Simon Spichak

As an overall note, people can say shit that sounds good but do absolutely nothing to improve people's quality of life. Take that into consideration on everything you hear.

For a story in @beingpatient.bsky.social, I'm looking to speak with #neurologists who've had healthy patients come to them because they believed they had Alzheimer's after taking a direct-to-consumer Alzheimer's blood test.

DM me or get in touch at simon (at) beingpatient (dot) com

Working on a story for @thesicktimes.org about the quality, utility, and risks of exercise trials in Long COVID (and how often they mention/measure PEM). Looking to speak with anyone that's participated in these trials and experienced PEM/other negative side effects as a result.

Digital Cognitive Tests: How AI Tools Could Transform Alzheimer's Diagnosis and Care | Being Patient AI-powered digital cognitive tests like Cognivue, Linus Health, and Altoida aim to detect early dementia and speed up Alzheimer’s diagnosis.

Digital Cognitive Tests: How AI Tools Could Transform Alzheimer’s Diagnosis and Care beingpatient.com/digital-cogn... by @spichaksimon.bsky.social @beingpatient.bsky.social

#Alzheimers #dementia #AI #ArtificialIntelligence

Digital Cognitive Tests: How AI Tools Could Transform Alzheimer's Diagnosis and Care | Being Patient AI-powered digital cognitive tests like Cognivue, Linus Health, and Altoida aim to detect early dementia and speed up Alzheimer’s diagnosis.

There's a suite of new AI-powered digital cognitive tests coming to market.

The goal: Speed up the assessment and spot people who might benefit from more extensive testing sooner.

My latest for @beingpatient.bsky.social

beingpatient.com/digital-cogn...

Read the full article for more details on why the study may have failed and what it means for the idea that the herpes virus could cause Alzheimer's.

6/6

The trial used valacyclovir, an antiviral drug for herpes infections. Ultimately, the study found valacyclovir failed to slow cognitive decline in people with mild cognitive impairment (MCI) or early Alzheimer’s over 78 weeks.

5/6

At the 2025 Alzheimer’s Association International Conference, Dr. Davangere P. Devanand, a psychiatrist and director of geriatric psychiatry at Columbia University Medical Center, presented data from the first randomized clinical trial to test this theory.

4/6

But so far, supporting evidence has come only from observational studies of electronic health records, and from animal or cell-based lab experiments, which don’t always translate to clinical care.

3/6

For decades, scientists have speculated that a common infection — the herpes virus — might contribute to, or even cause, Alzheimer’s disease later in life.

2/6

Cold Sore Virus Concept for Alzheimer’s Takes a Hit in Drug Trial | Being Patient A long-standing theory linked herpes to Alzheimer’s — but the first clinical trial of an antiviral herpes drug found no benefit.

Can antivirals against herpes treat early-stage Alzheimer's? At #AAIC2025, researchers presented data from a clinical trial showing no effect.

My latest for @beingpatient.bsky.social

beingpatient.com/cold-sore-vi...

1/6

Thanks for reading - I've followed up with the Ministry to get more info and the number of people affected

Delayed Disability Benefit Leaves Canadians Struggling, Facing Bank Fees They Can't Afford Customers banking with TD and BMO are affected but it isn't clear how many or when the situation will be fixed.

Delayed Disability Benefit Leaves Canadians Struggling, Facing Bank Fees They Can't Afford

Customers with TD and BMO are affected but it isn't clear how many or when the situation will be fixed.

My reporting on the #CDB delay.

simonspichak.beehiiv.com/p/canada-dis...

The vast majority of studies aren't breakthroughs and that's ok! Even negative findings, that might contradict what you expected, can be valuable if the study is conducted well.

11/11 🧵

Research is exciting - but overhyping and overselling the findings is not helpful for anyone.

Things that might help cells in a dish or mice probably won't translate to people. And biology doesn't always come with a neat, easy-to-understand story.

10/n 🧵

Some studies only have one or two statistically significant findings. Despite most of the data being negative, and a high likelihood their only positive finding may have occurred due to chance alone, the abstract/discussion may spin it as a success.

9/n 🧵

A lot of times these studies might propose an elegant biological idea in the introduction or discussion. And they can cite studies that support the idea and make it sound like a sure-thing. But these ideas, not matter how alluring still need to go through validation!

8/n 🧵

And then we also have multiple studies that find slightly different effects. It is important to see whether other studies looking in other groups of people see something similar.

7/n 🧵

These are a great way to look for potential things to test. But these studies come with a lot of confounding factors, like socioeconomic status, that influence the data. Often, when the drug or nutrient that reduces the risk in observational studies, it fails in trials.

6/n 🧵

The other thing to be careful about are observational studies. These often look across electronic health records to see if people who take a certain drug or medication are at a higher/lower risk of a disease outcome.

5/n 🧵

The other thing to consider is how well the model or a measure represents a pathological process in humans. Are the behaviors or processes researchers measuring something that is validated in the human version of the disease.

4/n 🧵

If researchers expose mice or rats to a drug or another exposure, it has to be at a physiologically relevant level. Some animal studies use very, very high dosages that don't tell us much about what might be happening in humans.

3/n 🧵

The vast majority of studies in cells or animal models don't translate to humans.

They might provide interesting insights into disease processes, but only if the model that they're using is validated. What does that mean?

2/n 🧵

When researchers and press offices overhype their findings it can undermine trust and distort expectations for people living with these diseases.

A few things I learned as a researcher and now a science journalist.

A🧵
1/n

Great summary! I'd argue they also deviated from their preregistered protocol by including studies with patients who didn't have Long COVID (i.e. acute infection or recovery after hospitalization) www.crd.york.ac.uk/PROSPERO/vie...

Blood tests in healthy people lead to inconclusive results and false positives. But there are many other lifestyle factors that are more predictive and potentially modifiable -- like getting cardiovascular disease under control.

7/n

But only 43 percent of people who received a positive test result developed Alzheimer’s.

When you have a large population being tested for something that is relatively rare, you'll have way more false positives (even with a >85% accurate test) than true positives.

6/n

A study published in March tested how well six different blood biomarkers predicted Alzheimer’s disease in a 2,148 healthy older adults in Sweden. Over the course of 10 years, over 90 percent of the people who tested negative never developed Alzheimer’s or dementia.

5/n

The other issue: At this stage, testing even when you're healthy even with a great blood test doesn't work.

4/n

At the end of the study, researchers offered to show them the results of their blood test and brain scan. And by this point, a quarter of those people had changed their minds: With the information now readily available, only 60 percent said yes.

3/n