The recruitment of study participants was a joint effort involving local and regional fire departments in Oklahoma, along with the Oklahoma State University Fire Protection & Safety Engineering Technology program. In addition, we attended routine meetings with volunteer fire chiefs to introduce our research and recruit firefighters. Forty-seven firefighters (28 volunteers and 19 careers) were recruited from 11 fire departments in Oklahoma. All participants gave written informed consent after a detailed explanation was provided about the benefits and risks involved with the investigation. Institutional Review Board approval was obtained for this study. Participants who met the following inclusion criteria were eligible for participation: (1) at least 18 years of age, (2) a volunteer or career firefighter in Oklahoma, and (3) a minimum of one year of firefighting experience. Each participant completed a self-administered questionnaire at baseline, in which they were asked to provide details including demographic information, PPE practices, and firefighting-related practices.

We visited each participating fire department and conducted baseline samplings of all study participants. Prior to baseline sampling, none of the participants had engaged in firefighting activities for a period of seven consecutive days. This approach helped to establish a standardized baseline for each participant. Participants also completed a questionnaire that provided additional information about their PAH exposure. For the post-firefighting samples, we visited each fire department within 24 h of the fire, but before the gear was washed, to collect PAH and blood samples.

PAHs deposited on the surface of each firefighter's turnout gear and skin were collected using a 10 cm × 10 cm polyester fabric wipe (Alpha Wipes, Texwipe, Kernersville, NC) saturated with 1 ml of 99% isopropyl alcohol. We applied the wipe with even firm pressure in a consistent area on both the gear and skin. The selection of sampling sites was based on the existing literature, which highlights the neck as the primary site of dermal exposure among firefighters, as PPE hoods offer limited protection (19, 20). Additionally, the wrist of firefighters have high levels of PAH exposure (21). Gear wipes were taken from the right front sleeve and back neck area on the outside surface of the jacket (Supplementary Figure S1). Skin wipes were collected from each firefighter's right forearm and back of the neck (Supplementary Figure S2).

In addition to the wipe samples, a phlebotomist drew 3 ml of blood from each firefighter, storing the samples in EDTA-containing tubes (BD Vacutainer, Franklin Lakes, NJ) to avoid coagulation. The collected blood samples were analyzed at the University of Oklahoma Medical Center laboratory, while the collected wipe samples were analyzed on a Varian 450 Gas Chromatography (GC) equipped with a 200 Mass Spectrometry (MS) detector (Agilent Technologies, Santa Clara, CA), as described previously (22).

For the wipe samples, we excluded a PAH analyte from the analysis if its concentration fell below the limit of detection (LOD) in more than 70% of the samples. For PAHs with concentrations below the LOD, but in less than 70% of the samples, we addressed the issue of missing data by employing regression imputation to estimate each missing value. We employed the IARC group classifications to predict missing values for PAHs detected on gear. For example, phenanthrene exhibited no missing data, while pyrene had 15% missing data. Leveraging the available data from phenanthrene, we imputed the missing values for pyrene using firefighter groups and fire departments as independent variables in pre- and post-firefighting scenarios, respectively. We applied the same approach to impute the missing values for the remaining PAHs.

We compared baseline demographic characteristics between volunteer and career firefighters using Student's t-tests or Mann-Whitney U tests for continuous variables and Pearson Chi-squared tests or Fisher's exact tests for categorical variables. For the analysis of PAHs and CBC parameters, we initially assessed the normality of the data. Based on the data's normality, we selected either parametric or non-parametric statistical methods for data analysis. To compare differences before and after firefighting, we employed paired sample t-tests or Wilcoxon signed-rank tests. For comparing intergroup differences between volunteer and career firefighters, we utilized Student's t-tests or Mann-Whitney U tests.

For the blood samples, a heatmap was created to analyze correlations between 10 PAHs on gear and 16 CBC parameters using the “pheatmap” package (version 1.0.12) in R (version 4.3.0). Spearman correlation coefficients were calculated to measure correlations. The heatmap used a color gradient (red for positive, blue for negative) to depict these correlations, with asterisks indicating statistical significance (p < 0.05). Generalized estimating equations (GEE) models, which can account for the correlation between repeated measurements made on the same subject over time, were used to estimate the association between the CBC parameters and the ∑PAHs deposited on gear or skin. ∑PAHs refers to the sum of the individual PAH concentrations measured separately on the skin or gear samples for each firefighter. Note that in the ∑PAH on skin model, none of the variables exhibited statistical significance. CBC parameters were log-transformed and GEE models were fitted for each CBC parameter with the covariates, which included age, career/volunteer firefighters, pre/post firefighting, BMI, smoking status, years served as a firefighter, and hours per month on a fire scene. Covariates were selected into the models in a forward approach, and any covariate that had a significance level of p < 0.05 was retained. Then, the final models were obtained by adding ∑PAHs to these selected models. The quasi-likelihood information criterion (QIC) was used to compare the final model with the null model, and the model with the smaller QIC was the best working model. The correlations between each type of blood cell from the CBC and each individual PAH were evaluated by the Spearman rank correlation coefficient. Statistical significance was taken at 2-sided P < 0.05. Statistical analyses were performed using SAS 9.4 (SAS Institute Inc., Cary, NC).

In this study, 28 of the firefighters were volunteers, 19 were career, and a majority (94.7%) of the participants were male (Table 1). The average number of years served as a firefighter was 11.6 for volunteers and 10.8 for careers (p = 0.64). All participating firefighters reported engaging in fire-related activities an average of four times per month. However, participation in non-fire-related activities varied significantly between the volunteer and career groups, with volunteers engaging in an average of 1.5 non-fire-related activities per week and career firefighters engaging in an average of eight per week (p < 0.05). Most of the volunteer firefighters did not have a typical work shift; rather, they were always on call (p < 0.05). Regarding health status, body mass index (BMI) showed notable differences between career and volunteer firefighters (p < 0.05). Among career firefighters, 36.84% were overweight and 26.32% were obese category. In contrast, among volunteer firefighters, 21.43% were overweight and a higher percentage, 64.29%, were obese category. Volunteer firefighters reported higher rates of respiratory diseases compared to career firefighters (Supplementary Table S1). Specifically, 14.29% reported asthma, 7.14% reported chronic bronchitis, 10.71% reported coughing, and 3.57% reported seasonal allergies.

Tables 2, 3 present the results of the quantitative analysis of PAHs on the gear and skin, respectively, of firefighters before and after firefighting activities. One post-test gear sample from a career firefighter was not available due to issues with the laboratory analytical extraction. The majority of the studied PAHs showed increased concentration on gear after firefighting, with significant increases observed in acenaphthylene, fluoranthene, fluorene, 1-methyl-naphthalene, 2-methyl-naphthalene, and pyrene. However, there were no significant differences in any PAH levels on the gear between the two groups of firefighters, volunteer and career. Regarding PAH deposition on the skin, we found a significant difference in fluorene after firefighting, while acenaphthene was significantly higher on the skin of volunteer firefighters than on career firefighters.

Table 4 compares the six common PAHs found on the firefighters' gear and skin. At baseline, the gear had significantly lower levels of acenaphthene, fluorene, phenanthrene, and pyrene compared to skin. 2-methyl-naphthalene was significantly higher on gear vs. skin at baseline. After firefighting, the gear continued to have lower levels of acenaphthene, fluorene, and phenanthrene vs. skin (p < 0.05). Again, 2-methyl-naphthalene was significantly higher on gear after firefighting compared to skin. There was no significant difference in pyrene levels between gear and skin after firefighting.

Table 5 displays the PAH levels on gear and skin before and after firefighting for both career and volunteer firefighters. For total PAH on gear, there was a statistically significant difference for career firefighters, with an increase from 1.87 ng/cm² to 4.56 ng/cm², while the increase for volunteers from 2.80 ng/cm² to 3.76 ng/cm² was not statistically significant. Regarding total PAH on skin, the amount on volunteers increased slightly after firefighting, whereas career firefighters showed a decrease from 2.05 ng/cm² to 1.45 ng/cm² (p < 0.05).

Table 6 illustrates the response of CBC parameters to firefighting activities for both volunteer and career firefighters. Significant differences were observed in career firefighters for mean corpuscular volume (MCV), mean corpuscular hemoglobin concentration (MCHC), red cell distribution width (RDW), white blood cell (WBC), granulocyte (GRAN) count, and lymphocyte (LYMPH) count (p < 0.05) after firefighting compared to baseline, whereas no significant changes were observed in any CBC parameters for volunteer firefighters.

Figure 1 presents the CBC parameters measured for each group at baseline and after firefighting. At baseline, there were significant differences between volunteer firefighters and career firefighters in MCHC, RDW, WBC, LYMPH count, GRAN count, mononucleosis (MONO) count, and immature granulocytes (IG) count (p < 0.05). Among these, volunteer firefighters had higher levels of all significantly different indicators except for MCHC. However, no significant group difference in CBC parameters was detected after firefighting.

Table 7 shows the results regarding generalized estimating equations (GEE) modeling between each CBC parameter and the ∑PAHs deposited on gear. Among all firefighters, a 10-unit (ng/cm²) increase in ∑PAHs was associated with a 0.013 (95% CI: 0.002–0.024; p = 0.0155) and 0.028 (95% CI: 0.000–0.057; p = 0.0470) increase in log(MCH) and log(RDW), respectively. A 10-unit increase in ∑PAHs was associated with a 0.096 (95% CI: 0.012–0.180; p = 0.0248) and 0.285 (95% CI: 0.109–0.462; p = 0.0015) increase in log(WBC) and log(IG), respectively. This data suggests that increased ∑PAHs deposition on firefighters’ gear was associated with alterations in hematological and immune parameters. While the changes in MCH and RDW were relatively modest, the notable increases in WBC and IG levels warrant further attention and investigation, as they might be indicative of a significant immune response or inflammation due to PAH exposure.

We employed heatmap to illustrate the correlations between CBC parameters and individual PAH (Figure 2). Despite the use of multiple testing corrections, our limited sample size had low statistical power to detect certain correlations. Therefore, we present the unadjusted correlation results, with a warning against potential Type I errors. Among the PAHs deposited on gear, we found that the concentration of acenaphthylene, a PAH that increased in firefighters after firefighting, was positively correlated with WBC, GRAN count, LYMPH count, MONO count, and IG count (p < 0.05). Phenanthrene was positively correlated with PLT, LYMPH count, and MONO count. Fluorene was found to be positively correlated with RBC and HCT (p < 0.05). Notably, 1-methyl-naphthalene was positively correlated with RDW (p < 0.05) while negatively correlated with MCH (p < 0.05). However, for PAHs deposited on the skin, we found a different correlation pattern with CBC parameters (Supplementary Figure S3). Acenaphthene was positively correlated with RBC, RDW, WBC, and GRAN count (p < 0.05). Phenanthrene and pyrene were positively correlated with MCHC (p < 0.05), while negatively correlated with MPV, although the latter correlation was not statistically significant.

MCHC is a measure of the concentration of hemoglobin in cells. A low concentration of hemoglobin within a given volume of red blood cells results in a reduced capacity to carry oxygen to the tissues and affects the cardiovascular system through oxygen supply (23). In our study, we observed a decrease in MCHC levels in both career and volunteer firefighters following firefighting activities. Notably, prior to firefighting, volunteer firefighters exhibited a significantly lower MCHC compared to career firefighters, suggesting potential pre-existing hematological differences that could have implications for their overall health for firefighting duties. Moreover, we identified a negative correlation between MCHC levels and acenaphthene concentrations in the gear wipe samples (p < 0.05). Acenaphthene, a type of PAH, albeit present in low concentrations, may still play a role in reducing MCHC levels in firefighters. This area warrants further investigation, especially considering that acenaphthene is known to follow various metabolic pathways, including oxidation to acenaphthenequinone (24). On the other hand, MCHC levels have been shown to be affected in various hematological conditions (25, 26). Understanding the specific mechanisms through which PAH affects hematological parameters like MCHC could have significant implications for firefighter health.

In our research, RDW was significantly higher in volunteer firefighters after firefighting, while there was no significant difference in career firefighters. Elevated RDW levels have been associated with various cardiovascular conditions (27) and increased risk of cancer mortality (28). A comprehensive meta-analysis has underscored RDW's potential as a prognostic marker in cancer patients, linking high RDW levels to poor outcomes (29). The concentration of 1-methyl-naphthalene on gear had a positive correlation with RDW (p < 0.05). Our wipe sample results showed that the concentration of 1-methyl-naphthalene increased after firefighting compared to the baseline for both career and volunteer firefighters. While 1-methyl-naphthalene is a derivative of naphthalene, which has been studied for its potential chronic effects (30), its specific impact on RDW levels remains unclear. The observed correlation between 1-methyl-naphthalene and RDW may also be helpful in estimating heart disease or cancer risk (14, 31, 32) and warrants further investigation. One study revealed a linear association between RDW and the risk of myocardial infarction (MI), indicating that individuals with a higher RDW were more likely to experience MI, with each 1% increase in RDW associated with a 13% elevated risk (33).

WBC is associated with infection and inflammation in the body. This association highlights the significance of regularly monitoring WBC levels as a valuable marker for overall health assessment and the early detection of potential health risks. A previous study reported a substantial increase in WBC counts, rising from 4.94 K/mm³ to 9.15 K/mm³ immediately after firefighting activities and subsequently stabilizing at 8.12 K/mm³ after 90 min, as serving as markers of systemic inflammation and physiological stress (34). In our study, we observed an increase in WBC counts among career firefighters, rising from 5.46 K/mm³ to 6.33 K/mm³, a statistically significant finding. In contrast, volunteer firefighters experienced a more modest increase, going from 8.07 K/mm³ to 8.21 K/mm³. It is noteworthy that volunteer firefighters exhibited significantly higher baseline WBC counts compared to career firefighters.

These findings collectively contribute to our understanding of the dynamic changes in leukocyte parameters between career and volunteer firefighters in their response to firefighting-related stressors. Positive correlation coefficients were observed between the concentration of acenaphthylene on gear and WBC, GRAN count, LYMPH count, and IG count in firefighters. High LYMPH and GRAN counts could indicate several issues, including infection, as elevated leukocyte counts have been associated with possible infections (35); blood cell cancer, as lymph node counts have been linked to the stage of illness and patient outcome in cancers like pancreatic ductal adenocarcinoma (36); or some type of autoimmune disease, as lymph node alterations have been observed in conditions like chronic kidney disease (37). The increase in leukocytes in firefighters after firefighting may be associated with an elevated level of 1-methyl-naphthalene. The lymphocytes have been proposed as an inflammatory biomarker potential predictor of risk and prognosis in CVD (38). Therefore, 1-methyl-naphthalene exposure may increase the risk of CVD, cancer, and other diseases in firefighters, warranting further investigation.

Our results from the GEE model indicate that the elevated MCH and RDW levels could signify underlying hematological changes that may have long-term health implications for firefighters. A modest 10-unit increase in the sum of PAHs that led to these significant changes is a cause for concern, particularly given the cumulative nature of occupational hazards faced by firefighters. Even low-level chronic exposure to toxic chemicals may result in adverse health outcomes, emphasizing the need for ongoing vigilance. Earlier research has indicated that changes in erythrocyte physiology, such as alterations in MCH, could serve as early markers for systemic issues like anemia or other blood disorders. For instance, the erythrocyte zinc-protoporphyrin/heme (ZPP/H) ratio has been identified as a valid point-of-care biomarker to diagnose iron deficiency anemia (39). Additionally, erythrocyte destruction has been noted as a major factor in the early stages of anemia or inflammation (40). As mentioned earlier, elevated RDW has been identified as an independent prognostic marker for systemic inflammatory responses (41) and specific cardiovascular conditions (42). A stronger increased in sum of PAHs was associated with an elevated WBC and IG. Elevated WBC and IG levels have been associated with an increased risk for infection (43). Such elevations in WBC and IG may not only signify an ongoing inflammatory process but could also be indicative of compromised immune functioning over time.

Most PAHs on gear, including acenaphthylene, fluoranthene, fluorene, 1-methyl-naphthalene, 2-methyl-naphthalene, and pyrene, exhibited significant increases after firefighting, potentially attributable to variations in firefighters' turnout gear practices. While turnout gear is designed to serve as a critical barrier against immediate hazards, our study adds to the existing literature by revealing that it may not provide comprehensive protection against the dermal absorption of PAHs. Specifically, we found compelling evidence of significant disparities in PAH levels between firefighter gear and skin, both at baseline and following firefighting activities. Previous research has shown that PAHs most likely enter firefighters' bodies through their skin, with the neck being a primary site of exposure due to lower dermal protection (20). Moreover, dermal absorption has been identified as a significant contributor to biological levels of PAH metabolites in firefighters (44). The turnout gear, which is treated with per- and polyfluoroalkyl substances (PFAS) chemicals for water and oil resistance, undergoes a shedding process in which PFAS migrate from the outer shell to the thermal liner, the layer in direct contact with the skin (45). This migration could potentially compromise the gear's protective barrier, facilitating the absorption of PAHs through the skin. As PFAS degradation occurs, the chemicals could be absorbed through skin contact with the thermal liner, thereby creating a pathway for PAHs to penetrate as well. Also, the composition of turnout gear has also been found to affect moisture and thermal management, which could influence the protection offered to firefighters (46, 47). Therefore, while turnout gear is designed to protect firefighters, its material properties and degradation over time may inadvertently facilitate the skin absorption of PAHs. Given these findings, our results underscore the need for further research to elucidate the mechanisms leading to differential PAH absorption on gear and skin, as well as to optimize the design and material composition of turnout gear to minimize such risks.

Our study investigating the exposure of firefighters to PAHs has provided valuable insights into the potential health risks associated with firefighting activities. However, it is important to acknowledge several limitations that should be considered when interpreting the findings. Firstly, the study design involved collecting baseline samples from firefighters who had not engaged in firefighting activities for seven consecutive days. While this approach aimed to establish a standardized baseline, it may not fully capture the day-to-day variations in PAH exposure that firefighters experience during their active duty. Additionally, the post-firefighting sampling was conducted within 24 h before the gear was washed. This timing constraint, coupled with the need to coordinate with different fire departments, resulted in multiple visits and potential variations in the post-firefighting data collection process. Secondly, the sample size in this study was relatively small. While efforts were made to recruit a diverse group of participants, the limited sample size may impact the generalizability of the results to a broader population of firefighters, especially considering the potential variations in firefighting practices and PAH exposure levels across different regions. Although PAHs are the primary combustion byproducts in fire smoke, their presence in ambient air is widespread. Our detection of PAHs was unable to disentangle the effects of other confounding factors. Thirdly, the study used self-administered questionnaires to collect information on demographic information, PPE practices, and firefighting-related practices. Self-reported data can be subject to recall bias and may not always accurately reflect participants' actual behaviors and exposures. Finally, the study primarily focused on the association between PAH exposure and CBC parameters, which provide important insights into potential health effects. However, other relevant health indicators or long-term health outcomes were not assessed, limiting the comprehensive understanding of the health implications of PAH exposure among firefighters.