Alzheimer's brain markers tell different stories depending on race and ethnicity

Most of what we know about the earliest brain changes in Alzheimer's disease comes from studying white populations. That is not a critique of individual researchers - it is a structural feature of how Alzheimer's research has been funded, recruited, and designed for decades. A new study from USC makes the case that this uniformity has left significant blind spots.

The diversity problem in Alzheimer's diagnostics

Alzheimer's disease is defined by two proteins: amyloid beta, which forms plaques between brain cells, and tau, which forms tangles inside them. The conventional model treats these biomarkers as universal signals. Amyloid accumulates first, tau follows, memory declines, and the progression is presumed to look roughly the same regardless of who the patient is.

The Health and Aging Brain Study-Health Disparities (HABS-HD), one of the largest and most racially diverse brain-imaging studies of aging in the United States, tested that assumption. Using advanced PET brain scans capable of detecting abnormal protein buildup years before symptoms appear, researchers examined more than 1,500 cognitively normal older adults and those with mild cognitive impairment across Black, Hispanic, and non-Hispanic white populations.

Higher tau, different patterns

The findings challenge the one-size-fits-all model. Black and Hispanic participants showed higher levels of tau in the medial temporal lobe - a brain region critical for memory and considered an early warning site for Alzheimer's progression - compared to non-Hispanic white participants. This elevated tau appeared even in the absence of significant amyloid plaque buildup, which is noteworthy because the standard model positions amyloid as the trigger for tau accumulation.

The relationship between these brain proteins and actual memory performance also varied across groups. Higher tau levels were generally linked to worse memory across all participants. But amyloid buildup strengthened this connection only in non-Hispanic white and Hispanic participants, not in Black participants.

That divergence matters clinically. If tau and amyloid interact differently with memory across populations, diagnostic thresholds and risk assessments calibrated primarily on white cohorts may misidentify risk in other groups.

Beyond proteins: what else drives decline

The study raises a critical question: if amyloid and tau do not fully explain memory changes in Black older adults, what does? The researchers point to several candidates. Vascular health - blood pressure, cerebrovascular disease, stroke history - is more prevalent in Black communities and is independently associated with cognitive decline. Chronic stress exposure, including the cumulative physiological burden of racism and discrimination, may contribute through pathways that standard Alzheimer's biomarkers do not capture.

Co-morbid health conditions, differences in educational access and quality, and socioeconomic factors may all interact with biological aging in ways that vary across populations. Alzheimer's is almost certainly not one disease with one pathway, and the biomarker framework may need to accommodate that complexity.

A tracer with its own limitations

The study also flagged a technical concern. Some of the observed differences in tau levels may reflect limitations of the imaging tracer itself rather than true biological variation. PET tracers designed to bind tau can occasionally produce signals in nearby brain regions unrelated to actual tau accumulation - a phenomenon called off-target binding. If this occurs at different rates across populations, it could confound comparisons.

This is not a flaw unique to this study. It is an industry-wide challenge. Imaging tools developed and validated predominantly in white populations may not perform identically in diverse cohorts. Careful validation across populations is essential before these tools can be used equitably in clinical settings.

Where the field goes from here

Future studies from the HABS-HD cohort will follow participants over time, tracking how tau, amyloid, vascular health, genetics, and social determinants interact to influence cognitive aging across communities. The goal is not simply to document differences but to understand the biological pathways that produce them, enabling diagnostic and treatment strategies that work for everyone.

The study does not overturn the amyloid-tau model of Alzheimer's disease. But it demonstrates that the model is incomplete, and that its gaps fall disproportionately along lines of race and ethnicity. Precision medicine for Alzheimer's cannot be precise if it is calibrated to only part of the population.