Sign in

Sarah Ackley

@sfackley.bsky.social
650 followers 1.3K following 87 posts

Assistant Professor of Epidemiology at Brown University. Dementia Pharmacoepidemiology. Dynamical Modeling. Statistics. Metascience & Knowledge Representation. Cat Mom. She/her.

PostsRepliesMedia
Reposted by Sarah Ackley
MedPage Today @medpagetoday.com · 23/05/2026
A statistical approach used to support amyloid-targeting treatment for Alzheimer's disease may lead to overstated claims about amyloid-cognition relationships, an analysis suggested. @sfackley.bsky.social www.medpagetoday.com/neurology/alzh…
medpagetoday.com
Study Questions Methods Used in Alzheimer's Drug Analysis
Quantile aggregation may lead to overstated claims about amyloid-cognition relationships
081
Sarah Ackley @sfackley.bsky.social · 22/05/2026
Check out our new research letter!
081
Reposted by Sarah Ackley
Mike Greicius @mikegreicius.bsky.social · 18/05/2026
🧪 #alzheimer #amyloid Amazing new paper from @sfackley.bsky.social that exposes the statistical alchemy Lilly applied to their donanemab data: jamanetwork.com/journals/jam... There is little to no correlation between amyoid removal and clinical outcomes with donanemab (or lecanemab).
jamanetwork.com
Methodological Considerations for Quantile Aggregation in Alzheimer Disease Trials
This study evaluates the quantile-aggregation method using both simulated data and publicly available Alzheimer disease randomized trial data.
1227
Reposted by Sarah Ackley
Isabel Elaine Allen @datacooker.bsky.social · 18/05/2026
Beware the reanalysis -
052
Reposted by Sarah Ackley
AcadeMax @academaxjournals.bsky.social · 18/05/2026
Appreciation to @mikegreicius.bsky.social for sharing this and to @sfackley.bsky.social for this important paper. It carefully examines whether amyloid removal in donanemab aligns with meaningful clinical benefit. Follow @academaxjournals.bsky.social for relevant research. #Science #Neuroscience
071
Reposted by Sarah Ackley
Brown University School of Public Health @brownpublichealth.bsky.social · 18/05/2026
Professor @sfackley.bsky.social finds prior methods of studying cognition and amyloid, which builds up in the brains of Alzheimer’s patients "made a failed trial look like it had successfully removed amyloid." Learn more 🧠 sph.brown.edu/news/2026-05-18/alzhe…
Visualization of brain neurons coated in amyloid protein, a feature of Alzheimer's disease.Smiling headshot of Professor Sarah Ackley
051
Reposted by Sarah Ackley
Mike Greicius @mikegreicius.bsky.social · 18/05/2026
Lilly used "quantile-aggregation" to transform individual level data with 100s of data points and a correlation of 0.08 (see figure) into 10 data points with a correlation of 0.87! Dr. Ackley clearly debunks this approach with simulations and real-world data. Check it out. Spread the word.
A scatter plot showing the minimal correlation between amyloid removal and clinical outcomes in the phase 3 trial for donanemab. R-squared is less than 1% (meaning less than 1% of the variance in the clinical measure can be explained by the amyloid removal).
051
Sarah Ackley @sfackley.bsky.social · 18/05/2026
Eli Lilly’s reanalysis of TRAILBLAZER-ALZ 2 proposes a new method for evaluating amyloid as a surrogate endpoint and argues the resulting high R-squared values are “unlikely to be artificially inflated.” Our response now out in JAMA Neurology: they are. jamanetwork.com/journals/jam...
jamanetwork.com
Methodological Considerations for Quantile Aggregation in Alzheimer Disease Trials
This study evaluates the quantile-aggregation method using both simulated data and publicly available Alzheimer disease randomized trial data.
041
Reposted by Sarah Ackley
Washington Post Opinions @postopinions.bsky.social · 28/02/2026
"I am a clinical epidemiologist whose research focuses on assessing interventions for the treatment and prevention of dementia," @sfackley.bsky.social writes. "I believe it is premature to promote coffee for the prevention of dementia." wapo.st/4s5WRiD
051
Reposted by Sarah Ackley
Alt The National Institute on Aging @altthenia.bsky.social · 06/11/2025
For those of you attending #GSA2025 in Boston, consider this panel of #NIH awardees, Drs. Alden Gross, @mariaglymour.bsky.social , @sfackley.bsky.social , and @bryandjames.bsky.social. Please remember to tag #geronsky and @geronsociety.bsky.social in your posts.
The "biggest baddest" bias in aging research: A spirited GSA debate.
0157
Reposted by Sarah Ackley
Alzheimer's & Dementia Journals @alzdemjournals.bsky.social · 03/11/2025
🔔 #TRCI Special Issue 🔔 Marking the 10th anniversary of AD Drug Development Pipeline articles led by Dr. Jeffrey Cummings! Submit your research highlighting advances in ADRD clinical trials and drug development! ⏰Oct. 31, 2026 alz-journals.onlinelibrary.wiley.com/hub/journal/...
alz-journals.onlinelibrary.wiley.com
Alzheimer's & Dementia: Translational Research & Clinical Interventions
<em>Alzheimer's & Dementia: Translational Research & Clinical Interventions</em> journal bridges drug discovery research and clinical studies for dementia & Alzheimer's.
163
Reposted by Sarah Ackley
Maria Glymour @mariaglymour.bsky.social · 14/09/2025
Such a cool paper. #EpiSky @amjepi.bsky.social #DataScience #Reproducibility Led by @sfackley.bsky.social at @brown-epi.bsky.social Research on reproducible research funded by NIH/NIA - thank you!
0102
Sarah Ackley @sfackley.bsky.social · 28/08/2025
There is a lot of literature with highly visible peaks just below 0.05. And the easiest way to get that is to p-hack under the null bc that's the scenario with the most density in the neighborhood of 0.04.
000
Sarah Ackley @sfackley.bsky.social · 28/08/2025
Before writing this I had somehow assumed if you had lowish power you'd be more likely to get 0.04+delta than 0.01 + delta. But the p-curves, increase as you move towards zero. So a lot of p-hacking practices will show up in ways that aren't particularly visible--e.g. 0.01.
200
Sarah Ackley @sfackley.bsky.social · 28/08/2025
This work illustrates how large-scale text analysis, combined with extensions of p-curve methods to variable-power settings, can determine how efforts to improve research rigor are influencing the quality and credibility of large bodies of the scientific literature.
030
Sarah Ackley @sfackley.bsky.social · 28/08/2025
The fraction of borderline significant P values has modestly declined. But this appears to reflect increased study power, not reductions in P hacking or publication bias.
110
Sarah Ackley @sfackley.bsky.social · 28/08/2025
Check out our recently published paper at AJE! academic.oup.com/aje/advance-... We analyzed 21,332 abstracts (2000-2024) from four major epidemiology journals, extracting 25,288 P values using ChatGPT’s 4o model with few-shot in-context prompting. #episky #reproducibility #metasci #metascience
academic.oup.com
Validate User
152
Reposted by Sarah Ackley
Triangulation of Innovative Methods to End Alzheimer’s Disease @timead.bsky.social · 25/07/2025
#TIME-AD is heading to #AAIC25! 📊 We’ll be sharing results from our latest preprint examining whether reductions in brain amyloid—a key Alzheimer’s biomarker—can meaningfully predict changes in cognition and functioning. 🔗 Read the preprint: tinyurl.com/5n8psura
021
Sarah Ackley @sfackley.bsky.social · 25/07/2025
Looking forward to presenting this renanalysis of the A4 study, as well as a simulation study evaluating SILA's ability to detect ceilings in amyloid accumulation at #AAIC25. #ENDALZ www.medrxiv.org/content/10.1...
medrxiv.org
Evaluation of Amyloid Removal as a Surrogate for Cognitive Decline: Pilot Analysis in Individual-Level Data from the A4 Study of Solanezumab
Background Amyloid removal has been used as a surrogate outcome in Alzheimer’s disease trials, allowing accelerated approval of aducanumab and lecanemab. The A4 (Alzheimer’s Clinical Trial Consortium ...
020
Reposted by Sarah Ackley
Triangulation of Innovative Methods to End Alzheimer’s Disease @timead.bsky.social · 23/07/2025
🧠 TIME-AD is now on Bluesky! #TIME-AD brings together results from different study designs and populations to find what really works to prevent and delay dementia. Follow along as we share what we’re learning and check out our website to learn more: www.time-ad.org #EndALZ #EpiSky #NeuroSky
195
Sarah Ackley @sfackley.bsky.social · 22/07/2025
This link might be better!
medrxiv.org
Evaluation of Amyloid Removal as a Surrogate for Cognitive Decline: Pilot Analysis in Individual-Level Data from the A4 Study of Solanezumab
Background: Amyloid removal has been used as a surrogate outcome in Alzheimer's disease trials, allowing accelerated approval of aducanumab and lecanemab. The A4 (Alzheimer's Clinical Trial Consortium...
000
Sarah Ackley @sfackley.bsky.social · 22/07/2025
These methods should be replicated in trials with greater clinical decline in placebo groups and larger treatment effects (e.g., lecanemab, donanemab) to assess whether amyloid truly works as a surrogate.
100
Sarah Ackley @sfackley.bsky.social · 22/07/2025
... and does not find evidence that amyloid is a good surrogate marker. But: A4 had minimal progression and modest amyloid removal.
100
Sarah Ackley @sfackley.bsky.social · 22/07/2025
FDA accelerated approvals for Alzheimer’s drugs have relied on amyloid reduction as a surrogate marker, but does it actually lead to cognitive benefit? This preprint uses individual-level A4 trial data to ask that question rigorously ...
100
Sarah Ackley @sfackley.bsky.social · 22/07/2025
New preprint up! Evaluation of Amyloid Removal as a Surrogate for Cognitive Decline: Pilot Analysis in Individual-Level Data from the A4 Study of Solanezumab www.medrxiv.org/content/10.1...
medrxiv.org
131
Reposted by Sarah Ackley
Andrea Gilmore Bykovskyi @algilmore.bsky.social · 25/06/2025
Excellent talk at #MELODEM2025 by @sfackley.bsky.social evaluating amyloid as a surrogate marker with individual level A4 trial data One big take away — 5 years was not enough time to see cognitive decline in the placebo group == a major challenge
121
Reposted by Sarah Ackley
Jennifer Weuve @epidancer.bsky.social · 25/06/2025
Now at #MELODEM2025 Is reducing/removing amyloid from the brain a reasonable surrogate for slowing/stopping cognitive decline in dementia? @sfackley.bsky.social uses data from the A4 RCT (of solanezumab) to answer this q, ... @melodem.bsky.social
131
Sarah Ackley @sfackley.bsky.social · 28/05/2025
Thanks!
000
Sarah Ackley @sfackley.bsky.social · 27/05/2025
I was invited to comment for AlzForum on GLP-1s and dementia. GLP-1s show promise, but RCTs like GRADE have not demonstrated clear cognitive benefits. Observational studies may be confounded by socioeconomic factors and prescribing patterns. Read the full comment: www.alzforum.org/papers/gluco...
alzforum.org
Glucose-Lowering Medications, Glycemia, and Cognitive Outcomes: The GRADE Randomized Clinical Trial. | ALZFORUM
020
Sarah Ackley @sfackley.bsky.social · 24/04/2025
In addition to today's NIH-grant-related anxiety, a cat vomited a fur ball in my shoe.
020
Sarah Ackley @sfackley.bsky.social · 15/04/2025
www.nytimes.com/2025/04/14/u...
nytimes.com
Harvard Will Not Comply With a List of Trump Administration Demands (Gift Article)
Federal officials said Harvard must enact “merit-based reform” in hiring and admissions and report international students who broke rules, among other steps. Harvard called the demands unlawful.
120
Sarah Ackley @sfackley.bsky.social · 14/04/2025
I wrote 1/3 of a page of my grant!
headings only: significance, innovation, approach, future directions
120
Sarah Ackley @sfackley.bsky.social · 07/04/2025
This would seem to contradict that, at least for the most recent death. Looking into the other 2. bsky.app/profile/kaka...
000
Sarah Ackley @sfackley.bsky.social · 07/04/2025
So many more people than I realized!
010
Sarah Ackley @sfackley.bsky.social · 06/04/2025
& Not sure what to make of the hospitalizations here--seems much lower than the oft cited 1 in 3 cases. People aren't going or can't go to the hospitial when they need to? Or people aren't that sick? bsky.app/profile/mega...
100
Sarah Ackley @sfackley.bsky.social · 06/04/2025
Note on my previous calculation--arguably 1/1000 is too high.
010
Sarah Ackley @sfackley.bsky.social · 06/04/2025
Fair point. 1/1000 could be wrong for this setting. It seems like (from some sloppy googling) that since 2003, there has only been one other US measles death: www.usatoday.com/story/news/2...
100
Sarah Ackley @sfackley.bsky.social · 06/04/2025
7/ Without knowing more about the deaths(like if they had other health conditions) and with small Ns, I would be cautious drawing any strong conclusions. But--as I think most in PH would agree--concern about loss of elimination is increasingly warranted.
020
Sarah Ackley @sfackley.bsky.social · 06/04/2025
6/ This could indicate--among other possibilities--that there are cases that are not being reported. Mild cases w/o rash among the vaccinated? Avoidance of healthcare/interaction with public health workers?
120
Sarah Ackley @sfackley.bsky.social · 06/04/2025
5/ Other post-elimination US outbreaks have reached 100s of cases (tho typically in more isolated religious communities), but the number of deaths in this one is starting to seem high.
110
Sarah Ackley @sfackley.bsky.social · 06/04/2025
4/ Today, with 3 deaths and 544 cases, the probability is even lower: 1-dbinom(0,544,1/1000) - dbinom(1,544,1/1000) - dbinom(2,544,1/1000) [1] 0.01788214 That’s just a 1.8% probability of seeing 3+ deaths.
110
Sarah Ackley @sfackley.bsky.social · 06/04/2025
3/ Calculation in R: 1-dbinom(0,350,1/1000) - dbinom(1,350,1/1000) [1] 0.04859081 There’s a 0.048 probability (4.8%) of seeing 2+ deaths in 350 cases.
110
Sarah Ackley @sfackley.bsky.social · 06/04/2025
2/ A few weeks ago, when there were 2 deaths and ~350 cases, a friend suggested the high death rate indicated underreporting. Here’s why they were concerned:
110
Sarah Ackley @sfackley.bsky.social · 06/04/2025
1/ The number of deaths (3) in this measles outbreak seems high given its size (480 TX + 50 OK + 10 NM = 544). Typically, ~1/1000 die from measles based on past data (pre- and post-elimination). Is it possible cases are underreported? #episky
242
Sarah Ackley @sfackley.bsky.social · 05/04/2025
Hilbert’s funeral in 1943 was sparsely attended, with only a handful of people present, including physicist Arnold Sommerfeld. News of his death spread months later.
010
Sarah Ackley @sfackley.bsky.social · 05/04/2025
By the time Hilbert died in 1943, the University of Göttingen had been nearly completely restaffed by the Nazis. Many former faculty members had been Jewish or married to Jews.
100
Sarah Ackley @sfackley.bsky.social · 05/04/2025
At a banquet a year later, Minister of Education Bernhard Rust asked Hilbert if the Mathematical Institute had suffered from the Jews' departure. Hilbert replied, "Suffered? It doesn't exist any longer, does it?"
100
Sarah Ackley @sfackley.bsky.social · 05/04/2025
Some math/physics history I happened across yesterday— David Hilbert lived to see the Nazis purge many prominent faculty members at the University of Göttingen in 1933, including Hermann Weyl, Emmy Noether, and Edmund Landau. (Wikipedia)
110
Reposted by Sarah Ackley
The New York Times @nytimes.com · 03/04/2025
Breaking News: The Trump administration intends to block $510 million in federal contracts and grants for Brown University.
nytimes.com
Trump Administration Set to Pause $510 Million for Brown University
The administration has now targeted five schools’ federal funding as part of a pledge to combat what it considers to be antisemitism on university campuses.
103473269
Sarah Ackley @sfackley.bsky.social · 02/04/2025
Picture of MIT's great dome
000