Slowing Alzheimer's decline means the disease gets worse more slowly than it otherwise would. It does not mean memory comes back.
In the trials behind today's anti-amyloid drugs, every group declined — the treated group simply declined less over the same stretch of time. That distinction decides what a family should expect from treatment. The Alzheimer's Association states plainly that there is no evidence to date that any treatment can restore or reverse memory loss or cognitive function due to Alzheimer's; anti-amyloid drugs aim only to slow decline in memory, thinking and function.
Medical information disclaimer: This article is for general educational purposes only and does not provide medical advice, diagnosis, or treatment. Always consult a physician or other qualified health professional about symptoms, medications, tests, or treatment decisions.
Table of Contents
- What the trial numbers actually showed
- Why lost memory does not come back
- Would a family notice the difference?
- Who these drugs are actually for
- The risks that come with the slowing
- Why timing of diagnosis is the part you can control
- Frequently Asked Questions
What the trial numbers actually showed
The clearest way to see the difference is to read the scores. CLARITY AD tested lecanemab (brand name Leqembi) using CDR-SB, a clinician-rated scale of memory, judgment, home life and personal care, where a rising score means a worsening person. Over 18 months, the lecanemab group's score rose 1.21 points and the placebo group's rose 1.66, as reported in the New England Journal of Medicine. Both numbers are positive. Both groups ended the trial worse than they started.
The famous "27% slowing" is the ratio between those two amounts of worsening — 0.45 divided by 1.66 — not a gain over where anyone began. Donanemab followed the same pattern. In TRAILBLAZER-ALZ 2, published in JAMA, CDR-SB worsened 1.20 points on the drug versus 1.88 on placebo in the low-to-medium tau group, a 35% slowing on the trial's primary measure. One detail there does stand out: 47% of treated participants had no measurable decline at one year, against 29% on placebo. That is a held position, not a recovered one.
Why lost memory does not come back
The biology explains the ceiling. The National Institute on Aging describes Alzheimer's as a progressive, currently irreversible disorder in which neurons lose their connections to each other and die, and brain tissue shrinks. Amyloid plaques are protein clumps that build up between neurons. Anti-amyloid drugs are antibodies that clear those clumps.
But the damage already done is structural loss — cells that are gone and circuits that no longer connect. Removing plaque addresses what may be driving further loss. It does not rebuild what has already died. That is why a drug can slow the slope of the line and never lift it back toward baseline. The neurons that carried a particular wedding, a particular route home, a particular grandchild's name are not waiting to be reactivated.
Would a family notice the difference?
Possibly not, on an individual level. Researchers set benchmarks called minimal clinically important differences — the smallest change a patient or family would actually perceive. A published analysis in Alzheimer's & Dementia put those thresholds at 0.98 CDR-SB points for mild cognitive impairment and 1.63 for mild Alzheimer's dementia. Lecanemab's 0.45-point difference sits below both.
On average, the change may fall under what a spouse or adult child can detect in daily life, even while the statistical effect on the group is real. A different framing may be more useful than points. A 2025 simulation study in Alzheimer's & Dementia: TRCI translated CLARITY AD's effect into time, estimating roughly 5.3 months of worsening avoided at 18 months, with treated patients projected to reach the placebo group's 18-month severity about 7.5 months later. That is a delayed arrival at the same place — worth something to many families, but not a reversal.
Who these drugs are actually for
The evidence applies to a narrow group, and being outside it matters. Someone with substantial existing memory loss is outside the evidence entirely. There is no trial showing what these drugs do at that stage, which is a different statement from showing they do nothing.
- **Stage.** The FDA label restricts Leqembi to patients started at the mild cognitive impairment or mild dementia stage. It has not been tested in moderate or severe Alzheimer's.
- **Confirmed amyloid.** Treatment requires confirmed amyloid pathology, established by imaging or spinal fluid testing — an Alzheimer's diagnosis by symptoms alone is not enough.
- **Genetic testing first.** ApoE ε4 genotyping is required before treatment begins, because that gene variant raises the risk of side effects.
The risks that come with the slowing
The FDA prescribing information for Leqembi carries a boxed warning — the agency's strongest — for amyloid-related imaging abnormalities, or ARIA. These are brain swelling (ARIA-E) and small brain bleeds (ARIA-H) that can be serious, life-threatening and fatal. Monitoring has since been tightened.
The FDA now recommends an additional, earlier MRI between the second and third infusion to catch these changes sooner. This is the trade that should be weighed honestly: a measured slowing that may fall below what a family perceives, against a real risk requiring repeated brain scans and infusion visits. Different people will weigh it differently, and both answers can be reasonable.
Why timing of diagnosis is the part you can control
Benefit is measured as decline avoided from the moment treatment starts. Nothing before that point is recoverable. That makes every month of delayed diagnosis permanently lost ground — and the National Institute on Aging notes these drugs slowed decline only in early-stage participants over 18 months and do not cure the disease. It is worth understanding how the older drugs differ, because they are often confused with the new ones.
Cholinesterase inhibitors such as donepezil raise levels of acetylcholine, a chemical messenger involved in memory. The Alzheimer's Society reports they stabilise symptoms for about six to twelve months in roughly half of the people who take them, after which decline resumes. They ease symptoms without altering the underlying disease. For a family noticing early changes, the practical step is to raise them now rather than waiting to see if they worsen — and to ask specifically whether amyloid testing is appropriate, since that test, not the symptom conversation alone, is what opens the door to anti-amyloid treatment.
Frequently Asked Questions
If both groups got worse, is the drug doing anything at all?
Yes. The treated groups in CLARITY AD and TRAILBLAZER-ALZ 2 worsened measurably less than placebo over 18 months, and in the donanemab trial 47% of treated participants had no measurable decline at one year versus 29% on placebo. The effect is real; it is a slower rate of loss, not a recovery.
Can donepezil restore memory if the newer drugs cannot?
No. Cholinesterase inhibitors stabilise symptoms for roughly six to twelve months in about half of those who take them, then decline resumes. They do not change the disease itself.
My relative is at a moderate stage. Can they take Leqembi?
The FDA label restricts it to patients started at mild cognitive impairment or mild dementia, and it has not been tested in moderate or severe Alzheimer's. Their care team can discuss symptom-focused treatment and clinical trial options instead.
What does the required genetic test change?
ApoE ε4 genotyping is required before treatment because that variant raises the risk of ARIA — brain swelling and microbleeds carrying a boxed warning. The result informs how a clinician weighs the risk, not whether the drug works.
