ENDOTOXIN LEVELS AND THEIR ASSOCIATION WITH RESPIRATORY OUTCOME IN TWO SASKATCHEWAN COMMUNITIES

There are a number of studies that investigate the association between endotoxin and respiratory outcomes, but few examine ambient endotoxin particularly in smaller industrial cities in Canada. In epidemiological studies of respiratory diseases, environmental factors are of major concern to health. The aims of this study are: (1) to assess ambient endotoxin levels in two communities in southern Saskatchewan (SK), Canada; and (2) to examine the association between ambient endotoxin levels and forced expiratory volume in one second (FEV1) among older adult residents of the two communities. This work is part of the larger Air Quality and Lung Health Study conducted in the cities of Estevan and Swift Current, SK. Ambient environmental and personal respiratory health measures were collected in three periods (phases) of three month durations. To assess ambient endotoxin levels in the two communities, samples were collected in spring 2013 (Phase 1), fall 2013 (Phase 2), and spring 2014 (Phase 3). The phase sampling strategy was employed to account for potential seasonal variations in endotoxin. Consecutive full-week (7 day accumulation) particle mass size fractions (PM2.5 and PM10) were gravimetrically collected for each of the communities for the 3-month duration of each phase (1-3) and analyzed for endotoxin load (EU/µg) and endotoxin concentration (EU/m3). Geometric means were calculated for endotoxin load and concentration for each particle size fraction (PM2.5 and PM10), each community, and each phase. Differences were tested between particle size fractions, and between communities. The highest levels of endotoxin (EU/µg) found in Estevan were 0.02 (Phase 1), 0.03 (Phase 2) and 0.01(Phase 3), while endotoxin (EU/m3) were 0.04 (Phase 1), 0.07 (Phase 2) and 0.02 (Phase 3). Similar trends were found in Swift Current with 0.04 (Phase 1), 0.10 (Phase 2), and 0.05 (Phase 3) in endotoxin (EU/µg), while 0.07 (Phase 1), 0.13 (Phase 2), and 0.08 (Phase 3) were found in endotoxin (EU/m3). Estevan had no significant differences in endotoxin load (EU/µg) between size fractions for any of the phases. However, endotoxin load was significantly higher in the PM2.5 size fraction as compared to the PM10 for both Phases 2 and 3 for Swift Current. For both communities, in all phases, there was significantly greater endotoxin concentration (EU/m3) in PM2.5 as compared to PM10. Comparing communities, Swift Current had significantly greater endotoxin load (EU/µg) in the PM2.5 size fraction (in all phases) and the PM10 size fractions in Phases 2 and 3 as compared to Estevan. Similar trends were observed for endotoxin concentration (EU/m3) in PM2.5 where mean concentrations were greater in Swift Current as compared to Estevan for all phases, but the difference between communities was significant only in Phase 3 (p<0.0001). For PM10 only in Phase 3 were endotoxin concentrations significantly different between communities with Swift Current having higher mean levels (p<0.02). Endotoxin load and concentration were evaluated against pulmonary function measures (FEV1) in older adults in the two communities to test the association between ambient endotoxin levels and forced expiratory volume in one second (FEV1) among older adult residents of the two communities. Twice daily (morning and evening) FEV1 were collected from older adults (>50 years) from both communities over the duration of each of the three phases. Weekly (7 days) morning and evening mean, minimum and maximum FEV1 were calculated to match endotoxin collection/analysis periods. Despite the relatively low levels of endotoxin in these communities, personal respiratory health measures revealed that there were significant associations in Swift Current Phase 3 measures between mean and max FEV1 and PM2.5 endotoxin load and concentration, and these were the highest levels that were measured for either community or any phase. Taken together these results indicate that endotoxin levels in ambient measures of particulate vary by size fraction with higher concentrations attributed to the smaller size fraction (PM2.5). Endotoxin in ambient air particulate has the potential to influence respiratory outcomes. Even at what is considered low levels a continuous exposure to endotoxin can be associated with respiratory health effects as was seen in Swift Current during Phase 3