Wildfire smoke impacts respiratory health more than fine particles from other sources: observational evidence from Southern California

Fine particulate matter (PM2.5) is a major component of wildfire smoke and significantly impacts public health. While PM2.5 levels in the US have generally decreased due to regulations, wildfire-prone areas show an exception, and climate change is projected to worsen this issue. Current air quality standards do not differentiate between PM2.5 sources, assuming equal harm from all sources. However, recent toxicological studies suggest that wildfire PM2.5 may be more toxic than equivalent doses of ambient PM2.5. This study aims to quantify the differential impacts on respiratory health of wildfire-specific PM2.5 compared to PM2.5 from other sources in Southern California (SoCal). SoCal's unique wildfire season, driven by Santa Ana winds (SAWs), presents an ideal setting for this investigation. The study will utilize multiple statistical approaches to isolate wildfire-specific PM2.5 and analyze its effect on respiratory hospital admissions over a 14-year period (1999-2012), focusing on the Santa Ana wind season (September to May). Understanding this difference is crucial for developing effective air quality policies that protect public health.

Existing research demonstrates the detrimental effects of PM2.5 from wildfire smoke on respiratory and cardiovascular health. However, studies often lack the spatial and temporal resolution to isolate the impact of wildfire-specific PM2.5 from other PM2.5 sources. Previous studies have been limited by relying on computationally intensive models or focusing on single wildfire events. While some studies suggest that wildfire PM2.5 is more toxic than other sources due to its chemical composition and increased oxidative potential, a comprehensive analysis comparing the health impacts of wildfire-specific PM2.5 against non-wildfire PM2.5 at a fine spatial scale over an extended period is lacking. This study addresses this gap by applying multiple statistical approaches to differentiate and quantify the health impacts of wildfire-specific PM2.5.

This study utilized daily hospital admission data for respiratory diseases (from the California Office of Statewide Health Planning and Development) and daily zip code-specific PM2.5 concentrations (from the US EPA Air Quality System) for Southern California (1999-2012, excluding summer months). Four analytical approaches were used to isolate wildfire-specific PM2.5: (i) an instrumental variable approach with two-stage regression using Santa Ana wind strength and upwind fire presence as instruments; (ii) a spatio-temporal multiple imputation approach; (iii) an interaction effect approach; and (iv) a seasonal interpolation method. Two distinct exposure definitions were employed: (i) the occurrence of strong SAWs and the presence of fire upwind; and (ii) smoke plume datasets (NOAA Hazard Mapping System) within a 160 km buffer. The study area encompassed 696 zip code polygons within the Santa Ana wind domain. Regressions included controls for flu admissions, weather covariates, day-of-week, month-of-year, zip code fixed effects, and a time trend. The key analysis involved regressing respiratory admission rates on the estimated wildfire-specific PM2.5 concentrations, allowing for comparison across different methods and exposure definitions.

The study found that a 10 µg/m³ increase in wildfire-specific PM2.5 was associated with a 10% increase in respiratory hospital admissions (95% CI: 3.5–16.5), using the spatio-temporal imputation approach and the 'fire upwind and strong SAW' exposure definition. This represents a substantially higher impact than that observed with non-wildfire PM2.5 (0.67% to 1.3% increase). Results were consistent across multiple analytical approaches, although the magnitude of the effect varied depending on the method and exposure definition used. Overall, the findings strongly indicate that wildfire-specific PM2.5 is up to ten times more harmful to respiratory health than PM2.5 from other sources. Even when considering comparable exposure levels, the study demonstrated the significantly higher risk associated with wildfire PM2.5. A case study focusing on the impactful October 2007 wildfires further supported these findings. The maps and figures display the spatial distribution of PM2.5 and respiratory admissions, highlighting the impact of wildfire PM2.5 on highly populated coastal areas.

The findings clearly demonstrate that wildfire-specific PM2.5 poses a substantially greater risk to respiratory health than PM2.5 from other sources. This contradicts the current assumption in air quality regulations that PM2.5 from all sources has equal toxicity. The significantly higher impact of wildfire PM2.5, even at similar concentration levels, suggests inherent differences in chemical composition and toxicity. This underscores the need for a shift in air quality policies to incorporate source-specific toxicity assessments for PM2.5, particularly considering the increasing frequency and intensity of wildfires. The study's findings are particularly relevant for Southern California and other regions prone to wind-driven wildfires. The methodology provides a robust framework for future research on the health impacts of wildfire smoke in other geographical settings.

This study provides strong evidence that wildfire-specific PM2.5 exerts a significantly greater impact on respiratory health compared to PM2.5 from other sources. The consistent findings across various analytical methods highlight the importance of source-specific considerations in air quality management. Future research should investigate the chemical composition of wildfire PM2.5 from different ecosystems and combustion temperatures to better understand the mechanisms behind its enhanced toxicity. Moreover, exploring the effectiveness of different strategies to mitigate the public health impacts of wildfire smoke is crucial.

Several limitations should be considered when interpreting the results. Using patient home addresses to estimate exposure may introduce inaccuracies, and community-level PM2.5 measurements might not fully capture individual exposures. The reliance on visible satellite data for smoke plume identification may underestimate actual exposure due to limitations in detecting ground-level smoke and potential systematic biases. The study's definition of upwind fire exposure could also lead to misclassification of some PM2.5 as smoke or non-smoke. Finally, the absence of lagged effects in the models and the lack of ozone adjustment are acknowledged as potential limitations.