Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Apr 7.
Published in final edited form as: Int J Hyg Environ Health. 2026 Jan 30;273:114746. doi: 10.1016/j.ijheh.2026.114746

Differential metabolic profiles by training fire exposure in female firefighters

Tuo Liu a, James Hollister b, Krystal J Kern a, Michelle Valenti a,b, Shawn C Beitel a, John J Gulotta e, Sara A Jahnke f, Heather Buren g, John Haseney h, Brian O’Neill i, Caitlin St Clair j, Yiwen Liu b, Frank von Hippel a, Catherine E Mullins c, Douglas I Walker c, Jaclyn M Goodrich d, Jefferey L Burgess a, Melissa A Furlong a,*
PMCID: PMC13052827  NIHMSID: NIHMS2147059  PMID: 41619359

Abstract

Background:

Female firefighters face elevated risks for cancer and reproductive disorders, but the underlying metabolic mechanisms remain unclear.

Objectives:

This study aimed to identify urinary metabolites and metabolic processes associated with training fire exposure among female municipal firefighters.

Methods:

High-resolution metabolomics (HRM) was conducted on urine samples collected before and after live-fire training from female firefighters enrolled in the Fire Fighter Cancer Cohort Study. Linear mixed-effects models, adjusting for age, education, and Hispanic ethnicity, were used to identify differentially expressed metabolites (DEMs) with false discovery rate correction. Functional enrichment analysis (FEA) via metabolite-set enrichment analysis (MSEA) from MetaboAnalyst was performed to identify enriched metabolic processes. A stratified analysis examined the influence of fire types on post-fire metabolic profiles.

Results:

One hundred female firefighters donated a total of 200 urine samples (100 pre-, 100 post-fire). HRM was performed in four modes including HILIC(+), HILIC(−), C18(+), and C18(−). We identified 200, 300, 280, and 306 metabolites and 10, 9, 23, and 19 post-training fire DEMs from the four modes, respectively. FEA highlighted enrichment of glycerophospholipid metabolism (p < 0.05). Stratified analysis identified 11 DEMs by fire type with greater changes observed following burn room/tower exposures compared to flashover fires.

Conclusion:

Training fire exposure induced widespread metabolic alterations in female firefighters, particularly in pathways related to oxidative stress and cell damage. These findings suggest potential biological pathways linking repeated fire exposure to chronic inflammation and disease risk. Burn room/tower burn exercises elicited more pronounced metabolic shifts than flashover fires.

Keywords: Female firefighter, Fire exposure, Urinary metabolomics, Health risks

1. Introduction

Firefighters are exposed to known or probable carcinogens (Daniels et al., 2015), including but not limited to polycyclic aromatic hydrocarbons (PAHs) (Kenneth et al., 2014), benzene (Kenneth et al., 2014), formaldehyde (Keith, 2011), and per-and polyfluoroalkyl substances (PFAS) (Mitchell et al., 2025). As a result, they experience a higher risk for some cancers, such as skin melanoma (Lee et al., 2020), lung (Daniels et al., 2015), leukemia (Daniels et al., 2015), kidney (Rebecca et al., 2015) and prostate cancer (Lee et al., 2020; Rebecca et al., 2015; Grace et al., 2006). These findings come from studies conducted in Florida (Lee et al., 2020; Fangchao et al., 2006), Washington (Paul et al., 1994), California (Rebecca et al., 2015; Michael, 2007), and other parts of the United States (US) (Daniels et al., 2015; Robert et al., 2014; Dongmug et al., 2008) Firefighters also face excess overall cancer mortality (SMR = 1.14; 95 % CI 1.10–1.18) as compared to the general US population (Robert et al., 2014; Demers et al., 2022). In addition, the International Agency for Research on Cancer (IARC) recently classified firefighters’ occupational exposure as carcinogenic to humans with sufficient evidence for mesothelioma and bladder cancer (Demers et al., 2022). Despite an increasing number of studies on the health effects of firefighting, evidence on its impact among female firefighters remains sparse. A limited number of US studies have reported that female firefighters had significantly increased bladder mortality (SMR = 33.51; 4.06–121.05, sample size <5), brain (aOR = 2.54; 1.19–5.42), and thyroid (aOR = 2.42; 1.56–3.74) cancer risk (Lee et al., 2020; Fangchao et al., 2006; Robert et al., 2014). However, a study on cancer incidence and mortality among Australian female firefighters showed a similar cancer rate compared with the general population (Deborah et al., 2019).

The relationship between fireground exposure and adverse health effects in female firefighters is not well understood. In addition to elevated cancer risks, fireground exposure may also contribute to reproductive issues in female firefighters. Previous epidemiological studies have shown that female firefighters have high rates of adverse reproductive outcomes, including miscarriage (Davidson et al., 2022), preterm birth (Davidson et al., 2022), and lower levels of anti-müllerian hormone (AMH) (Davidson et al., 2022), an important measure directly associated with reproductive reserve. However, the link between toxic exposure, disturbed metabolism, and adverse health outcomes is yet to be established. Our previous study in male firefighters identified a large set of differentially expressed metabolites (DEMs) comparing metabolites in urine samples before and after municipal structure fire (MSF) exposure (Furlong et al., 2023). Firefighters are also exposed to products of combustion during live-fire training, where they generally participate in flashover and burn room/tower fires. Activities during fire training can vary depending on departmental protocols and fire types. For example, flashover training may involve firefighters remaining stationary to observe fire behavior or also include active tasks. In contrast, burn room or tower burn exercises more commonly involve dynamic activities such as search and rescue. Additionally, the materials burned in each type of training fire often differ.

Given the notable lack of comprehensive research on firefighting-related metabolic changes in female firefighters, our study aimed to analyze changes in urine metabolic profiles before and after training fires that simulate some of the exposures firefighters face during municipal structure fires (MSFs). We hypothesized we would identify distinct DEMs specific to female firefighters following training fire exposure, with some overlap in DEMs between the female firefighters and the prior research on male firefighters participating in MSFs. We further anticipated discovering fire type-specific DEMs among female firefighters, and that these metabolic changes are expected to have important implications for firefighter health.

2. Methods

2.1. Study population

Building on the existing Fire Fighter Cancer Cohort Study (FFCCS) framework (Burgess et al., 2024), this study enrolled female firefighters into the Federal Emergency Management Agency (FEMA) funded Female Firefighter Sub-Study: Evaluation of Exposures and Toxicity (FEMA EMW EMW-2021-FP-00141). Female career and volunteer firefighters at least 18 years of age were enrolled from 11 fire agencies across 21 fire departments and 8 states. Participants were enrolled into the study prior to starting their live-fire training. Standardized surveys were administrated at enrollment to collect information on participant demographics, cancer and behavioral risk factors, and firefighting history. Study data were collected and managed using REDCap. The institutional review board at participating institutes approved this research, and all participants provided informed consent.

2.2. Sample collection & preparation

All pre- and post-fire samples were collected from October 2022 to July 2024. Pre-fire samples were collected on the same day of the live fire training before entering the training fire. Post-fire samples were aimed to be collected 2–4 h after completing the training fire. Post-fire samples were collected within 1.8–11.0 h after the training fire exposure. Participants completed a detailed survey with each collection that included information on the type of training fire (flash over, room/tower burn, other), medication and supplement use, and menstrual cycle questions, and perceived gear fit. All urine was collected in 150 mL polypropylene collection cups, initially frozen at −20 °C, then shipped on dry ice to the University of Arizona. Upon receipt, samples were stored at −80 °C until processed. Urine was thawed in a rotary incubator at 25 °C for 2 h or until fully thawed. Samples were mixed via inversion, then aliquoted into polypropylene cryovials comprising of 1.8 mL and 12.0 mL aliquots. During sample processing, specific gravity was measured using an Atago pocket refractometer. Processed samples were then stored at −80 °C.

2.3. High-resolution metabolomics

All urine samples were shipped on dry ice to the Comprehensive Laboratory for Untargeted Exposome Science at Emory University for high-resolution metabolic profiling. Urine samples were prepared in batches that included both study samples and quality assurance/quality control (QA/QC) samples using an Opentrons OT2 automated liquid handler and 96-well plates. Prior to processing, samples were thawed at 4 °C. A 40 μL aliquot of each sample was extracted with 120 μL of acetonitrile containing 13C-labeled internal standards (ISs). Treated samples were vortexed for 2 min, left to equilibrate at 4 °C for 30 min, and subsequently centrifuged at 3220×g for 15 min at 4 °C. Following centrifugation, two 30 μL aliquots of the supernatant were transferred to separate 96-well plates, each containing either 60 μL of water (for Reverse Phase Chromatography (C18)) or 60 μL of a 1:1 acetonitrile/water solution (for Hydrophilic Interaction Liquid Chromatography (HILIC)). These were vortexed for another 2 min and stored in a refrigerated autosampler until they were analyzed. Untargeted analysis was performed using two separate systems configured for C18 or HILIC analysis. These consisted of a Vanquish Duo Ultra Performance Liquid Chromatography (UPLC) unit (Thermo Fisher Scientific, Rockford, IL, USA) paired with an Exploris 120 High Resolution Mass Spectrometry (HRMS) system (Thermo Fisher Scientific, Rockford, IL, USA). The LC column temperatures were maintained at 40 °C for HILIC and 30 °C for C18, while the autosampler was kept at 5 °C. Samples were analyzed using dual column chromatography with mobile phases optimized for positive or negative ionization. For C18 chromatography, a Higgins TARGA C18 5 μm 50 × 2.1 mm column (Higgins Analytical, Inc., Mountain View, CA, USA) was used in both positive and negative ionization modes. HILIC chromatography utilized a Waters Atlantis Premier BEH Z-HILIC, 4.6 × 50mm, 2.5 μm particle size column (Waters Corporation, Milford, MA, USA) for positive mode and an XBridge Amide 3.5 μm 3.0 × 50mm column (Waters Corporation, Milford, MA, USA) for negative mode. The mobile phases for the reverse phase C18 analysis included water containing 0.1 % formic acid and 95/5 (v/v) acetonitrile water containing 0.1 % formic acid for positive mode, and 10 mM ammonium acetate in water and 95/5 (v/v) acetonitrile/water for negative mode. HILIC mobile phases included water containing 0.1 % formic acid and 10 mM ammonium formate, and 95/5 (v/v) acetonitrile/water containing 0.1 % formic acid for positive mode, and 10 mM ammonium acetate in water adjusted to pH 9.5 and 95/5 (v/v) acetonitrile/water for negative mode. Flow rates ranged from 0.3 mL/min to 0.6 mL/min, and the total run time was 7.5 min.

2.4. Feature preprocessing and metabolite annotation

Following analysis of the study and QAQC samples, raw instrument files were converted to mzML with peak picking and extracted using XCMS. Detected m/z were grouped using RamClustR to identify m/z’s corresponding to the same compound. Batch correction was performed using log2-transformed raw intensities with WaveICA 1.0, which corrects for both inter- and intra-batch effects.

Metabolites were identified by comparing detected m/z and retention time to a database of 1200 standards analyzed using the same method parameters that included a wide range of environmental and endogenous compounds. Metabolite identifications were achieved uniquely by matching m/z and retention time with a tolerance of 5 ppm and 15 s, respectively.

2.5. Data preprocessing and statistical analysis

We followed the previously published metabolomics pipeline in this study (Liu et al., 2025a) for adjustment of urinary dilution and removal of unwanted variations. We restricted differential analysis to metabolites with level 1 annotation confidence where metabolites’ structure was confirmed against standards and present in at least 75 % of the samples. All metabolic features’ ion intensities were log2 transformed and standardized to stabilize variation and facilitate linear model assumptions. Differential analysis was performed using a linear mixed effects model fitted on the relative intensity of each metabolite identified with exposure status (pre-fire/postfire) being the major predictor while adjusting for demographic covariates including age, education, Hispanic ethnicity, and type of training fires. We performed a complete-case analysis in this study and participants with missing demographics were excluded from modeling. Missing here refers to missing values due to participants not responding. The model slope was then extracted from the modeling results and interpreted as changes in the expression of metabolite with fire exposure where positive slope indicated increased expression and negative slope indicated decreased expression. P-values were adjusted by controlling for a false discovery rate (FDR) at 5 %. We also conducted exploratory analyses where we relaxed the cut-off to evaluate DEMs as those with a raw p-value less than 0.05. A volcano plot was made to showcase the overall differential status of the metabolome by fire exposure, with identified DEMs highlighted. For easy comparison with previous studies, fold changes were also calculated and presented in a circular bar chart for DEMs comparing postfire and pre-fire average ion intensities to showcase the changes at a metabolite level, together with a two-sample t-test (Supplemental Table 2).

Additionally, we performed a secondary analysis to compare female firefighters’ post-fire metabolic profiles by fire type. A multiple linear regression model was fitted on preprocessed intensities of each metabolite with fire type being the main predictor while adjusting for covariates including Hispanic ethnicity, age, and education. P-values were extracted, and multiple testing was adjusted for with a threshold of FDR q < 0.05, with exploratory findings reported for raw p < 0.05.

All statistical analyses were performed in the R programming environment (version 4.4.1)(R Core Team, 2023).

2.6. Functional enrichment analysis

Following preprocessing and differential analysis, we performed functional enrichment analysis (FEA) to investigate the difference in metabolic profiles by live-fire exposure at the biological pathway level. Targeted quantitative enrichment analysis was performed using MetaboAnalyst (version 6.0)(Pang et al., 2024; Xia et al., 2009). Sum normalization and auto-scaling were applied to remove systematic variability. Only pathways with at least two entries were included. We defined the statistical significance for enrichment as a raw p-value of 0.05.

3. Results

3.1. Study population and sample

Among the 100 female firefighters in the study, 31 % were Hispanic, and 61 % had a college degree. Roughly half of the participants responded to flashover training fires and the other half to burn room/tower burn training fires (Table 1). Among participants with available data (n = 53), 11 (21 %) reported having smoked at least 100 cigarettes in their lifetime; smoking history was missing for 47 participants. Alcohol consumption varied, with 39 % reporting no alcohol use in the past week. A family history of cancer was reported by 21 % of the participants. Regarding acute health factors during the study period, 4 % were taking prescribed medications and 9 % were using over the counter (OTC) medications.

Table 1.

Summary statistics of demographics for female firefighters participating in training fires (N = 100) for the female firefighter study.

N = 100a
Age 30.0 (21.0, 40.0)
Hispanic ethnicity 31/100 (31 %)
Education
 Some College or lower 39/100 (39 %)
 College graduate 61/100 (61 %)
Type of fire
 Burn Room/Tower Burn 49/100 (49 %)
 Flash Over 51/100 (51 %)
Family cancer history 21/100 (21 %)
Prescribed medication current use 4/100 (4.0 %)
OTC medication current use 9/100 (9.0 %)
Smoked at least 100 cigarettes
 Yes 11/53 (21 %)
 (Missing) 47
Any alcohol in past week
 No 39/100 (39 %)
 Yes 61/100 (61 %)
Open in a new tab
a

Median (IQR) or Frequency (%).

3.2. High-resolution metabolomics

We identified a total of 247, 202, 321, and 291 urinary metabolites with level 1 annotation confidence from HILIC(−), HILIC(+), C18(−), and C18(+) modes, respectively. Most of the identified metabolites were endogenous with a small proportion of them being markers for per- and polyfluoroalkyl substances (PFAS), pesticides, phenol, phthalates, and nicotine. After removing metabolites that were present in less than 75 % of samples, a total of 224, 180, 300, and 263 features were retained from HILIC(−), HILIC(+), C18(−), and C18(+) modes, respectively.

3.3. Differential analysis and statistical analysis

After adjustment for age, education, Hispanic ethnicity, and fire type, the main model identified no metabolites that were enriched after multiple testing adjustment. In exploratory analyses with a relaxed raw p <0.05, we identified 10, 9, 23, and 19 DEMs by fire exposure for HILIC (+), HILIC(−), C18(+), and C18(−) mode, respectively, when comparing pre-fire and post-fire urine samples of female firefighters participating in training fires (Table 2). Fig. 1A presents the overall differential status for this exploratory analysis with the top 10 DEMs annotations for each of the separation-ionization mode. Overall, we saw more upregulations than downregulations after training fire exposure in female firefighters. A complete list of differential status by fire exposure can be found in Supplemental Table 1.

Table 2.

DEMs by statistical significance within each separation (ionization) mode comparing post-fire to pre-fire urine samples for female firefighters participating in training fires.

Metabolite Molecular Formula PubChem CID MZ RT Adduct Slope P-Value Mode
Nicotine C10H14N2 942 163.12 333.80 M +H 0.38 0.01 HILIC(+)
Valine C5H11NO2 6287 118.09 263.40 M +H 0.17 0.01 HILIC(+)
Fenobucarb C12H17NO2 19588 208.13 67.70 M +H 0.32 0.02 HILIC(+)
4-Quinolinecarboxylate C10H7NO2 10243 174.06 94.30 M +H 1.14 0.02 HILIC(+)
2,4-Dihydroxypteridine C6H4N4O2 10250 165.04 97.40 M +H 1.51 0.02 HILIC(+)
Guanidinoacetate C3H7N3O2 763 118.06 306.90 M +H 0.53 0.03 HILIC(+)
N-Acetylphenylalanine C11H13NO3 74839 249.12 73.20 M + ACN + H 0.63 0.04 HILIC(+)
L-Carnitine C7H15NO3 10917 162.11 294.30 M +H 0.37 0.04 HILIC(+)
L-Cartinine C7H15NO3 2724480 162.11 297.10 M +H 0.37 0.04 HILIC(+)
Sphinganine C18H39NO2 91486 302.31 203.50 M +H −0.07 0.05 HILIC(+)
Succinic acid C4H6O4 1110 117.02 306.50 M-H 0.19 0.03 HILIC(−)
[C8.1]-2-Octenoic acid C8H14O2 5282713 141.09 68.00 M-H 0.33 0.03 HILIC(−)
Benzoate C7H6O2 243 121.03 82.02 M-H 0.41 0.04 HILIC(−)
4-Hydroxybenzaldehyde C7H6O2 126 121.03 67.70 M-H 0.41 0.04 HILIC(−)
Benzoic acid C7H6O2 243 121.03 70.40 M-H 0.41 0.04 HILIC(−)
Methyl Vanillate C9H10O4 19844 181.05 70.69 M-H 0.44 0.04 HILIC(−)
4-Hydroxyphenylpyruvic acid C9H8O4 979 179.04 70.60 M-H 0.64 0.04 HILIC(−)
3-Hydroxybenzaldehyde C7H6O2 101 121.03 64.03 M-H 0.45 0.04 HILIC(−)
Alpha-ketoisovaleric acid C5H8O3 49 231.09 69.00 M-H 1.44 0.05 HILIC(−)
7-Ketochenodeoxycholate C24H38O4 53477693 391.28 270.17 M +H 0.19 0.00 C18(+)
Valine C5H11NO2 6287 118.09 21.60 M +H 0.20 0.01 C18(+)
5-Aminovaleric acid C5H11NO2 138 118.09 19.48 M +H 0.20 0.01 C18(+)
Deoxycarnitine C7H15NO2 725 146.12 22.37 M +H 0.79 0.02 C18(+)
Acetylcholine C7H16NO2+ 187 146.12 20.31 M+ 0.79 0.02 C18(+)
N,N-Dimethyl-1,4-Phenylenediamine C8H12N2 7472 137.11 22.97 M +H 0.82 0.02 C18(+)
Mannitol C6H14O6 6251 205.07 21.41 M + Na 1.16 0.02 C18(+)
Histamine C5H9N3 774 112.09 17.45 M +H 0.84 0.02 C18(+)
N-acetyl-tyrosine C11H13NO4 68310 224.09 44.85 M +H 0.88 0.02 C18(+)
Glycochenodeoxycholic acid C26H43NO5 12544 450.32 255.50 M +H 0.83 0.02 C18(+)
O-Acetylcarnitine C9H17NO4 439756 204.12 23.00 M +H 0.77 0.03 C18(+)
Acetyl-DL-carnitine C9H17NO4 7045767 204.12 19.97 M +H 0.77 0.03 C18(+)
Salsolinol C10H13NO2 91588 180.10 21.41 M +H 0.47 0.03 C18(+)
Dihydrouracil C4H6N2O2 649 132.08 22.29 M + NH4 1.09 0.03 C18(+)
Creatine C4H9N3O2 586 132.08 22.29 M +H 1.09 0.03 C18(+)
Pterin C3H6N2S 135398660 164.06 23.74 M +H 1.21 0.03 C18(+)
Testosterone C19H28O2 6013 289.22 262.32 M +H 0.86 0.04 C18(+)
Aminoisobutanoate C4H9NO2 64956 104.07 21.30 M +H 0.19 0.04 C18(+)
Gamma-Aminobutyrate C4H9NO2 119 104.07 20.73 M +H 0.19 0.04 C18(+)
Alpha-aminobutyric acid C4H9NO2 80283 104.07 19.72 M +H 0.19 0.04 C18(+)
2,6-Dihydroxypyridine C5H5NO2 69371 112.04 25.93 M +H 0.40 0.04 C18(+)
Propionyl-L-carnitine C10H19NO4 107738 218.14 21.55 M +H 0.63 0.04 C18(+)
Propanoylcarnitine C10H19NO4 188824 218.14 22.26 M +H 0.63 0.04 C18(+)
[C10.1]-Cis-2-Decenoic acid C10H18O2 5356596 169.12 237.80 M-H 0.20 0.01 C18(−)
2′,4′-Dihydroxyacetophenone C8H8O3 6990 151.04 191.70 M-H 0.19 0.01 C18(−)
Resorcinol Monoacetate C8H8O3 5055 151.04 190.10 M-H 0.19 0.01 C18(−)
[C19.1]-10Z-Nonadecenoic acid C19H36O2 5312513 295.26 397.20 M-H 0.21 0.01 C18(−)
4-Hydroxybenzoate; 4-Hydroxybenzoic acid C7H6O3 135 137.02 21.00 M-H 0.44 0.02 C18(−)
[C18.2]-11-Octadecen-9-ynoic acid C18H30O2 5312688 277.22 325.00 M-H 0.49 0.02 C18(−)
3-(4-hydroxyphenyl)propanoic acid C9H10O3 10394 165.06 20.00 M-H 0.39 0.03 C18(−)
N-Acetylglucosamine C8H15NO6 439174 256.06 20.90 M + Cl 0.26 0.03 C18(−)
N-Acetylglutamate C7H11NO5 70914 188.06 18.00 M-H 0.25 0.03 C18(−)
[C12.1]-5-Dodecenoic acid C12H22O2 5312378 197.15 263.30 M-H 0.08 0.03 C18(−)
Alpha-Muricholic acid C24H40O5 53477700 407.28 208.30 M-H 0.82 0.04 C18(−)
Ursocholic acid C24H40O5 122340 407.28 199.30 M-H 0.82 0.04 C18(−)
4-Aminobenzoate C7H7NO2 978 136.04 20.60 M-H 0.20 0.04 C18(−)
2-Deoxy-D-Glucose C6H12O5 439268 163.06 20.70 M-H 0.25 0.04 C18(−)
Porphobilinogen C10H14N2O4 1021 225.09 19.30 M-H 0.72 0.04 C18(−)
Glutamate C5H9NO4 33032 146.05 18.60 M-H 0.29 0.04 C18(−)
2-Keto-3-Deoxy-D-Gluconic Acid C6H10O6 194024 177.04 19.00 M-H 0.19 0.04 C18(−)
Theophylline C7H8N4O2 2153 179.06 34.50 M-H 1.28 0.04 C18(−)
2,6-Dihydroxypyridine C5H5NO2 69371 110.02 19.80 M-H 0.43 0.05 C18(−)
Open in a new tab

MZ: mass to charge ratio; RT: retention time in seconds; Slope: the coefficient estimates for the exposure status (‘pre-fire’-‘post-fire’: 0–1) term of the main model. Positive slopes indicated an increase in the relative ion intensity of metabolites after fire exposure whereas negative slopes indicated a decrease.

Fig. 1.

Fig. 1.

Open in a new tab

Combo plot for differential analysis and functional enrichment analysis. A Volcano plot of DEMs comparing post-fire to pre-fire urine samples from female firefighters participating in training fires. The slope term was from the main model where exposure status (post-fire/pre-fire) was the main predictor and preprocessed ion intensity was the response. Upregulation is marked in red whereas downregulation in blue. The horizontal dashed line marked “p-value = 0.05″ is the boundary that served as the threshold for differential status in this study. P-values were −log10 transformed. B Circular bar chart for fold changes of DEMs identified by DA. Unique CID for each of the DEMs was used for visualization purposes and a reference to metabolite name can be found in Table 2. Fold change is reflected by the sizes of the circles and mode is indicated in 4 colors. C Enrichment plot for the comparison between post-fire vs. pre-fire urine samples among female firefighters participating in training fires, by four separation(ionization) modes. Statistical significance was determined by MSEA. The vertical dashed line in black and grey marked “p-value = 0.05″ and “p-value = 0.1”, respectively, is the boundaries that served as the thresholds for enrichment status in this study. Multiple testing was adjusted with a FDR threshold at 0.05 to avoid false positives. The enrichment ratio was defined as the ratio of the number of significant hits from the user input list of differential metabolites to the number of expected metabolites in each pathway. Separation-ionization is marked in purple, blue, dark green, and light green. The enrichment ratio is reflected by the size of the dot where large sizes indicate larger enrichment. The p-values were −log10 transformed. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

3.4. Functional enrichment analysis

Metabolite-set enrichment analysis (MSEA) by mapping all DEMs to KEGG pathways revealed no pathways that were enriched at an FDR<0.05, and one significantly enriched pathway at a relaxed p-value< 0.05: glycerophospholipid metabolism. Enrichment status from each of the 4 modes was presented in Fig. 1C and detailed statistical tests for top 5 KEGG terms can be found in Supplemental Table 3. At a metabolite level, the only hit from our identified DEMs in the glycerophospholipid metabolism was acetylcholine (slope = 0.79, p-value = 0.02). By the hard threshold of p-value 0.05 or FDR 0.05, no other significant enrichments were observed in this study. We also observed a modest increase in acetylcholine and nicotine levels following fire exposure (Fig. 3).

Fig. 3.

Fig. 3.

Open in a new tab

Boxplot of selected metabolite expression at pre-fire and post-fire. Ion intensities were log2 transformed. Solid dots represent individual measurements from participants where outliers are marked in a different shape.

3.5. Secondary analysis: effect of fire type

After adjusting for age, Hispanic ethnicity, and education, stratified analysis by fire type (burn room/tower vs. flashover) on the metabolome using multiple linear regression, there were no significant differences after adjusting for multiple testing. In exploratory analyses using a raw p < 0.05, we identified a total of 17 DEMs by fire type, including 3, 2, 5, and 1 DEMs, from HILIC(+), HILIC(−), C18(+), and C18(−) modes, respectively. A direct comparison of the relative ion intensity between burn tower/burn room fires and flash over fires was plotted using boxplots for each of the DEMs identified by the secondary model in Fig. 2C. Most DEMs by fire type saw increased intensities in burn room/burn tower fires compared to flashover fires. A complete list of DEMs for sub-analysis with model details can be found in Supplemental Table 4. Burn room/tower fires introduced a larger number of DEMs, and thus a more intense differential profile than flashover fires in female firefighters (Fig. 2A and B).

Fig. 2.

Fig. 2.

Open in a new tab

A Volcano plot of DEMs comparing post-fire to pre-fire urine samples from female firefighters participating in flashover training fires. The slope term was from the main model where exposure status (postfire/pre-fire) was the main predictor and preprocessed ion intensity was the response. Upregulation is marked in red whereas downregulation in blue. The horizontal dashed line marked “p-value = 0.05″ is the boundary that served as the threshold for differential status in this study. P-values were −log10 transformed. B Volcano plot of DEMs comparing post-fire to pre-fire urine samples from female firefighters participating in burn room/tower training fires. The slope term was from the main model where exposure status (postfire/pre-fire) was the main predictor and preprocessed ion intensity was the response. Upregulation is marked in red whereas downregulation in blue. The horizontal dashed line marked “p-value = 0.05″ is the boundary that served as the threshold for differential status in this study. P-values were −log10 transformed. C Boxplots comparing metabolites that were differentially expressed at p-value 0.05 level by training fire types, organized by separation(ionization) mode. Two sample t-tests were performed to derive statistical significance. Yellow boxes represented flash-over fires while blue boxes indicated burn room/tower burn fires. Dots on the top indicated the separation(ionization) mode with orange, pink, blue, and purple representing C18(+), C18(−), HILIC(+), and HILIC(−), respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

4. Discussion

In this study, we performed targeted metabolomics to better understand the health risk imposed by fire exposures in female firefighters. This approach allowed us to profile the metabolome and identify differential metabolic expressions across 100 female firefighters. Out of identified urinary metabolites, we identified 10, 9, 23, and 19 DEMs by fire exposure from HILIC(+), HILIC(−), C18(+), and C18(−) mode, respectively. These differences were enriched in glycerophospholipid metabolism, alterations of which are associated with inflammation (Zhu et al., 2022). Identification of DEMs and enrichments in the biological processes may help explain the increased cancer risk by fire exposure among female firefighters. Additionally, the post-fire metabolic profiles vary by fire type, suggesting metabolic responses to fire exposure may be strongly related to type of training fires and activities at the fire scene. However, results were not significant when adjusting for multiple comparisons.

4.1. Training fire related enrichment in metabolism

Firefighters are exposed to high levels of oxidative stress due to inhaling smoke which includes toxic chemicals and particulate matter (Williams et al., 2013; Olorunfemi et al., 2013). Oxidative stress burden and chronic inflammation are related to multiple disorders, including but not limited to cancer (Lowe and Laher, 2014), cardiovascular diseases (Lowe and Laher, 2014), and renal disease resulting from the excessive accumulation of reactive free radicals (Reddy, 2023). Additionally, oxidative stress, chronic inflammation, and cancer are closely related. Long term oxidative stress can lead to chronic inflammation, which in turn leads to tumorigenesis, tumor cell survival, proliferation, and more (Reuter et al., 2010).

Existing evidence has implicated that oxidative stress alters cellular lipid metabolism (Bartolacci et al., 2021). Glycerophospholipids are major components of cell membranes, maintaining membrane integrity, fluidity, and are involved in signaling processes (Hishikawa et al., 2014). Oxidative stress can damage cellular membranes, prompting increased turnover and repair via enhanced glycerophospholipid metabolism (Zhang et al., 2024).

The increased activity in glycerophospholipid metabolism may represent an adaptive response to counteract damage induced by fire exposure. The body ramps up lipid turnover and repair mechanisms to restore normal cellular functions. While this may initially be protective, chronic or repeated exposure and the resulting persistent metabolic changes might contribute to persistent cellular stress. Over time, such changes could potentially be linked to an increased risk of inflammatory conditions (Reuter et al., 2010), cardiovascular disease (Dubois-Deruy et al., 2020), or cancer (Reuter et al., 2010; Singh et al., 2019) observed at higher rates in firefighters.

The hit from our identified DEMs with the metabolite set in glycerophospholipid metabolism was acetylcholine. Choline is the precursor for acetylcholine, a neurotransmitter synthesized by the enzyme choline acetyltransferase and a product of glycerophospholipid metabolism (Kanehisa, 2000). An increase in glycerophospholipid turnover can elevate the availability of choline, thereby potentially increasing acetylcholine synthesis. We saw a slightly higher level of acetylcholine after fire exposure (Fig. 3), which might be indicative of activated glycerophospholipid metabolism in response to cellular damage from oxidative stress caused by fire exposure.

Beyond global pathway enrichment, individual DEMs identified in this study provide a molecular link to the increased incidence of female-specific malignancies and reproductive disorders. Some metabolites display properties directly related to oxidative stress or inflammation or play documented roles in oncogenesis. 4-hydroxybenzoic acid (4-HBA) is a phenolic antioxidant that scavenges free radicals, but it also functions as a specific HDAC6 inhibitor that can modulate the tumor microenvironment in breast cancer models (Wang et al., 2018). N-acetyl-tyrosine is an acetylated amino acid that may mitigate oxidative damage (Dubois-Deruy et al., 2020), likely acting as a buffer against the acute reactive oxygen species (ROS) burden generated during fire exposure. Additionally, histamine is a potent mediator released during inflammation that synergizes with oxidative stress; notably, histamine signaling through the H4 receptor, which is highly expressed in breast and endometrial tissues, is associated with increased tumor cell proliferation and immune evasion (Medina and Rivera, 2010). Finally, 4-hydroxyphenylpyruvic acid is a precursor to tyrosine metabolism, which is closely linked to systemic inflammation (Singh et al., 2019). These identified oxidative and inflammatory markers provide a potential biological pathway for the elevated health risks observed in this population, including higher rates of miscarriage and reduced ovarian reserve documented among female firefighters (Davidson et al., 2022; Jahnke et al.). All in all, the enrichment of glycerophospholipid metabolism, together with other DEMs with oxidation-related properties observed in firefighters likely reflects the response to environmental stress, primarily oxidative stress, induced by fire-related toxicants. This metabolic shift is part of the cellular effort to repair damaged membranes and to regulate inflammatory responses. Understanding these changes not only sheds light on the acute biological responses to fire exposure but also highlights a potential pathway for understanding and improving long-term occupational health in female firefighters.

Nicotine was one of the DEMs that was significantly higher in post-fire samples. Nicotine is generally considered a biomarker for tobacco use, but given the short half-life of nicotine, cotinine has been used as the biomarker for exposure to cigarette smoke (Schick et al., 2017). Nicotine was detected at both pre-fire and post-fire in female firefighters and found to be increased post-fire (Fig. 3), without the presence of cotinine, possibly indicative of other sources for nicotine than cigarette smoke. This is supported by the results of a survey study conducted in 2015 finding only 5.1 % of women firefighters were current smokers and 1.2 % using smokeless tobacco (Jitnarin et al., 2019). Furthermore, the National Fire Protection Association (NFPA) Standard 1500 requires that all fire department facilities be designated as smoke-free and most of the study participants were recruits in career departments.

4.2. Burn room/tower fires induced differential profiles in the post-fire metabolome

In addition to the differences in burning materials, burn room/tower burn fire simulations typically involve more physically demanding tasks such as active searches and rescues, whereas flashover fire scenarios generally require less overall activity. As expected, a broader differential metabolic profile was observed in firefighters exposed to burn room/tower burn fires compared to those exposed to flashover fires. However, without additional data, we cannot determine the relative contributions of burning materials and physical activity to the observed metabolic differences. Further research is needed to disentangle these factors and better understand their individual impacts on firefighters’ metabolic responses.

4.3. Comparison with male firefighters

To better contextualize the findings in female firefighters, we provide a preliminary benchmarking comparison with previously studied male firefighters responding to MSF and WUI fires. We previously reported metabolic differences before and after fire exposure in male firefighters who responded to MSFs and identified a larger number of DEMs (N = 268), and thus a broader metabolic response associated with MSF exposure (Furlong et al., 2023). In male firefighters participating in WUI fires (Liu et al., 2025b), we also observed extensive metabolic responses post-fire, identifying N = 176 DEMs at FDR <0.05. Several metabolites, including glutamate (amino acid), N-acetylphenylalanine (acetylated amino acid), 4-aminobenzoate (vitamin-related compound), histamine (biogenic amine), and glycochenodeoxycholic acid (bile acid), were consistently elevated after fire exposure across all three fire exposure scenarios. The more pronounced response observed in male firefighters, compared to their female counterparts, is likely attributed to the nature of their fire exposure. In our sample, male firefighters responded to actual fire incidents, whereas the female samples were collected during training fires. This difference in exposure intensity may explain the weaker metabolic responses observed among female firefighters in this study.

4.4. Strengths and limitations

This study leveraged high-throughput metabolomics to analyze a relatively large cohort of 100 female firefighter participants using a well-curated library of approximately 1200 metabolite standards. This allowed for comprehensive metabolic profiling and strengthened confidence in our findings and interpretations. An additional strength was the pre- and post-fire sampling design, which enabled within-individual comparisons and helped control for time-invariant confounders. However, several limitations should be noted. First, we lacked detailed information on the duration of each training fire and the specific materials burned, both of which are critical for characterizing exposure. The potential for acute confounding from behaviors occurring between the pre- and post-fire sample collection (1.8–11.0 h) should be noted. For instance, while chronic smoking is controlled for by the within-subject design, an increase in nicotine levels was observed post-exposure (p = 0.01), which may reflect recent tobacco use or environmental uptake during training. Second, training fires varied across departments and academies in terms of structure, intensity, and procedures, introducing heterogeneity that may have influenced metabolic responses. Third, the controlled nature of training fire exposures and the modest enrichment significance observed limit the generalizability of our findings to real-world fire scenes, where exposures may be more intense and variable. Four, while we provide a preliminary comparison to male firefighter cohorts, this analysis is limited by the lack of harmonized training protocols across the different studies. A systematic, side-by-side study using identical exposure scenarios is needed to fully elucidate sex-based metabolic differences. Finally, it is important to note that this untargeted metabolomics approach is inherently exploratory and hypothesis-generating. While we have identified metabolic signatures that align with known cancer and reproductive health pathways, these do not establish causality. Future longitudinal studies and targeted metabolic assays are required to validate the link between the acute metabolic disruptions identified here and the long-term clinical health conditions observed in female firefighters.

5. Conclusion

Female firefighters exposed to training fires exhibited a set of metabolic changes, particularly involving metabolites and biological processes related to cellular damage from oxidative stress. These changes suggest a potential pathway for chronic inflammation with long-term exposure, which may help explain the higher prevalence of certain health conditions observed in female firefighters. The type of fire appears to play a significant role in shaping metabolic responses. Future research is needed to clarify the connections between these metabolic changes and the development of target diseases, as well as further investigate the specific contributions of different fire types to the changes in metabolome and the health outcomes.

Supplementary Material

MMC1

Acknowledgements

We thank our fire service research liaisons that facilitated enrollments and sample collections and thank all study participants who provided their time and samples for this study. We are also grateful for our funding agencies.

Funding

The Federal Emergency Management Agency (FEMA) funded the research through EMW-2015-FP-00213 and EMW-2021-FP-00141. The research was also supported by NIEHS P30ES006694.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijheh.2026.114746.

Footnotes

CRediT authorship contribution statement

Tuo Liu: Writing – review & editing, Writing – original draft, Visualization, Methodology, Formal analysis, Data curation. James Hollister: Writing – review & editing, Resources, Project administration, Data curation. Krystal J. Kern: Writing – review & editing, Resources, Project administration, Data curation. Michelle Valenti: Writing – review & editing, Resources, Project administration, Investigation, Data curation. Shawn C. Beitel: Writing – review & editing, Supervision, Resources, Project administration, Investigation, Data curation, Conceptualization. John J. Gulotta: Writing – review & editing, Resources, Project administration, Investigation. Sara A. Jahnke: Writing – review & editing, Supervision, Resources, Investigation. Heather Buren: Writing – review & editing, Resources, Investigation. John Haseney: Writing – review & editing, Investigation. Brian O’Neill: Writing – review & editing, Resources, Investigation. Caitlin St Clair: Writing – review & editing, Investigation. Yiwen Liu: Writing – review & editing, Supervision, Funding acquisition. Frank von Hippel: Writing – review & editing, Supervision, Resources. Catherine E. Mullins: Writing – review & editing, Software, Methodology, Data curation. Douglas I. Walker: Writing – review & editing, Supervision, Resources, Methodology, Data curation. Jaclyn M. Goodrich: Writing – review & editing, Supervision, Funding acquisition. Jefferey L. Burgess: Writing – review & editing, Supervision, Methodology, Investigation, Funding acquisition, Data curation, Conceptualization. Melissa A. Furlong: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Conceptualization.

References

  1. Bartolacci C, Andreani C, El-Gammal Y, Scaglioni PP, 2021. Lipid metabolism regulates oxidative stress and ferroptosis in ras-driven cancers: a perspective on cancer progression and therapy. Front. Mol. Biosci 8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Burgess J, Shawn CB, Miriam MC, Melissa AF, Paola Louzado F, Jamie Kolar G, Grant C, Jaclyn MG, Judith MG, Olivia H, James H, Jeff H, Sara AJ, Krystal JK, Frank AL, Caban-Martinez A, Alexander CM, Russell O, Cynthia P, Sreenivasan R, Heather MS, Solle NS, Christine T, Derek JU, Michelle V, Gulotta J, 2024. The fire fighter cancer cohort study: protocol for a longitudinal occupational cohort study of United States firefighters (Preprint). JMIR Res. Protoc. 10.2196/70522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Daniels RD, Bertke S, Dahm MM, Yiin JH, Kubale TL, Hales TR, Baris D, Zahm SH, Beaumont JJ, Waters KM, Pinkerton LE, 2015. Exposure–Response relationships for select cancer and non-cancer health outcomes in a cohort of Us firefighters from San Francisco, Chicago and Philadelphia (1950–2009). Occup. Environ. Med 72 (10), 699. 10.1136/oemed-2014-102671. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Davidson S, Jahnke S, Jung AM, Burgess JL, Jacobs ET, Billheimer D, Farland LV, 2022. Anti-Müllerian hormone levels among female firefighters. Int. J. Environ. Res. Publ. Health 19 (10). 10.3390/ijerph19105981. Epub 20220514 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Deborah Catherine G., Anthony Del M, Sabine P, Stephen Vander H, Malcolm Ross S, 2019. Mortality and cancer incidence among female Australian firefighters. Occup. Environ. Med. 10.1136/oemed-2018-105336. [DOI] [PubMed] [Google Scholar]
  6. Demers PA, DeMarini DM, Fent KW, Glass DC, Hansen J, Adetona O, Andersen MHG, Freeman LEB, Caban-Martinez AJ, Daniels RD, Driscoll TR, Goodrich JM, Graber JM, Kirkham TL, Kjaerheim K, Kriebel D, Long AS, Main LC, Oliveira M, Peters S, Teras LR, Watkins ER, Burgess JL, Stec AA, White PA, DeBono NL, Benbrahim-Tallaa L, de Conti A, El Ghissassi F, Grosse Y, Stayner LT, Suonio E, Viegas S, Wedekind R, Boucheron P, Hosseini B, Kim J, Zahed H, Mattock H, Madia F, Schubauer-Berigan MK, 2022. Carcinogenicity of occupational exposure as a firefighter. Lancet Oncol. 23 (8), 985–986. 10.1016/S1470-2045(22)00390-4. [DOI] [PubMed] [Google Scholar]
  7. Dongmug K, Letitia D, Phillip RH, David K, 2008. Cancer incidence among Male Massachusetts firefighters, 1987–2003. Am. J. Ind. Med 10.1002/ajim.20549. [DOI] [PubMed] [Google Scholar]
  8. Dubois-Deruy E, Peugnet V, Turkieh A, Pinet F, 2020. Oxidative stress in cardiovascular diseases. Antioxidants 9 (9), 864. 10.3390/antiox9090864. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Fangchao M, Lora EF, David JL, Edward JT, Terence AG, 2006. Cancer Incidence in Florida Professional Firefighters, 1981 to 1999. J. Occup. Environ. Med 10.1097/01.jom.0000235862.12518.04. [DOI] [PubMed] [Google Scholar]
  10. Furlong MA, Liu T, Snider JM, Tfaily MM, Itson C, Beitel S, Parsawar K, Keck K, Galligan J, Walker DI, Gulotta JJ, Burgess JL, 2023. Evaluating changes in firefighter urinary metabolomes after structural fires: an untargeted, high resolution approach. Sci. Rep. 13 (1), 20872. 10.1038/s41598-023-47799-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Grace KL, Ash G, Paul S, James AD, Tarek MS, Heriberto B-V, Kari D, James EL, 2006. Cancer risk among firefighters: a review and meta-analysis of 32 studies. J. Occup. Environ. Med. 10.1097/01.jom.0000246229.68697.90. [DOI] [PubMed] [Google Scholar]
  12. Hishikawa D, Hashidate T, Shimizu T, Shindou H, 2014. Diversity and function of membrane glycerophospholipids generated by the remodeling pathway in Mammalian cells. JLR (J. Lipid Res.) 55 (5), 799–807. 10.1194/jlr.r046094. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Jahnke SA-O, Poston WSC, Jitnarin N, Haddock CK. Maternal and Child Health Among Female Firefighters in the U.S(1573–6628 (Electronic)) [DOI] [PMC free article] [PubMed]
  14. Jitnarin N, Poston WSC, Haddock CK, Jahnke SA, 2019. Tobacco use among women firefighters. Womens Health Issues 29 (5), 432–439. 10.1016/j.whi.2019.05.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kanehisa M, 2000. Kegg: kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 28 (1), 27–30. 10.1093/nar/28.1.27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Keith TP, 2011. Iarc Monographs on the Evaluation of Carcinogenic Risks to Humans. Volume 98: Painting, Firefighting and Shiftwork. International Agency for Research on Cancer. Occupational Medicine. 10.1093/occmed/kqr127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kenneth WF, Judith E, John S, Deborah LS, Joachim DP, Matthew AS, Matthew AS, Charles M, Gavin PH, James D, 2014. Systemic exposure to pahs and benzene in firefighters suppressing controlled structure fires. Ann. Occup. Hyg. 10.1093/annhyg/meu036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Lee DJ, Koru-Sengul T, Hernandez MN, Caban-Martinez AJ, McClure LA, Mackinnon JA, Kobetz EN, 2020. Cancer risk among career Male and female Florida firefighters: evidence from the Florida firefighter cancer registry (1981–2014). Am. J. Ind. Med 63 (4), 285–299. 10.1002/ajim.23086. [DOI] [PubMed] [Google Scholar]
  19. Liu T, Furlong MA, Snider JM, Tfaily MM, Itson C, Beitel SC, Gulotta JJ, Parsawar K, Keck K, Galligan J, Walker DI, Goodrich JM, Burgess JL, 2025a. Differential metabolic profiles by Hispanic ethnicity among Male Tucson firefighters. Metabolomics 21 (2), 37. 10.1007/s11306-024-02198-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Liu T, Furlong MA, Snider JM, Beitel S, Mullins CE, Walker DI, Goodrich JM, Urwin DJ, Gabriel J, Hughes J, Gulotta JJ, Calkins MM, Liu Y, Von Hippel FA, Beamer P, Burgess JL, 2025b. Evaluating urinary metabolic profiles with wildland-urban-interface (Wui) fire exposure among Male firefighters: a comparison with municipal structure fires (Msf). Environ. Health 24 (1). 10.1186/s12940-025-01239-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Lowe F, 2014. Biomarkers of oxidative stress. In: Laher I (Ed.), Systems Biology of Free Radicals and Antioxidants. Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 65–87. [Google Scholar]
  22. Medina VA, Rivera ES, 2010. Histamine receptors and cancer pharmacology. Br. J. Pharmacol. 161 (4), 755–767. 10.1111/j.1476-5381.2010.00961.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Michael NB, 2007. Registry-based case-control study of cancer in California firefighters. Am. J. Ind. Med. 10.1002/ajim.20446. [DOI] [PubMed] [Google Scholar]
  24. Mitchell CL, Hollister J, Fisher JM, Beitel SC, Ramadan F, O’Leary S, Fan ZT, Lutrick K, Burgess JL, Ellingson KD, 2025. Differences in serum concentrations of per-and Polyfluoroalkyl substances by occupation among firefighters, other first responders, healthcare workers, and other essential workers in Arizona, 2020–2023. J. Expo. Sci. Environ. Epidemiol 10.1038/s41370-025-00753-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Olorunfemi A, Jim Z, Jim Z, Daniel BH, Jia-Sheng W, John EV, Luke PN, 2013. Occupational exposure to woodsmoke and oxidative stress in wildland firefighters. Sci. Total Environ. 10.1016/j.scitotenv.2013.01.075. [DOI] [PubMed] [Google Scholar]
  26. Pang Z, Lu Y, Zhou G, Hui F, Xu L, Viau C, Spigelman Aliya F, MacDonald Patrick E., Wishart David S., Li S, Xia J, 2024. Metaboanalyst 6.0: towards a unified platform for metabolomics data processing, analysis and interpretation. Nucleic Acids Res. 10.1093/nar/gkae253 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. gkae253. Paul AD, Paul AD, Harvey C, Thomas LV, Noel SW, Nicholas JH, Linda R, 1994. Cancer Incidence among firefighters in Seattle and Tacoma, Washington (United States). Cancer Causes Control. 10.1007/bf01830258. [DOI] [PubMed] [Google Scholar]
  28. R Core Team, 2023. R: a Language and Environment for Statistical Computing.
  29. Rebecca JT, Sara EL, Pam S, Rosemary DC, Dennis MD, Geoffrey MC, 2015. Risk of cancer among Firefighters in California, 1988–2007. Am. J. Ind. Med 10.1002/ajim.22466. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Reddy VP, 2023. Oxidative stress in health and disease. Biomedicines 11 (11), 2925. 10.3390/biomedicines11112925. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Reuter S, Gupta SC, Chaturvedi MM, Aggarwal BB, 2010. Oxidative stress, inflammation, and cancer: how are they linked? Free Radic. Biol. Med. 49 (11), 1603–1616. 10.1016/j.freeradbiomed.2010.09.006. Epub 20100916 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Robert DD, Travis LK, James HY, Matthew MD, Thomas H, Dalsu B, Shelia Hoar Z, Shelia Hoar Z, James JB, Kathleen MW, Lynne EP, 2014. Mortality and cancer incidence in a pooled cohort of Us firefighters from San Francisco, Chicago and Philadelphia (1950–2009). Occup. Environ. Med. 10.1136/oemed-2013-101662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Schick SF, Blount BC, Jacob P, Saliba NA, Bernert JT, El Hellani A, Jatlow P, Pappas RS, Wang L, Foulds J, Ghosh A, Hecht SS, Gomez JC, Martin JR, Mesaros C, Srivastava S, St Helen G, Tarran R, Lorkiewicz PK, Blair IA, Kimmel HL, Doerschuk CM, Benowitz NL, Bhatnagar A, 2017. Biomarkers of exposure to new and emerging tobacco delivery products. Am. J. Physiol. Lung Cell. Mol. Physiol. 313 (3), L425–L452. 10.1152/ajplung.00343.2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Singh N, Baby D, Rajguru JP, Patil PB, Thakkannavar SS, Pujari VB, 2019. Inflammation and cancer. Ann. Afr. Med 18 (3). [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Wang X-N, Wang K-Y, Zhang X-S, Yang C, Li X-Y, 2018. 4-Hydroxybenzoic acid (4-Hba) enhances the sensitivity of human breast cancer cells to Adriamycin as a specific Hdac6 inhibitor by promoting Hipk2/P53 pathway. Biochem. Biophys. Res. Commun. 504 (4), 812–819. 10.1016/j.bbrc.2018.08.043. [DOI] [PubMed] [Google Scholar]
  36. Williams KM, Franzi LM, Last JA, 2013. Cell-Specific oxidative stress and cytotoxicity after wildfire coarse particulate matter instillation into mouse lung. Toxicol. Appl. Pharmacol. 266 (1), 48–55. 10.1016/j.taap.2012.10.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Xia J, Psychogios N, Young N, Wishart DS, 2009. Metaboanalyst: a web server for metabolomic data analysis and interpretation. Nucleic Acids Res. 37 (Web Server issue), W652–W. 10.1093/NAR/GKP356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Zhang Y, Gong J, Hu X, He L, Lin Y, Zhang J, Meng X, Zhang Y, Mo J, Day DB, Xiang J, 2024. Glycerophospholipid metabolism changes Association with ozone exposure. J. Hazard Mater. 475, 134870. 10.1016/j.jhazmat.2024.134870. [DOI] [PubMed] [Google Scholar]
  39. Zhu Q, Wu Y, Mai J, Guo G, Meng J, Fang X, Chen X, Liu C, Zhong S, 2022. Comprehensive metabolic profiling of inflammation indicated key roles of glycerophospholipid and arginine metabolism in coronary artery disease. Front. Immunol. 13. 10.3389/fimmu.2022.829425. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

MMC1
NIHMS2147059-supplement-MMC1.xlsx (306.6KB, xlsx)

RESOURCES