Background and aims: While PM₁₀ and PM_2.5 exposures have been found to pose substantial health risks in prior studies, few studies have evaluated the health impacts of ultrafine particles (UFPs or PM_0.1, ≤ 0.1µm in diameter). This study estimated the effect of long-term UFPs exposure on mortality in New York State (NYS) and evaluated the sociodemographic disparities and seasonal patterns of this relationship.

Methods: Daily UFPs and other pollutants were obtained from GEOS-Chem-APM, a state-of-the-art chemical transport model with detailed treatments of aerosol microphysics. Interquartile range (IQR) and high-level UFPs (>75^th percentile) were used as the two exposure indicators. Mortality data was obtained from Vital Records for Upstate NY and NY City, respectively. Outcomes included total (nonaccidental) mortality and cause-specific mortality due to respiratory and cardiovascular (CVD) diseases, mental disorders, neoplasms, and diseases of nervous systems. We adopted a modified difference-in-differences model to examine the UFPs-mortality relationship at county-subdivision levels by controlling for time-invariant spatial confounders (sociodemographic variables), temporal factors (annual and secular trend), and NH₃, O₃, SO₂, temperature, and relative humidity.

Results: The excess risk (ER_IQR) for long-term UFPs exposure on total nonaccidental mortality was 3.67% (95%CI: -0.65%, 8.18%). The risk ratio (RR) for high UFPs exposure (>= 75%) was 1.10 (95%CI: 1.05, 1.17). UFPs-associated risks were statistically significantly increased for CVD mortality (RR=1.11, 95%CI: 1.05, 1.18) and neoplasms (RR=1.09, 95%CI: 1.02, 1.16). Compared to non-Hispanic Whites, non-Hispanic Blacks (RR=1.40, 95%CI: 1.16, 1.68) and Hispanics (RR=1.13, 95%CI: 1.00, 1.29) experienced significantly higher mortality risk after long-term exposure to high-level UFPs. The risk estimates were also increased for non-NYC residents, during winter and the largest among children under 5.

Conclusions: Long-term UFPs exposure was significantly associated with increased total nonaccidental mortality, and mortality for CVD and neoplasms. We also identified UFPs-mortality disparities by race/ethnicity, age, and urbanicity.