Psychiatric Conditions and Symptoms After Toxic Environmental Exposures During Military Service

Abstract

Background:

US service members are often exposed to a range of service-related hazards. To date, there has been limited synthesis of the existing research conducted on military environmental exposures and subsequent psychiatric conditions and symptoms.

Objective:

To systematically review and characterize the main features of studies examining associations between military exposures and mental health outcomes.

Methods:

We used evidence mapping methodology to systematically search MEDLINE, Embase, PsycINFO, and PTSDpubs for studies of toxic exposure during military service and psychiatric outcomes, which included psychiatric diagnoses, psychiatric symptoms, and neurocognitive functioning.

Results:

We identified 49 studies; most were comprised of predominantly White, male veteran samples. Chemical exposures, including chemical munitions from the Gulf War era and Agent Orange from the Vietnam War era, were the most frequently examined military toxic exposures. Symptoms of depression, PTSD, and anxiety were the most commonly examined psychiatric outcomes. Only 9 studies assessed neurocognitive functioning. We found extensive variation in how exposures and outcomes were defined and measured. Most exposure and symptom data were based on self-reports. Overall, available evidence suggests that veterans reporting environmental toxic exposures may report relatively high levels of mental health needs.

Conclusions:

We found broad evidence that toxic exposure was associated with poorer mental health outcomes, though the ability to draw stronger conclusions is limited by the quality of the current literature. Future research should focus on longitudinal studies of toxic exposure and mental health that include more broadly representative military populations, including diverse samples and more recent service cohorts.

US service members are at risk of environmental exposures to airborne hazards (eg, burn pits, diesel fumes), radioactive materials, toxic biological and chemical agents (eg, pesticides, chemical weapons), and other occupational hazards.^1–4 While some military environmental exposures (MEE; eg, exposure to heat, infectious agents, noise, and chemical solvents) have been experienced across all military eras, other exposures are specific to a service era. For instance, Agent Orange is an herbicide used during the Vietnam War,⁵ and oil well fires and chemical agents (eg, sarin from the Khamisiyah Munitions Depot demolition) were exposures unique to the wars in the Persian-Gulf and Iraq and Afghanistan.^6,7 Gulf War service has also come to be associated with a wide variety of medically unexplained symptoms affecting cognitive, emotional, and physical functioning. This has come to be known as Gulf War Illness or Chronic Multi-Symptom Illness.^8,9

Military service members can be exposed to a wide range of toxic agents while on active duty, including diesel, jet propellants (eg, JP-8), lead, chemical solvents, fire retardants, and radiation.^10,11 Some of these exposures are specific to military theaters, whereas other MEEs are linked to military bases. For example, US service members stationed at Camp Lejeune were exposed to water contaminated with industrial solvents (eg, trichloroethylene, perchloroethylene) and benzene from leaking underground fuel storage tanks between 1953 and 1987. This exposure has been linked to female infertility, liver and esophageal cancers, Parkinson disease, and at least 12 other adverse health conditions.¹²

MEEs have been associated with a wide range of adverse health effects in studies of military service members. Most studies examined the physical or physiological effects of MEEs, and some reported genetic and epigenetic changes associated with MEEs.^13,14 Specific MEEs have been associated with the development of respiratory diseases, liver and kidney damage, blood disorders, movement disorders, and various types of cancer.^13,14 While the research on mental health (MH) and cognitive implications of these exposures is more scant, increasing evidence has linked military deployment-related MEEs to adverse mental health and cognitive effects.

There has been increased scientific and political attention to the links between MEE and postdeployment health over the past 2 decades,^3,4,15 culminating in the passage of the Sergeant First Class Heath Robinson Honoring Our Promise to Address Comprehensive Toxics (PACT) Act in 2022.¹⁶ The PACT Act expanded Veteran Affairs (VA) health care eligibility for veterans with medical conditions presumed to result from MEEs and mandated MEE screening of all veterans. To date, over 5 million veterans have been screened, with almost half (43%) reporting at least one MEE.¹⁶

The PACT Act represents the largest expansion of VA benefits in history, including MH- and counseling-related benefits, yet could increase the demand for MH-related services. Several prior studies have suggested that MEEs might be associated with poorer MH.^17,18 To date, no studies have synthesized the existing literature linking MEEs to MH outcomes. In this review, we systematically organize and describe peer-reviewed literature studying associations between MEEs and MH and neurocognitive outcomes (ie, diagnoses and symptoms). Our review focused on 2 key questions (KQs): for which MEEs has the co-occurrence of psychiatric conditions or symptoms been evaluated (KQ1)? What is the range of prevalence, incidence, or association of co-occurring psychiatric conditions or symptoms among individuals with a history of MEEs (KQ2)?

METHODS

This manuscript summarizes a VA-commissioned Evidence Synthesis Program (ESP) report on MEEs and MH outcomes.¹⁹ The full report was intended to provide detailed information for VA clinical teams and community care partners developing national standards for the screening, assessment, and treatment of veterans with MEE. This manuscript provides the most relevant data, oriented toward a broader research and clinical audience in a concise form.

Searching and Screening

We searched MEDLINE (via Ovid), Embase (via Elsevier), APA PsycINFO (via Ovid), and PTSDpubs (via ProQuest) through July 27, 2023, using terms for military personnel, MEEs, toxic exposures, and MH and substance use (Table 1, Supplemental Digital Content 1, http://links.lww.com/MLR/C985 for complete search strategies). Two investigators independently screened titles, abstracts, and full-text articles; disagreements were resolved by consensus.

Data Abstraction

An individual investigator extracted data from eligible studies; data were then overread by a second investigator. For KQ1, we extracted key characteristics of the study population (eg, military service era, age, branch), exposure (eg, types, duration, dose, time since exposure), and types of co-occurring psychiatric or neurocognitive conditions or symptoms measured. For KQ2, we extracted the reported effect size of the associations (eg, odds/risk ratios, measures of variability).

Synthesis

We organized our results into 3 broad categories, informed by the Diagnostic and Statistical Manual-5-Text Revision (DSM-5-TR) and clinical application. These groups included psychiatric diagnoses (eg, posttraumatic stress disorder, PTSD; major depressive disorder, MDD), psychiatric symptoms (eg, low mood and anxiety), as well as neurocognitive diagnoses (eg, major and mild neurocognitive disorder), and neurocognitive symptoms (eg, executive dysfunction and impaired memory). Neurocognitive symptoms and diagnoses were combined into a single “Neurocognitive” category due to the small number of studies identified. These groupings were designed to assist with the clinical utility and interpretation of the results. Supplemental Digital Content Table 2, http://links.lww.com/MLR/C986 describes the coding schema used for this evidence-mapping study.

For KQ1, we summarized key characteristics of the included studies, such as study design, population demographics, exposure characteristics, and outcomes (MH and MEEs). We then identified commonalities across the studies to summarize the literature and to identify which exposures and MH outcomes had limited or no available data (ie, gap analysis). We then used visual approaches to display key patterns (eg, exposures and outcomes) across included studies. For this evidence map, we did not aggregate estimates or generate summary effect estimates. For KQ2, we abstracted effect estimates when possible and categorized the magnitude of the effect as none, small, medium, or large, in line with established guidelines in the literature (Supplemental Digital Content Table 3, http://links.lww.com/MLR/C987).²⁰ We also grouped estimates for distinct psychiatric and neurocognitive conditions and symptoms by exposure type and reported the ranges of study effect sizes across these groups.

We coded exposures using the structure described in Supplemental Digital Content Table 4, http://links.lww.com/MLR/C988. Similar chemical types were grouped together (eg, studies reporting JP-8/diesel exposure) as there was a lack of information regarding exposures across studies, including the route of exposure (eg, inhalation, skin). We used the term “herbicides” to refer to studies that examined Agent Orange or other herbicides, and used the term “pesticides” when a study did not report what type of pesticide was used.

Many included studies reported composite outcomes (eg, combined neuropsychiatric symptoms and physical symptoms) as part of the Gulf War Illness construct. For our analysis of associations between exposures and conditions, we did not include these studies unless the neuropsychiatric symptoms were disaggregated in the primary study. We intended to define the co-occurrence of relevant MH outcomes, but not to establish causation. We mapped the patterns of associations across the included studies and provided visual displays of the exposure and co-occurrence of neuropsychiatric symptoms and disorders.

RESULTS

After removing duplicate citations, our search identified 1929 unique articles. After title and abstract screening, 161 potentially relevant articles remained. After screening the full text, 49 of these studies met eligibility criteria and were included in KQ1. Forty-one reported disaggregated psychiatric or neurocognitive outcomes relevant to KQ2 (Fig. 1), of which 22 studies reported neurocognitive symptoms or diagnostic outcomes, 19 reported psychiatric diagnostic outcomes, and 30 reported psychiatric symptoms (categories not mutually exclusive).

FIGURE 1.

Literature flow.

Overview of Included Studies

The mean sample size of included studies was 69,627 (median: 1779; range: 100–1,368,148). Thirteen were cohort studies with control groups, 8 were cohort studies without a control group, 5 were case-control, and 23 were cross-sectional. Study populations included the following service eras: Persian Gulf War (N=29), Vietnam War (N=7), Iran-Iraq War (N=4), Operation Iraqi Freedom (OIF; N=5), Operation Enduring Freedom/Operation New Dawn (OEF/OND; N=3), and other (eg, peacetime, multiera, or not reported, N=5; Table 1). While demographic data was not consistently reported, nearly all the identified studies included mostly male (range: 62%–100%) and mostly White samples (range: 52.6%–98.6%). Only 8 studies included a sample of at least 20% females,^1,21–27 none of which conducted sex-stratified analyses specifically looking at sex differences in relation to MEE. Forty-one studies included members or veterans of the US armed forces, and 8 included a non-US sample (3 United Kingdom, 4 Iran, 1 South Korea).

KQ1: Mapping of MH and Exposure Outcomes

The relationship between MEEs and MH (symptoms or diagnoses) was evaluated by 45 of the 49 included studies. Most of those studies conducted multiple analyses. When possible, all analyses relevant to the present study were included in this synthesis, which resulted in a far greater number of associations than the articles reviewed. Figure 2 depicts the number of associations between MEE type and outcomes analyzed (N=228) as well as the service era of the sample. Gulf War veterans were the most frequently studied participants in the literature and accounted for over half of the extracted analyses (N=123). In terms of associations, Vietnam-era veterans were the second most common cohort of participants (10.1%), followed by OEF, OIF, and OND cohorts (9.2% of association). Veterans from peacetime eras were the least common (1.3%). Chemical weapons were the most studied MEE, representing 23.7% of the associations. While chemical weapons were studied across multiple military cohorts, other exposures such as herbicides (ie, Agent Orange) were predominately evaluated in Vietnam veterans. Contaminated water from Camp Lejeune had the least number of associations (N=3). Composite MH variables, inclusive of multiple diagnoses in combination with psychiatric and/or physical symptoms were the most common MH outcomes identified (N=49). This was closely followed by depression outcomes (N=46), and PTSD and posttraumatic stress symptoms (N=40).

FIGURE 2.

Visualization of associations across included studies by Era of Service, Military Toxic Exposure, and Psychiatric Symptoms and Diagnoses.

Depression

Depressive Symptoms

Seventeen studies examined associations of depressive symptoms with various MEEs.^{1,6,17,21,22,28–39} In these studies, symptoms were most often measured by using a wide variety of self-report measures. Chemicals were the most-examined exposure, with chemical weapons most commonly reported (ie, sarin gas; N=8),^{1,6,30,34,36–38} followed by pesticides (N=3),^6,32,39 herbicides (ie, Agent Orange; N=3),^17,28,35 and organophosphate esters (N=1).²⁹ Three studies examined exposures to air pollutants from oil well fires.^32,33,38 Two studies examined other types of exposures, including JP-8 and/or diesel.^29,32 Three studies examined multiple/mixed exposures during the Gulf War,²¹ and another from OEF and OIF.²¹ No studies examined exposures related to Camp Lejeune, heavy metals, or radiation in relation to depressive symptoms. Nine of these studies included Gulf War-era veterans.^{1,6,22,31–33,37–39} An additional 3 studies were conducted among US Vietnam-era veterans,^17,28,35 3 among Iranian veterans,^30,34,36 one from OEF and OIF,²¹ and one study included aircraft maintenance veterans of all eras.²⁹

In addition, one cohort study of Gulf War-era veterans (N=598,269) examined the association between suicide and chemical weapons exposure at the Khamisiyah munitions depot using mortality data from the US Department of Defense records.⁴⁰

Depressive Disorders

Five studies that examined associations between depressive disorders and MEEs.^{18,33,41–43} A variety of depressive disorders were included in this group, with the most common being MDD^33,42,43 Depressive disorders were assessed using a variety of methods, with only one using diagnostic codes¹⁸ and 2 using structured clinical interviews.^33,41 The most commonly studied exposure was chemicals (N=4), including chemical weapons (1 of sulfur mustard,⁴¹ 1 of sarin⁴³), the herbicide Agent Orange,⁴² and Camp Lejeune water contamination.¹⁸ One additional study assessed exposure to oil well fires in Kuwait.³³ Four studies included US military veterans: 2 Gulf War era,^33,43 one Vietnam era,⁴² and one among veterans stationed at Camp Lejeune.¹⁸ The last study was conducted among Iranian Military veterans of the Iran-Iraq War.⁴¹

Anxiety

Anxiety Symptoms

Nine studies examined associations between anxiety symptoms and various military exposures, using various self-report measures.^{1,17,22,30,34,35,38} The relationship between anxiety symptoms and chemicals was most commonly studied (N=6). Anxiety symptoms were studied in relation to: herbicides (ie, Agent Orange; N=2)^17,35 and chemical weapons (ie, sulfur mustard gas, sarin, Khamisiyah munitions depot explosion; N=4).^1,30,34,38 Two studies examined exposure to tent heater exhaust.²² No data were available on the relationship between anxiety symptoms and airborne hazards, contaminated water from Camp Lejeune, radiation, or heavy metals. Study samples included veterans from the Persian Gulf (N=3),^1,22,38 Vietnam (N=2),^17,35 and the Iran-Iraq war (N=2).^30,34

Anxiety Disorders

Four studies examined anxiety disorders in relation to chemical exposures^18,41,43 This included sulfur mustard gas (Gulf War era),⁴¹ sarin nerve agents (Iran-Iraq War era),⁴³ and the contaminated water at Camp Lejeune.¹⁸ Two studies assessed obsessive-compulsive disorder, which was confirmed by a structured clinical interview⁴¹ or self-report.⁴³ One study assessed any anxiety disorder (based on a diagnostic codes search),¹⁸ and another utilized a semi-structured interview to identify people with hypochondriasis.⁴³

Posttraumatic Stress

Symptoms of Posttraumatic Stress

Nine studies included symptoms of PTSD as an outcome^{1,22,28,34,44–48} using various self-report measures. One study used a clinician-administered PTSD scale.⁴⁶ Chemical exposures, including pesticides,^45,47,48 herbicides,^28,44 solvents, or paints,⁴⁵ and chemical-based weapons^1,34,47,48 were the most commonly assessed toxins (N=7). Two studies assessed PTSD symptoms in relation to burn pits,^45,47 one also included exposure to oil well fires and exhaust from heaters and generators.³⁸ An additional study included exposure to tent heaters or diesel fuel.¹⁹ Symptoms of PTSD were also assessed in relation to radiation (depleted uranium),⁴⁵ mixed Gulf War-related exposures,^46,47 and multiple exposures during the Bosnia era.⁴⁵ Most of these studies (N=5) were conducted among veterans of the Gulf War Era, followed by Vietnam veterans (N=2). No data was available concerning PTSD symptoms and heavy metals or Camp Lejeune-related exposures.

Posttraumatic Stress Disorder (PTSD)

Six studies assessed the association between PTSD and military exposures,^{25,30,36,41–43} of which half used a structured clinical interview to identify PTSD,^30,36,41 and 2 used self-report measures.^25,43 Five studies explored chemical weapons exposures (3 sulfur mustard gas,^30,36,41 2 sarin nerve agent^36,43). One included self-reported exposure to biological or chemical warfare agents,²⁵ and one explored exposure to the herbicide Agent Orange.⁴² These studies were conducted with veterans from the Vietnam era (N=1),⁴² the Gulf War (N=1),⁴³ or a combination of personnel from the Gulf War or Bosnia era (N=1).²⁵ The remaining studies were conducted with Iranian military veterans from the Iran-Iraq War and were conducted with all-male samples (N =3).^30,36,41

Sleep Dysfunction

Sleep Problems

Seven studies were identified that examined sleep problems in relation to MEEs.^{1,6,21,31,32,37,38} All studies identified sleep problems using a dichotomous item (yes/no) on a self-report questionnaire. Sleep problems were most commonly assessed in relation to chemicals, including exposure to Khamisiyah munitions depot detonation (N=2),^31,38 chemical weapons (N=4),^1,6,31,37 and pesticides (N=2).^6,32 Other MEEs explored were air pollution from oil well fires (N=1³²) and multiple or mixed exposures (including tent heaters and diesel fuel or nuclear and biological exposures).^31,32,38 Except for one study of OIF/OEF veterans,²¹ all sleep problems were assessed in Gulf War cohorts.

Sleep Disorders

Two studies assessed sleep disorders and both were in relation to chemical exposures.¹⁸ One study of Agent Orange-exposed Vietnam veterans utilized a composite sleep disorder variable obtained using self-report.⁴² The other study assessed rapid eye movement sleep behavior disorder in a sample of veterans who had been exposed to contaminated water at Camp Lejeune and utilized ICD codes for diagnostic identification.¹⁸

Psychotic Disorders and Related Symptoms

Symptoms of Psychosis

Symptoms of psychosis were assessed in 3 studies using multiscale measures of personality^28,35 or the DSM-III criteria.⁴¹ All participants reported chemical exposures.^28,35,41 Two of the studies included Vietnam veterans exposed to Agent Orange, and the other study assessed Iraq veterans exposed to sulfur mustard gas during the Iran-Iraq War.

Psychotic Disorders

Two studies assessed psychotic disorders. The mean age of participants was 58.8 years. One included a sample of Korean veterans, exposed to Agent Orange in Vietnam.⁴⁹ The other study assessed Iranian military veterans exposed to sulfur mustard gas during the Iran-Iraq War.⁴¹ These studies relied on self-report as well as pre-existing records to identify exposure variables.

Substance use Symptoms and Disorders

Substance Use Symptoms

In 3 studies, Vietnam-era veterans exposed to Agent Orange were assessed for symptoms of substance use disorders through self-report.²⁸ No studies assessed the relationship between substance use symptoms and exposure to air pollution, burn pits, fumes, or heavy metals.

Substance Use Disorders (SUDs)

Three studies examined associations between SUDs and chemical exposures.^41–43 One looked at exposure to Agent Orange,⁴² and 2 assessed anticholinesterase agents and sulfur mustard (chemical weapons).^41,43 The SUD variable was determined based on pre-existing data (clinical medical records, historical diagnoses),^42,43 or by a structured interview.⁴¹ Exposure variables were based on military records^41,43 or self-report.⁴²

Other Individual Psychiatric Symptoms

Seven studies assessed MH symptoms not captured within the psychiatric categories described thus far.^{22–24,27,34,50,51} These studies included a broad “other” category^24,34,50,51 or included symptoms related to mood lability or personality.^22,23,34 Chemicals exposures were the most common,^{22–24,34,50} followed by multiple exposures⁵¹ and air pollution.²⁷

Composite Psychological Outcomes

Composite Psychiatric Symptoms

Six studies analyzed composite MH outcomes comprised of multiple psychiatric symptoms (eg, depression, anxiety).^{2,17,33,39,51,52} One of the studies assessed an associations with the OIF/OEF cohort, and the remaining 5 assessed associations among Persian-Gulf veterans.

Composite Psychiatric Diagnoses

Eight other studies assessed the relationship between toxins and psychological disorders. Seven of these studies utilized a composite MH diagnosis inclusive of a combination of various psychological disorders. These aggregate MH outcomes were examined in relation to various exposure variables, including oil fires^47,53 pesticides,^36,38,54 Khamisiyah-related chemicals,⁴⁸ and chemical weapons.⁴⁹ One study looked at only a finite set of MH conditions (ie, any mood disorder, bipolar disorder, and somatoform disorder) in relation to sulfur mustard exposure.²⁶ MH diagnoses were determined using pre-existing records or self-report. One study⁵⁵ utilized a semi-structured clinical interview.

Neurocognitive Symptoms and Diagnoses

Twenty-two studies examined neurocognitive diagnoses and symptoms in relation to MEEs.^{1,6,18,21,24,26,31,32,35,37–39,41,42,52–60} Most of these studies identified neurocognitive dysfunction based on self-report measures.^{1,6,21,24,31,32,37,38,53–55,57–60} Two studies relied on pre-existing diagnoses,^18,42 4 utilized standardized objective neuropsychological measures,^26,35,39,56 and 1 utilized a mental status examination.⁴¹ Within the 22 neurocognitive studies, 158 associations were described. Figure 3 visualizes associations across included studies by era of service, MEEs, and neurocognitive symptoms and diagnoses. Thirteen of the studies utilized a composite variable to assess neurocognition, and this variable included data based on physiological, psychological, and behavioral symptoms. This included 11 Gulf War exposure studies,^{1,6,31,32,37,38,53,54,57,59,60} and 2 Iraq-Afghanistan studies^21,24 that analyzed subjective cognitive complaints as part of a combined multisystem physiological-psychological construct. It was unclear whether an underlying neurocognitive disorder was driving associations found in these analyses, as these composite variables included unrelated constructs. Only 9 studies utilized a clear disaggregated neurocognitive construct;^{18,26,35,39,41,42,55,56} 6 assessed neurocognitive symptoms, 2 assessed dementia diagnoses,^18,42 and 1 examined an undefined “impaired cognition syndrome.” Neurocognitive functioning was examined in relation to contaminated water in Camp Lejeune,¹⁸ Agent Orange exposure in Vietnam,^35,42,56 sulfur mustard gas in Iraq war veterans (chemicals),⁴¹ and Gulf War veterans who had mixed exposures (Khamisiyah-related),^26,55 exposed to chemicals (pesticides), mixed exposures (poison gas or germ warfare), or other exposures (antinerve gas pills).^26,39 Data on exposure variables were obtained using self-report,^{6,31,32,35,38,39,53–55,57,59,60} biospecimen (blood⁵⁶), or pre-existing documentation (ie, deployment records).^{1,6,18,26,37,41,42,60} Diversity in participants was lacking, with mostly White and male samples reported. Figure 3 provides a visualization of the research on MEEs and neurocognitive outcomes across the eras of service.

FIGURE 3.

Visualization of associations across included studies by Era of Service, Military Toxic Exposure, and Neurocognitive Symptoms and Diagnosis.

KQ2: Evidence Mapping of Associations

For KQ2, we analyzed the size of associations across the most common exposures and most common MH outcomes. Of the 49 included studies, 41 provided information about the association between MEEs and neurocognitive or psychiatric symptoms and diagnoses. The remaining 8 studies reported composite outcomes (including irrelevant physical symptoms). Figure 4 provides a visualization of the number of associations across studies categorized into sizes within outcome category and separated by 4 major groups of MEEs. Some studies only reported whether the association was significant without reporting the strength of these associations.

FIGURE 4.

Size of associations across most common exposures and most common symptoms and condition.

Most analyses examining the relationship between MEE and MH outcomes (symptoms and diagnoses) focused on chemical weapons exposure. This pattern was observed across the following MH outcome categories: depression symptoms and diagnoses, anxiety symptoms and diagnoses, PTSD and symptoms, and neurocognitive symptoms and diagnoses. For neurocognitive, anxiety, and depression outcomes, the size of the associations with exposures was most frequently not reported (ie, descriptive data only), or no association was observed (ie, only nonsignificant P-values reported). Across PTSD analyses, large and moderate associations were frequently observed among a wide range of MEEs; nearly half of the associations with chemical weapons produced large effects. There was no data in the literature evaluating the relationship between burn pits, depleted uranium, and air pollution in relation to depression, neurocognitive, or anxiety outcomes. There was no data on the association between anxiety outcomes and JP-8/jet fuel or between Gulf War-related exposures and anxiety or depression symptoms. Across all MH symptoms and diagnoses, some large associations were noted, mostly regarding exposure to chemicals; however, no large associations were identified across the neurocognitive outcomes.

Descriptive-only data or nonsignificant results were the second most reported association for chemical exposures and MH outcomes. Twenty-four associations between MH outcomes and mixed or multiple exposures were observed, with P-values being the common statistic reported. For neurocognitive variables, nonsignificant P-values were most common, followed by significant P-values. Fourteen associations were reported between MH outcomes and exposure to air pollution, burn pits, or fumes. All other associations with airborne hazards varied in terms of type and strength of association. Four associations were identified between radiation and MH outcomes.

Within the 41 included studies, there was significant variation in construct formation and measurement. Overall, there were few patterns in terms of the strength in associations between MEEs and psychological diagnoses. The statistical analyses used and methodological approach to study design all varied greatly. Moreover, several of these studies ran multiple (eg, ≥20) analyses, and it is unclear whether these studies adjusted for multiple comparisons. Effect sizes were rarely reported, making it difficult to determine the strength of associations. Composite variables of multiple MH disorders and composite exposure variables also made it difficult to discern what was responsible for driving correlations. Many studies used convenience samples that did not include equal sample sizes across groupings, which may have inflated or minimized the effect. In addition, studies were mostly cross-sectional in nature, and as such causality of these associations cannot be inferred. Symptom validity testing was not reported as being completed by any of the studies, so the accuracy of self-reported symptoms and exposures should be interpreted with caution. Recall bias, also may have affected the integrity of the self-reported data. Across the literature, studies lacked data on the mechanism, duration, quantity, and route of exposures. Data collection approaches appeared to rely on dichotomous exposure constructs (present or not present) as opposed to collecting variables inclusive of detailed exposure data. While this may have allowed researchers to collect larger amounts of data, there is a current gap in knowledge regarding exposure outcomes.

DISCUSSION

In this evidence map, we reviewed the empirical literature that assesses the association of MEEs with MH and neurocognitive symptoms and diagnoses among US and non-US military veterans. Two major findings emerged. First, there is broad evidence that toxic exposures during military service are associated with poorer MH. For example, exposure to chemicals and air pollution appears to be associated with a high occurrence of PTSD and posttraumatic stress symptoms. Nevertheless, such findings should be understood within the context of our second broad finding—there are several methodological limitations in the literature examining environmental exposure during military service and MH outcomes, including the imprecise measurement of MEEs and MH symptoms or diagnoses.

Despite these limitations, the overall pattern of associations between MEE and MH has important clinical implications. Given the broad associations between MEE and MH, it would be useful for clinical care to include screening for symptoms of depression, anxiety, and PTSD among veterans reporting MEEs. Assessing the MH needs of this group may help ensure this higher-risk population receives the neuropsychological and mental health care they need. In published studies, exposures often occurred while on active duty, which could have been years, or decades before the study being conducted. This reality made the measurement of self-reported toxic exposure more subject to recall bias yet also gave ample time for the effects of MEEs to emerge over. This long latency period between exposure and illness has support from research on both veteran and civilian samples. For example, higher rates of Parkinson disease dementia/major neurocognitive disorder were found decades after exposure to Agent Orange³⁵ and veterans with MEE appear to show worsening MH over time when compared with veterans reporting fewer exposures.¹⁹ Proactive and routine screenings could improve the opportunity for earlier identification and treatment of MH symptoms or disorders if they emerge over time. More provider training on occupational health, including awareness and education on screening and addressing military MEE is needed.

Our review highlighted limitations in the existing literature linking MEE and MH, suggesting approaches that might improve future research. The most pressing limitations of existing research included: (1) lack of diverse representation in samples (eg, sex, race, ethnicity, service era); (2) lack of precision measurement for MEE, MH, and neurocognitive outcomes; and (3) few longitudinal studies. In addition, previous research has failed to consider adjusting for the primary effect of stress (eg, combat, deployment, poor living conditions), which may have exacerbated the effects of the MEE on MH outcomes. While its impossible to disentangle these effects retroactively, future research could prospectively and more precisely examine the biological mechanisms and psychological processes underlying these MH and neurocognitive changes. It is possible that MH and neurocognitive outcomes may be causally unrelated to MEE, yet still yield robust associations if MEEs and MH/neurocognitive decline have co-occurred for different reasons (eg, stress, aging, lifestyle factors, pre-existing conditions, genetic factors). ​

Studying more diverse samples will better reflect the changing veteran population and improve the generalizability of research. In addition, most studies used self-report measures, which can be subject to recall bias. There is great value in self-report, but they should be administered with a symptom validity measure and, when available, corroborated by objective measures (eg, military records showing proximity to exposure). Complementary data collection approaches (eg, electronic health and deployment records) could also provide valuable data. Real-time measurements (eg, biological, ecological momentary assessment) could also be used as a baseline to prospectively follow military service members postdeployment. Longitudinal studies that account for changes in MH and neurocognitive functioning over time would be beneficial to characterize how MEE might lead to or contribute to this longitudinal decline. MH changes over the life course; better accounting for this change as veterans age is important to accurately gauge future MH needs. Better-defined control groups would also be helpful in disentangling the role of psychological stress and pre-existing genetic predisposition for MH and neurocognitive outcomes. More specificity in terms of outcome variables should be considered in future studies to provide more specificity into causal mechanisms that might explain associations (or the lack of associations). Neuropsychiatric variables of interest should be disaggregated into clearly defined and distinct constructs. Most studies have relied on self-report data to assess neurocognitive changes. Future studies should incorporate objective neuropsychological tests to evaluate these outcomes in a standardized manner. A comprehensive neuropsychological evaluation would help researchers understand the degree of cognitive decline (if any) and track the severity of these changes over time.

While the original report was intended to provide detailed information for VA providers, we know that veterans receive care from community providers and are not the only group of people who may have been exposed to environmental exposures. As such, we envision our findings may generalize to civilian providers and researchers. There are several non-military occupations (eg, mechanics, painters, carpenters) who by the nature of their job may have high exposure to toxins. As such, civilian providers may wish to employ screening measures like that of the PACT ACT initiated MEE Screener, to get a better understanding of the prevalence of environmental exposures in civilian settings. This will help researchers better understand the prevalence of environmental exposures and begin to explore whether they are associated with MH.

CONCLUSION

We summarize the empirical literature assessing environmental exposures during military service in relation to MH outcomes, including psychological symptoms and disorders, and neurocognitive outcomes. Our review has 2 primary findings. First, at the broadest level, veterans who report exposure to toxins during military service appear to be at risk of poor MH and might constitute an at-risk group that would benefit from psychological assessment and treatment. Second, the existing literature in this area largely lacks methodological rigor—including in terms of measurement, study design, and specificity—that would be important to support the implementation of specific interventions to improve MH among veterans with service-related MEE. Higher quality data accompanied by plausible biological hypotheses (including highlighting the biological mechanisms underlying the associations and more mechanistic studies) are needed, to avoid more publications showing vague “associations” between a multitude of complex factors, without identifying causal mechanisms.

Supplementary Material

mlr-64-s137-s001.docx ^{(25.5KB, docx)}

mlr-64-s137-s002.docx ^{(14.4KB, docx)}

mlr-64-s137-s003.docx ^{(13.8KB, docx)}

mlr-64-s137-s004.docx ^{(15.7KB, docx)}