age, sex, year of enrollment, smoking (status, duration, intensity, and ${\text{intensity}}^2$), BMI categories, marital status, employment status, and 2001 neighborhood-level mean income. In two-pollutant models, ${\text{PM}}_{2.5}$ mass and ${\text{NO}}_2$ exposures were estimated using SLR only. Note: BMI, body mass index; CI, confidence interval; Cu, copper; Fe, iron; HR, hazard ratio; K, potassium; Ni, nickel; ${\text{PM}}_{2.5}$, fine particulate matter; RF, random forest; S, sulfur; Si, silicon; SLR, supervised linear regression; V, vanadium; Zn, zinc.