Simple blood tests can detect dementia in underrepresented Latin American populations

A recent study published in Nature Aging demonstrates that simple blood tests can accurately detect Alzheimer’s disease and other forms of dementia across highly diverse populations in Latin America. By combining these blood markers with standard memory tests and brain scans, clinicians can dramatically improve their ability to diagnose memory-robbing conditions in historically underrepresented regions.

Dementia is an umbrella term for a variety of conditions that cause a progressive loss of thinking and memory skills. Alzheimer’s disease is the most frequent cause of dementia in older adults. Frontotemporal lobar degeneration is another distinct type of dementia that primarily affects the front and sides of the brain, often leading to changes in personality and language.

Diagnosing these conditions traditionally requires expensive brain scans or invasive spinal taps. These traditional methods are often completely unavailable in lower-income regions. Medical professionals in these areas often have to rely solely on observing a patient’s behavior and administering basic memory quizzes.

Recently, researchers developed blood tests that look for specific proteins associated with these brain diseases. These proteins act as biological markers, indicating what is happening deep inside the brain tissue. The primary proteins measured are amyloid and tau, which tend to misfold and clump together in the brains of people with Alzheimer’s disease.

In healthy brains, a protein called amyloid beta normally floats around and is cleared away without issue. In Alzheimer’s disease, a specific version of this protein begins to stick together and form hard plaques. Researchers often measure the ratio of two different versions of this protein, as a lower ratio in the blood usually indicates that sticky plaques are accumulating in the brain.

The tau protein also plays a major role in brain health by acting as a structural support system for nerve cells. When the brain is diseased, tau proteins can gain extra chemical tags in a process called phosphorylation. The researchers specifically looked at two tagged versions of this protein to track the progression of the disease.

Another important marker is neurofilament light chain, a structural protein found inside nerve fibers. When nerve cells are damaged or die, this protein leaks out of the brain and into the bloodstream. Doctors can measure the concentration of this protein to gauge the overall level of active brain damage.

Medical professionals are already beginning to use these blood tests in the United States and Europe. Yet their performance in other parts of the world remained largely untested. Latin American populations possess vast genetic, social, and environmental diversity.

This immense variety can easily influence how diseases develop and how they appear in biological tests. A team led by first authors Ariel Caviedes and Felipe Cabral-Miranda wanted to know if these tests worked equally well in these understudied populations. Caviedes is a researcher at the Latin American Brain Health Institute in Chile, while Cabral-Miranda is based at the Federal University of Rio de Janeiro in Brazil.

The researchers recruited 605 volunteers from memory clinics across Argentina, Brazil, Chile, Colombia, Mexico, and Peru. This group included individuals diagnosed with Alzheimer’s disease, patients with frontotemporal lobar degeneration, and healthy older adults. The team collected blood samples from all participants and measured the concentrations of the target brain proteins.

Alongside the blood tests, the researchers administered a battery of standard memory and thinking assessments. A portion of the participants also underwent magnetic resonance imaging. This brain scanning technique allows doctors to see the physical structure of the brain and measure tissue shrinkage in specific areas.

The scientists also conducted genetic testing on a subset of the participants to determine their global ancestry. The vast majority of the participants had predominantly Amerindian genetic backgrounds. The team also looked for a specific variation of the apolipoprotein E gene, which is a known genetic risk factor for memory disorders.

To make sense of the vast amount of data, the research team used a technique called machine learning. Machine learning involves training a computer program to recognize complex patterns in massive datasets. The program learns which combinations of proteins and test scores best predict a specific medical diagnosis.

Using computer algorithms, the team analyzed the data to see if the blood proteins could accurately separate the sick patients from the healthy volunteers. They found that the blood tests alone performed quite well. The protein profiles correctly identified Alzheimer’s disease with 83 percent accuracy and frontotemporal lobar degeneration with 88 percent accuracy.

Specific proteins proved better at identifying certain diseases than others. A modified version of the tau protein was the strongest indicator of Alzheimer’s disease. In contrast, the neurofilament light chain protein was the most accurate marker for frontotemporal lobar degeneration.

The blood tests also aligned with the physical changes seen inside the patients’ heads. High levels of these disease-linked proteins matched up with visible tissue loss on the brain scans. The proteins also corresponded to lower scores on the memory and thinking evaluations.

The brain scans revealed clear anatomical differences between the two main patient groups. In people with Alzheimer’s disease, elevated blood proteins correlated with shrinkage in the back and sides of the brain. These areas are primarily responsible for storing new memories and processing visual information.

For patients with frontotemporal lobar degeneration, the protein markers told a different story. Elevated protein levels in these individuals corresponded to tissue loss in the frontal lobes and anterior temporal lobes. These specific brain regions govern executive functions, such as decision-making, social behavior, and language comprehension.

The memory and thinking assessments echoed these physical changes perfectly. In Alzheimer’s patients, higher levels of the modified tau protein were strongly linked to worse scores on memory tests. In contrast, high levels of the neurofilament light chain protein in the other patient group predicted severe declines in behavioral control and daily functioning.

The researchers achieved the highest success rates when they combined the blood tests with the brain scans and cognitive assessments. This combined approach raised the diagnostic accuracy to 90 percent for Alzheimer’s disease and 96 percent for frontotemporal lobar degeneration. The results show that relying entirely on a single blood test could lead to diagnostic errors.

“Combining biomarkers with cognitive and neuroimaging markers in diverse populations is essential to avoid misdiagnosis and to ensure fair access to care,” said Agustin Ibanez, co-senior author of the study. “Otherwise, even the most advanced tools risk reinforcing existing health inequalities.”

“These findings reinforce the enormous potential of blood-based tests to transform dementia diagnosis,” said Claudia Duran-Aniotz, senior author of the study. Integrating biological, cognitive, and physical measurements provides a much clearer picture of a patient’s health. The results suggest that these accessible screening tools can successfully identify neurodegenerative conditions across varied genetic backgrounds.

The study does have some limitations that require further investigation. The research team only evaluated the participants at a single point in time. This cross-sectional design prevents them from seeing how the protein levels might change as the diseases naturally progress over several years.

The researchers also noted that older adults often have multiple co-occurring health conditions, such as diabetes or heart disease. The team recorded the presence or absence of these conditions but did not analyze their individual impacts in deep detail. It remains possible that certain untreated medical issues could slightly alter the concentration of proteins in the blood.

Additionally, the researchers did not compare the blood tests against spinal fluid samples or physical brain tissue examined after death. These are considered the definitive standards for confirming the presence of these brain diseases. The lack of these gold-standard measurements means the researchers cannot completely rule out the presence of other overlapping brain conditions.

Future studies will need to track patients over longer periods and incorporate these stricter verification methods. Medical researchers must also continue to test these diagnostic tools in varied populations globally. Doing so will help ensure that modern medical advancements benefit people in all regions, regardless of their background or geographic location.

The study, “Blood-based AT(N) biomarkers for Alzheimer’s disease and frontotemporal lobar degeneration in Latin America,” was authored by Ariel Caviedes, Felipe Cabral-Miranda, Paulina Orellana, Hernán Hernández, Fernando Henríquez, Raul Gonzalez-Gomez, Matias Pizarro, Joaquin Migeot, Carolina Ochoa-Rosales, Carolina Gonzalez-Silva, Nickole Marin-Diaz, Carlos Coronel-Oliveros, Hernando Santamaría-García, Danilo Carmona, Adolfo M. García, Andrea Slachevsky, Andrew Singleton, Andy Yue Qi, Brian Lawlor, Bruce Miller, Catherine Dhooge, Caroline Pantazis, Chinedu T. Udeh-Momoh, David Aguillón, Diana L. Matallana, Eduardo R. Zimmer, Elisa de Paula França Resende, Francesca R. Farina, Francisca Cabello, Francisco Lopera, Henrik Zetterberg, José Alberto Avila-Funes, Juliana Acosta-Uribe, Katherine L. Possin, Kenneth S. Kosik, Kun Hu, Leonel T. Takada, Maira Okada de Oliveira, Marcelo Adrian Maito, Marc Suárez-Calvet, Maria E. Godoy, Maria Isabel Behrens, Mario A. Parra, Marta del Campo, Martin Bruno, Nancy Gelvez, Natalia Pozo-Castro, J. Nicholas Cochran, Nilton Custodio, Rodrigo Ortega, Rodrigo Santibanez, Rolando de la Cruz, Rosa Montesinos, Sarah McDonagh, Sara Bandres-Ciga, Shireen Javandel, Sonia M. D. Brucki, Stefanie D. Pina-Escudero, Victor Valcour, Ziyi Li, Jennifer S. Yokoyama, Agustin Ibañez, and Claudia Duran-Aniotz on behalf of the Multi-Partner Consortium to Expand Dementia Research in Latin America.