Sources of personal PM2.5 exposure during pregnancy in the MADRES cohort

Abstract

Background: Personal exposure to fine particulate matter (PM_2.5) is impacted by different sources each with different chemical composition. Determining these sources is important for reducing personal exposure and its health risks especially during pregnancy.

Objective: Identify main sources and their contributions to the personal PM_2.5 exposure in 213 women in the 3rd trimester of pregnancy in Los Angeles, CA.

Methods: We measured 48-hr integrated personal PM_2.5 exposure and analyzed filters for PM_2.5 mass, elemental composition, and optical carbon fractions. We used the EPA Positive Matrix Factorization (PMF) model to resolve and quantify the major sources of personal PM_2.5 exposure. We then investigated bivariate relationships between sources, time-activity patterns, and environmental exposures in activity spaces and residential neighborhoods to further understand sources.

Results: Mean personal PM_2.5 mass concentration was 22.3 (SD = 16.6) μg/m³. Twenty-five species and PM_2.5 mass were used in PMF with a final R² of 0.48. We identified six sources (with major species in profiles and % contribution to PM_2.5 mass) as follows: secondhand smoking (SHS) (brown carbon, environmental tobacco smoke; 65.3%), fuel oil (nickel, vanadium; 11.7%), crustal (aluminum, calcium, silicon; 11.5%), fresh sea salt (sodium, chlorine; 4.7%), aged sea salt (sodium, magnesium, sulfur; 4.3%), and traffic (black carbon, zinc; 2.6%). SHS was significantly greater in apartments compared to houses. Crustal source was correlated with more occupants in the household. Aged sea salt increased with temperature and outdoor ozone, while fresh sea salt was highest on days with westerly winds from the Pacific Ocean. Traffic was positively correlated with ambient NO₂ and traffic-related NO_x at residence. Overall, 76.8% of personal PM_2.5 mass came from indoor or personal compared to outdoor sources.

Impact: We conducted source apportionment of personal PM_2.5 samples in pregnancy in Los Angeles, CA. Among identified sources, secondhand smoking contributed the most to the personal exposure. In addition, traffic, crustal, fuel oil, fresh and aged sea salt sources were also identified as main sources. Traffic sources contained markers of combustion and non-exhaust wear emissions. Crustal source was correlated with more occupants in the household. Aged sea salt source increased with temperature and outdoor ozone and fresh sea salt source was highest on days with westerly winds from the Pacific Ocean.

Keywords: PM2.5; Personal exposure; Secondhand smoking source; Source apportionment; Traffic source; pregnancy.

Fig. 1. PMF-predicted source loading profiles (in % of species).

Sources are color coded as shown: .

Fig. 2. Relationship between source mass contributions (y-axis) and environmental or home characteristics.

a Secondhand smoking and home type, b aged sea salt and window opening time in the 48-h monitoring period, c fresh sea salt and average wind direction in the 48-h monitoring period, and d crustal and number of household occupants (*The relationships in (b d) were significant with Kruskal-Wallis test p value  < 0.05) .