9

Concluding Remarks

The committee was tasked with assessing possible relationships between toxic military exposures and mental, behavioral, and neurologic outcomes. The Department of Veterans Affairs (VA) directed the committee to focus on veterans who served in the Southwest Asia Theater of Operations or Afghanistan after September 11, 2001. Outcomes specified in the task include posttraumatic stress disorder, depression, psychosis, schizophrenia, bipolar disorder, chronic multisymptom illness (CMI), traumatic brain injury (TBI), neurocognitive disorders (e.g., dementia), attention-deficit/hyperactivity disorder, and suicide attempts and suicide deaths. The committee added adjustment disorders, substance use disorders, amyotrophic lateral sclerosis, multiple sclerosis, and Parkinson’s disease due to veteran interest in their etiology and based on committee expertise. The committee did not have data on suicide fatalities and instead assessed nonfatal suicide attempts and intentional self-harm. TBI, while included in the committee’s charge, was used as a covariate because it is caused by external force and not toxic exposures.

The committee interpreted “toxic exposures” to refer to military-related environmental and occupational exposures. Environmental exposures refer to hazardous exposures experienced at home, work, or during daily activities; occupational exposures refer to hazardous exposures sustained in work settings or while performing work activities. During military deployment, environmental and occupational exposures may be combined because service members live and work in the same or nearby locations. Specific exposure groups were identified through an iterative process refining the available data and are burn pits, dust and particulate matter (PM) (and other specific pollutants), jet and diesel exhaust, jet and

diesel fuels, incinerator emissions, metals (including lead and manganese), mold, radiation, and solvents. Although other military exposures such as pesticides, ultrafine particles, and per- and polyfluoroalkyl substances may be of interest due to their prevalence in military settings or their known risks for mental health or neurological outcomes, data on these exposures were insufficient or not available.

The committee approached its task by conducting data analyses and a structured literature review and then integrating evidence from these two sources to develop conclusions. For the analyses, the committee conducted a series of case-control studies to assess possible relationships between the exposures and outcomes of interest using exposure data from the Individual Longitudinal Exposure Record (ILER), a joint Department of Defense (DoD)–VA platform that documents service-related toxic exposures of veterans and service members, linked to health outcomes data from the Veterans Health Administration (VHA). This dataset comprised post-9/11 veterans who received a diagnosis at VHA between 2017 and 2023 and who had an ILER record. Importantly, this study sample only generalizes to veterans who have a DoD-recorded exposure and an outcome diagnosis at VA during 2017–2023; it does not generalize to all post-9/11 service members or veterans.

The committee recognizes many limitations of its analysis, resulting primarily from inadequacies and weaknesses in data quality and availability as related to the measurement of exposures and outcomes as well as gaps in data on potential confounders. Limitations of the exposure data stem from only having data on exposures recorded in ILER, which is incomplete, periodic, and inconsistent. In addition, these data are recorded at the area level, which may lead to misclassifying a given individual as exposed or not and does not permit assessment of an individual’s risk of developing a condition. These data also do not include reliable information on exposure concentration, duration, and frequency, which are necessary to assess concentration-response relationships. Measurement of health outcomes was limited by only having VHA data for the study period 2017–2023, which does not capture outcomes diagnosed outside of VHA or within VA before 2017. Additionally, relevant conditions may have remained undiagnosed, misdiagnosed, or otherwise miscoded, as is common in electronic health records, particularly for psychiatric disorders. The committee also lacked data on potential confounding variables such as preexisting conditions or trauma exposures, which are important to adjust for in order to isolate the unique effect of military exposures on health. These data limitations could either magnify or attenuate the observed relationships between the exposures and outcomes of interest.

Given these limitations in the data analysis along with the general limitations of relying on any singular analysis, the committee conducted

a structured (but not systematic) review of epidemiologic literature to contextualize the quantitative results. The committee was not tasked with assessing the weight of evidence and instead used this literature as existing evidence to support the data analysis. Notably, owing to the scarcity of available literature on military-specific exposures or among military and veteran populations, the committee also relied on literature examining analogous environmental and occupational exposures in the general population. The committee acknowledges that this literature from civilian settings may have limited relevance to military exposures. For example, exposures during deployment are often higher intensity, more frequent, and last for the duration of deployment. Thus, it is reasonable to expect that some exposures that occur in the general population may reflect the lower range of effects compared to military exposures. Relatedly, low-level and long-term environmental exposure such as to PM from air pollution may not produce the same consequences arising from higher intensity but shorter duration exposure such as to jet and diesel exhaust during military deployments. In addition, extrapolating literature from the general population to military contexts may not capture the complexity of the military experience. While both the general population and military populations are concurrently and sequentially exposed to multiple toxicants and other stressors, the added stressful aspects of military deployment may particularly enhance vulnerability.

To draw conclusions, the committee synthesized these two streams of evidence and made determinations about possible risk-conferring relationships. The results of the analyses and the literature review in some cases did not align. The committee considered four scenarios of results, which it categorized in two conclusions. The committee concluded that there was a possible risk-conferring relationship between a given exposure and a given outcome if (1) both the evidence from the analyses and the epidemiologic literature showed risk-conferring effects; (2) the evidence from the analyses showed a risk-conferring relationship and the literature was mixed, unclear, or otherwise inconclusive; or (3) the evidence from the analyses did not show a risk-conferring relationship, but the literature was suggestive. The committee concluded there was inadequate or insufficient evidence of a possible risk-conferring relationship between a given exposure and outcome if both the analyses and the literature did not show evidence of a risk-conferring relationship.

Based on the results of its analyses and literature review, the committee found that military exposures may be related to the subsequent development of mental, behavioral, and neurologic outcomes and CMI. Specifically, the committee concluded that possible risk-conferring relationships exist between dust and PM, exhaust, incinerator emissions, PM

alone,¹ or solvents and specific mental, behavioral, and neurologic health outcomes and CMI. The committee concluded, that inadequate/insufficient evidence of possible risk-conferring relationships exists for burn pits, fuels, metals, mold, or radiation and any outcome.

Despite the limitations of the committee’s analysis, it is possible to obtain a more comprehensive understanding of the health effects of military exposures via scientific research. For example, prior National Academies committees have established evidence of an association between airborne hazards encountered in the Southwest Asia theater and respiratory symptoms (NASEM, 2020) and industrial solvents in drinking water at U.S. Marine Corps Base Camp Lejeune, North Carolina, and health outcomes including cancers, reproductive outcomes, and neurobehavioral effects (NRC, 2009). The committee considered lines of inquiry that could be conducted with existing data but discontinued these efforts due to time and resource constraints. This included investigating suicide fatalities in addition to nonfatal suicide behaviors, accounting for differences in VHA utilization to understand how veterans’ frequency and type of engagement in VHA services are associated with opportunities to diagnose the outcomes, and evaluating potential mediators or effect modifiers such as homelessness. In addition, service members generally experience multiple toxic exposures rather than single exposures, and these multiple exposures may interact to result in enhanced toxicity. Thus, understanding cumulative risk of such exposures in relation to mental, behavioral, and neurologic health outcomes would be particularly important.

This report offers an exploratory step toward establishing which military exposures might increase risk of mental, behavioral, and neurologic health outcomes. The committee’s conclusions can be interpreted as signals of potentially meaningful relationships that may be of interest for additional investigation. When more and higher quality data are available, researchers will be able to conduct studies that better characterize the relationship between military exposures and mental, behavioral, and neurologic outcomes. Current gaps in the data include lack of reliable information on additional military exposures on known neurotoxicants (e.g., pesticides) and full pre-deployment exposure data as well as on confounders before military exposure such as preexisting health conditions and trauma exposures. Where data are available, improved quality would help to minimize sources of bias and error and enhance the precision and accuracy of

___________________

¹ Analytic results on exposure to dust and PM, exhaust, and incinerator emissions were supported by literature on PM. Conclusions relying on the data analyses alone or together with the literature refer to dust and PM, exhaust, and incinerator emissions; conclusions relying solely on the literature refer to PM alone. The committee did not make conclusions on inadequate/insufficient evidence on the literature alone, so there are no conclusions on whether inadequate or insufficient evidence exists for PM alone.

estimates of these associations. With respect to exposure data, improvements include more precise measurements that capture specific chemicals or toxicants and not only sources of exposures, which would facilitate deeper understanding of the biologic mechanisms by which military exposures cause disease. Similarly, information on the concentration, duration, and frequency of exposures would facilitate estimating concentration-response relationships. Furthermore, person-level measures such as those collected via biologic monitoring, biomarker measurement of exposures, or exposomics-based profiling would reduce misclassification associated with the use of area-based measures and increase statistical power. For improved outcomes data, use of verified health data would better capture health outcomes than diagnostic codes in administrative data. Data on symptoms in addition to diagnoses would enable understanding how military exposures not only cause new onset of disease but also exacerbate existing disease. Use of symptom-based measures would also improve measurement of CMI, which lacks a standardized case definition using diagnostic codes. Data on both exposures and outcomes not restricted to those recorded by DoD and VA would enable generalizability of findings to all post-9/11 service members and veterans.

Data improvements would allow for improved future research designs. For example, biomarker measurement of relevant exposures and verified health data on a sub-sample of veterans could facilitate deeper understanding of the observed relationships between coarse exposure categories and actual health measurements and also help with imputation and joint analysis in a two-phase design framework. For such a study, the verified health data could be collected from a sample of veterans that is not restricted to those with a documented exposure in ILER or who receive care at VHA, which may help to improve the generalizability of findings. Once associations between exposures and outcomes are identified through such a study, comprehensive environmental monitoring to gather data sources and routes of exposure could later inform prevention and intervention measures. In addition, multiple military exposures and mental, behavioral, and neurologic health outcomes overlap. A multi-morbidity mixtures analysis of the sequence of mental, behavioral, and neurologic outcomes and toxicant mixtures could illuminate the role of combinations of exposures and the orders in which a cascade of outcomes occurs in producing disease. Such future research could improve the accuracy and precision of estimates of the relationship between military exposures and mental, behavioral, and neurologic health outcomes. Futhermore, once there is strong evidence of a true association, rigorous research designs together with guidelines for evidence of causality between military exposures and health among veterans, supported by a deep understanding sub-group effects and heterogeneity, may permit future causal inference.

REFERENCES