Advertisement

Risk Factors Associated with Miscarriage and Impaired Fecundity among United States Servicewomen during the Recent Conflicts in Iraq and Afghanistan

Open AccessPublished:February 01, 2017DOI:https://doi.org/10.1016/j.whi.2016.12.012

      Abstract

      Background

      Research on the reproductive health of U.S. servicewomen deployed in support of the recent operations in Iraq and Afghanistan is sparse. The objective of this study was to evaluate whether military experiences, including combat deployment, deployment length, and life stressors during the recent conflicts, were associated with increased odds for miscarriage or impaired fecundity among U.S. servicewomen.

      Methods

      We used data from the Millennium Cohort Study, a large longitudinal military study that began in 2001 and includes military personnel from all service branches, including active duty and Reserve/National Guard personnel. Participants for this study included women aged 18 to 45 years who had completed two questionnaires (2004–2006 and 2007–2008). Separate multivariable logistic regression models were performed to estimate the odds of reporting miscarriage and impaired fecundity by military experiences that adjusted for covariates. Subanalyses were conducted using International Classification of Diseases, Ninth Revision, Clinical Modification codes found in the Military Health System Data Repository for both outcomes among servicewomen on active duty.

      Results

      Overall, 31% and 11% of military servicewomen reported miscarriage and impaired fecundity, respectively, during the approximate 3-year follow-up period. After adjusting for demographic, behavioral, and military characteristics, deployment experiences and life stressors were not associated with miscarriage or perceived impaired fecundity. Subanalyses using medical record data confirmed these results.

      Conclusions

      Overall, these results suggest that military deployments do not increase risk for miscarriage and impaired fecundity among U.S. servicewomen. However, because the point estimates for many of the exposures were elevated, more research is needed to better understand the potential risks associated with environmental exposures and specific types of combat exposures.
      Women account for 14.6% of the U.S. Armed Forces, with more than 200,000 women deploying in support of the recent conflicts in Iraq and Afghanistan (“

      Women in Military Service for America Memorial Foundation Inc. (2010). Available: www.womensmemorial.org/PDFs/StatsonWIM.pdf. Accessed: March 15, 2013.

      ). Before these conflicts, the number of women who deployed was much smaller, and women servicemembers often served in noncombat roles (“
      Iraq and Afghanistan Veterans of America
      Women warriors: Supporting she “who has borne the battle”.
      ). However, during the recent conflicts, servicewomen have reported in-theater combat experiences. This is in part due to the absence of traditional battle lines (
      • Vogt D.
      • Vaughn R.
      • Glickman M.E.
      • Schultz M.
      • Drainoni M.
      • Elwy R.
      • Eisen S.
      Gender differences in combat-related stressors and their association with postdeployment mental health in a nationally representative sample of US OEF/OIF veterans.
      ) as well as recent policy changes allowing women to take on a wider range of duties, including infantry roles on combat teams, thereby increasing their exposures to combat and intense stress (
      Joint Chiefs of Staff
      Elimination of the 1994 direct ground combat definition and assignment rule.
      ). With an increasing number of servicewomen deploying and experiencing combat, there is a growing need to evaluate if these experiences adversely affect reproductive health (“
      Iraq and Afghanistan Veterans of America
      Women warriors: Supporting she “who has borne the battle”.
      ).
      Deployments to the recent conflicts in Iraq and Afghanistan may have exposed women to both military-unique experiences as well as traditional factors associated with miscarriage and/or infertility. One known factor that has been associated with miscarriage and infertility is stress. The exact biological mechanism for which stress is believed to cause these adverse outcomes has not been confirmed, but the literature suggests that the immune and hormone alterations caused by stress may lead to reproductive failure (
      • Agarwal A.
      • Aponte-Mellado A.
      • Premkumar B.J.
      • Shaman A.
      • Gupta S.
      The effects of oxidative stress on female reproduction: A review.
      ,
      • Madhappan B.
      • Kempuraj D.
      • Christodoulou S.
      • Tsapikidis S.
      • Boucher W.
      • Karagiannis V.
      • Theoharides T.C.
      High levels of intrauterine corticotropin-releasing hormone, urocortin, tryptase, and interleukin-8 in spontaneous abortions.
      ,
      • Nakamura K.
      • Sheps S.
      • Arck P.C.
      Stress and reproductive failure: Past notions, present insights and future directions.
      ). Given that research in nonmilitary populations indicates that experiencing life stressors has an adverse effect on reproductive outcomes, it is important to determine if the unique stressors of military deployments are associated with adverse reproductive outcomes, such as miscarriage and impaired fecundity.
      Previous research investigating the association of deployment with miscarriage and impaired fecundity among U.S. servicewomen is sparse and most of this research has focused on veterans (
      • Araneta M.
      • Kamens D.
      • Zau A.
      • Gastanaga V.
      • Schlangen K.
      • Hiliopoulos K.
      • Gray G.
      Conception and pregnancy during the Persian Gulf War: The risk to women veterans.
      ,
      Armed Forces Health Surveillance Center
      Health of women after wartime deployments: Correlates of risk for selected medical conditions among females after initial and repeat deployments to Afghanistan and Iraq, active component, U.S. Armed Forces.
      ,
      • Doyle P.
      • Maconochie N.
      • Davies G.
      • Maconochie I.
      • Pelerin M.
      • Prior S.
      • Lewis S.
      Miscarriage, stillbirth and congenital malformation in the offspring of UK veterans of the first Gulf war.
      ,
      • Doyle P.
      • Maconochie N.
      • Ryan M.
      Reproductive health of Gulf War veterans.
      ,
      • Kang H.
      • Magee C.
      • Mahan C.
      • Lee K.
      • Murphy F.
      • Jackson L.
      • Matanoski G.
      Pregnancy outcomes among U.S. Gulf War veterans: A population-based survey of 30,000 veterans.
      ,
      • Katon J.
      • Cypel Y.
      • Raza M.
      • Zephyrin L.
      • Reiber G.
      • Yano E.M.
      • Schneiderman A.
      Self-reported infertility among male and female veterans serving during Operation Enduring Freedom/Operation Iraqi Freedom.
      ,
      • Mattocks K.
      • Kroll-Desrosiers A.
      • Zephyrin L.
      • Katon J.
      • Weitlauf J.
      • Bastian L.
      • Brandt C.
      Infertility care among OEF/OIF/OND women veterans in the Department of Veterans Affairs.
      ,
      • Wells T.
      • Wang L.
      • Spooner C.
      • Smith T.
      • Hiliopoulos K.
      • Kamens D.
      • Sato P.
      Self-reported reproductive outcomes among male and female 1991 Gulf War era US military veterans.
      ). The majority of the miscarriage studies did not find an association of deployment to the 1991 Gulf War with miscarriage in pregnancies conceived after the war by women veterans (
      • Doyle P.
      • Maconochie N.
      • Davies G.
      • Maconochie I.
      • Pelerin M.
      • Prior S.
      • Lewis S.
      Miscarriage, stillbirth and congenital malformation in the offspring of UK veterans of the first Gulf war.
      ,
      • Doyle P.
      • Maconochie N.
      • Ryan M.
      Reproductive health of Gulf War veterans.
      ,
      • Kang H.
      • Magee C.
      • Mahan C.
      • Lee K.
      • Murphy F.
      • Jackson L.
      • Matanoski G.
      Pregnancy outcomes among U.S. Gulf War veterans: A population-based survey of 30,000 veterans.
      ,
      • Wells T.
      • Wang L.
      • Spooner C.
      • Smith T.
      • Hiliopoulos K.
      • Kamens D.
      • Sato P.
      Self-reported reproductive outcomes among male and female 1991 Gulf War era US military veterans.
      ). Although infertility among servicemembers and veterans who deployed in support of the operations in Iraq and Afghanistan has been studied, to our knowledge, only one previous report examined infertility among the current era of U.S. servicewomen (
      Armed Forces Health Surveillance Center
      Health of women after wartime deployments: Correlates of risk for selected medical conditions among females after initial and repeat deployments to Afghanistan and Iraq, active component, U.S. Armed Forces.
      ). Two of the previous studies described the prevalence of infertility (∼2% have received a diagnosis of infertility, 16% reported history of infertility) among women veterans who received care at a VA facility, although neither study reported the effects of specific deployment related-stressor (
      • Katon J.
      • Cypel Y.
      • Raza M.
      • Zephyrin L.
      • Reiber G.
      • Yano E.M.
      • Schneiderman A.
      Self-reported infertility among male and female veterans serving during Operation Enduring Freedom/Operation Iraqi Freedom.
      ,
      • Mattocks K.
      • Kroll-Desrosiers A.
      • Zephyrin L.
      • Katon J.
      • Weitlauf J.
      • Bastian L.
      • Brandt C.
      Infertility care among OEF/OIF/OND women veterans in the Department of Veterans Affairs.
      ). One previous report found longer duration of deployment associated with an increased risk for infertility (
      Armed Forces Health Surveillance Center
      Health of women after wartime deployments: Correlates of risk for selected medical conditions among females after initial and repeat deployments to Afghanistan and Iraq, active component, U.S. Armed Forces.
      ). Moreover, the majority of these previous studies were limited by being cross-sectional or data linkage studies.
      Exploring the relationship of recent deployments and combat-related experiences with miscarriage and impaired fecundity may help to better understand whether serving in the military has a negative effect on women's reproductive health. Our longitudinal study aimed to evaluate if U.S. military servicewomen who deployed and experienced combat during the recent conflicts in Iraq and Afghanistan subsequently experienced an increased odds of miscarriage or impaired fecundity. The Millennium Cohort Study, as the largest longitudinal military study in U.S. history, has the unique ability to prospectively investigate women's reproductive health in relation to military experiences.

      Material and Methods

      Study Population and Data Sources

      The study population was derived from the Millennium Cohort Study, a longitudinal study designed to evaluate the impact of military service on the long-term health of U.S. servicemembers. Using an in-depth questionnaire, military, health, occupational factors, and life experiences were assessed prospectively at 3-year intervals. The methodology of the Millennium Cohort Study has been described previously (
      • Ryan M.
      • Smith T.
      • Smith B.
      • Amoroso P.
      • Boyko E.
      • Gray G.
      Millennium Cohort: Enrollment begins a 21-year contribution to understanding the impact of military service.
      ,
      • Smith T.C.
      The Millennium Cohort Study
      The US Department of Defense Millennium Cohort Study: Career span and beyond longitudinal follow-up.
      ). The study currently includes more than 200,000 participants who enrolled during four separate cycles between 2001 and 2013.
      The population for the current study consisted of women from the first two enrollment cycles, or panels, who completed surveys during the 2004 through 2006 and 2007 through 2008 survey cycles. To be eligible, women had to be aged 18 to 45 years and to not have separated from the military before September 2001 (n = 14,480). After exclusion criteria, 3,366 women were included in the miscarriage analysis and 11,183 were included in the impaired fecundity analysis (Figure 1). The population for each outcome was stratified by panel owing to differing enrollment criteria. Exposures and covariates were measured on the 2004 through 2006 questionnaire, while outcomes were measured on the 2007 through 2008 questionnaire; therefore, for these analyses, the data collected from 2004 through 2006 are called baseline and the data collected from 2007 through 2008 are called follow-up (Figure 2).
      Data sources included the Millennium Cohort Study questionnaire and electronic Department of Defense (DoD) personnel data provided by the Defense Manpower Data Center. The Millennium Cohort Study questionnaire was used to assess self-reported data, including information on self-reported miscarriage, self-perceived impaired fecundity, combat exposures, and other lifestyle and health metrics. Electronic personnel data from Defense Manpower Data Center was used to assess deployments in support of the recent operations in Iraq and Afghanistan, and military and demographic characteristics. The study was approved by the institutional review boards Naval Health Research Center and San Diego State University, and informed consent was obtained from all participants.

      Outcomes

      We assessed two outcomes using data from the Millennium Cohort Study questionnaire: self-reported miscarriage and self-perceived impaired fecundity. Miscarriage was assessed using the question, “Have you had a miscarriage within the last 3 years?” to which participants could respond, “Yes,” “No,” or “Does Not Apply,” on the follow-up questionnaire. Miscarriage was assessed among women who reported a pregnancy during the study time and had responded “Yes” or “No” to the miscarriage question.
      Self-perceived impaired fecundity was assessed using the question on the follow-up questionnaire, “During the last 3 years have you tried and been unable to become pregnant?”, with the same response options as the miscarriage question. Self-perceived impaired fecundity was assessed among women who responded “Yes” or “No” to the self-perceived impaired fecundity question.
      In a subanalysis, medical encounter data obtained from the Military Health System Data Repository and methodology used by the DoD Birth and Infant Health Registry was used to assess the outcomes (
      • Ryan M.A.
      • Pershyn-Kisor M.A.
      • Honner W.K.
      • Smith T.C.
      • Reed R.J.
      • Gray G.C.
      The Department of Defense Birth Defects Registry: Overview of a new surveillance system.
      ). Miscarriage was determined by having at least one of the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnostic codes that indicate a pregnancy loss (ectopic/molar pregnancy: 630.xx–633.xx; other pregnancy with abortive outcome: 634.xx–639.xx; intrauterine death: 656.4x; outcome of delivery: V27.[1,3,4,6,7]), procedure (removal of ectopic/molar pregnancy: 66.62, 68.0, 74.3), or Current Procedural Terminology codes (treatment of molar/ectopic pregnancy: 59100, 59120, 59121, 59130, 59135, 59136, 59140, 59150, 59151, 59870) between 2004 and 2007. Pregnancies ending in elective abortions and those thought to have a miscoded pregnancy loss (owing to a live birth outcome up to 36 weeks after the suspected false code) were excluded from analyses. To assess infertility in the medical encounter data, the ICD-9-CM diagnostic codes used were 628.x (infertility diagnosis from a medical professional) (
      • Conlin A.M.
      • Bukowinski A.T.
      • Sevick C.J.
      • DeScisciolo C.
      • Crum-Cianflone N.F.
      Safety of the pandemic H1N1 influenza vaccine among pregnant U.S. military women and their newborns.
      ).

      Main Exposures of Interest

      The three main exposures assessed in these analyses were deployment status, cumulative days deployed, and life stressors. The two deployment-related exposures were determined using electronic DoD personnel files that include deployment dates for those deployed since September 11, 2001. Participants who returned from deployment before completing their baseline questionnaire were considered deployed. Deployment status was categorized as nondeployed, deployed without combat, and deployed with combat. Among those deployed, combat experience was determined based on at least one positive response of being personally exposed to dead and/or decomposing bodies, maimed soldiers or civilians, prisoners of war/refugees, witnessing death, or witnessing physical abuse (torture, beating, rape). Cumulative deployment length was calculated as the total number of days deployed within the 3-year period before their baseline questionnaire submission and was categorized as 0, 1 to 180, and more than 180 days. The third exposure, life stressors, was assessed at baseline among panel 1 participants only using questions on the Millennium Cohort questionnaire. Panel 1 participants indicated if they had experienced any of the listed stressors in the past 3 years, such as divorce or separation, major financial problems, and forced sexual relations or sexual assault. Participants were classified as having low/mild or moderate/major life stressors based on a modified version of the Social Readjustment Rating Scale scoring system (
      • Hobson C.
      • Kamen J.
      • Szoste J.
      • Nethercut C.M.
      • Tiedmann J.
      • Wojnarowicz S.
      Stressful life events: A revision and update of the Social Readjustment Rating Scale.
      ,
      • Holmes T.H.
      • Rahe R.H.
      The social readjustment rating scale.
      ). Life stressors were not examined among the panel 2 participants because these items were assessed as lifetime stressors.

      Covariates

      Covariates were selected based on previous research indicating an association with reproductive health outcomes or because they were of interest to the military. Demographic characteristics were assessed using electronic personnel files and included sex, birth date, highest education level, marital status, race/ethnicity, pay grade, service component, service branch, and occupation at baseline.
      Several characteristics were assessed using self-reported data from the baseline questionnaire. These characteristics are modifiable behaviors that are known to have negative reproductive health effects. Current, past, and nonsmokers were identified based on self-report of having smoked at least 100 cigarettes (five packs) in their lifetime and if they had successfully quit smoking (
      • Smith B.
      • Ryan M.
      • Wingard D.
      • Patterson T.
      • Slymen D.
      • Macera C.
      Cigarette smoking and military deployment: A prospective evaluation.
      ). Heavy weekly drinking was defined as consuming more than seven drinks per week (
      • Jacobson I.
      • Ryan M.
      • Hooper T.
      • Smith T.
      • Amoroso P.
      • Boyko E.
      Alcohol use and alcohol-related problems before and after military combat deployment.
      ). Alcohol-related problems were assessed using the Patient Health Questionnaire alcohol screening tool; participants who reported at least one risky drinking behavior, such as driving under the influence more than once over the past 12 months, screened positive for alcohol-related problems (
      • Jacobson I.
      • Ryan M.
      • Hooper T.
      • Smith T.
      • Amoroso P.
      • Boyko E.
      Alcohol use and alcohol-related problems before and after military combat deployment.
      ,
      • Spitzer R.
      • Williams J.
      • Kroenke K.
      • Hornyak R.
      • McMurray J.
      Validity and utility of the PRIME-MD Patient Health Questionnaire in assessment of 3000 obstetric-gynecologic patients: The PRIME-MD Patient Health Questionnaire Obstetrics Gynecology Study.
      ,
      • Spitzer R.
      • Williams J.
      • Kroenke K.
      • Linzer M.
      • deGruy 3rd, F.
      • Hahn S.
      • Johnson J.
      Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study.
      ).
      Posttraumatic stress disorder (PTSD) was assessed using data from the PTSD Checklist-Civilian Version. A positive PTSD screen was based on criteria established in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (
      • Brewin C.
      Systematic review of screening instruments for adults at risk of PTSD.
      ,
      • Spitzer R.
      • Kroenke K.
      • Williams J.
      Primary PHQ. Validation and utility of a self-report version of PRIME-MD - The PHQ primary care study.
      ). Using a standardized Patient Health Questionnaire scoring mechanism, major depression, panic syndrome, and other anxiety syndrome were assessed (
      • Kroenke K.
      • Spitzer R.L.
      • Williams J.B.
      The PHQ-9: Validity of a brief depression severity measure.
      ,
      • Spitzer R.
      • Kroenke K.
      • Williams J.
      Primary PHQ. Validation and utility of a self-report version of PRIME-MD - The PHQ primary care study.
      ,
      • Spitzer R.
      • Williams J.
      • Kroenke K.
      • Hornyak R.
      • McMurray J.
      Validity and utility of the PRIME-MD Patient Health Questionnaire in assessment of 3000 obstetric-gynecologic patients: The PRIME-MD Patient Health Questionnaire Obstetrics Gynecology Study.
      ,
      • Spitzer R.
      • Williams J.
      • Kroenke K.
      • Linzer M.
      • deGruy 3rd, F.
      • Hahn S.
      • Johnson J.
      Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study.
      ). Participants were classified as having a positive psychiatric history if they screened positive for one or more of the following: PTSD, major depression, panic syndrome, or other anxiety syndrome. Psychotropic medication use was assessed based on self-report of medication for anxiety, depression, or stress on the baseline questionnaire.
      Prior medical conditions that may be associated with adverse reproductive outcomes were assessed based on self-report of provider-diagnosed conditions and included hypertension, thyroid condition other than cancer, seizures, lupus, diabetes or sugar diabetes, rheumatoid arthritis, or frequent bladder infections. Participants were classified as having none of these conditions or one or more condition(s). Data on body mass index were analyzed by using self-reported weight and height and categorized as underweight/normal (≤24.9 kg/m2), overweight (25–29.9 kg/m2), or obese (≥30 kg/m2).
      Exposures to pesticides and chemicals were assessed based on self-report of being personally exposed to pesticides (including creams, sprays, or uniform treatments) or with pesticides applied in the environment or around living spaces; occupational hazards requiring protective equipment, such as respirators or hearing protection; or routine skin contact with paint and/or solvents and/or substances. The number of pesticide and chemical exposures reported was summed and categorized as zero, one, or two.

      Statistical Methods

      Univariate analyses were used to assess the relationship between the two outcomes of miscarriage and self-perceived impaired fecundity, the three main exposures, and all covariates. Separate analyses were performed for each panel. Multivariable logistic regression models were performed using separate models for each outcome (miscarriage and self-perceived impaired fecundity) and each of the three main exposures (deployment status, cumulative days deployed, and life stressors). Each model was adjusted for covariates that were found to be significant (p < .05), confounders (>10% change in the exposure estimate), and covariates that are of interest regardless of significance (a priori defined as age, occupation, and pesticide and chemical exposures). The Hosmer–Lemeshow test was used to test model fit and multicollinearity was assessed using a variance inflation factor of four or higher.
      Subanalyses were conducted using objective medical encounter data obtained from the Military Health System Data Repository to define both miscarriage and infertility using a subset of active duty study participants. Survival analysis using Cox proportional hazard models was performed because these data included date of diagnosis. Participants were censored at the diagnosis date, date of separation from military service, or follow-up survey date, whichever occurred first. All analyses were completed using SAS statistical software, version 9.3 (SAS Institute, Inc., Cary, NC).

      Results

      Miscarriage

      A total of 1,045 of 3,366 servicewomen (31.0%) between the ages of 18 and 45 years who had a pregnancy reported a miscarriage during the approximate 3-year follow-up period. Notably, women in panel 1 who reported a miscarriage were proportionally more likely to be older, college educated, and officers; in addition, they were more likely to report one or more prior medical conditions, prior miscarriage, and previous impaired fecundity (Table 1). Women in panel 2 who reported a miscarriage were proportionally more likely to be past smokers; they were also more likely to report psychiatric symptoms, prior miscarriage, and impaired fecundity.
      Table 1Demographic, Military, Behavioral, and Health Characteristics of Millennium Cohort Servicewomen by Self-Reported Miscarriage
      Self-report of miscarriage on the follow-up survey.
      Status
      Characteristics
      All characteristics assessed from the baseline survey, except for age, which was assessed using the follow-up survey.
      Panel 1 (n = 1,511)Panel 2 (n = 1,855)
      No Miscarriage (n = 1,038)Miscarriage (n = 473)No Miscarriage (n = 1,283)Miscarriage (n = 572)
      Mean ± SD age32.8 ± 4.634.4 ± 5.427.1 ± 3.827.4 ± 4.5
      Education
       Some college or less673 (64.8)280 (59.2)1,072 (83.6)489 (85.5)
       Bachelor's degree or higher365 (35.2)193 (40.8)211 (16.4)83 (14.5)
      Marital status
       Not married336 (32.4)162 (34.3)827 (64.5)380 (66.4)
       Married702 (67.6)311 (65.8)456 (35.5)192 (33.6)
      Race/ethnicity
       Non-Hispanic White667 (64.3)296 (62.6)852 (66.4)374 (65.4)
       Non-Hispanic Black166 (16.0)84 (17.8)203 (15.8)91 (15.9)
       Other205 (19.7)93 (19.7)228 (17.8)107 (18.7)
      Service component
       Reserve/National Guard621 (59.8)291 (61.5)452 (35.2)199 (34.8)
       Active duty417 (40.2)182 (38.5)831 (64.8)373 (65.2)
      Service branch
       Army549 (52.9)243 (51.4)609 (47.5)290 (50.7)
       Navy/Coast Guard183 (17.6)80 (16.9)244 (19.0)113 (19.8)
       Marine Corps23 (2.2)11 (2.3)42 (3.3)10 (1.7)
       Air Force283 (27.3)139 (29.4)388 (30.2)159 (27.8)
      Occupation
       Administration and supply451 (43.4)200 (42.3)502 (39.1)228 (39.9)
       Combat specialist73 (7.0)31 (6.6)78 (6.1)48 (8.4)
       Health care218 (21.0)118 (24.9)229 (17.8)103 (18.0)
       Other296 (28.5)124 (26.2)474 (36.9)193 (33.7)
      Pay grade
       Enlisted770 (74.2)315 (66.6)1,131 (88.2)508 (88.8)
       Officer268 (25.8)158 (33.4)152 (11.8)64 (11.2)
      Pesticide/chemical exposures
      Variable measures exposure to pesticides and chemical exposures. The variable was based on self-report of exposure to occupational hazards requiring protective equipment, routine skin contact with paint and/or solvent and/or substances, pesticides applied to the environment, living spaces, and/or to yourself and/or clothing.
       None528 (50.9)230 (48.6)557 (43.4)225 (39.3)
       1280 (27.0)126 (26.6)385 (30.0)188 (32.9)
       2230 (22.2)117 (24.7)341 (26.6)159 (27.8)
      Body mass index
       Underweight/normal562 (54.1)266 (56.2)838 (65.3)363 (63.5)
       Overweight355 (34.2)151 (31.9)360 (28.1)180 (31.5)
       Obese121 (11.7)56 (11.8)85 (6.6)29 (5.1)
      Psychotropic medication use
      Self-reported use of medication to treat anxiety, depression, or stress.
       No974 (93.8)433 (91.5)1,212 (94.5)534 (93.4)
       Yes64 (6.2)40 (8.5)71 (5.5)38 (6.6)
      Psychiatric symptoms
      Assessed using screening tools for posttraumatic stress disorder, depression, anxiety, or panic.
       No951 (91.6)419 (88.6)1,148 (89.5)488 (85.3)
       Yes87 (8.4)54 (11.4)135 (10.5)84 (14.7)
      Prior medical history
      Self-report of physician diagnosis to any of the following: hypertension, thyroid condition other that cancer, seizures, lupus, diabetes or sugar diabetes, rheumatoid arthritis, and/or frequent bladder infections.
       None904 (87.1)379 (80.1)1,130 (88.1)494 (86.4)
       1 or more condition(s)134 (12.9)94 (19.9)153 (11.9)78 (13.6)
      Smoking status
       Never smoker681 (65.6)316 (66.8)784 (61.1)336 (58.7)
       Past smoker113 (10.9)60 (12.7)211 (16.4)126 (22.0)
       Current smoker244 (23.5)97 (20.5)288 (22.4)110 (19.2)
      Heavy weekly drinking
      Heavy weekly drinking is defined as consuming more than seven drinks per week.
       No954 (91.9)433 (91.5)1,162 (90.6)509 (89.0)
       Yes84 (8.1)40 (8.5)121 (9.4)63 (11.0)
      Alcohol-related problems
      Alcohol-related problems was assessed using the Patient Health Questionnaire.
       No986 (95.0)449 (94.9)1,165 (90.8)511 (89.3)
       Yes52 (5.0)24 (5.1)118 (9.2)61 (10.7)
      Previously reported miscarriage
      Self-reported miscarriage at baseline.
       No932 (89.8)320 (67.7)1,168 (91.0)410 (71.7)
       Yes106 (10.2)153 (32.3)115 (9.0)162 (28.3)
      Previously reported impaired fecundity
      Self-reported infertility at baseline is not included in main analyses for the miscarriage population. Frequencies are given as a part of subanalyses. Number of people missing this variable in Panel 1 = 3, Panel 2 = 1.
       No903 (87.0)386 (81.6)1,142 (89.0)482 (84.3)
       Yes133 (12.8)86 (18.2)141 (11.0)89 (15.6)
      Self-report of miscarriage on the follow-up survey.
      All characteristics assessed from the baseline survey, except for age, which was assessed using the follow-up survey.
      Variable measures exposure to pesticides and chemical exposures. The variable was based on self-report of exposure to occupational hazards requiring protective equipment, routine skin contact with paint and/or solvent and/or substances, pesticides applied to the environment, living spaces, and/or to yourself and/or clothing.
      § Self-reported use of medication to treat anxiety, depression, or stress.
      || Assessed using screening tools for posttraumatic stress disorder, depression, anxiety, or panic.
      Self-report of physician diagnosis to any of the following: hypertension, thyroid condition other that cancer, seizures, lupus, diabetes or sugar diabetes, rheumatoid arthritis, and/or frequent bladder infections.
      # Heavy weekly drinking is defined as consuming more than seven drinks per week.
      ∗∗ Alcohol-related problems was assessed using the Patient Health Questionnaire.
      †† Self-reported miscarriage at baseline.
      ‡‡ Self-reported infertility at baseline is not included in main analyses for the miscarriage population. Frequencies are given as a part of subanalyses. Number of people missing this variable in Panel 1 = 3, Panel 2 = 1.
      In the adjusted model, deployment experiences, with or without combat, were not associated with reporting miscarriage in either panel 1 or 2 (Table 2). In separate models, cumulative days deployed was not associated with reporting miscarriage in either panel 1 or 2 and reporting life stressors was not associated with reporting miscarriage in panel 1 (Table 2). Subanalyses completed using medical record data showed 136 miscarriage events among 1,041 eligible women. Consistent with the self-report findings, no associations were found between any of our main exposures and miscarriage in the multivariable models (data not shown).
      Table 2Frequencies, Prevalence, and AOR of Miscarriage
      Number and 3-year cumulative prevalence of self-reported miscarriage at follow-up.
      among Servicewomen in the Millennium Cohort
      ModelsPanel 1
      Models are adjusted for age, marital status, occupation, pesticide and chemical exposures, adverse prior medical history, and prior miscarriage at baseline.
      (n = 1,511)
      Panel 2
      Models adjusted for age, occupation, pesticide and chemical exposures, smoking status, and prior miscarriage at baseline.
      (n = 1,855)
      n (%)
      Number and 3-year cumulative prevalence of self-reported miscarriage at follow-up.
      AOR95% CIn (%)
      Number and 3-year cumulative prevalence of self-reported miscarriage at follow-up.
      AOR95% CI
      Deployment status
      Combat was defined as self-report of having personally experienced at least one of the following combat or combat-like situations: witnessing a person's death due to war, disaster, or tragic event; witnessing instances of physical abuse (torture, beating, rape); dead and/or decomposing bodies; maimed soldiers or civilians; and prisoners of war or refugees. This was assessed on the baseline survey as having occurred during the past 3 years.
       Nondeployed357 (30.9)1.00416 (30.7)1.00
       Deployed, without combat53 (27.8)1.05(0.73–1.51)54 (26.1)0.86(0.61–1.22)
       Deployed, with combat63 (38.7)1.41(0.96–2.07)102 (34.7)1.16(0.87–1.56)
      Cumulative days deployed
      Cumulative length deployed was assessed as the total number of days deployed from 3 years before the baseline assessment to the day the baseline survey was completed.
       0357 (30.9)1.00416 (30.7)1.00
       1–18055 (32.0)1.22(0.84–1.77)68 (30.0)1.01(0.73–1.39)
       >18061 (33.5)1.18(0.82–1.71)88 (32.1)1.05(0.77–1.42)
      Life stressors
      Life stressors was assessed using scoring mechanisms adapted from the Holmes and Rahe Social Readjustment Rating Scale.
       Low/mild441 (30.6)1.00
       Moderate/major32 (45.1)1.67(0.99–2.80)
      Abbreviations: AOR, adjusted odds ratio; CI, confidence interval.
      Number and 3-year cumulative prevalence of self-reported miscarriage at follow-up.
      Models are adjusted for age, marital status, occupation, pesticide and chemical exposures, adverse prior medical history, and prior miscarriage at baseline.
      Models adjusted for age, occupation, pesticide and chemical exposures, smoking status, and prior miscarriage at baseline.
      § Combat was defined as self-report of having personally experienced at least one of the following combat or combat-like situations: witnessing a person's death due to war, disaster, or tragic event; witnessing instances of physical abuse (torture, beating, rape); dead and/or decomposing bodies; maimed soldiers or civilians; and prisoners of war or refugees. This was assessed on the baseline survey as having occurred during the past 3 years.
      Cumulative length deployed was assessed as the total number of days deployed from 3 years before the baseline assessment to the day the baseline survey was completed.
      Life stressors was assessed using scoring mechanisms adapted from the Holmes and Rahe Social Readjustment Rating Scale.
      Factors associated with miscarriage in panel 1 included age (for every 5-year incremental increase in age, adjusted odds ratio [AOR], 1.49; 95% confidence interval [CI], 1.31–1.70), marital status (married women: AOR, 0.76; 95% CI, 0.60–0.98), prior medical condition (AOR, 1.44; 95% CI, 1.05–1.97), and prior history of a miscarriage (AOR, 3.95; 95% CI, 2.96–5.26). Among women in panel 2, factors associated with higher odds of miscarriage included smoking (past smokers, AOR, 1.34; 95% CI, 1.02–1.76) and prior history of miscarriage (AOR, 3.83; 95% CI, 2.93–5.00; data not shown in tables).
      To account for the possibility that some factors may mediate or confound the relationship between the main exposures and miscarriage, we reran the models without prior medical conditions and smoking status. Findings were consistent with the final models for both the combat exposure and cumulative days deployed, where miscarriage was not found to be associated with these exposures. However, life stressors were found to be marginally significant when these variables were removed from the panel 1 model (AOR, 1.75; 95% CI, 1.04–2.94; data not shown).

      Self-Perceived Impaired Fecundity

      A total of 1,268 of 11,183 women (11.3%) between the ages of 18 and 45 years self-reported impaired fecundity. Notably, women in panel 1 who reported self-perceived impaired fecundity were proportionally more likely to be married and report a prior miscarriage or previous impaired fecundity (Table 3). Women in panel 2 who reported self-perceived impaired fecundity were proportionally more likely to be married, active duty servicemembers, and past smokers; in addition, they were more likely to report psychiatric symptoms, one or more prior medical conditions, prior miscarriage, and previous impaired fecundity.
      Table 3Demographic, Military, Behavioral, and Health Characteristics of Millennium Cohort Servicewomen by Self-Perceived Impaired Fecundity
      Self-report of perceived impaired fecundity on the follow-up survey.
      Characteristics
      All characteristics assessed from the baseline survey, except for age, which was assessed using the follow-up survey.
      Panel 1 (n = 5,940)Panel 2 (n = 5,243)
      No Impaired Fecundity (n = 5,284)Impaired Fecundity (n = 656)No Impaired Fecundity (n = 4,631)Impaired Fecundity (n = 612)
      Mean ± SD age35.7 ± 5.735.4 ± 5.427.9 ± 4.728.4 ± 5.0
      Education
       Some college or less3,398 (64.3)414 (63.1)3,656 (78.9)506 (82.7)
       Bachelor's degree or higher1,886 (35.7)242 (36.9)975 (21.1)106 (17.3)
      Marital status
       Not married2,378 (45.0)255 (38.9)3,391 (73.2)405 (66.2)
       Married2,906 (55.0)401 (61.1)1,240 (26.8)207 (33.8)
      Race/ethnicity
       Non-Hispanic White3,388 (64.1)372 (56.7)3,100 (66.9)381 (62.3)
       Non-Hispanic Black985 (18.6)155 (23.6)696 (15.0)115 (18.8)
       Other911 (17.2)129 (19.7)835 (18.0)116 (19.0)
      Service component
       Reserve/National Guard3,080 (58.3)357 (54.4)2,113 (45.6)248 (40.5)
       Active duty2,204 (41.7)299 (45.6)2,518 (54.4)364 (59.5)
      Service branch
       Army2,662 (50.4)335 (51.1)2,203 (47.6)306 (50.0)
       Navy/Coast Guard914 (17.3)124 (18.9)861 (18.6)117 (19.1)
       Marine Corps126 (2.4)17 (2.6)119 (2.6)13 (2.1)
       Air Force1,582 (29.9)180 (27.4)1,448 (31.3)176 (28.8)
      Occupation
       Administration and supply2,391 (45.2)294 (44.8)1,687 (36.4)222 (36.3)
       Combat specialist336 (6.4)36 (5.5)304 (6.6)31 (5.1)
       Health care1,105 (20.9)149 (22.7)889 (19.2)125 (20.4)
       Other1,452 (27.5)177 (27.0)1,751 (37.8)234 (38.2)
      Pay grade
       Enlisted3,901 (73.8)485 (73.9)3,905 (84.3)538 (87.9)
       Officer1,383 (26.2)171 (26.1)726 (15.7)74 (12.1)
      Pesticide/chemical exposures
      Variable measures exposure to pesticides and chemical exposures. The variable was based on self-report of exposure to occupational hazards requiring protective equipment, routine skin contact with paint and/or solvent and/or substances, pesticides applied to the environment, living spaces, and/or to yourself and/or clothing.
       None2,676 (50.6)306 (46.6)1,929 (41.7)228 (37.3)
       11,426 (27.0)198 (30.2)1,458 (31.5)199 (32.5)
       21,182 (22.4)152 (23.2)1,244 (26.9)185 (30.2)
      Body mass index
       Underweight/normal2,893 (54.8)359 (54.7)3,143 (67.9)382 (62.4)
       Overweight1,863 (35.3)215 (32.8)1,279 (27.6)190 (31.0)
       Obese528 (10.0)82 (12.5)209 (4.5)40 (6.5)
      Psychotropic medication use
      Self-reported use of medication to treat anxiety, depression, or stress.
       No4,836 (91.5)604 (92.1)4,361 (94.2)557 (91.0)
       Yes448 (8.5)52 (7.9)270 (5.8)55 (9.0)
      Psychiatric symptoms
      Assessed using screening tools for posttraumatic stress disorder, depression, anxiety, or panic.
       No4,832 (91.4)587 (89.5)4,174 (90.1)503 (82.2)
       Yes452 (8.6)69 (10.5)457 (9.9)109 (17.8)
      Prior medical history
      Self-report of physician diagnosis to any of the following: hypertension, thyroid condition other that cancer, seizures, lupus, diabetes or sugar diabetes, rheumatoid arthritis, and/or frequent bladder infections.
       None4,457 (84.3)537 (81.9)4,177 (90.2)526 (85.9)
       1 or more condition(s)827 (15.7)119 (18.1)454 (9.8)86 (14.1)
      Smoking status
       Never smoker3,446 (65.2)419 (63.9)3,017 (65.1)347 (56.7)
       Past smoker686 (13.0)95 (14.5)732 (15.8)143 (23.4)
       Current smoker1,152 (21.8)142 (21.6)882 (19.0)122 (19.9)
      Heavy weekly drinking
      Heavy weekly drinking is defined as consuming more than seven drinks per week.
       No4,742 (89.7)570 (86.9)4,037 (87.2)530 (86.6)
       Yes542 (10.3)86 (13.1)594 (12.8)82 (13.4)
      Alcohol-related problems
      Alcohol-related problems was assessed using the Patient Health Questionnaire.
       No4,976 (94.2)620 (94.5)4,119 (88.9)541 (88.4)
       Yes308 (5.8)36 (5.5)512 (11.1)71 (11.6)
      Previously reported miscarriage
      Self-reported miscarriage at baseline is not included in main analyses for the infertility population. Frequencies are given as a part of subanalyses. Number of people missing this variable in Panel 1 = 92, Panel 2 = 70.
       No4,888 (92.5)563 (85.8)4,271 (92.2)522 (85.3)
       Yes315 (6.0)82 (12.5)302 (6.5)78 (12.7)
      Previously reported impaired fecundity
      Self-reported infertility was assessed at baseline.
       No5,045 (95.5)387 (59.0)4,380 (94.6)425 (69.4)
       Yes239 (4.5)269 (41.0)251 (5.4)187 (30.6)
      Self-report of perceived impaired fecundity on the follow-up survey.
      All characteristics assessed from the baseline survey, except for age, which was assessed using the follow-up survey.
      Variable measures exposure to pesticides and chemical exposures. The variable was based on self-report of exposure to occupational hazards requiring protective equipment, routine skin contact with paint and/or solvent and/or substances, pesticides applied to the environment, living spaces, and/or to yourself and/or clothing.
      § Self-reported use of medication to treat anxiety, depression, or stress.
      || Assessed using screening tools for posttraumatic stress disorder, depression, anxiety, or panic.
      Self-report of physician diagnosis to any of the following: hypertension, thyroid condition other that cancer, seizures, lupus, diabetes or sugar diabetes, rheumatoid arthritis, and/or frequent bladder infections.
      # Heavy weekly drinking is defined as consuming more than seven drinks per week.
      ∗∗ Alcohol-related problems was assessed using the Patient Health Questionnaire.
      †† Self-reported miscarriage at baseline is not included in main analyses for the infertility population. Frequencies are given as a part of subanalyses. Number of people missing this variable in Panel 1 = 92, Panel 2 = 70.
      ‡‡ Self-reported infertility was assessed at baseline.
      In the adjusted models, deployment experiences, with or without combat, were not associated with reporting self-perceived impaired fecundity in panel 1 or 2 (Table 4). In separate models, cumulative days deployed was not associated with reporting self-perceived impaired fecundity in either panel 1 or 2, nor was life stressors in panel 1 (Table 2). Subanalyses using medical record data found 213 events of infertility in 4,079 eligible women. Consistent with the self-reported findings, no associations were found between any of the main exposures and infertility in multivariable models (data not shown).
      Table 4Frequencies, Prevalence, and AOR of Self-Perceived Impaired Fecundity
      Number and 3-year cumulative prevalence of self-reported perceived impaired fecundity at follow-up.
      among Servicewomen in the Millennium Cohort
      ModelsPanel 1
      Models adjusted for age, race/ethnicity, occupation, pesticide and chemical exposures, heavy drinking, and prior perceived impaired fecundity at baseline.
      (n = 5,940)
      Panel 2
      Models adjust for age, race/ethnicity, occupation, pesticide and chemical exposures, smoking status, psychiatric symptoms, and prior perceived impaired fecundity at baseline.
      (n = 5,243)
      N (%)
      Number and 3-year cumulative prevalence of self-reported perceived impaired fecundity at follow-up.
      AOR95% CIN (%)
      Number and 3-year cumulative prevalence of self-reported perceived impaired fecundity at follow-up.
      AOR95% CI
      Deployment status
      Combat was defined as self-report of having personally experienced at least one of the following combat or combat-like situations: witnessing a person's death due to war, disaster, or tragic event; witnessing instances of physical abuse (torture, beating, rape); dead and/or decomposing bodies; maimed soldiers or civilians; and prisoners of war or refugees. This was assessed on the baseline survey as having occurred during the past 3 years.
       Nondeployed485 (10.7)1.00418 (11.2)1.00
       Deployed, without combat87 (11.0)0.96(0.73–1.27)67 (9.9)0.86(0.64–1.15)
       Deployed, with combat84 (13.6)1.28(0.95–1.72)127 (15.1)1.23(0.96–1.57)
      Cumulative days deployed
      Cumulative length deployed was assessed as the total number of days deployed from 3 years before the baseline assessment to the day the baseline survey was completed.
       0485 (10.7)1.00418 (11.2)1.00
       1–18087 (12.2)1.08(0.82–1.43)91 (12.6)1.10(0.84–1.42)
       >18084 (12.1)1.10(0.83–1.47)103 (12.9)1.02(0.79–1.32)
      Life stressors
      Life stressors was assessed using scoring mechanisms adapted from the Holmes and Rahe Social Readjustment Rating Scale.
       Low/mild612 (10.8)1.00
       Moderate/major44 (15.5)1.24(0.85–1.82)
      Abbreviations: AOR, adjusted odds ratio; CI, confidence interval.
      Number and 3-year cumulative prevalence of self-reported perceived impaired fecundity at follow-up.
      Models adjusted for age, race/ethnicity, occupation, pesticide and chemical exposures, heavy drinking, and prior perceived impaired fecundity at baseline.
      Models adjust for age, race/ethnicity, occupation, pesticide and chemical exposures, smoking status, psychiatric symptoms, and prior perceived impaired fecundity at baseline.
      § Combat was defined as self-report of having personally experienced at least one of the following combat or combat-like situations: witnessing a person's death due to war, disaster, or tragic event; witnessing instances of physical abuse (torture, beating, rape); dead and/or decomposing bodies; maimed soldiers or civilians; and prisoners of war or refugees. This was assessed on the baseline survey as having occurred during the past 3 years.
      || Cumulative length deployed was assessed as the total number of days deployed from 3 years before the baseline assessment to the day the baseline survey was completed.
      Life stressors was assessed using scoring mechanisms adapted from the Holmes and Rahe Social Readjustment Rating Scale.
      Among women in panel 1, factors associated with higher odds of perceived impaired fecundity included race/ethnicity (non-Hispanic Black women in comparison with non-Hispanic White women: AOR, 1.34 [95% CI, 1.07–1.68]; other race/ethnicity: AOR, 1.27 [95% CI, 1.01–1.61]), heavy weekly drinking (AOR, 1.40; 95% CI, 1.07–1.84), and previous impaired fecundity (AOR, 14.45; 95% CI, 11.78–17.74). Among the women in panel 2, factors associated with higher odds of perceived impaired fecundity included race/ethnicity (non-Hispanic Black women, in comparison to non-Hispanic White women: AOR, 1.34; 95% CI, 1.04–1.72), age (for every 5-year incremental increase in age: AOR, 1.16; 95% CI, 1.01–1.33), psychiatric symptoms (AOR, 1.48; 95% CI, 1.15–1.90), smoking (past smokers, AOR, 1.52; 95% CI, 1.21–1.92), and previously reported self-perceived impaired fecundity (AOR, 6.91; 95% CI, 5.55–8.61; data not shown in tables).
      To account for the possibility that some factors may mediate or confound the relationship between perceived impaired fecundity, we reran the models without heavy weekly drinking, psychiatric symptoms, and smoking status. Findings were consistent with the final models for the combat exposure, cumulative days deployed, and life stressors, where perceived impaired fecundity was not found to be associated with these exposures (data not shown).

      Discussion

      Our findings suggest experiencing deployment and combat-related exposures are not associated with miscarriage and self-perceived impaired fecundity among a large cohort of military women who served during the recent conflicts. Moreover, self-reported findings in all models were consistent with objective outcome data when assessed using medical records. Although none of the main exposures were found to be statistically significant in association with miscarriage or impaired fecundity, the point estimate for those deployed with combat experience were increased for all models. Overall, these findings indicate that recent deployment and combat experiences do not seem to have significant adverse effects at a population level on servicewomen's reproductive health. However, because the point estimates for many of the exposures were increased, more research is needed to understand potential risks associated with specific exposures, such as combat. Our findings indicate that life stressors, which have historically been associated negatively with reproductive health outcomes, and military-specific characteristics (such as service branch or component, occupation, and rank) were not associated with miscarriage or impaired fecundity after adjusting for additional factors.
      Previous studies from the first Gulf War have found minimal associations between serving in combat-related roles and negative reproductive health conditions among men (
      • Araneta M.
      • Kamens D.
      • Zau A.
      • Gastanaga V.
      • Schlangen K.
      • Hiliopoulos K.
      • Gray G.
      Conception and pregnancy during the Persian Gulf War: The risk to women veterans.
      ,
      • Doyle P.
      • Maconochie N.
      • Davies G.
      • Maconochie I.
      • Pelerin M.
      • Prior S.
      • Lewis S.
      Miscarriage, stillbirth and congenital malformation in the offspring of UK veterans of the first Gulf war.
      ,
      • Doyle P.
      • Maconochie N.
      • Ryan M.
      Reproductive health of Gulf War veterans.
      ,
      • Kang H.
      • Magee C.
      • Mahan C.
      • Lee K.
      • Murphy F.
      • Jackson L.
      • Matanoski G.
      Pregnancy outcomes among U.S. Gulf War veterans: A population-based survey of 30,000 veterans.
      ,
      • Wells T.
      • Wang L.
      • Spooner C.
      • Smith T.
      • Hiliopoulos K.
      • Kamens D.
      • Sato P.
      Self-reported reproductive outcomes among male and female 1991 Gulf War era US military veterans.
      ). With the exception of one study (
      • Araneta M.
      • Kamens D.
      • Zau A.
      • Gastanaga V.
      • Schlangen K.
      • Hiliopoulos K.
      • Gray G.
      Conception and pregnancy during the Persian Gulf War: The risk to women veterans.
      ), our findings are consistent with the few previous studies that found no association between deployment and miscarriages. Although one previous report found an association between duration of deployment and infertility, our study did not find any significant associations between deployment-related factors and infertility (
      Armed Forces Health Surveillance Center
      Health of women after wartime deployments: Correlates of risk for selected medical conditions among females after initial and repeat deployments to Afghanistan and Iraq, active component, U.S. Armed Forces.
      ). Our prospective study, however, was one of the first to be able to incorporate the timing of events and adjust for numerous military, occupational, demographic, and behavioral characteristics that are known to be risk factors for these outcomes. This ability allowed us to account for the potential effects of these characteristics, and to isolate the effects that deployment status and duration of deployment may have on the reproductive health of women Service members. However, we are not suggesting that all types of various deployment exposures, such as burn pits or blast exposures, have no effect on reproductive health. More research is needed to examine if specific deployment experiences may impact reproductive health of servicewomen and women veterans.
      A recent study among women who served in Iraq and Afghanistan found an association between mental disorders and adverse reproductive health outcomes (
      • Cohen B.E.
      • Maguen S.
      • Bertenthal D.
      • Shi Y.
      • Jacoby V.
      • Seal K.H.
      Reproductive and other health outcomes in Iraq and Afghanistan women veterans using VA health care: Association with mental health diagnoses.
      ). This recent study indicates that there may be a possible indirect relationship between combat and negative reproductive health outcomes, owing to combat-related mental disorders. Our study specifically examined deployment experiences after adjusting for several health factors, including mental disorders, and found no association. However, because combat experiences are known to be associated with mental disorders, we performed a subanalysis to investigate if deployment may have an indirect association with these negative reproductive health outcomes. Even after removing some of these potential mediators from our models, only one model indicated a significant association (life stressors with impaired fecundity). The results from these additional analyses further suggest that deployment and combat are neither directly nor indirectly associated with miscarriages or impaired fecundity.
      We found prior miscarriage or impaired fecundity was the strongest risk factor associated with subsequent reporting of miscarriage or impaired fecundity at follow-up among both panels. Additionally, a prior history of medical problems (e.g., hypertension, diabetes, hypertension, thyroid conditions) was significantly associated with reporting miscarriage in panel 1. Incremental increase of age, smoking, and heavy alcohol consumption were also found to be significantly correlated with the outcomes. These are known risk factors of adverse reproductive health of women. The fact that our findings are consistent with the previous literature helps validate our study (
      Armed Forces Health Surveillance Center
      Health of women after wartime deployments: Correlates of risk for selected medical conditions among females after initial and repeat deployments to Afghanistan and Iraq, active component, U.S. Armed Forces.
      ,
      • Artini P.
      • Uccelli A.
      • Papini F.
      • Simi G.
      • Di Berardino O.
      • Ruggiero M.
      • Cela V.
      Infertility and pregnancy loss in euthyroid women with thyroid autoimmunity.
      ,
      • Bhattacharya S.
      • Townend J.
      • Bhattacharya S.
      Recurrent miscarriage: Are three miscarriages one too many? Analysis of a Scottish population-based database of 151,021 pregnancies.
      ,
      • Campbell S.
      • Lynch J.
      • Esterman A.
      • McDermott R.
      Pre-pregnancy predictors linked to miscarriage among Aboriginal and Torres Strait Islander women in North Queensland Australian and New Zealand.
      ,
      • Chiodo L.
      • Bailey B.A.
      • Sokol R.J.
      • Janisse J.
      • Delaney-Black V.
      • Hannigan J.H.
      Recognized spontaneous abortion in mid-pregnancy and patterns of pregnancy alcohol use.
      ).
      This study has several limitations. Although we were able to adjust for many known risk factors of adverse reproductive outcomes, other factors known to cause miscarriage or impair fecundity, such as information on the male partner's characteristics and comprehensive female obstetrical/gynecological information, were not available. It is important to consider that women who are pregnant are not allowed to deploy, which may affect the associations between our outcomes and exposures. Our study also lacked precise timing between return from deployment and incidence of miscarriage or onset of impaired fecundity, which was mitigated by the subanalysis performed using medical record data to compare the results obtained from using the self-reported data. The use of medical record data also mitigated the limitation of using a non-standard self-reported measure of impaired fecundity. In addition, self-reported data, including the miscarriage and impaired fecundity data, are subject to recall bias, especially among women who are concerned over the exposures they experienced on their deployment. Self-reported miscarriage data may also be underreported, although many women often do no not visit a health care provider when a miscarriage is experienced or when fertility issues arise, making self-reported data potentially more complete than data from medical records.
      This study has several strengths. To our knowledge, this study is one of the first to prospectively analyze the association between military factors and the reproductive outcomes of miscarriage and self-perceived impaired fecundity among women Service members who served in support of the recent conflicts in Iraq and Afghanistan. Further, our sample included all branches of the U.S. military, Reserve and National Guard members, and women from all military occupations. The survey tool also allowed capture of behaviors not found in medical records, such as alcohol consumption and tobacco use, which are known risk factors for these outcomes, as well as military-specific exposures, including combat (
      • Campbell S.
      • Lynch J.
      • Esterman A.
      • McDermott R.
      Pre-pregnancy predictors linked to miscarriage among Aboriginal and Torres Strait Islander women in North Queensland Australian and New Zealand.
      ,
      • Chiodo L.
      • Bailey B.A.
      • Sokol R.J.
      • Janisse J.
      • Delaney-Black V.
      • Hannigan J.H.
      Recognized spontaneous abortion in mid-pregnancy and patterns of pregnancy alcohol use.
      ).

      Implications for Practice and/or Policy

      Our findings of no associations between deployment and these reproductive health outcomes may provide reassurance to women who deploy as well as to DoD and Department of Veterans Affairs leaders and policymakers. Findings associated with adverse reproductive health events in this study were similar to those in the general population. As such, policies and guidelines should continue to focus on the importance of smoking cessation programs, alcohol abuse counseling, and the prevention/treatment of chronic medical conditions.

      Conclusions

      Women servicemembers who deployed in support of the recent operations in Iraq and Afghanistan did not have an increased odds of miscarriage or perceived impaired fecundity after deployment in our study sample compared with those who had not deployed. Because reproductive health issues include other outcomes in addition to miscarriage and impaired fecundity, and because it was not possible to examine specific environmental exposures associated with deployments or specific combat exposures, additional research on the reproductive health of military personnel is essential to ensure the ongoing health and well-being of our servicemembers and their families.

      Acknowledgments

      Andrea Ippolito had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. In addition to the authors, the Millennium Cohort Study Team includes Richard Armenta, PhD; Lauren Bauer, MPH; Deborah Boggs Bookwalter, PhD; Ania Bukowinski, MPH; Madeline Cross; James Davies; Carrie Donoho, PhD; Alex Esquivel; CDR Dennis Faix, MD; Lt Col Susan Farrish, MD; Toni Rose Geronimo; Isabel Jacobson, MPH; Claire Kolaja, MPH; Joyce Kong, PhD; William Lee; Kyna Long, MS; Gordon Lynch; Denise Lovec-Jenkins; Rayna Matsuno, PhD; Danielle Mitchell; Kristin Motylinski; Anna Rivera, MPH; Chiping Nieh, PhD; Chris O'Malley, MPH; Serguey Parkhomovsky; Anet Petrosyan; Jackie Pflieger, PhD; Christopher Phillips, MD; Ben Porter, PhD; Rudy Rull, PhD; Kari Sausedo, MA; Chelsea Schafer; Beverly Sheppard; Steven Speigle; Evelyn Sun, MPH; Valerie Stander, PhD; Daniel Trone, PhD; and Jennifer Walstrom from the Deployment Health Research Department, Naval Health Research Center, San Diego, California. The authors also thank the professionals from the U.S. Army Medical Research and Materiel Command, especially those from the Military Operational Medicine Research Program, Fort Detrick, Maryland; Scott L. Seggerman from the Management Information Division, Defense Manpower Data Center, Monterey, California; and the Millennium Cohort Study participants.

      References

        • Agarwal A.
        • Aponte-Mellado A.
        • Premkumar B.J.
        • Shaman A.
        • Gupta S.
        The effects of oxidative stress on female reproduction: A review.
        Reproductive Biology and Endocrinology. 2012; 10: 49
        • Araneta M.
        • Kamens D.
        • Zau A.
        • Gastanaga V.
        • Schlangen K.
        • Hiliopoulos K.
        • Gray G.
        Conception and pregnancy during the Persian Gulf War: The risk to women veterans.
        Annals of Epidemiology. 2004; 14: 109-116
        • Armed Forces Health Surveillance Center
        Health of women after wartime deployments: Correlates of risk for selected medical conditions among females after initial and repeat deployments to Afghanistan and Iraq, active component, U.S. Armed Forces.
        MSMR. 2012; 19: 2-10
        • Artini P.
        • Uccelli A.
        • Papini F.
        • Simi G.
        • Di Berardino O.
        • Ruggiero M.
        • Cela V.
        Infertility and pregnancy loss in euthyroid women with thyroid autoimmunity.
        Gynecological Endocrinology. 2013; 29: 36-41
        • Bhattacharya S.
        • Townend J.
        • Bhattacharya S.
        Recurrent miscarriage: Are three miscarriages one too many? Analysis of a Scottish population-based database of 151,021 pregnancies.
        European Journal of Obstetrics, Gynecology, and Reproductive Biology. 2010; 1: 24-27
        • Brewin C.
        Systematic review of screening instruments for adults at risk of PTSD.
        Journal of Traumatic Stress. 2005; 18: 53-62
        • Campbell S.
        • Lynch J.
        • Esterman A.
        • McDermott R.
        Pre-pregnancy predictors linked to miscarriage among Aboriginal and Torres Strait Islander women in North Queensland Australian and New Zealand.
        Journal of Public Health. 2011; 4: 343-351
        • Chiodo L.
        • Bailey B.A.
        • Sokol R.J.
        • Janisse J.
        • Delaney-Black V.
        • Hannigan J.H.
        Recognized spontaneous abortion in mid-pregnancy and patterns of pregnancy alcohol use.
        Alcohol. 2012; 3: 261-267
        • Cohen B.E.
        • Maguen S.
        • Bertenthal D.
        • Shi Y.
        • Jacoby V.
        • Seal K.H.
        Reproductive and other health outcomes in Iraq and Afghanistan women veterans using VA health care: Association with mental health diagnoses.
        Womens Health Issues. 2012; 22: e461-e471
        • Conlin A.M.
        • Bukowinski A.T.
        • Sevick C.J.
        • DeScisciolo C.
        • Crum-Cianflone N.F.
        Safety of the pandemic H1N1 influenza vaccine among pregnant U.S. military women and their newborns.
        Obstetrics and Gynecology. 2013; 121: 511-518
        • Doyle P.
        • Maconochie N.
        • Davies G.
        • Maconochie I.
        • Pelerin M.
        • Prior S.
        • Lewis S.
        Miscarriage, stillbirth and congenital malformation in the offspring of UK veterans of the first Gulf war.
        International Journal of Epidemiology. 2004; 33: 74-86
        • Doyle P.
        • Maconochie N.
        • Ryan M.
        Reproductive health of Gulf War veterans.
        Philosophical transactions of the Royal Society of London. Series B, Biological Sciences. 2006; 361: 571-584
        • Hobson C.
        • Kamen J.
        • Szoste J.
        • Nethercut C.M.
        • Tiedmann J.
        • Wojnarowicz S.
        Stressful life events: A revision and update of the Social Readjustment Rating Scale.
        International Journal of Stress Management. 1998; 1: 1-23
        • Holmes T.H.
        • Rahe R.H.
        The social readjustment rating scale.
        Journal of Psychosomatic Research. 1967; 11: 213-218
        • Iraq and Afghanistan Veterans of America
        Women warriors: Supporting she “who has borne the battle”.
        (Available:) (Accessed: March 15, 2013)
        • Jacobson I.
        • Ryan M.
        • Hooper T.
        • Smith T.
        • Amoroso P.
        • Boyko E.
        Alcohol use and alcohol-related problems before and after military combat deployment.
        JAMA. 2008; 6: 663-675
        • Joint Chiefs of Staff
        Elimination of the 1994 direct ground combat definition and assignment rule.
        (Available:) (Accessed: September 10, 2014)
        • Kang H.
        • Magee C.
        • Mahan C.
        • Lee K.
        • Murphy F.
        • Jackson L.
        • Matanoski G.
        Pregnancy outcomes among U.S. Gulf War veterans: A population-based survey of 30,000 veterans.
        Annals of Epidemiology. 2001; : 504-511
        • Katon J.
        • Cypel Y.
        • Raza M.
        • Zephyrin L.
        • Reiber G.
        • Yano E.M.
        • Schneiderman A.
        Self-reported infertility among male and female veterans serving during Operation Enduring Freedom/Operation Iraqi Freedom.
        Journal of Womens Health. 2014; 23: 175-183
        • Kroenke K.
        • Spitzer R.L.
        • Williams J.B.
        The PHQ-9: Validity of a brief depression severity measure.
        Journal of General Internal Medicine. 2001; 16: 606-613
        • Madhappan B.
        • Kempuraj D.
        • Christodoulou S.
        • Tsapikidis S.
        • Boucher W.
        • Karagiannis V.
        • Theoharides T.C.
        High levels of intrauterine corticotropin-releasing hormone, urocortin, tryptase, and interleukin-8 in spontaneous abortions.
        Endocrinology. 2003; 144: 2285-2290
        • Mattocks K.
        • Kroll-Desrosiers A.
        • Zephyrin L.
        • Katon J.
        • Weitlauf J.
        • Bastian L.
        • Brandt C.
        Infertility care among OEF/OIF/OND women veterans in the Department of Veterans Affairs.
        Medical Care. 2015; 53: S68-S75
        • Nakamura K.
        • Sheps S.
        • Arck P.C.
        Stress and reproductive failure: Past notions, present insights and future directions.
        Journal of Assisted Reproduction and Genetics. 2008; 25: 47-62
        • Ryan M.
        • Smith T.
        • Smith B.
        • Amoroso P.
        • Boyko E.
        • Gray G.
        Millennium Cohort: Enrollment begins a 21-year contribution to understanding the impact of military service.
        Journal of Clinical Epidemiology. 2007; 2: 181-191
        • Ryan M.A.
        • Pershyn-Kisor M.A.
        • Honner W.K.
        • Smith T.C.
        • Reed R.J.
        • Gray G.C.
        The Department of Defense Birth Defects Registry: Overview of a new surveillance system.
        Teratology. 2001; 64: S26-S29
        • Smith B.
        • Ryan M.
        • Wingard D.
        • Patterson T.
        • Slymen D.
        • Macera C.
        Cigarette smoking and military deployment: A prospective evaluation.
        American Journal of Preventive Medicine. 2008; 6: 539-546
        • Smith T.C.
        • The Millennium Cohort Study
        The US Department of Defense Millennium Cohort Study: Career span and beyond longitudinal follow-up.
        Journal of Occupational and Environmental Medicine. 2009; 51: 1193-1201
        • Spitzer R.
        • Kroenke K.
        • Williams J.
        Primary PHQ. Validation and utility of a self-report version of PRIME-MD - The PHQ primary care study.
        JAMA. 1999; 282: 1737-1744
        • Spitzer R.
        • Williams J.
        • Kroenke K.
        • Hornyak R.
        • McMurray J.
        Validity and utility of the PRIME-MD Patient Health Questionnaire in assessment of 3000 obstetric-gynecologic patients: The PRIME-MD Patient Health Questionnaire Obstetrics Gynecology Study.
        American Journal of Obstetrics and Gynecology. 2000; 183: 759-769
        • Spitzer R.
        • Williams J.
        • Kroenke K.
        • Linzer M.
        • deGruy 3rd, F.
        • Hahn S.
        • Johnson J.
        Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study.
        JAMA. 1994; 272: 1749-1756
        • Vogt D.
        • Vaughn R.
        • Glickman M.E.
        • Schultz M.
        • Drainoni M.
        • Elwy R.
        • Eisen S.
        Gender differences in combat-related stressors and their association with postdeployment mental health in a nationally representative sample of US OEF/OIF veterans.
        Journal of Abnormal Psychology. 2011; 120: 797-806
        • Wells T.
        • Wang L.
        • Spooner C.
        • Smith T.
        • Hiliopoulos K.
        • Kamens D.
        • Sato P.
        Self-reported reproductive outcomes among male and female 1991 Gulf War era US military veterans.
        Maternal and Child Health Journal. 2006; 10: 501-510
      1. Women in Military Service for America Memorial Foundation Inc. (2010). Available: www.womensmemorial.org/PDFs/StatsonWIM.pdf. Accessed: March 15, 2013.

      Biography

      Andrea C. Ippolito, MPH, is a Research Analyst at American Specialty Health. Her research areas focus on maternal and child health, population health and hematological malignancies. She was with the Naval Health Research Center while working on this project.
      Amber D. Seelig, MPH, a Research Health Science Specialist at the VA Puget Sound, has focused on mental and behavioral health research in military and Veteran populations. She was with the Naval Health Research Center while working on this project.
      Teresa M. Powell, MS, is a Biostatistician for the Deployment Health Research Department at the Naval Health Research Center. Her research areas focus on Gulf War illness, reproductive, mental, and behavioral health in military populations.
      Ava Marie S. Conlin, DO, MPH, is a Senior Medical Epidemiologist for the Deployment Health Research Department at the Naval Health Research Center, leading the DoD Birth and Infant Health Registry. Her interests are maternal/child health, vaccine safety, and military medicine.
      Nancy F. Crum-Cianflone MD, MPH, is an Internal medicine/infectious disease physician and epidemiologist with extensive experience conducting military health-related research. She is a U.S. military veteran who has cared for military members and studied military-related health for over a decade.
      Hector Lemus, DrPH, MS, is a professor in the Graduate School of Public Health at San Diego State University. He was a Senior Biostatistician for the Deployment Health Research Department at the Naval Health Research Center while working on this project.
      Carter J. Sevick, MS, is a Biostatistician/statistical programmer for the Colorado Health Outcomes Program. He was a biostatistician with the DoD Birth and Infant Health Registry at the Naval Health Research Center while working on this project.
      Cynthia A. LeardMann, MPH, is a Senior Epidemiologist for the Deployment Health Research Department at the Naval Health Research Center. Her research areas focus on mental health and women's health in military populations.