checkedThe Diary of a CEO2026-11-12last verified 2026-11-13

did the cochrane review find amyloid-targeting alzheimer's drugs don't meaningfully slow decline?

verdict: confirmed
the primary source says it, at that magnitude
“They found that the drugs basically have no meaningful effect in slowing the progression, much less leading to improvement.”

a guest, a neurologist said on The Diary of a CEO, 2026-11-12, at 36:20. hear it in context → the speaker cited: the Cochrane analysis.

first, the part that holds

the speaker has the review right. the 2026 cochrane review of amyloid-targeting antibodies called their effect on cognition and dementia severity at 18 months "trivial", and said removing amyloid "does not appear" to bring clinically meaningful improvement. people on the drugs kept declining in every trial, so "much less leading to improvement" holds as well.

what we found when we went and looked

we read the april 2026 cochrane review (CD016297). it pooled 17 placebo-controlled trials with 20,342 people across seven amyloid-targeting antibodies, and every trial was funded by the drug's manufacturer. at 18 months the pooled effect on memory and thinking (ADAS-Cog) was a standardized mean difference of -0.11 (95% CI -0.16 to -0.06; 13 trials, moderate certainty). the effect on dementia severity (CDR-SB) was -0.12 (-0.24 to -0.00; 9 trials, low certainty). the authors called these effects "trivial" and the effect on daily function "small at best". the effects are not zero, since the confidence intervals mostly exclude zero, but the review judged them too small to matter clinically. that is the claim the speaker made. the caveat is that this is a pooled result across seven drugs, five of them abandoned. in their own pivotal trials, the two approved antibodies (lecanemab and donanemab) slowed decline on CDR-SB by 0.45 and 0.70 points more than placebo, and whether that much is meaningful is still argued.

assertion by assertion

assertionverdictwhat the record says
the cochrane analysis of amyloid-targeting drugs found no meaningful effect on slowing alzheimer's progressionconfirmedthe review's conclusion: the effect on cognitive function and dementia severity at 18 months is "trivial", the effect on functional ability is "small at best", and removing amyloid does not appear to bring clinically meaningful improvement. pooled SMDs were -0.11 for ADAS-Cog (moderate certainty) and -0.12 for CDR-SB (low certainty, upper confidence limit -0.00). the limitation is that the effects are statistically detectable rather than zero, and the dementia-severity estimate is low certainty ("may make little to no difference"), so the authors' 'not meaningful' is a clinical judgment, not a finding of zero effect. [1][2]
the drugs did not lead to improvementconfirmedthe review measured how much people declined compared with placebo, and no improvement was reported. in the lecanemab trial, CDR-SB worsened by 1.21 points on the drug and 1.66 on placebo. in the donanemab trial it worsened by 1.72 and 2.42 points (combined population). treated groups got worse, just more slowly. [2][3][4]
popular framing: this means the two approved drugs showed no effect in their own trialsmisattributedthe cochrane number is a class rollup across seven antibodies, and only one trial each of lecanemab and donanemab is in it. in its 18-month trial, lecanemab beat placebo on CDR-SB by -0.45 (95% CI -0.67 to -0.23). in its 76-week trial, donanemab beat placebo by -0.70 (-0.95 to -0.45). by our arithmetic that is about 27% and 29% slower decline relative to placebo. both were statistically significant. whether differences of that size matter to patients is the real dispute, and it is a separate question from whether any effect exists. [3][4]
side detail: the drugs carry brain swelling and bleeding riskconfirmedcochrane estimated about 107 more cases of ARIA-E (brain swelling) per 1,000 people (11 trials, moderate certainty) and 29 more symptomatic cases per 1,000 (2 trials). it found no detectable increase in serious adverse events or deaths (high certainty). in the donanemab trial, ARIA-E occurred in 24.0% on the drug versus 2.1% on placebo. [1][2][4]

both directions

for the speaker: the review's own conclusion matches the claim: a "trivial" effect on cognition and dementia severity at 18 months, function "small at best", and amyloid removal "does not appear" to bring clinically meaningful improvement. for cognition the finding is moderate certainty. the review also found the drugs raise the risk of brain swelling, by about 107 more people per 1,000 with any ARIA-E, and all 17 trials were funded by the manufacturers, which makes a small pooled effect harder to wave away. in the individual trials, people on the drugs still declined, so "much less leading to improvement" is accurate.

against: the effects are small but not zero. the pooled ADAS-Cog confidence interval excludes zero, and the CDR-SB interval just touches it. the review rated the dementia-severity estimate low certainty ("may make little to no difference"), so the speaker states it more flatly than the evidence grade does. most of the pooled trials tested antibodies that failed and were dropped (bapineuzumab, solanezumab, crenezumab, gantenerumab and others). that pulls the class average toward zero, and the result says less about the two approved drugs than the speaker implies. in their own pivotal trials, lecanemab and donanemab slowed CDR-SB decline by 0.45 and 0.70 points more than placebo, and both results were statistically significant. "the drugs basically have no meaningful effect" is fair as a summary of the review. read as a verdict on the two approved drugs specifically, it goes further than the review does.

what this cannot say

the cochrane library full text returned an error, and pubmed and pmc pages showed a cookie or browser check. we read the review's plain-language summary on cochrane.org and its abstract through the national library of medicine's e-utilities service, but not the full forest plots or any per-drug subgroup analysis. we cannot say whether the review analyzed lecanemab or donanemab separately. the abstract we read gives no specific minimal-clinically-important-difference threshold. the percent-slowing figures for the two approved drugs are our own arithmetic from the published changes, not numbers stated in the papers. "meaningful" is a clinical judgment that no dataset settles. the review covers 18 months and people with mild cognitive impairment or mild dementia, and six ongoing trials could change its conclusions. we did not verify the timestamp or exact wording of the episode.

sources, each one fetched and read

  1. cochrane, plain language summary: are anti-amyloid monoclonal antibodies effective? (CD016297), published 16 april 2026, search current to 7 august 2025 What we read there: 17 studies, 20,342 people, all company-funded. the antibodies "probably result in little to no difference in the decline in memory functioning and thinking ability" at 18 months, and may make little to no difference to dementia severity. brain swelling 119 vs 12 per 1,000. no increase in other serious adverse events or deaths. authors' conclusion: the effect on cognition and dementia severity is "trivial" and the effect on function "small at best". Fetched 2026-11-13 (HTTP 200). https://www.cochrane.org/evidence/CD016297_are-medicines-anti-amyloid-monoclonal-antibodies-reduce-build-abnormal-proteins-brain-effective
  2. cochrane database of systematic reviews 2026;4:CD016297, amyloid-beta-targeting monoclonal antibodies for people with mild cognitive impairment or mild dementia due to alzheimer's disease, doi:10.1002/14651858.CD016297, PMID 41985900 (abstract via NLM e-utilities) What we read there: ADAS-Cog SMD -0.11 (-0.16 to -0.06), 13 trials, 9,895 people, moderate certainty. CDR-SB SMD -0.12 (-0.24 to -0.00), 9 trials, 8,053 people, low certainty. ADCS-ADL SMD 0.09 (0.03 to 0.16), moderate certainty. any ARIA-E 107 more per 1,000 (77 to 148 more). serious adverse events and mortality not increased (high certainty). the antibodies "increase the risk of amyloid-related imaging abnormalities." Fetched 2026-11-13 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/41985900/
  3. lecanemab in early alzheimer's disease, new england journal of medicine 2023;388(1):9-21, PMID 36449413 (abstract via NLM e-utilities) What we read there: CDR-SB change at 18 months: 1.21 on lecanemab vs 1.66 on placebo. adjusted difference -0.45 (95% CI -0.67 to -0.23; P<0.001). ARIA-E 12.6% on lecanemab. the authors call the result moderately less decline and say longer trials are needed. Fetched 2026-11-13 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/36449413/
  4. donanemab in early symptomatic alzheimer disease: the TRAILBLAZER-ALZ 2 randomized clinical trial, JAMA 2023;330(6):512-527, PMID 37459141 (abstract via NLM e-utilities) What we read there: combined population CDR-SB change at 76 weeks: 1.72 on donanemab vs 2.42 on placebo, difference -0.70 (-0.95 to -0.45). iADRS difference 2.92 (1.51 to 4.33). ARIA-E 24.0% vs 2.1%. 3 treatment-related deaths on donanemab vs 1 on placebo. Fetched 2026-11-13 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/37459141/

speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam