In brief, the main reason I don’t think the argument works that autosomal-dominant Alzheimer’s has a different etiology than sporadic Alzheimer’s is that they look, in so many respects, like essentially the same disease, with the same sequence of biomarkers and clinical symptoms:
Amyloid pathology starts in the default mode network, and gradually spreads throughout the brain over 15-20 years.
It eventually reaches the medial temporal region, where Primary Age-Related Tauopathy is lying in wait.
At this point, tau pathology, a prion-like pathology which in Alzheimer’s has a very specific conformation, starts spreading from there. The tau protein misfolds in the exact same way in both forms of the disease (Falcon et al (2018). Tau filaments from multiple cases of sporadic and inherited Alzheimer’s disease adopt a common fold), however it misfolds in a different way in the large majority of other known tau pathologies, of which there are a dozen or so (Shi et al (2021). Structure-based classification of tauopathies).
Then, neurodegeneration follows in lockstep throughout the brain with the presence of tau pathology, with cognitive deficits matching those expected from the affected brain regions. In particular, since the hippocampal formation is located in the medial temporal region, anterograde amnesia is typically the first symptom in both types of Alzheimer’s (unlike many other forms of neurodegeneration, in which other clinical symptoms dominate in the early stages).
It’s as if two bank robberies occurred two hours apart in the same town, conducted in almost exactly the same manner, and in one we can positively ID the culprit on camera. It’s a reasonable conclusion that the culprit in the other case is the same.
Some further evidence:
There has been extensive causal mediation modeling, e.g. Hanseeuw et al (2019). Association of Amyloid and Tau With Cognition in Preclinical Alzheimer Disease, which so far as I’m aware always fits the amyloid → tau → neurodegeneration (ATN) model of the disease, and generally doesn’t fit contradictory models.
There have been extensive in vitro and in vivo studies showing that amyloid pathology can directly induce tau pathology (which, as mentioned above, correlates with the location and severity of neurodegeneration). He et al (2018). Amyloid-β plaques enhance Alzheimer’s brain tau-seeded pathologies by facilitating neuritic plaque tau aggregation and Lodder et al (2021). CSF1R inhibition rescues tau pathology and neurodegeneration in an A/T/N model with combined AD pathologies, while preserving plaque associated microglia are just two of dozens of examples.
The main genetic risk factors of sporadic, late-onset Alzheimer’s disease are shown to impair amyloid-β clearance or compaction (e.g. Castellano et al (2011). Human apoE Isoforms Differentially Regulate Brain Amyloid-β Peptide Clearance, among many others), although through less well-understood mechanisms, often involving lipid processing, so by itself this isn’t smoking gun evidence, but it is consistent with everything else that is known.
As for the evidence from amyloid-targeting therapies, a few things can be said. I’ll focus on monoclonal antibodies, which are the most-favored approach in the research community today. I’m aware of seven such antibodies: aducanumab, donanemab, lecanemab, solanezumab, crenezumab, gantenerumab, and bapineuzumab. Of these, three have had promising, though not stellar, findings in clinical trials:
Aducanumab passed its cognitive endpoint in its phase 2 trial (Sevigny et al (2016). The antibody aducanumab reduces Aβ plaques in Alzheimer’s disease), and one of two phase 3 trials (Haeberlein et al (2022) Two Randomized Phase 3 Studies of Aducanumab in Early Alzheimer’s Disease).
Donanemab passed its primary endpoint in its phase 2 trial. Mintun et al (2021). Donanemab in Early Alzheimer’s Disease (It hasn’t yet reported from a phase 3 trial.)
Lecanemab was found, in a Bayesian analysis of its phase 2 trial, to have a 64% chance of slowing cognitive decline by at least 25% (however, the primary endpoint was an 80% chance, so it technically failed its trial). Swanson et al (2021). A randomized, double-blind, phase 2b proof-of-concept clinical trial in early Alzheimer’s disease with lecanemab, an anti-Aβ protofibril antibody (It also hasn’t yet reported from phase 3. [Update Sep 27, 2022: Now it has. Slowdown of cognitive decline by 27% with a p-value of 0.00005.])
In the above cases, the reduction in the pace of cognitive decline is generally around 30% or so, with a fairly wide range around that. [Removed claim about the other antibodies “almost always” showing a nonsignificant directional effect, after reviewing the data again.] Furthermore, some of the failed studies skirted the edge of statistical significance, and when they have looked at earlier vs. later intervention have typically found that earlier intervention is more effective (e.g. Doody et al (2014). Phase 3 Trials of Solanezumab for Mild-to-Moderate Alzheimer’s Disease).
This is all what we expect if amyloid is causally far upstream of the more proximate causes of neurodegeneration: if you only start intervention in the clinical phase, then you’re 15-20 years into the disease and the tau pathology is already active and spreading and causing neurodegeneration on its own, thus you’ve effectively taken the gun out of the shooter’s hand after they’ve already pulled the trigger. This is helpful (and in Alzheimer’s, it appears to slow decline by ~30%). On the other hand, you either need to intervene much earlier (not yet tested, although the first results from such trials are expected later this year), or in a different manner (I favor tau antibodies for the clinical phase) if you expect to do more than that.
(I know I didn’t provide references to all my claims, but I can dig them up from my notes for anything specific if you’re curious.)
I apologize if this is piling on, but I would like to note that this error strikes me as very similar to another one made by the same author in this comment, and which I believe is emblematic of a certain common failure mode within the rationalist community (of which I count myself a part). This common failure mode is to over-value our own intelligence and under-value institutional knowledge (whether from the scientific community or the Amazon marketplace), and thus not feel the need to tread carefully when the two come into conflict.
In the comment in question, johnswentworth asserts, confidently, that there is nothing but correlational evidence of the role of amyloid-β in Alzheimer’s disease. However, there is extensive, strong causal evidence for its role: most notably, that certain mutations in the APP, PSEN1, and PSEN2 genes deterministically (as in, there are no known exceptions for anyone living to their 80′s) cause Alzheimer’s disease, and the corresponding proteins are well understood structurally and functionally to be key players in the production of amyloid-β. Furthermore, the specific mutations in question are shown through multiple lines of evidence (structural analysis, in vitro experiment, and in vivo experiments in transgenic mice) to lead directly (as opposed to indirectly, via a hypothetical other Alzheimer’s-causing pathway) to greater production of amyloid-β.
A detailed summary of this and further evidence can be found in section 1.1 “Rationale for targeting Aβ and tau” of Plotkin and Cashman (2020). Passive immunotherapies targeting Aβ and tau in Alzheimer’s disease. A good general survey on amyloid-β production is Haass et al (2012). Trafficking and Proteolytic Processing of APP.
(My background: I have a family member with Alzheimer’s and as a result I spent five months studying the scientific literature on the subject in detail. I am posting under a pseudonym to protect my family member’s privacy.)