For years, my family noticed my father making small mistakes. We did what most families do: We explained it away. The stress of his schedule, we said — he was working constantly, under real pressure.
But when he came to meet my second daughter the week after she was born, I could no longer explain it away. As he entered the gate to our yard, he looked down at our small dog — an animal who had been a fixture in my home for years, a face he should have known well — and his expression went uncertain. Then he asked if I had gotten a new dog.
I am a specialist in Alzheimer’s disease. I knew, in that moment, what I was seeing.
My mother had raised concerns with his primary care physician more than once before this. My father denied any problem, and he could still pass a standard cognitive screen. His cognitive reserve was high. His physician was not dismissive out of negligence. The system gave him no tools to act on my mother’s concerns.
By then, the disease already had been advancing for a long time. I know this now not just as a daughter, but as a physician who studies this process. The biology of Alzheimer’s disease begins 15 to 20 years before a family starts noticing something is off. My father’s disease began in midlife, silently, while he was still working and raising his family. I missed it. Not because I lacked training, but because I was trained to wait for unmistakable decline before acting.
That is exactly what our clinical system teaches. It is the wrong lesson.
Alzheimer’s is not primarily a disease of old age. It is a decades-long biological process that reaches clinical visibility only late in its course, when the brain has already sustained substantial and often irreversible injury. Amyloid accumulates. Tau pathology spreads through vulnerable neural networks. By the time memory fails and a diagnosis is made, that cascade has been underway for well over a decade.
The clinical system is calibrated to the endpoint of disease. We do not diagnose Alzheimer’s disease at its origin. We diagnose its aftermath.
I see this in my clinic regularly. Patients arrive only after symptoms have declared themselves. By the time they reach me, the window for intervention has already narrowed. What I rarely see is a patient who has been identified as being at increased risk in midlife and referred before decline becomes undeniable. That patient almost never presents, because there is no system to identify them. I am regularly asked about prevention by my patients’ children, healthy adults who are watching a parent decline and want to understand their own risk. I have no formal pathway to offer them.
There is no standardized framework for assessing neurodegeneration risk before symptoms appear. Primary care physicians lack the tools and guidelines to identify who is at elevated risk. Neurologists see patients after symptoms emerge. No specialty owns prevention. The people most likely to benefit from early intervention move through the health care system without ever being told the disease process may be already underway.
This is not for lack of scientific foundation. Large population studies have identified modifiable risk factors, including cardiovascular and metabolic health, sleep, physical activity, and social engagement, which together account for a substantial proportion of dementia risk. Multidomain intervention trials have shown that targeting these factors can preserve cognitive function in at-risk populations.
The tools to act on this knowledge are arriving faster than the systems to deploy them. In 2025, the Food and Drug Administration cleared the first blood tests for Alzheimer’s disease, including one designed specifically for use in primary care. These tests are currently indicated for patients already showing symptoms, but they are a harbinger of what is coming. We are moving toward a world in which a routine blood draw can detect Alzheimer’s-related pathology years before symptoms appear.
Disease-modifying therapies have demonstrated efficacy in early symptomatic disease. The expectation is that intervening earlier will yield greater benefit. Anti-amyloid therapies are now being tested in presymptomatic people with elevated amyloid burden in the AHEAD 3-45 clinical trial, which completed recruitment in 2024.
What we need is not simply earlier diagnosis but a fundamental shift in how we think about brain health. The model already exists in other areas of medicine: identify risk early, monitor change over time, and intervene before irreversible damage occurs. For brain health, this means structured risk assessment in primary care, referral pathways for those at elevated risk, and specialist evaluations that go beyond cognitive testing to assess the full range of factors that shape brain aging. It means a longitudinal plan, adjusted as evidence and individual trajectories evolve.
The case for acting earlier is strong, but this comes with two risks worth naming: continued reliance on reactive care that delays intervention, and premature adoption of interventions that exceed the supporting evidence. A responsible prevention model must hold both in view. Clinical action should be proportionate to the strength of available evidence. Biomarkers should be used selectively, where results would meaningfully change management. Uncertainty should be named clearly rather than obscured.
My father is not an edge case. He is what this disease looks like from the inside: a brain’s resilience slowly eroded by years of accumulating pathology. His physician had no framework to address this process early. The clinical system gave my family no pathway for thinking about his risk before his decline became unmistakable.
That is a failure of systems, not of individuals. The biology of this disease will not wait for the infrastructure to catch up.
Elizabeth A. Bevins, M.D., Ph.D., is a faculty neurologist at the University of California, San Diego specializing in Alzheimer’s disease and a member of the Alzheimer’s Disease Cooperative Study. She is building a prevention program at UC San Diego called Brain Longevity.
