Association of Long-Term Residentially Assigned Speciated Atmospheric Mercury Emission Flux with All-Cause and Cause-Specific Mortality

Abstract

Abstract Mercury (Hg) is a persistent global pollutant, but population evidence on residentially assigned species-specific atmospheric Hg emission flux and mortality is limited. We analyzed UK Biobank cohorts for all-cause/other-cause (N = 322,666), cardiovascular disease (N = 344,333), and cancer mortality (N = 381,432). Annual EDGAR v8.1 fluxes of total Hg, gaseous elemental Hg (Hg0), gaseous oxidized Hg (Hg2+), and particle-bound Hg (Hg–P) were assigned to geocoded residences. Cox models estimated hazard ratios (HRs) across quartiles; restricted cubic splines and quantile g-computation assessed exposure-response and mixture associations. During follow-up, 20,370 deaths occurred. Higher residentially assigned Hg emission flux was associated with higher all-cause and cause-specific mortality. For all-cause mortality, HRs (95% CIs) for Q3 and Q4 versus Q1 were 1.16 (1.11, 1.22) and 1.13 (1.07, 1.18) for total Hg; 1.13 (1.08, 1.19) and 1.14 (1.08, 1.19) for Hg0; 1.19 (1.13, 1.24) and 1.13 (1.07, 1.19) for Hg2+; and 1.17 (1.11, 1.23) and 1.18 (1.12, 1.25) for Hg–P, respectively. Hg–P received the largest positive quantile g-computation weight for several outcomes, but the Hg species were strongly correlated. In conclusion, higher residential Hg emission flux was associated with mortality in primary analyses, but sensitivity to temporal exposure specification warrants cautious interpretation and confirmation.