ABSTRACT
Since September 11, 2001, over 2.7 million United States service members have deployed to South-West Asia and the Middle East and have been exposed to environmental hazards and psychological trauma. Many of these service members have returned with medical and psychological illnesses, some of which have proved complex and resistant to treatment. One notable constellation of symptoms is post-deployment respiratory illness, which has become a focus of research and policy efforts. The present study sought to examine the impact of post-deployment psychological distress on respiratory symptom severity. Data were obtained from the Veterans Affairs Airborne Hazards and Open Burn Pit Registry (AHOBPR) health surveillance database (N =107,403). Psychological factors were compared against common organic and environmental predictors of post-deployment respiratory distress. Psychological distress following deployment was a stronger predictor of 12-month shortness of breath severity than general respiratory pathology or level of exposure to environmental hazards, controlling for gender, age, race, and tobacco use. Additionally, psychological distress was a better predictor of shortness of breath severity than documented respiratory illnesses including asthma, chronic obstructive pulmonary disease, and chronic bronchitis. Implications and directions for future research are discussed, as well as potential alterations to existing treatment and health surveillance paradigms.
KEYWORDS: Military, deployment, dyspnea, anxiety, hyperventilation
What is the public significance of this article?—The present study examined the relationship between psychological factors, such as combat exposure and psychological distress, and respiratory symptoms in post-deployment Veterans. Environmental factors such as burn pit exposure or poor air quality are often emphasized as causal factors for post-deployment respiratory symptoms; however, our research suggests that psychological factors may interact with environmental hazards to affect the severity of post-deployment respiratory symptoms. We also highlight emerging research that suggests that this interaction process may be far more complex than previously thought, and may involve a wide breadth of processes and bodily systems.
Since September 11th, 2001, over 2.7 million United States military personnel have deployed in support of operations in Afghanistan and Iraq. Many of these individuals were exposed to environmental, physical, and psychological hazards and returned home with short- and long-term health effects that are still not entirely understood (VA.gov: Veterans Affairs, 2018). One of the signature post-deployment illnesses of the conflicts in Afghanistan and Iraq is lung-related injury or disease, with studies finding that between 14–69% of returning Veterans reported post-deployment respiratory symptoms (Roop et al., 2007; Sanders et al., 2005; Szema et al., 2017). Research also has demonstrated an increased incidence of chronic lung disease in these post-deployment Veterans compared to their non-deployed peers (Roop et al., 2007; Szema et al., 2017). Research has supported significant associations between deployment to Afghanistan or Iraq and new post-deployment diagnoses of asthma, chronic bronchitis, and chronic obstructive pulmonary disease (Abraham et al., 2012).
Early studies identified airborne hazards, specifically burn pit products, as potential etiological factors driving the increased incidence of respiratory conditions and morbidity following deployment to Iraq and Afghanistan. Burn pits are open-air sites that were historically used in-theater to dispose of trash, medical waste, and other refuse in-theater, where traditional waste disposal operations were difficult or impossible (Weese & Abraham, 2009). Military bases in both the Iraq and Afghanistan theaters used burn pits to dispose of waste products; in addition, military personnel were frequently exposed to sandstorms, munitions smoke, and other airborne hazards (Szema et al., 2017). Research into these exposures found that particulate matter (PM) levels at some military bases in Iraq and Afghanistan frequently exceeded Military Exposure Guidelines and the Environmental Protection Agency’s (EPA) 24-hour PM limit and reached levels considered sufficient to cause illness (Abraham et al., 2012; Blasch et al., 2016; Weese & Abraham, 2009). PM, generally, is a heterogeneous mixture of toxic and nontoxic particles that are suspended in the air and, depending on its size, easily inhaled into the lowest portions of the respiratory system (Weese & Abraham, 2009). Exposure to high levels of PM have been linked to lower respiratory problems, COPD, and reductions in lung function and life expectancy (Weese & Abraham, 2009).
The broad links between airborne hazard exposure and some respiratory symptoms and conditions are well-established and recognized by the Veterans Affairs Administration (VA), which is responsible for evaluating those connections for compensation and disability purposes (VA.gov: Veterans Affairs, 2018). However, researchers have had difficulty identifying specific, data-driven linkages between theoretically relevant exposures and pathological biological mechanisms or processes (Wauters et al., 2019). Most studies that have demonstrated a compelling pathway from identified exposures to specific symptoms or diseases have focused on Veterans with severe documented exposures that are likely not germane to the entire population (King et al., 2011; Wauters et al., 2019). A large portion of the variance in post-deployment respiratory condition onset and severity is still unexplained, which has hindered efforts to effectively identify and treat these conditions (Wauters et al., 2019).
One factor that could influence the course and severity of these respiratory conditions that has not been well examined is the presence of psychological correlates and comorbidities (Wauters et al., 2019). Occasionally, researchers and clinicians have dubiously attributed diffuse, multi-system illnesses in Veterans to psychiatric illness or psychosomatic conversion Porter et al., 2020. However, this explanation errs by conflating psychological processes that can influence organic disease with pure psychogenic illness (Chenivesse et al., 2014; Porter et al., 2020; Wauters et al., 2019). The field of psychology has developed a more granular understanding of how psychological distress can affect, at a biological level, the processes that relate to and perpetuate various organic symptoms and illnesses (Koren et al., 2021; Sapolsky, 1994). Of particular importance to the current study, research has illuminated strong linkages between psychological distress, acute and chronic hyperventilation, and biological correlates of respiratory illness including bronchoconstriction and suffocation alarm (Chenivesse et al., 2014). These processes affect the severity and stability of respiratory illnesses such as COPD (Alius et al., 2013; Bailey, 2004; Chenivesse et al., 2014; Slatore et al., 2018; Wauters et al., 2019; Wilhelm et al., 2001). Recent research has identified ways that psychological factors could impact post-deployment respiratory illness: 1) by mimicking organic pulmonary illness, 2) exacerbating organic pulmonary illness, and 3) affecting how Veterans perceive and react to symptoms of respiratory distress (Abdel-Hamid, 2018; Leventhal et al., 2016; Slatore et al., 2018; Wauters et al., 2019).
Researchers are beginning the complex process of differentiating physical and psychological contributors to respiratory distress. For example, in a retrospective analysis of over 180,000 service members seeking services from the VA, mental health diagnoses were associated with an increased risk of subsequent diagnosis with any respiratory illness (Slatore et al., 2018). Interestingly, this relationship was negligible for respiratory illnesses that are more frequently diagnosed through objective measures (e.g., pulmonary function tests; Slatore et al., 2018). The authors noted that these results were consistent with multiple historical studies that found associations between anxiety and trauma disorders and increased incidence of respiratory illness (Slatore et al., 2018). At the same time, they specified that this interaction could be attributable to multiple causal mechanisms, including shared etiological factors (Slatore et al., 2018). Mounting evidence suggests that psychological factors play some role in the etiology (origin and path of development) or maintenance of respiratory distress in some circumstances (Porter et al., 2020; Slatore et al., 2018; Wauters et al., 2019). For example, a study of 9/11 first responders who were exposed to both psychological trauma and environmental toxins found that a diagnosis of posttraumatic stress disorder was a better predictor of later respiratory symptoms and symptom severity than objective pulmonary function (Slatore et al., 2018). These findings are difficult to account for using etiological models focused entirely on exposure to environmental toxins (Slatore et al., 2018).
While most of the previously cited studies have focused on the exacerbating effect of psychological distress on organic illness and respiration, psychological distress also can be a potent standalone etiological factor for respiratory distress (Abdel-Hamid, 2018; Slatore et al., 2018). For example, hyperventilation syndrome (HVS) – a pattern of deep, sigh, and over breathing in excess of metabolic demands that causes characteristic physical symptoms, including chronic dyspnea – is widely understood to be underpinned by psychologically mediated autonomic processes (Chenivesse et al., 2014). Research has found that chronic hyperventilation in the absence of physical dysfunction can still be profoundly disabling (Chenivesse et al., 2014). A study of HVS patients found that they had lower reported quality of life than patients with COPD, cystic fibrosis, or asthma (Chenivesse et al., 2014). This included lower functioning scores on physical functioning, social functioning, physical roles, emotional roles, mental health, vitality, body pain, and general health (Chenivesse et al., 2014). The authors concluded that HVS can be profoundly disabling and difficult to differentiate, both diagnostically and functionally, from severe organic lung disease (Chenivesse et al., 2014).
There have been broad and substantial changes in our understanding of how and to what extent psychological factors may influence respiratory illness and symptoms (Sapolsky, 1994). Additionally, it is increasingly clear that these factors can have at least as profound an influence on respiratory symptoms and distress as organic disease processes (Chenivesse et al., 2014; Slatore et al., 2018; Wauters et al., 2019). The goal of the present study was to begin to clarify how, and to what extent, psychological factors may be associated with Veterans’ experiences of post-deployment respiratory distress and respiratory symptoms. Specifically, we compared the relative utility of Veteran-reported psychological and organic variables in accounting for variance in post-deployment shortness of breath severity, while controlling for known confounding variables. We hypothesized that psychological distress would have equal predictive power for shortness of breath severity when compared to grouped organic lung disease, burn pit exposure, and respiratory illness variables. We also hypothesized that this predictive power would be unique to psychological distress and would not be otherwise statistically accounted for by the other grouped factors. Finally, we predicted that psychological distress would significantly predict shortness of breath severity when controlling for all other measured organic variables, and that the effect of psychological distress on shortness of breath severity would be greater than any other single variable in the analysis.
Method
Participants
The present study retrospectively examined digital survey data provided by United States Veterans (N =211,218) to the VA’s Airborne Hazards and Open Burn Pit Registry (AHOBPR) health-surveillance program between 2014 and 2019. For the purposes of this paper, the word “Veterans” includes any individual who previously served, or is currently serving, in the United States Armed Forces. In keeping with the conventions of the VA and AHOBPR, “Veterans of Iraq and/or Afghanistan,” refers to any previously or currently serving service member who was deployed to Iraq and surrounding countries after August 2, 1990, or to Afghanistan or Djibouti after September 11, 2001 (VA.gov: Veterans Affairs, 2018). The only exception to this phrasing in the current study is that participants who identified as Gulf War Veterans were excluded from our analyses, to limit heterogeneity of exposures across conflicts. Regardless, the majority of AHOBPR registrants (around ~90% at time of data extraction) were deployed in support of operations after 2001. A minority (~10%) participated in Operations Desert Shield and/or Desert Storm and self-identified as Gulf War Veterans.
The AHOBPR is a congressionally mandated health surveillance data collection effort housed under the VA’s Health Outcomes of Military Exposures (HOME) branch (National Academies of Sciences, Engineering, and Medicine, 2020). The goal of the program is to collect data on Veterans’ deployments, environmental exposures, and health outcomes to illuminate any associations between deployment-related variables and negative health outcomes (National Academies of Sciences, Engineering, and Medicine, 2020). Participants were encouraged to provide data for the AHOBPR through a variety of different methods, including public messaging campaigns, contact with the VA health system, and a public-facing website (National Academies of Sciences, Engineering, and Medicine, 2020). Participants self-selected into the AHOBPR by completing an online form provided by the VA on their public health website (VA.gov: Veterans Affairs, 2018). Data are collected on a rolling basis via a VA-hosted internet survey platform (VA.gov: Veterans Affairs, 2018). Participants self-select into the AHOBPR by navigating to the online survey site and answering a series of multiple choice and scale response questions about their service, health, deployment, and exposure histories.
As mentioned above, inclusion criteria for the AHOBPR include a history of deployment to specific countries in the Middle East and South-West Asia after 1990 and 2001, respectively (VA.gov: Veterans Affairs, 2018). For the present study, we excluded participants who did not report experiencing shortness of breath within 12 months prior to their registration. We also excluded Veterans who self-identified as having served in the Gulf War and Veterans who reported being diagnosed with a lung condition prior to being deployed. Our final sample, after exclusions, included 97,911 Veterans’ survey response sets.
Measures
Registrants with the AHOBPR complete a series of questions related to their service, deployment, and exposure history. Items relevant to the present study are described below.
Demographic & service variables
Participants self-reported their sex, branch of service, age, and race.
Years spent smoking
Participants self-reported whether they had ever smoked tobacco and, if so, the number of years they had smoked. Nonsmokers were coded as “0 years” spent smoking, and the remainder of participants were coded on a continuous scale based on their years spent smoking.
Psychological/emotional distress
Participants were asked whether they had experienced a “depression/anxiety/emotional problem,” that impacted their ability to perform activities such as walking or running. This is a crude measure of psychological distress and would be unlikely to capture all registrants with psychological distress or illness; however, based on the question’s wording, participants who responded affirmatively are likely to have experienced significant psychological distress and impairment. Importantly, while there are several reasons why these psychological difficulties could be related to activities of daily living (ADL) impairment, the question does not assume or imply that these psychological factors are associated with shortness of breath.
Lung conditions and diagnoses
Veterans who reported impairment in ADLs also were asked whether they experienced any “Lung [or] breathing problem[s] (for example, asthma and emphysema).” This item did not require a medical diagnosis. Veterans who responded affirmatively were coded as having respiratory symptoms, distress, or disease.
Veterans also self-reported confirmed medical diagnoses of allergies, asthma, COPD, emphysema, chronic bronchitis, constrictive bronchiolitis, and idiopathic pulmonary fibrosis. These conditions were recorded as either present or not present. As opposed to the general lung disease variable, these items specified that the condition must have been diagnosed and documented by a medical professional. Previous research has found an almost perfect correlation between similar self-report measures and information in electronic health records (Porter et al., 2020).
Highest burn pit exposure
Burn pit exposure was quantified as the participants’ self-reported daily exposure to smoke or fumes from burn pits while deployed. Participants reported the average number of hours per day they were exposed to burn pit smoke during a specified deployment (0–24). For Veterans who reported multiple deployments, we included only their highest reported daily exposure, and eliminated the data from any other reported deployments. This decision was made based on previous studies that found an association between any airborne hazard exposures and respiratory symptoms, but not between the length of that exposure and respiratory symptoms (Slatore et al., 2018).
Other environmental exposures
The VA included in the AHOBPR questions pertaining to various deployment-related exposures, other than burn-pits, that they recognize as being associated with higher levels of airborne particulate matter. These include exposure to improvised explosive device blasts, combat smoke, convoy dust, fumes from vehicles, construction byproducts, pesticides, dust storms, and poor air quality. The current study did not include items pertaining to IED explosions and combat smoke exposures because they were expected to be related to higher overall emotional distress and posttraumatic stress disorder. The current study also excluded the air quality items because these items asked about respiratory symptoms associated with poor air quality, not the air quality itself. Within the AHOBPR, the remaining items were all measured on a scale of 0–31 corresponding to the average number of days per month the Veteran reported being exposed to the hazards during their deployment.
Shortness of breath (Dyspnea) severity
Shortness of breath severity was quantified as the participants’ self-reported average breathlessness over the preceding year on a scale of 0–4. Participants were coded as 0 if they reported not being troubled by breathlessness or being bothered by shortness of breath during heavy exercise; 1 if they reported shortness of breath hurrying on level ground or walking up a slight hill; 2 if they reported walking slower than most people on level ground, or had to stop at 1 mile or 15 minutes at their own pace; 3 if they reported stopping for breath after walking about 100 yards; and 4 if they reported being too breathless to leave the house, or breathless while dressing or undressing.
Data analyses
All statistical analyses were performed using STATA version 16 software. STATA 16 is a robust statistical software package that is particularly well-suited to analyzing large datasets. The software was selected due to the researchers’ familiarity with the program and its processing capabilities.
First, to identify blocks of terms that were significantly affecting the predictive power of our model, we conducted a hierarchal regression analysis with five blocks and 22 total predictor terms. This allowed us to examine the relative influence of different types of exposures based on theoretically relevant groupings. Next, we conducted a multiple linear regression using the original terms in order to determine whether there was a significant relationship between the various predictor variables and the dependent variable, and, if so, the strength of that relationship. We used G*Power 3.1.9.4 to calculate our required power for this analysis, and retroactively confirm the sufficiency of our sample size. The power analysis identified 21,759 participant records as the cutoff to be adequately powered to detect a small effect size (f2 =.001) with an α =.050, β =.800, and 22 predictor variables. This provided an actual calculated power of approximately 80%. Finally, we calculated η2 for each term within the regression in order to determine the effect size of each predictor variable on the dependent variable.
Results
Demographics
The sample demographics for the current study are depicted in Table 1. In total, 88.7% of our Veteran sample was male, 78.9% identified as white, 14.2% identified as African American or Black, 2.5% identified as Asian, and 4.2% identified as another race, or declined to answer. The majority of our sample reported that their primary service occurred in the Army (63.8%), with the next highest branches being the Air Force (17.8%), the Marine Corps (11.9%), and the Navy (6.3%). The average age at time of registration was 38-years-old, with a standard deviation of 8 years. The average number of years spent smoking was 4.0 across the entire sample (μ =4.00, σ =7.60).
Table 1.
Demographics of participants in study sample.
| Variable | Level | Observations | % (*Mean/Std. Dev) |
|---|---|---|---|
| Sex | Male | 83,915 | 88.7% |
| Female | 10,716 | 11.3% | |
| Race | White | 58,401 | 78.9% |
| African American | 10,564 | 14.2% | |
| Asian | 1,888 | 2.5% | |
| Other | 3,128 | 4.2% | |
| Branch of Service | Army | 62,432 | 63.8% |
| Air Force | 17,423 | 17.8% | |
| NavyMarines | 6,20311,612 | 6.34%11.9% | |
| Other | 215 | 0.2% | |
| Age at Registration | N/A | 94,265 | (*38.4/8.4) |
| Average Burn Pit Exposure Hours | N/A | 97,911 | (*12.4/9.5) |
| Avg. Convoy Exposure Days/Mon | N/A | 91,727 | (*16.3/11.8) |
| Avg. Refueling Exposure Days/Mon | N/A | 89,687 | (*13.8/12.2) |
| Avg. Engine Maintenance Days/Mon | N/A | 90,773 | (*9.1/12.5) |
| Avg. Construction Exposure Days/Mon | N/A | 86,452 | (*5.4/9.0) |
| Avg. Pesticide Exposure Days/Mon | N/A | 74,766 | (*1.4/5.1) |
| Avg. Dust Storm Days | N/A | 84,929 | (*9.3/7.4) |
| Number of Years Smoking Tobacco | N/A | 97,911 | (*4.0/7.6) |
| Reported Lung Condition | Yes | 38,213 | 39.0% |
| No | 59,698 | 61.0% | |
| Dx of Allergies | Yes | 39,399 | 43.4% |
| No | 51,330 | 56.6% | |
| Dx of Asthma | Yes | 17,239 | 18.9% |
| No | 74,327 | 81.2% | |
| Dx of Emphysema | Yes | 1,436 | 1.6% |
| No | 90,928 | 98.5% | |
| Dx Chronic Bronchitis | Yes | 3,840 | 4.2% |
| No | 86,741 | 95.8% | |
| Dx of COPD | Yes | 14,978 | 16.9% |
| No | 74,026 | 83.2% | |
| Dx CB | Yes | 1,267 | 1.3% |
| No | 96,644 | 98.7% | |
| Dx IPF | Yes | 290 | 0.3% |
| No | 97,621 | 99.7% | |
| Reported Mood Condition | Yes | 27,554 | 28.1% |
| No | 70,356 | 71.9% |
In terms of exposures, the average number of hours per day registrants were exposed to smoke from burn pits was 12.5 (μ =12.46, σ =9.49). Participants also were exposed to convoy operations on 16.3 days per month (μ =16.34, σ =11.82), refueling operations for 13.8 days (μ =13.84, σ =12.19), engine maintenance for 9.1 days (μ =9.13, σ =12.50), construction for 5.4 days (μ =5.44, σ =9.01), pesticides for 1.4 days (μ =1.43, σ =5.14), and dust storm for 9.3 days (μ =9.34, σ =7.75). Among registrants, 39.0% reported having a lung condition that interfered with their activities of daily living, and 28.1% reporting having an anxiety/depression/emotional condition or problem.
For medically confirmed diagnoses, 43.4% of participants reported being diagnosed with allergies, 18.9% reported being diagnosed with Asthma, 1.6% reported being diagnosed with emphysema, 4.2% reported being diagnosed with chronic bronchitis, 16.9% reported being diagnosed with chronic obstructive pulmonary disease (COPD), 1.3% reported being diagnosed with constrictive bronchiolitis, and 0.3% reported being diagnosed with idiopathic pulmonary fibrosis. In accordance with our exclusion criteria, any participants who reported being diagnosed with any of these lung conditions prior to deployment were not included in our sample.
Analysis 1: Hierarchal regression model
We conducted a hierarchal regression with shortness of breath regressed on blocks of related variables added in each step to assess the magnitude and significance of ΔR2. The iterations of the model and their related terms are presented in Table 2. The covariates were added in step one. In step two, the lung condition variable was added to the model (ΔR2 = .029, p < .001). In step three, average burn pit exposure was added to model (ΔR2 = .010, p < .001). Step four included the other exposures variables (ΔR2 = .030, p < .001). Step five included all diagnosed lung conditions; however, this step did not result in a significant change in R2 (ΔR2 = .013, p ~ 0.999). In the final step, psychological distress was added (ΔR2 =.023, p < .001). The covariates alone, including sex, race, branch, age, and smoke years accounted for less than 1% of the total variance in shortness of breath severity (R2 = .004). The entire model (at step 6) accounted for about 11% of the total variance in shortness of breath severity (R2 = .109).
Table 2.
Hierarchical regression and model comparisons.
| Model | R2 | F(df) | p | Δ R2 | F(df) Change | p |
|---|---|---|---|---|---|---|
| Model 1 | .00 | 54.70(5,72,296) | < .001 | N/A | N/A | N/A |
| Model 2 | .03 | 406.86(6, 72,295) | < .001 | .03 | 2159.52(1,72,295) | < .001 |
| Model 3 | .04 | 464.55(7, 72,294) | < .001 | .01 | 784.20(1,72,294) | < .001 |
| Model 4 | .07 | 294.32(13, 48,514) | < .001 | .03 | 72.62(6,48,514) | < .001 |
| Model 5 | .09 | 193.76(20, 41,160) | < .001 | .01 | −15.03(7,41,160) | 1.000 |
| Model 6 | .11 | 240.19(21, 41,159) | < .001 | .02 | 1068.32(1,41,159) | < .001 |
Terms added by step include 1) Model 1: Sex, Race, Branch, Age, Smoke Years 2) Model 2: Self-Reported Lung Condition 3) Model 3: Avg. Burn Pit Exposure 4) Model 4: Avg. Engine Maintenance, Convoy, Refueling, Construction, Pesticide, Dust Storm Exposure 5) Model 5: Dx’d Allergies, Asthma, Emphysema, Chronic Bronchitis, COPD, C.B., Idiopathic Pulmonary Fibrosis 6) Model 6: Self-Reported Psychological Distress
Analysis 2: Multiple linear regression model and coefficient comparisons
Next, to examine the contributions of individual predictors, a multiple linear regression model was fitted to the data with shortness of breath regressed on all of the twenty-one predictor variables. All of the included variables, shown in Table 3, significantly predicted shortness of breath severity except branch of service, number of years spent smoking, and average exposure to engine maintenance. The predictors with the largest regression coefficients were psychological distress (Std. β = .155, p < .001), self-reported lung conditions (Std. β = .112, p < .001), a diagnosis of chronic bronchitis (Std. β = .076, p < .001), and average dust storm exposure (Std. β = .071, p < .001). All model terms were evaluated against a corrected α of .002 to account for α inflation.
Table 3.
Multiple linear regression model & coefficients for association between predictors and shortness of breath severity.
| Variable | Beta | Std. Error | t | p |
|---|---|---|---|---|
| Sex | .023 | .013 | 4.82 | < .001* |
| Race | .021 | .005 | 4.46 | < .001* |
| Age at Registration | .021 | .000 | 4.29 | < .001* |
| Branch | .001 | .003 | 0.22 | .83 |
| Smoke Years | .008 | .000 | 1.68 | .09 |
| Avg. Burn Pit Exp. | .057 | .000 | 11.71 | < .001* |
| Avg. Convoy Exp. | .033 | .000 | 6.02 | < .001* |
| Avg Refueling Exp. | .040 | .000 | 6.65 | < .001* |
| Avg. Engine Exp. | .011 | .000 | 2.09 | .04 |
| Avg. Construction Exp. | .017 | .000 | 3.33 | < .01* |
| Avg. Pesticide Exp. | .048 | .000 | 9.36 | < .001* |
| Avg. Dust Storm Exp. | .071 | .000 | 14.25 | < .001* |
| Lung Conditions | .112 | .009 | 22.12 | < .001* |
| Psychological Distress | .155 | .009 | 33.69 | < .001* |
| Allergies | −.033 | .008 | −6.88 | < .001* |
| Asthma | .045 | .011 | 8.55 | < .001* |
| Emphysema | .019 | .034 | 3.85 | < .001* |
| Chronic Bronchitis | .076 | .021 | 14.79 | < .001* |
| COPD | .048 | .012 | 9.64 | < .001* |
| Const. Bronchiolitis | .029 | .034 | 6.03 | < .001* |
| IPF | .017 | .066 | 3.62 | < .001* |
*Significant at a corrected α =.002. Overall R2 =.11. IPF refers to idiopathic pulmonary fibrosis.
Analysis 3: Effect size comparisons
To better evaluate the actual and comparative effect of each variable on shortness of breath severity, partial η2 was calculated for each term within our linear regression model. The effect size results for each variable are described in Table 4. Two variables rose to the level of a small effect size (η2 > .01). These variables were psychological distress (η2 =.025) and self-reported lung condition (η2 =.012). The remainder of the variables fell below the η2 threshold for a small effect.
Table 4.
Effect sizes of variables on shortness of breath severity.
| Variable | Partial η2 |
|---|---|
| Sex | .001 |
| Race | .000 |
| Age at Registration | .000 |
| Branch During Service | .000 |
| Years Spent Smoking Tobacco | .000 |
| Avg. Burn Pit Exposure | .003 |
| Avg. Convoy Exposure | .001 |
| Avg. Refueling Exposure | .001 |
| Avg. Engine Exposure | .000 |
| Avg. Construction Exposure | .000 |
| Avg. Pesticide Exposure | .002 |
| Avg. Dust Storm Exposure | .004 |
| Self-Reported Lung Conditions | .012* |
| Self-Reported Psychological Distress | .025* |
| Dx of Allergies | .001 |
| Dx of Asthma | .002 |
| Dx of Emphysema | .000 |
| Dx of Chronic Bronchitis | .005 |
| Dx of COPD | .002 |
| Dx of Const. Bronchiolitis | .001 |
| Dx of IPF | .000 |
Partial η2 values with asterisk have small effect sizes on Shortness of Breath Severity. IPF refers to idiopathic pulmonary fibrosis.
Discussion
Our findings provide support for the idea that psychological variables are associated with increased respiratory symptoms in post-deployment Veterans above and beyond the symptom severity explained by identified biological disease processes. Within a biopsychosocial framework, it is expected that multiple mechanisms can exert simultaneous, compounding effects on a given symptom or condition; that was the case in the present study where multiple factors had small but statistically significant impact on respirator health. This is consistent with the extant literature on the impact of psychological factors on disease processes and provides a potential mechanism, through hyperventilation, for psychological factors to influence physiological breathing processes and sensations (Chenivesse et al., 2014; Koren et al., 2021; Sapolsky, 1994).
In this sample of Veterans registered with the AHOBPR, psychological distress was a meaningful predictor of shortness of breath severity across a series of analyses. First, within our hierarchical regression analysis, psychological distress significantly increased the amount of variance explained by the model, even when added after demographics, tobacco use, all exposure data, self-reported lung conditions, and all reported lung diagnoses. Previous studies have used similar analyses to argue against a purely psychogenic model of post-deployment respiratory distress, by showing the residual impact of organic factors and processes (Porter et al., 2020). Within our biopsychosocial framework, both biological and psychological factors exert an influence, but there is a residual impact of psychological factors after controlling for the organic factors collected in the AHOBPR.
Analyses found that psychological distress was a significant predictor of shortness of breath severity and had the highest regression coefficient and effect size of all the variables examined. The next most influential variable was the presence of a self-reported lung condition, which would be a natural etiological explanation for anyone experiencing shortness of breath. In contrast, medically diagnosed lung diseases were weaker predictors of shortness of breath severity. Even constrictive bronchiolitis, a very severe restrictive lung disease, was a weaker predictor of shortness of breath severity than self-reported lung conditions, and its associated effect size was substantially smaller than that of psychological distress. Indeed, when controlling for all other variables, only psychological distress and self-reported lung conditions were associated with effects that reached the common η2 threshold for a small effect size with psychological distress having an effect roughly twice as large as self-reported lung conditions.
The lack of strong effects attributable to common deployment-related exposures, including tobacco use, burn pits, and dust storms, are counter to the popular understanding of the etiology of post-deployment shortness of breath, but these results are not unprecedented (National Academies of Sciences, Engineering, and Medicine, 2020). In a gold-standard review of the current evidence surrounding the effects of exposures on post-deployment health, the National Academies of Sciences, Engineering, and Medicine (2020) found that only general respiratory distress, not specific pulmonary conditions, was associated with deployment to the Middle East and South-West Asia. Interestingly, they found good toxicological and laboratory evidence for the pathogenic properties of common deployment-related exposures (e.g., heavy metals in dust), but limited to no epidemiological association between those exposures and health effects following deployment (National Academies of Sciences, Engineering, and Medicine, 2020). The reason for this discrepancy is a topic of strong debate. One potential explanation within the domain of post-deployment respiratory distress is that some deployment exposures may serve as temporary, but highly salient, dyspnea-inducing factors, but that the effects of those hazards on respiratory function may wane over time and be superseded by learned/psychological contributors (Cao et al., 2016; Vlemincx & Luminet, 2020). Importantly, this approach has some good support in adjacent literature, and avoids the pitfalls of holding constant the influence of organic factors at all points in the symptom development and maintenance processes (Cao et al., 2016; Veidal et al., 2016; Vlemincx & Luminet, 2020).
The present results are correlational, and so we cannot draw conclusions as to whether psychological distress is causing increased shortness of breath severity in Veterans or identify specific mechanisms. Furthermore, because the data were collected subsequent to the onset of respiratory symptoms, we cannot determine whether psychological distress was involved in the etiology of the shortness of breath or is simply associated with increased shortness of breath once symptoms have developed. Still, while acknowledging these limitations, our results suggest a unique and independent impact of psychological distress on shortness of breath experienced by veterans enrolled in the AHOBPR. This is consistent with the robust body of literature that indicates that psychological distress, through biopsychological feedback mechanisms and physiological changes from hyperventilation, could cause or exacerbate shortness of breath (Abdel-Hamid, 2018; Bailey, 2004; Chenivesse et al., 2014; Slatore et al., 2018). The present findings highlight the importance of a systematic examination of the role of psychological factors, including but not limited to distress, in understanding the maintenance and exacerbation of respiratory symptoms apparently arising out of deployment experiences. If the psychological distress included in the AHOBPR was simply a result of stress caused by a physical pulmonary illness, we would not expect the relationship between psychological distress and shortness of breath severity to be nearly as robust as it is, accounting for various exposures, medical diagnoses, tobacco use, and demographics. Importantly, despite the design limitations of the present study, our results provide preliminary support for future research into psychological distress as a potential etiological factor in post-deployment shortness of breath.
The present findings have implications for public health and treatment approaches for addressing the needs of the population of Veterans with deployment-related breathing problems. Most current interventions and public health approaches emphasize identifying and treating underlying pulmonary conditions that are presumed to be related to burn pit exposure and to drive the bulk of the reported respiratory distress. In this study, the relationship between burn pit (and other environmental) exposures and respiratory distress was not fully accounted for by diagnosed respiratory diseases. This suggests that some of that effect of exposure on distress may not be reflected by current diagnostic and medical approaches.
Though the effect sizes within our study were small, we were able to compare the influence of psychological distress to the majority of organic variables that previously have been associated with post-deployment respiratory distress in previous research (Abraham et al., 2012; Blasch et al., 2016; Weese & Abraham, 2009). Psychological distress only had a small statistical effect on shortness of breath severity, but that effect was much larger than any reported exposures, medical diagnoses, or demographic features. The potential role of psychological factors has been largely neglected in efforts to understand and treat the breathing issues in this population. For example, we were able to identify a large number of items in the AHOBPR that assessed exposure and physical health domains. In contrast, only one question on the AHOBPR database directly queries about Veteran mental health. Similarly, some treatments for pulmonary illnesses, such as fast-acting albuterol, may be counterproductive in that they can exacerbate anxiety and psychological distress, thereby having a paradoxical effect on symptom severity. Future tracking and research efforts with this population would benefit from more fully integrating and attending to psychological issues and distress.
The present study utilized what is likely the largest repository of detailed data regarding Veterans’ deployment related exposures and respiratory health outcomes. The size of this sample provided sufficient power to detect even very small effects, and to compare those effects among variables of interest. Although the effects identified in the present study are too small to suggest that we have pinpointed the key to developing effective treatment options for this population, they do provide a basis for informing changes to current treatment approaches, namely the incorporation of psychological factors into our clinical conceptualizations. However, the cross-sectional and retrospective nature of the present data limit our conclusions. Therefore, it is critical that the relationships and potential mechanisms identified in the present study be examined through longitudinal and repeated measures research with a data set that includes expanded and more detailed questions regarding psychological variables.
At this time, research into the etiology of post-deployment respiratory distress in service members and Veterans remains quite limited. As previously mentioned, there remains a level of hesitancy and stigma around exploring these types of factors in Veterans, especially given the historical all-or-nothing approach that has fueled debates over etiology (NASEM, 2020). Statements such as the one in the previously mentioned NASEM report, which specifically identifies psychological and stress-related factors as pieces in the post-deployment health puzzle, may help reverse some of these trends (NASEM, 2020).
The more research can sensitively and compellingly highlight the well-documented complexity in these post-deployment conditions, the more providers and patients may be open to exploring the value of potentially helpful adjunct or complementary treatments. There remains a myriad of un- or under-explored explanations and mechanisms to be examined and much work to be completed in order to adequately address the clinical needs of this population.
There remains a myriad of un- or under-explored explanations and mechanisms to be examined and much work to be completed in order to adequately address the clinical needs of this population. Indeed, the fact that the full set of variables examined in the present study accounted for only about 11% of the variance in reported shortness of breath indicates that there is substantially more unknown than is known in this area. The results of this study can be useful in driving future exploration into those mechanisms. In the meantime, it is important for providers and patients to be aware of the gaps in our current knowledge of post-deployment respiratory distress and symptoms.
Correction Statement
This article has been corrected with minor changes. These changes do not impact the academic content of the article.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data Availability Statement
Data was provided by the US Veterans Affairs Administration and is available for review through the VA subject to administrative vetting and approval.
References
- Abdel-Hamid, H. (2018). Vocal cord dysfunction: An often misdiagnosed condition. The Egyptian Journal of Chest Diseases and Tuberculosis, 67(1), 1. 10.4103/ejcdt.ejcdt_12_18 [DOI] [Google Scholar]
- Abraham, J. H., Debakey, S. F., Reid, L., Zhou, J., & Baird, C. P. (2012). Does deployment to Iraq and Afghanistan affect respiratory health of US military personnel? Journal of Occupational and Environmental Medicine, 54(6), 740–745. 10.1097/jom.0b013e318252969a [DOI] [PubMed] [Google Scholar]
- Alius, M. G., Pané-Farré, C. A., Leupoldt, A. V., & Hamm, A. O. (2013). Induction of dyspnea evokes increased anxiety and maladaptive breathing in individuals with high anxiety sensitivity and suffocation fear. Psychophysiology, 50(5), 488–497. 10.1111/psyp.12028 [DOI] [PubMed] [Google Scholar]
- Bailey, P. H. (2004). The dyspnea-anxiety-dyspnea cycle — COPD patients’ stories of breathlessness: “It’s scary /when you can’t breathe.” Qualitative Health Research, 14(6), 760–778. 10.1177/1049732304265973 [DOI] [PubMed] [Google Scholar]
- Blasch, K., Kolivosky, J., & Hill, B. (2016). Occupational exposures among personnel working near combined burn pit and incinerator operations at Bagram Airfield, Afghanistan. Inhalation Toxicology, 28(5), 216–225. 10.3109/08958378.2016.1145768 [DOI] [PubMed] [Google Scholar]
- Cao, L., Zhang, X. G., Wang, J. G., Wang, H. B., Chen, Y. B., Zhao, D. H., Shi, W. F., & Xie, L. X. (2016). Pulmonary function test findings in patients with acute inhalation injury caused by smoke bombs. Journal of Thoracic Disease, 8(11), 3160–3167. 10.21037/jtd.2016.11.94 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chenivesse, C., Similowski, T., Bautin, N., Fournier, C., Robin, S., Wallaert, B., & Perez, T. (2014). Severely impaired health-related quality of life in chronic hyperventilation patients: Exploratory data. Respiratory Medicine, 108(3), 517–523. 10.1016/j.rmed.2013.10.024 [DOI] [PubMed] [Google Scholar]
- Institute of Medicine (US) Committee on the Initial Assessment of Readjustment Needs of Military Personnel, Veterans, & Families, and T . (1970, January 1). Operation enduring freedom and Iraqi freedom: Demographics and impact. https://www.ncbi.nlm.nih.gov/books/NBK220068/
- King, M. S., Eisenberg, R., Newman, J. H., Tolle, J. J., Harrell, F. E., Jr, Nian, H., Ninan, M., Lambright, E. S., Sheller, J. R., Johnson, J. E., & Miller, R. F. (2011). Constrictive bronchiolitis in soldiers returning from Iraq and Afghanistan. The New England Journal of Medicine, 365(3), 222–230. 10.1056/NEJMoa1101388 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Koren, T., Yifa, R., Amer, M., Krot, M., Boshnak, N., Ben-Shaanan, T. L., Azulay-Debby, H., Zalayat, I., Avishai, E., Hajjo, H., Schiller, M., Haykin, H., Korin, B., Farfara, D., Hakim, F., Kobiler, O., Rosenblum, K., & Rolls, A. (2021). Insular cortex neurons encode and retrieve specific immune responses. Cell, 184(24), 5902–5915.e17. 10.1016/j.cell.2021.10.013 [DOI] [PubMed] [Google Scholar]
- Leventhal, H., Phillips, L. A., & Burns, E. (2016). The Common-Sense Model of self-regulation (CSM): A dynamic framework for understanding illness self-management. Journal of Behavioral Medicine, 39(6), 935+. 10.1007/s10865-016-9782-2 [DOI] [PubMed] [Google Scholar]
- National Academies of Sciences, Engineering, & Medicine . (2020). Respiratory Health Effects of Airborne Hazards Exposures in the Southwest Asia Theater of Military Operations. The national academies of sciences, engineering, & medicine. Retrieved March 2, 2022. from https://www.nap.edu/catalog/25837/respiratory-health-effects-of-airborne-hazards-exposures-in-the-southwest-asia-theater-of-military-operations [PubMed]
- Porter, B., Long, K., Rull, R. P., Dursa, E. K., & Millennium Cohort Study Team . (2020). Prevalence of chronic multisymptom illness/Gulf War illness over time among millennium cohort participants, 2001 to 2016. Journal of Occupational and Environmental Medicine, 62(1), 4–10. 10.1097/JOM.0000000000001716 [DOI] [PubMed] [Google Scholar]
- Roop, S. A., Niven, A. S., Calvin, B. E., Bader, J., & Zacher, L. L. (2007). The prevalence and impact of respiratory symptoms in asthmatics and nonasthmatics during deployment. Military Medicine, 172(12), 1264–1269. 10.7205/milmed.172.12.1264 [DOI] [PubMed] [Google Scholar]
- Sanders, J. W., Frenck, R. W., Frankart, C., Putnam, S. D., Sharp, T. W., Riddle, M. S., Monteville, M. R., Rockabrand, D. M., & Tribble, D. R. (2005). Impact of illness and non-combat injury during operations Iraqi freedom and enduring freedom (Afghanistan). The American Journal of Tropical Medicine and Hygiene, 73(4), 713–719. 10.4269/ajtmh.2005.73.713 [DOI] [PubMed] [Google Scholar]
- Sapolsky, R. M. (1994). Why zebras don’t get ulcers: A guide to stress, stress related diseases, and coping. W.H. Freeman. [Google Scholar]
- Slatore, C. G., Falvo, M. J., Nugent, S., & Carlson, K. (2018). Afghanistan and Iraq war Veterans: Mental health diagnoses are associated with respiratory disease diagnoses. Military Medicine, 183(5–6), 5–6. 10.1093/milmed/usx108 [DOI] [PubMed] [Google Scholar]
- Szema, A., Mirsaidi, N., Patel, B., Viens, L., Forsyth, E., Li, J., Dang, S., Dukes, B., Giraldo, J., Kim, P., & Burns, M. (2017). Proposed Iraq and Afghanistan War-Lung Injury (IAW-LI) clinical practice recommendations: National academy of sciences’ institute of medicine burn-pits workshop. American Journal of Mens Health, 11(6), 1653–1663. 10.1177/1557988315619005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- US Department of Veterans Affairs . (n.d.). FY2022 budget submission: Medical programs and information technology programs. https://www.va.gov/budget/docs/summary/fy2022VAbudgetVolumeIImedicalProgramsAndInformationTechnology.pdf
- US Department of Veterans Affairs, Veterans Health Administration . (2015, April 27). “Public Health.” Protect Your Health. www.publichealth.va.gov/exposures/burnpits/registry.asp
- VA.gov: Veterans Affairs . (2018, September 19). https://www.ptsd.va.gov/understand/types/combat_exposure.asp
- Veidal, J., Sverrild, M., Backer, A., Porsbjerg, C., & Porsbjerg, C. (2016). The impact of dysfunctional breathing on the assessment of asthma control. Respiratory Medicine, 123, 42–47. 10.1016/j.rmed.2016.12.008 [DOI] [PubMed] [Google Scholar]
- Vlemincx, E., & Luminet, O. (2020). Sighs can become learned behaviors via operant learning. Biological Psychology, 151, 107850. 10.1016/j.biopsycho.2020.107850 [DOI] [PubMed] [Google Scholar]
- Wauters, R. H., Foster, B. E., & Banks, T. A. (2019). Environmental exposures and asthma in active duty service members. Current Allergy and Asthma Reports, 19(9), 9. 10.1007/s11882-019-0873-3 [DOI] [PubMed] [Google Scholar]
- Weese, C. B., & Abraham, J. H. (2009). Potential health implications associated with particulate matter exposure in deployed settings in Southwest Asia. Inhalation Toxicology, 21(4), 291–296. 10.1080/08958370802672891 [DOI] [PubMed] [Google Scholar]
- Wilhelm, F. H., Gerlach, A. L., & Roth, W. T. (2001). Slow recovery from voluntary hyperventilation in panic disorder. Psychosomatic Medicine, 63(4), 638–649. 10.1097/00006842-200107000-00017 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data was provided by the US Veterans Affairs Administration and is available for review through the VA subject to administrative vetting and approval.