INTRODUCTION

Many lesbian and bisexual (LB) veterans have been targets of victimization in the military based on their gender and presumed sexual orientation. Under Don’t Ask, Don’t Tell (DADT), thousands of LB women were discharged from military service, while countless others continued to serve in silence, and, as veterans, sought care from the Department of Veterans Affairs (VA).1,2 Prior research has shown that many LB veterans experience discrimination, rejection and/or poor care following disclosure of their sexuality to healthcare providers,3 and may engage in strategies to avoid conversations regarding sexual identity. These experiences may be particularly harmful for LB veterans returning from military deployments with substantial physical and mental health problems,47 and possibly compounded by lingering effects of targeted sexual assault and harassment experienced during military service based on perceived sexual orientation.8 Recent research9 indicates that 15.1 % of female OEF/OIF veterans report experiencing sexual trauma during military service. Given these healthcare needs among LB veterans, and the potential for underuse/care avoidance, understanding the healthcare needs of this population is crucial if the VA is to provide comprehensive care to all women veterans, regardless of sexual orientation.

METHODS

Study Design

The Women Veterans Cohort Study (WVCS) is an ongoing prospective cohort study involving male and female OEF/OIF veterans receiving care at two VA facilities in the U.S, one in the northeast and one in the midwest8.

Sample

Letters describing the study were sent to 3,251 female OEF/OIF patients enrolled at each facility. Veterans expressing interest in the study contacted the research coordinator, read a study description, were consented and then, if enrolled, were screened for eligibility. Between July 2008 and October 2011, baseline surveys were completed by 11 % of female veterans who were invited to participate (n = 365). For this study, data were obtained from two linked sources: participant surveys and VA electronic medical records.

Participant Surveys

Our analyses focused on questions that explored sexual orientation, physical and mental health status, combat and sexual trauma exposure, and satisfaction with VA care, using the measures below.

Sexual Orientation

Participants were asked to identify the sexual orientation category that best described them: heterosexual, gay or lesbian, bisexual, celibate or asexual, or not sure. The gay or lesbian and bisexual categories were combined for these analyses.

Post-Deployment Health Status

Post-deployment health was measured by asking participants to rate both their current physical and mental health as: much better than before deployment, slightly better than before deployment, about the same, slightly worse than before deployment, or much worse than before deployment.

Access to Care/Utilization

We asked participants whether they had private or public insurance and what type of private (e.g., employer-sponsored) or public (e.g., Medicare, Medicaid, Tricare) insurance they had. We also asked whether they had a regular provider, and whether that provider was a VA provider. Participants with a regular VA provider were asked if that provider was located in a Primary Care or Women’s Health clinic.

Combat Trauma

Combat trauma was measured using the Combat Exposure Scale (CES), a seven-item self-report measure that has been shown to have a high degree of validity and reliability.10

Military Sexual Trauma

Sexual trauma during military service was assessed with the following two questions: “While you were in the military, did you receive uninvited and unwanted sexual attention, such as touching, cornering, pressure for sexual favors, or sexual remarks?”, and “While you were in the military, did someone ever use force, or threat of force, to have sexual contact with you against your will?”. Response categories included “yes” and “no”.

Childhood Sexual Trauma

Childhood sexual trauma was assessed with specific questions about the presence and frequency of sexual abuse at different times in childhood, including childhood and adolescence prior to the age of 18.11 Response categories included: never, 1–2 times, 3–5 times, more than 5 times.

Smoking

Smoking status was ascertained by a question asking respondents about frequency of smoking cigarettes and data was recoded for respondents who smoked “everyday” and “some days” as current smokers, and those who smoked “not at all” as nonsmokers.

Hazardous Drinking

Hazardous drinking (drinking associated with possible harm) was defined as a score of eight or more on the Alcohol Use Disorders Identification Test (AUDIT).12

VA Administrative Data Measures

We used VA administrative records to assess 17 common women’s health conditions (Appendix Table 5) for which both LB and heterosexual veterans might seek care. We used the Agency for Healthcare Research and Quality’s (AHRQ) Clinical Classifications Software (CCS) framework to map ICD-9 codes to conditions; specific conditions were grouped into broad categories.13 A patient was considered to have one of the designated medical conditions if she had at least one ICD-9 code for that condition category assigned by a VA provider during the study period (2008–2011). We used the same methodology to assess mental health conditions (depression, bipolar disorder, post traumatic stress disorder [PTSD], and anxiety disorder). We derived a count of primary and mental health care visits during the study period from clinic stop codes in VA administrative files.

Analysis

We used the χ 2 test to compare the demographic, health care utilization, and clinical characteristics of LB and heterosexual veterans. Statistical analyses were performed using SAS version 9.1.3 (SAS, Inc., Cary, North Carolina).

RESULTS

Demographic characteristics of the study sample are presented in Table 1. Of the 365 OEF/OIF women veterans enrolled in the study, 35 women (9.6 %) identified as either gay or lesbian (4.7 %) or bisexual (4.9 %). Thirty women identified as asexual or celibate, and were excluded from the analysis. LB and heterosexual veterans did not differ significantly on demographic characteristics, including age, branch of service, race/ethnicity, or service component, though LB veterans were less likely to be married than heterosexual veterans. Most women veterans had private insurance, but LB women were significantly less likely to have government-sponsored insurance (e.g., Medicaid) (17 % vs. 32 %, p = .03).

Table 1 Demographic Characteristics of OEF/OIF Women Veterans (n = 335)

Overall, LB veterans were more likely to have been the victims of some form of childhood sexual abuse than heterosexual veterans (60 % vs. 36 %, p < .001). LB veterans were significantly more likely to have experienced sexual abuse by an adult prior to their 13th birthday (46 % vs. 26 %, p = .02) and to have experienced sexual contact without consent between their 13th–18th birthdays (34 % vs. 17 %, p = .02) (Table 2).

Table 2 Combat and Military/Childhood Sexual Trauma (n = 335)

Fifty percent of the LB veterans and 35 % of the heterosexual veterans had a diagnosed mental health condition of PTSD, anxiety disorder, depression, or bipolar disorder (p = 0.10) (Table 3). Since return from deployment, LB veterans were more likely than heterosexual veterans to rate their current mental health as worse than before deployment (35 % vs. 16 %, p < .001), but there were no differences in post-deployment physical health ratings. LB veterans were more likely to be current smokers (43 % vs. 23 %, p = .008) and hazardous drinkers (32 % vs. 16 %, p = .03) than heterosexual veterans. There were no statistically significant differences in diagnosed women’s health conditions between the two groups.

Table 3 Health Conditions Among LB and Heterosexual Veterans (n = 335)

Differences in experiences with and perceptions of VA healthcare are in Table 4. LB veterans were more likely to use VA providers for their healthcare than heterosexual veterans (31 % vs. 14 %, p = .01), and were more likely to plan to use the VA in the future (100 % vs. 88 %, p = .03). There were no statistically significant differences between the two groups in perceptions of VA quality, availability of services, or ability to treat women veterans, although LB rated the latter two criteria lower than heterosexual veterans.

Table 4 VA Healthcare Utilization, Satisfaction, and Perceptions of Quality

DISCUSSION

This is one of the first studies to examine health conditions and healthcare utilization among LB women veterans in VA care. In our study, LB veterans had higher rates of mental health problems, smoking, and poorer self-rated mental health. As echoed in a recent study,14 a striking finding was that LB veterans had experienced significantly higher rates of military sexual trauma than heterosexual veterans, and had higher rates of hazardous drinking, both consistent with other studies showing a high correlation between childhood sexual abuse and adult substance abuse disorder among lesbian women.15,16 Prior studies have noted high rates of antigay harassment in the military, ranging from verbal abuse to physical abuse to death threats, as well as sexual victimization, particularly among lesbian-identified service members.17 Universal screening for military service-related sexual trauma has been implemented within VA and has increased rates of mental health treatment.18 Health care providers working with female veterans should also be aware of high rates of combat exposure and childhood abuse and refer women to appropriate VA treatment and support groups for sequelae of these experiences.

This study has several limitations. Though only 35 veterans identified as gay, lesbian or bisexual, this self-report represents 10 % of our OEF/OIF survey cohort, which is slightly higher than population estimates of lesbians in the military2. Furthermore, because women veterans could enroll in the cohort study between 2008 and 2011, and Don’t Ask Don’t Tell hadn’t yet been repealed, there is a possibility that the number of women identifying as LB in the study is an underestimate of the true population of women who self-identify as LB. In addition, we do not have data on the rates of mental health and substance use disorders of the veterans studied at the time they entered service. Elevated rates of mental health and substance use disorders may have placed them at increased risk for military sexual trauma. Other limitations of this study include that the original study focused on broad issues affecting all OEF/OIF women veterans and did not include a comprehensive assessment specific to LB veterans as an underserved population, nor any type of examination regarding the degree to which LB veterans may have felt marginalized or discriminated against in healthcare. Finally, we chose to combine the gay/lesbian category with the bisexual category for these analyses, which could have led to an overestimation or underestimation of associations reported in the analyses.

Very little research on the health and health care needs of LB veterans has been published to date. Research in non-veteran populations has demonstrated that LB persons often fear negative consequences of disclosing their sexual orientation to health care providers. These disclosure-related fears may have been amplified among veterans due to DADT; in our anecdotal experience, many veterans mistakenly believed DADT was a policy that VA shared with DOD. With the repeal of DADT, LB veterans may begin to feel more comfortable disclosing their sexuality to their VA healthcare providers without fear of reprisal. In turn, it is essential that VA healthcare providers create a healthcare environment free of assumed heterosexuality (e.g. not assuming that all female veterans require birth control for sexual activity) and ensure that they are knowledgeable about LB health issues.

Future research should focus on expanding this study to include a larger and more diverse sample of lesbian, gay, bisexual, and transgender veterans receiving care at VA facilities across the country. Questions regarding sexual orientation and behavior should also be included in all VA surveys to ensure that the needs of this population are being met across all areas of VA care.