Health Nerd
@gidmk.bsky.social
30K followers 400 following 2.3K posts
Epidemiologist. Research Fellow. Doctor of Spreadsheets. Writer (Slate, TIME, Guardian, etc). PhD, MPH. Host of senscipod Email [email protected] he/him. Find my writing on Substack and Medium.
Posts Media Videos Starter Packs
gidmk.bsky.social
To paraphrase @michaelhobbes.bsky.social, if you're going to compare ultra-processed candy with minimally-processed salads, your study isn't very useful.

bsky.app/profile/gidm...
gidmk.bsky.social
And if you look at the diets...lol.

You could eat a salad with steamed/poached protein or a sandwich with CHOCOLATE MILK AND CANDY.

Totally shocking that one of these did worse in the experiment. Who'd have thunk.
gidmk.bsky.social
And if you look at the diets...lol.

You could eat a salad with steamed/poached protein or a sandwich with CHOCOLATE MILK AND CANDY.

Totally shocking that one of these did worse in the experiment. Who'd have thunk.
gidmk.bsky.social
They provided similar calories on both diets (it appears higher in the UPF arm, but no statistical test to check significance). The self-reported median intake of calories was higher on the UPF diet (and adherence was lower on minimally-processed) but this was not statistically significant.
gidmk.bsky.social
Step 1: prove that people eat a little bit more food when it tastes better.

Step 2: reword "tastes better" to "Ultra-processed".

Step 3: published in Cell Metabolism.
tamarhaspel.bsky.social
This may be the worst nutrition trial I've ever seen.

Subjects got a minimally processed diet or an ultra-processed diet, ostensibly to test the effect of processing.

But the UPF diet was higher in sat fat, sugar, refined grains, & lower in fiber.

What happened?
www.cell.com/cell-metabol...
gidmk.bsky.social
In a world where funding and career advancement are tied to citation counts (i.e. this world) it is problematic.
gidmk.bsky.social
Alternatively, aggressively go after co-authorship on as many guidelines as possible.
gidmk.bsky.social
Definitely one useful take home!
gidmk.bsky.social
Here's an example. Nearly 30k(!) citations. Of those, I'd estimate about 20-22k are from being a middle author on GBD and related publications.

This author also has nearly a dozen retracted RCTs due to significant data issues.

scholar.google.com/citations?us...
scholar.google.com
gidmk.bsky.social
There are researchers who are in the top 0.01% of academics worldwide because they send a few hundreds lines of data to the IHME once every year or two and get onto the GBD studies.

This is...problematic.
gidmk.bsky.social
A useful post on the bizarre citation cabal that is the IHME and the Global Burden of Disease study.
ikashnitsky.phd
I wrapped up my scattered comments on IHME into a blog post

🔗 ikashnitsky.phd/2025/ihme-bibl

tl;dr: avoid getting unwarranted co-authorship recognition and do apply some sort of contribution-weighted thinking when evaluating someone's publication record based on bare numbers
Beyond Fraud: How IHME Distorts Academic Metrics
Dr. Ilya Kashnitsky is a demographer @ Statistics Denmark.
ikashnitsky.phd
gidmk.bsky.social
My new piece is on why we don't really know if soft drinks are giving people liver disease despite the headlines, and also a fascinating and problematic metascience issue.

gidmk.substack.com/p/are-fizzy-...
Are Fizzy Drinks Giving You Liver Disease?
The latest headlines about sugar and artificially sweetened beverages continue to mislead.
gidmk.substack.com
gidmk.bsky.social
My 2yo, trying to convince me to carry her: I just a little girl!

My 2yo, trying to convince me to give her my coffee: I just a big girl!
gidmk.bsky.social
The Cass review into children's gender services in the UK continues to be terrible.

A new report from Australia led by the fantastic Dr. Julia Moore - and featuring me - looks into the issues in detail.

My new piece.

gidmk.substack.com/p/the-cass-r...
The Cass Review Continues To Be Severely Flawed
A new paper from an Australian team shines a light on the problems with the review
gidmk.substack.com
gidmk.bsky.social
Ah yes, the exact opposite of science.

Very much the quiet part out loud.
atrupar.com
RFK Jr on Tylenol and autism: "It is not proof. We're doing the studies to make the proof."
gidmk.bsky.social
Yep. Structural barriers>individual motivation every time
gidmk.bsky.social
Realistically though the most likely outcome is that people will not change their diets anyway because that's generally how dietary recommendations fail.
gidmk.bsky.social
We don't know whether this is caused by COVID-19. All of the participants in the study had almost certainly had COVID-19, so it's incredibly uncertain. It's a very specific population, so not really generalizable. Etc, etc, etc.
gidmk.bsky.social
I'm not trying to attack the researchers or research. This is interesting, and with enough of these studies you might get to the point where you could make some real comments on how Long COVID may impact the brain.

But the press release and media buzz about it are absurd.
gidmk.bsky.social
Pretty much what I expected. In this very small group of patients, there was a modest increase in AMPA receptor density in Japanese Long COVID patients who had persistent cognitive impairment.
gidmk.bsky.social
Haven't read the study or press release yet.

My guess is that this is going to be a tiny, arbitrary sample of people where there is a very modest but statistically significant increase in some random biomarker in the Long COVID group that may or may not be replicated in larger trials.
gidmk.bsky.social
Of the many, many silly things in nutrition research, the fact that the Korean government regularly funds research that shows kimchi in a positive light is one of my favourites.

Hilarious stuff.
gidmk.bsky.social
Unlikely. I'll take a look.
gidmk.bsky.social
Astonishingly, since posting this comment I've already identified more errors in the paper. This statement does not match the very first row of the table directly underneath it, for example: