Contributions of particulate and gas phases of simulated burn pit smoke exposures to impairment of respiratory function

Abstract

Objective:

Inhalation of smoke from burning of waste materials on military bases is associated with increased incidences of cardiopulmonary diseases. This study examined the respiratory and inflammatory effects of acute inhalation exposures in mice to smoke generated by military burn pit-related materials including plywood (PW), cardboard (CB), mixed plastics (PL), and a mixture of these three materials (MX) under smoldering (0.84 MCE) and flaming (0.97 MCE) burn conditions.

Methods:

Mice were exposed nose-only for one hour on two consecutive days to whole or filtered smoke or clean air alone. Smoldering combustion emissions had greater concentrations of PM (~40 mg/m³) and VOCs (~5-12 ppmv) than flaming emissions (~4 mg/m³ and ~1-2 ppmv, respectively); filtered emissions had equivalent levels of VOCs with negligible PM. Breathing parameters were assessed during exposure by head-out plethysmography.

Results:

All four smoldering burn pit emission types reduced breathing frequency (F) and minute volumes (MV) compared with baseline exposures to clean air, and HEPA filtration significantly reduced the effects of all smoldering materials except CB. Flaming emissions had significantly less suppression of F and MV compared with smoldering conditions. No acute effects on lung inflammatory cells, cytokines, lung injury markers, or hematology parameters were noted in smoke-exposed mice compared with air controls, likely due to reduced respiration and upper respiratory scrubbing to reduce the total deposited PM dose in this short-term exposure.

Conclusion:

Our data suggest that material and combustion type influences respiratory responses to burn pit combustion emissions. Furthermore, PM filtration provides significant protective effects only for certain material types.

Introduction

Emissions from the combustion of natural and anthropogenic sources may induce cancer and promote heart and lung diseases. In particular, fine particulate matter (PM_2.5) produced from biomass burning in wildfires has been associated with increases in respiratory and cardiovascular morbidity and all-cause mortality (Adetona et al., 2016; Cascio, 2018; Chen et al., 2021; Karanasiou et al., 2021). Respiratory morbidities such as asthma, chronic obstructive pulmonary disease (COPD) and rhinosinusitis are exacerbated by PM_2.5. (Heinrich et al., 2018; Leland et al., 2021; Lu et al., 2021; Orellano et al., 2017). Studies show that exposure to wildfire smoke increases hospitalizations and emergency visits for respiratory conditions such as COPD and asthma (Karanasiou et al., 2021). Wildfire smoke was associated with a 10.5% increase in respiratory morbidities across multiple studies, and overall asthma risk increased by 9.2% per 10 μg/m³ increase in wildfire-related PM_2.5 (Karanasiou et al., 2021). While some inconsistencies exist in the literature for cardiovascular incidents caused by wildfire exposure, in general for PM_2.5 each 10 μg/m³ increase was associated with a 4.5% increase in cardiovascular mortality (Cascio, 2018; Chen et al., 2021; Karanasiou et al., 2021). Although these associations have been made for wildfire smoke and biomass burning, less is known about the environmental and human health effects of burning man-made materials.

Military bases may produce large quantities of solid waste, with few options for environmentally safe waste management practices, necessitating the use of other disposal methods such as burn pits. Burn piles or pits may contain materials such as plastic, metals, rubber, solvents, munitions, and wood, often ignited with diesel or jet fuel accelerants (Baird, 2012; Mallon et al., 2016). Burn pits were commonly used in Iraq, Saudi Arabia, and Afghanistan, with more than 200 active burn pits from 2002 to 2009 (Mallon et al., 2016). The U.S. Department of Defense estimated that 200 tons of waste was burned per day at Balad Air Force Base, Iraq, during the peak years of 2005 and 2007 (Bith-Melander et al., 2021). Fine PM (PM_2.5) is of particular interest due to high levels from open burning (average >45 μg/m³, maximum of 150 μg/m³ (Li et al., 2021), the predominance of the inhalable size fraction, and the ability to carry metals and other aerosols with toxic potential including silica (fine desert sand), polycyclic aromatic hydrocarbons (PAHs), and dioxins (Berman et al., 2021; Engelbrecht et al., 2009; Li et al., 2021; Xia et al., 2016).

Several health studies of soldiers exposed to burn pit emissions found increased rates of cardiopulmonary illness, including asthma and obstructive lung disease (Falvo et al., 2015; Liu et al., 2016; Pugh et al., 2016; Rivera et al., 2018; Szema et al., 2010; Wauters et al., 2019). The U.S. Department of Defense concluded that 13% of all U.S. Army Medic visits in Iraq were due to new-onset respiratory-related ailments, including dyspnea and asthma (Szema et al., 2010). Findings from the Millennium Cohort study showed that combat deployment during post-2001 military conflicts was associated with a 24-30% higher risk of new-onset asthma (Rivera et al., 2018). Veterans deployed to Iraq and Afghanistan had increased rates of COPD and asthma, even after controlling for demographics and tobacco use (Pugh et al., 2016). Veterans exposed to burn pit smoke during deployment have also shown higher rates of allergic rhinitis, chronic rhinosinusitis, and nasal polyposis (Hill et al., 2022). Studies of veterans presenting with new onset cough and dyspnea suggest deployment was associated with respiratory disease (Garshick et al., 2019; King et al., 2011). An analysis of Veteran Affairs (VA) healthcare provided to veterans deployed from 2002-2014 found that among the newly diagnosed COPD cases only 31% presented with concordant spirometry (Schneiderman et al., 2017). Similarly, a recent study of veterans presenting with dyspnea with no abnormal spirometry found an increase in ventilatory dysfunction and slow compartment filling as determined by magnetic resonance imagining (Mammarappallil et al., 2020). These studies highlight that respiratory effects related to burn pit exposures are still not well characterized, may be non-traditional in their symptomology, and require further research.

We previously investigated the respiratory effects of isolated emission condensates from burn pit-related materials in an animal aspiration model (Kim et al., 2021). Adult female CD1 mice were exposed by oropharyngeal aspiration to emission condensates from smoke generated from common burn pit materials, including plywood (PW), cardboard (CB), plastics (PL), or a mixture of all three (MX) burned under low temperature smoldering or high temperature flaming conditions (Kim et al., 2021). Mice exposed to smoldering PW and CB PM had reduced breathing frequency (F) at 4 hours post-exposure, while flaming MX PM reduced F at 24 hours post-exposure. Flaming PL PM promoted neutrophilic lung inflammation at both time points. This study used standard doses of PM (100 μg per mouse) for all exposures but did not test the complete emission mixture including any gaseous components. Therefore, in this study we assessed the direct real-time effects of exposures to simulated burn pit smoke emissions in mice. Emissions from PW, CB, PL, or MX were generated under smoldering or flaming burn conditions and delivered to mice whole or filtered to remove PM and assess effects of the gaseous components, including volatile organic compounds (VOCs). Our results show that all smoldering burn pit material emissions caused significant impairment of respiratory function, to a greater degree than the corresponding flaming emissions. Particle filtration of smoldering emissions significantly improved respiratory function, except in the case of CB, indicating that PM filtration provides significant protective effects only for certain material types.

METHODS

Burn Materials

Burn materials used in this study were identified from a U.S. Army waste management study (Aurell et al., 2019) and consisted of military grade plywood (PW) and cardboard (CB) sourced from ActionPak, Inc., Bristol, PA as well as a mix of plastics (PL) comprised of low-density polyethylene, high-density polyethylene, polyethylene terephthalate, and polystyrene pellets (Edwards Industrial Surplus, Robards, KY). A burn materials mixture (MX) containing 52% cardboard, 27% plastic and 21% plywood was also tested; this composition was previously assessed (Kim et al., 2021) and determined to be representative of military burn pit waste streams (Aurell et al., 2019).

Automated Combustion System

A quartz tube furnace system was used to generate smoke emissions under constant conditions for 60 minutes, as previously described (Kim et al., 2021). Approximately 15 g of each type of material was placed inside the quartz tube and burned under controlled combustion conditions for 60 min. Using an automated mass flow controller, smoke was generated under smoldering (~510 °C) or flaming (~640 °C) conditions with a modified combustion efficiency (MCE) of 0.82 for smoldering condition and 0.97 for flaming conditions and PM levels of ~40 mg/m³ for smoldering and ~4 mg/m³ for flaming conditions. Exposure atmosphere CO₂, CO, nitrogen oxides (NO and NO₂), VOCs, PM concentrations, particle size, temperature, relative humidity, static pressure, and flow rate were analyzed as previously described (Hargrove et al., 2019; Kim et al., 2019). Smoke exposures were diluted with medical grade air and transported to a 64-port nose-only inhalation chamber (Laboratory Products Inc., Seaford DE) at a flow of 1 L/min at each port. All exposures were repeated with the addition of a high-efficiency particulate air (HEPA) filter placed just before the nose only exposure tower to remove PM and test the effects of gases alone. Air samples were collected from the exposure chamber at one of the nose-only ports using stainless steel evacuated canisters to measure a range of speciated VOCs using 2,4-dinitrophenylhydrazine (DNPH)-coated silica cartridges to quantify carbonyl compounds. After sampling was completed, the canister samples were analyzed by gas chromatography mass spectrometry in accordance with EPA method TO-15 as previously described (George et al., 2014). The DNPH cartridges were extracted with 6 mL of carbonyl-free acetonitrile after sampling and analyzed by high-performance liquid chromatography following EPA method TO-11A. The VOCs targeted in these analyses represent many important air toxics but do not represent a total speciation of all gas-phase organic compounds produced in burn pit smoke. Particle size distributions in the range of 10 nm to 10 μm were monitored using a scanning mobility particle sizer (NanoScan SMPS, Model 3910; TSI Inc., Shoreview, MN) and an optical particle sizer (OPS, Model 3330; TSI Inc.). Since these two methods provide differing measures of particle sizes for the same emission type, both are presented for clarity.

Smoke Exposure and Real Time Respiratory Monitoring

All animal use protocols were approved by the local EPA Institutional Animal Care and Use Committee. Female Balb/cJ mice aged 6-8 weeks (Jackson Laboratories, Bar Harbor, ME) were weight randomized and housed in groups of 4 with food (Purina 5001) and water ad libitum. All mice were adapted to nose-only restraint tubes over increasing time in the two days prior to smoke exposures. Mice contained in slotted restraint tubes paired with thoracic head-out plethysmography (HOP) chambers (Emka Technologies, Falls Church, VA) were attached directly to the exposure tower, while control groups were concurrently exposed to filtered air in nose-only tubes in a separate chamber, as previously described (Hargrove et al., 2019). Exposures began with 20 minutes of clean air exposure to collect baseline respiratory data followed by 1 hour of smoke exposure and a 10-minute clean air post-exposure to monitor recovery or return to respiratory baselines. The control mice exposed to air alone were only assessed for blood, lung lavage, and lung histopathology endpoints; in addition to these data, smoke-exposed mice were also evaluated for respiratory responses to the various exposure atmospheres in comparison to their individual pre-exposure clean air control baseline values. All animals were exposed on two consecutive days to smoke or filtered air and monitored continuously. Respiratory responses were assessed using Emka IOX 2.10.0.40 software (Emka Technologies) including tidal volume (TV), frequency (F), peak inspiratory and expiratory flow rates (PIF, PEF), inspiratory and expiratory times (Ti, Te), minute volume (MV), and relaxation time (RT). Pre- and post-exposure averages and 10-minute averages during exposure were computed for statistical comparisons. Since no meaningful differences were observed for these 10-minute periods over each hour of exposure on both days, data were combined and averaged for the two total hours of exposure for each mouse for statistical comparisons. Flaming PL and MX emissions contained high levels of black soot which penetrated through the nose-only ports, passing through to the thoracic plethysmograph chamber, and notably interfering with respiratory measures by clogging the mesh screen pneumotachographs within 15 to 20 minutes of exposure. Consequently, although mice were exposed to one full hour of flaming PL or MX each day, only the first 15 minutes of respiratory measures were utilized in computing average responses.

Blood and BALF Collection and Analysis

Four hours after the final exposure, mice were euthanized with a lethal dose of phenytoin/pentobarbital (>25 and 200 mg/kg, respectively; Virbac AH Inc., Fort Worth, TX) and blood was drawn via cardiac puncture for complete blood cell analysis. The left lung was clamped closed, and right lobes were lavaged with three aliquots of Hanks balanced salt solution (0.6 mL each). Bronchoalveolar lavage fluid (BALF) was analyzed as previously described (McGee et al. 2015) for total cell numbers and differentials. Clinical chemistry analyses of BALF supernatants were conducted using a Konelab 30 analyzer (Thermo Clinical Lab Systems, Espoo, Finland), including lactate dehydrogenase (LDH) and protein (reagents from Thermo Fisher diagnostics, Rockford, IL), N-acetyl-β-D-glucosaminidase (NAG, Roche Diagnostics, Indianapolis, IN), and γ-glutamyl transferase (GGT, Thermo Fisher Scientific, Pittsburgh, PA). Inflammation-related cytokines including macrophage inflammatory protein-2 (MIP-2), tumor necrosis factor-alpha (TNF-α), interferon-gamma (IFN-γ) and interleukin-6 (IL-6) were analyzed using kits from MilliporeSigma (Burlington, MA) according to the manufacturer’s instructions.

Lung Histopathology

Left lungs were fixed at 22 cm H₂O of pressure using 10% buffered formalin, and paraffin-embedded 5 μm longitudinal sections were stained with hematoxylin and Eosin (H&E). These sections were analyzed via light microscopy by a board certified pathologist (Experimental Pathology Laboratories, Research Triangle Park, NC) for inflammatory, degenerative, metaplastic, and proliferative changes based on previously defined criteria (Renne et al., 2009). Findings were graded on a semi-quantitative scale based on severity.

Statistical Analysis

GraphPad Prism 6.07 (San Diego, CA) software was used to perform statistical analyses. Values were expressed as means and standard error of the mean (SEM). Statistical analyses were conducted using one-way analysis of variance (ANOVA) among smoldering or flaming conditions, followed by Tukey or Sidak post-hoc test. Comparisons of burn condition (smoldering or flaming) and effects of filtration on endpoints from the same combustion emission type were performed using a two-tailed t-test. A P value < 0.05 was considered statistically significant.

RESULTS

Smoke Composition

Previous analyses of biomass combustion determined that modified combustion efficiency (MCE; calculated as % MCE = (ΔCO₂ / (ΔCO₂ + ΔCO)) × 100), was >95% for flaming and 65-85% for smoldering combustion conditions (Urbanski, 2014). MCE calculations showed each of the burn pit combustion atmospheres met these criteria for flaming and smoldering conditions (Table 1). PM levels produced by smoldering and flaming combustion were ~40 mg/cm³ and ~4 mg/cm³, respectively. Particle mass median aerodynamic diameter (MMAD) measured by Nanoscan had a range of 200 – 280 nm in 6 of the 8 emission atmospheres, while flaming PL and MX had smaller sizes (129 and 144 nm, respectively). In contrast, MMADs of flaming PL and MX measured by OPS were much larger (4.10 and 2.52 μm, respectively), greater than smoldering PL (2.27 μm), smoldering MX (0.85 μm), and PW and CB under either combustion condition (range of 0.33 to 0.55 μm; Table 1). These data indicate a significantly greater level of particle agglomeration during combustion of plastics, which was more extensive under flaming conditions, generating notable levels of soot. PM levels in filtered smoldering and flaming emissions were ≤ 0.2 mg/m³, equivalent to levels found in control air exposures.

Table 1:

Characteristics of whole burn pit material smoke emissions from plywood (PW), cardboard (CB), plastic (PL) and a mixture (MX) of PW, CB, and PL. Data represent mean and SEM of real-time values throughout exposures (CO, CO₂), or time-integrated concentrations using gravimetric analysis (PM) or mass spectrometry for captured volatiles via vacuum canisters/DNPH-coated silica cartridges (VOCs) in accordance with EPA methods TO-15/TO-11A. Particle size data (PM nano, PM micro) represent mass median aerodynamic diameter and geometric standard deviation.

	Smoldering				Flaming
	PW	CB	PL	MX	PW	CB	PL	MX
MCE a	0.82	0.82	0.85	0.85	0.96	0.97	0.99	0.98
	0.84	0.79	0.85	0.84	0.96	0.97	0.99	0.98
CO (ppmv)	51.6 [1.0]	70.5 [2.8]	35.7 [0.6]	55.9 [1.4]	116.9 [4.0]	103.4 [4.9]	19.9 [0.7]	38.6 [2.0]
	55.2 [0.8]	75.7 [0.7]	39.2 [0.5]	55.7 [0.6]	95.6 [3.7]	113.6 [4.5]	15.6 [0.5]	26.8 [1.9]
CO2 (ppmv)	269.9 [7.3]	318.2 [9.4]	240.6 [6.3]	347.9 [8.8]	2787.0[59.7]	4010.1[64.9]	1665.0[22.5]	1871.3[37.9]
	302.7 [6.4]	310.1 [6.8]	246.4 [5.4]	308.1 [6.0]	2739.3[67.8]	3423.0[54.7]	1682.7[27.5]	2161.9[68.6]
VOCs (ppmv)	5.127	11.438	7.317	5.515	1.810	2.393	1.129	1.171
	4.738	13.673	5.065	4.043	1.731	2.437	0.863	1.032
PM (mg/m3)	41.6 [0.6]	39.0 [1.1]	40.8 [0.5]	41.5 [0.6]	4.1 [0.1]	3.6 [0.1]	4.7 [0.1]	4.0 [0.1]
	0.08 [0.01]	0.12 [0.01]	0.19 [0.02]	0.23 [0.01]	0.09 [0.03]	0.15 [0.02]	0.23 [0.01]	0.24 [0.01]
PM nano b	244.2 [1.3]	201.0 [1.30]	249.4 [1.33]	213.7 [1.31]	279.3 [1.43]	208.6 [1.42]	129.4 [1.30]	143.7 [1.20]
PM micro c	0.55 [1.48]	0.51 [1.49]	2.27 [1.44]	0.85 [1.58]	0.43 [1.36]	0.33 [1.14]	4.10 [2.22]	2.52 [1.83]

Italicized rows are filtered smoke exposure values

^aModified combustion efficiency (MCE) = ΔCO₂/(ΔCO₂ + ΔCO)

^bPM size (nm) measured with NanoScan SMPS Model 3910.

^cPM size (μm) measured with optical particle sizer (OPS) Model 3330.

Smoldering combustion conditions produced far greater concentrations of VOCs than did flaming conditions, with smoldering CB having the highest concentration of measured VOCs for any exposure group (~11.4 ppmv total whole smoke and ~13.7 ppmv filtered smoke, Table 1). CB also produced the highest levels of the measured air toxics (Figure 1, Supplementary Table S1), including formaldehyde, acetaldehyde, and acrolein (~3.7, 3.6, and 0.8 ppmv respectively). Filtration only slightly altered the MCE, CO, CO₂, and VOC levels for all groups (Table 1, Supplementary Table S1). Total VOC levels in the filtered smoldering CB atmosphere were approximately 20% higher than in the whole CB atmosphere.

Figure 1:

The top 5 most prevalent species of VOCs measured for each whole smoke exposure condition of burn pit materials under smoldering and flaming conditions. Values represent average concentrations measured under each burn condition in accordance with EPA methods TO-15 and TO-11A.

Respiratory Effects

Breathing frequency (F) rapidly declined in the first 10 minutes of exposure before leveling out at a depressed respiratory rate with no evidence of desensitization in all exposure groups (Figure 2). All mice partially recovered in the 10 minutes of post-exposure to clean air, and there were no statistical differences in the pre-exposure values from day 1 to day 2 (Figure 2). With no major trends or statistical difference across the two days of exposure, the two hours of respiratory data were averaged into single values for each animal. All four smoldering burn pit emission types reduced F an average of 145 breaths per minute (bpm) compared with baseline exposures to clean air (Figure 3). Smoldering PL emissions produced the greatest decrement in F (−167.5 bpm), compared with PW, CB, and MX (−151.1, −126.4 and −136.8 bpm, respectively). HEPA filtration significantly abrogated the effects of smoldering PW, PL, and MX on F by 51%, 26%, and 22%, respectively, indicating that PM was partially responsible for altering respiratory function. However, filtering the CB emission only improved the change in F by a non-significant 7%, suggesting most of the effect was caused by gas phase products in this atmosphere. Mice exposed to flaming emissions (containing ~10x lower PM concentrations and lower VOCs) had significantly less suppression of F compared with corresponding smoldering conditions. Flaming PW and CB produced significantly greater decrements in F (28.7% and 27.2% decrease respectively) than flaming PL (10%) or MX (13%). HEPA filtration significantly reduced the effects of flaming PW, CB, and MX, but not PL on F.

Figure 2:

Mice (n=8/group) were exposed nose-only to smoldering emissions of plywood (PW), cardboard (CB), plastic (PL) or a mixture of all three (MX) for 1 hour a day on 2 consecutive days. Results show average frequency in 10-minute intervals for 20 minutes of clean air pre-exposure and 60 minutes of smoke followed by 10 minutes clean air post-exposure.

Figure 3:

Mice (n=8/group) were exposed nose-only to smoldering or flaming emissions of plywood (PW), cardboard (CB), plastic (PL) or a mixture of all three (MX) for 1 hour a day on 2 consecutive days. Results show changes in average breathing parameters compared with baseline values (clean air control prior to exposure) and represent means and SEM for 8 mice per group. Hatched bars represent whole smoke exposure, and clear bars represent HEPA-filtered smoke. *P < 0.05 vs. whole smoke of same fuel and temp, †P < 0.05 vs CB, ‡P < 0.05 vs PL, §P < 0.05 vs MX. #P < 0.05 vs. corresponding smoldering whole smoke combustion condition.

Smoldering and flaming PW reduced MV to equivalent levels (-17.4 mL and −15.3 mL, respectively; Figure 3), and both were significantly improved by HEPA filtration (−6 and −3.6 mL respectively). Smoldering PL and MX emissions also reduced MV (−16.3 and −15.1 mL), and both reductions were improved significantly by HEPA filtration, however flaming PL and MX emissions had no effects on MV. As with F and several other breathing parameters, the reductions in MV by CB emissions were not ameliorated by filtration for smoldering or flaming conditions. Other breathing parameters related to timing (Te, Ti, RT), airflows (PIF, PEF), and volumes (TV) had changes which corresponded to the effects seen with F and MV (Table 2), though generally with less sensitivity.

Table 2:

Respiratory parameters for mice (n=8/group) were exposed to flaming and smoldering burn pit materials plywood (PW), cardboard (CB), plastic (PL) and a mixture (MX) for 1 hour on 2 consecutive days, presented as change from baseline average ± SEM. *P < 0.05 compared to all other fuel types, #P < 0.05 compared to matched fuel smoldering conditions. ^{CB, PL, MX}P < 0.05 when compared to that fuel type. Bolded text: P < 0.05 when compared to matched whole smoke condition.

Smoldering		Ti (Δ msec)	Te (Δ msec)	RT (Δ msec)	PIF (Δ mL/s)	PEF (Δ mL/s)	TV (Δ mL)
PW	whole	43.3 ± 7.6	291.1 ± 26.1	271.6 ± 23.2	−0.71 ± 0.06CB	−0.59 ± 0.07*	−0.02±0.01CB,PL
	filtered	12.7 ± 2.6	116.4 ± 9	104.7 ± 9.1	−0.04 ± 0.08	−0.07 ± 0.09	0.02 ± 0.01
CB	whole	48.9 ± 5.8	240.7 ± 19.9PL	211.4 ± 19.6	−0.37 ± 0.05	−0.26 ± 0.03	0.02 ± 0.01
	filtered	31.0 ± 2.9	205.2 ± 7.5	180.6 ± 4.7	−0.29 ± 0.04	−0.24 ± 0.03	0.02 ± 0.01
PL	whole	58.6 ± 6.5	358.2 ± 26.2	248.1 ± 27.8	−0.46 ± 0.08	−0.37 ± 0.07	0.04 ± 0.01MX
	filtered	38.2 ± 5.9	253.2 ± 10.3	225.8 ± 9.0	−0.07 ± 0.07	−0.07 ± 0.1	0.04 ± 0.01
MX	whole	50.4 ± 7.5	290.9 ± 16.4	278.1 ± 16.6	−0.54 ± 0.1	−0.27 ± 0.06	0.01 ± 0.01
	filtered	27.6 ± 3.4	157.9 ± 14.3	146.0 ± 13.4	−0.14 ± 0.1	−0.04 ± 0.07	0.03 ± 0.01
Flaming		Ti (Δ msec)	Te (Δ msec)	RT (Δ msec)	PIF (Δ mL/s)	PEF (Δ mL/s)	TV (Δ mL)
PW	whole	11. 5 ±2.1#MX	99.2±10.6#PL,MX	97.9 ± 10#PL,MX	−0.78 ± 0.09*	−0.6 ± 0.1*	−0.05 ± 0.01*
	filtered	7.3 ± 1.3	40.3 ± 2.7#	32.6 ± 2#	−0.14 ± 0.05	−0.18 ± 0.05	0.01 ± 0.01
CB	whole	14.4 ± 2.2#MX	108.3±5.8#PL,MX	101.5±4.9#PL,MX	−0.11 ± 0.05#	−0.16 ± 0.06PL	0.02 ± 0.01
	filtered	5.4 ± 0.8#	58.8 ± 6.1#	53.1 ± 5.4#	−0.19 ± 0.04	−0.27 ± 0.04	−0.01 ± 0.01#
PL	whole	8.5 ± 1#	24.1 ± 2.1#	26.2 ± 2.1#	0.13 ± 0.06#	0.21 ± 0.07#	0.03 ± 0.01
	filtered	1.8 ± 1.7#	17.5 ± 3.4#	9.5 ± 3.3#	0.12 ± 0.04#	0.05 ± 0.07	0.02 ± 0.01#
MX	whole	1.7 ± 2.5#	9 ± 4.8#	6.4 ± 4.1#	0.12 ± 0.09#	0.11 ± 0.07#	0.02 ± 0.01
	filtered	0.7 ± 2.3#	16.5 ± 5.4#	14.2 ± 4#	−0.05 ± 0.04	−0.11 ± 0.04	0.01 ± 0.01#

Acute Toxicity

Mice were euthanized four hours after the final exposure to burn pit emissions. Markers of lung inflammation were assessed in BAL fluid, and histopathology of lung sections were evaluated. BAL fluid cells, (neutrophils, macrophages, and lymphocytes), inflammatory cytokines (MIP-2, IFN-γ, and TNF-α) and BAL fluid biochemistry (protein, LDH, NAG, GGT) were unaffected by exposure to any of the burn pit smoke emission atmospheres (Supplementary Tables S2 and S3). Blood cell parameters (cell numbers and size characteristics) were also unaffected (Tables S2 and S3). Examination of the left lung from animals exposed to smoldering emissions found no exposure-related histopathological changes beyond incidental or background findings, as may be expected based on this acute exposure (2 days, 1 hour per day) and post-exposure necropsy time (4 hours). Using the average PM concentrations and MV calculated during exposure, and a deposition fraction of 0.32 for the tracheobronchial and pulmonary regions of a 0.65 μm aerosol (Foster et al., 2001), we estimate that mice in our study inhaled at most 20 μg smoldering PM or 4 μg flaming PM into the tracheobronchial and pulmonary regions over the 2-day exposure protocol. These deposited doses are significantly lower than the dose of 100 μg of burn pit PM condensates given by oropharyngeal aspiration in our previous study (Kim et al., 2021), which likely accounts for the lack of inflammatory responses observed with inhalation of the burn pit emissions. Since VOCs are also present with inhaled burn pit emissions and not with aspirated PM, the results also suggest that exposure to these concentrations of burn pit emissions in totality do not contribute to an acute inflammatory response in mice after short-term exposure.

Discussion

In this study we determined the physical and chemical characteristics of emissions from combustion of three burn pit materials and a mixture under smoldering and flaming conditions. In general, these smoke emission profiles produced varying degrees of respiratory depression during exposure, which were partially ameliorated by HEPA filtration. These data show that both gaseous and particulate components are important in the acute respiratory response to the simulated burn pit emissions. Smoldering emissions produced stronger responses than flaming emissions, which are most likely related to the higher levels of PM (~10-fold) and VOCs (~5-fold). The degree of respiratory F depression was very similar to our previous work on the effects of wildfire emission exposure which found that smoldering oak and eucalyptus produced reductions in F of ~150-200 bpm and reductions in MV of ~5-10 mL (Hargrove et al., 2019). Our previous study showed that a bolus dose of 100 μg of PM condensate from these emissions can produce reductions in F of mice measured post-exposure, and can also promote some markers of lung inflammation in a few of the groups (Kim et al., 2021). The present study advances these findings through exposure of mice to freshly generated emissions, producing varying quantities of PM, gases, and VOCs, and measuring respiratory physiological parameters during exposure. This route of exposure more accurately reflects real-world exposure conditions and was enhanced by measurement of real-time respiratory responses.

Analysis of the burn pit smoke emission atmospheres revealed distinct differences in gaseous and particulate components depending on burn conditions and source materials. CO has been associated with altered lung function in humans (Canova et al., 2010) as well as increased anti-inflammatory cytokines and inhibition of lower airway inflammation in mice (Otterbein et al., 2000). While exposures were intended to produce equivalent CO concentrations, in practice they varied by fuel and temperature (~15-115 ppmv). These levels were consistent with previous biomass smoke exposures and are not believed to be acutely toxic or anti-inflammatory in mice (Hargrove et al., 2019; Mayr et al., 2005). Formaldhyde, acetaldehyde, and acrolein all cause sensory irritation as measured by decreases in breathing frequency in rats (Cassee et al., 1996). All smoldering emission types exceeded the OSHA or NIOSH short-term exposure limits (STEL) of 2 ppmv formaldehyde and 0.3 ppmv acrolein (OSHA, 2021b, 2022) except for smoldering PW, which almost met the STEL at 1.9 ppmv formaldehyde. Formaldehyde and acetaldehyde are both well-known irritants and carcinogens (Morris, 1997; Nishikawa et al., 2021). These comparisons to occupational safety recommendations indicate that the simulated burn pit exposures produce significantly elevated levels of irritant gases relevant to human health. A meta-analysis of studies examining the effects of indoor pollutants found strong evidence that exposure to VOCs in the indoor home environment increased the risk of asthma and asthma-related symptoms (Paterson et al., 2021). Previous work has documented that exposure to acrolein causes reduction in F as well as cardiovascular impacts in rats, including alterations in blood pressure and decreased left ventricular contractility (Perez et al., 2015). Exposure to a mixture of acrolein, formaldehyde, and acetaldehyde together showed a synergistic effect in reducing F when compared to each chemical individually (Cassee et al., 1996). Our study highlights that even in cases where exposed individuals may be provided particle filtration masks, they are not fully protected against inhaling burn pit smoke emissions due to the presence of irritant gases, emphasizing the inadequate nature of these commonly used methods of respiratory protection. Benzene, a known carcinogen in humans (OSHA, 2021a) was a primary constituent produced by flaming combustion of all burn pit materials, but was not significantly present under smoldering combustion conditions. Although appreciable levels of benzene (0.48, 0.5, 0.24 and 0.43 ppmv for flaming PW, CB, PL and MX, respectively) were present, these did not exceed the STEL (1-5 ppmv for NIOSH/OSHA) but did exceed the NIOSH recommended exposure limit of 0.1 ppmv for a 10 hour work shift (OSHA, 2021a). Our data show that sensory irritant chemicals are present in smoldering burn pit emissions at levels which could potentially contribute to acute respiratory changes during exposure.

Depression of F caused by smoldering PW, PL and MX emissions was ameliorated by the addition of HEPA filtration, and PW was the most affected (~60% decrease in whole smoldering vs ~28% decrease for filtered). We previously showed that filtration of eucalyptus wood smoke similarly alleviated depression of respiration in mice (Hargrove et al., 2019). We found a relative lack of effect of filtration on respiratory responses to smoldering CB smoke emissions. Smoldering CB produced the highest concentrations of irritant aldehydes among all smoldering burn pit emissions, indicating that the respiratory effects, including a ~51% and 45% decrease in F for whole and filtered smoke respectively, is likely carried predominately by these VOCs as opposed to PM.

We saw no changes in markers of systemic or deep lung inflammation, injury, or histopathology, which is likely due to the acute nature of this inhalation exposure. It is important to note that while our previous study utilized a consistent dose of burn pit material PM deposited directly into the lower respiratory tract via oropharyngeal aspiration, in the inhalation exposure system we used here, changes in respiration as well as upper respiratory scrubbing may reduce the total deposited dose which could also contribute to the lack of inflammatory changes observed in the instillation study (Kim et al., 2021). The VA recognizes that sinonasal disease is associated with burn pit exposure (Hill et al., 2022), and while our study focused on lower respiratory effects, it is possible that inflammatory changes occurred in the upper respiratory tract of exposed animals. Regardless, our real-time plethysmography system is a sensitive indicator of respiratory system response to inhaled smoke in a mouse model (Hargrove et al., 2019; Kim et al., 2019; Pauluhn et al., 2021). Biomass smoke may also affect irritant responses as determined by locomotor responses in zebrafish, in which smoldering biomass emissions were more irritating than flaming emissions when normalized to the amount of PM produced per mass of fuel burned (Martin et al., 2021). In mice, activation of TRPA1 receptors by agonist in the upper airways induces a breathing frequency reduction of ~30%, which was partially mediated by trigeminal ganglion neurons (Inui et al., 2016). These responses are similar to those observed in our burn pit exposures, indicating sensory irritation may be a likely mechanism for the acute respiratory depression we observed. In addition to upper airways irritant responses, a mechanism of action has been suggested for respiratory regional responses to burn pit-associated PM_2.5, whereby inhalation of burn pit and desert-associated PM disrupts alveolar epithelial cell metabolism, increases reactive oxygen species and inflammation, and activates cell signaling pathways leading to bronchoconstriction and symptoms reported in returning veterans (Berman et al., 2021).

Interestingly, while we saw the least physiological changes with flaming PL and MX exposures, oropharyngeal aspiration of emission condensates from flaming PL produced significant increases in BALF MIP-2 and neutrophil infiltration (Kim et al., 2021). This is likely an effect of the much greater dose (100 μg) in comparison to the present inhalation study in which we estimate that ~4 μg PM was deposited in the respiratory tract following flaming PL and MX exposures. PAH levels were highest in PM extracted from flaming PL and MX compared with flaming PW and CB, and PAHs in all flaming PM were at least an order of magnitude higher than in smoldering PM on a mass basis (Kim et al., 2021). PAHs are semi-volatile compounds derived as products of incomplete combustion and have been shown to be mutagenic and carcinogenic (Abdel-Shafy et al., 2016; Kim et al., 2021; Kim et al., 2018). PAHs and dioxins have been identified in increased levels in serum from veterans deployed to Iraq and Afghanistan, including benzo(ghi)perylene and 1,2,3,4,6,7,8- heptachlorodibenzo-p-dioxin (Xia et al., 2016). Smith et al. (2019) also demonstrated that exposure to these compounds isolated from burn pit smoke induced metabolic disruption in human lung fibroblasts. These studies show that PAHs present in smoke from the burn pit emissions tested herein could produce significant toxic effects following chronic exposures.

Conclusion

We found that all smoldering burn pit material emissions caused greater impairment of respiratory function than the corresponding flaming emissions in mice. Particle filtration of smoldering emissions provided significant protection from the loss of respiratory function, except for CB smoke, indicating that PM filtration provides significant protective effects only for certain material types. While the materials analyzed in this study are relevant for investigating in vivo and in vitro exposures, burn pits used at military facilities included even more complex mixtures, such as biological waste, electronic waste, munitions, and other wastes, and typically included accelerants such as diesel or JP-8 jet fuel, a benzene-based product (Mallon et al., 2016). Studies of additional mixtures are needed to fully characterize the components and health effects of burn pit emissions. Future analysis of mRNA gene changes in the nasal epithelium of exposed mice would allow a more sensitive assessment of the upper airways effects of acute exposures to burn pit materials.

Data Availability

The data that support the findings of this study are available upon reasonable request from the corresponding author S.H.G.