Abraham, Joseph H. ScD; DeBakey, Samar F. MD, MPH; Reid, Lawrence MPH; Zhou, Joey PhD; Baird, Coleen P. MD, MPH
Environmental exposures such as inhalation of dust, smoke, fumes, and aerosols are common in the deployed environment.1,2 Military deployment to Iraq and Afghanistan is suspected to adversely affect respiratory health among soldiers.2
Particulate matter (PM) in the region is a primary exposure concern of the US military due to blowing sand and dust, emissions from petrochemical and other industrial sites, vehicle traffic, burn pits used for waste disposal, and oil fields and fires.3,4 Deployed military service members are often exposed to PM levels exceeding typical levels in the United States.3,5,6 Measurements of PM less than or equal to 10 μm in aerodynamic diameter (PM10) collected during the 1991 Gulf War showed concentrations consistently 2 to 5 times higher than the Environmental Protection Agency (EPA)'s 24-hour criteria for PM10.6 For current theaters of operations in Southwest Asia, the US Army Center for Health Promotion and Preventive Medicine's Deployment Environmental Surveillance Program collected PM data, and levels routinely exceeded the EPA 24-hour standards.7 In fact, 75% of PM2.5 levels routinely exceeded 35 μg/m3, the cutoff above which levels are considered unhealthy for sensitive populations, according to the EPA's standardized air quality index.8
Military deployment to Iraq and Afghanistan has been associated with self-reported health outcomes. Survey results of military personnel deployed to Operation Enduring Freedom and Operation Iraqi Freedom (OEF/OIF) estimated that 69% of personnel reported a respiratory illness.4 Among participants in the Millennium Cohort Study,9 deployers had 1.4 times the rate of newly self-reported respiratory symptoms, relative to nondeployers, although they had similar rates of self-reported obstructive pulmonary disease diagnoses (chronic bronchitis, emphysema, and asthma). Length of deployment was also significantly associated with increased symptom reporting among Millennium Cohort Study participants in the Army, although the finding was not consistent among Air Force and Marine Corps personnel.
To assess the effect of deployment to OEF/OIF theaters on changes in the respiratory health of deployed personnel, we obtained deployment data as well as inpatient and outpatient medical encounter data for deployed active duty (AD) US military personnel.10 We then assessed the association between deployment and personnel health by comparing the before and after deployment rate of medical encounters for select conditions, and subsequently comparing individuals who deployed to OEF/OIF once with those who deployed multiple times.
To assess whether deployment influenced subsequent respiratory health, as indicated by short-term postdeployment medical encounters, we leveraged deployment and personnel data provided by the Defense Manpower Data Center (DMDC), and medical encounter data from the Defense Medical Surveillance System, including the Standard Inpatient Data Records, Health Care Service Records Institutional–Inpatient, Standard Ambulatory Data records, and the Health Care Service Records Institutional.
An extract of the DMDC OEF/OIF roster of all deployed personnel as of December 31, 2005 provided approximately 2.3 million deployment records for approximately 1.2 million service members. From this extract, duplicate records of those personnel still on deployment, and those personnel with less than 6 months of follow-up time after their last deployment, were removed, leading to a little more than 1 million records representing 820,042 service members. To derive a more accurate number of distinct deployments, records attributable to 21-day rest and relaxation or emergency temporary duty leave were collapsed into their respective interrupted deployment and were not counted as separate deployments. The number of deployments was then counted to classify service members into single- (deployment = 1) and multiple-deployer (deployment >1) groups. After selecting a 10% random sample of each of the single- and multiple-deployer groups, the study population consisted of 44,919 AD single deployers and 14,695 AD multiple deployers who were deployed to OEF/OIF and were redeployed by June 30, 2005, and had not deployed again between July 1, 2005, and December 31, 2005.
All inpatient and outpatient medical encounter data for members of the study population were obtained from the Defense Medical Surveillance System. Analyses were limited to primary diagnostic codes. Relevant diagnostic categories and diagnoses included diseases of the respiratory system (International Classification of Diseases–Ninth Revision–Clinical Modification [ICD-9-CM]: 460 to 519) that were considered both as a whole, and classified into the following six ICD-9-CM code–based categories: acute respiratory infections (ICD-9-CM: 460 to 466), other diseases of the upper respiratory tract (ICD-9-CM: 470 to 478), pneumonia and influenza (ICD-9-CM: 480 to 487), asthma/chronic obstructive pulmonary disease and allied conditions (ICD-9-CM: 490 to 496), pneumoconiosis and other lung diseases due to external agents (ICD-9-CM: 500 to 508), and other diseases of the respiratory system (ICD-9-CM: 510 to 519). We also examined rates of encounters with primary diagnoses of “symptoms involving respiratory system and other chest symptoms (ICD-9-CM: 786).”
We compared the pre- and postdeployment medical encounter rates of those deployed to OEF/OIF, overall. Then, to assess the effect of deployment, we stratified service members by deployment status. For single deployers, we examined up to 6 months before and after the deployment period, while accounting for available time at risk for a medical encounter. For multiple deployers, we examined up to 6 months before deployment, up to 6 months after last deployment, and all periods between deployments while service members were in garrison and at risk for a medical encounter. Therefore, all service members had at least 6 months of visits before and after deployment.
Case Definition and Control Selection
We then conducted a nested case–control study to further explore the independent effects of deployment status and cumulative time in theater on incident postdeployment obstructive pulmonary diseases' onset.
To select incident obstructive pulmonary disease cases and representative controls, we identified a total of 51,295 personnel who were free of any respiratory system or symptom diagnoses within the 6-month period before their deployment (38,490 single deployers and 12,805 multiple deployers). Within this group, we identified 533 cases with a postdeployment primary diagnosis for at least one obstructive pulmonary disease (ICD-9-CM: 490 to 496). We excluded one Coast Guard case. From the personnel free of respiratory system or symptom diagnoses within 6 months before their deployment, we randomly selected four controls for each case, matching on the year of case definition and the year of the last encounter during the study period for controls as well as the total number of postdeployment medical encounters, to control for patterns of care seeking. This resulted in 532 cases and 2128 matched controls.
Variables of Interest
Demographic (age and sex) and military (branch of service, grade, and occupation codes) characteristics were derived from the DMDC personnel records. Military characteristics corresponding to the start of the deployment of interest were used. Ages were determined on the basis of the time of diagnosis for cases and on the time of the last visit for controls, and categorized into quartiles based on the entire cohort (18 to 22, 23 to 26, 27 to 33, and more than 34 years). Occupations were dichotomized to indicate whether the personnel ever had a combat occupation code during the study period, as this information may indicate individuals who were more likely to be exposed to environmental hazards.
For the nested case–control study, we created two proxy variables for exposure; namely, number of deployments and cumulative time in theater, using the start and end dates of DMDC deployment records. The number of deployments was set to one for cases with an obstructive pulmonary disease encounter after their first deployment, and controls who had one deployment during the study period were also considered to have one deployment. Cumulative time in theater was calculated by summing the time for all the deployments before an obstructive pulmonary disease encounter for cases and before the last medical encounter for controls, and subsequently categorized into 5 levels (0 to 3, 4 to 6, 7 to 9, 10 to 12, and more than 13 months).
To partially account for smoking patterns, we noted the presence of any one of the following codes: ICD-9-CM code of 305.1 (tobacco use disorder), 989.84 (tobacco), V15.82 (history of tobacco use), or E869.4 (second-hand tobacco smoke).
For data management and analysis in this study, we used SAS/STAT software, version 9.1 of the SAS System for Windows (copyright 2002 to 2008, SAS Institute Inc, Cary, NC). Primary diagnoses in both inpatient and outpatient records for the 6-month period before and after deployment for single deployers and the periods in-between deployments for multiple deployers were examined for short-term outcomes after deployment. Encounters with the same primary diagnosis within 30 days of each other were collapsed to avoid double counting.
Rates of encounters within selected primary diagnoses and 95% confidence interval (CI) by deployment status for the pre- and postdeployment periods and by order of deployment were calculated. The selected categories included respiratory diagnoses and respiratory symptoms. Individuals' time in garrison, while at risk of a medical encounter, was considered in the denominators, and censored at separation from the military for any reason including death.
Conditional logistic regression analyses were employed to examine the independent effects of number of deployments at diagnosis and cumulative time in theater up to diagnosis on postdeployment obstructive pulmonary disease encounter, controlling for potential confounders in the model.
As a whole, the rate of respiratory system encounters decreased by eight percent from before deployment (encounters per 1000 person-years, 300.6, 95% CI, 294.4 to 306.9) to after deployment (276.2 encounters per 1000 person-years, 95% CI, 271.2 to 281.2). Nevertheless, the postdeployment rate of obstructive pulmonary disease (ICD-9-CM: 490 to 496) encounters was significantly higher than the rate before deployment (23.9 encounters per 1000 person-years compared with 19.0 encounters per 1000 person-years; relative rate, 1.25; 95%CI, 1.13 to 1.39). The postdeployment rate for encounters coded as other diseases of the upper respiratory tract (ICD-9-CM: 470 to 478) was similarly higher after deployment (66.2 encounters per 1000 person-years, 95% CI, 63.7 to 68.6), relative to predeployment (53.8 encounters per 1000 person-years; 95% CI, 51.1 to 56.4).
Overall, predeployment rate of encounters for primary respiratory system diagnoses (301 per 1000 person-years) was higher among single deployers when compared with multiple deployers (280.4 encounters per 1000 person-years; 95% CI, 268.3 to 292.6) (Fig. 1). The magnitude of the pre- to postdeployment decrease in respiratory system encounters was larger for multiple deployers. Among multiple deployers, the rate of respiratory system encounters increased minimally with increasing numbers of deployments after the first postdeployment period, but did not exceed the rate for single deployers.
The rate of respiratory symptom encounters (ICD-9-CM: 786) among single deployers (53.8 encounters per 1000 person-years; 95% CI, 50.7 to 56.8) was higher than those for the multiple deployers (37.0 encounters per 1000 person-years; 95% CI, 32.6 to 41.4) in the predeployment period. Respiratory symptom rates for both single and multiple deployers increased after deployment. (Fig. 1.)
Among single deployers, the rate of encounters for obstructive pulmonary disease encounters increased significantly from the predeployment to postdeployment periods (predeployment rate, 20.4 encounters per 1000 person-years, 95% CI, 18.5 to 22.3; postdeployment rate, 30.1 encounters per 1000 person-years, 95% CI, 27.8 to 32.5), although respiratory encounters decreased as a whole. No such increase was observed for multiple deployers when examining the first postdeployment period. An increase in the rate of obstructive pulmonary disease encounters was observed after the second deployment, although the increase was not statistically significant. A significant increase in “other diseases of the upper respiratory tract” (ICD-9-CM: 470 to 478), a subcategory that includes allergic rhinitis, was observed for both single and multiple deployers.
Among 532 cases with new obstructive pulmonary disease diagnoses, 50% (n = 266) had a bronchitis diagnosis (not specified as either acute or chronic, ICD-9-CM: 490), 46% (n = 245) had an asthma diagnosis (ICD-9-CM: 493), and 3% (n = 15) had a chronic bronchitis diagnosis (ICD-9-CM: 491). Obstructive pulmonary disease encounters constituted approximately 7% and 10% of all respiratory system diagnoses during the pre- and postdeployment periods, respectively, for single deployers, and it constituted 6%, 8%, and 9% in the predeployment, and first and second postdeployment periods, respectively, for multiple deployers. Twelve-and-a-half percent (62 of 495) of single deployers and 5.2% (8 of 154) of multiple deployers with a postdeployment medical encounter for an obstructive pulmonary disease during the study period had a medical encounter with a diagnosis in the same ICD-9 category before deployment.
Table 1 presents the distribution and crude and adjusted relative odds for selected characteristics of obstructive pulmonary disease cases and matched controls. Compared with their matched controls, cases were more likely to be female (15.4% vs 12.1%), enlisted (90.6% vs 86.9%), and Army (49.2% vs 39.6%) or Air Force (23.3% vs 21.2%). Female sex (adjusted odds ratio [OR], 1.38; 95% CI, 1.04 to 1.82), enlisted personnel (adjusted OR, 1.69; 95% CI, 1.19 to 2.39), and Army personnel (adjusted OR, 1.36; 95% CI, 1.03 to 1.79) remained independent predictors of having a new obstructive pulmonary disease encounter. Age and combat occupations were not statistically significantly associated with a postdeployment obstructive pulmonary disease diagnosis.
Extended cumulative time in-theater did not seem to increase the risk of becoming an obstructive pulmonary disease encounter case in this analysis. On the contrary, a statistically significant protective effect was observed for those with 13 or more months of cumulative deployment time, using 0 to 3 months as a reference group (adjusted OR, 0.65; 95% CI, 0.43 to 0.97). Similarly, we did not find significantly increased odds of having an obstructive pulmonary disease encounter for multiple deployers relative to single deployers (adjusted OR, 1.08; 95% CI, 0.82 to 1.42). No significant interactions between the number of deployments at diagnosis and cumulative time in theater up to diagnosis were found.
To further describe the obstructive pulmonary disease cases and matched controls, we estimated the prevalence of comorbid conditions during both the pre- and postdeployment periods (Table 2). There were no differences in the prevalence of most comorbid conditions examined between cases and controls during the predeployment period; however, cases exhibited higher prevalence of all comorbid conditions during the postdeployment period.
Deployment to Iraq and Afghanistan has been associated with increases in self-reported respiratory symptoms during4,11 and after deployment.9 In this study, we assessed pre- and postdeployment rates of inpatient and outpatient medical encounters for selected respiratory and disease/injury conditions among a group of AD military personnel who had been deployed to OEF/OIF areas of operation. Overall, the rate of respiratory system encounters decreased after deployment, a finding driven largely by a statistically significant drop in acute respiratory infections.
We focused on obstructive pulmonary disease encounters (ICD-9-CM: 490 to 496), as these conditions are potentially related to in-theater environmental exposures. The vast majority of obstructive pulmonary disease encounters were either for asthma (46%) or bronchitis (50%). Because of the study design and the need to examine visits in between multiple deployments, the length of follow-up was relatively short with a minimum of 6 months of follow-up. This time period would likely be insufficient for the development of some forms of chronic obstructive pulmonary disease such as emphysema. We observed statistically significant increases in the encounter rates for obstructive respiratory diseases from the pre- to postdeployment periods for single deployers, and a corresponding increase between the predeployment period and the last (fourth) postdeployment period. Medical encounters for obstructive pulmonary disease among multiple deployers did not increase between predeployment and the first postdeployment period.
Much of the evidence to date suggests that deployment may adversely impact short-term respiratory health. The potential for long-term respiratory health consequences of deployment to Iraq and Afghanistan is also of concern to troops. Roop et al (2007)11 surveyed troops deployed in support of OEF/OIF, and found that asthmatic and nonasthmatic patients reported statistically significantly increased respiratory symptoms during deployment relative to symptom prevalence before deployment. The authors noted a prevalence of asthma in five percent of the surveyed personnel. Sanders et al4 conducted a survey of military personnel deployed to OEF/OIF and found that 69% reported experiencing respiratory illnesses, of which 17% required medical care. Among survey respondents deployed to Iraq, respiratory conditions doubled in prevalence from precombat to combat periods of operations. Smith et al9 found that deployers had a higher rate of newly reported respiratory symptoms, relative to nondeployers (14% vs 10%), while observing similar rates of chronic bronchitis or emphysema (1% vs 1%) and asthma (1% vs 1%). In their survey-based study of Millennium Cohort Study participants, the authors found that deployment length was significantly associated with increased respiratory symptom reporting among Army personnel.12 King et al13 conducted a case series of 80 soldiers being evaluated for exertional dyspnea after deployment to Iraq and Afghanistan. Of these patients, 38 were subsequently diagnosed with constrictive bronchiolitis, a rare chronic lung disease. Others in this case series from Fort Campbell, Kentucky were assigned other chronic disease diagnoses, including respiratory bronchiolitis, sarcoidosis, respiratory bronchiolitis–associated interstitial lung disease, hypersensitivity pneumonitis, and asthma.
This study had several limitations. Primary among them is our lack of smoking pattern data for the cohort. Smoking is a major risk factor for obstructive pulmonary disease diagnoses.14,15 Unsurprisingly, tobacco-related ICD-9-CM codes were significantly and positively associated with obstructive pulmonary disease encounters in our study, although our ability to validate such codes in relation to actual tobacco use was limited by the nature of the data. In the survey conducted by Sanders et al,4 38.9% of the troops smoked at least half a pack of cigarettes per day. Among the smokers, 47.6% either began smoking or restarted smoking during the deployment. Moreover, a patient's smoking behavior may directly influence a physician's assignment of an obstructive pulmonary diagnosis. As such, it is conceivable that our observation of an increase in obstructive pulmonary encounters after deployment may be entirely due to smoking. Unfortunately, our data did not allow for analysis of individual smoking habits.
Because of the deployment and redeployment cycle, our follow-up period between deployments was likely too short for longer-term conditions such as emphysematous chronic obstructive disease to develop. Our study subjects had a minimum of 6 months' follow-up after or between deployments. Despite this, we still demonstrated some postdeployment increases. It would be anticipated that more cases might develop based on longer follow-up, as subclinical conditions might become evident.
Our study was further limited by a lack of specific deployment-related exposure assessments. We used deployment as a proxy for deployment-related environmental exposures, because comprehensive in-theater environmental data were not available during the period of this study. Deployment is only a rough proxy for environmental exposures, and, beyond speculation, we were unable to resolve the effects of specific environmental hazards such as PM, burn pit smoke, and combat-related exposures. In addition, we lacked specific information about individuals' precise deployment location(s); information that could conceivably shed light on relevant environmental exposures. In addition, information about stress and other relevant variables known to decrease immune system function would be valuable to future studies.16
Limiting the analysis to AD members increases our ability to capture inpatient and outpatient encounters within both the direct and purchased care systems, but limits our capacity to make inferences regarding the health status of Reserve and National Guard personnel.
The observed increase in obstructive pulmonary disease diagnoses after deployment among enlisted personnel might be attributed to occupational and consequently indoor/outdoor variations between enlisted and officer personnel. Our analyses did not address occupation because the documented occupations may not represent actual duties in theater.
This study defined cases of obstructive pulmonary disease diagnoses (including both asthma and bronchitis) based on primary ICD-9-CM codes generated from inpatient and outpatient medical encounters. Such an outcome metric is objective, relative to self-report, but our analysis may nevertheless be confounded by indication, if physicians' diagnoses are influenced by individuals' deployment status. Outcome status may also be misclassified, although we believe that any such errors are unlikely to be related to deployment. We did not have access to the results of specific medical evaluations such as spirometry, bronchoprovocation tests, response to bronchodilator, or radiographic examinations.
We observed an increase in the rate of respiratory symptom encounters among both single and multiple deployers and we recommend continued follow-up of respiratory related encounters, as they may reflect subclinical or preclinical manifestations of chronic respiratory disease.
This study confirms the need to explore independent risk factors for the onset of specific pulmonary/respiratory conditions potentially associated with in-theater environmental exposures. This can be achieved through the use of specific exposure assessments, longer follow-up periods, more stringent case definition, occupational/task assessments, and smoking histories. Our finding of an increased obstructive pulmonary disease encounter rate in the absence of an association with deployment count or duration is consistent with the conclusion of Smith et al9; namely, that specific environmental exposures, rather than deployment in general, are determinants of postdeployment respiratory illness.
1. US Government Accountability Office. Occupational and Environmental Health Surveillance Conducted During Deployments Needs Improvement. GAO- 05–903T. Washington, DC: US Government Accountability Office; 2005.
2. Weese CB, Abraham JH. Potential health implications associated with particulate matter exposure in deployed settings in southwest Asia. Inhal Toxicol. 2009;21:291–296.
3. Department of Defense. Environmental Exposure Report. Washington, DC: Department of Defense; 2000.
4. Sanders JW, Putnam SD, Frankart C, et al. Impact of illness and non-combat injury during Operations Iraqi Freedom and Enduring Freedom (Afghanistan). Am J Trop Med Hyg. 2005;73:713–719.
5. Hastings DL, Jardine S. The relationship between air particulate levels and upper respiratory disease in soldiers deployed to Bosnia (1997–1998). Mil Med. 2002;167:296–303.
6. Thomas R VT, Meagher J, McMullin C. Particulate exposure during the Persian gulf war. Washington, DC: Department of Defense; 2000.
7. Engelbrecht JP, McDonald EV, Gillies JA, Jayanty RK, Casuccio G, Gertler AW. Characterizing mineral dusts and other aerosols from the Middle East–Part 1: ambient sampling. Inhal Toxicol. 2009;21:297–326.
8. Environmental Protection Agency. Guidelines for Reporting of Daily Air Quality-Air Quality Index (AQI). EPA-454.B-06-001. Research Triangle Park, NC: Office of Air Quality Planning and Standards; 2006.
9. Smith B, Wong CA, Smith TC, Boyko EJ, Gackstetter GD. Newly reported respiratory symptoms and conditions among military personnel deployed to Iraq and Afghanistan: a prospective population-based study. Am J Epidemiol. 2009;170:1433–1442.
10. Rubertone MV, Brundage John F. The defense medical surveillance system and the department of defense serum repository: glimpses of the future of public health surveillance. Am J Public Health. 2002;92:1900–1904.
11. Roop SA, Niven AS, Calvin BE, Bader J, Zacher LL. The prevalence and impact of respiratory symptoms in asthmatics and nonasthmatics during deployment. Mil Med. 2007;172:1264–1269.
12. Dominici FPR, Peng RD, Bell ML, et al. Fine particulate air pollution and hospital admission for cardiovascular and respiratory diseases. JAMA. 2006;295:1127–1134.
13. King MS, Eisenberg R, Newman JH, et al. Constrictive bronchiolitis in soldiers returning from Iraq and Afghanistan. N Engl J Med. 2011;365:222–230.
14. Mannino DM, Braman S. The epidemiology and economics of chronic obstructive pulmonary disease. Proc Am Thorac Soc. 2007;4:502–506.
15. Stephens MB, Yew KS. Diagnosis of chronic obstructive pulmonary disease. Am Fam Physician. 2008;78:87–92.
16. Glaser R, Kiecolt-Glaser JK. Stress-induced immune dysfunction: implications for health. Nat Rev. 2005;5:243–251.