PROJECT SUMMARY
Though the majority of the military is composed of non-Hispanic white men, the number of racial/ethnic
minority and women veterans has steadily increased in recent years. African-American/Black and Hispanic
veterans make up the largest proportion of the minority veteran population (about 50% and 30% of minority
veterans, respectively) and the number of women veterans, especially racial/ethnic minority women, is
expected to grow. Noted differences in alcohol use and behavioral health symptoms of PTSD and depression,
as well as behavioral health care access, have been reported for post-9/11 African-American/Black, Hispanic
veterans, and women veterans. However, research on disparities in veterans’ behavioral health care service
utilization and behavioral health symptomology primarily focuses on veterans’ use of VA facilities, therefore
missing 50% of veterans not receiving VA care. Unfortunately, there is limited research on the associations
between perceived discrimination and quality of behavioral health care or behavioral health care access
among racial/ethnic minority and women veterans who receive care outside of the VA system. Further, the
studies that do exist on disparities among veterans tend to be cross-sectional and they limit their examination
of disparities to individual factors only and do not incorporate larger environmental factors. Thus, the current
study seeks to extend prior work by recruiting 2,000 non-VA attending veterans with 4 years of follow up (bi
annual assessments). First, in an attempt to better understand veterans’ experiences of discrimination, 65
veterans will be recruited for in depth qualitative interviews. Results of these interviews will inform (not
determine) measurement in the larger study. Once completed, recruitment will begin for the longitudinal cohort.
We will oversample for racial/ethnic minority veterans (70% of the total recruited sample; n =1,400) and women
veterans (40% of both racial/ethnic minority and non-Hispanic white recruited veterans; n = 800). Outcome
data will be collected on behavioral health symptoms, including alcohol use/disorder, PTSD, and depression,
as well as behavioral health care access (alcohol use treatment receipt, preparatory behaviors, attitudes about
treatment). Experiences of racial discrimination and sexism (e.g., dignity denial, microaggressions, gender
discrimination, sexual harassment, health care discrimination) as well as minority based and military specific
stressors will be collected each wave. In addition to individual level data, we will use publicly available datasets
(e.g., Census) to gather data on neighborhood deprivation, poverty, income inequality, segregation,
neighborhood violence, distance to health care facilities, and liquor/alcohol outlets. We will use machine
learning models that incorporate all hypothesized predictors across both individual and environmental domains
to determine which factors are most important in predicting behavioral health symptoms, care receipt, and
attitudes, and the valence of individual predictor effects on outcomes. The latter can provide a rank order of
importance of predictors to help guide future prevention, intervention, and policy efforts.
