Publications – Johns Hopkins University Disability Health Research Center

This open-access study examines how two common sets of disability survey questions—the American Community Survey (ACS-6) and the Washington Group Short Set (WG-SS)—capture different disability groups in U.S. federal data. Using data from the 2010–2018 National Health Interview Survey, the study looked at whether adults with eight types of disability (such as vision, hearing, intellectual/developmental, mental health, cancer, and diabetes) were identified as disabled by these question sets. 

The ACS-6 consistently identified more disabled adults than the WG-SS, across every disability type studied. For example, 91–94% of adults with vision or hearing disabilities were captured by the ACS, compared with only 66–68% by the WG-SS. The gap was even larger for groups not directly mentioned in the question prompts, such as people with mental health conditions, diabetes, or cancer. In these groups, fewer than half were identified by the WG-SS. 

While the ACS-6 performs better than the WG-SS, both undercount large segments of the disabled population. This undercounting matters because it shapes federal prevalence data and policy decisions. The study recommends expanding or supplementing the ACS-6 with additional questions to better capture intellectual and developmental disability, mental health, and physical health conditions. More inclusive questions will lead to stronger data, and ultimately, better-informed equity policies and programs. 

Citation: Landes, S. D., Hall, J. P., Swenor, B. K., & Vaitsiakhovich, N. (2025). Comparative performance of disability measures.PLOS ONE, 20(1): e0318745. https://doi.org/10.1371/journal.pone.0318745

This study examines the “disability mismatch”—the gap between how disability is measured in many US federal surveys and who is actually disabled. Current surveys often rely on the ACS-6 questions, which define disability only as having one of six functional limitations. But many disabled people either have limitations outside these domains, experience episodic or mitigated limitations, or do not experience limitations at all. This creates a narrow, reductionistic measure that excludes large portions of the disabled population. 

Using data from a nationally representative survey of 2,169 adults, the study tested a comprehensive disability status question. This question captures physical or mental health conditions, illnesses, impairments, or disabilities lasting 12 months or more. 

The comprehensive question identified disabled adults both with and without activity limitations and captured many people missed by the ACS-6. Across almost every self-described disability category—including mental health, physical health, neurological, musculoskeletal, and intellectual and developmental disabilities—the ACS-6 excluded a substantial share of respondents who said they had a disability. The comprehensive question consistently captured these individuals, demonstrating a more inclusive approach. 

The study suggests that a single comprehensive disability status question can serve as a more inclusive and accurate way to identify disabled people in federal surveys. Follow-up questions would still be needed to gather information about disability type, limitations, or accommodations, but beginning with a comprehensive question offers a more complete and representative measure of disability in the United States. 

Citation: Landes, S. D., Swenor, B. K., Hall, J. P., Forber-Pratt, A. J., Vaitsiakhovich, N., Caldwell, K., Kakara, M., Lefkowitz, D., Myers, A., & Popkin, S. J. (2025). The disability mismatch: the case for a comprehensive disability status measure. Health Affairs Scholar, 3(7), qxaf137. Doi: https://doi.org/10.1093/haschl/qxaf137 

Students with disabilities face many barriers in higher education, yet clear information about accessibility is often missing from university websites. This lack of transparency makes it harder for prospective students to evaluate whether schools can meet their access needs. To address this gap, researchers at the Johns Hopkins Disability Health Research Center created the University Disability Inclusion Dashboard, which scores universities on how well they share information about accessibility and disability inclusion. 

This study examined the top 50 universities in the United States that receive the most National Institutes of Health (NIH) funding. Each university was scored across 10 indicators, including accessibility, accommodations, grievance policies, and disability representation. Scores ranged from 0 to 100 and were converted into letter grades. Results showed poor performance overall: only 3 universities earned an “A,” while more than half scored a “D” or “F.” 

Importantly, no university characteristics—such as ranking, region, size, or public vs. private status—were linked to stronger inclusion scores. Even highly prestigious and well-funded institutions scored poorly. 

These findings highlight the urgent need for universities of all types to improve disability inclusion. Transparent, accessible policies are essential to ensure equity in higher education. 

Citation: Cerilli, C., Zhu, J., Varadaraj, V., Campanile, J., Sweeney, F., Smith, J., Yenokyan, G., & Swenor, B. (2025). Gaps in disability inclusion across universities in the United States. https://doi.org/10.1371/journal.pone.0317920

Explore the findings in action on the University Disability Inclusion Dashboard, an open-data platform built from this research. 

Falls are the leading cause of injury among older adults in the United States, and visual impairment (VI) is a known risk factor. This study used data from the 2021 and 2022 National Health and Aging Trends Study (NHATS) to explore how different aspects of vision relate to falls among community-dwelling Medicare beneficiaries aged 71 and older. 

The analysis included 2,822 participants, of whom more than one-third reported at least one fall in the previous year. At baseline, older adults with any VI—including reduced distance or near visual acuity, or impaired contrast sensitivity—had higher rates of falls and multiple falls compared to those without VI. However, when tracked over time, baseline VI alone did not predict higher risk in the following year. Instead, specific measures of declining vision were more telling: worsening contrast sensitivity was linked to a higher risk of multiple falls, while worsening distance and near visual acuity were associated with increased risk of single falls. 

These findings underscore the importance of monitoring not only whether an older adult has VI but also changes in specific aspects of their vision. Proactive screening and interventions targeting contrast sensitivity and acuity loss may help reduce fall risk as the population ages. 

Citation: Thomas, J., Almidani, L., Ramulu, P., & Varadaraj, V. (2025). Falls and multiple falls among United States older adults with vision impairment. American Journal of Ophthalmology, 271, 166–174. https://doi.org/10.1016/j.ajo.2024.11.012

Reliable disability data are essential for shaping effective and equitable public health policy. This Health Policy Brief emphasizes why expanding disability data collection at the state level is critical to understanding and addressing disparities affecting people with disabilities. 

The authors highlight that state governments play a pivotal role in data collection but currently lack standardized, comprehensive systems for tracking disability as a core demographic factor. Without consistent data, inequities in health care, housing, employment, and community participation remain under-addressed. 

Recommendations include adopting standardized disability questions, incorporating diverse data sources that reflect social determinants of health, strengthening community partnerships with people with disabilities, and building capacity within public health agencies to analyze and use disability data. The brief also calls for increased federal support and funding to help states implement these improvements. 

These actions would create the infrastructure needed to identify and close health gaps for people with disabilities, supporting the broader goal of health equity across the United States. 

Citation: Castro, F., Swenor, B. K., & Robles, J. (2025). Improved State-Level Disability Data Are Necessary for Advancing Health Equity. Health Policy Brief. https://doi.org/10.1377/hpb20250320.114325 

This article highlights ableism as a root barrier to health equity and argues that the recent designation of people with disabilities as a health disparity population marks an important but incomplete step toward meaningful change. While this policy shift reflects years of advocacy by the disability community, the authors note that public health—and epidemiology in particular—has historically overlooked the role of ableism in shaping health inequities.

The paper describes how epidemiology can play a central role in addressing ableism by reframing disability not as an outcome, but as a population-level experience shaped by discrimination and structural barriers. It emphasizes the need to move beyond medicalized models of disability and calls for more inclusive, modern frameworks that separate disability from disease and functioning, while better reflecting the perspectives and experiences of disabled people.

The paper also calls for concrete action across three domains: improving how disability is measured and conceptualized; collecting disability data as a core demographic variable across surveys, research, and health systems; and ensuring accessibility throughout data collection, analysis, and dissemination. Together, these efforts are presented as essential to preventing the reinforcement of ableism in public health research and to ensuring disability is fully integrated into health equity and social justice efforts.

Citation: Swenor, B.K., Varadaraj, V., & Castro, F.F. (2025). The Role of Epidemiology in Addressing Ableism. Epidemiology, 36(1), 76–78. https://doi.org/10.1097/EDE.0000000000001790

Between September 2023 and May 2024, three landmark policy milestones advanced disability health equity in the United States. This commentary highlights these achievements—while underscoring the ongoing need for comprehensive disability data to ensure progress is equitable and sustainable. 

The first milestone was the National Institute on Minority Health and Health Disparities’ (NIMHD) formal recognition of people with disabilities as a health disparity population. The second came when the U.S. Census Bureau reversed plans to alter disability questions in the American Community Survey after advocacy from the disability community. The third was the Department of Health and Human Services’ updated Final Rule strengthening Section 504 of the Rehabilitation Act of 1973, clarifying protections against discrimination in federally funded programs. 

While these actions represent major steps forward, the authors emphasize that gaps in disability data—including inconsistent measures and limited representation—continue to hinder full realization of health equity. They call for federal investment in developing inclusive, standardized disability measures that accurately reflect the diversity of people with disabilities and can be applied across national surveys and health systems. 

Citation: Landes, S. D., & Swenor, B. K. (2025). A Year of Disability Health Equity Milestones: Why Disability Data Is Still Needed. Disability and Health Journal, 18(2), 101723. https://doi.org/10.1016/j.dhjo.2024.101723 

This systematic review examined the extent to which autism interventions in the United States have been designed or culturally adapted for Black/African American individuals. Despite growing recognition of racial inequities in autism care, the specific landscape of culturally responsive interventions remains unclear. 

Researchers reviewed literature published between 1999 and 2023 and identified eight studies meeting inclusion criteria: four designed specifically for Black autistic individuals, one adapted from an existing intervention, and three evaluated in samples including Black participants. Using the Cultural Adaptation Checklist, the authors found that few interventions systematically incorporated cultural adaptations. While several studies demonstrated thoughtful engagement with Black families and communities, no shared framework guided these efforts. 

The findings highlight an urgent need for evidence-based autism interventions that are culturally responsive to the Black/African American population. Developing such interventions could improve treatment engagement, service quality, and equity in autism research and care. 

Citation: Davis, A. M., Burks-Abbott, G., Merecias, O., & Swenor, B. K. (2024). Autism interventions designed or adapted for the Black/African American population: A systematic review. Autism, 29(1), 26-40. https://doi.org/10.1177/13623613241259910 (Original work published 2025) 

This open-access study looks at how the National Health Interview Survey (NHIS) counts disabled adults—and how the wording of survey questions can change what we learn. The team compared two question sets: the American Community Survey (ACS) questions and the Washington Group Short Set (WG-SS) questions currently used in the NHIS. They analyzed NHIS data from 2011–2012 and and monitored participants over time, up to 2019. 

The big takeaway: the WG-SS missed more than half of the adults whom the ACS identified as disabled. When the WG-SS is used, the share of U.S. adults counted as disabled looks much smaller—about 8.1%—compared with 17.0% using the ACS. That means many people, especially those with just one disability (for example, only vision or only mobility), are not being counted. 

Because the WG-SS mainly captures people with multiple disabilities, it also makes disability look more strongly linked to death than it really is. In other words, death rates among “counted” disabled adults appear higher with the WG-SS than with the ACS, partly because the WG-SS is selecting a group with more complex needs. 

The study recommends re-evaluating how the NHIS measures disability and investing in better, broader questions. Right now, the ACS looks like the stronger starting point for counting a wider range of disabled adults, with added items needed to include groups often missed (such as intellectual/developmental disability, mental health conditions, and chronic illness). Better data means better policies, funding, and care. 

Citation: Landes, S., Swenor, B.K., & Vaitsiakhovich, N. (2024). Counting Disability in the National Health Interview Survey and its consequences: Comparing the American Community Survey to the Washington Group Disability. Disability and Health Journal, 17(2), 101553. https://doi.org/10.1016/j.dhjo.2023.101553

This chapter explores how artificial intelligence (AI) and data systems can either reinforce or challenge ableism—the discrimination and exclusion of people with disabilities. Historically, data collection and AI technologies have often excluded or misrepresented disabled people, shaping systems that ignore their lived experiences. As AI becomes increasingly central to how decisions are made, these patterns risk being automated on a larger scale. 

The authors outline what more equitable futures could look like. They describe “achievable imaginaries”: realistic, disability-led visions where AI and data are used to promote inclusion, justice, and self-determination. By centering disabled voices and leadership, the chapter shows how data and AI can be redesigned to support—not harm—disabled communities. 

Citation: Newman-Griffis, D., Swenor, B. K., Valdez, R., & Mason, G. (2024). Disability data futures: Achievable imaginaries for AI and disability data justice. ArXiv. preprint arXiv:2411.03885. Doi:https://doi.org/10.48550/arXiv.2411.03885 

Using national data from the National Health Interview Survey (2010–2018), this study examined how vision difficulty and diabetes affect eye care utilization among U.S. adults. The analysis included over 284,000 participants, exploring differences by health status, education, race, and other sociodemographic factors. 

Results showed that adults with both vision difficulty and diabetes were most likely to have received eye care in the past year, while those without either condition were least likely. However, disparities persisted—particularly among racial and ethnic minorities and individuals with lower education levels. Women and those with higher educational attainment were significantly more likely to report recent eye care. 

These findings highlight that, although people with diabetes and vision difficulty are more engaged with eye care systems, inequities in access remain across marginalized groups. The authors emphasize the need for targeted interventions and policies that reduce barriers and ensure equitable access to preventive eye care services for all people. 

Citation: Brinson, J., Kumar, P., Wang, J., Varadaraj, V., Swenor, B. K., & Scott, A. W. (2024). Disparities in Eye Care Utilization by Self-Reported Vision Difficulty and Diabetes Status in the United States. Ophthalmic Epidemiology, 31(3), 283–290. https://doi.org/10.1080/09286586.2023.2249540 

This study assessed how well federally administered, nationally representative U.S. surveys include disabled people. Researchers reviewed 201 surveys run by HHS and the Census Bureau and, using set criteria (nationally representative; fielded 2018–2023; person/household samples; adults ≥18), examined 30 surveys to see whether they: (1) oversampled disabled people, (2) had a data-collection accessibility protocol (a clear plan ensuring access under the ADA and Sections 504/508), and (3) offered multiple ways to participate (such as phone, paper, in-person, or online). 

Of the 30 surveys included, only one appeared to oversample disabled people, and none had a clear plan to make data collection accessible. Most surveys did not offer accommodations, and when they did, the support was informal or inconsistent. Some surveys even required respondents to see or hear well enough to complete the survey independently, effectively excluding many disabled adults. 

These findings show that many survey design and data-collection practices can exclude disabled people, despite disabled adults making up nearly 27% of the U.S. population. The study highlights the need for oversampling, formal accessibility protocols, universal design approaches, and transparent practices that comply with federal disability rights laws to ensure national surveys are truly representative and support equitable policy and resource allocation. 

Citation: Cerilli, C., Varadaraj, V., Choi, J., Sweeney, F., Castro, F., Landes, S. D., & Swenor, B. K. (2024). Disability inclusion in national surveys. Health Affairs Scholar, 2(12), qxae145. Doi: https://doi.org/10.1093/haschl/qxae145 

This study examines how U.S. federal surveys underreport disability by comparing two common question sets: the American Community Survey six-question set (ACS-6) and the Washington Group Short Set (WG-SS). While the ACS-6 performs better than the WG-SS, both fail to capture many disability statuses, including intellectual and developmental disabilities, birth defects, mental health conditions, diabetes, and cancer. Underreporting can lead to inadequate funding and support for people with disabilities. The authors recommend using the ACS-6 with expanded questions and developing new inclusive measures to fully represent the disabled population in federal surveys.

Citation: Vaitsiakhovich, N., Landes, S. D., Hall, J. P., & Swenor, B. K. (2024). Disability measures used in U.S. federal surveys significantly underreport disability status. Disability and Health Journal

Through qualitative interviews with 35 U.S. researchers who identify as having a disability, this study explores barriers to accessibility, inclusion, and career advancement in academia. Themes included identity and visibility, disclosure and stigma, ableism, accommodation access, time/leave and workload expectations, and advocacy/community. Participants reported inconsistent or delayed accommodations, pressure not to disclose disabilities, structural ableism in policies and culture, and added uncompensated mentoring burdens—while also describing peer networks and self-advocacy that help sustain careers. 

The authors call on academic leadership to allocate resources, embed disability in DEI, standardize and streamline accommodations, raise expectations beyond ADA compliance, and include people with disabilities in decision-making to advance equity across the research workforce. 

Citation: Castro, F., Cerilli, C., Hu, L., Iezzoni, L. I., Varadaraj, V., & Swenor, B. K. (2024). Experiences of researchers with disabilities at academic institutions in the United States. PLOS ONE, 19(8), e0299612. https://doi.org/10.1371/journal.pone.0299612 

This commentary makes the case for keeping the American Community Survey six-question set (ACS-6) as the US federal standard for measuring disability, rather than replacing it with the Washington Group Short Set (WG-SS). Drawing on US policy context and empirical evidence, Landes, Swenor, Hall, and Forber-Pratt explain that the WG-SS is not a singular “international standard” and that multiple international tools exist. Studies from several countries also show the WG-SS can undercount disabled people and produce inconsistent estimates. 

In US federal surveys, the WG-SS (at its prescribed cut point) yields more than 50% lower disability prevalence than the ACS-6 and misses many people who are blind, deaf, or have severe mobility limitations. Because the ACS-6 is the established federal standard, replacing it with the WG-SS would circumvent US policy, undermine the purpose of US data collection, and result in a loss of data continuity that would make historical trend analyses of the disabled population impossible. 

The takeaway: Maintain the ACS-6 as the US standard and invest in improved, more inclusive measures—developed under the leadership of disabled researchers and in partnership with the US disability community, aligned with US legal and policy standards (e.g., ADA). Building on the ACS-6 while modernizing where needed will produce more accurate, comparable, and equitable disability data for policy and research 

Citation: Landes, S.D., Swenor, B.K., Hall, J.P., Forber-Pratt, A. In Measuring Disability, Do Not Abandon Established US Standards, Health Affairs Forefront. DOI: 10.1377/forefront.20250623.803072 

Using 2010–2018 National Health Interview Survey data, this brief report tests whether the Washington Group Short Set (WGSS) questions accurately identify people who are blind or deaf. The authors find substantial under-identification: the WGSS places 35.7% of blind adults and 43.7% of deaf adults in categories that would not count them as having a disability at the WGSS’s recommended cut point. Notably, 9.4% of blind and 16.0% of deafrespondents were captured as having “no difficulty.” 

The paper argues these results challenge the WGSS’s face validity for severe vision and hearing disability in U.S. surveys and recommends halting WGSS use in federal surveys until causes are understood and measures improved. Policy implications include support for efforts (e.g., the proposed Deafblind DATA Act) to develop more inclusive, standardized measures that better capture blindness, deafness, and disability status. 

Citation: Scott D Landes, Bonnielin K Swenor, Jean P Hall, Performance of the Washington Group questions in measuring blindness and deafness, Health Affairs Scholar, Volume 2, Issue 11, November 2024, qxae131, https://doi.org/10.1093/haschl/qxae131 

This brief outlines practices that states and territories can implement to strengthen disability data collection and use as a foundation for advancing health equity. With more than one in four U.S. adults having a disability, the authors emphasize that meaningful inclusion of people with disabilities in data collection and decision-making is essential to designing effective policies and programs. 

The brief highlights examples from Colorado, Oregon, Ohio, and Massachusetts, illustrating strategies such as legislative mandates for disability data collection, recognizing disability as a demographic and health disparity population, and promoting interdepartmental collaboration. 

Key recommendations include: 

• Standardizing disability definitions and data collection methods across federal, state, and territory programs. 
• Ensuring accessibility and inclusion in data collection through training, community partnerships, and collaboration with disability-led organizations. 
• Developing centralized data hubs to improve sharing and analysis of disability data. 
• Promoting transparency while protecting privacy, and involving people with disabilities in every stage of data use. 
• Establishing legislation and funding to sustain these efforts and build state-level capacity. 

By implementing these practices, states and territories can create more accurate, inclusive, and actionable data systems that inform equitable public health and policy decisions. 

Citation: Swenor, Bonnielin, Varshini Varadaraj, Franz Castro, Caroline Cerilli, and Grace Jo. “Promising Practices for State and Territory Disability Data Collection and Usage.” National Governors Association. December 9, 2024. 

Using NHATS 2021 data from 2,767 community-dwelling adults aged ≥71 years, this study estimates the share of dementia prevalence potentially attributable to objectively measured vision impairments. The population attributable fraction (PAF) from at least one impairment (near acuity, distance acuity, or contrast sensitivity) was 19.0% (95% CI, 8.2–29.7). By impairment type: contrast sensitivity 15.0% (6.6–23.6), near acuity 9.7% (2.6–17.0), distance acuity4.9% (0.1–9.9). Vision impairment prevalence was 32.2%. PAFs were highest among ages 71–79 (24.3%), female participants (26.8%), and non-Hispanic White participants (22.3%). 

While not establishing causality, the findings support considering vision health—including contrast sensitivity and visual acuity—in dementia prevention strategies aimed at people with disabilities. 

Citation: Smith, J. R., Huang, A. R., Zhou, Y., Varadaraj, V., Swenor, B. K., Whitson, H. E., Reed, N. S., Deal, J. A., & Ehrlich, J. R. (2024). Vision impairment and the population attributable fraction of dementia in older adults. JAMA Ophthalmology, 142(10), 900–908. Doi: https://doi.org/10.1001/jamaophthalmol.2024.313 

Using national data from the Behavioral Risk Factor Surveillance System (2019–2021), this study examined the relationship between vision impairment and barriers to healthcare access among U.S. adults. 

Among more than 1.2 million participants, 5% reported vision impairment. Adults with vision impairment were significantly more likely than those without to lack health insurance coverage, a primary care provider, and a recent health checkup, and nearly twice as likely to report being unable to afford care. In addition, racial, ethnic, and sexual minorities reported greater challenges accessing care across all measures. 

These findings reveal that adults with vision impairment face substantial barriers to healthcare access—compounded by intersecting inequities related to race, ethnicity, and sexual orientation. Addressing these disparities is essential to promoting equitable access to preventive and primary healthcare for all people. 

Citation: Varadaraj, V., Castro, F., & Swenor, B. K. (2023). Disparities in Healthcare Access for Adults with Self-Reported Vision Impairment in the U.S. Investigative Ophthalmology & Visual Science, 64(8), 4241. https://doi.org/10.1167/iovs.64.8.4241 

In this article, Dr. Swenor calls for a framework of disability data justice, which ensures disability is prioritized across all public health data systems.  

She outlines six key recommendations:  

1. collect disability as a core demographic, 
2. ensure comprehensive disability data across health, housing, transportation, education, employment, and more,  
3. adopt a growth mindset as disability evolves (e.g., long COVID),  
4. partner with the disability community in data design and collection,  
5. make public health data and websites accessible, and 
6. expand diversity in public health leadership to include disabled professionals. 

Dr. Swenor highlights that building inclusive data systems is essential to dismantle ableism, improve pandemic preparedness, and advance equity for everyone. 

Citation: Swenor, B.K. (2022). A Need for Disability Data Justice. Health Affairs Forefront. https://doi.org/10.1377/forefront.20220818.426231

Using 2019 BRFSS data (418,268 adults), this study estimates that 26.8% of non-institutionalized U.S. adults—about 67.2 million people—had any disability, and 11.7% (~29.6M) had multiple disabilities. The most common types were mobility (13.3%, ~34.2M) and cognitive/mental (12.1%, ~28.6M), followed by independent living (7.2%, ~17.6M), hearing (6.1%, ~16.2M), vision (5.2%, ~12.8M), and self-care (3.9%, ~9.8M). 

Adults with disability were more likely to be older, women, Hispanic, have lower education (less than high school), lower income, and be unemployed. Disability was also more common among people identifying as bisexual, transgender, and gender nonconforming. Intersectional differences appeared across race/ethnicity (e.g., higher prevalence among Black women compared with other groups). 

Why it matters: More than 1 in 4 U.S. adults report disability. Estimates likely undercount true prevalence (BRFSS excludes institutionalized adults; the ACS 6-question set captures mainly more severe impairments; telephone mode may miss some groups). Ongoing, accessible surveillance and disability-disaggregated reporting are essential to target policies and programs that reduce inequities across intersecting groups. 

Citation: Varadaraj, V., Deal, J. A., Campanile, J., Reed, N. S., & Swenor, B. K. (2021). National prevalence of disability and disability types among adults in the US, 2019. JAMA Network Open, 4(10), e2130358. https://doi.org/10.1001/jamanetworkopen.2021.30358