Health Effects Associated With Jet Fuel Exposure

This systematic literature review investigates the health effects associated with jet fuel exposure in military settings. The methods used ensure transparency when synthesizing epidemiologic, animal toxicological, and mechanistic data, with adaptations to address exposure duration and the relationship between acute and long-term conditions.

Abstract

Objective

The aim of the study was to present the methods for a fit-for-purpose systematic review designed to assess the health effects of jet fuel exposure in military settings by synthesizing data from epidemiologic, animal toxicological, and mechanistic studies.

Methods

Methods followed the US Environmental Protection Agency’s Integrated Risk Information System and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines with adaptations to meet the needs for this review.

Results

A literature search was conducted across scientific databases, gray literature, and expert-identified sources, resulting in the inclusion of 279 relevant references. Results are presented in a systematic evidence map and companion articles.

Conclusions

The methods employed ensure transparency in the screening, evaluation, and synthesis of data, establishing a foundation for further research on the health impacts of jet fuel exposure in veterans.

LEARNING OUTCOMES

Upon reviewing this manuscript, readers will be able to:

• Describe the major routes of occupational and environmental exposure to jet fuel among military personnel and the factors influencing exposure levels.

• Explain the methodological framework used in this systematic review, including the adaptation of EPA IRIS and PRISMA guidelines, study selection criteria, and evidence synthesis approaches.

• Identify research gaps related to long-term health effects, exposure duration, and the role of acute symptoms in predicting chronic outcomes following jet fuel exposure.

Jet fuel is a significant exposure concern for many military personnel. Whether stationed at garrison locations or deployed settings, service members encounter jet fuels through various occupational tasks and environmental conditions. Although commonly referred to as “jet fuels,” aviation turbine fuels serve multiple functions beyond aviation. In the United States (US) military, these fuels are also widely used in generators and both wheeled and tracked vehicles, such as Mine-Resistant Ambush-Protected vehicles and A1 Abrams tanks.¹ Additionally, jet fuels have historically been used as accelerants in burn pits in deployed locations. Because of their versatility and extensive operational applications, jet fuel exposures represent a potential health risk factor for military personnel and veterans.^2–5 Service members may be particularly vulnerable to exposure while performing tasks such as aircraft and vehicle operation, fueling, and defueling; fuel tank entry and cleaning; aircraft fuel-cell maintenance; fuel performance testing; maintenance of military aircraft and machinery; transportation of jet fuel; and storage tank maintenance. Those present during spills or accidental releases may also experience exposure.

Occupational exposure to jet fuel can occur through multiple routes. Dermal contact may result from spills onto the skin or clothing, while inhalation of vapors can occur during maintenance activities or fueling operations. Several factors influence exposure intensity to jet fuel, including task location, ventilation adequacy, task duration, and use of personal protective equipment. These variables contribute to differing exposure of both liquid fuel and vapors. Though less common, ingestion and ocular contact are additional potential exposure pathways. Despite advancements in risk mitigation measures, such as protective clothing and ventilation, concerns about the health implications of jet fuel exposure for military personnel have persisted.^3,6

Beyond occupational hazards, environmental contamination events have also affected service members and their families. Two significant incidents occurred at Camp Lejeune, North Carolina, and the Red Hill Bulk Fuel Storage Facility near Joint Base Pearl Harbor-Hickam, Hawaii. At Camp Lejeune, widespread contamination of drinking water with jet fuel constituents and other volatile organic compounds from the 1950s to the 1980s raised substantial public health concerns.⁷ More recently, jet fuel leaks at Red Hill contaminated drinking water supplies for military and civilian communities, triggering immediate health advisories and ongoing remediation efforts.^8–11

US MILITARY JET FUEL

The jet fuels most commonly used in military operations are complex mixtures of aromatic and aliphatic hydrocarbons, along with nonhydrocarbon performance additives. Kerosene comprises most of the total volume. Although toxic hydrocarbons, including benzene, toluene, and naphthalene, are present only in minor quantities, their volatility increases the inhalation exposure potential among military personnel working with jet fuels.^12–16 Over time, the composition and operational use of jet fuels by the US Armed Forces have evolved to address safety, performance, and logistical requirements.⁶ Before the mid-1990s, multiple fuel types were in service, including Jet A, jet propellant-4 (JP-4), jet propellant-5 (JP-5), and jet propellant-8 (JP-8). JP-4, a highly volatile fuel, was phased out and replaced by JP-8 due to the latter’s lower volatility and improved safety characteristics. Currently, JP-5—with a higher flash point of 60°C—is primarily used by the US Navy to enhance shipboard handling safety. For land-based operations, the US Air Force and US Army predominantly rely on JP-8 (flash point: 38°C), while Jet A, a commercially available fuel similar in composition to JP-8, remains in use in certain domestic contexts. All jet fuels used by the military contain mission-specific additives to improve performance and stability, including antioxidants, static inhibitors, corrosion inhibitors, fuel system icing inhibitors, lubrication improvers, biocides, and thermal stability improvers.

PURPOSE

There is extensive information available from authoritative sources, including the International Agency for Research on Cancer,¹² National Toxicology Program,¹³ US Environmental Protection Agency (EPA),^14–16 and Agency for Toxic Substances and Disease Registry,^17–19 that characterizes the health effects of individual jet fuel constituents (eg, benzene, toluene, and xylene). However, less is known about the effects of these constituents when present at low concentrations within a heterogeneous fuel mixture. Therefore, this review focuses on the health effects of jet fuels as a mixture rather than on individual components.

Previous reviews assessing jet fuel as a mixture have provided valuable insights into acute health effects.^3,20,21 In a recent systematic review,²² we described the epidemiologic studies examining health outcomes associated with occupational jet fuel exposure in both military and nonmilitary settings. This review concluded that there was slight evidence for associations between jet fuel exposure and neurological, cognitive, and behavioral, respiratory, and cancer outcomes, while evidence for other health outcomes was considered indeterminate. The ability to make stronger conclusions was limited by several factors, including the small number and low quality of the relevant occupational exposure studies.

Recognizing these limitations, the current review aimed to expand the evidence base by further exploring these identified health outcomes while integrating toxicological and mechanistic data to address gaps where epidemiologic evidence is sparse. This broader approach enabled a more comprehensive evaluation of potential chronic effects and biological mechanisms that may underlie observed associations. Additionally, we used a systematic evidence map to identify areas that may have sufficient evidence for evaluation and identify gaps in evidence where future research would be best directed.²³ Here, we detail the methodology used for the fit-for-purpose systematic review to assess the health effects of jet fuel exposure. Results for cancer effects,²⁴ neurological, cognitive, and behavioral effects,²⁵ respiratory effects,²⁶ immune effects,²⁷ hematologic and cardiovascular effects,²⁸ hepatic and renal effects,²⁹ and endocrine and reproductive effects³⁰ are available in companion articles.

METHODS

Study Design and Scope

The methods used in this systematic literature review were adapted from those employed by the EPA for Integrated Risk Information System (IRIS) assessments.^23,31 The IRIS methodology was selected because it is an established framework used to identify health hazards associated with chemicals in the environment in support of scientific decision-making. The IRIS methodology considers the entire body of scientific literature by incorporating epidemiologic studies, animal toxicological studies, and mechanistic studies. Additionally, this review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.³² The PRISMA statement provides an evidence-based, 27-item checklist designed to ensure transparency in reporting systematic reviews, thereby enhancing review quality and enabling methodological replication (Supplemental Digital Content, Table S1, http://links.lww.com/JOM/C263).

This body of work expands on a previous review of the occupational evidence with a broadened scope.^22,33 The fit-for-purpose aspects of this protocol incorporate modifications to the IRIS approach necessitated by the need for efficiency to adhere to congressional timelines³³ when the review was conceived. As detailed below, these include single independent extractions with quality control by a secondary reviewer, and a modified study quality approach.

The a priori scope of the systematic review is to evaluate any associations between exposure duration and health outcomes, recognizing that prolonged and consistent exposures may influence risk differently than intermittent or short-term exposures. Furthermore, this review sought to examine associations between immediate symptoms and long-term health outcomes to assess both short- and long-term health outcomes associated with jet fuel exposure. Additional study protocol details are provided in the Supplemental Digital Content (http://links.lww.com/JOM/C263).

PECO Criteria

The Population, Exposure, Comparison, and Outcome (PECO) framework provided the basis for the literature search strategies. These criteria were used to define the inclusion and exclusion parameters for screening literature search results, enabling the identification and prioritization of epidemiologic, animal toxicological, and mechanistic studies related to jet fuel exposure and associated health effects. Separate PECO criteria were applied to each category of literature (epidemiologic, animal toxicological, and mechanistic). The full PECO terms are presented in Tables 1–3 and were selected to ensure the capture of all relevant literature on health effects resulting from acute, subchronic, and chronic exposures through any exposure pathway (eg, inhalation, oral, dermal, and intraperitoneal injection).

TABLE 1.

Populations, Exposures, Comparators, and Outcomes Criteria for a Systematic Review on the Health Effects From Exposure to Jet Fuel Within Epidemiologic Studies

PECO Element	Inclusion Criteria
Population	Any population and life stage (eg, military, occupational, or general population, including children and other sensitive populations), not limited by country.
Exposure	Jet fuels Other names: jet exhaust, jet engine exhaust, aircraft exhaust, aircraft engine exhaust, aircraft fuel, aviation fuel, aviation turbine fuel, jet propellant, aviation propellant, aircraft propellant, jet A fuel, jet A-1 fuel, Jet B fuel, TS-1 fuel, JP-1 fuel, JP1 fuel, JP-2 fuel, JP2 fuel, JP-3 fuel, JP3 fuel, JP-4 fuel, JP4 fuel, JP-5 fuel, JP5 fuel, JP-6 Fuel, JP6 Fuel, JP-7 Fuel, JP7 Fuel, JP-8 fuel, JP8 fuel, JP-9 fuel, JP9 fuel, JP-10 fuel, JP10 fuel, JPTS fuel, zip fuel, JP5 jet fuel, S-8 fuel, JP8 aviation fuel, JP4 aviation fuel, alternative jet fuel, synthetic jet fuel Any exposure to jet fuels, listed above, via any exposure route (eg, oral, dermal, inhalation, or unknown/multiple routes). Includes exposures to jet fuels through accidental spills such as ground water/drinking water.
Comparator	A comparison or referent population not exposed, or exposed to lower levels, of jet fuels.
Outcome	All health outcomes (both cancer and noncancer). Epidemiologic studies with self-reported diagnosed disease and self-reported symptoms are included.

This PECO expands on the PECO used in a previous review.^22,33

TABLE 3.

Populations, Exposures, Comparators, and Outcomes Criteria for a Systematic Review on the Health Effects From Exposure to Jet Fuel within Mechanistic Studies

PECO Element	Inclusion Criteria
Population	Human: any population and life stage (eg, military, occupational, or general population, including children and other sensitive populations), not limited by country. Animal: vertebrates (eg, mammals, avian species, fish, reptiles) of any life stage (preconception, in utero, lactation, peripubertal, and adult stages). Ex vivo, in vitro, in silico: Cultures of human or animal primary, immortalized, or transformed cells from relevant animal models (vertebrates, see Animal Hazard PECO above), organ slices, organotypic culture, in vitro molecular or biochemical assay systems (ie, cell-free). Nonvertebrate and bacterial models used to study molecular or cellular effects (eg, bacterial reverse mutation [Ames] assay, Drosophila melanogaster DNA damage or repair response). Data from whole tissue or organ system exposures ex vivo. In silico modeling, data are relevant if they inform mechanism(s) of jet fuel-related health effects (eg, QSAR models for DNA reactivity).
Exposure	Jet fuels Other names: jet exhaust, jet engine exhaust, aircraft exhaust, aircraft engine exhaust, aircraft fuel, aviation fuel, aviation turbine fuel, jet propellant, aviation propellant, aircraft propellant, jet A fuel, jet A-1 fuel, Jet B fuel, TS-1 fuel, JP-1 fuel, JP1 fuel, JP-2 fuel, JP2 fuel, JP-3 fuel, JP3 fuel, JP-4 fuel, JP4 fuel, JP-5 fuel, JP5 fuel, JP-6 Fuel, JP6 Fuel, JP-7 Fuel, JP7 Fuel, JP-8 fuel, JP8 fuel, JP-9 fuel, JP9 fuel, JP-10 fuel, JP10 fuel, JPTS fuel, zip fuel, JP5 jet fuel, S-8 fuel, JP8 aviation fuel, JP4 aviation fuel, alternative jet fuel, synthetic jet fuel Any exposure to jet fuels, including in vitro, in vivo (by various routes of exposure, see Exposure Sections of Human Hazard and Animal Hazard PECO criteria above), and ex vivo. In silico studies will also be included if the model system can be linked to a PECO-relevant species and is associated with jet fuel exposure.
Comparator	Human: a comparison or referent population not exposed, or exposed to lower levels, of jet fuels. Animal, ex vivo, in vitro, in silico: a concurrent control group exposed to vehicle only or untreated control.
Outcome	Cellular, biochemical, and molecular changes related to jet fuel exposure and toxicity. This may include molecular initiating events or downstream key events that inform the mode of action linking jet fuel exposure to disease.

This PECO expands on the PECO used in a previous review.^22,33

TABLE 2.

Populations, Exposures, Comparators, and Outcomes Criteria for a Systematic Review on the Health Effects From Exposure to Jet Fuel within Animal Toxicological Studies

PECO Element	Inclusion Criteria
Population	Vertebrates (eg, mammals, avian species, fish, reptiles) of any life stage (preconception, in utero, lactation, peripubertal, and adult stages).
Exposure	Jet fuels Other names: jet exhaust, jet engine exhaust, aircraft exhaust, aircraft engine exhaust, aircraft fuel, aviation fuel, aviation turbine fuel, jet propellant, aviation propellant, aircraft propellant, jet A fuel, jet A-1 fuel, Jet B fuel, TS-1 fuel, JP-1 fuel, JP1 fuel, JP-2 fuel, JP2 fuel, JP-3 fuel, JP3 fuel, JP-4 fuel, JP4 fuel, JP-5 fuel, JP5 fuel, JP-6 Fuel, JP6 Fuel, JP-7 Fuel, JP7 Fuel, JP-8 fuel, JP8 fuel, JP-9 fuel, JP9 fuel, JP-10 fuel, JP10 fuel, JPTS fuel, zip fuel, JP5 jet fuel, S-8 fuel, JP8 aviation fuel, JP4 aviation fuel, alternative jet fuel, synthetic jet fuel Any exposure to jet fuels via any routes, including oral, inhalation, dermal, injection, or unknown/multiple routes. Any length of exposure. Co-exposures to chemical or physical exposures typically found during military service (eg, burn pits, noise) will be considered relevant.
Comparator	A concurrent control group exposed to vehicle only or untreated control.
Outcome	All health outcomes (both cancer and noncancer).

This PECO expands on the PECO used in a previous review.^22,33

Literature Searches

Four categories of bibliographic sources were used for this review: traditional scientific databases, gray literature, expert-identified publications, and references identified through a reference crosswalk. The aim of the literature search was to identify studies relevant to the three core evidence streams assessed in this review: epidemiologic, animal toxicological, and mechanistic studies. These study types were defined as primary evidence and were prioritized for full-text screening, data extraction, study quality evaluation, and weight-of-evidence synthesis.

PubMed (National Library of Medicine) and EBSCOhost (EBSCO Information Services) were identified as major literature repositories to search for primary scientific and medical evidence. The databases were searched using the search strings provided in Supplemental Digital Content, Table S2 (http://links.lww.com/JOM/C263). These searches targeted the title and abstract fields unless otherwise noted. A first search for occupational epidemiologic literature was conducted in January 2023,²² followed by a broadened search to identify animal toxicological and mechanistic data as well as any additional epidemiologic data in September 2023. Finally, an updated search was conducted on March 26, 2025. References retrieved from the literature searches were de-duplicated and imported into Litstream software (ICF).

In parallel with searches of the peer-reviewed literature, searches were conducted to collect relevant domestic and international gray literature from nonperiodical sources. These searches were performed using automated keyword searches across databases maintained by the Defense Technical Information Center, Occupational Safety and Health Administration, World Health Organization, and others. The keyword search string and full list of sites searched are also provided in Supplemental Digital Content, Table S2 (http://links.lww.com/JOM/C263).

Although the literature search strategies were designed to be broad, some relevant references may have been missed. This could occur, for example, if the exposure of interest was not mentioned in the title, abstract, or keyword fields, or if gray literature was not indexed in the searched databases or web sites. To reduce the likelihood of missing pertinent studies, subject matter experts (SMEs) nominated key publications and reports for inclusion in the review. Additionally, a reference crosswalk was conducted using select reviews to identify references that may not have been captured by the database searches or SME recommendations. Seven reviews were included for this crosswalk.^{3,6,20,34–37} The crosswalk assessed whether references within these reviews had already been identified at the title/abstract level. Any references that had not been previously identified were further evaluated for PECO relevance to ensure that no critical studies were overlooked.

Literature Screening

References retrieved from the literature searches were de-duplicated and imported into Litstream software for title and abstract screening based on PECO relevance. Each reference deemed PECO-relevant during title and abstract screening proceeded to full-text review, where the study report and any supplemental materials were evaluated to confirm relevance. All non-English references were excluded. References with uncertain relevance following title and abstract screening underwent full-text retrieval and further assessment. PECO relevance screenings were conducted by two independent reviewers per reference using structured forms in Litstream, see Supplemental Digital Content, Tables S3 and S4 (http://links.lww.com/JOM/C263). References determined to be irrelevant during screening were excluded from further review. In cases where the two reviewers disagreed on PECO relevance, a third review was conducted by an SME to resolve discrepancies.

In addition to the primary evidence streams, other reference types—such as case reports, case series, and reviews (secondary evidence)—were tagged as supplemental evidence. These supplemental sources were considered in the broader synthesis of findings to provide contextual information or highlight plausible biological mechanisms. However, they did not contribute to formal weight-of-evidence determinations. Reviews that summarized toxicological studies using animal models or mechanistic data were referenced but were not independently evaluated as primary sources for health effect conclusions.

Data Extraction

A standardized data extraction process was applied to all included references to ensure consistency, accuracy, and appropriate classification of study findings. Each reference was classified by type (eg, primary epidemiologic, animal toxicological, or mechanistic studies, meta-analysis, case study, case report, or case series), and relevant data were extracted using structured forms in Litstream. Health effects were categorized into one or more of 18 predefined outcome groups, and specific jet fuels assessed were documented. All health outcomes were extracted regardless of effect magnitude, statistical significance, or study quality. For case studies, case reports, secondary reviews, and mechanistic studies, only qualitative information was captured.

One reviewer conducted data extraction, with independent verification by at least one additional reviewer to verify accuracy. Any conflicts or discrepancies in data extraction were resolved through discussion with SMEs and the evaluation team, ensuring accuracy. Data extraction was performed using structured forms in Litstream, and extracted data elements are available in Supplemental Digital Content, Tables S5–S9 (http://links.lww.com/JOM/C263). Briefly, data extracted from each epidemiologic study included details on the population, location, study design, jet fuel type, and health outcomes assessed. For animal toxicological studies, data extracted included species, sex, study design, jet fuel type, exposure route, life stage when exposure and assessment(s) occurred, and health endpoints assessed.

In some instances, multiple references described the same epidemiologic or animal toxicological evidence. To prevent double counting, these references were identified as overlapping datasets and assigned “parent” and “child” designations. The parent reference was selected based on several factors, including peer-review status, sample size, accuracy of outcome measures, and comprehensiveness of reporting. Only the parent reference underwent data extraction and study quality evaluation (described below), while the child references were tracked for reference throughout the review process, but not extracted, ensuring each unique study was counted only once. In cases where the child reference provided unique information not provided in the parent reference, that information was extracted and evaluated as a unique data point and is cited accordingly. Otherwise, only the parent reference was considered during evidence synthesis. This approach minimized the risk of misinterpreting the quantity of available evidence for a given health outcome.

For animal toxicological studies, some references reported findings from the same animal cohort across multiple publications but addressed different endpoints or outcomes. These references were classified as shared datasets. Unlike overlapping studies, all references representing shared datasets were included in data extraction. This approach ensured that clustered references were appropriately considered during evidence synthesis, allowing for a comprehensive comparison and review.

The extracted data were used to create an interactive systematic evidence map by evidence stream in Tableau to provide an inventory of the relevant literature, following approaches also used by EPA.^23,38 Separate heatmaps depict epidemiologic and animal toxicological references. The heatmaps allow users to investigate the availability of evidence by body system and jet fuel type, as well as study quality (see below) and other study characteristics and to identify evidence gaps.

Study Quality Evaluation

All PECO-relevant, primary, epidemiologic, and animal toxicological studies underwent study quality evaluation, independent of the direction or magnitude of their findings. Key evaluation considerations included potential sources of bias, reflecting the assessment of internal validity (ie, factors that might affect the magnitude or direction of an effect) and sensitivity (ie, factors that limit the ability of a study to detect a true effect; low sensitivity introduces a bias toward the null when an effect exists). These considerations were assessed across multiple study domains. No study was excluded from the review based on concerns about the risk of bias.

Two experienced team members conducted the study quality evaluations. A reviewer assessed the reliability of the study results. A senior second reviewer performed a critical evaluation of the primary reviewer’s judgments and made final determinations, a deviation from the IRIS handbook³¹ for efficiency. Conflicts or discrepancies in study quality evaluations were resolved through consensus-based discussion within the evaluation team. Study quality evaluation domains for epidemiologic studies included participant selection, exposure measurement, outcome ascertainment, potential confounding, analysis, selective reporting, and study sensitivity. For animal toxicological studies, evaluation domains included reporting; allocation; observational bias/blinding; confounding/variable control; reporting and attrition bias; chemical administration and characterization; exposure timing, frequency, and duration; endpoint sensitivity and specificity; and results presentation. For each domain, the reviewer assigned a rating of good, adequate, deficient, or critically deficient (Table 4 and Supplemental Digital Content, Tables S10–S27, http://links.lww.com/JOM/C263). For animal studies, an additional rating of “not reported” was available. This rating was interpreted as adequate if the missing information not reported was unlikely to have a severe or notable impact on the results. However, it was interpreted as deficient if the missing information was likely to have a substantial impact on the results or their interpretation. Following the domain-specific assessments, overall study confidence (ie, high, medium, low, or uninformative) was determined based on the identified strengths and limitations (Table 5 and Supplemental Digital Content, Tables S17 and S27, http://links.lww.com/JOM/C263). For studies that examined multiple health outcomes, the evaluation process was outcome-specific, because study quality could vary depending on the outcome assessed. For example, an epidemiologic study could be considered medium confidence for health outcomes assessed using a validated test (if outcome ascertainment was rated as adequate) but considered uninformative for outcomes assessed via self-report (if outcome ascertainment was rated as critically deficient). Thus, variation in any domain could lead to different confidence ratings for individual outcomes, noted in the results as mixed judgments for a domain within a study.

TABLE 4.

Possible Domain Ratings for Study Quality Evaluation

Good	Intended to represent a judgment that there was appropriate study conduct relating to the domain and any minor deficiencies that were noted would not be expected to influence interpretation of the study findings.
Adequate	Indicates a judgment that there were study design limitations relating to the domain, but that those limitations are not likely to be severe and are expected to have minimal impact on interpretation of the study findings.
Reported (interpreted as adequate)	Indicates that the information necessary to evaluate the domain was not available in the study, but any limitations are not expected to be severe and are expected to have minimal impact on study finding interpretation. This rating is applicable to animal toxicological studies only.
Deficient	Denotes identified biases or limitations that are interpreted as likely to have had a substantial impact on the results or that prevent reliable interpretation of the study findings.
Not reported (interpreted as deficient)	Indicates that the information necessary to evaluate the domain was not available in the study, but any limitations are expected to have had a substantial impact on the results or to prevent reliable interpretation of the study findings. This rating is applicable to animal toxicological studies only.
Critically deficient	Reflects a judgment that the study design limitations relating to the domain introduced a flaw so serious that the study should not be used without exceptional justification (eg, it is the only study of its kind and may highlight possible research gaps). This judgment should only be used if there is an interpretation that the limitation(s) would be the primary driver of any observed effect(s), or if it makes the study findings uninterpretable.

These descriptions are provided in Environmental Protection Agency’s Integrated Risk Information System (IRIS) Handbook.³¹

TABLE 5.

Overall Study Confidence Classifications

High confidence	No notable concerns were identified (eg, most or all domains rated good).
Medium confidence	Some concerns are identified but expected to have minimal impact on the interpretation of the results (eg, most domains rated adequate or good; may include studies with deficient ratings if concerns are not expected to strongly impact the magnitude or direction of the results). Any important concerns should be carried forward to evidence synthesis.
Low confidence	Identified concerns are expected to significantly impact the study results or their interpretation (eg, generally deficient ratings for one or more domains). The concerns leading to this confidence judgment must be carried forward to evidence synthesis.
Uninformative	Serious flaw(s) make the study results unusable for informing hazard identification (eg, generally critically deficient rating in any domain; many deficient ratings).

These descriptions are provided in Environmental Protection Agency’s Integrated Risk Information System (IRIS) Handbook.³¹

Bold emphasis applied as intended.

To the extent possible, the overall confidence rating reflected an interpretation of the potential influence of study limitations on the results, including possible impact on the direction and/or magnitude of results. The rationale for the overall confidence rating was recorded and a brief summary of key strengths and limitations was provided. The limitations identified as part of the study quality evaluation were incorporated into the synthesis of findings for each body of evidence related to a given health effect. The summary for each study evaluation, including risk-of-bias judgments, is available in the Tableau dashboard at: https://public.tableau.com/app/profile/vha.home/viz/SupportingInformationforVHAJetFuelsReport_17002413903760/ReadMe?publish=yes. Case reports, case series, secondary reviews, overlapping “child” references, and mechanistic studies were not included in the study quality evaluation.

Evidence Synthesis and Integration

This review considered evidence synthesis and evidence integration as distinct but related processes. For each assessed health effect, the evidence synthesis summarized the body of available studies and incorporated conclusions from individual study quality evaluations. Syntheses of human and animal health effects focused on studies with high and medium confidence, while findings from low confidence results were given lower priority. However, in certain cases, such as when few or no higher confidence studies were available for a given health outcome, low confidence studies were considered to help evaluate consistency. Additionally, when low confidence studies addressed specific gaps or uncertainties in the higher confidence evidence base, the findings from these studies were also considered as sources of information for hazard identification. Because of the limited availability of epidemiologic evidence, studies of all confidence levels (high, medium, low, and uninformative) were included in the synthesis; however, uninformative studies were excluded from the synthesis of animal evidence in alignment with the IRIS handbook.³¹

The available epidemiologic and animal toxicological evidence on the potential health effects of jet fuels was synthesized separately, and a summary discussion was developed for each evidence stream. Mechanistic evidence was also incorporated into the synthesis when relevant. Strength-of-evidence judgments were made for each health outcome within each evidence stream (epidemiologic and animal toxicological) using standard terminology and definitions (ie, robust, moderate, slight, indeterminate), following the framework described in the IRIS Handbook.³¹ These judgments considered several factors, such as consistency across studies, the presence of a biological gradient (dose-response relationships), and coherence of findings across different evidence streams applying the causality approach introduced by Austin Bradford Hill.³⁹

Evidence integration combined the epidemiologic and animal toxicological evidence synthesis judgments and accounted for various factors, such as the human relevance of animal findings, coherence across evidence streams, information on susceptible populations or life stages, biological plausibility, mode-of-action, and other critical inferences informed by mechanistic and supplemental evidence. Mechanistic evidence, including in vitro and ex vivo studies, was examined for insights into potential biological pathways underlying observed health outcomes, thereby strengthening overall conclusions. These studies were categorized according to signaling pathways relevant to each outcome.

Integrated weight of the evidence judgments were developed through a structured review using established causality considerations. Categories and definitions are available in Supplemental Digital Content, Tables S28–S30 (http://links.lww.com/JOM/C263). These considerations included risk of bias, sensitivity, consistency, effect magnitude and precision, biological gradient (dose-response), coherence, and mechanistic evidence related to biological plausibility. In a deviation from the EPA IRIS protocol,³¹ this review also considered possible associations between exposure duration and health outcomes and between acute and chronic effects in the synthesis of epidemiologic studies.

During evidence integration, the strength of evidence judgments from the epidemiologic and animal toxicological studies were considered, alongside relevant mechanistic or mode-of-action evidence. Final evidence integration conclusions summarized whether the overall body of evidence indicated that jet fuel exposure has the potential to cause adverse health effects in humans.

RESULTS

Screening Overview

The results of the study identification process are summarized in Figure 1. The literature search yielded 3482 references from published literature databases and 476 gray literature references. Expert identification contributed an additional 87 references, and the cross-reference process added 246 references, resulting in a total of 4291 references. Of these, 621 references were deemed relevant during title and abstract screening. Full-text screening identified 279 references that met the inclusion criteria. Some of these references reported mechanistic endpoints in addition to epidemiologic or animal toxicological findings. The final total of relevant studies derived from the 279 references included 42 primary epidemiologic studies, 17 case reports or case series, 125 animal toxicological studies, 113 mechanistic studies, and 46 secondary resources. After accounting for overlapping references that describe the same data, 36 unique epidemiologic and 118 unique animal toxicological studies were identified for inclusion in the systematic review.

FIGURE 1.

Reference flow diagram of the search, screening, and selection of studies included in the review. This figure provides an overview of the study identification process results. The literature search yielded 4291 references. After completing title and abstract screening and full-text screening, 279 references were deemed relevant, with 42 epidemiologic references and 125 animal toxicological references. ^a A study may have reported on multiple model types; therefore, the total for the category was greater than the total relevant references. ^b Six references reported on the same epidemiologic study. After accounting for multiple references, there were 36 unique epidemiologic studies in this body of literature. ^c Seven references reported on the same animal toxicological study. Seven references were classified as partial overlaps as they reported on a portion of the same health outcome endpoints as another reference; however, unique endpoints reported were considered individually. After accounting for multiple references, there were 118 unique animal toxicological studies in this body of literature.

Most of the references (n = 3868) retrieved in the search were deemed not relevant. Many excluded references were case reports focused on kerosene-related accidents or incidents, air pollution from kerosene combustion in households, or kerosene storage practices unrelated to occupational, military, or environmental exposure scenarios. Other excluded references described the physical properties of jet fuels or engineering advances without addressing health effects. Additionally, abstracts and references published in languages other than English were also excluded.

Data from all steps of the systematic review process, including excluded references, were exported from Litstream into Microsoft Excel and visualized using Tableau (Seattle, WA). The Tableau Dashboard features a diagram following the Interactive REFerence Flow (I-REFF) approach to promote transparency and traceability of literature review results,⁴⁰ study quality evaluation heatmaps with detailed rationales for confidence scores for epidemiologic and animal toxicological studies, and evidence maps for all study types (epidemiologic, animal toxicological, and mechanistic). The visuals are available for download from Tableau’s interactive dashboard: https://public.tableau.com/app/profile/vha.home/viz/SupportingInformationforVHAJetFuelsReport_17002413903760/ReadMe?publish=yes.

Characteristics of Relevant Literature

Epidemiologic Studies

The review identified 36 unique epidemiologic studies assessing five types of jet fuel: JP-4, JP-5, JP-8, Jet A, and “other” (Table 6). The “other” category included studies in which the specific fuel type was either not identified or was a variation not commonly used by the US military. Overall, JP-5 and Jet-A were the least frequently assessed among epidemiologic studies, while JP-8 and the “other” category each accounted for approximately one-third of the total (Table 6). Some studies reported findings related to multiple fuel types or exposure characteristics. Therefore, these categories are not mutually exclusive, and in some cases, the total number of fuel-specific assessments exceeded the number of unique studies.

TABLE 6.

Study Characteristics Summary of Unique Primary Epidemiologic Studies by Jet Fuel Exposure

Study Characteristicsa	Jet Fuel Type
	JP-4	JP-5	JP-8	Jet A	Otherb
Overall study qualityc
High	0	0	0	0	0
Medium	1	0	6	0	1
Low	4	2	5	0	10
Low, uninformative	0	0	1	0	0
Uninformative	2	3	2	1	5
Sex
Both	2	1	10	1	3
Female	2	3	2	0	1
Male	6	0	2	1	11
Not reported	0	1	1	0	3
Military status
Military	6	5	12	1	6
Nonmilitary occupational	1	0	0	0	8

^aOne reference may report on multiple study characteristics or jet fuels so these categories are not mutually exclusive; therefore, the total number of references for each category may be greater than the total number of references.

^bIncludes nonpriority jet fuels.

^cBecause study quality is evaluated for each endpoint reported and one reference could have multiple endpoints that were given different overall confidence ratings, one reference could have multiple overall confidence ratings.

None of the epidemiologic studies were considered high confidence. Most studies were low confidence or uninformative studies, with concerns included potential for selection bias due to inadequate recruitment details, outcome misclassification from self-reported symptoms, residual confounding, or limited study sensitivity. Among medium confidence studies, JP-8 was the most frequently assessed jet fuel (n = 6). JP-4 and “other” each had one medium confidence study, whereas studies on JP-5 and Jet-A were low confidence or uninformative. Most jet fuel exposure assessments focused on military personnel, with males being the most represented study population.

Animal Toxicological Studies

The literature review identified 118 relevant animal toxicological studies that assessed exposure to six types of jet fuel: JP-4, JP-5, JP-8, Jet A, synthetic jet fuel, and “other” (Table 7). The “other” category included studies where the specific fuel type was either not identified or was an uncommon variant. Similar to the epidemiologic studies, JP-8 was the most commonly studied fuel. JP-5 and Jet-A were underrepresented in the reviewed studies. Very few studies assessed chronic, reproductive, or developmental exposures. Instead, most studies were designed to assess the effects of short-term exposures in mature animals. Additionally, when both sexes were not assessed, male animals were more commonly evaluated. The effects of jet fuel exposure were studied across multiple species, with rats, mice, dogs, pigs, and rabbits being the most frequently assessed. Exposures occurred primarily through inhalation, dermal, or oral routes.

TABLE 7.

Study Characteristics Summary of Unique Primary Animal Toxicological Studies by Jet Fuel Exposure

Study Characteristicsa	Jet Fuel
	JP-4	JP-5	JP-8	Jet-A	Synthetic Jet Fuelb	Otherc
Overall study qualityd
High	0	1	5	1	2	8
High, medium	0	1	3	0	1	3
High, low	0	0	1	0	0	0
High, medium, uninformative	0	0	0	0	0	2
High, medium, low, uninformative	0	0	0	0	0	0
Medium	4	4	33	10	3	14
Medium, low	1	2	9	2	0	2
Medium, low, uninformative	0	1	2	0	0	0
Medium, uninformative	1	1	1	1	0	0
Low	8	4	26	5	0	8
Low, uninformative	5	0	5	1	0	2
Uninformative	5	2	11	2	0	3
Sex
Female	6	3	30	7	3	10
Male	10	8	40	4	4	18
Both	7	5	17	1	2	9
Not specified	3	1	12	2	0	6
Species
Rat	9	6	37	4	2	13
Mouse	7	3	37	9	2	10
Dog	2	1	0	0	0	2
Pig	0	0	3	2	0	2
Rabbit	1	1	6	0	1	8
Othere	4	3	3	0	0	4
Study design f
Acute	2	3	24	4	3	11
Short-term	5	7	49	9	2	15
Subchronic	5	1	10	0	1	4
Chronic	3	2	3	4	0	3
Developmental	0	0	5	0	0	0
Reproductive	0	1	1	0	0	0
Route of exposure
Inhalationg	8	5	41	3	3	16
Dermal	2	1	29	11	2	13
Oralh	1	4	8	0	0	1
In utero/placental transfer	0	1	3	0	0	0
Rearing water	2	1	0	0	0	1
Otheri	2	1	3	0	1	4

^aOne reference may report on multiple study characteristics or jet fuels so these categories are not mutually exclusive; therefore, the total number of references for each category may be greater than the total number of references.

^bSynthetic jet fuel, also known as sustainable aviation fuel or synthetic kerosene, was categorized as any jet fuel that does not utilize fossil fuels.

^cIncludes nonpriority jet fuels.

^dBecause study quality is evaluated for each endpoint reported and one reference could have multiple endpoints that were given different overall confidence ratings, one reference could have multiple overall confidence ratings.

^eSpecies includes domestic animals (eg, pet), fathead minnows, guinea pigs, hamsters, mini-pigs, and Top Smelt Larvae.

^fStudy designs were determined as following: acute (<24 hours), short-term (1–30 days), subchronic (30–90 days), chronic (>90 days), developmental (exposure occurs during gestation and endpoints are measured in the fetuses or pup), and reproductive (study begins treatment before mating and continues through birth and in some cases through a second generation).

^gInhalation exposure includes nose only, whole body, and whole head exposure.

^hOral exposure includes drinking water and lavage.

ⁱAlternative route of exposure includes Intraperitoneal injections.

The study assessment also included a review of exposure duration. The study designs were categorized based on total exposure duration into the following periods: acute (<24 hours), short-term (1–30 days), subchronic (30–90 days), chronic (>90 days), developmental (exposure occurring during gestation, with endpoints measured in the fetuses or pups), and reproductive (exposure beginning before mating and continuing through birth, and in some cases, through a second generation). These classifications were applied to facilitate comparisons across studies and to evaluate health outcomes associated with different exposure windows.

Since study quality was evaluated for each endpoint reported in primary animal toxicological studies. For studies that assessed multiple endpoints, each endpoint was assigned a different confidence rating. Furthermore, because a single study could evaluate the same endpoint across multiple fuel types, the confidence rating for that endpoint could vary within a single reference, leading to multiple overall confidence ratings within a single study. Eight studies were identified in which all assessed endpoints were rated as high confidence.

Health Effect Outcomes Assessed in Relevant Studies

This fit-for-purpose, systematic literature review categorized health outcomes into 19 major categories (Supplemental Digital Content, Table S5, http://links.lww.com/JOM/C263). The categories were neoplasia, renal, respiratory, dermal, nervous, digestive, hepatic, immune, cardiovascular, male reproductive, female reproductive, hematologic, dental, developmental, endocrine, metabolic, musculoskeletal/connective tissue, systemic, and “other.” The “other” category encompassed all health outcomes that did not fit within these predefined categories. Nervous system endpoints were further divided into two distinct categories: neurologic outcomes (eg, sensory-related effects, motor coordination, deficits in memory) and cognitive and behavioral health outcomes (eg, mental health conditions, changes in attention, cognitive decrements). Each health effect category was further analyzed based on study design, population characteristics, and exposure details, ensuring a comprehensive evaluation of health outcomes.

Notable patterns in the evidence base emerged across certain categories. For example, respiratory and nervous system effects were among the most frequently reported across all study types. Findings on the weight of evidence as it relates to certain health outcomes are examined in greater detail in companion articles.^24–30 However, across the evidence base, there is a notable absence of data to allow for the comparison of effects associated with different durations of exposure or data to examine potential correlation between immediate symptoms (acute effects) and longer-term outcomes.

Epidemiologic Studies

The health outcome categories assessed in the epidemiologic studies are presented in Table 8. The five most frequently studied health outcomes included effects on the nervous system (n = 18), occurrence of neoplasms (n = 10), effects on the respiratory system (n = 9), effects on the digestive system (n = 8), and effects on cognitive and behavioral health (n = 7). Nine additional body systems were addressed in only one to four studies. Additionally, six health effect outcome categories were not examined in any of the identified epidemiologic studies.

TABLE 8.

Health Effect Outcome Category Summary of Unique Primary Epidemiologic Studies by Jet Fuel Type

Health Effect Outcome Categorya	Jet Fuel Type					Total Unique Studies
	JP-4	JP-5	JP-8	Jet A	Other
Neoplasm	3	0	0	0	7	10
Nervous	2	5	7	1	5	18
Respiratory	2	2	1	0	5	9
Digestive	0	3	1	0	4	8
Cognitive and behavioral health	0	2	3	0	2	7
Dermal	0	2	1	0	1	4
Cardiovascular	0	2	0	0	2	4
Renal	0	1	2	0	1	4
Hepatic	0	0	2	0	1	3
Immune	0	0	2	0	1	3
Reproductive (female)	0	0	2	0	0	2
Hematologic	0	0	1	0	0	1
Reproductive (male)	1	0	0	0	0	1
Dental	0	0	0	0	0	0
Developmental	0	0	0	0	0	0
Endocrine	0	0	0	0	0	0
Metabolic	0	0	0	0	0	0
Musculoskeletal/connective Tissue	0	0	0	0	0	0
Systemic	0	0	0	0	0	0
Otherb	1	2	0	0	1	4

^aOne reference may report on multiple study characteristics or jet fuels so these categories are not mutually exclusive; therefore, the total number of references for each category may be greater than the total number of references.

^bHealth outcome endpoints that do not fit within any of the other health effect outcome categories.

Animal Toxicological Studies

The characteristics of the animal toxicological studies are summarized in Table 9. The literature review identified seven health outcomes that were assessed in at least 30 studies: systemic effects (n = 52), immune system effects (n = 50), respiratory system effects (n = 44), dermal effects (n = 43), nervous system effects (n = 37), hepatic effects (n = 34), and renal effects (n = 32). Neoplasia was assessed across 12 unique references, placing it among the least frequently evaluated outcomes with the exception of dental effects (n = 2) and “other” (n = 2), all remaining categories were analyzed in at least eight individual studies.

TABLE 9.

Health Effect Outcome Category Summary of Unique Primary Toxicological Studies by Jet Fuel Type

Health Effect Outcome Categorya	Jet Fuel Type						Total Unique References
	JP-4	JP-5	JP-8	Jet A	Synthetic Jet Fuelsb	Other
Neoplasm	4	2	2	3	0	2	12
Systemic	9	10	25	6	3	17	52
Immune	7	4	27	7	3	10	50
Respiratory	11	5	19	2	3	11	44
Dermal	7	6	23	8	2	16	43
Nervous	7	7	19	3	1	9	37
Hepatic	9	4	11	3	2	8	34
Renal	8	4	10	3	2	8	32
Hematologic	5	2	6	1	1	8	22
Digestive	5	5	4	1	1	5	21
Cardiovascular	4	4	4	1	2	6	21
Endocrine	3	4	3	1	2	6	19
Reproductive, male	4	5	2	0	2	4	17
Reproductive, female	4	3	3	1	1	4	16
Metabolic	5	1	3	0	0	5	14
Musculoskeletal/connective Tissues	3	3	2	0	0	1	9
Developmental	0	1	7	0	0	0	8
Dental	1	0	1	0	0	0	2
Otherc	0	0	1	1	0	0	2

^aOne reference may report on multiple study characteristics or jet fuels so these categories are not mutually exclusive; therefore, the total number of references for each category may be greater than the total number of references.

^bSynthetic jet fuel, also known as sustainable aviation fuel or synthetic kerosene, was categorized as any jet fuel that does not utilize fossil fuels.

^cHealth outcome endpoints that do not fit within any of the other health effect outcome categories.

Mechanistic Studies

The characteristics of the mechanistic studies are summarized in Table 10. The literature review identified 113 relevant mechanistic studies, which examined six types of jet fuel exposures: JP-4, JP-5, JP-8, Jet A, synthetic jet fuel, and “other.” As observed in other evidence streams, JP-8 was the most frequently studied fuel type. The identified mechanistic studies included three types of models: animal (n = 57), in vitro/in silico/ex vivo (n = 56), and human (n = 4). The six most frequently assessed health outcomes included effects on the immune system (n = 31), dermal effects (n = 28), effects on the respiratory system (n = 26), occurrence of neoplasia (n = 15), effects on the hepatic system (n = 9), effects on the nervous system (n = 9), and systemic effects (n = 9). Seven other outcome categories were evaluated in one to eight studies each, while five categories were not assessed in any of the relevant mechanistic studies.

TABLE 10.

Study Characteristics Summary of Mechanistic Studies by Jet Fuel Exposure

Study Characteristicsa	Jet Fuel Type						Total Unique References
	JP-4	JP-5	JP-8	Jet A	Synthetic Jet Fuelb	Otherc
Study type
Animal	0	3	48	8	6	5	57
In vitro/in silico/ex vivo	1	3	50	8	6	1	56
Human	1	0	3	1	0	1	4
Health effect outcome category
Neoplasm	1	1	9	2	5	1	15
Immune	0	2	29	6	4	0	31
Dermal	0	0	28	4	3	0	28
Respiratory	0	0	23	2	1	1	26
Hepatic	0	0	8	1	0	1	9
Nervous	0	1	8	1	0	2	9
Systemic	2	2	7	2	0	2	9
Renal	0	0	4	0	3	1	8
Developmental	0	1	5	0	0	0	6
Cardiovascular	0	0	3	0	0	0	3
Endocrine	0	1	1	0	0	1	1
Hematologic	0	0	1	0	0	0	1
Metabolic	1	0	1	1	0	0	1
Reproductive, male	0	0	1	0	0	0	1
Dental	0	0	0	0	0	0	0
Digestive	0	0	0	0	0	0	0
Musculoskeletal/connective Tissue	0	0	0	0	0	0	0
Reproductive, female	0	0	0	0	0	0	0
Otherd	0	0	0	0	0	0	0

^aOne reference may report on multiple study characteristics or jet fuels so these categories are not mutually exclusive; therefore, the total number of references for each category may be greater than the total number of references.

^bSynthetic jet fuel, also known as sustainable aviation fuel or synthetic kerosene, was categorized as any jet fuel that does not utilize fossil fuels.

^cIncludes nonpriority jet fuels.

^dHealth outcome that does not fit within any of the other health effect outcome categories.

DISCUSSION

Jet fuel exposure has been recognized as a significant environmental hazard in military settings, with potential acute and chronic health effects that have been incompletely characterized.²² By adapting a well-established systematic approach for characterizing the health hazards of environmental chemicals,³¹ this fit-for-purpose systematic review adhered to rigorous scientific standards, promoting transparency, while also providing a comprehensive assessment of the available data on the health effects associated with jet fuel exposure. The review identified potential associations between jet fuel exposure and a broad range of adverse health outcomes.^24–30 However, the strength of evidence varied, with some outcomes supported by a moderate evidence base of epidemiologic and toxicological data, while others lacked sufficient information to establish clear causal relationships. The review also highlighted gaps in the existing literature that limit the current understanding of the relationship between exposure duration and chronic health outcomes and the potential role of acute symptoms as predictors of long-term illness.

This review included a broad and systematic literature search that identified epidemiologic studies, animal toxicity data, and mechanistic research from various sources, including US and international databases published through March 2025. Key sources included government repositories, such as the DoD-owned DTIC, as well as publicly available peer-reviewed journals. Studies meeting the PECO framework were selected for further analysis. This review included both US and international studies to support a comprehensive evaluation of findings from diverse occupational and environmental settings.

Because of the limited number of epidemiologic studies identified, all eligible human studies were incorporated in the evidence synthesis, regardless of their confidence level, with uninformative studies being considered as supplementary information rather than primary sources of evidence. This approach differs from the traditional IRIS methodology, which typically synthesizes findings only from studies with low, medium, and high confidence ratings.³¹ However, a stricter evaluation framework following IRIS guidance without deviation was applied for animal toxicological data, ensuring that only studies with at least low confidence were integrated into the main synthesis.

Strengths

The methodology employed in this systematic review of the health risks associated with jet fuel exposure had several distinct strengths that enhanced its thoroughness and rigor. A notable feature was the broad scope of the literature search, which included peer-reviewed studies, gray literature, and expert-identified reports. By omitting date restrictions, the review was able to capture both historical and recent data, providing a comprehensive and longitudinal perspective on jet fuel exposure. This inclusive approach ensured that relevant sources, including those less accessible or not typically indexed in traditional scientific databases, were considered. As a result, the review compiled a diverse and extensive body of evidence that was sufficient for assessments of several outcomes.

Another strength was the application of three PECO criteria to adjust sensitivity for study inclusion or exclusion by evidence stream. Individualized PECO criteria were applied to epidemiologic, toxicological, and mechanistic evidence streams, allowing for the inclusion of a maximum number of relevant studies. This broadly inclusive approach was particularly important in addressing gaps in epidemiologic studies, especially within military populations. The structured inclusion of animal and mechanistic studies helped compensate for the limited availability of human data, ensuring that the review could derive meaningful conclusions regarding potential health risks.

The rigorous study quality evaluation process, adapted from the IRIS framework,³¹ further strengthened the methodology. Each study underwent assessment across multiple domains, including exposure measurement and confounding, with a focus on outcome-specific evaluations. This structured approach allowed for a more nuanced assessment of the reliability of each study, thereby enhancing the overall validity of the conclusions. Even studies with significant limitations contributed valuable insight if they met specific criteria in other domains. The structured assessment of study quality improved confidence in the findings of the review.

The integration of multiple evidence streams also bolstered the strength of the methodology. By synthesizing findings from epidemiologic, animal toxicological, and mechanistic studies, the review provided a comprehensive evaluation of potential health effects. This multi-stream integration was particularly valuable for addressing the limited availability of epidemiologic data on long-term health outcomes. The systematic integration of findings, assessment of biological plausibility, and evaluation of cross-stream coherence ensured that the conclusions were scientifically rigorous. This approach aligned with established best practices in environmental health risk assessment and reinforced the credibility of the conclusions drawn.

Challenges and Limitations

There are a few limitations related to the conduct of this systematic review. For efficiency, data extraction and study quality evaluations were performed by single reviewers with confirmation by a more senior reviewer as opposed to dual independent extractions and evaluations then verified by a third reviewer. It is possible this approach may introduce some errors. However, because the extracted data were used to categorize evidence, identify data gaps, were not intended for meta-analysis, and additional quality control was implemented throughout the reporting phases, it is unlikely that conclusions were significantly impacted.

This review identified several key gaps in the literature. The comprehensive search strategy yielded only 36 unique epidemiologic studies, none of which were rated as high confidence during study quality evaluation. Only eight studies received medium confidence ratings, while the remaining studies were classified as low confidence or uninformative due to methodological challenges that introduced uncertainty in their findings. Several studies did not perform statistical comparisons. Most epidemiologic studies categorized exposure groups based solely on job title rather than directly measuring exposure levels. Quantitative assessments of jet fuel exposure through occupational measurements (eg, personal sampling of total hydrocarbons) and biomarkers of exposure (eg, urinary 1-naphthol and 2-naphthol or benzene, toluene, ethylbenzene, and xylenes in exhaled breath), or exposure assignment based on job exposure matrices with supporting industrial hygiene data were considered in exposure assessment; however, this information was not available in the majority of studies included in this review and remains a limitation across studies of military personnel. Furthermore, few studies allowed for follow-up periods long enough to be able to assess long-term health outcomes that may occur months or years after exposure, with most studies evaluating health effects in actively working individuals concurrently exposed to jet fuels. Many studies examined exposures and outcomes at the same time, limiting the ability to make causal inferences regarding observed associations. Future high-quality studies are needed that prioritize longitudinal designs with sufficient follow-up to observe effects after exposure has ceased and to observe what effects persist or resolve over time. Furthermore, studies that examine effects based on the duration of jet fuel exposure and cohorts of occupational populations are warranted to advance the understanding of how the risk of adverse health outcomes changes with increased years of exposure.

The reviewed epidemiologic studies also did not adequately represent the broader veteran or military population. Data describing health outcomes in females were particularly lacking. Additional research is needed to confirm health outcomes associated with jet fuel exposure, especially in underrepresented groups. Additionally, important individual characteristics potentially associated with exposures and health outcomes, such as race and ethnicity, behaviors (eg, smoking, alcohol use), and co-morbid health conditions, were infrequently studied. Most epidemiology study participants were active workers or service members, who tend to be healthier than the general population; therefore, healthy worker bias may have obscured true associations between exposure and adverse health outcomes.⁴¹ Observational occupational studies are also susceptible to healthy worker bias, in which exposed individuals may leave their jobs due to illness, retirement, or death, thereby underestimating exposure-related risks. Finally, epidemiologic evidence assessing whether immediate symptoms or acute effects resulting from jet fuel exposure progress into more chronic health outcomes was absent. Studies are needed to determine whether immediate symptoms can be used as predictors of more serious health effects, which could improve preventive care for those with occupational exposures going forward.

Although animal toxicological studies included more high and medium confidence ratings than epidemiologic studies, the inherent challenge of extrapolating animal findings to human health outcomes remains.⁴² Beyond this issue, most of the animal toxicological studies utilized a variety of jet fuel types and exposure design paradigms, making it difficult to draw definitive conclusions. This was particularly challenging when the direction of effects varied across study conditions. Additionally, relatively few animal toxicological studies assessed health outcomes following chronic exposure to jet fuels, despite the fact that these exposure scenarios are the most relevant to occupational settings for humans.

The effects of individual types of jet fuels could not be adequately assessed due to the limited epidemiologic, animal toxicological, and mechanistic evidence on specific types. Furthermore, several studies did not identify the type of jet fuel being evaluated. Among those that did, JP-8 was the most frequently studied fuel type across all three major data streams, though the evidence was still lacking in quality and quantity. There was a notable lack of studies assessing JP-5 exposure, which may limit the applicability of the findings of this systematic review to US Navy personnel, as JP-5 is the primary aviation fuel used aboard naval vessels. In addition, none of the reviewed epidemiologic studies conducted thorough exposure assessments to account for co-exposures commonly found in those settings, such as solvents, that may contribute to the observed outcomes.

The reviewed case studies and case series primarily described short and intense exposures (eg, direct skin contact from fuel spills). While these reports provided interesting clinical observations, their findings are not generalizable to broader populations. Most of the case studies and case series indicated that symptoms diminished or resolved with treatment or removal of exposure and were not indicative of long-term, irreversible toxicity. As a result, these data offered limited insight into the persistence of health effects following exposure cessation and the potential for long-term disease development.

Lastly, there were several limitations in the evaluation and inclusion of mechanistic evidence within this review. Similar to the epidemiologic and animal toxicological evidence, mechanistic studies captured during the literature review spanned several jet fuel types, study designs, and models. The mechanisms of action underlying jet fuel-related health effects were not comprehensively investigated, contributing to uncertainty regarding how jet fuels induce adverse human health effects. Additionally, the available mechanistic evidence relied on a variety of in vitro and in vivo models, which further complicated the extrapolation of these health outcomes in determining human plausibility.

Future Implications

The findings of this systematic review provide a foundation for informing policies related to preventing and reducing jet fuel exposure, improving healthcare interventions, and guiding benefits eligibility assessments for military personnel and veterans. While epidemiologic evidence on long-term health effects is still limited, the review highlights potential health risks associated with jet fuel exposure across multiple body systems. These findings underscore the need for enhanced exposure monitoring, improved occupational safety measures, and expanded medical surveillance for affected populations. Additionally, the identification of data gaps, such as the long-term consequences of acute exposure events, emphasizes the importance of targeted research efforts to better characterize exposure-outcome relationships and inform future risk mitigation strategies.

Beyond the direct implications for exposure prevention and healthcare policy, this systematic review also offers an example of how the IRIS methodological framework³¹ can be applied to improve the understanding of potential health risks related to environmental exposures in military and occupational settings. The application of PECO criteria and thorough study quality evaluation strengthens the rigor, transparency, and consistency of the assessment, particularly in data-limited contexts. This adaptable approach supports more informed assessments of environmental and occupational hazards where human data may be scarce, supporting broader risk assessment and regulatory efforts.