Characteristics of Participants During the First Three Years of the National Firefighter Registry for Cancer, United States, January 2023–December 2025

ABSTRACT

Background

Firefighting is a known human carcinogen, but knowledge gaps remain regarding cancer burden and risk factors among U.S. firefighters. The Firefighter Cancer Registry Act of 2018 directed the development of the National Firefighter Registry (NFR) for Cancer, a voluntary exposure registry managed by the National Institute for Occupational Safety and Health. This manuscript describes participation during the first three years of NFR enrollment (January 2023–December 2025).

Methods

We descriptively calculated frequencies and percentages of participants across demographic and occupational characteristics and stratified by enrollment completion status, work status, and method of recruitment.

Results

There were 39,058 firefighters from all 50 states and several territories who participated in the NFR, including 4019 reporting a history of cancer. Participants were predominantly male (91%), White (83%), actively serving full time (75%), and reported various firefighting assignments. Overall, 72% of participants submitted the enrollment questionnaire, 24% did not submit the questionnaire but provided enough information to include in state cancer registry linkages, and 4% had incomplete enrollment. The largest attrition occurred at the work history section of the questionnaire, but completion rates were generally similar by demographic and occupational characteristics. A department roster‐based recruitment method resulted in a slightly lower self‐reported cancer prevalence through exploratory analyses, suggesting possible reduction of selection bias by cancer status.

Conclusions

Findings demonstrate national engagement and high completion rates in a large, voluntary occupational registry. Continued recruitment, particularly among volunteer firefighters and other underrepresented groups, will strengthen the NFR's utility for analyzing cancer and cancer risk factors among U.S. firefighters.

1. Background

There are over one million firefighters working in the United States (U.S.) [1]. Through their work on and off the fireground, firefighters are exposed to polycyclic aromatic hydrocarbons, volatile organic compounds, and hundreds of other chemicals released from burning materials [2, 3, 4, 5]. Many of these substances are known or probable human carcinogens, such as asbestos, benzene, benzo[a]pyrene, formaldehyde, 1,3‐butadiene, styrene, and acrolein [5, 6]. Large epidemiological studies suggest that firefighters in the U.S. have an increased risk for both cancer incidence and mortality [7, 8, 9, 10]. A rapid increase in studies on cancer in firefighters in the U.S. and abroad, which include epidemiological, mechanistic, and exposure studies, warranted a re‐evaluation of the carcinogenicity of firefighting by the International Agency for Research on Cancer (IARC) in 2022 [11],which previously classified the occupation as possibly carcinogenic to humans in 2010 [12] The agency concluded that the occupation of firefighting is a human carcinogen (Group 1 classification), with sufficient evidence to support causation for mesothelioma and bladder cancer; and additional limited (i.e., emerging) evidence to support causation with melanoma, non‐Hodgkin lymphoma, and cancers of the colon, prostate, and testes [5, 13].

Many knowledge gaps related to the burden of cancer and cancer‐related risk factors among U.S. firefighters remain. Much of the research was conducted using data from decades ago, which is often necessary given the latency of cancer; but this complicates our understanding of risk because of changes in workplace practices, occupational exposures, and workforce demographics. Additionally, cancers that appear associated with firefighting seem to vary by geography and demographic characteristics like sex and race/ethnicity [7, 14, 15, 16, 17, 18]. However, analyses of sub‐groups like women and non‐White firefighters have been underpowered because study samples have consisted predominantly of White male populations. Additional research could shed light on how cancer, including rarer cancers, affects a current and diverse workforce.

Furthermore, research suggests that workplace activities, practices, and exposures vary between and within firefighting specializations. [3, 19, 20, 21]. But few epidemiological studies have evaluated cancer outcomes among groups other than career structural firefighters [18, 22, 23, 24, 25, 26], with most on non‐U.S. populations. Additional information on cancer rates and risk factors for volunteer firefighters—who make up about 65% of the U.S. firefighting workforce [1]—wildland firefighters, and other members of the fire service may benefit this field of research.

Lastly, while much of the past research has evaluated cancer related to occupational status as a firefighter, there remains a lot to be learned about cancer risk factors within the occupation. A better understanding of the prevalence and impacts of exposures (e.g., incident types and frequency of responses), workplace characteristics and practices (e.g., use of personal protective equipment, decontamination practices, shift work), and other cancer risk factors (e.g., alcohol and smokeless tobacco use) could inform the development and evaluation of practices and recommendations that aim to reduce cancer in the fire service.

The primary challenge to conducting epidemiological research on cancer among firefighters in the U.S. is the lack of a central and comprehensive source of data that provides details about firefighters' occupation, other cancer‐related risk factors, and cancer status on a national level. Many of the recent large studies have been records‐based, relying on data sources such as cancer registry, fire incident, or employment/union records [7, 8, 14, 16, 27]. However, these secondary data sources do not always provide accurate or comprehensive work or exposure history, let alone information about lifestyle and covariate exposures [28, 29, 30, 31, 32]. Epidemiologic studies that collect data using surveys designed for general population health research also lack comprehensive firefighting information [9]. Several important, ongoing, cohort studies specific to U.S. firefighters collect survey data, but are somewhat limited in geographic representation or sample size due to the scope of their objectives (e.g., collecting biological materials for biomonitoring) [33, 34].

In light of knowledge gaps and inadequate data sources for evaluating firefighter cancer, the U.S. passed the Firefighter Cancer Registry Act of 2018 (i.e., the Act) (and later the Firefighter Cancer Registry Reauthorization Act of 2023) [35, 36]. The Act directs the Department of Health and Human Services (HHS) to “develop and maintain… a voluntary registry of firefighters to collect relevant health and occupational information of such firefighters for purposes of determining cancer incidence.” HHS delegated authority to the Centers for Disease Control and Prevention (CDC)'s National Institute for Occupational Safety and Health (NIOSH) in developing this new public health surveillance system, which is called the National Firefighter Registry (NFR) for Cancer. The NFR is collecting and maintaining comprehensive information about participants' demographics, work history, exposures, workplace practices, lifestyle, and health to monitor cancer and evaluate cancer risk factors among U.S. firefighters.

Enrollment of NFR participants began in January 2023. Recruitment has since involved dissemination of communication material and social media, presence at fire service conferences, department roster‐based outreach, and other channels. The objective of this manuscript is to describe participation in the NFR during the first three years of enrollment, including participant characteristics by enrollment completion, active employment status, and method of recruitment. These analyses are useful for evaluating the initial success of the NFR program, identifying areas for improvement, and could be beneficial in designing other voluntary, occupational exposure and health registries.

2. Materials and Methods

2.1. Design

The NFR is a continuous, voluntary exposure and risk factor surveillance system of firefighters, designed to be linked with health outcome surveillance systems, including population‐based cancer registries and the National Death Index, prospectively over time. The NFR is generating detailed knowledge needed by fire service and public health communities to inform cancer prevention activities, with a long‐term goal of reducing cancer in firefighters. Participants enroll electronically through a web portal that includes, in order, informed consent, a user profile, and an enrollment questionnaire. The web portal will also accommodate voluntary follow‐up questionnaires for participants to complete longitudinally. The primary data collection activities of the NFR described in the scientific protocol [37] include:

• –Collection of self‐reported information from firefighters on employment/workplace characteristics, exposure, demographics, lifestyle factors, co‐morbidities, and other cofactors related to cancer.
• –Obtaining records from fire departments or agencies to track trends and patterns of exposure as it relates to cancer in firefighters.
• –Monitoring cancer in firefighters by linking with health outcome information databases to assess cancer incidence and mortality.

NIOSH will use data from the NFR to investigate many topics related to cancer and cancer risk factors among firefighters, such as the burden of specific cancer types; variation in cancer rates among different groups of firefighters (e.g., demographic, geographic, and subspecialty groups); and patterns in exposure, workplace practices/characteristics, and health behaviors and their associations with cancer rates. Data collected in the NFR web portal are protected by an Assurance of Confidentiality and stored and shared in a secure and confidential manner that protects the privacy of participants. This activity was reviewed by CDC, deemed not research, and was conducted consistent with applicable federal law and CDC policy. ¹ Nonetheless, to emphasize and ensure that participation is voluntary, which is a stipulation of the Act, all NFR participants undergo informed consent in the web portal prior to enrolling.

2.2. Participants

All current and former U.S. firefighters at least 18 years of age are eligible to participate in the NFR, regardless of their health or cancer status. Multiple recruitment methods are used to enroll firefighters into the NFR, but generally fall into two approaches or recruitment routes: roster‐based sampling and Open Enrollment. Special attention is being given to ensure participation from large numbers of women, non‐White, and volunteer firefighters, so that these understudied groups can be analyzed as individual sub‐populations to the extent possible, as instructed by the Act and outlined in the scientific protocol. NIOSH is also encouraging participation from subspecialty groups of firefighters including but not limited to wildland firefighters, fire investigators, and fire instructors.

Roster‐based sampling of active firefighters is employed to encourage participation among groups of firefighters specified in the Act, promote participation among firefighters that otherwise might be less likely to participate, and improve generalizability of analytic findings. Roster‐based sampling is being carried out through the Targeted Enrollment and Gold Helmet Department programs.

Targeted Enrollment involves recruiting from municipal structural fire departments (career, volunteer, and combination) selected by NIOSH based on geographic, demographic, or other workforce characteristics, ascertained through national data or stakeholder discussions, to increase representation of understudied groups. All active fire personnel from selected departments are individually invited by NIOSH via fire department leadership to participate in the NFR according to a standard procedure. During the first three years, Targeted Enrollment departments were selected based on a large female workforce, identified using information from the National Fire Protection Association or department leadership reporting.

The Gold Helmet Department program is an outreach campaign that recognizes fire departments that achieve NFR participation of 50% or more of their active firefighters (or more than 300 active firefighters) [38]. It was developed to encourage fire departments to actively promote participation among their own members by tying public recognition to a collective departmental goal. Public recognition of qualifying departments then generates social proof across the broader fire service, motivating other departments to champion participation from within their own rosters. Any fire department can become a Gold Helmet Department. Department leadership interested in this program contact NIOSH (by emailing NFRegistry@cdc. gov) and self‐report their active roster size and conduct their own recruitment.

Open Enrollment involves convenience sampling, whereby firefighters self‐identify as eligible, and is not otherwise carried out through recruitment from a department roster. Open Enrollment is employed to capture a larger and more inclusive sample (e.g., former firefighters and firefighters not affiliated with structural municipal fire departments are eligible) that aids in hypothesis generation such as identifying cancer risk factors. Initial Open Enrollment recruitment activities included a soft rollout campaign with select fire departments or at select events that occurred prior to national launch of the NFR. Following national launch (with a newsletter and large conference announcement in April 2023), recruitment activities included word‐of‐mouth communication; wide dissemination of communication products (e.g., social media posts, fact sheets, newsletters, podcasts); and exhibits or presentations at fire service conferences and fire departments.

Regardless of the recruitment route, the enrollment process is the same for all individuals. For the purposes of this analysis, individuals are considered NFR participants if they completed informed consent; and enrollment generally refers to participants' interaction with the web portal (e.g., participants were characterized by varying levels of enrollment completion based on their progress through the enrollment web portal). NFR participants were included in the current analysis if they enrolled between January 17, 2023 (the date of first enrollment) and December 31, 2025.

2.3. Analysis

We used participant‐reported information from the NFR enrollment portal—including from the informed consent agreement, user profile, and enrollment questionnaire—to describe the characteristics of participants who enrolled during the first three years of operation (January 17, 2023–December 31, 2025).

First, we evaluated the distribution of participation in the NFR across the U.S. Participants were assigned to a state based on their reported residential state at the time of enrollment. Total participant frequencies were calculated for each state and divided into quartiles for geographic mapping. Additionally, the frequency of participants for territories were calculated, but not included in the calculation of quartiles because they were presented separately from the map.

Second, we examined retention throughout the enrollment process to identify steps that could be barriers to completion. We calculated the number and percent of participants that were retained at each of several sequential sections of the enrollment process including the completion of informed consent, initiation of the user profile, initiation of selected sections of the questionnaire (i.e., demographics, work [and exposure] history, and health history) and questionnaire submission. Initiation was defined as having responded to at least one question within that section.

Lastly, we conducted descriptive bivariate analyses in which we calculated and compared frequencies and percentages of firefighters enrolled in the NFR by various participation‐, work‐, and demographic‐related characteristics overall and stratified by three variables of interest: completion status, work status, and recruitment route. For all three analyses, we present column percentages to assess differences in characteristics by categories of each stratification variable. For the analysis of completion status, we additionally present row percentages to assess whether completion status varied for different categories of our descriptive variables (i.e., differences in completion status vs. differences by completion status). Missing values were excluded from percentage calculations and results were not shown where nearly all observations were missing for a variable in a given stratum. Interpretations of results were based on the size of differences in the descriptive results (i.e., differences in percentages), but noted differences were confirmed via post hoc tests (chi‐square or Fisher's exact) with an α of 0.05 (not shown).

Completion status represents participants' progress toward completing the entire enrollment process, in sequential order, through questionnaire submission, as of December 31, 2025. Completion status is defined by three categories: incomplete, in‐progress, and completed. Participants were classified in the incomplete group if they consented to participate (requiring name and email address) but discontinued the process at either sex or date of birth collection (i.e., either were left missing), which are mandatory fields collected sequentially in the user profile and variables needed for pursuing cancer registry linkages for prospective follow‐up. Those in the in‐progress group provided at least their full name, email, sex, and date of birth but did not complete enrollment questionnaire submission (i.e., discontinued at any point of the user profile or questionnaire following date of birth collection). Lastly, firefighters were placed in the completed group if they submitted the enrollment questionnaire (i.e., selected “submit”). Participants in the in‐progress and completed groups are eligible to include in prospective cancer registry linkages.

Work status indicates participants' current working status in the fire service. Participants can select multiple options in responding to the user profile question “What is your current work status in the fire service (select all that apply)?” For comparison purposes in the current analysis, participants were hierarchically assigned to mutually‐exclusive work status groups using a categorical variable with four levels ([i] active, full‐time service; [ii] active, volunteer or part‐time service; [iii] active, other service; and [iv] former service) and a dichotomous variable with two groups for stratified analysis ([A] active service and [B] former service) (Table SI). Participants that selected, at a minimum, “full‐time paid” were categorized as [1] active, full‐time service; this selection superseded all other options. Participants were considered [2] active, volunteer or part‐time service if they selected at least one of the following categories: “part‐time, paid,” “volunteer (full or part time),” or “paid on call or paid per call”; these selections superseded all options except “full‐time paid.” Participants were categorized as [3] active, other service if they selected at least one of the following categories: “seasonal” or “academy student”; these selections superseded all other uncategorized options (but were superseded by “full‐time paid” and active, volunteer or part‐time service options). A selection of “other” by itself was also categorized under [3] active, other service; otherwise, a selection of “other” was superseded by any other additional options selected. Participants were considered [4] former service if they selected “retired,” “no longer working in the fire service,” or “out on long‐term disability” without selecting any options categorized above except “other.” In the analysis stratified by work status to compare active versus former firefighters (i.e., using the dichotomous variable), only participants who completed the entire questionnaire and reported their work status were included for the most complete data.

Recruitment route was analyzed as an exploratory variable for classifying the presumed outreach method used to recruit participants based on the department reported for their current/most recent affiliation. Specifically, firefighters were categorized as presumably recruited through roster‐based sampling if, by December 31, 2025, they were currently active at a fire department that was either (A) a selected Targeted Enrollment department with considerable management or union participation in recruiting participants (n = 15 departments), or, (B) a recognized Gold Helmet Department (n = 94 departments). Active firefighters reporting any other affiliation were categorized as presumably‐recruited through Open Enrollment sampling. In the analysis stratified by recruitment route, only participants who completed the entire questionnaire, reported an active work status, and listed a department, agency, or organization were included.

The characteristics that we analyzed descriptively included residential region (by nine census‐defined geographic divisions); sex; age at enrollment; race/ethnicity; education; marital status; social security number (SSN) reporting (whether participants provided their partial SSN for health follow‐up through data linkages); work status (four‐level); decade started in the fire service; tenure in the fire service (in years); job assignments ever worked (i.e., a non‐mutually‐exclusive characteristic including structural firefighter, structural firefighting leadership, wildland firefighter, wildland firefighting leadership, fire instructor, fire investigator, or emergency medical technician [EMT] and other assignments); self‐reported cancer history (yes/no); and recruitment route (defined previously).

Data were extracted for analysis on December 31, 2025. Analyses were completed using R Studio. State‐level counts for participants based on their residential state were visualized using a choropleth map created in Microsoft Power BI.

3. Results

3.1. Overall

Overall, 39,058 participants joined the NFR in the first 3 years of enrollment (January 2023 through December 2025). Figure 1 displays the geographic distribution of NFR participation across the U.S. states. Firefighters from all 50 states participated, with substantial variability in state‐level frequencies. States in the highest quartile had between 1176 and 3550 participants, whereas those in the lowest quartile had between 20 and 220 participants. Some participants also resided in the District of Columbia (n = 22) and Puerto Rico and Guam (n < 10 each). In general, more populous states had more participants.

Figure 1.

Frequency of National Firefighter Registry for Cancer participants by self‐reported U.S. state of residence grouped by quartiles, January 2023–December 2025 (Participants missing residential state information excluded. The frequencies of participants in several U.S. territories were not included in the quartiles calculated for mapping; participants enrolled from the District of Columbia [n = 22], Guam [n = 6], and Puerto Rico [n = 5].).

Table 1 presents the descriptive demographic and occupational characteristics of NFR participants overall and by enrollment completion status. Of 39,058 participants, nearly half were concentrated in the South Atlantic (24%) or Mountain (20%) regions of the U.S. Most participants were male (91%), between the ages of 25–44 (50%) or 45–64 (38%), White (83%), held more than a high school education (89%), and were married (72%). Ninety‐three percent of participants reported their partial SSN.

Table 1.

Demographic and occupational characteristics of National Firefighter Registry for Cancer participants overall and by enrollment completion status, January 2023–December 2025.

	Totala		Incompleteb			In progress			Completed
Characteristic	n	%	n	Col %	Row %	n	Col %	Row %	n	Col %	Row %
Overall	39,058	100.0	1639	100.0	4.2	9320	100.0	23.9	28,099	100.0	71.9
Residential division
South Atlantic	9056	24.3	160	19.8	1.8	1936	21.4	21.3	6960	25.4	76.6
Mountain	7535	20.2	118	14.6	1.6	2051	22.7	27.2	5366	19.6	71.1
Pacific	4433	11.9	110	13.6	2.5	1456	16.1	32.8	2867	10.4	64.6
West North Central	4273	11.5	71	8.8	1.7	1047	11.6	24.5	3155	11.5	73.8
East North Central	3976	10.7	128	15.9	3.2	843	9.3	21.2	3005	10.9	75.5
Middle Atlantic	3317	8.9	84	10.4	2.5	613	6.8	18.5	2620	9.5	78.9
West South Central	2077	5.6	64	7.9	3.1	590	6.5	28.4	1423	5.2	68.4
New England	1318	3.5	39	4.8	3.0	246	2.7	18.7	1033	3.8	78.3
East South Central	1282	3.4	33	4.1	2.6	241	2.7	18.8	1008	3.7	78.5
U.S. territory	11	0.0	0	0.0	0.0	3	0.0	27.3	8	0.0	72.7
Missing	1780
Sex
Male	34,017	90.9	—	—	—	8488	91.1	25.0	25,510	90.8	75.0
Female	3423	9.1	—	—	—	832	8.9	24.3	2589	9.2	75.6
Missing	1618
Age (years)
< 25	2478	6.6	—	—	—	639	6.9	25.8	1839	6.5	74.2
25–44	18,527	49.5	—	—	—	4868	52.2	26.3	13,656	48.6	73.7
45–64	14,167	37.9	—	—	—	3424	36.7	24.2	10,740	38.2	75.8
65+	2255	6.0	—	—	—	389	4.2	17.3	1864	6.6	82.7
Missing	1631
Race/ethnicity
White	29,072	82.8	—	—	—	5665	78.3	19.5	23,393	84.0	80.5
Black	994	2.8	—	—	—	232	3.2	23.3	761	2.7	76.6
Hispanic	1455	4.1	—	—	—	362	5.0	24.9	1093	3.9	75.1
Asian	478	1.4	—	—	—	173	2.4	36.2	303	1.1	63.4
American Indian or Alaska Native	243	0.7	—	—	—	59	0.8	24.3	184	0.7	75.7
Native Hawaiian or Pacific Islander	183	0.5	—	—	—	84	1.2	45.9	99	0.4	54.1
Middle Eastern or North African	27	0.1	—	—	—	3	0.0	11.1	24	0.1	88.9
Multiple	2654	7.6	—	—	—	660	9.1	24.9	1992	7.2	75.1
Missing	3952
Education
Less than high school	72	0.2	—	—	—	19	0.3	26.4	53	0.2	73.6
High school	3756	10.7	—	—	—	800	11.1	21.3	2953	10.6	78.6
Some college	16,652	47.5	—	—	—	3505	48.6	21.0	13,139	47.2	78.9
College or more	14,577	41.6	—	—	—	2884	40.0	19.8	11,685	42.0	80.2
Missingc	4001
Marital status
Married	24,779	71.7	—	—	—	4994	70.5	20.2	19,771	72.0	79.8
Not married	9791	28.3	—	—	—	2090	29.5	21.3	7696	28.0	78.6
Missingc	4488
SSN reported
Yes	36,159	92.6	—	—	—	8959	96.1	24.8	27,186	96.8	75.2
No	2899	7.4	—	—	—	361	3.9	12.5	913	3.2	31.5
Work status
Active, full‐time service	28,503	75.3	529	65.5	1.9	7146	78.8	25.1	20,828	74.5	73.1
Active, volunteer or part‐time service	5468	14.5	186	23.0	3.4	1161	12.8	21.2	4121	14.7	75.4
Active, other service	491	1.3	27	3.3	5.5	117	1.3	23.8	347	1.2	70.7
Former service	3368	8.9	66	8.2	2.0	643	7.1	19.1	2659	9.5	78.9
Missing	1228
Start year
< 1980s	1795	5.2	—	—	—	325	4.7	18.1	1469	5.3	81.8
1980s	3130	9.0	—	—	—	586	8.5	18.7	2543	9.1	81.2
1990s	6386	18.4	—	—	—	1327	19.2	20.8	5057	18.2	79.2
2000s	9206	26.5	—	—	—	1954	28.2	21.2	7250	26.1	78.8
2010s	8066	23.2	—	—	—	1611	23.3	20.0	6453	23.2	80.0
2020s	6174	17.8	—	—	—	1123	16.2	18.2	5050	18.2	81.8
Missing	4301
Tenure (years)
< 10	10,728	30.8	—	—	—	1985	28.6	18.5	8740	31.4	81.5
10–19	8673	24.9	—	—	—	1793	25.9	20.7	6879	24.7	79.3
20–29	9004	25.9	—	—	—	1887	27.2	21.0	7116	25.5	79.0
30+	6393	18.4	—	—	—	1270	18.3	19.9	5119	18.4	80.1
Missing	4260
Job assignments (ever held)d
Structural firefighter	28,513	87.4	—	—	—	4293	86.2	15.1	24,212	87.6	84.9
Structural fire leadership	16,004	49.0	—	—	—	2448	49.1	15.3	13,550	49.0	84.7
Wildland firefighter	5418	16.6	—	—	—	879	17.6	16.2	4538	16.4	83.8
Wildland supervisor	1883	5.8	—	—	—	324	6.5	17.2	1558	5.6	82.7
Fire investigator	1353	4.1	—	—	—	251	5.0	18.6	1102	4.0	81.4
Instructor	2420	7.4	—	—	—	388	7.8	16.0	2029	7.3	83.8
Other and EMT	6635	20.3	—	—	—	987	19.8	14.9	5646	20.4	85.1
Missing	6422
Cancer history
No	24,295	85.3	—	—	—	472	84.6	1.9	23,823	85.3	98.1
Yes	4019	14.1	—	—	—	79	14.2	2.0	3934	14.1	97.9
Unsure	178	0.6	—	—	—	7	1.3	3.9	171	0.6	96.1
Missing	10,566
Recruitment route
Roster‐based	9848	27.5	—	—	—	2597	32.3	26.4	7249	26.1	73.6
Open enrollment	25,924	72.5	—	—	—	5448	67.7	21.0	20,474	73.9	79.0
Missing	3286

^a Missing frequencies only presented for the total sample and excluded from percentage calculations.

^b Results not reported where nearly all observations were missing values (non‐missing n < 21).

^c Missing values for education and marital status include a response of “prefer not to answer” (n = 156 and 375, respectively).

^d Participants may fall into more than one category for job assignment.

Regarding work characteristics, about 75% of participants reported an active full‐time service in the fire service, 14% were active volunteer or part‐time service, and 1% were active as “other”; 9% of participants were former (e.g., retired) firefighters. Over half of participants started their fire service career since the year 2000 (67%). Participation was similar across tenure in the fire service (25%–31%), except for a slightly smaller proportion of people reporting 30+ years (18%). The most commonly reported job assignments that participants worked were structural firefighter (87%), structural firefighting leadership (49%), EMT and other assignments (20%), and wildland firefighter (17%).

Fourteen percent of participants reported a history of cancer. About three quarters of participants were presumably recruited through Open Enrollment, meaning they were not necessarily at a department where active roster‐wide recruitment was occurring.

3.2. Completion Status

Most participants completed all components of the enrollment process (n = 28,099; 72%), while 24% (n = 9320) were “in progress,” and 4% (n = 1639) were “incomplete” as of December 31, 2025 (Table 1). When looking at differences by completion status (i.e., column percents), the distributions of nearly all characteristics were similar (not considering the incomplete group for variables where few provided a response, i.e., n ≤ 21 for all variables except residential region, work status, and sex). However, a slightly higher proportion of participants whose enrollment was in‐progress or completed reported an active full‐time status than participants in the incomplete group (75%–79% vs. 65%); and a higher proportion of participants in the incomplete group reported an active volunteer or part‐time status (23% vs. 13%–15%).

Completion rates (i.e., row percents) were inherently highest among participants that provided information about characteristics ascertained toward the end of the enrollment process (e.g., 81%–98% in the completed group for job assignments and cancer history vs. smaller percentages in the completed group for other variables). When looking at completion rate differences within variables, a slightly higher proportion of participants who were aged 65 and older completed the entire process compared to those under the age of 45 (83% vs. 74%). Smaller proportions of some non‐White groups of participants completed the questionnaire (54%–63%) than White participants (80%); however, sample sizes were smaller and therefore less stable. A higher proportion of participants who reported an SSN completed the enrollment process (75%) compared to those that did not report an SSN (31%), indicating that a lack of SSN reporting may be driven by a failure to complete the process and not an unwillingness to provide the information.

Figure 2 presents the number and percentage of firefighters who made it to selected stages of the enrollment process. The final section reflects only those who submitted the questionnaire (72%). The largest drop in completed enrollments occurred between the work history and health history sections; nearly 17% of participants did not make it past work history and onto health history (90% and 73%, respectively). Of participants who started the health history section, very few dropped out (1%) while responding to questions on health and lifestyle/health behavior prior to submitting the questionnaire.

Figure 2.

Distribution of National Firefighter Registry for Cancer participants retained at selected sequential steps of the enrollment process, January 2023–December 2025 (Retention at each step was defined as completion of informed consent, initiation of the user profile, initiation of selected sections of the questionnaire [i.e., demographics, work history, and health history] and questionnaire submission.).

3.3. Work Status

Table 2 displays the distribution of characteristics by work status among participants who completed the enrollment process (n = 27,955 excluding those missing work status). Many differences were seen, largely driven by age at enrollment because most former firefighters were retirees (not shown). Roughly 60% of active participants were under 45 years of age compared to 6% of former firefighters; and 40% of former firefighters were aged 65 or over compared to 3% of active firefighters. Unsurprisingly, more former firefighters were married, started their fire service career earlier, and reported a longer tenure than the active firefighters. Slightly more former firefighters were White than active firefighters (91% vs. 83%). While the distribution of most job assignments ever held was similar between active and former firefighters, substantially more former firefighters had ever worked in structural firefighting leadership than active firefighters, unsurprisingly (67% vs. 47%). The proportion of participants reporting a history of cancer was drastically higher among former firefighters (46% vs. 11%). The distribution of recruitment route slightly differed by work status, suggesting that active firefighters were more successfully recruited than former firefighters through roster‐based sampling (27% vs. 17%, respectively), and, former firefighters were more successfully recruited than active firefighters through the Open Enrollment route (83% vs. 73%, respectively). Furthermore, the distribution suggests that roster‐based recruitment, although aimed only at active firefighters, might be reaching firefighters formerly affiliated with the engaged departments.

Table 2.

Demographic and occupational characteristics of National Firefighter Registry for Cancer participants ^a by work status, January 2023–December 2025.

Characteristic	Active service		Former service
	n b	%	n b	%
Overall	25,296		2659
Residential division
South Atlantic	6238	25.2	713	27.0
Mountain	4953	20.0	403	15.3
Pacific	2475	10.0	388	14.7
West North Central	3003	12.1	147	5.6
East North Central	2645	10.7	351	13.3
Middle Atlantic	2390	9.7	224	8.5
West South Central	1250	5.0	169	6.4
New England	892	3.6	141	5.3
East South Central	901	3.6	106	4.0
U.S. territory	8	0.0	0	0.0
Sex
Male	22,942	90.7	2434	91.5
Female	2354	9.3	225	8.5
Age (years)
< 25	1826	7.2	4	0.2
25–44	13,444	53.1	147	5.5
45–64	9247	36.6	1441	54.2
65+	779	3.1	1067	40.1
Race/ethnicity
White	20,880	83.3	2398	90.9
Black	680	2.7	75	2.8
Hispanic	1046	4.2	45	1.7
Asian	293	1.2	10	0.4
American Indian or Alaska Native	163	0.6	19	0.7
Native Hawaiian or Pacific Islander	94	0.4	5	0.2
Middle Eastern or North African	24	0.1	0	0.0
Multiple	1900	7.6	87	3.3
Education
Less than high school	46	0.2	7	0.3
High school	2706	10.8	227	8.6
Some college	11,790	47.0	1289	48.8
College or more	10,519	42.0	1116	42.3
Marital status
Married	17,509	70.9	2166	82.4
Not married	7203	29.1	462	17.6
SSN reported
Yes	24,466	96.7	2592	97.5
No	830		67
Start year
< 1980s	656	2.6	796	30.2
1980s	1552	6.2	980	37.2
1990s	4478	17.9	554	21.0
2000s	7003	27.9	212	8.1
2010s	6348	25.3	78	3.0
2020s	5023	20.0	13	0.5
Tenure (years)
< 10	8525	34.0	181	6.9
10–19	6590	26.3	260	9.9
20–29	6339	25.3	741	28.2
30+	3642	14.5	1446	55.0
Job assignments (ever held)c
Structural firefighter	21,858	87.9	2233	84.7
Structural fire leadership	11,706	47.0	1775	67.4
Wildland firefighter	4036	16.2	486	18.4
Wildland supervisor	1315	5.3	232	8.8
Fire investigator	906	3.6	192	7.3
Instructor	1679	6.7	336	12.8
Other and EMT	4880	19.6	741	28.1
Cancer history
No	22,320	88.8	1397	52.7
Yes	2687	10.7	1223	46.2
Unsure	141	0.6	29	1.1
Recruitment route
Roster‐based	6783	27.2	443	16.9
Open enrollment	18,186	72.8	2183	83.1

^a Including only participants who completed the enrollment questionnaire and reported information on work status.

^b Missing values excluded.

^c Participants may fall into more than one category for job assignment.

3.4. Recruitment Route

Table 3 shows the distribution of characteristics by presumed recruitment route among participants who were in an active work status and completed the enrollment process (n = 24,969 excluding those missing department affiliation and/or work status). We observed no notable differences by recruitment route for residential region, sex, age, education, marital status, and SSN reporting. Roster‐based recruitment appeared to be more effective for recruiting non‐White firefighters in general (74% White among roster‐based group vs. 87% among Open Enrollment group). There were more active full‐time firefighters among the roster‐based recruitment group (95% vs. 78%); and more active volunteer or part‐time firefighters in the Open Enrollment recruitment group (21% vs. 4%). A higher proportion of newer firefighters, and subsequently those with less tenure, enrolled via the roster‐based recruitment route. The proportions of participants who ever worked in wildland firefighting/supervising or structural firefighting leadership were larger in the Open Enrollment recruitment group. Lastly, the proportion of participants reporting a history of cancer was slightly smaller in the roster‐based recruitment group than in the Open Enrollment recruitment group (7% vs. 12%).

Table 3.

Demographic and occupational characteristics of National Firefighter Registry for Cancer participants ^a by recruitment route, January 2023–December 2025.

Characteristic	Roster‐based		Open enrollment
	n b	%	n b	%
Overall	6783		18,186
Residential division
South Atlantic	2045	31.0	4127	23.1
Mountain	1503	22.8	3399	19.0
Pacific	830	12.6	1622	9.1
West North Central	1125	17.1	1857	10.4
East North Central	353	5.4	2224	12.4
Middle Atlantic	249	3.8	2118	11.8
West South Central	205	3.1	1036	5.8
New England	81	1.2	805	4.5
East South Central	196	3.0	698	3.9
U.S. territory	0	0.0	8	0.0
Sex
Male	6259	92.3	16,396	90.2
Female	524	7.7	1790	9.8
Age (years)
< 25	652	9.6	1115	6.1
25–44	3986	58.8	9294	51.1
45–64	2081	30.7	7079	38.9
65+	64	0.9	698	3.8
Race/ethnicity
White	4991	74.3	15,662	86.7
Black	262	3.9	405	2.2
Hispanic	507	7.5	523	2.9
Asian	165	2.5	125	0.7
American Indian or Alaska Native	45	0.7	113	0.6
Native Hawaiian or Pacific Islander	62	0.9	30	0.2
Middle Eastern or North African	14	0.2	10	0.1
Multiple	673	10.0	1193	6.6
Education
Less than high school	7	0.1	39	0.2
High school	753	11.2	1891	10.5
Some college	3295	49.1	8370	46.4
College or more	2652	39.5	7756	43.0
Marital status
Married	4423	67.1	12,922	72.4
Not married	2172	32.9	4917	27.6
SSN reported
Yes	6522	96.2	17,675	97.2
No	261		511
Work statusa
Active, full‐time service	6432	94.8	14,167	77.9
Active, volunteer or part‐time service	279	4.1	3793	20.9
Active, other service	72	1.1	226	1.2
Start year
<1980s	49	0.7	599	3.3
1980s	181	2.7	1361	7.5
1990s	883	13.1	3566	19.7
2000s	1860	27.6	5090	28.1
2010s	1885	28.0	4400	24.3
2020s	1870	27.8	3067	17.0
Tenure (years)
< 10	2901	43.0	5487	30.3
10–19	1799	26.7	4729	26.2
20–29	1488	22.1	4808	26.6
30+	556	8.2	3060	16.9
Job assignments (ever held)c
Structural firefighter	6152	92.5	15,551	86.2
Structural fire leadership	2360	35.5	9282	51.4
Wildland firefighter	496	7.5	3510	19.5
Wildland supervisor	82	1.2	1228	6.8
Fire investigator	119	1.8	783	4.3
Instructor	300	4.5	1375	7.6
Other and EMT	1238	18.6	3610	20.0
Cancer history
No	6274	93.1	15,774	87.1
Yes	443	6.6	2223	12.3
Unsure	25	0.4	112	0.6

^a Including only participants who completed the enrollment questionnaire, reported an active work status (i.e., excluding former service), and selected a department, agency, or organization.

^b Missing values excluded.

^c Participants may fall into more than one category for job assignment.

4. Discussion

During the first 3 years of enrollment (2023–2025), nearly 40,000 current and former U.S. firefighters voluntarily joined the National Firefighter Registry for Cancer. Firefighters joined from all 50 states and several territories and a range of firefighting assignments. Roughly 4000 firefighters with a self‐reported history of cancer enrolled.

Completion rates were high even though the estimated length of time it takes to complete all components of the enrollment process (i.e., informed consent, user profile, and enrollment questionnaire) is 40 min or longer. Over 70% of participants who provided consent completed the questionnaire; and participants who had not provided enough information to conduct cancer status follow‐up via cancer registry linkages (i.e., incomplete group) accounted for less than 5% of the sample. As participants can save their progress and return later to complete enrollment, the process was not completed in one sitting by all participants. “Push notifications” with reminders to complete enrollment were sent via email or text message to participants who had not submitted the questionnaire as a tool to promote completion and will continue to be used. It is possible that a portion of participants from the incomplete and in‐progress groups could still return and complete enrollment after data were exported for this analysis.

The section that provided the largest barrier to completion appeared to be the work history portion of the questionnaire; roughly one in six participants failed to complete this section and move to the health history questions. The work history section was designed to estimate exposures by incident type for each job assignment at every fire department reported by firefighters. As a result, individuals who held several job assignments at multiple departments throughout their careers were asked to answer more questions than those with fewer job assignments or departments. This may have increased the burden for those with more extensive work histories and could have contributed to incomplete registration. However, we did not observe differences in completion rates by job tenure or start year. In fact, a higher proportion of individuals aged 65 and older completed enrollment in comparison to younger participants. These findings suggest that attrition in the work history section was distributed across the cohort and did not disproportionately affect firefighters with more comprehensive work histories.

Active, volunteer or part‐time firefighters made up a slightly larger portion of the incomplete enrollments than they did the in‐progress and completed enrollments; although, most completed the enrollment process. Studies have shown that volunteer firefighters report multiple barriers to participating in some health and safety practices, including time constraints with work or personal schedules, less predictable shifts, and less department‐level health and safety infrastructure [39, 40, 41]. These barriers could also explain why some firefighters do not complete NFR enrollment process. Lastly, although many of the work‐related questionnaire items pertained primarily to structural firefighting (e.g., department‐focused language, selected workplace practices, etc.), enrollment completion rates were no lower among those who had ever worked in wildland firefighting.

The abovementioned barriers to enrollment completion could also be barriers to NFR participation, in addition to a lack of awareness about the NFR. Volunteer firefighters were substantially underrepresented in the first three years of enrollment. While active volunteer or part‐time firefighters composed only 14% of the overall sample (and 16% of active participants), the National Fire Protection Association (NFPA) estimates that they make up about 65% of the national, municipal firefighting workforce [1].

On the other hand, the proportion of women, overall and among active firefighters, enrolled in the NFR thus far is the same as the NFPA‐estimated proportion of women in the national workforce of current municipal firefighters (9%). The Bureau of Labor Statistics (BLS) estimated that roughly 6% of firefighters were women in 2024, but these data represent jobs worked for pay only (and is similar to NFPA's estimate for the career firefighting workforce—5% women) [42]. Consistent with NFPA and BLS trends, an ad hoc analysis showed that, overall, a larger proportion of active, volunteer or part‐time participants were women than were active, full‐time participants (14% vs. 8%; data not shown).

While the NFPA and BLS estimate that firefighters under the ages of 45 or 50 represent about 70% of the national workforce, they made up a slightly smaller proportion—but still a majority—of active NFR participants.

While estimates on race/ethnicity are not available from the NFPA, the BLS estimated that 83% of paid firefighters were White, about 10% were Black, 3% were Asian, and 15% were Hispanic in 2024. A similar proportion of active NFR participants was White, but the proportions for non‐White groups (especially Black and Hispanic) were smaller among active NFR participants than in the BLS‐estimated workforce. We examined participants reporting multiple races as a separate category, but a post hoc calculation of non‐mutually‐exclusive groups (i.e., eliminating the multiple races category) produced slightly larger proportions of most non‐White groups, including a doubled proportion of Hispanic firefighters (9%; data not shown) among active participants. Lastly, we found that the active firefighting group contained a smaller proportion of White participants than the former firefighting group. This difference might reflect a national shift in workforce demographics over time.

Interestingly, the Mountain region of the U.S. (Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah, and Wyoming) accounted for one of the largest groups of participants even though it is one of the least populous [43]. By contrast, the West South Central region (Arkansas, Louisiana, Oklahoma, and Texas) was underrepresented in the NFR given its larger proportion of the U.S. population. Few participants from U.S. territories enrolled in the NFR during the first 3 years of enrollment. In general, comparisons between NFR participation rates and general population size should be interpreted with caution, as general population size does not necessarily correspond to firefighting workforce size (particularly for more granular units of geography); population‐to‐firefighter ratios vary across municipalities.

About 14% of participants overall, and 46% of former firefighting participants, reported a history of cancer. Note that cancer ascertainment was not limited to diagnoses made after participants began a fire service career, but rather any history of cancer. The complete prevalence of cancer in the general U.S. population was about 5% among all ages and 21% among those ages 65+ in 2022 according to data from the Surveillance, Epidemiology, and End Results (SEER) Program [44]. Although studies show that firefighters have higher rates of cancer (incidence) than observed in the general U.S. population [7], it is likely that there is some degree of unmeasured selection bias based on cancer history among the NFR sample—meaning that firefighters with a personal history of cancer were more likely to participate—due to the voluntary design and the cancer‐focused objective.

However, the exploratory evaluation of recruitment tactics suggested that roster‐based recruitment might reduce possible selection bias by cancer status among active firefighters since the proportion of participants reporting a history of cancer appeared smaller among this group. This is plausible because roster‐based recruitment could have been more successful at promoting participation among firefighters who would have otherwise been less motivated to join than those with a history of cancer. Although the difference was somewhat small (5%) and based on assumptions in defining recruitment route, this trend is encouraging and suggests that recruitment from rosters or other inventories of personnel might help address selection bias that is inherent to voluntary occupational health/exposure registries.

The exploratory evaluation of presumed recruitment route also suggested that both methods performed equally well for sex. This was somewhat surprising because initial Targeted Enrollment departments were selected to oversample for women. However, Gold Helmet departments contributed most roster‐based‐recruited participants, and these were not departments selected by NIOSH for any workforce characteristics. Further, Open Enrollment activities likely reached more women than originally anticipated, since the proportion of female participants was no smaller than in the national workforce. Roster‐based recruitment appeared to provide small improvements in the recruitment of most non‐White groups; relying on this method more heavily to recruit from these groups in the future could lend promising results. On the other hand, Open Enrollment recruitment appeared, thus far, more effective for recruiting wildland firefighters and structural firefighting leadership. The decentralized nature of wildland firefighting agency records can make roster‐based recruitment challenging, which is why initial roster‐based recruitment efforts targeted municipal fire departments.

However, it should be noted that the recruitment route comparison is vulnerable to misclassification because many participants in the roster‐based group may have actually been recruited through Open Enrollment means (e.g., conference outreach, social media, word of mouth, etc.); and some participants in the Open Enrollment group may be at departments that are involved in active roster‐wide recruitment efforts either unbeknownst to NIOSH or not yet meeting the threshold of response rate/engagement at the time of this analysis. Such misclassification is a limitation and might mask real/larger differences.

Another source of potential misclassification is the hierarchical method used to categorize current work status into mutually exclusive groups, particularly among participants who selected multiple options. Thus, the characteristics of those with overlapping roles could be masked. However, the categorization system was developed in consultation with fire service subject matter experts and designed to reflect firefighters' current, primary occupational activities, putting more weight on any active service.

Additional limitations to the dataset include multiple other potential sources of bias that may affect the generalizability of results and could be considered in the interpretation of future analyses. Examples include survivor bias (i.e., an underrepresentation of firefighters with cancers that have lower survival rates) that could coincide with differing levels of exposure or cancer risk factors, and healthy worker effects in general population comparisons. Furthermore, self‐reported information can be vulnerable to recall errors, particularly for historical information and for participants with a long work history. Missingness is also a potential concern, particularly for the larger sample that will be linked with state cancer registry records (i.e., the in‐progress and completed groups combined). Proportions of missing values for this group are as high as 6‐8% for race/ethnicity, education, marital status, start year, and tenure; 13% for job assignment; and 24% for self‐reported cancer history (not shown). Among only participants who completed the enrollment process (i.e., submitted the questionnaire), missingness for any characteristics that we analyzed was very small (< 2%; not shown). Non‐random patterns in missing data cannot be ruled out and could bias results. Cancer registry linkages are expected to improve cancer ascertainment among the in‐progress group over time. Unexpectedly, SSN reporting was very high—unlike what has been observed in general population studies [45, 46]—and did not differ by completion status, work status, or recruitment route.

In general, the NFR data provide ample opportunity for varied cross‐sectional analyses and for comprehensive, prospective, cancer follow‐up through cancer registry and other database linkages. Opportunities might also exist to further develop the exposure/health profiles of NFR participants through linkages with other firefighting studies or systems, such as the Fire Fighter Cancer Cohort Study or the National Emergency Response Information System [33, 47].

In addition to ongoing recruitment to increase overall participation and to capture new firefighters over time, particular recruitment efforts can target volunteer and non‐White firefighters and underrepresented states/territories. Furthermore, focused communication efforts and technical assistance (e.g., push notifications, in‐person guidance, “train the trainer” material, etc.) could improve participants' completion past the work history section of the enrollment questionnaire. Lastly, vigorous, roster‐based recruitment within departments or agencies can continue being employed to increase the number of participants and potentially offset selection bias within the NFR sample. Scientists can consider similar models as a tool for other voluntary, occupational exposure and health registries. The suggestive patterns that we observed with roster‐based recruitment could also be used to encourage other fire departments and agencies to become engaged in recruitment. NIOSH can partner with fire service organizations, public health groups, and other localand state agencies to increase outreach.

5. Conclusion

The NFR for Cancer provides a large and rich surveillance dataset for studying workforce characteristics, workplace practices, and cancer risk factors among a national sample of firefighters. The distribution of demographic characteristics of NFR participants is similar to the national firefighter workforce in some respects, but generalizability to the entire U.S. fire service population cannot be inferred due to multiple potential sources of selection bias. Over time, ongoing recruitment and retention efforts are expected to result in sizable increases in participation from groups of firefighters who have been underrepresented in cancer surveillance and research, a study group that might be less vulnerable to selection bias by cancer status, and a large dataset for analyzing cancer and cancer risk factors among the U.S. Fire Service for the foreseeable future.

Author Contributions

M.R.S. and A.C.M. contributed to conception and design of the analysis and interpretation of results. A.C.M. led the data analysis. M.R.S. led the drafting of the manuscript. M.R.S., A.C.M., K.W.F., and A.F.W. contributed to the acquisition of data and critical review of the manuscript and approved journal submission.

Disclosure

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the National Institute for Occupational Safety and Health or Centers for Disease Control and Prevention.

Consent

This activity was reviewed by CDC, deemed not research, and was conducted consistent with applicable federal law and CDC policy (See e.g., 45 C.F.R. part 46.102(l)(2), 21 C.F.R. part 56; 42 U.S.C. §241(d); 5 U.S.C. §552a; 44 U.S.C. §3501 et seq.). Nonetheless, due to the voluntary nature of participation, all NFR enrollees undergo written informed consent in the web portal prior to enrolling.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting Table S1

Data Availability Statement

The data that support the findings of this study are in preparation for public data sharing in accordance with an Assurance of Confidentiality and procedures currently under development. For more information email NFRegistry@cdc. gov.