Highlight
Non-Hispanic Black and Hispanic older adults with dementia were significantly less likely than non-Hispanic White adults to receive a timely clinical diagnosis in Medicare claims.
Educational attainment explained the largest share of the observed disparity, accounting for 48.4% of the gap for non-Hispanic Black individuals and 61.9% for Hispanic individuals.
Neighborhood affluence and local healthcare infrastructure, including the density of health practitioner and physician offices, also explained meaningful proportions of the disparity.
The findings suggest that improving dementia equity will require interventions that extend beyond specialist care and address structural differences in education, neighborhood resources, and access to diagnostic pathways.
Background
Dementia remains one of the most consequential syndromes in aging medicine, affecting cognition, function, caregiver burden, healthcare use, and long-term care needs. Timely diagnosis matters for several reasons. It creates an opportunity to identify reversible contributors to cognitive decline, initiate symptomatic treatments when appropriate, implement safety planning, support caregivers, align care with patient preferences, and connect patients to social and community resources. The importance of diagnosis has increased further with the arrival of disease-modifying therapies for selected patients with early Alzheimer disease, which depend on recognition of cognitive impairment before advanced decline.
Yet dementia has long been underdiagnosed in routine practice. Delayed recognition may be especially common in populations facing barriers related to healthcare access, educational disadvantage, language discordance, mistrust, structural racism, or geographic underinvestment in healthcare services. Racial and ethnic disparities in dementia diagnosis are therefore not simply documentation differences; they may reflect unequal opportunities to enter the diagnostic pathway at every step, from symptom recognition to primary care evaluation, specialist referral, biomarker workup, and claims-based coding.
Prior studies have shown that Black and Hispanic older adults experience both a higher burden of dementia and lower rates of formal diagnosis. What has remained less clear is which modifiable factors account for the disparity. In particular, the relative contribution of individual-level factors such as education versus neighborhood-level factors such as affluence and local healthcare supply is highly relevant to policy and practice. The study by Qian, Li, and Chen addresses this question using nationally representative, Medicare-linked survey data spanning more than two decades.
Study Design and Methods
Design and data sources
This was a retrospective observational study using 1998-2020 Medicare-linked data from the Health and Retirement Study, a major longitudinal survey of older U.S. adults. The investigators identified adults aged 65 years or older who were classified as having dementia in the Health and Retirement Study and linked those participants to Medicare claims.
Population
The analytic sample included 3,429 U.S. older adults with dementia as identified in the survey data. The central comparison was across racial and ethnic groups, specifically non-Hispanic Black, Hispanic, and non-Hispanic White individuals.
Outcome definition
The primary outcome was timely diagnosis of dementia. The investigators defined this as a clinical diagnosis of dementia recorded in Medicare claims within a window extending from 3 years before to 1 year after the month of the Health and Retirement Study interview at which dementia was first observed. Participants without a claim-based diagnosis in that interval were classified as having delayed or missed diagnosis.
This definition is clinically meaningful because it aims to capture whether the healthcare system identified dementia around the time it became apparent in structured survey-based assessment, while allowing for some lead and lag time in real-world detection.
Covariates and explanatory factors
The study first adjusted for non-modifiable demographics and health conditions, then examined the extent to which additional individual- and neighborhood-level factors explained the racial and ethnic gaps. Among the explanatory variables, educational attainment emerged as the dominant individual-level factor. Neighborhood-level measures included affluence and the density of healthcare offices, including health practitioner offices and physician offices. Dementia specialist evaluation was also examined as a contributor.
Key Results
Primary disparity in timely diagnosis
After adjustment for non-modifiable demographics and health conditions, non-Hispanic Black and Hispanic individuals remained less likely than non-Hispanic White individuals to receive a timely dementia diagnosis. The adjusted odds ratio was 0.71 for non-Hispanic Black individuals, with a 95% confidence interval of 0.57 to 0.89. For Hispanic individuals, the adjusted odds ratio was 0.61, with a 95% confidence interval of 0.44 to 0.85.
These effect sizes indicate clinically and statistically significant disparities. Put simply, the odds of having dementia recognized in a timely manner were about 29% lower for non-Hispanic Black individuals and about 39% lower for Hispanic individuals compared with non-Hispanic White individuals, even after accounting for baseline demographic and health differences.
What explained the disparity?
Educational attainment explained the largest proportion of the disparity in both groups. For non-Hispanic Black individuals, education explained 48.4% of the gap. For Hispanic individuals, it explained 61.9% of the gap. This is a striking result because it points to a life-course determinant with implications for health literacy, cognitive test performance, communication with clinicians, navigation of healthcare systems, and possibly clinician interpretation of cognitive symptoms.
Neighborhood affluence was the next most important contributor, explaining 27.3% of the disparity for non-Hispanic Black individuals and 17.8% for Hispanic individuals. This finding underscores that dementia diagnosis is not determined solely within the exam room. The socioeconomic characteristics of where a person lives likely shape access to consistent primary care, transportation, preventive services, caregiver support, specialist availability, and the administrative capacity of local practices.
Healthcare infrastructure also mattered. Density of health practitioner offices explained 15.8% of the disparity for non-Hispanic Black individuals and 14.6% for Hispanic individuals. Density of physician offices explained 9.6% and 12.4%, respectively. Dementia specialist evaluation explained a smaller but still measurable share, 6.6% for non-Hispanic Black individuals and 5.5% for Hispanic individuals.
These rankings are important. They suggest that disparities are not driven primarily by specialist access alone. Instead, more upstream and broader structural factors, especially education and neighborhood resources, appear to shape whether older adults ever reach diagnostic evaluation in a timely way.
Clinical interpretation of the decomposition findings
The study does not imply that education itself is a simple biologic cause of delayed diagnosis. Rather, education may serve as a marker of cumulative advantage or disadvantage across the lifespan. Lower educational attainment can influence baseline cognitive reserve, performance on screening instruments, expectations about aging, patient-clinician communication, trust, and health-system navigation. In some cases, cognitive symptoms may be normalized or misattributed. In others, standard testing may be less valid across educational and linguistic backgrounds, potentially delaying recognition.
Similarly, neighborhood affluence likely stands in for a broader ecosystem of opportunity: access to longitudinal primary care, stable insurance navigation, transportation, broadband access, caregiver flexibility, and local clinician workforce. The observed effect of healthcare office density further supports the view that the practical availability of entry points into care influences whether dementia gets diagnosed on time.
Why These Findings Matter Now
The timing of this study is highly relevant. Dementia diagnosis is moving from a primarily supportive label toward a gateway for increasingly specific management strategies. In addition to established nonpharmacologic and supportive interventions, emerging disease-modifying treatments for early Alzheimer disease require substantial diagnostic workup, including confirmation of clinical stage, discussion of benefit-risk tradeoffs, and often access to specialty evaluation and biomarker testing. If timely diagnosis remains inequitable at the front end, innovations in treatment may widen downstream disparities.
From a health-systems perspective, the study argues for a shift in focus. Expanding memory clinics and specialist consultations is useful, but insufficient on its own. The larger explanatory role of education and neighborhood affluence suggests that equitable dementia care starts earlier and broader: in public health, primary care, community outreach, culturally and linguistically appropriate assessment, and neighborhood-level investment in access.
Expert Commentary
This study adds granularity to a well-recognized inequity in dementia care. Its strength lies in combining a population-based survey signal of dementia with claims-based ascertainment of diagnosis, thereby addressing the gap between probable disease and formal recognition. The decomposition approach is especially useful because it identifies where the disparity may be concentrated.
Several mechanisms are plausible. First, cognitive screening and recognition are influenced by educational and cultural context. Lower educational attainment can alter performance on cognitive tests without necessarily mapping neatly onto disease severity, and clinicians may hesitate to code dementia if diagnostic certainty is low. Second, neighborhoods with fewer healthcare resources may reduce opportunities for repeated observation, collateral history, and specialist referral. Third, Hispanic older adults may face additional language barriers or variable access to culturally concordant care, which can delay both recognition and documentation.
The findings also fit with broader literature showing that social determinants strongly influence dementia risk, presentation, and care quality. They are aligned with the principle that diagnosis equity is not only a neurology issue, but also a primary care, geriatrics, and health policy issue.
Study limitations
As with any observational analysis, causality cannot be established. The explanatory proportions should not be interpreted as proof that changing one factor would eliminate the same proportion of disparity. Some variables may act as proxies for unmeasured constructs such as healthcare trust, language concordance, family advocacy, clinician bias, transportation barriers, or regional differences in coding practices.
The outcome relied on Medicare claims, which measure recorded diagnosis rather than necessarily clinician suspicion or full diagnostic evaluation. A person may have cognitive impairment recognized in practice without a claim coded as dementia in the specified interval. Conversely, coding practices may vary by setting and clinician specialty. The definition of timely diagnosis, while reasonable and clinically grounded, remains partly dependent on the timing of survey-observed dementia rather than symptom onset.
There is also potential heterogeneity within broad racial and ethnic categories, particularly within the Hispanic population, where country of origin, language preference, immigration history, and local healthcare context may influence diagnosis patterns. Finally, because the sample was restricted to older adults linked to Medicare and classified as having dementia in the Health and Retirement Study, generalizability to all community settings and insurance arrangements may be incomplete.
Implications for Clinical Practice and Policy
For clinicians
Primary care teams should recognize that delayed dementia diagnosis may reflect structural barriers rather than patient reluctance alone. Cognitive concerns in patients from historically underserved backgrounds may warrant proactive follow-up, use of validated and culturally appropriate cognitive assessment tools, interpreter services when needed, and collateral input from family or caregivers. Clinicians should also be careful not to dismiss early symptoms as normal aging, particularly when educational background complicates interpretation of brief screening tests.
For health systems
Systems aiming to improve dementia equity should invest in front-line detection pathways. Practical steps include embedding cognitive assessment protocols in primary care, expanding community health worker and care navigator programs, improving referral pathways to geriatrics and neurology, and targeting outreach to neighborhoods with lower affluence and lower healthcare office density. Telehealth may help some patients, but it cannot substitute fully for local diagnostic infrastructure, particularly in older populations with digital barriers.
For policymakers
The results argue for policies that treat dementia diagnosis as part of a broader equity agenda. Payment and quality programs could incentivize equitable cognitive assessment and follow-up. Geographic maldistribution of outpatient care resources deserves attention, as do transportation, interpreter access, and caregiver support services. Over the longer term, the dominant role of education in explaining disparities reminds us that dementia care inequities are partly rooted in social disadvantage accumulated across decades.
Funding and Registration
The abstract provided does not report funding information or a ClinicalTrials.gov registration number. This is expected for many retrospective observational analyses using existing survey and claims data, which are generally not prospectively registered as clinical trials. Readers should consult the full Journal of General Internal Medicine article for complete disclosures, author conflicts of interest, and funding details.
Conclusion
In this national study of older U.S. adults with dementia, non-Hispanic Black and Hispanic individuals were substantially less likely than non-Hispanic White individuals to receive a timely clinical diagnosis. Educational attainment explained the largest share of the disparity, followed by neighborhood affluence and local healthcare office density, with dementia specialist evaluation contributing a smaller portion.
The central message is clear: inequitable dementia diagnosis is driven less by a single point failure and more by layered differences in social and healthcare opportunity. Timely dementia detection will not be improved equitably through specialty expansion alone. It will require earlier, broader, and more structurally informed strategies that strengthen primary care recognition, improve neighborhood access to healthcare, and address educational and socioeconomic disadvantage across the life course.
As dementia care becomes more diagnostically and therapeutically complex, ensuring that timely diagnosis is distributed fairly is no longer optional. It is a prerequisite for equitable access to modern dementia care.
References
1. Qian Y, Li F, Chen X. Associations of Individual- and Neighborhood-Level Factors with Racial and Ethnic Disparities in Timely Dementia Diagnosis. Journal of General Internal Medicine. 2026-05-01. PMID: 42067745. Available at: https://pubmed.ncbi.nlm.nih.gov/42067745/
2. Alzheimer’s Association. 2024 Alzheimer’s disease facts and figures. Alzheimers Dement. 2024;20(5):3708-3821.
3. US Preventive Services Task Force. Screening for Cognitive Impairment in Older Adults: US Preventive Services Task Force Recommendation Statement. JAMA. 2020;323(8):757-763.
4. Bradford A, Kunik ME, Schulz P, Williams SP, Singh H. Missed and delayed diagnosis of dementia in primary care: prevalence and contributing factors. Alzheimer Dis Assoc Disord. 2009;23(4):306-314.
5. Lin PJ, Daly AT, Olchanski N, et al. Dementia diagnosis disparities by race and ethnicity among Medicare beneficiaries. Med Care. 2021;59(8):679-686.