A review of medical records and scientific images has raised questions about studies of the APOE3 Christchurch and Reelin-COLBOS genetic variants. However, independent studies in mice and cells still suggest a possible biological effect—especially when two copies of Christchurch are present.
For nearly a decade, a few extraordinary carriers of a hereditary Alzheimer's-causing mutation have raised hopes that a natural way to protect the brain has been found. Despite carrying a mutation that should cause cognitive decline as early as their forties, they have remained healthy for decades beyond what was expected.
The explanation offered by the researchers was the existence of rare genetic variants—primarily apoe3 Christchurch andReelin-COLBOS — which may inhibit the spread of the protein Tao and protect nerve cells. The discoveries have already led to the development of antibodies and experimental gene therapies.
But thatResearch published in the journal Science On August 13, 2026, the investigation raises serious questions about some of the evidence supporting the promise. The investigation does not establish scientific fraud or prove that the genetic variants have no effect, but it does point to controversial changes in patient data and suspicious imagery in several articles.
The family that became key in Alzheimer's research
The story has its roots in the Antioquia region of Colombia. An extended family of about 6,000 people lives in the area, about a fifth of whom carry the E280A mutation in the gene. PSEN1The mutation, also known as “paisa,” causesAlzheimer's Autosomal dominant: Mild cognitive impairment usually appears in carriers around age 44, anddementia Develops about five years later.
Colombian neurologist Francisco Lopera and his team at the Antioquia Neuroscience Group (GNA) have been following the family for decades. The study has given scientists a rare opportunity to examine the stages of the disease before symptoms appear.
In 2019, an international team reported on Elyria Rosa Piedraita de Viegas. She carried the Paisa mutation but did not develop cognitive impairment until she was in her seventies—about three decades later than expected. Researchers found that she carried two copies of a rare variant of the APOE3 gene, called Christchurch.
Brain imaging showed a very large amount of Amyloid-beta, but relatively limited levels of tau and neurodegeneration. The researchers suggested that Christchurch does not necessarily prevent amyloid buildup, but rather interferes with the stages in which tau spreads in the brain and damages nerve cells. The case was published inNature Medicine.
Second case and the Reelin-COLBOS version
In 2023, the same collaboration published another case. A man carrying the Paisa mutation remained cognitively unaffected until the age of 67. This time, he was not found to have Christchurch, but rather a rare variant in the RELN gene, which encodes the protein Reelin.
The researchers called it Reelin-COLBOS and claimed that it might boost signaling that protects the entorhinal cortex, an area of the brain that plays a key role in memory and is damaged in the early stages of Alzheimer’s. The man’s sister also carried the variant and developed symptoms at age 58.
The article, which was also published inNature Medicine, emphasized that this was a very small number of cases and that the results were not conclusive. Despite the reservations, the variant was quickly presented as a candidate for treatment development.
Is a single copy of Christchurch also protective?
The key question was whether a person needed two copies of Christchurch, as Elyria had, or whether one copy would provide partial protection.
In 2024, researchers reported inNew England Journal of Medicine In 27 Christchurch carriers out of 1,077 family members who carried the Paisa mutation, the analysis found that a single copy of Christchurch delayed the onset of cognitive impairment and dementia by a median of about five years.
But other researchers have noted that an earlier analysis of the same data did not find a clear protective effect. In a letter published in the journal, Nicholas Cochrane, Alison Gotta and Michael Gracius argued that the study's result was affected by retrospective changes in the age at which the disease onset was determined in some of the carriers.
According to the Science study, Christchurch carriers were 20 times more likely than non-carriers to have their age of onset of symptoms moved up. In one case, the age of onset of cognitive impairment was moved up from 42 to 52, even though the medical record had already documented impairment at age 42. In several other cases, the date of dementia diagnosis was moved up by several years.
The researchers responded withLetter published in NEJM The reassessment was done systematically by a multidisciplinary team that was unaware of the subjects' genetic status. However, sources interviewed for the investigation claimed that at least two of the study leaders knew who the carriers of the variant were, and that the review committee did not recommend some of the changes.
The controversy does not prove that the data was deliberately altered, but it weakens confidence in the claim that a single copy of Christchurch is enough to inhibit the disease.
Suspected duplications in scientific images
Science asked three experts in scientific image analysis—Matthew Schrag, Kevin Patrick, and Elizabeth Bick—to review 12 articles about Christchurch and Reelin-COLBOS.
In three of the papers, they identified what appeared to be duplications or reuse of parts of images that supported the study's hypotheses. Among other things, photographs of brain organoids and mouse brain sections were examined. In some of the images, samples that were presented as coming from different organoids or animals appeared to experts to be parts of the same sample.
Image duplication can be due to an editing error, mislabeling, or a side effect of digital processing. It does not in itself prove forgery. To determine the source, the original files, complete experimental data, and a review of the journals or research institutions are usually required.
According to Science, the researchers did not provide the original, uncropped images they were asked to provide for the investigation. However, the journalistic request is no substitute for an institutional review process or a formal investigation by a journal.
The independent evidence has not disappeared.
The doubts about the human data and some of the images do not invalidate all research on Christchurch.
A relatively independent study led by Yidong Huang of the University of California, San Francisco found that introducing Christchurch into cells and mice carrying APOE4 reduced the accumulation of tau. The imaging experts who reviewed the article for Science found no abnormalities. The study was published inNature Neuroscience.
Further studies in mice, cell cultures, and brain organoids have suggested a potential effect of Christchurch on tau, amyloid, and neuroinflammation. In 2026, researchers reported that Christchurch expression in astrocytes reduced signs of amyloid pathology in mice. Another study found that expression of the variant in the liver of APOE4-carrying mice increased amyloid clearance and reduced cognitive impairment.
These are interesting findings, but mouse and cell models are not proof of protection against Alzheimer's in humans. Many therapeutic candidates that have been successful in such models have subsequently failed in clinical trials.
Even among critics, there is a distinction between two copies and a single copy. David Holtzman of Washington University in St. Louis believes that two copies of Christchurch may indeed inhibit the disease, but he says the existing human data are not convincing that one copy is enough.
“Macondo”: The promise that has not yet been made public
The third version the team worked on is called Macondo, after the fictional town in Gabriel Garcia Marquez's book. It was found in the CNTF garden of the writer's sister, who the researchers say lived into her nineties without developing the disease that afflicted other family members.
The findings have not yet been published in a peer-reviewed paper. According to the investigation, two researchers who contributed to the draft have removed their names from it. One of them, Diego Spolevda-Paya, found that the Macondo carriers in the Colombian family developed dementia over a very wide range—from the late 1930s to the early 1970s. He concluded that the version alone does not provide reliable protection.
בPreliminary research that has not yet been peer-reviewed Spolveda-Paia suggested that the effect attributed to Reelin-COLBOS may also result from a combination of several genetic variants, rather than a single protective gene.
This possibility fits with the understanding that resistance to Alzheimer's may be a multigenic trait: each person has thousands of rare variants, and sometimes only a certain combination of them, along with environmental and health factors, determines when the disease will break out.
The response of researchers and institutions
Joseph Arboleda-Velásquez and Yaquil Quiros, two of the study leaders, declined Science's requests to be interviewed or to respond in writing to the detailed questions.
The hospitals and academic bodies associated with their work said they were conducting confidential internal reviews of research concerns, but could not comment on matters that may involve private medical information or unpublished research.
The GNA group defended its work. In response to Science, it said that the alleged problems with the images concern a limited portion of the publications and do not, in its opinion, compromise the authenticity of the original data, the quantitative analyses, or the conclusions. The group added that if corrections or clarifications are required, they will be provided through official channels.
As of the time of publication of the investigation, no official determination of fraud or institutional determination of misconduct had been reported. The allegations require examination of the journals and research institutions, including examination of the original data and images.
Hope remains, but the evidence needs to be stronger
The story does not end with the conclusion that Christchurch and Reelin-COLBOS are unimportant. They may indeed participate in mechanisms that moderate the spread of tau or protect certain areas of the brain. However, the distance between an effect in cells or mice and an effective treatment in humans is still large.
More convincing proof would require additional carriers of both Christchurch copies who remain healthy beyond the expected age, transparent databases, validation by independent groups, and controlled clinical trials.
The broader lesson is not that there is no room for hope, but that great promise requires an exceptionally high level of evidence. When a rare genetic finding becomes the basis for multimillion-dollar investments and human trials, rigorous scientific review is not an obstacle to a cure—it is a condition for the path to the right place.
FAQ
Did the Science investigation prove that the studies were falsified?
No. The investigation pointed to controversial changes in the data and images that experts saw as duplications. A determination of forgery or scientific misconduct requires a formal review of the research institutions or journals and access to the original material.
Could APOE3 Christchurch still protect against Alzheimer's?
Yes, it's still a plausible scientific possibility. One human case and several studies in cells and mice support a possible effect. However, the evidence is stronger for two copies of the variant than for a single copy.
What is the difference between familial Alzheimer's and the common disease?
Autosomal dominant familial Alzheimer's is caused by rare mutations, such as PSEN1 E280A, and sometimes appears at a young age. Most cases of Alzheimer's are not caused by a single mutation, and develop from a combination of age, Genetics, health status and environmental factors.
Is there already a Christchurch-based treatment?
There is no approved treatment. Antibodies and gene therapies based on the putative mechanism are in early stages of research. Success in cells, mice, or monkeys does not guarantee that a treatment will be safe or effective in humans.
More on the subject on the science website
- Brainwashing — Alzheimer's: A New Therapeutic Direction
- Adjust the biological clock and protect the brain from Alzheimer's
- From cancer to Alzheimer's: Engineered immune cells reduce protein deposits in the brain
- Israeli study: Electron spin affects the formation of Alzheimer's proteins
- Gene therapy may cure Alzheimer's
For the original publication: Opening the original publication
