Separate the diseases. Separate pathology from function.
This research program tracks what is known, what failed, what remains uncertain, and which experiments could most reduce uncertainty about prevention, disease modification, resilience, and functional recovery.
Independent, AI-assisted evidence synthesis for scientific review. Not independently peer reviewed.
Research is organized by disease and stage rather than treating dementia as one biological condition. Mixed pathology is common enough that a single diagnostic label may not explain every symptom or treatment response. Public summaries therefore preserve subtype, stage, population, and outcome context.
What counts as meaningful progress
Earlier detection and better biomarkers can improve diagnosis and research, but they are not interchangeable with preventing decline, restoring function, or producing durable clinical benefit. The program distinguishes risk prediction, pathology change, cognitive outcomes, functional outcomes, adverse effects, and durability.
Prevention, interception, and resilience
The research agenda includes risk reduction, preclinical interception, vascular and metabolic contributors, immune and glial biology, proteostasis, mitochondrial quality, synaptic and network function, and why some people remain functionally resilient despite pathology. Each line is treated as a question to test rather than a treatment recommendation.
Failed and negative evidence matters
Trials or mechanisms that do not produce the expected benefit are preserved because they narrow the search space. A useful research record should state the strongest evidence for an idea, the strongest evidence against it, major confounders, safety or feasibility limits, and what result would justify narrowing or abandoning the idea.
Public-safe sources
For general background on Alzheimer’s disease and related dementias, diagnosis, biomarkers, and ongoing research, see the U.S. National Institute on Aging.
Historical research: a rare Reelin variant and resilience in inherited Alzheimer's disease
Research update — AI-assisted summary of research by the cited authors.
In May 2023, Lopera and colleagues described a man with an inherited Alzheimer's mutation whose cognitive impairment appeared much later than expected. Despite extensive amyloid plaques, tau pathology was limited in his entorhinal cortex. He carried a rare Reelin variant. Laboratory work linked that variant to stronger Reelin signaling and reduced tau phosphorylation in a mouse model. These observations suggest a research direction; they do not demonstrate that changing Reelin signaling prevents dementia in people.
Limits of the evidence
A rare individual case and laboratory findings cannot establish general protection or treatment benefit. Authors disclosed biotechnology company, consulting and advisory interests. Based on published abstracts; details beyond those abstracts have not been checked.
This is research information, not personal medical advice.
Historical research: US POINTER compared two approaches to cognitive health
Research update — AI-assisted summary of research by the cited authors.
The US POINTER randomized trial compared structured and self-guided lifestyle programs in older adults at increased risk of cognitive decline. Both groups improved on a composite cognitive test measure during follow-up, with a statistically greater improvement in the structured group. The study compared two active programs; it did not compare a program with receiving no intervention.
Limits of the evidence
The reported outcome was cognitive-test performance over two years, not proof that dementia was prevented. The authors call for further investigation of functional outcomes and longer follow-up to establish clinical relevance and durability. This historical research summary is not an individual prevention or treatment recommendation. The Alzheimer's Association funded the study and participated in its design, conduct and reporting. Based on published abstracts; details beyond those abstracts have not been checked.
This is research information, not personal medical advice.
Historical research: PLCG2 variants and microglia in Alzheimer's disease
Research update — AI-assisted summary of research by the cited authors.
A 2023 study investigated PLCG2, a protein involved in immune signaling in the brain's microglia. Researchers identified the M28L variant as a potential Alzheimer's risk variant and compared it with the previously identified protective P522R variant. In an amyloid-producing mouse model, the variants had opposite effects: P522R lessened disease-related changes, while M28L increased plaque burden. The study also linked the variants to different microglial gene-expression patterns and changes in uptake and clearance of amyloid.
Limits of the evidence
These findings combine genetic associations with laboratory and mouse-model experiments. They do not establish a treatment benefit in people or show that changing PLCG2 activity would safely prevent Alzheimer's disease. This is an account of a historical study, not a new clinical result. Based on published abstracts; details beyond those abstracts have not been checked.
This is research information, not personal medical advice.
Historical research: genetic patterns in cognitively healthy centenarians
Research update — AI-assisted summary of research by the cited authors.
Tesi and colleagues compared Alzheimer's-associated genetic variants in cognitively healthy centenarians, people with Alzheimer's disease, and a control group. Variants previously associated with lower Alzheimer's risk were more common in the cognitively healthy centenarians. The authors highlighted patterns involving genes associated with immune function and the cellular systems that process and recycle material.
Limits of the evidence
This was a genetic association study, not a treatment trial. It does not establish that changing a particular gene prevents Alzheimer's disease, nor does it provide an individual diagnosis or guarantee of protection. Findings in a selected group reaching extreme old age should not be assumed to apply to everyone. Based on published abstracts; details beyond those abstracts have not been checked.
This is research information, not personal medical advice.
Historical research: a brain-delivery formulation studied in rats
Research update — AI-assisted summary of research by the cited authors.
In a 2019 laboratory study, Jojo and colleagues developed lipid-based nanoparticles carrying pioglitazone for delivery through the nose. They reported sustained drug release in laboratory tests, improved passage through nasal tissue ex vivo, and transport to the brain in rats. The work studied formulation and drug distribution; it did not demonstrate that this approach treats Alzheimer's disease in people.
Limits of the evidence
This is preclinical formulation research, not a human Alzheimer's treatment trial. Drug transport alone does not establish clinical benefit or safety. The paper concerns pioglitazone, not insulin. This historical study is not presented as a new discovery or as instructions for treatment. Based on published abstracts; details beyond those abstracts have not been checked.
Research review: PLCG2 and immune-cell resilience in Alzheimer's disease
Research update — AI-assisted summary of research by the cited authors.
A 2026 review by Tsai and colleagues examines PLCG2, a signaling protein involved in the brain's immune cells, called microglia. The authors discuss a rare genetic variant associated with lower Alzheimer's risk and evidence from cell and animal models. They describe how microglia can support tissue repair or contribute to damaging inflammation, and discuss early laboratory efforts to find molecules that activate PLCG2.
Limits of the evidence
This is a review of existing research, not a new clinical trial. A protective genetic association does not establish that a drug acting on the same protein prevents or treats Alzheimer's disease. Laboratory findings do not establish benefit or safety in patients. This summary uses the published abstract and metadata; it does not independently assess every cited study or provide treatment advice. Based on published abstracts; details beyond those abstracts have not been checked.
This is research information, not personal medical advice.
Research archive: brain-tissue signals after an early Alzheimer's vaccine study (2018)
Research update — AI-assisted summary of research by the cited authors.
A 2018 study by Paquet and colleagues examined post-mortem brain tissue from people with Alzheimer's disease who had received AN1792, an experimental amyloid-targeting immunization, and from people who had not. Several proteins involved in programmed cell death were less abundant in the previously immunized group. The authors interpreted this as reduced apoptosis-related signaling in remaining neurons and other cells. They also noted greater neuronal loss following immunization, so lower protein signals should not be read as proof that treatment preserved neurons.
Limits of the evidence
This was a small observational comparison of brain tissue after death, not a new randomized trial demonstrating clinical benefit. The results do not establish improved memory, daily functioning or safety, and they cannot be generalized to current anti-amyloid treatments. This historical summary covers the published abstract and bibliographic metadata; it is not treatment advice. Based on published abstracts; details beyond those abstracts have not been checked.
This is research information, not personal medical advice.
Research archive: verubecestat did not demonstrate clinical benefit in prodromal Alzheimer's disease
Research update — AI-assisted summary of research by the cited authors.
Investigators tested verubecestat, an inhibitor of amyloid production, in participants with memory problems and brain amyloid evidence who did not meet dementia criteria. The randomized, placebo-controlled trial was stopped for futility. Verubecestat did not improve the primary clinical rating, and some assessments suggested worse cognition and daily function. Adverse events were more common with the drug.
Limits of the evidence
This trial studied one drug in a particular prodromal population. It does not establish that all amyloid-directed treatments share the same effects. This summary covers the published abstract and bibliographic metadata, not a full-report assessment or treatment recommendation. Based on published abstracts; details beyond those abstracts have not been checked.
This is research information, not personal medical advice.
Research archive: intranasal insulin showed no cognitive or functional benefit in the primary trial cohort
Research update — AI-assisted summary of research by the cited authors.
Researchers compared intranasal insulin with placebo in people with amnestic mild cognitive impairment or Alzheimer's disease dementia. An initial delivery device was replaced after reliability problems; participants using the replacement device formed the primary analysis cohort. That analysis found no observed cognitive or functional benefit and no difference in the measured cerebrospinal-fluid outcomes.
Limits of the evidence
These findings concern the trial's primary cohort and delivery setup. Device reliability and the change in the primary cohort are important interpretation limits. This archive summary covers the published abstract and bibliographic metadata, not a full-report assessment or treatment recommendation. Based on published abstracts; details beyond those abstracts have not been checked.
This is research information, not personal medical advice.
Trial analysis examines age and blood-pressure treatment effects on brain imaging
Research update — AI-assisted summary of research by the cited authors.
Researchers analysed brain-imaging measures of small vessel disease in participants from the SPRINT and ACCORD blood-pressure trials. The abstract reports a pattern suggesting greater reductions in white matter injury with intensive blood-pressure treatment at younger ages. In SPRINT, however, the overall age-by-treatment interactions were not statistically significant; ACCORD reported a difference between its younger and older groups.
Limits of the evidence
These analyses concern imaging markers, not direct evidence that treatment prevents Alzheimer's disease or dementia. Age-related patterns differed by trial and outcome, so they should not be treated as a settled age-specific treatment rule. This summary covers the published abstract and bibliographic metadata only, not a full-report review. It gives no blood-pressure targets or treatment advice. Based on published abstracts; details beyond those abstracts have not been checked.
Research update — AI-assisted summary of research by the cited authors.
Secondary analyses of trials in early symptomatic Alzheimer's disease examined ARIA, imaging-detected abnormalities that can involve brain swelling or bleeding-related findings. These abnormalities were more frequent among participants receiving donanemab than among those receiving placebo in the controlled trials. The researchers also examined associations between ARIA and participants' characteristics and brain-imaging findings.
Limits of the evidence
The report combines prespecified and exploratory analyses of randomized trials with an uncontrolled open-label addendum. The reported risk-factor associations do not establish that changing a particular factor prevents ARIA. This summary covers the published abstract and bibliographic metadata, not a full-report review or an assessment of an individual's treatment benefits and risks. Based on published abstracts; details beyond those abstracts have not been checked.
Author disclosures
The PubMed record reports that several authors were employees and shareholders of Eli Lilly and Company during the study.
This is research information, not personal medical advice.
Brain-tissue case study describes findings after verubecestat treatment
Research update — AI-assisted summary of research by the cited authors.
Researchers examined brain tissue from one person with Alzheimer's disease who had received verubecestat, a BACE-1 inhibitor. Compared with untreated sporadic Alzheimer's disease cases of similar age, the treated case had smaller plaques, less abnormality in nerve processes around plaques, and less synapse-associated amyloid-beta. These are tissue observations from one treated case.
Limits of the evidence
One treated case and comparison tissue cannot establish that the drug caused the differences or improved cognition, daily functioning or survival. This summary covers the published abstract and bibliographic metadata, not the full methods. These findings do not establish the treatment's effectiveness or safety. Based on published abstracts; details beyond those abstracts have not been checked.
Citation correction (November 5, 2026). The journal issue was published in 2022. The paper first appeared online in 2021.
This is research information, not personal medical advice.
APOE3 Christchurch mouse study finds less tau buildup around plaques
Research update — AI-assisted summary of research by the cited authors.
In a Cell study, researchers engineered mice to carry a human version of the APOE3 Christchurch variant and studied them in a model that forms amyloid plaques. After introducing tau from Alzheimer's brain tissue, they observed less tau pathology near the plaques and a stronger response by microglia, immune cells in the brain, around those plaques. In related cell experiments, the researchers reported increased uptake and breakdown of tau aggregates. They also observed changes in blood lipids in the mice.
Limits of the evidence
The paper examined experimental models. It did not test whether a treatment improves memory or daily function in people. The human case that motivated the study does not establish that the variant alone caused protection. The model results do not establish human benefit or safety. One author disclosed being an inventor on a university-licensed anti-tau antibody patent. Based on published abstracts; details beyond those abstracts have not been checked.
Citation correction (November 5, 2026). The journal issue was published in 2024. The paper first appeared online in 2023.
This is research information, not personal medical advice.
What one Christchurch case reported about amyloid and tau in Alzheimer’s
Research update — AI-assisted summary of research by the cited authors.
A 2019 case report described a person with an inherited form of Alzheimer’s disease who developed mild cognitive impairment much later than was typical in that family. The person carried two copies of a rare version of APOE called Christchurch. Brain scans showed substantial amyloid buildup and relatively limited tau signal. The authors reported that the tau signal was concentrated in certain brain regions, with other regions usually affected in clinical Alzheimer’s relatively spared.
Limits of the evidence
This is a report about one person. It cannot establish that the Christchurch variant caused the later onset, estimate what happens in other carriers, or show that a treatment works. Some authors reported industry relationships in the paper’s competing interests statement.
This is research information, not personal medical advice.
Childhood experiences and later thinking in one Alzheimer’s cohort
Research update — AI-assisted summary of research by the cited authors.
Researchers asked whether childhood experiences that offered more opportunities to learn were linked to how people’s thinking changed as they aged. In this study, people who reported more of those opportunities tended to decline more slowly on thinking tests and had fewer Alzheimer’s-related brain changes at autopsy.
Technical details. This observational analysis used 813 autopsied participants in the Rush Memory and Aging Project. Participants reported childhood experiences at study entry and received annual cognitive assessments. The analysis linked a composite measure of early-life experiences to a lower overall Alzheimer’s pathology score and slower cognitive decline. The authors noted that earlier studies had reported mixed results on changes in thinking over time.
Limits of the evidence
Participants recalled childhood experiences many years later. The selected autopsy sample and incomplete measurement of childhood circumstances limit how broadly the results apply. This study reports associations and cannot establish that a specific activity prevents dementia. This article summarizes published research for educational purposes.
This is research information, not personal medical advice.
Starting scores and later decline are different measures
Research update — AI-assisted summary of research by the cited authors.
Two studies of aging asked related questions about learning experiences and later thinking. One looked at years of formal education. The other combined four kinds of childhood experience and resources. The studies measured different things, which matters when interpreting their results.
Years of education. Wilson and colleagues analyzed annual cognitive tests in 2,899 older adults from the Religious Orders Study and Rush Memory and Aging Project. More years of education were associated with a higher starting score on the overall thinking measure. Across the full group, years of education were unrelated to the rate of change on that measure. Among participants who died during follow-up, more schooling was associated with an earlier start to the period of faster decline before death. Years of schooling were unrelated to the rate of decline during that faster period.
A broader childhood measure. Oveisgharan and colleagues analyzed 813 autopsied Rush Memory and Aging Project participants. Their measure combined childhood socioeconomic circumstances, access to learning resources at age 12, participation in mentally stimulating activities, and foreign-language instruction. A higher combined score was associated with slower cognitive decline and a lower overall Alzheimer’s pathology score at autopsy.
Limits of the evidence
The first analysis used years of schooling. The second used a broader measure and a selected group with autopsy data. Both draw on the Rush Memory and Aging Project, so the findings should not be treated as independent replication. The published descriptions do not establish exactly how many people appeared in both analyses. These observational studies do not establish that a specific experience prevents dementia.
This is research information, not personal medical advice.
POINTER imaging substudy found no difference on four main brain-scan measures
Research update — AI-assisted summary of research by the cited authors.
Researchers used brain scans to compare two lifestyle programs in a subgroup of the U.S. POINTER trial. Over two years, they found no detectable difference between the groups on the four main scan measures. The scan results answer a different question from the cognitive test results reported in the larger trial.
Of 1,943 parent-trial participants assessed for imaging eligibility, 983 completed at least one MRI or PET scan. The four primary imaging outcomes were global amyloid burden, entorhinal tau burden, hippocampal volume, and white matter hyperintensity volume. The structured and self-guided groups did not differ significantly on these outcomes or on longitudinal cognition within the imaging subgroup.
Limits of the evidence
Both groups received an active program. Imaging participants were a selected subset, and completing one scan did not imply complete longitudinal data. The abstract reports two exploratory findings about differences in cognitive outcomes. Both need confirmation in other studies. The results neither establish effects on dementia diagnoses nor rule out changes the study did not measure. Based on published abstracts; details beyond those abstracts have not been checked.
This is research information, not personal medical advice.
Small trial left the effects of a balance program on falls and balance uncertain
Research update — AI-assisted summary of research by the cited authors.
Researchers tested a balance-focused exercise program in older adults with mild cognitive impairment. They wanted to know whether the program improved balance or reduced falls. The study did not find clear differences on those main measures. It enrolled too few people to determine reliably whether the program helped.
The single-blind randomized trial enrolled 128 participants before withdrawals and reported data for 126. The intervention lasted six months, followed by six months of maintenance. Outcomes were assessed at six and twelve months. The abstract reports no significant between-group differences in either primary outcome, balance or falls, at either assessment.
Limits of the evidence
The study did not reach its planned sample size for either primary outcome. An exploratory analysis of people who adhered to the program cannot establish a randomized treatment effect. This report does not establish effects on dementia diagnoses. Based on published abstracts; details beyond those abstracts have not been checked.
This is research information, not personal medical advice.
A POINTER MRI analysis reported a difference on one measure by age group
Research update — AI-assisted summary of research by the cited authors.
Researchers compared brain scans from older adults assigned to a structured or a self-guided lifestyle program in U.S. POINTER. One measure, called free water, increased more slowly in the structured group among participants younger than 70. The analysis did not find a clear difference in that measure among participants aged 70 or older. This was a measure from brain scans, and the study did not show that either program prevented dementia.
The secondary analysis included 959 participants with evaluable MRI data. For the prespecified age groups, the interaction between follow-up time, program and age group was 0.050 (standard error 0.018, p=0.006). The modeled comparison within the younger group was −0.031 (standard error 0.012, p=0.009). When the researchers modeled age as a continuous measure, the interaction was weaker (p=0.06). The study found no clear difference between programs on the other imaging measures it tested, including the occurrence of new small brain bleeds.
Limits of the evidence
Both groups received active programs. The analysis was limited to trial participants with evaluable imaging, and the age-specific comparisons within the model were post hoc. The results do not establish an age threshold for treatment benefit or explain the parent trial’s cognitive results. The analysis did not measure whether the programs reduced dementia diagnoses.