Reviewed by the Help Dementia Editorial Team — our editors review every article for accuracy against guidance from the National Institute on Aging, the Alzheimer’s Association, and peer-reviewed sources.
When Alzheimer’s disease is diagnosed after the cognitive decline has progressed beyond mild cognitive impairment, many newer disease-modifying drugs become unavailable—a window that some patients lose without realizing it. The newest anti-amyloid monoclonal antibodies, including lecanemab (Leqembi) and donanemab (Kisunla), work only in early symptomatic stages because they target amyloid buildup before neurons have suffered irreversible damage. A 73-year-old woman in the later stages of Alzheimer’s, unable to manage finances or recognize family members reliably, would not qualify for these treatments even though she has the amyloid pathology the drugs are designed to clear. Late diagnosis means years of potential therapeutic benefit are lost to neurodegenerative processes that cannot be reversed.
The critical window for new Alzheimer’s drugs hinges on cognitive thresholds, not just the presence of amyloid in the brain. A person can have amyloid accumulation years before symptoms emerge, remaining in a phase when treatment could theoretically prevent decline. Once cognitive symptoms progress from mild to moderate, most anti-amyloid therapies no longer offer clinical benefit. The FDA-approved drugs work because they intervene when the brain still has neuronal reserves—the ability to compensate for pathology. In advanced stages, the damage is done, and clearing amyloid cannot restore lost cognitive function.
Table of Contents
- Why Alzheimer’s Drugs Only Work Early
- Biomarker Testing and Cognitive Eligibility Criteria
- The Cost of Late Diagnosis and Loss of Cognitive Time
- Access, Cost, and Practical Barriers to Early Testing
- What Happens in Moderate and Advanced Dementia
- The Preclinical Amyloid Stage and Early Intervention Trials
- Pathology Burden and Cognitive Reserve
Why Alzheimer’s Drugs Only Work Early
Anti-amyloid monoclonal antibodies were developed based on the amyloid hypothesis: removing amyloid-beta plaques from the brain slows cognitive decline if caught early. Lecanemab, approved by the FDA in January 2023, demonstrated that it slowed cognitive decline by approximately 35 percent over 18 months in people with mild cognitive impairment or mild dementia caused by Alzheimer’s disease. However, this benefit only emerged in people with measurable amyloid and tau pathology combined with early symptoms—not those with advanced cognitive impairment. The neurobiology explains why timing matters so dramatically.
In early Alzheimer’s, amyloid accumulation triggers a cascade of neuroinflammation, tau tangles, and neuronal death, but the brain’s functional networks remain largely intact. Removing amyloid at this stage allows the brain to recover some reserve and slow further decline. In late-stage Alzheimer’s, widespread neuronal loss has already occurred, synaptic connections are damaged, and the hippocampus and cortex have atrophied. Clearing amyloid from a brain that has lost half its neurons cannot restore those dead cells or rebuild lost connections. A person diagnosed at age 82 with moderate dementia may have decades of unrecognized amyloid accumulation behind them, but by the time diagnosis occurs, the therapeutic window has closed.
Biomarker Testing and Cognitive Eligibility Criteria
Receiving an anti-amyloid drug requires three things: confirmed amyloid pathology, confirmed tau pathology, and cognitive impairment within a specific range. Amyloid can be detected through PET imaging, cerebrospinal fluid (CSF) biomarkers in spinal taps, or blood-based biomarkers like phosphorylated tau and phosphorylated tau-181. A person with biomarker evidence of amyloid and tau but no cognitive symptoms yet—those in the asymptomatic preclinical phase—cannot access lecanemab or donanemab outside of research trials, because the drugs are not approved for prevention. Conversely, a person with cognitive decline too severe for these drugs may still have detectable amyloid.
The cognitive thresholds used in trials defined “mild cognitive impairment” or “mild dementia” using the Montreal Cognitive Assessment (MoCA) or Clinical Dementia Rating (CDR) scales. Lecanemab trials enrolled people with MoCA scores of 20-26 (out of 30) or CDR scores suggesting mild impairment. Once a person’s MoCA score drops below 16, or their CDR indicates moderate dementia, they fall outside the evidence base. At that point, the person has too much neurodegeneration to benefit—not because amyloid isn’t present, but because the neural substrate is too damaged. This creates a cruel paradox: the amyloid is still there, but the patient has aged out of treatment eligibility.
The Cost of Late Diagnosis and Loss of Cognitive Time
Early diagnosis of cognitive impairment is not routine in primary care. Many people experience memory problems, get told “it’s normal aging” or “it’s stress,” and don’t receive formal neuropsychological testing for years. By the time a formal diagnosis happens—often prompted by family concern or a fall—the person may already be in moderate stages. In a 2020 U.S. study, the average time from first symptom to diagnosis was 2-3 years, during which cognitive decline continues unchecked.
The lost time is quantifiable in cognitive terms. If someone experiences mild cognitive impairment for 5 years before diagnosis, during which decline progresses at 3-4 points per year on cognitive scales, that person enters treatment trials 15-20 points lower than they might have. Even if lecanemab could slow their decline by 35 percent from that point, they have already lost years of potential benefit. Some countries with better diagnostic infrastructure—like the Netherlands and parts of Scandinavia—have higher rates of early Alzheimer’s detection through systematic cognitive screening programs. The result is that their populations have broader access to newer drugs simply because diagnosis happens sooner. A person diagnosed with mild cognitive impairment at age 65 in a country with systematic screening may benefit from a drug that offers no hope to someone diagnosed at age 75 in a country without such infrastructure, even if both have amyloid pathology.
Access, Cost, and Practical Barriers to Early Testing
Even where early diagnostic testing exists, it often isn’t pursued. Comprehensive amyloid biomarker testing—especially amyloid PET imaging—costs thousands of dollars and is rarely covered by insurance for asymptomatic people. Blood-based biomarkers are cheaper and increasingly accessible but still not standard in primary care. Without insurance coverage for preventive testing, people rely on symptom-triggered evaluation, which inherently comes late. For those who do get diagnosed in the therapeutic window, cost becomes another barrier. Lecanemab, administered as an intravenous infusion every two weeks, costs approximately $26,500 annually.
Donanemab is similar. Even for people who qualify cognitively and biomarker-wise, affording or accessing these drugs is not guaranteed. Rural patients may need to travel hours for infusions. People with limited English proficiency may struggle to understand informed consent. Older adults with other chronic diseases may not tolerate the drugs well; lecanemab carries a 21 percent risk of amyloid-related imaging abnormalities (ARIA), a form of microhemorrhage or microinfarction visible on brain MRI, which can cause cognitive symptoms. For a person on the edge of cognitive eligibility, the risk-benefit calculus shifts quickly.
What Happens in Moderate and Advanced Dementia
Once Alzheimer’s progresses to moderate or advanced stages, anti-amyloid drugs are not recommended because clinical trials showed no benefit and risks remain. The cognitive decline accelerates, behavioral symptoms emerge—agitation, wandering, aggression—and care needs increase dramatically. People in moderate stages often cannot dress themselves, manage hygiene, or recognize family. Advanced stages involve loss of speech, inability to swallow without aspiration risk, and dependence on 24-hour supervision. The absence of disease-modifying therapy does not mean no treatments exist.
Cholinesterase inhibitors like donepezil and memantine (an NMDA antagonist) offer modest cognitive support in moderate stages, though their benefit is small compared to amyloid-targeting drugs. Behavioral management—antipsychotics, antidepressants, anti-anxiety medications—addresses symptoms but does not slow disease progression. A person in moderate dementia receiving these medications is managing symptoms, not treating the underlying pathology. The distinction matters: symptom management extends quality of life and comfort, but disease modification would extend cognitive function. Those diagnosed late miss the only point in the disease where disease modification is possible.
The Preclinical Amyloid Stage and Early Intervention Trials
A growing body of research focuses on the preclinical phase of Alzheimer’s: people who have amyloid and tau accumulation but no cognitive symptoms. These individuals are the theoretical ideal for intervention, since pathology can be modified before neuronal loss begins. Multiple trials are underway to test whether early intervention in asymptomatic people prevents or delays symptom onset. Donanemab trials (AHEAD) and lecanemab trials have enrolled asymptomatic amyloid-positive individuals to determine whether catching pathology before symptoms offers durable protection.
However, mass screening for asymptomatic amyloid is not yet standard practice. Identifying cognitively normal people with amyloid requires either brain imaging (expensive, limited access) or blood biomarker screening programs (increasingly available but not universally offered). A person who has no symptoms, no awareness of cognitive change, and no family history of dementia is unlikely to seek amyloid testing. Even if they had a blood test, they would need genetic counseling to understand what amyloid positivity means for their future and what interventions might help. The infrastructure to support asymptomatic screening and early intervention in the general population does not yet exist in most healthcare systems.
Pathology Burden and Cognitive Reserve
The relationship between brain pathology and cognitive symptoms is not linear. Some people maintain cognitive function despite significant amyloid and tau accumulation because they have higher cognitive reserve—the brain’s ability to use alternate networks and compensate for damage. Cognitive reserve is built through education, complex occupational work, cognitive engagement, physical fitness, and social connection. A retired professor with extensive education and lifelong intellectual activity might tolerate substantial amyloid accumulation without manifesting symptoms, while someone with less education and cognitive engagement might show symptoms at lower amyloid levels. This creates unpredictability in diagnosis timing.
Two 70-year-olds with identical amyloid PET scans and identical tau levels might be at very different stages of cognitive disease because their reserve differs. The person with higher reserve may remain asymptomatic while the other experiences mild cognitive impairment. When diagnosis finally occurs—triggered by symptoms, not by pathology burden—the timing varies. Higher cognitive reserve also means a later symptomatic diagnosis, which means later eligibility assessment for amyloid-targeting drugs. By the time symptoms are obvious enough to prompt testing and diagnosis, the brain may have already progressed beyond the therapeutic window for drugs that depend on neural plasticity and surviving synaptic connections to work.
