Veteran participation in the All of Us Research Program: applying an intersectionality lens to evaluate participant diversity | BMC Medical Research Methodology

Diversity and inclusion in biomedical research are necessary to ensure advances in healthcare can be tailored to meet individual needs. U.S. Military Veterans are well established as an underrepresented group with unique exposures and health experiences of key importance to developing and implementing healthcare advances. Recent studies conducted at the VA have demonstrated that targeted recruitment strategies can promote Veteran representativeness in research [32, 33]. However, developing tailored approaches to drive Veteran representation requires an examination and understanding of Veteran diversity and identity focused on characteristics beyond age, sex, and race/ethnicity descriptors. Alongside these often-used measures of diversity, geographic differences and lived experience in health and wellness are of growing importance to research, particularly in combination with each other. Here we present evidence supporting previously undescribed Veteran intersectionality including 75 unique UBR combinations, and highlight the value of employing community-tailored methods and practices to better engage diverse participants in health research.

VA recruited and enrolled Veterans broadly for All of Us since 2018. Initial comparison of national Veteran demographics (age, sex, race/ethnicity) show our enrolled Veterans to be reflective of the U.S. Veteran population with a few differences (Fig. 3). Notable areas of difference between enrollee and national demographics are age and HLS identity. All of Us Veterans skew to those over the age of 65, likely due to their frequent use of VA facilities and availability during daytime clinic hours. Veterans identifying as HLS are noted in the All of Us data but not in the national demographics. All of Us allows participants to self-identify as HLS independent of other race/ethnicity categories. National Veteran demographics do not present HLS as a sole identity. We acknowledge differences in data collection and reporting methods for the two datasets and detail our efforts to make them comparable for the purposes of this analysis while still reporting key features unique to the All of Us dataset. While analysis of age, sex, and race/ethnicity demographics of VA All of Us participants detail a group that reflect the diversity of Veterans nationally, it is not necessarily a representative sample as the program is open to all Veterans [31].

Use of engagement approaches to meet distinct challenges (Table 1) allowed us to recruit and enroll a cohort of Veterans with complex identities (Figs. 1 and 2). For example, approximately one third of our participants met the criteria for three or more UBR categories (Fig. 1). Notable to us, 29.4% of those considered UBR by geography also met the criteria for three or more additional UBR categories, the most sizeable of which were age, disability, and income (Fig. 2), giving insight into often overlooked marginalization in more rural settings. While we have only piloted and conducted a small number of each type of engagement approach and therefore cannot report on comprehensive outcomes, the deployment of community-specific engagement and enrollment methods provided research opportunities to a broad segment of the Veteran population. Through our experience, we acknowledge that some engagement approaches appeared to be better received regionally and within specific groups than others, and have learned it is important to explore a variety of methods and seek regular community feedback to find what works best. Regardless of whether eventual enrollment stemmed from remote engagement, partnership with specific groups and community hubs, or our participant champions, to name a few examples, each of our engagement approaches took into account important factors (e.g., distance to an enrolling clinic, mobility, time, high-touch support, participant-to-participant referral, and trust building), which ultimately enabled us to spread program awareness in unique ways to greater and more diverse facets of the Veteran community.

Using All of Us operational data, our findings are the first to elucidate the complexity of Veteran identity beyond the scope of national statistics, with well over 75 unique UBR combinations identified among our participants. Though basic Veteran demographics are largely shaped by military recruitment and draft policies, shedding light on why many Veterans alive today identify as older, cisgender men [34], the broad range of datapoints (e.g., gender identity, access to healthcare, disability status, educational attainment, income, and geographical residence) collected by All of Us and viewed in combination demonstrates greater depth in demographics and health/wellness experiences than is often considered in enrollment reporting. Insights presented by Sabatello et al. [1] and Mapes et al. [6] suggest that identity and lived experience have a profound impact on health outcomes, lending weight to the need to consider intersectionality as a key variable in biomedical research.

The data presented here serve as a starting point for considering intersectional identity when planning and developing Veteran-specific programs. In the near future, ongoing changes to military recruitment strategy and policy will result in a more diverse Veteran landscape, including an increase in the proportion of Veterans from non-White racial and ethnic backgrounds (defined by VA as racial and ethnic minorities) from 27.5% in 2024 to 39.1% in 2053, women Veterans from 11.7% (2024) to 18.7% (2053), and younger Veterans (under the age of 45) from 21.9% (2024) to 25.5% (2053) [28], as well as growth in transgender and gender diverse-identifying Veterans receiving care at the VA [35]. Taken together, projections of Veteran population shifts, and the increasing importance of multi-faceted identity and lived experience for health outcomes, stress the importance of examining and understanding Veteran intersectionality. Considering complex and changing intersectional identities can ensure purposeful approaches to research representation and development of policy and practice to meet changing healthcare needs.

In response to shifts in Veteran identity over the years, VA has made great strides to meet the needs of specific communities, including creation of the Office of Rural Health in 2007, which works to reduce barriers that rural-living Veterans experience in receiving healthcare services [36], implementing the PACT Act passed in 2022, which expands eligibility for VA services to Veterans exposed to toxins during their military service [37], and the establishment of the Inclusion, Diversity, Equity, and Access Council in 2023 [38] and the Office of Health Equity LGBT Workgroup [39], both intended to identify inequities and offer solutions to delivery of benefits and services. VA efforts also include increased healthcare services offered to women Veterans and assignment of a Women Veterans Program Manager at every VA medical center [40]. We partnered with several of these groups to create community-based approaches to engage Veterans in the All of Us Research Program.

Our effort to engage and enroll a diverse group of Veterans in the All of Us Research Program was not without limitations. Funding constraints and protocol requirements limited our options for engagement, leading us to limit the frequency of outreach requiring staff travel, and sometimes seek out less costly solutions. Accessibility to digital devices and broadband internet required for All of Us enrollment also presented challenges, leaving some Veterans unable to complete program registration and enrollment without in-person support at an enrollment clinic. We also encountered Veterans who did not have access to transportation to a participating clinic and/or lived too far away but required in-person support, limiting our ability to assist them. Trust proved to be another obstacle to participation, underlining the importance of providing transparency throughout the research process, working to alleviate privacy concerns, and establishing localized personal relationships between program staff and participants.

While ours is just one research study and accounts for a small percentage of the Veteran population, Veterans are an important representation of the larger U.S. population and greater understanding of their intersectional identities is necessary for development of equitable and inclusive policies and programs that will better serve them [17]. In particular, more rigorous engagement of wide-ranging Veteran communities and groups – going beyond standard diversity considerations like age, sex, and race/ethnicity for research participation – is necessary to accurately identify their unique perspectives and needs. We have shared approaches to broaden participation aimed at increasing inclusion and present an intersectional view of the diversity of Veterans, but significant work is still needed to bring the concept of multi-faceted identity and the needs of the multiply marginalized to the forefront of planning for population-focused research. Ongoing program efforts are focused on further engaging women, rural-living, and sexual and gender minority identifying Veterans to better understand additional meaningful facets in their complex identities aiming to ensure inclusion in research.