A New Blueprint for Alzheimer’s

By Sarah C.P. Williams | Illustration by Keith Negley | Portraits by Jörg Meyer

When Connie Walterscheidt sent her family a “Happy Mother’s Day” text on Valentine’s Day in 2017, her sister Debbie Gritz couldn’t just shrug it off. Debbie and Connie, then 53, had worked side by side as phlebotomists at the Oklahoma Blood Institute for two decades. Lately, Debbie had noticed Connie making careless mistakes and forgetting how to use lab equipment.

“Connie had spent all these years being top-tier at work, and then suddenly she was really struggling,” says Debbie. “It was hard to watch.”

Shortly after the Mother’s Day text, Debbie encouraged Connie to talk to a neurologist. Later that year came the diagnosis: early-onset Alzheimer’s.

For Debbie and sister Vickie Elmer, Connie’s disease was just the beginning. Despite no known family history of Alzheimer’s, their easygoing father, Leon, received the same diagnosis a few months later.

“We all knew there was something wrong with Dad, but we had put it off to old age,” recalls Vickie. “Once Connie was diagnosed and we learned more about what Alzheimer’s looked like, we realized we should get him tested.”

Then, their mother, Ruby, started showing the same symptoms and, ultimately, received the same diagnosis.

“It hit us like dominoes,” Debbie says. “Full force.”

For the two healthy sisters, the experience was overwhelming.

“Connie was the hardest. She wasn’t herself anymore,” Debbie says. “She became belligerent, angry, and very vocal and hurtful toward us.”

Until their sister’s death in 2024 and their parents’ deaths in 2021 and 2023, Debbie and Vickie—along with one of Connie’s daughters— became full-time caretakers. At the same time, they began to worry about their own futures, their children, and the 34 great-grandchildren that Ruby and Leon left behind.

“We were walking in there blind when Connie and our parents were diagnosed,” Vickie says. “If it shows up in us, I want to know early and I want the rest of our family to at least have an inkling of what to do.”

So they turned to VP&S researchers, joining a family study that tracks people at high risk for Alzheimer’s. When their father passed away in 2023, they donated his brain to the study. But so far, they have few answers; their family has none of the genetic variants currently known to be associated with Alzheimer’s.

“The known risk genes only explain around half the genetic contribution to Alzheimer’s,” says Christiane Reitz, MD, PhD, a neurologist and genetic epidemiologist who works with families like Debbie and Vickie’s. “There’s much more to be discovered.”

That lack of answers for so many individuals and families is driving a fundamental shift in Alzheimer’s research. At Columbia, scientists are no longer betting everything on a single, simple explanation for the disease. Instead, they’re attacking it from many angles: probing the genetics of the disease, asking questions about the immune system’s role, analyzing millions of brain cells to understand when and how damage begins, and studying why some people’s brains resist disease.

The work is pointing toward a complex picture. Alzheimer’s likely has many causes that differ from person to person, which means it will require many different treatments. Basic research must uncover each of those triggers and reveal how they work.

“You’re not going to fix something unless you know what’s fundamentally broken,” says Scott Small, MD, director of Columbia’s Alzheimer’s Disease Research Center (ADRC). “I think now we’re getting closer to that. The field is becoming more optimistic.”

Moving Past Amyloid

For a short time in the 1980s, Alzheimer’s researchers thought they had the disease figured out. Pathologists had observed that most people who died of Alzheimer’s had clumps (called plaques) of a sticky protein called amyloid beta clogging up their brains. At the same time, neurologist Peter St George-Hyslop, MD, discovered that some patients had mutations in the amyloid precursor protein (APP) gene—the very gene that produces amyloid protein. It all seemed to be pointing to one common thread: Amyloid plaques in the brain were causing Alzheimer’s.

“We immediately oversimplified things,” admits Dr. St George-Hyslop, now the Belle and Murray Nathan Professor of Neurology at Columbia. “We thought that was it.”

But Dr. St George-Hyslop and others soon found that the APP mutation was relatively rare, and that amyloid plaques in the brain didn’t always correspond directly with dementia symptoms. More recently, drugs that clear amyloid plaques from the brain have shown only modest benefits for Alzheimer’s patients.

“There’s no question that amyloid plaques are bad. But we’re coming around to the idea that amyloid is the smoke, and the fire is what’s going on inside neurons,” says Dr. Small. “If you want to intervene, you can’t just clear the smoke; you have to put out the fire.”

To find out what else contributes to Alzheimer’s, researchers at VP&S continue to probe the genetics of the disease. Gene variations that make a person more likely to develop Alzheimer’s not only point toward ways to gauge people’s risks, but also help to explain what can go wrong to trigger the disease.

“Genetics is one of the most direct ways of finding out what the underlying problem is,” says Dr. Reitz.

The Genetic Foundation

Richard Mayeux, MD, chair of the Department of Neurology and one of the field’s most prominent genetic epidemiologists, has spent more than 35 years building the infrastructure to ask increasingly complex questions about what causes Alzheimer’s. In 1992, he established the Washington Heights-Inwood Columbia Aging Project, which continues to track aging and dementia in thousands of families in Northern Manhattan.

He also leads the National Institute on Aging Alzheimer’s Disease Family Based Study, which follows 2,500 families around the country—including Debbie and Vickie’s—that are affected by the disease. The approach: Find genes that appear only in affected family members, never in healthy relatives.

“We’ve taken a very restrictive approach in finding these gene variants,” Dr. Mayeux explains. “To meet our criteria, a variant had to be present in every affected individual in a family, and in nobody in the family who doesn’t have the disease.”

The Columbia team has been at the forefront of the field for a long time, says Dr. St George-Hyslop, who joined the University in 2022. “We have made many of the critical discoveries on the genetics and biology of Alzheimer’s and related diseases, including the discovery of and characterization of genes that have been game changers for the field because they shed light on the basic processes of the disease.” These include the genes for APP, which is cleaved to create the sticky peptides that form the amyloid plaques; PSEN1 and PSEN2, enzymes that cleave APP; and SORL1, which helps control the trafficking of APP to sites where it is cleaved.

Turning to more diverse populations than ever before is helping to reveal new genes associated with Alzheimer’s. Dr. Reitz has carried out studies around the world to find genes that might be more prevalent in underrepresented populations. In 2013, she revealed that ABCA7—a gene that has a minor influence on Alzheimer’s risk in European whites—is the second-strongest risk factor in African Americans. In 2024, she identified MPDZ, another gene that increases risk specifically in people of African ancestry. She’s now co-leading a large National Institutes of Health-funded initiative recruiting participants from Africa, South America, and underrepresented U.S. populations.

“You can’t just take a variant in a gene that’s found in one ethnic group and assume it will work the same way in another ethnic group,” says Dr. Reitz. “You need to actually do the work in different populations.”

Down the road, what’s found in one population could have ripple effects for everyone with Alzheimer’s, even those who don’t share the same genetic backgrounds.

By following people over decades, tracking families at risk of Alzheimer’s, and turning to new populations, Drs. Mayeux and Reitz and colleagues have uncovered a list of rare variants in genes, and many have already pointed toward a few key systems that go wrong in the brain in Alzheimer’s disease: inflammation, the movement of proteins between cells, cholesterol metabolism, and how molecules are broken down and recycled.

With Dr. Small at the helm, the Columbia ADRC focuses on two of those areas: the immune system’s role in Alzheimer’s, and how the molecular recycling program in brain cells goes awry with the disease.

Immune Betrayal

As the gene list grew, a pattern emerged. Genes like CD33 and TREM2 weren’t involved in making or processing amyloid. Instead, they were unique to microglia, the brain’s immune cells. Dr. St George-Hyslop’s group played a key role in discovering such genes and dissecting how they contribute to the disease.

“The Alzheimer’s field was for a long time solely focused on neurons, which are the primary cells that die during the disease,” says Philip De Jager, MD, PhD, chief of neuroimmunology. “But neurons are in an environment that is defined and supported by other cells including immune cells.”

Microglia, together with astrocytes, are supposed to be the brain’s support and cleanup crew. When bacteria or viruses invade, microglia attack. After a stroke or trauma, they sweep away dead cells and help reestablish connections. When Alzheimer’s screens started turning up genetic variation in microglia genes, Dr. De Jager wanted to know how the cells were involved in the disease.

In a project spanning the last five years, Dr. De Jager analyzed 1.6 million brain cells from more than 400 donors at different stages of Alzheimer’s disease. His team identified different types of cells that appear at different points in the progression of disease: One type of microglia accumulates early with amyloid plaques, another shows up alongside clumps of a protein called tau, and a subtype of astrocyte emerges only before cognitive decline.

“If different types of microglia and astrocytes are doing different things at different times, we can’t just take any Alzheimer’s patient and stimulate all these cells at once to treat the disease,” Dr. De Jager explains. “There might be points in the sequence of disease-related events where we want to turn down their activity instead, or only stimulate one very particular subtype of microglia or astrocyte.”

Dr. De Jager is now following up on the nuances between astrocyte and microglia types that appeared in his study. Since each brain represents only one single snapshot in time, his team is modeling how groups of immune cells change over time to create a more dynamic picture by considering the 400 brains simultaneously. Then he wants to know whether blocking just one type of cell might stop the progression of cognitive decline.

Andrew Teich, MD, PhD, a neuropathologist at Columbia, has taken a complementary approach to studying how immune signals in the brain change throughout the progression of Alzheimer’s. Working with Guy McKhann, MD, a neurosurgeon at Columbia, they recognized a unique opportunity to study brain tissue from Alzheimer’s patients early in their disease. Dr. McKhann is an expert in the treatment of a particular type of cerebrospinal fluid buildup in the brain—normal pressure hydrocephalus (NPH)—that affects aging patients. NPH is treated by placing a brain “shunt,” a tube that drains the spinal fluid to another part of the body. Columbia is one of the first centers to routinely biopsy the brain at the time of shunt replacement to examine pathology in NPH patients, who sometimes have early-stage Alzheimer’s changes, independently from their hydrocephalus.

“It was a perfect opportunity to look at the early inflammatory responses in Alzheimer’s disease,” Dr. Teich says. “We usually only have the chance to see brains postmortem at the very end stage of disease.”

The biopsies not only had expected signs of disease, like amyloid plaques and tau tangles, but also changes to immune molecules. High levels of YKL-40—a molecule produced by astrocytes and microglia during inflammation—were predictive of diseaseassociated changes in these cell types, suggesting YKL-40 could serve as an early biomarker of how these processes play out in atrisk individuals. Drs. Teich and McKhann are following up on the role of the molecule in the progression of Alzheimer’s.

Together, the findings show that Alzheimer’s involves a cascade of changes in the brain: Immune cells switch their behavior, produce inflammation, and trigger the brain’s destruction. But the immune system isn’t the only culprit. Other genes pointed toward a fundamental problem inside neurons themselves.

When the Brain's Recycling System Breaks

Among the genes Dr. Mayeux’s team discovered through family studies was SORL1, which is involved in how cells recycle proteins. In parallel, by asking why Alzheimer’s starts where it does in the brain, SORL1 and the recycling system it regulates emerged as an answer. Since then, Drs. Small and St George-Hyslop and other researchers have been investigating how SORL1 dysfunction leads to disease.

Dr. Small’s work focuses on a structure inside cells called the endosome—a sorting center where cells decide what molecules to recycle, what to break down, and what to send back to the cell surface. Dr. Small worked out the details of how SORL1 is needed to help proteins exit the endosome. When the gene for SORL1 is faulty, proteins get stuck inside.

That traffic jam happens first in cells in the entorhinal cortex, one of the earliest brain regions affected by Alzheimer’s. Dr. Small’s lab discovered why: This region has high levels of cellular trafficking because it’s one of the most hyperconnected areas in the brain.

“This part of the brain is effectively Grand Central Station,” Dr. Small says. “Imagine if you block all the exiting trains. You get chaos.”

When the recycling system breaks down, neurons themselves stop working. At the same time, microglia—which also need SORL1 to function properly— can’t respond normally to the neuronal damage, accelerating disease progression. Dr. Small calls it “a double whammy.”

Dr. Small co-founded a company developing compounds to restore the endosome’s function. But he’s also still solving a fundamental genetics puzzle: SORL1 has around 500 variants, and scientists still don’t know which ones truly cause dysfunction.

“Ultimately, we want clinicians to be able to test which gene for SORL1 someone has and use that to assess their risk of Alzheimer’s or guide what treatment might work,” Dr. Small says. “First, we have to figure out which variants are important.”

Multiple Diseases, Multiple Solutions

The dozens of genes linked to Alzheimer’s tell a story of complexity. Each person likely has a different combination of genetic vulnerabilities—some with immune problems, some with cellular recycling failures, and others with both, or another problem entirely. Ultimately, each molecular error converges on the same outcome: neurons that degenerate, causing dementia.

“Until a few years ago, we were focused on Alzheimer’s as a unitary disease. It turns out it’s not,” says Dr. Mayeux. “My suspicion is that the heterogeneity explains why we have so many different genes and none of them are clear right now.”

The realization echoes what happened in cancer research. Doctors no longer end their diagnosis at “lung cancer”—they identify molecular subtypes and treat each differently.

“If someone says, ‘I have lung cancer’ or ‘I have heart disease,’ there are lots of different types,” Dr. Mayeux says. “Those fields have moved forward by focusing on signatures that tell you, ‘This is the type of lung cancer you have,’ and then develop specific treatments for each subtype. We’re not there yet with Alzheimer’s, but that’s where we’re headed.”

The idea that there are many converging causes for Alzheimer’s means that different research approaches aren’t competing—they’re complementary.

“There are a lot of good ideas that are not mutually exclusive,” says Dr. Teich. “I don’t think we’re in a situation where somebody’s going to win and everyone else is wrong. I think there are a lot of people playing different musical instruments, now realizing we’re part of the same orchestra.”

The goal of each researcher is to make Alzheimer’s a disease that can be screened for, caught early, and treated so it doesn’t progress.

“Sooner or later, when we go see our annual primary care provider, we’ll not only be testing for cholesterol and diabetes, but also for Alzheimer’s,” Dr. Small predicts.

Hope For Patients

The timeline of basic research—unraveling how individual molecules within brain cells work together to protect against damage or to trigger dementia—is slow. But the pieces are coming together.

“We’ve gone from not knowing anything about the disease to having simplistic ideas to now beginning to understand how these things come together,” Dr. St George-Hyslop says. “As you see more pieces of the puzzle, it’s easier to solve the rest of it.”

For families like Debbie and Vickie’s, the research represents more than scientific progress. Participating in Columbia’s family studies connects them to educational resources and caregiver support groups.

“My team establishes direct connections to the communities we work with,” explains Dr. Reitz. “For a lot of these families—especially those with many people affected—they really don’t know where else to get help.”

Debbie and Vickie say they’ve gotten some relief from the support and knowledge gained through their Columbia study. But, without answers about why their sister and parents developed Alzheimer’s, they remain on edge.

“We worry for our kids—not only that they’ll have to take care of us, but what if they develop it even before us? The not knowing is hard,” says Debbie.

For them, and other families affected by Alzheimer’s disease, the answers, tests, and new treatments can’t come quickly enough.