Correlates of Frequent Mental Distress among Active and Former Military Personnel

By Eleanor DiBiasio , Melissa A Clark and Samantha R Rosenthal In   Issue Volume 22 No. 2 Doi No https://doi-ds.org/doilink/11.2021-61868836/JMVH Vol 22 No 2

INTRODUCTION

There are currently over 20 million veterans living in the United States (1) and nearly ten percent are receiving care for mental illness (2).  In 2007, the Department of Veteran’s Affairs (VA) spent over $3.5 billion dollars in mental health and substance abuse treatment for veterans (3).  That number is increasing as more veterans utilize VA mental health services (2,4). Due to the wars in Iraq and Afghanistan, nearly 40% of veterans return home with mental health problems, the most common of which are post-traumatic stress disorder (PTSD) and depression (5).  These returning veterans add to the millions who served in the World Wars, the Korean War, Vietnam, and the Gulf War, each of which had a substantial mental health impact on its respective generation of soldiers (6).

Despite the high prevalence of stress and mental illness, many veterans still battle their mental illness privately due to the stigma of these conditions in the military and at home.  Of veterans who meet the screening criteria for a mental health problem, only 38-45% express interest in receiving help and only 23-40% actually obtain professional help (7).  Understanding the correlates of mental illness among veterans can help the Veterans Administration (VA) and other veteran support organizations target mental health treatment to individuals who need it, whether they report their needs or not.

This study examines the association between potential risk factors and mental distress among current and formerly active duty military personnel using a nationally representative sample of veterans of all ages, races, incomes and backgrounds.  We define “active duty” according to the U.S. Census Bureau definition:

“Active duty military service includes full-time service, other than training, as a member of the U.S. Army, Navy, Air Force, Marine Corps, Coast Guard or as a commissioned officer of the Public Health Service or the National Oceanic and Atmospheric Administration… Active duty also applies to a person who is a cadet attending one of the five United States Military Service Academies. Active duty applies to service in the military Reserves or National Guard only if the person has been called up for active duty, mobilized, or deployed.” (8)

The nationally representative sample provides a novel opportunity to determine the characteristics of veterans more likely to suffer from mental distress and to make recommendations to prevent and treat the condition.

 

METHODS

Sample

The data were from the 2010 Behavioral Risk Factor Surveillance System (BRFSS), a national telephone survey administered annually by the Centers for Disease Control and Prevention (CDC).  Further information about the methods of data collection and survey administration used in the BRFSS can be found elsewhere (9).

Eligibility for the study included those who had served or were currently serving in the active duty U.S. military, or in an activated unit of the Reserves or National Guard. All BRFSS respondents were asked, “Have you ever served on active duty in the United States Armed Forces, either in the regular military or in a National Guard or military reserve unit? Active duty does not include training for the Reserves or National Guard, but DOES include activation, for example, for the Persian Gulf War.” Possible responses were “yes, now on active duty;” “yes, on active duty during the past twelve months, but not now;” “yes, on active duty in the past, but not during the last 12 months;” “no, training for Reserves or National Guard only;” and “no, never served in the military.” Those who reported being currently on active duty, active duty in the past year, or active duty over one year ago were included in the analytic sample.  Those with invalid information (i.e., those who answered ‘don’t know/not sure,’ who refused to answer, or who had missing data) were excluded from the analytic sample, as were those who had never served in active duty forces.  Also excluded was 1.9% of the eligible population with invalid information for the outcome of interest, days of poor mental health in the past month.  The final analytic sample included 59,310 respondents. Henceforth the current and former active duty soldiers in the analytic sample will be referred to as “veterans.”

 

Measures

Outcome. We used a validated Centers for Disease Control and Prevention (CDC) measure of frequent mental distress (FMD), defined as having fourteen or more days of poor mental health in the past thirty days (10). This two-week cut-off matches diagnostic tools used by clinicians to diagnose mental health problems such as anxiety and depression (11,12).

Potential risk factors for FMD included demographic characteristics (age, gender, race/ethnicity, marital status, education level, income, number of dependent children, and employment), as well as risk factors identified in previous literature about veterans (time since service, physical health status, days without adequate sleep in past month, and emotional support (13-16)). Age was categorized into quartiles. Race/ethnicity was organised so that the ‘other’ group included American Indian, Alaskan Native, Pacific Islander, Asian and other small minority groups.  Marital status was categorized into “never married,” “widowed,” “separated or divorced” and “married or the member of an unmarried couple.” Education was categorized as “did not finish high school,” “finished high school,” “some college,” and “finished college.” The income categories were chosen because they create four roughly equal income intervals.  Because 11.0% of respondents in our analytic sample had missing information on income, we added a “missing” category to the analysis. We dichotomized respondents as employed or not employed. Homemakers, students and retirees were included as ‘employed’ because they are often voluntarily unemployed and involuntary unemployment has been shown to correlate more highly with mental illness than voluntary unemployment (17, 18).  Finally, we categorized respondents as having either zero, one, or two or more dependent children under the age of 18 years.

Time since service, poor physical health, inadequate sleep, and lack of emotional support are all possible causes of emotional stress and psychological problems (13-16) and therefore were included as potential risk factors in the analysis.  We created a three-category variable for time since service: currently active duty, active duty in the past twelve months, and active duty over twelve months ago.

Poor physical health was defined using the question, “for how many days during the past 30 days was your physical health not good?” Responses ranged from zero to thirty days. We then created two categories based on the validated CDC measure of frequent physical distress (19): 1) no frequent physical distress (physical health not good 0-14 days) and 2) frequent physical distress (physical health not good 15 or more days).

To quantify the amount of inadequate sleep received, we used the question, “During the past 30 days, for about how many days have you felt you did not get enough rest or sleep?” This variable was treated as continuous to compute point estimates of the average number of days without adequate sleep among those with and without frequent mental distress.

Finally, we estimated the availability of emotional support using the question, “How often do you get the social and emotional support you need?” with responses of Always, Usually, Sometimes, Rarely, or Never. We collapsed the possible responses into two categories: “Always or usually has adequate emotional support” and “Sometimes, rarely or never has adequate emotional support.”

 

Analysis

Data were weighted to adjust for non-response and non-coverage in order to obtain U.S. population estimates and to account for the complex survey design of the 2010 BRFSS. For all potential risk factors, missing and invalid observations were excluded.  For all variables except income, the number of observations with invalid information was less than 10% of the total observations in the sample. Frequencies and weighted percentages were calculated for all variables (Table 1). Bivariate analyses were computed to compare the prevalence of frequent mental distress across demographic characteristics and potential risk factors. To determine if combat exposure was specifically related to frequent mental distress, we conducted a sub-analysis (N=2,556) on the relationship between combat exposure and frequent mental distress using bivariate analysis.  The smaller sample size was a result of only having data on combat exposure from respondents in two states (Nebraska and Tennessee).

Multiple logistic regression was employed to characterize the relationship between potential risk factors and frequent mental distress.  We obtained crude and adjusted odds ratios of frequent mental distress for all potential risk factors (Table 2).  STATA version 12 was used for all analyses (20).

 

RESULTS

Overall, only 25% of the study population was under age 45, and the largest age group was those over seventy years (29%).  The study sample was predominately male (92%), non-Hispanic white (78%), and married (75%). The vast majority reported completing high school or a higher level of education (96%) and almost one-third had household incomes of more than $75,000 (32%). Ninety-three percent reported being employed and 91% served more than 12 months ago, with only 5.0% on active duty at the time of survey administration. Most had no dependent children (74%), and 86% percent reported good physical health.  Respondents reported an average of 7.1 days of inadequate sleep (95% CI 7.0-7.3) in the past month, and most respondents had at least adequate emotional support in their lives (81%).  It is worthwhile to note that our sample of current and former active duty personnel in this study is significantly older than the general U.S. civilian population.  Moreover, our sample consisted of a higher proportion of males, had on average a lower reported income, and had more physical distress than non-active duty personnel who answered the BRFSS 2010, indicating that those with active duty experience in the U.S. have different distributions of demographic and health variables than the general population.

Almost 10% of the study sample reported experiencing frequent mental distress (FMD).  As shown in Table 1, those with FMD were more likely to be African American, Hispanic, and non-Hispanic multiracial respondents.  Results of the multivariable analysis of risk factors for frequent mental distress (Table 2) showed that younger adults and females had increased odds of FMD.  Those who were widowed, less educated, and had lower incomes, had higher odds of FMD. Those who were unemployed had three times the odds of FMD compared to those who were employed, and individuals in poor physical health had much greater odds of FMD than individuals with better physical health. Those with less than adequate sleep also experienced slightly higher odds of FMD than those reporting adequate sleep.  Individuals who reported always or usually having adequate emotional support had decreased odds of FMD compared to those who reported sometimes, rarely or never having adequate emotional support.

Because not every active duty individual in the U.S. military participates in combat, we conducted a sub-analysis to determine if participating in active duty combat was related to frequent mental distress.  We found no significant results in this regard (p=0.36).

 

DISCUSSION

This study characterizes frequent mental distress among current and former active duty U.S. military personnel. Using the CDC measure of frequent mental distress (FMD), we found that almost 10% of current and former active duty soldiers have FMD.  This is higher than the prevalence of FMD estimated for the general population (7.1%) (21). This is one of the largest studies to date of a random sample of veterans, and as a result our findings are highly dependable.

Current and former active duty soldiers with FMD are primarily of low income, limited education, and high unemployment.  This is not a surprising finding, as it is well established that low socioeconomic status is associated with high psychiatric morbidity (22). Improving education is one way to improve socioeconomic status because individuals who earn a post-secondary degree experience greater financial rewards than those who do not attend post-secondary school (23).   Studies of the post-World-War-II GI bills clearly established that lowering tuition costs for veterans substantially increased enrollment in post-secondary education (24, 25). Therefore, programs that increase veterans’ education have the potential to reduce poverty, especially since young adults from economically disadvantaged families are more likely than others to enrol in the military (26). We recommend an increase in programs to assist veterans in obtaining higher education, especially scholarship programs that reduce the cost of attending college for those who cannot afford it.  This will have a positive impact on education, income and employment, which were three important risk factors for FMD in this study.  For veterans who lack the qualifications or desire for post-secondary education, employment and economic security can still be promoted with workforce re-entry and job training programs.

Younger veterans were also more likely to have FMD, even after controlling for time since service.  Seal and colleagues found similar results in their analyses of Iraq and Afghanistan veterans (5, 27), and suggest that due to their lower military ranks, younger veterans may have greater combat exposure leading to poorer mental health outcomes.  Because there is evidence that stress, such as that of combat or more generally being of lower rank, can have lifelong effects on mental health (28), screening and treating our youngest veterans for FMD early may help prevent chronic, lifelong mental health problems.

In our sample, females in active duty service were particularly likely to suffer from FMD.  This finding is consistent with other military research (29-31). Females may be particularly at risk due to the high prevalence of sexual assault against women in the U.S. military (32), as well as marginalization by other service members (33). Moreover, females have about twice the risk for depression and anxiety disorders compared to males, regardless of military service history (34).  Male underreporting of mental health symptoms due to a culture of stigma may also at least partially explain the gender differences we observed (35).  As with young veterans, timely screening and treatment can prevent long-term psychological problems for females in the military.  In addition, efforts to increase reporting of mental health symptoms among males may be helpful in reaching men who otherwise feel stigmatized.  One way to increase reporting among men is to train primary care physicians to screen for symptoms that present more often in men than women–including irritability and risk-taking–when screening for depression (36).

Systems to ensure emotional support for veterans returning home may also be a key method of mental illness reduction in this population.  We found that having adequate emotional support was associated with lower odds of FMD.  This confirms the already well-established connection between social support and mental health (37).  There are many evidence-based social support interventions to improve mental health (38-40), including support groups, one-to-one interventions, and interventions to enhance natural networks.  We recommend establishing strong support programs for service members, especially for those who are widowed.

Compared to those without FMD, rates of frequent physical distress and the number of days without sleep were higher among those with FMD.  This finding is consistent with literature reporting a connection between physical health and mental wellbeing (41-45).  Physical therapy and pain management programs have been shown to improve depressive symptoms for individuals in chronic pain (46).  Physical distress was the risk factor most strongly associated with FMD in our results, therefore it is especially important that veterans have access to services addressing this issue.  In addition to reducing FMD by alleviating disability and pain, these programs present a prime opportunity for mental health screenings and treatment referrals that may help veterans address other factors leading to their mental distress.  Since our results show that those with frequent physical distress had over four times the odds of reporting frequent mental distress, screening for mental distress in physical therapy programs may be an effective way to identify veterans who may be experiencing FMD and refer them to appropriate services.  Our finding that inadequate sleep and FMD are correlated is not particularly surprising, since FMD serves as a proxy for several mental disorders with symptoms that include sleep disturbance, depression and PTSD (10, 47).  It is therefore difficult to determine if FMD is a cause of sleep disturbance or a result.  Although our data did not allow us to examine cause-and-effect relationships, future studies should question whether, in some veterans, sleeping difficulties are resulting in preventable FMD.

In summary, education and training programs, job fairs, and physical therapy offices would be excellent sites to screen for mental distress in veteran populations, provide information on mental health services, and refer individuals who may be suffering to appropriate services.  Our results show that individuals who are likely to take advantage of employment training and physical therapy are also the individuals at highest risk for FMD.  The staff that is performing the screenings should be trained to pay special attention to younger veterans and females, as they have higher odds of FMD.  Another group with higher likelihood of mental distress is those who lack emotional support.  These individuals may be harder to reach because they are less likely to be involved in programs and services for veterans, but are still likely to utilize basic services such as primary care or physical therapy.  Future research should determine how best to reach veterans who may be socially isolated.

Our study has several limitations.  First, our analyses are based on self-report data.  Studies have shown that military personnel are less likely to report mental illness than the non-military population due to a culture of stigma (48). The anonymous nature of the BRFSS may have somewhat mitigated the effect of stigma, but the true prevalence of mental distress in this population may be underestimated by our study. The self-reported nature of the BRFSS also limited our ability to study the physical health status of participants. There are clear limitations to the variable used to measure frequent physical distress because “good physical health” is a subjective term. However, we dichotomized this variable in a way that has been validated by CDC.  Next, our data were cross-sectional.  We cannot assume that any covariates in the final model have a causal relationship with frequent mental distress in the current and former active duty population.  For example, poor physical health is likely to be highly correlated with poor mental health, so the relationships observed could be due to reverse causation or residual confounding.  In addition, the BRFSS is inherently limited by low response rates, which is an indicator of potential bias in the data collected.  For the 2010 BRFSS, the median response rate was between 35.8% (if unknown records are assumed eligible) and 54.60% (49). However, these rates are slightly higher than other years of the BRFSS (50-52), and regardless of its low response rate the BRFSS is considered one of few available large, nationally representative surveys (53).  Our analysis was also limited by the variables available for analysis in the BRFSS.  There may be other risk factors for FMD among active and former service members that we did not include, such as housing insecurity or living with an income below the federal poverty level.  Moreover, time since service and age are highly related (p<0.001 in chi squared), so there is likely some collinearity in our model. Therefore it is difficult to determine if younger age is a risk factor for FMD or if age acts as a proxy for time since active duty service.  Further studies should determine if the association we observed between age and FMD is due primarily to vulnerability of youth or recency of combat.  Next, the sampling methods of the BRFSS are limited to those with telephones.  Therefore, those who are homeless, institutionalized, or who have fragile living situations are more likely to have been excluded from our sample. Finally, our sub-analysis results found that exposure to combat was not significantly related to FMD, however we were not able to analyze this relationship in our entire sample. Our sub-analysis sample size (n=2,556) was limited by the fact that only two states, Nebraska and Tennessee reported information about combat exposure in the 2010 BRFSS.   This finding is supported by recent literature, which finds that suicide rates among U.S. military personnel are independent of combat exposure (54).  However, there is potential of residual confounding by combat exposure in the final adjusted model of the full sample.

 

CONCLUSION

Active duty military experience increases the risk that individuals will experience psychiatric distress (55, 56).  Thus, we have a responsibility to give the best standard of care for these individuals. Our findings show that poor physical health, unemployment, low income/education, and a lack of emotional support are all associated with frequent mental distress in this population.  Based on these results, we recommend that mental health resources in the Department of Veterans Affairs (VA) system be primarily targeted toward early and accurate mental health screenings, physical therapy and pain management, emotional support structures, and programs to address unemployment and poverty.  It may be particularly important that younger veterans and female veterans receive these services.  Future studies should examine these relationships to determine causality.

 

Table 1: Characteristics* of 2010 U.S. population of individuals who were formerly or currently active duty military

 

 

 

     Characteristic

Frequent Mental Distress

n=5,189 (8.8%)*

No Frequent Mental Distress

n=54,121 (91.3%)*

Total

N=59,310*

p-values

Age (in quartiles)       18-45 years       46-58 years

       59-69 years

       70-99 years

788 (34)

1,368 (26)

1,737 (25)

1,263 (15)

5,463 (24)

8,244 (19)

17,318 (26)

22,725 (31)

6,251 (25)

9,612 (19)

19,055 (26)

23,988 (29)

p<0.0001

Gender         MaleFemale

4,490 (86)

699 (14)

50,035 (92)

4,086 (7.6)

54,525 (92)

4,785 (8.1)

p<0.0001

Race/ethnicityNon-Hispanic whiteNon-Hispanic black

Hispanic

Non-Hispanic multiracial

Other

4,011 (73)

428 (11)

239 (8.6)

176 (3.1)

226 (4.8)

45,516 (79)

3,345 (9.9)

1,787 (6.3)

999 (1.8)

1,550 (3.2)

49,527 (78)

3,773 (10)

2,026 (6.5)

1,175 (1.9)

1,776 (3.3)

p=0.0001

Education levelDid not finish high schoolFinished high school

Some college

Finished college

483 (8.0)

1,641 (29)

1,822 (39)

1,228 (24)

2,976 (4.1)

15,632 (28)

15,560 (31)

19,804 (37)

3,459 (4.5)

17,273 (28)

17,382 (32)

21,032 (36)

p<0.0001

Income level        <$25,000$25,000-$50,000

$50,000-$75,000

³$75,000

Don’t know or refused

2,107 (36)

1,372 (25)

577 (13)

597 (15)

536 (9.9)

11,076 (17)

15,535 (26)

8,640 (17)

12,864 (30)

6,006 (10)

13,183 (18)

16,907 (26)

9,217 (16)

13,461 (29)

6,542 (10)

p<0.0001

Employment       Employed (or homemaker,student, retiree)

Unemployed

3,343 (82)

483 (18)

49,715 (94)

2,182 (5.6)

53,058 (93)

2,665 (6.5)

p<0.0001

Marital statusMarried or member ofunmarried couple

Divorced or separated

Widowed

Never married

2,645 (61)

1,421 (11)

648 (6.9)

446 (9.4)

35,645 (76)

7,967 (22)

6,868 (6.7)

3,469 (6.3)

38,290 (75)

9,388 (12)

7,516 (6.7)

3,915 (6.6)

p<0.0001

Number of children <18 years       0       1

       2+

Time since service

Currently active duty

Active duty past 12 months

Active duty over 12 months ago

4,232 (67)

400 (10)

531 (23)

101 (3.8)

119 (4.0)

4,969 (92)

46,924 (75)

3,016 (9.7)

4,061 (16)

1,297 (5.2)

1,242 (3.6)

51,582 (91)

51,156 (74)

3,416 (9.8)

4,592 (16)

1,398 (5.0)

1,361 (3.6)

56,551 (91)

p<0.0001

p=0.2130

Physical health status        No frequent physical distressFrequent physical distress

2,453 (52)

2,617 (48)

46,632 (89)

6,426 (11)

49,085 (86)

9,043 (14)

p<0.0001

Days without enough sleep in Past month? (mean in days[95%CI])

17.4 [16.8-18.0]

6.1 [5.9-6.2]

7.1 [7.0-7.3]

p<0.0001

Emotional support        Sometimes, rarely or neverUsually or always

2,261 (43)

2,567 (57)

9,539 (16)

40,951 (84)

11,800 (19)

43,518 (81)

p<0.0001

*unweighted totals (weighted percentages)

€ Within each covariate, the frequencies may not add up to the total number of respondents in that category (N) because of missing and/or invalid observations, which have been excluded from the table.

Frequent physical distress defined as >14 days poor mental health in past 30 days

‡Frequent mental distress was defined as >14 days poor mental health in past 30 days

 

Table 2 – Crude and Adjusted Odds Ratios for Frequent Mental Distress among 2010 US Currently or Formerly Active Duty Adults (N=58,906)

 

Demographic   characteristics and risk factors

                                      Odds   of frequent mental distress          Crude OR (95% CI)                          Adjusted OR (95%   CI)

Age (in quartiles)       18-45 years       46-58 years

       59-69 years

       70-99 years

1.00 (ref)

1.00 (0.84-1.18)

0.68 (0.58-0.80)

0.36 (0.31-0.43)

1.00 (ref)

0.79 (0.62-1.02)

0.69 (0.53-0.91)

0.38 (0.27-0.52)

Gender         MaleFemale

0.52 (0.44-0.62)

1.00 (ref)

0.56 (0.44-0.71)

1.00 (ref)

Race/ethnicityNon-Hispanic whiteNon-Hispanic black

Hispanic

Non-Hispanic multi

Other

1.00 (ref)

1.21 (1.00-1.46)

1.48 (1.10-2.00)

1.85 (1.40-2.45)

1.63 (1.16-2.30)

1.00 (ref)

0.76 (0.57-1.01)

0.99 (0.63-1.54)

0.80 (0.53-1.20)

0.81 (0.47-1.42)

Income level        <$25,000<$50,000

<$75,000

³$75,000

Don’t   know or refused

4.41 (3.70-5.25)

1.95 (1.63-2.34)

1.56 (1.28-1.91)

1.00 (ref)

1.93 (1.54-2.41)

1.83 (1.43-2.35)

1.57 (1.24-1.99)

1.42 (1.11-1.80)

1.00 (ref)

1.53 (1.16-2.03)

Employment       Employed (or homemaker,student, retiree)

Unemployed

1.00 (ref)

3.68 (2.92-4.64)

1.00 (ref)

2.19 (1.69-2.85)

Education levelDid not finish high schoolFinished high school

Some college

Finished college

3.06 (2.43-3.85)

1.67 (1.43-1.95)

1.99 (1.73-2.29)

1.00 (ref)

1.15 (0.87-1.51)

0.89 (0.73-1.09)

1.23 (1.02-1.48)

1.00 (ref)

Time since serviceCurrently active dutyActive duty past 12 months

Active duty over 12 months ago

0.73 (0.51-1.04)

1.11 (0.75-1.64)

1.00 (ref)

0.73 (0.45-1.19)

0.77 (0.51-1.19)

1.00 (ref)

Marital statusMarried or member ofunmarried couple

Divorced or separated

Widowed

Never married

0.54 (0.44-0.66)

1.35 (1.06-1.73)

0.68 (0.54-0.87)

1.00 (ref)

0.97 (0.72-1.33)

1.24 (0.87-1.78)

1.51 (1.06-2.15)

1.00 (ref)

Number of children <18 years       0       1

       2+

Physical health status

       No frequent physical distress

Frequent physical distress

1.00 (ref)

1.14 (0.95-1.36)

1.59 (1.32-1.92)

1.00 (ref)

7.69 (6.81-8.68)

1.00 (ref)

0.86 (0.66-1.13)

1.14 (0.86-1.52)

1.00 (ref)

4.71 (3.95-5.62)

Days without enough sleep in Past month? (mean in days[95%CI])

1.09 [1.08-1.09]

1.07 [1.06-1.07]

Emotional support        sometimes, rarely or neverhas adequate   emotional support

always or usually has

adequate   emotional support

1.00 (ref)

0.26 (0.23-0.30)

1.00 (ref)

0.38 (0.32-0.45)

Frequent mental distress defined as reporting poor mental health on 14 or more days out of the past 30 days

CI = confidence interval

§Frequent physical distress defined as reporting poor physical health

References

(1)     United States Department of Veterans Affairs. Veteran Population. National Center for Veterans Analysis and Statistics. January 12, 2010. Available at https://www1.va.gov/vetdata/Veteran_Population.asp. Accessed September 26, 2013. (2)     United States Government Accountability Office.  VA Mental Health: Number of Veterans Receiving Care, Barriers Faced, and Efforts to Increase Access; 2011. (3)     Wagner TH, Sinnott P, Siroka AM.  Mental Health and Substance Use Disorder Spending in the Department of Veterans Affairs, Fiscal Years 2000-2007.  Psychiatric Services.  2011;62(4):389-395. (4)     OECD. Health Policies and Data. A System of Health Accounts. 2013. Available athttps://www.oecd.org/els/health-systems/asystemofhealthaccounts.htm#Data_collection. Accessed September 26, 2013. (5)     Seal KH, Metzler TJ, Gima KS, et al. Trends and Risk Factors for Mental Health Diagnoses Among Iraq and Afghanistan Veterans using Department of Veterans Affairs Health Care, 2002-2008.  Am J Public Health.  2009;99(9):1651-1658. (6)     Pols H, Oak S.  War and Military Mental Health.  Am J Public Health.  2007;97(12):2132-2142. (7)     Hoge CW, Castro CA, Messer SC, et al.  Combat Duty in Iraq and Afghanistan, Mental Health Problems, and Barriers to Care.  N Engl J Med.  2004;351(1):13-22. (8)     US Census Bureau.  Veterans: Definitions and Concepts.  May 21, 2013.  Available at https://www.census.gov/hhes/veterans/about/definitions.html#Q2.  Accessed February 27, 2014. (9)     CDC. 2010 Health Risks in the United States: Behavioral Risk Factor Surveillance System At-a-Glance. Chronic Disease Prevention and Health Promotion.  February 23, 2010.  Available at https://www.cdc.gov/chronicdisease/resources/publications/AAG/brfss.htm. Accessed September 26, 2013. (10) CDC.  Self-reported frequent mental distress among adults: United States 1993-2001.  Morbidity and Mortality Weekly Report.  2004; 53(41): 963-966. (11) Mayo Clinic staff. Depression (major depression). Health Information. February 10, 2012. Available at https://www.mayoclinic.com/health/depression/DS00175/DSECTION=tests. Accessed September 26, 2012 (12) Mayo Clinic staff.  Generalized Anxiety Disorder.  Health Information.  September 8, 2011.  Available at https://www.mayoclinic.com/health/generalized-anxiety-disorder/DS00502/DSECTION=tests-and-diagnosis (13) Harvard Medical School. Sleep and Mental Health. Harvard Mental Health Letter.  July 2009.  Available at https://www.health.harvard.edu/newsletters/Harvard_Mental_Health_Letter/2009/July/Sleep-and-mental-health. Accessed September 26, 2012. (14) Brown GW, Andrews B, Harris T, et al.  Social Support, Self Esteem, and Depression. Psychol Med. 1986;16(4):813-31. (15) Turner RJ, Noh S.  Physical Disability and Depression: A Longitudinal Analysis.  J Health Soc Behav.  1988;29(1):23-37. (16) Gade DM, Wenger JB.  Combat Exposure and Mental Health: the Long-Term Effects among US Vietnam and Gulf War Veterans.  Health Econ.  2011;20(4):401-416. (17) Paul KI, Moser K.  Unemployment Impairs Mental Health: Meta-Analyses.  J Vocat Behav.  2009;74(3):264-282. (18) Pharr JR, Moonie S, Bungum TJ.  The Impact of Unemployment on Mental and Physical Health, Access to Health Care, and Health Risk Behaviors.  ISRN Public Health.  2012;1-7. (19) CDC. Frequently Asked Questions. Health Related Quality of Life. March 15, 2011. Available athttps://www.cdc.gov/hrqol/faqs.htm-and-diagnosis. Accessed September 26, 2012. (20) STATA [computer program].  Version 12.  College Station, TX.  Statacorp; 2011 (21) Brown DW, Balluz LS, Ford ES, et al.  Associations Between Short- and Long-Term Unemployment and Frequent Mental Distress Among a National Sample of Men and Women.  J Occup Env Med.  2003;45(11):1159-1166. (22) Lorant V, Deliège D, Eaton W, et al.  Socioeconomic Inequalities in Depression:  A Meta-Analysis.  Am J Epidemiol.  2003;157(2):98-112. (23) Williams A, Swail WS.  Is more better?: The impact of post-secondary education on the economic and social well-being of American society.  Educational Policy Institute: American Higher Education Report Series.  2005. (24) Angrist JD. The Effect of Veterans Benefits on Education and Earnings. Industrial and Labor Relations Review. 1993;46(4):637-52. (25) Bound J, Turner M. Going to War and Going to College: Did World War II and the G.I. Bill Increase Educational Attainment for Returning Veterans? Journal of Labor Economics. 2002;20(4):784-815. (26) Lutz A.  Who joins the military?: A look at race, class and immigration status.  Journal of Political and Military Sociology.  2008;36(2):167-88. (27) Seal KH, Bertenthal D, Miner CR, et al.  Bringing the War Back Home: Mental Health Disorders Among 103,788 US Veterans Returning From Iraq and Afghanistan Seen at Department of Veterans Affairs Facilities.  Arch Intern Med.  2007;167(5):476-482. (28) Spiro A, Schnurr PP, Aldwin CM.  Combat-related posttraumatic stress disorder symptoms in older men.  Psychol Aging.  1994;9(1):17-26. (29) Goodman GP, DeZee KJ, Burks R, et al. Epidemiology of Psychiatric Disorders Sustained by a U.S. Army Brigade Team During the Iraq War. Gen Hosp Psychiatry. 2011;33(1):51-57 (30)  Hoge CW, Auchterlonie JL, Milliken CS. Mental Health Problems, use of Mental Health Services and Attrition from Military Service after Returning from Deployment to Iraq or Afghanistan. JAMA.  2006;295(9):1023-1032. (31) Kehle SM, Reddy MK, Ferrier-Auerbach AG, et al. Psychiatric Diagnoses, Comorbidity, and Functioning in National Guard Troops Deployed to Iraq. J Psychiatr Res. 2011;45(1):126-132. (32) Skinner KM, Kressin N, Frayne S, et al.  The Prevalence of Military Sexual Assault Among Female Veterans’ Administration Outpatients.  J Interpers Violence.  2000;15(3):291-310. (33) Booth-Kewley S, Schmied EA, Highfill-McRoy RM, et al.    Predictors of Psychiatric Disorders in Combat Veterans.  BMC Psychiatry.  2013;13:130. (34) Mayo Clinic staff.  Depression in women: Understanding the gender gap.  Diseases and conditions: In-depth.  January 19, 2013.  Available at < https://www.mayoclinic.com/health/depression/MH00035>.  Accessed October 28, 2013. (35) Winerman L.  Helping men to help themselves.  Monitor Psychol.  2005;36:57-59. (36) Harvard Medical School.  Recognizing depression in men.  Harvard Mental Health Letter.  June 2011.  Available at < https://www.health.harvard.edu/newsletters/Harvard_Mental_Health_Letter/2011/June/recognizing-depression-in-men>.  Accessed October 28, 2013. (37) Kawachi I, Berkman LF.  Social Ties and Mental Health.  J Urban Health.  2001;78(3):458-467. (38) Mittleman MS, Ferris SH, Shulman E, et al. A Comprehensive Support Program: Effect on Depression in Spouse-Caregivers of AD Patients. Gerontologist. 1995;35:792-802. (39) Harris T, Brown GW, Robinson R. Befriending as an Intervention for Chronic Depression among Women in an Inner City. 1: Randomized Controlled Trial. Br J Psychiatry. 1999;174:219-224. (40) Brand EF, Lakey B, Berman S. A Preventive, Psychoeducational Approach to Increase Perceived Social Support. Am J Community Psychol. 1995;23:117-135 (41) Paluska SA, Schwenk TL.  Physical Activity and Mental Health.  Sports Medicine.  2000;29(3):167-180. (42) Emerson E, Honey A, Llewellyn G. The Well-Being and Aspirations of Australian Adolescents and Young Adults with a Long-Term Health Condition, Disability or Impairment. Canberra: Australian Research Alliance for Children and Youth.  Working paper. (43) Honey A, Emerson E, Llewellyn G. in press. The mental health of young people with disabilities: Impact of social conditions. Social Psychiatry and Psychiatric Epidemiology. 2011;46(1):1-10. (44) Miauton L, Narring F, Michaud PA. Chronic Illness, Life Style and Emotional Health in Adolescence: Results of a Cross-Sectional Survey on the Health of 15-20-year-olds in Switzerland. Eur J Pediatr. 2003;162(10):682-689. (45) Wolman C, Resnick MD, Harris LJ, et al. Emotional Well-Being among Adolescents with and without Chronic Conditions. J Adolescent Health. 1994;15(3):199-204. (46) Maruta T.  Depression Associated with Chronic Pain: Incidence, Characteristics, and Long-Term Outcome.  Keio J Med.  1989;38(4):403-412. (47) Mayo Clinic staff. Post-traumatic stress disorder (PTSD). Health Information. April 8, 2011. Available at https://www.mayoclinic.com/health/post-traumatic-stress-disorder/DS00246/DSECTION=symptoms. Accessed November 7, 2013. (48) Britt TW, Greene-Shortridge TM, Castro CA.  The Stigma of Mental Health Problems in the Military.  Mil Med.  2007;172(2):157-61(5). (49) CDC.  Behavioral risk factor surveillance system 2010 summary data quality report.  May 5, 2011.  Available at ftp://ftp.cdc.gov/pub/data/brfss/2010_summary_data_quality_report.pdf.  Accessed February 10, 2014. (50) CDC.  Behavioral risk factor surveillance system 2009 summary data quality report.  February 18, 2011.  Available at ftp://ftp.cdc.gov/pub/data/brfss/2009_Summary_Data_Quality_Report.pdf.  Accessed February 10, 2014. (51) CDC.  Behavioral risk factor surveillance system 2008 summary data quality report.  February 18, 2011.  Available at ftp://ftp.cdc.gov/pub/data/brfss/2008_Summary_Data_Quality_Report.pdf.  Accessed February 10, 2014. (52) CDC.  Behavioral risk factor surveillance system 2007 summary data quality report.  March 9, 2008.  Available at https://www.cdc.gov/brfss/annual_data/annual_2007.htm.  Accessed February 10, 2014. (53) Powell-Griner E, Town M, Nelson DE, et al.  National estimates of risks/behaviors: Behavioral Risk Factor Surveillance System and National Health Interview Survey.  Proceedings of the Annual Meeting of the American Statistical Association.  August 5-9, 2001.  Available at https://www.amstat.org/sections/srms/proceedings/y2001/Proceed/00397.pdf.  Accessed February 10, 2014. (54) LeardMann CA, Powell TM, Smith TC, et al.  Risk factors associated with suicide in current and former US military personnel.  JAMA.  2013;310(10):1076. (55) Pflanz S.  Occupational Stress and Psychiatric Illness in the Military: Investigation of the Relationship between Occupational Stress and Mental Illness among Military Mental Health Patients.  Mil Med.  2001;166(6):457-462. (56) Friedman MJ. Veterans’ Mental Health in the Wake of War. N Engl J Med. 2005;352(13):1287-1290.

Acknowledgements

Reader Feedback