Lack of spatial variation of endotoxin in ambient particulate matter across a German metropolitan area

Abstract

In this study, we describe the spatial variation of endotoxin across an urban setting using Geographic Information Systems (GIS) methods. We also identify potential sources of endotoxin that account for between-site variability and compare endotoxin levels in particulate matter with a 50% aerodynamic cut-off diameter of 2.5 μm (PM_2.5) and of 10 μm (PM₁₀). In 1999–2000, we collected PM_2.5 and PM₁₀ in Munich urban air and measured soluble endotoxin concentrations in both particle fractions. Using Teflon filters and Harvard impactors, PM_2.5 was collected at 40 outdoor monitoring sites across Munich and PM₁₀ at a subset of these sites ($n=12$). Approximately four samples were collected at each site for a total of 158 PM_2.5 samples and 48 PM₁₀ samples. We visited and characterized the surrounding 100 m of each site for potential endotoxin sources. The geometric mean endotoxin concentration for all sites was 1.46 EU mg⁻¹ PM_2.5 (95% confidence intervals (CI): 1.21–1.77) and at the subset of the sites was 1.30 EU mg⁻¹ PM_2.5 (95% CI: 1.01–1.67 EU mg⁻¹ PM_2.5). Endotoxin levels in PM₁₀ were higher, 3.91 EU mg⁻¹ PM₁₀ (95% CI: 3.03–5.03 EU mg⁻¹ PM₁₀), than in PM_2.5 and were moderately correlated, $r=0.51$.

All endotoxin concentrations measured in this study were <5.5 EU m⁻³ and thus lower than the accepted thresholds for acute adverse health effects for occupational exposures. Sites with more potential sources ($n⩾3$) had slightly higher mean endotoxin levels (MR: 1.30 for EU mg⁻¹ PM_2.5 and 1.13 for EU m⁻³ PM_2.5) than sites with no identified sources. Based on the ranges of endotoxin levels at the different sites, we found very little spatial variation in ambient endotoxin concentrations across the metropolitan area of Munich using inverse distance weighting method (IDW) methods ($R^2=0.013$ for EU mg⁻¹ PM_2.5 and $R^2=0.020$ for EU m⁻³ PM_2.5). Potential sources of endotoxin surrounding the sites only partly explained the variation seen.

Introduction

Endotoxins are part of the outer membrane of the cell wall of Gram-negative bacteria. Although the term endotoxin is occasionally used to refer to any cell-associated bacterial toxin, it is properly reserved to refer to the lipopolysaccharide complex associated with the outer membrane of Gram-negative bacteria such as Escherichia coli, Salmonella and other leading pathogens. Inhalation of bacterial endotoxin or its chemically pure form called lipopolysaccharide (LPS) in doses as low as 4–15 ng m⁻³ has been associated with acute and chronic airways inflammation and lung function decrements (Douwes and Heederik, 1997; Milton et al., 1996). Given its toxicity and pro-inflammatory effects, endotoxin has been studied in specific occupational settings and in house dust for more than 30 yr (Douwes et al., 2000; Rylander, 2002). Ambient exposure to endotoxin, however, has not well been characterized; at present only three studies have measured systematically endotoxin concentrations in ambient particles (Carty et al., 2003; Heinrich et al., 2003b; Mueller-Anneling et al., 2004). Carty et al. (2003) found that endotoxin levels were significantly related to ambient temperature and relative humidity. Heinrich et al. (2003b) collected fine and coarse fractions of particles in the two small German towns of Hettstedt and Zerbst, which are approximately 100 km apart. The average endotoxin levels were not significantly different between the two towns, but endotoxin content in PM₁₀ was approximately 10-fold higher than the PM_2.5 content at both locations. The study of Mueller-Anneling et al. (2004) provides a characterization of endotoxin concentration across a large metropolitan area, but only in relation to PM₁₀. They reported that the desert and mountain communities had the highest endotoxin levels, whereas samples collected in rural areas had mid-range endotoxin levels. The city of Los Angeles itself had the lowest endotoxin results.

Recent studies have focused on the physical properties, e.g. surface size or chemical compostion of ambient particles and bioaersols. However, the exact constituents of air pollution that cause disease and the precise mechanisms involved are complex. Numerous studies have been conducted to determine which components of particulate matter (PM) may contribute to airway inflammation and irritation (Soukup and Becker, 2001). Bacterial endotoxin is one such biologically active constituent of PM that has been shown to trigger a complex inflammatory response in humans. The other justifying point of this investigation is related to the characterization of ambient particle properties, which might be involved in causing the known adverse health effects.

Research on spatial patterns of airborne endotoxin levels may be useful in determining certain outdoor endotoxin sources and might improve assessment of exposure to airborne endotoxin. Several sources of indoor endotoxin in settled house dust have been described (Heinrich et al., 2001; Park et al., 2001b). Park et al. described that it is likely that outdoor endotoxin levels have an influence on indoor levels, e.g. when windows are open (Park et al., 2000). Comparisons of indoor and outdoor endotoxin concentrations in factories (Su et al., 2002), farms (Chang et al., 2001) and in urban homes (Long et al., 2001; Park et al., 2000) have also been described in the literature. In general, levels of endotoxin are higher indoors and inside factories than in the nearby outdoor environments (Long et al., 2001; Park et al., 2000; Su et al., 2002) or in the surrounding areas (includes administrative building, farm perimeters) (Chang et al., 2001). High levels of endotoxin have also been found in agriculture and related industries such as garbage handling and compost facilities (Sigsgaard et al., 1994), cotton mills (Christiani et al., 1993) and in dairy barns (Kullman et al., 1998).

As part of an international collaborative study on the impact of Traffic-Related Air Pollution on Childhood Asthma (TRAPCA), we sampled fine and inhalable PM with a 50% aerodynamic cut-off diameter of 2.5 μm (PM_2.5) at 40 monitoring sites and of 10 μm (PM₁₀) at 12 monitoring sites in ambient air across Munich, Germany for 17 months (Gehring et al., 2002; Hoek et al., 2002). In the current study, we used these sampled particles and determined endotoxin concentrations in PM_2.5 and PM₁₀.

We report endotoxin levels in PM_2.5 and PM₁₀ per milligram and per cubic meter of sampled air and to describe spatial variation of endotoxin across Munich, a German metropolitan area.

The goals of our study were to determine ambient endotoxin levels in PM_2.5 and PM₁₀ in the metropolitan area of Munich and describe the spatial varation of endotoxin in PM_2.5 across the city. We also describe the correlation of endotoxin in PM_2.5 and PM₁₀ measured at the same sites and investigate whether potential sources of endotoxin are useful in predicting endotoxin ambient levels.