Indoor Endotoxin Exposure and Ambient Air Pollutants Interact on Asthma Outcomes

It is now quite well established that ambient (outdoor) air pollution is a health risk and contributes substantially to the global burden of disease. The mortality and morbidity burden resulting from ambient air pollution has increased during the last 2 decades (1). Human populations breathe complex mixtures of air pollutants, allergens, and irritants. It has been known for some years from small controlled clinical trials that ambient air pollutants such as ozone (O₃) (2), nitrogen dioxide (NO₂) (3), and combinations of these gaseous pollutants (4) interact with allergens on physiological outcomes in people with allergic asthma. Much more has been learned in recent years about the effects of fine particles (particulate matter <2.5 μm in aerodynamic diameter [PM_2.5]) on asthma and other chronic diseases (5). A major constituent of PM_2.5 in urban environments is diesel exhaust particles, which have been shown to bind the major grass pollen allergen (6).

However, in Western societies, most people spend the vast majority of their time indoors, where the mix of air pollutants, allergens, and irritants is different. The predominant indoor allergens include those derived from house dust mites, pets, cockroaches, and molds. Endotoxin is a constituent of gram-negative bacteria, often found in house dust and some occupational settings, which increases the risks for asthma and chronic bronchitis, but perhaps paradoxically also appears to confer some protection against the development of asthma in children (7, 8).

In this issue of the Journal, Mendy and colleagues (pp. 712–720) report a large cross-sectional analysis of data from the National Health and Nutrition Examination Survey (9). This paper presents some novel findings of synergistic interactions between house dust endotoxin and ambient air pollutants on asthma in both children and adults. The participants in dust sampling were reasonably representative of the US population, although the Midwest region was slightly oversampled. A well-standardized, highly sensitive endotoxin assay was used.

A two-stage model strategy was used to estimate ambient air pollutants (PM_2.5, NO₂, and O₃). At the first stage, meteorological data and chemical reaction kinetics were used to estimate the annual average concentrations of air pollutants at grid level (36 × 36 km) by the Community Multiscale Air Quality model. At the second stage, a Downscaler model and monitored air pollution data were used to downscale the estimated air pollution concentrations into a finer grid level (12 × 12 km). This method has been validated and is useful for health risk assessment. However, the introduction of remote sensing data, land use information, and road information as predictors, as well as application of a machine learning method, could further improve exposure assessment for future studies (10). The measurement error would be smaller, leading to unbiased effect estimates (11).

Well-validated asthma outcomes from the National Health and Nutrition Examination Survey were analyzed by Mendy and colleagues (9). A wide range of serum allergen-specific IgE was measured in vitro, so the classification of atopy was as good as possible. The statistical analysis allowed for the relevant potential confounders including age, sex, ethnicity, socioeconomic status, household smoking, body mass index, and urbanization, which were fitted to multivariate logistic models.

Importantly, the geometric mean concentrations of PM_2.5, O₃, and NO₂ were not particularly high; all were well below the U.S. Environmental Protection Agency National Ambient Air Quality standards for annual averages. More interactions between endotoxin and air pollutants were found on emergency room (ER) visits for asthma in the last 12 months than would have been expected by chance. Specifically, higher endotoxin and PM_2.5 concentrations were associated with ER visits in all participants, and higher endotoxin and NO₂ concentrations were associated with ER visits in children. Interactions were mostly confined to those participants sensitized to aeroallergens.

These important findings are both biologically and epidemiologically plausible. We agree with the authors that further prospective cohort studies are needed to confirm whether similar interactions can be demonstrated in patients with other conditions such as chronic obstructive pulmonary disease, and we would add bronchiectasis, cystic fibrosis, and idiopathic pulmonary fibrosis. New strategies are needed to reduce ambient PM_2.5 and NO₂ concentrations, such as reductions in primary emissions and ammonia (12), through incentives for greater use of electric vehicles and increasing pollution taxes (13). Indoor endotoxin concentrations could be reduced by measures such as household hygiene and cleaning, avoidance of biomass fuels (14), choice of home flooring, and keeping pets such as dogs outdoors (15). There may also be some role for home air purification units, including high efficiency particulate air filters to reduce airborne concentrations (16).

At a time when some politicians appear determined to ignore the evidence and water down long-standing protections to public health, demonstration of new interactions between indoor and outdoor air pollutants serve as an important reminder of how much still remains to be understood. The confirmation of health effects at concentrations below current standards emphasizes the importance of adequately protecting our patients with chronic lung diseases. If anything, after considering other recent studies, ambient air quality standards need to be further strengthened.