Young adulthood is associated with the initiation and escalation of health risk behaviors that are among the leading causes of preventable morbidity and premature death in the United States. Rates of excessive alcohol use, sexually transmitted infections (STIs), and motor vehicle accidents peak during this life stage (
1–
3). Approximately 40% of young adults self-report binge drinking (
4), which contributes to over one-third of premature deaths in this age group (
5). Over half of the 19 million STIs diagnosed annually afflict 15- to 24-year-olds (
6). Tobacco use accounts for 443,000 deaths annually (
6), and 90% of smokers initiate tobacco use before age 21.
Although these risk behaviors are modifiable through preventive care interventions (
7–
9), approximately 30% of U.S. young adults (18–29 years of age) lack health insurance (
3,
10). Young adults are least likely to have a regular medical provider or facility, and their annual rate of emergency department utilization (8%) is second only to those over age 75 (
2). An integral part of Healthy People 2010 and 2020, a set of national health objectives, involves enhancing young adults’ access to preventive care to reduce substance use–related injuries and STIs (
6). Thus knowledge about health disparities and access to services among subgroups of young adults is needed to inform targeted prevention and outreach efforts.
Young adults who have served in the military have unique health care needs compared with the general young adult population (
11–
13). Veterans have disproportionately high rates of posttraumatic stress disorder (PTSD), depression, and substance abuse (
14–
17). Although younger age and male gender are associated with greater health risk behaviors and lower health care utilization among veterans (
18,
19) and civilians (
2,
20), comparatively less attention has focused on risk behaviors and health care access and utilization among young veterans, particularly female veterans, compared with civilians.
Population-based study findings are not consistent with regard to the impact of veteran status and gender on health indicators among young adults. Evidence suggests higher rates of heavy drinking among veterans compared with civilians of similar age and gender (
21,
22), although a recent nationally representative study of men observed comparable rates between young adult veterans and civilians (
23). Nationally representative data indicate that smoking rates among male and female veterans (27% and 23%, respectively) are higher compared with their civilian counterparts (23% and 18%, respectively) (
11). Studies of veterans seeking health care at U.S. Department of Veterans Affairs (VA) facilities suggest that veterans are more likely than civilians in similar age samples to be overweight (
16,
24). Although studies (
25–
27) show high rates of risk behaviors (such as drinking and driving and risky sexual behavior) among soldiers after a military deployment, no research has documented population-based estimates of risk behaviors in association with gender and veteran status.
Limitations of previous research include samples comprising only men (
22,
23) or only women (
28,
29), thereby not allowing gender comparisons. Studies of veterans have typically included those engaged in the VA health care system (
18,
24), so findings may not generalize to veterans who are patients in non-VA care (
30). Given the growing numbers of veterans separating from military service, a more comprehensive understanding of health risks and access to care is needed to inform targeted health interventions for young adult men and women to prevent future chronic disease. This study evaluated a nationally representative sample of U.S. young adult veterans and civilians on general health, risk behaviors, access to care, and health care utilization. We hypothesized that male and female veterans would have poorer health and greater health risk behaviors despite greater access to and utilization of services than civilian counterparts. Further, we hypothesized that among both veterans and civilians, women would have greater health care utilization and lower rates of health risk behaviors compared with men.
Methods
Study design, data source, and study population
This retrospective, cross-sectional study used data obtained from the Centers for Disease Control and Prevention (CDC) Behavioral Risk Factor Surveillance System (BRFSS), a national telephone survey conducted annually to monitor health conditions of U.S. adults. BRFSS data are publicly available (
31), and the survey uses state-level sampling plans and weighting to obtain a representative sample from households with telephones. Each respondent is assigned a final sampling weight based on probability of selection and a poststratification factor to ensure that the distribution of the weighted sample agrees with population estimates from the U.S. Census Bureau. We analyzed 2010 BRFSS data from all states that included items on health care access, health status, alcohol consumption, and tobacco use. In addition, respondents from 13 states completed items on depression and anxiety. A total of 451,075 adults completed the 2010 survey.
Respondents were asked whether they had “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.” Among 28,183 men and women surveyed between the ages of 19 and 30, 42 were missing data on military status. Because of the 12-month time frame for measures of health care access and utilization, we excluded respondents who reported “current active duty” (N=315), “active duty within the past 12 months but not currently” (N=169), and “service in the National Guard or Reserves” (N=186). The final sample was 27,471 respondents categorized into two groups: veteran (631 endorsed “active duty in the past but not during the last 12 months”) and civilian (26,840 endorsed “never served in the military”). Study approval was obtained from the VA Puget Sound Institutional Review Board.
Measures
Demographic characteristics.
The BRFSS collects self-reported age, race and ethnicity, marital status, education, employment status, and income. To create a variable reflecting where respondents lived, we grouped states into Census Bureau Regions (Northeast, Midwest, South, and West) by their Federal Information Processing Standard codes, and we grouped participants living in Guam, Puerto Rico, and the Virgin Islands into an “Islands” region.
Health care access and utilization.
Respondents were asked whether they had any kind of health care coverage (such as health insurance or Medicare) and whether there was a time in the past 12 months when they needed to see a physician but could not because of cost. Respondents reported whether they had a personal physician or health care provider and how much time had elapsed since they visited a physician for a routine checkup (within the past year or more than one year). Eligibility and enrollment in VA care were not assessed in the 2010 BRFSS.
General medical health and mental distress.
Poor general medical health and mental distress were assessed with two items: number of days of poor general medical health in the past 30 days and number of days of mental distress in the past 30 days. Consistent with a recent BRFSS study validating the use of a six-day cutoff point (
32), poor general medical health and mental distress were dichotomized (six or more days versus fewer than six days).
Respondents were asked whether a health care provider ever told them they had a depressive disorder (such as major depression) or an anxiety disorder (such as generalized anxiety disorder or PTSD). Overweight and obesity status were derived by calculating body mass index (BMI) based on self-reported height and weight: normal weight (BMI <25 kg/m
2) and overweight or obese (BMI ≥25 kg/m
2) (
33). Respondents indicated whether they exercised regularly in the past month, defined as physical activity or exercise outside of regular job duties.
Health risk behaviors.
Respondents reported whether they had engaged in the following behaviors: smoking and other tobacco use (currently smoking or using tobacco at least some days in past month), binge drinking (four or more drinks for women and five or more drinks for men at least once during the past month), heavy drinking (one or more drinks for women and two or more drinks for men per day in the past month), drinking and driving in the past month, HIV/AIDS risk behaviors (injection drug use or unprotected sex in the past year), and lack of seat belt use (not always using a seat belt when driving or riding in a car).
Data analysis
Analyses were performed in Stata/IC version 11.2 (
34) and accounted for survey design and weighted sampling probabilities. Respondents with missing data on income were included in a separate income category (
35). Descriptive statistics, stratified by gender, were used to compare veterans and civilians on demographic characteristics and health indicators; t tests compared means, and chi square tests compared proportions. Multivariable logistic regression was used to generate odds ratios and 95% confidence intervals for health indicators to identify main effects and the interaction of gender and veteran status, with adjustment for sociodemographic characteristics.
Results
Sample characteristics
We identified 631 veterans (74.6% were male, N=471) and 26,840 civilians (37.6% were male, N=10,080) ages 19 to 30. Weighted proportions of sociodemographic characteristics are presented by gender for veterans and civilians in
Table 1. Compared with civilian counterparts, veterans were older, more likely to be married or partnered, and more likely to have graduated from high school or to have a GED but were less likely to have completed college.
Health care access and utilization
Weighted proportions for health care access and utilization according to veteran status are presented in
Table 2. Among men, veterans were more likely than civilians to have health insurance, although no significant differences were found in health care utilization. Female veterans and civilians did not differ on rates of health insurance coverage or health care utilization. Veteran status was not associated with differences in financial barriers to receiving health care for either men or women.
General medical and mental health
Compared with their civilian counterparts, veterans were more likely to report poor general medical health in the past month and were more likely to be overweight or obese (
Table 2). Male veterans had higher rates of lifetime depression and anxiety disorders than civilian men. Among women, veterans had significantly higher rates of anxiety disorders than civilians and had higher absolute rates of depression (33.9% versus 20.4%), but the difference for the latter was not statistically significant.
Health risk behaviors
Male and female veterans had higher rates of current smoking compared with civilians (
Table 2). No differences between men and women were found in rates of the other health risk behaviors for veterans and their civilian counterparts.
Adjusted analyses of health outcomes
Table 3 presents findings from multivariate logistic regression models testing the main effects and interaction of veteran status and gender with adjustment for sociodemographic covariates.
Health care access and utilization.
There were main effects for gender predicting health care access and utilization. Women were more likely than men to have insurance and to have had a routine checkup and a regular provider in the past year. Women were more likely than men to report not being able to see a provider due to cost. There were no main effects for veteran status or interactions between veteran status and gender.
General medical and mental health.
Main effects for gender indicated that women were more likely than men to report frequent poor general health, frequent mental distress, and lifetime depression or anxiety disorder. Although men were more likely than women to report having exercised in the past month, men were more likely to be overweight or obese. Main effects for veteran status indicated that veterans had poorer general health and were more likely than civilians to have a lifetime anxiety disorder. There were no two-way interactions between veteran status and gender.
Health risk behaviors.
Compared with women, men were more likely to report binge-drinking, heavy drinking, drinking and driving, smoking, and tobacco use, and less likely to always to wear a seat belt. There were no main effects for veteran status or interactions between veteran status and gender.
Discussion
This study sought to provide population-based estimates of health risk behaviors and health care access and utilization in a representative sample of adult veterans and civilians ages 19–30. Given the lack of preventive care guidelines for young adults (
8), identification of health disparities within this group informs the provision of care. Notable gender differences were identified across measures of health care access, utilization, and health status among the overall sample of young adults. Targeted outreach efforts may need to account for barriers to health care access and utilization that may differ for young adult men and women. Although women had higher rates of insurance coverage and health care utilization than men in our sample, women reported greater financial barriers to receiving care. Young adult women may be unable to access needed care due to competing school, work, and family demands (
36). With regard to young adult men, lower rates of health care utilization in conjunction with greater risk-taking behaviors compared with women may reflect men’s negative attitudes and stigma toward help seeking or perceived invulnerability to health consequences (
37).
After analyses adjusted for sociodemographic factors, veterans had poorer general health and higher rates of lifetime anxiety disorders compared with civilians. Exposure to trauma and military-related stress heightens the risk of experiencing comorbid psychiatric and general medical symptoms, particularly in the context of PTSD (
38,
39). Moreover, approximately one-third of veterans had a diagnosed depressive disorder. With research suggesting that rates of severe psychiatric distress and suicide peak during young adulthood, these findings support extending mental health assessment and early intervention efforts beyond the VA health system to other health care settings that serve veterans (
37). Contrary to expectations, being a veteran (versus a civilian) was not associated with higher rates of hazardous alcohol use and drinking and driving. Although prior work suggests higher rates of heavy drinking among military service members compared with civilians (
40), our findings highlight the importance of accounting for age, in addition to gender, when evaluating the association between veteran status and alcohol use (
23). Veterans in our sample were older and more likely than civilian counterparts to be married or partnered, which may serve as protective factors for hazardous drinking (
41,
42). It is also possible that heavy drinking among military personnel represents a temporary escalation that upon transition to veteran status returns to levels more similar to the general young adult population.
Most health care studies of veterans have used health care utilization information captured from VA databases, thereby selecting treatment-seeking patients who may have poorer health compared with nonusers of VA health care services (
43). Our finding that young adult veterans had poorer general health compared with their civilian peers suggests that veteran health disparities extend beyond those seeking VA services. Although these data did not include a measure of VA and non-VA health care utilization, our results underscore the need for clinicians to be cognizant of the mental and general medical problems affecting young adult veterans within and outside of the VA health care system. Despite these disparities in health status, rates of health care access and utilization did not differ between young adult veterans and civilians in adjusted analyses. In light of the availability of VA care at no cost to most veterans for five years after separating from military service in Afghanistan or Iraq, we expected veterans to report greater health care utilization compared with civilians. Thus the VA may need to further address patient- and system-level barriers to utilization of VA services for young adult veterans.
A number of challenges impede the delivery of evidence-based preventive care (
8). A recent study using a national sample of U.S. young adults found that an estimated 70% of visits to primary care providers did not include preventive care counseling; of particular concern were low rates of screening for mental problems, substance use, STIs, frequency of exercise, and obesity (
2). The use of clinical decision-making support tools for screening high-risk behaviors within primary care settings may assist providers faced with competing time demands (
44,
45) and allow for more targeted prevention approaches for young adult men, who appear to be particularly at risk for engaging in such behaviors regardless of veteran status. For example, Web-based behavioral health screening tools that recommend brief interventions depending on the results of screening may be particularly useful for young adults in busy primary care settings (
46,
47) and have been associated with patients’ self-reported knowledge gains and improved health outcomes (in regard to obesity and smoking, for example) (
48). Other innovative technologies to support self-management of substance use and mental health problems among veterans include the VA Web site “My HealtheVet” in addition to smartphone applications such as PTSD Coach from the U.S. Department of Defense and the National Center for Telehealth and Technology (
49).
Our study had several limitations. The cross-sectional nature of the BRFSS precludes evaluation of temporal relationships between health risk behaviors and health care utilization in relationship to veteran status. Longitudinal studies are necessary to evaluate health conditions before and after deployment and before and after other related developmental milestones, such as leaving home for the first time. In addition, the number of female veterans in the sample was relatively low, potentially limiting power to detect statistically significant differences between female veterans and civilians that may be clinically important, highlighting the need for prospective studies that include larger numbers of female veterans. Only a select number of states included BRFSS items assessing psychiatric issues, so those results should be interpreted with caution. The BRFSS also relies on single-item, self-reported measures, thus preventing a more thorough assessment of general medical and mental health conditions. Finally, rates of health care access and utilization reported in this study may underestimate current rates among U.S. young adults, because 2010 BRFSS data were collected before the recent implementation of the Affordable Care Act that now enables young adults to remain on their parents’ health insurance plan until age 26. As a result, greater numbers of young adults have coverage, and additional provisions of this law will likely increase utilization, including expanded coverage of screening services for young adults and greater funding for programs targeting health promotion and preventive care for young adults.
Conclusions
Strengths of this research include use of a U.S. population–based data set, assessment of multiple health indicators, and inclusion of both young adult veterans and civilians. Findings suggesting differences across veteran status and gender on chronic health risk factors—namely general medical and mental health problems, tobacco use, and being overweight or obese—may inform providers’ and policy makers’ efforts to improve young adults’ engagement in care, particularly those engaging in health risk behaviors. Results highlight the need for innovative outreach programs tailored to the needs of young adult veterans and civilians.
Acknowledgments and disclosures
This work was supported by grant TP 61-025 from the VA Puget Sound Health Services Research and Development Postdoctoral Fellowship Program. This article is the result of work supported by resources from the VA Puget Sound Health Care System. The views expressed in this article are those of the authors and do not necessarily reflect the views of the Department of Veterans Affairs.
The authors report no competing interests.