Barriers to Mental Health Treatment for Military Wives

Qualitative interview data from a convenience sample of 17 women currently or previously married to (or romantically involved with) male service members revealed unique concerns that informed questionnaire development. [This stage of the study is described in more detail in an online data supplement to this report.] Quantitative work involved 569 women (average age 29, range 18 to 56). Most were white non-Hispanic (N=379, 85% of the 444 respondents who provided data on race-ethnicity) and were recruited primarily through social networking tools, including Facebook, Twitter, online forums, chat rooms, and Web sites oriented toward military spouses. In addition, 56 Army Family Assistance Centers and Army Community Services facilities programs around the country were asked to distribute flyers describing the study. The flyers were also distributed at Oregon armories.

Social media were by far the most productive recruitment strategy. The project’s Web site quickly acquired over 4,000 Facebook “friends” located in more than 45 states and at least eight foreign countries (in Asia, Europe, and Latin America). Recruitment was conducted from May 2010 through July 2011.

Participants were asked to interact with a Web site entitled Military Wives Matter that included screening questions and health status measures. Depression in the past week was measured with the 20-item Center for Epidemiological Studies–Depression (CES-D) scale (Cronbach’s α=.93 in this sample). Possible CES-D scores range from 0 to 60; scores of 16 to 26 are considered indicative of mild depression, and scores ≥27 suggest major depression.

Nonspecific psychological distress in the previous month was measured with the six-item K6 instrument (Cronbach’s α=.89 in this sample). K6 scores range from 0 to 24. Scores >12 correspond to clinician-rated measures of serious mental illness.

Barriers to mental health services were measured by items from the National Survey on Drug Use and Health (NSDUH) (10). All participants were asked, “During the past 12 months, was there any time when you needed mental health treatment or counseling for yourself but didn’t get it?” Those answering in the affirmative were presented with 13 barriers (for example, “couldn’t afford the cost”) and one open-ended item (“Some other reason, please specify”). Items included military-specific barriers that were suggested by participants in the qualitative interviews (for example, “couldn’t find a provider who understands the needs of military spouses” and “couldn’t find a provider you could trust”).

Comparative data on the overall population of people married to military personnel were obtained from U.S. Department of Defense demographic reports (9). Numbers of deployments for service members were estimated from an Institute of Medicine report (2).

General (civilian) population data were obtained from the 2011 NSDUH (N=70,109; response rate, 74%). Comparison respondents were married women ages 18 to 64 who satisfied diagnostic criteria for major depressive disorder at some time during the year before the interview (N=567 respondents).

Data analysis employed SPSS, version 20, and accounted for NSDUH complex sampling. The project was approved and overseen by the Institutional Review Board at Oregon Health and Sciences University. Informed consent was obtained from all participants.

Similarities and differences were noted when data for study participants were compared with national data describing military spouses (9). For example, all study participants were women, whereas females represented 91% of military spouses nationwide (9). Participants (average age 29±8, range 18 to 56) were younger than military spouses overall (average age 33), and whites were overrepresented among participants, compared with the service member population. Racial-ethnic distribution for the study participants was as follows: Caucasian, N=379, 85%; African American, N=11, 2%; Asian/Pacific Islander, N=12, 3%; American Indian, N=8, 2%; Hispanic/Latino, N=29, 7%; and other, N=5, 1%. Nearly all participants (N=402, 91%) participants were married, whereas by definition (9), all military spouses were married. Slightly less than two-thirds of participants (N=269, 61%) were parents, compared with slightly more than two-thirds (69%) of service personnel. Among participants, 37% (N=161) were employed, which is similar to the proportion of military spouses overall (42%). The vast majority of participants (N=400, 94%) lived in the United States, as did nearly all (91%) service members. Most participants rated their finances as comfortable or adequate, but 26% (N=115) reported poor financial circumstances (data were not available for military spouses overall). Slightly over half the participants (N=254, 58%) described their location as urban (data were not available for military spouses overall).

Slightly over half the participants (N=249, 58%) were married to Army personnel, compared with 50% of military spouses overall. The vast majority of participants’ husbands (N=366, 87%) were on active duty, compared with about two-thirds (65%) for military spouses overall. Most participants’ husbands (N=349, 83%) were enlisted (versus officers), and the proportion was similar for military spouses overall (79%). Nearly half of participants (N=181, 46%) said that their husbands were deployed, whereas deployment was 6% for military spouses overall. Participants’ husbands had a mean±SD of 2.4±2.2 deployments to Iraq or Afghanistan, compared with 1.6±1.1 for military personnel overall (2). This difference was highly statistically significant (p<.001 by Z test). Half the participants (N=212, 50%) lived more than 50 miles from their husbands’ duty stations (data were not available for military spouses overall). It should be noted that percentages presented for participants were calculated on the basis of the number who answered each question, rather than the total sample. [A table summarizing data on sample characteristics is available in the online data supplement.]

Substantial numbers of participants endorsed moderate or severe mental health problems. On the CES-D scale most (N=200, 51%) had scores higher than 26, which suggested major depression; another 27% (N=104) had scores in the mild depression range (scores of 16–26). The CES-D mean score was 27±13. NSDUH data indicated a prevalence rate of past-year major depressive disorder of 8% among married women (complex SE=<1%).

On the K6 scale, 37% (N=145) scored higher than 12, suggesting serious psychological distress. The comparable percentage for NSDUH respondents was 29% (complex SE=3%). Thus participants were more likely than participants in the national survey to have had serious psychological distress within the past month (p<.001 by Z test).

Table 1 lists barriers to mental health services experienced by participants and national survey respondents, respectively. Substantial percentages of both groups said that they had needed but did not receive mental health treatment during the previous 12 months. A larger proportion of the study participants described unmet treatment need (N=162, 44%), compared with national survey respondents (30%) (p<.001 by Z test). Logistic regression models that included variables common to both data sets showed that increases in age and K6 score were the most powerful predictors of unmet treatment need. In both populations, each unit increase in the K6 score elevated the chances of reporting unmet treatment need by 20% (p<.001).

Substantial differences in the presence of barriers were noted between the military wives in the study and the national survey respondents. As shown in Table 1, military wives reported being unable to get away during the day as the most common barrier (endorsed by 38% of participants), whereas this concern was less pertinent for the general population (24% of respondents). Conversely, over half (53%) of the general population described cost barriers, whereas only 19% of military wives did so (p<.01 by Z test for both comparisons).

About a quarter of military wives (26%) endorsed the item about confidentiality, compared with only 4% in the national survey (p<.001 by Z test). In logistic regression models, the K6 score predicted confidentiality concerns for military wives (p<.001). No statistically significant predictors of confidentiality concerns were noted for the national survey respondents.

The military wives also had concerns about negative opinion in the community (19%) and about being committed or forced to take medication (17%); these concerns were not shared by the general population of women (5% and 8%, respectively, p<.01 by Z test for both comparisons). In addition, large percentages of military wives expressed concern that mental health providers would not understand military spouses (35%) or said that they could not find a counselor they could trust (29%). A quarter of military wives said they did not know where to go for mental health services, compared with only 13% of the survey respondents (p=.002 by Z test).

Among the study participants, endorsement of the item about needing but not getting treatment did not vary by the participant’s rural versus urban location, branch of service, or active versus reserve status.

Key findings were twofold. First, military wives with mental health concerns faced numerous barriers to treatment. Second, reported barriers differed substantially from those experienced by women in the general population.

The project had representation of military wives from throughout the United States as well as overseas. The study compared participants with the overall military spouse population and with women in the general population. The methods were developed in collaboration with military wives, and the project achieved its goal of recruiting a sample with substantial psychological distress. On the other hand, the study relied on the Internet for access to potential participants. Military wives lacking Internet access were not eligible. Another limitation is that the study recruited few military wives from racial-ethnic minority groups. Participants’ husbands were intentionally not included. The recruitment methods and survey mode were designed to increase participation of women with depressive symptoms. Thus the estimated symptom levels of participants presumably exceed those of military wives overall.

Some participants began but declined to complete the survey. It was not possible to compare these women with those who completed the survey. Also, the questionnaire was intentionally brief. Thus some topics could not be explored in detail, such as the use of specific mental health services (for example, inpatient care). Moreover, military spouses have stressors and needs in several nonclinical areas, such as financial management (1). It is likely that the questionnaire design may have led to overrepresentation of military wives concerned about clinical needs. Military wives with needs in nonclinical areas may have been underrepresented.

These limitations notwithstanding, several conclusions follow from this work. In addition to typical barriers to accessing care, such as logistical challenges (for example, lack of child care), participants encountered military-specific barriers, such as not being able to find a provider who understands military culture. Although several military programs focus on spouse mental health (14), many military wives seek services from civilian providers (12–14). Participants felt that few civilian providers could understand the concerns of military families.

Thus it is important to educate civilian clinicians about these military-specific concerns, including the possible impact of entering care, trust issues, and damage to service members’ careers. For example, the federal government has supported pilot projects pairing programs for active duty service members or veterans with civilian agencies (13). Military programs such as RESPECT-Mil (15) include training that could be adapted for civilian providers.

One approach might be to expand the educational offerings aimed at civilian providers by the military’s TRICARE program and its contractors. [A table in the online data supplement lists pertinent resources.] Such strategies may facilitate access to mental health services for military wives. Internet-based services may be particularly helpful for military wives. Strategies employing social media (such as Facebook) might also be used to inform military wives about availability of mental health services tailored to their needs and could facilitate provision of respectful services to a worldwide population.

This work was supported by grants R34-MH080765 and R34-MH083494 from the National Institute of Mental Health and by the Medical Research Foundation of Oregon. The authors gratefully acknowledge the work of Staff Sergeant Eddie Black, B.S.

The authors report no competing interests.