How Alzheimer's Is Diagnosed Before Memory Problems Appear

Research-informed explainer — last updated April 12, 2026

Alzheimer's disease begins in the brain 15 to 20 years before the first memory complaint. The amyloid plaques and tau tangles that define the disease accumulate silently during this window, causing no noticeable symptoms while the damage spreads. New biomarker tests — from PET scans to spinal fluid analysis to blood-based markers — can now detect this process years before clinical symptoms appear. For patients and families who want to understand their risk, or for those enrolled in prevention trials, knowing what happens before memory problems start has become one of the most important questions in neurology.

This explainer draws on research from five specialists in the Convene directory. Ronald Petersen at Mayo Clinic defined mild cognitive impairment (MCI) as a diagnostic entity — establishing the clinical concept that sits between normal aging and Alzheimer's dementia and that has become the target for early detection. William Jagust at UC Berkeley contributed to the 2018 NIA-AA Research Framework that redefined Alzheimer's as a biological condition detectable before symptoms, and to the hypothetical biomarker cascade model showing the sequence in which amyloid, tau, and neurodegeneration markers become abnormal. Reisa Sperling at Harvard and Brigham and Women's led the landmark work defining preclinical Alzheimer's and what it means to have the pathology of the disease without its symptoms. Steven DeKosky at the University of Florida co-authored the NIA-AA criteria for MCI due to Alzheimer's. Bradley Hyman at Massachusetts General Hospital and Harvard Medical School has published extensively on the neuropathological assessment of Alzheimer's, establishing the gold-standard framework for how the disease is confirmed at autopsy and how that informs biomarker interpretation in living patients.

What Alzheimer's disease actually is

Alzheimer's disease is defined by two abnormal protein deposits in the brain: amyloid-beta plaques that accumulate outside neurons, and tau neurofibrillary tangles that accumulate inside neurons. For decades, a definitive diagnosis required examining brain tissue after death. The key shift in the past 15 years has been learning to detect these biological changes in living people using imaging and fluid biomarkers.

Bradley Hyman and colleagues at the National Institute on Aging established the neuropathological assessment guidelines for Alzheimer's — the framework for what brain pathology must look like to confirm a diagnosis at autopsy [12, 13]. These guidelines use a scoring system called ABC (Amyloid, Braak stage for tau, CERAD score for neuritic plaques). What made these frameworks significant beyond pathology labs is that they gave researchers targets to detect with biomarkers: if you can measure amyloid and tau in a living brain or in body fluids, you can assess the same biological changes that define the disease.

The biological staging of Alzheimer's

The 2018 NIA-AA Research Framework — a collaboration involving William Jagust and many other researchers — redefined Alzheimer's disease based on biology rather than symptoms [4]. Under this framework, a person has Alzheimer's disease if they have the biological markers of amyloid and tau pathology, regardless of whether they have cognitive symptoms. This was a fundamental shift: it separated the disease (the biology) from its clinical consequences (the symptoms).

The framework uses three categories of biomarkers, sometimes called A/T/N:

• A (Amyloid): Amyloid-PET scan or cerebrospinal fluid amyloid levels
• T (Tau): Tau-PET scan or CSF phosphorylated tau levels
• N (Neurodegeneration): MRI showing brain atrophy, FDG-PET showing reduced glucose metabolism

A person who is A-positive, T-positive, and N-positive has the biological hallmarks of Alzheimer's disease. A person who is A-positive but T-negative and N-negative may be in a very early stage, before tau pathology has spread widely. This staging system has reshaped how clinical trials select participants and how early treatment is being considered.

The biomarker cascade: what comes first

The hypothetical model of Alzheimer's biomarkers, originally proposed in 2009 by Jack and colleagues including William Jagust, described the sequence in which different biomarkers become abnormal [5]. The sequence goes roughly like this: amyloid accumulation begins first, detectable by PET or CSF many years before symptoms. Tau pathology and neurodegeneration follow. Subtle memory changes emerge. Then full clinical MCI, and eventually dementia. Jagust and colleagues refined this model in 2013 to incorporate better data from longitudinal studies [6].

The practical implication is that amyloid is the earliest detectable change. A positive amyloid PET scan or low CSF amyloid-42 may appear 10 to 15 years before the first cognitive symptom. This detection window is the reason the field is focused so heavily on amyloid-based treatments — if you can intervene before tau spreads and neurons die, you have the best chance of preventing dementia from developing.

Mild cognitive impairment: the bridge stage

Ronald Petersen at Mayo Clinic defined mild cognitive impairment in a landmark 2004 paper as an abnormal cognitive state that falls between normal aging and dementia [1]. MCI involves measurable cognitive decline — often detectable on formal neuropsychological testing — without the functional impairment required for a dementia diagnosis. People with MCI can still manage their finances, drive, and live independently. Their cognitive symptoms are real, but not yet disabling.

MCI is clinically important because it represents the stage when most people first seek evaluation, and it is the population in which most early Alzheimer's clinical trials now enroll. The conversion rate from MCI to Alzheimer's dementia is approximately 10 to 15 percent per year — far higher than in the general population — and biomarker status (amyloid-positive vs. negative) predicts that risk substantially [3].

The 2011 NIA-AA workgroup guidelines, developed by Petersen, DeKosky, and colleagues, formalized the concept of MCI due to Alzheimer's disease [11] — distinguishing MCI caused by Alzheimer's pathology from MCI caused by other processes (vascular disease, Lewy body pathology, etc.). This distinction matters because the underlying cause determines prognosis and eventually treatment options.

Preclinical Alzheimer's: before any symptoms

Reisa Sperling's landmark 2011 paper with the NIA-AA workgroup on preclinical Alzheimer's proposed that the disease begins long before MCI — in a phase when biomarkers are abnormal but cognition is entirely normal [8]. This concept has been operationalized in several large prevention trials, including the A4 Study (Anti-Amyloid treatment in Asymptomatic Alzheimer's), which enrolled cognitively normal older adults with elevated amyloid on PET scanning.

The preclinical concept also introduced a new framework: if a person is amyloid-positive but has no symptoms, are they sick? There is no single answer. Many amyloid-positive cognitively normal individuals will eventually develop symptoms; some may not progress during their lifetime. This uncertainty makes the communication of preclinical biomarker results to patients one of the most complex challenges in modern neurology.

How diagnosis happens in practice today

A clinical evaluation for Alzheimer's typically begins with cognitive testing — structured assessments like the Mini-Mental State Examination (MMSE) or the Montreal Cognitive Assessment (MoCA). These can detect impairments but cannot distinguish Alzheimer's from other causes of cognitive decline, and they are normal in the preclinical stage by definition.

Brain imaging follows when there is clinical concern. MRI can show hippocampal atrophy and patterns of brain shrinkage consistent with Alzheimer's. FDG-PET can show reduced glucose metabolism in the pattern typical of Alzheimer's. Amyloid-PET scanning directly images amyloid deposits in the brain and provides the clearest presymptomatic biological information. CSF sampling — a lumbar puncture — measures amyloid-42, total tau, and phosphorylated tau; this is less expensive than PET and has been available longer, though it requires a spinal procedure.

Blood-based biomarkers are the emerging frontier. Plasma phosphorylated tau 217 (p-tau217) and related markers have shown high accuracy for detecting Alzheimer's pathology in research settings and are moving rapidly toward clinical use. The 2024 Alzheimer's Association Workgroup revised the diagnostic and staging criteria to incorporate blood-based biomarkers as valid evidence of the disease [7].

What this means for you

Most people reading this are not in a clinical trial or seeking a presymptomatic diagnosis. The practical takeaway is this: if you are worried about cognitive changes in yourself or a family member, the time to see a neurologist is now rather than later. Early evaluation can establish a baseline, rule out treatable causes (thyroid disorders, vitamin deficiencies, sleep apnea, medication effects), and — if Alzheimer's pathology is confirmed — connect you with centers that offer access to approved treatments and trials. Early-stage disease is precisely when treatment may slow progression most effectively.

Questions to ask your doctor

• What cognitive assessments do you use, and what would an abnormal result mean for my next steps?
• Is amyloid PET scanning or CSF biomarker testing appropriate for me at this point?
• If biomarker testing shows I have Alzheimer's pathology but no symptoms, what does that mean for my prognosis?
• Are there any clinical trials I qualify for based on my cognitive and biomarker status?
• What is the most important thing I can do to reduce my risk of progression if I am in an early stage?
• How often should I have follow-up cognitive testing to track any changes?