![Abnormal levels of the beta-amyloid protein clump together to form plaques (seen in brown) that collect between neurons and disrupt cell function. Abnormal collections of the tau protein accumulate and form tangles (seen in blue) within neurons, harming synaptic communication between nerve cells. [Photo credit: NIH]](https://sp-ao.shortpixel.ai/client/to_webp,q_glossy,ret_img,w_750,h_375/https://www.psypost.org/wp-content/uploads/2021/10/Beta-Amyloid-Plaques-and-Tau-in-the-Brain-750x375.jpg)
People who develop Alzheimer’s disease before age 65 tend to have more widespread buildup and faster spread of toxic tau protein in the brain than those who develop the condition later in life, according to a brain-imaging study published in Science Translational Medicine.
These differences persisted even after accounting for how cognitively impaired patients were at their initial scan, suggesting that early-onset Alzheimer’s involves a more aggressive form of tau accumulation. The researchers suggest the findings could have implications for how younger patients are treated, including the timing of amyloid-lowering drugs and the testing of therapies that target tau.
Symptom patterns and brain proteins in early- and late-onset Alzheimer’s
Alzheimer’s disease is a progressive neurodegenerative disorder in which brain cells gradually deteriorate and die, causing worsening problems with memory, thinking, and daily functioning. Clinicians classify the condition as early-onset Alzheimer’s when symptoms begin before age 65 and late-onset Alzheimer’s when symptoms appear after 65. Early-onset cases account for roughly 5 percent of all diagnoses, and only about 5 percent of those younger cases stem from known inherited genetic mutations, meaning most arise sporadically without a single known trigger.
Older patients typically experience trouble forming new memories as their main early symptom, whereas younger patients more often develop early difficulties with language, visual and spatial perception, or planning and decision-making, alongside faster cognitive decline. Inside the brain, Alzheimer’s is marked by two abnormal protein deposits: sticky plaques of amyloid-beta between nerve cells and twisted fibers, or tangles, of tau inside nerve cells.
Don’t miss the next big psychology finding
Every morning, PsyPost sends a short briefing on the most interesting new peer-reviewed research in psychology and neuroscience.
Researchers can map both proteins in living people using positron emission tomography, or PET, a scan in which a small amount of radioactive tracer is injected into the bloodstream and binds to target brain proteins so a scanner can record their three-dimensional distribution. Using this approach, an earlier brain-imaging study of living adults indicated that tau buildup in the medial temporal lobe, the inner part of the temporal lobe that is home to key memory structures, tracked with poorer episodic memory, or memory for personal events. In that earlier work, tau spread into broader regions accompanied amyloid plaques and worse overall thinking.
Autopsy studies also suggest that younger Alzheimer’s patients carry heavier amyloid and tau loads, whereas older patients more frequently harbor additional age-related conditions, such as blood vessel injury and clumps of other abnormal proteins.
Untangling delayed diagnosis from aggressive disease biology
Smaller PET studies had suggested greater tau buildup and faster tau increases over time in younger patients, yet people with early-onset Alzheimer’s are often diagnosed later in their illness because the condition is less suspected in middle age, leaving open whether heavier tau simply reflected more advanced disease at first evaluation. Konstantinos Chiotis, an assistant professor in the Department of Neurobiology, Care Sciences and Society at Karolinska Institutet and a cognitive neurologist at Karolinska University Hospital in Stockholm, led the study. He is also affiliated with the University of California San Francisco’s Memory and Aging Center. To examine at scale whether tau behaves more aggressively in younger brains even after accounting for clinical stage, Chiotis and colleagues across multiple institutions compared initial and follow-up scans from two large cohorts.
“As a neurologist, I wanted to understand the biological basis of differences we have long recognized clinically: younger patients with Alzheimer’s often have a broader range of symptoms and a faster disease course,” Chiotis told PsyPost. “It was unclear how much these differences reflected underlying disease biology and how much could be explained by delayed diagnosis or greater disease severity when patients were assessed. Tau PET allowed us to compare the amount, distribution and subsequent accumulation of tau in large early- and late-onset cohorts.”
Comparing brain scans across two large Alzheimer’s cohorts
The team analyzed 390 sporadic early-onset patients, with an average age of 59, from the Longitudinal Early-Onset Alzheimer’s Disease Study, or LEADS, which excludes carriers of known familial dementia mutations, alongside 211 late-onset patients, with an average age of nearly 78, from the Alzheimer’s Disease Neuroimaging Initiative, or ADNI. All 601 patients had amyloid plaques on PET and either mild cognitive impairment, an early stage where thinking declines noticeably without preventing independent living, or mild dementia. Researchers also examined 25 impaired patients under age 65 from the second cohort as an independent validation group and 547 cognitively unimpaired adults for comparison, using anonymized data accessible upon request through both study repositories.
Repeat tau PET scans using a tracer called flortaucipir were available for 245 early-onset patients over roughly two years and 99 late-onset patients over nearly three years, excluding scans taken after any participant began plaque-removing antibody therapy. Each scan was paired with magnetic resonance imaging, or MRI, which uses magnetic fields and radio waves to map brain anatomy across standardized regions of the cerebral cortex, the brain’s outer layer governing complex thought, perception, and language. Because nearly three-quarters of early-onset patients already had dementia at baseline compared with just over a third of late-onset patients, analyses controlled for baseline cognitive scores, overall amyloid burden, sex, and education, and checked results after correcting for brain shrinkage.
Widespread cortical tau appears early in younger patients
At their initial scan, early-onset patients carried substantially higher tau levels than late-onset patients across nearly all cortical regions, at both the mild cognitive impairment and dementia stages. This gap held after adjusting for cognitive severity and amyloid burden and after excluding the roughly 20 percent of early-onset patients with non-memory syndromes, and a version of the gap appeared in the 25-person validation subgroup, whose tau levels fell between those of the two main groups. While tau levels in the medial temporal lobe overlapped considerably between groups, early-onset patients showed widespread tau across frontal, parietal, and temporal cortices as early as mild cognitive impairment, whereas tau in late-onset patients remained largely confined to temporal and parietal regions even in dementia.
“Most younger participants still had memory problems, but around one in five had a presentation dominated by other difficulties, including language or visuospatial symptoms,” Chiotis noted. “The wider distribution of tau helps explain this clinical diversity and reinforces the importance of assessing cognitive abilities beyond memory.”
Viewed continuously across both impaired groups, younger age corresponded to heavier cortical tau, whereas older patients showed higher tracer signal only in deep brain structures where the tracer binds to non-tau targets during normal aging. Among cognitively unimpaired adults without amyloid plaques, older age instead went with higher medial temporal tau signal. Across impaired patients, women had modestly higher baseline tau than men, and carriers of the apolipoprotein E epsilon-4 genetic risk variant showed more widespread tau increases over time, yet neither sex nor genetic status modified the gap between early- and late-onset disease, as those interactions were not statistically significant.
Tau burden also tracked more steeply with both amyloid plaques and cognitive impairment in early-onset disease. Among younger patients, whole-cortex tau rose more steeply with amyloid levels than it did among older patients. Worse thinking and daily functioning scores also corresponded to widespread cortical tau in early-onset patients, compared with mainly temporal cortex tau in late-onset patients. Although early-onset patients entered the study with modestly higher amyloid plaque levels than late-onset patients, that difference was far smaller than the disparity in tau.
Faster yearly tau accumulation over multi-year follow-up
Across follow-up scans, tau rose faster per year in early-onset than in late-onset Alzheimer’s in every target region, independent of starting tau levels, amyloid load, and baseline cognitive scores. In late-onset patients, longitudinal tau increases centered on temporoparietal areas and were not statistically significant in the frontal lobe, whereas early-onset patients accumulated tau across broad cortical regions, including frontal and occipital areas typically affected late in the illness. Mathematical trajectory modeling estimated that once patients crossed the threshold for cortical tau positivity, late-onset patients took more than five additional years to reach tau levels comparable to those in early-onset disease.
“This is the largest tau PET study to date comparing early- and late-onset Alzheimer’s, and it substantially strengthens evidence from earlier, smaller studies,” Chiotis explained. “The longitudinal data extend that evidence by showing that tau accumulated more than twice as fast in younger patients during follow-up. These differences persisted after accounting for clinical severity and amyloid burden, supporting the view that age at symptom onset is associated with meaningful biological differences in Alzheimer’s disease. The findings reinforce the importance of earlier recognition in younger patients and of considering their underlying disease burden when designing treatment trials.”
“The increasingly divergent tau trajectories were particularly striking,” he observed. “Younger patients already had more tau at study entry and continued to accumulate it faster, making delayed diagnosis alone an unlikely explanation for the differences. I was also struck by how much more pronounced the differences were for tau than for amyloid, the other major protein involved in Alzheimer’s disease.”
These results are consistent with an earlier brain-imaging and biomarker study of healthy and cognitively impaired adults, which indicated that early tau buildup on PET scans predicted rapid declines in episodic memory better than amyloid plaques or spinal fluid markers. Whereas that earlier research compared different biomarker testing methods across cognitively unimpaired and impaired individuals, the current study specifically contrasted tau dynamics between younger- and older-onset patients with amyloid-confirmed cognitive impairment.
What the findings could mean for treatment
In the authors’ reading, symptoms in late-onset Alzheimer’s may outpace cortical tau spread because older brains frequently harbor coexisting vascular injury or other abnormal proteins that impair cognition alongside modest tau levels. In younger patients, by contrast, they propose that aggressive and potentially faster-seeding tau pathology is the main driver of decline.
Because clinical trials of anti-amyloid antibody drugs such as lecanemab and donanemab, which remove amyloid plaques from the brain, suggest greater benefit when treatment starts at lower tau levels, the researchers reason that early-onset patients may need earlier intervention to benefit equally from amyloid-lowering medications. They also argue that early-onset patients, often excluded from trials due to age cutoffs or non-memory symptoms, may be strong candidates for testing therapies targeting tau directly, alone or combined with anti-amyloid drugs.
“Our findings make earlier diagnosis particularly important in younger patients,” Chiotis emphasized. “They can already have extensive tau pathology when symptoms are still mild, and tau continues to accumulate rapidly. At the same time, diagnosis is often delayed because difficulties with planning, language or work performance are attributed to depression or burnout. During that delay, the disease can continue to advance biologically, making timely recognition important for access to care, support and appropriate treatment.”
Disclosures, study caveats, and group-average limits
As with all research, there are some caveats to consider. Several authors disclosed financial ties to imaging and pharmaceutical companies: one co-author is an inventor on patents assigned to Eli Lilly, whose subsidiary supplied the flortaucipir tracer, and holds stock or stock options in the company, and another founded and leads a brain-imaging firm. A third co-author is a full-time employee of the Alzheimer’s Association, which helped fund the early-onset cohort, and was among those who designed the study. The tracer suppliers did not fund the work and had no role in analyzing or interpreting the data.
“The more-than-twofold difference refers to tau accumulation measured by PET; it does not mean that every younger patient experiences cognitive decline twice as fast,” Chiotis cautioned. “These are group averages, and individual patients vary considerably. Finally, this study did not test treatments and cannot establish whether younger patients benefit more or less from amyloid-lowering drugs.”
The study also did not directly measure coexisting vascular damage or non-Alzheimer’s proteins, so future work combining tau PET with blood, fluid, and vascular markers will be needed to test how much mixed pathologies shape cognitive decline in older patients. Because the two age groups came mainly from separate cohorts, age-related and cohort-related differences cannot be fully separated, though the authors note that the validation subgroup and earlier studies point the same way. Late-onset status was assigned by age at the first scan, since symptom-onset age was not reliably recorded, a choice the authors expect would shrink rather than inflate group differences.
The flortaucipir tracer also has limited sensitivity for early-stage tau deposits. In addition, more than 80 percent of participants in both cohorts were non-Hispanic white adults who averaged about 16 years of schooling, leaving open how well these patterns generalize to broader populations.
Next steps in tracking early-onset Alzheimer’s biology
Follow-up work on these questions involves the study’s senior authors, Renaud La Joie and Gil Rabinovici of the University of California San Francisco.
“We want to understand why tau accumulates so aggressively in younger patients, including possible genetic and molecular explanations,” Chiotis said. “Together with Renaud La Joie and Gil Rabinovici, we are also examining treatment outcomes among LEADS participants receiving amyloid-lowering therapies. The longer-term goal is to combine brain imaging and fluid biomarkers to recognize disease earlier and guide treatment according to each patient’s underlying biology.”
He also highlighted the contributions of the study volunteers and public and nonprofit research sponsors. “I would also like to acknowledge the participants and families in LEADS and ADNI, and the sustained support from the US National Institute on Aging/NIH and the Alzheimer’s Association that made this work possible,” he added.
The study, “Distinct tau PET dynamics in early versus late age-of-onset Alzheimer’s disease,” was authored by Konstantinos Chiotis, Ganna Blazhenets, David N. Soleimani-Meigooni, Paul S. Aisen, Alinda Amuiri, Alireza Atri, Laurel Beckett, Michael Brickhouse, David G. Clark, Jeffrey L. Dage, Gregory S. Day, Ranjan Duara, Ani Eloyan, Tatiana Foroud, Neill R. Graff-Radford, Ian M. Grant, Dustin B. Hammers, Lawrence S. Honig, Erik C. B. Johnson, David T. Jones, Kala Kirby, Robert Koeppe, Joel H. Kramer, Walter A. Kukull, Julien Lagarde, Antoine Leuzy, Piyush Maiti, Joseph C. Masdeu, Mario F. Mendez, Erik Musiek, Kelly N. Nudelman, Chiadi U. Onyike, Meghan Riddle, Salma Rocha, Emily Rogalski, Stephen Salloway, Daniel R. Schonhaut, Sharon J. Sha, Ranjani Shankar, Alexander Taurone, Maryanne Thangarajah, Arthur W. Toga, Alexandra Touroutoglou, Raymond Scott Turner, Prashanthi Vemuri, Thomas S. Wingo, David A. Wolk, Kyle Womack, Jiaxiuxiu Zhang, Maria C. Carrillo, Bradford C. Dickerson, Liana G. Apostolova, Renaud La Joie, and Gil D. Rabinovici.
