Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Crémieux Profile picture
Crémieux
@cremieuxrecueil
I write about genetics, 'metrics, and demographics. Read my long-form writing at https://t.co/8hgA4nNS2A.
43 subscribers
May 19 • 5 tweets • 1 min read
The beta blocker propranolol reliably impairs memory consolidation. Image The above result is for healthy volunteer samples. This is the result for clinical samples.

It's effectively the same picture. Image
May 15 • 18 tweets • 7 min read
Today the New England Journal of Medicine published the second big win for lifesaving N-of-1 gene therapies.

They might have just saved a baby's life from being snuffed out by a fatal, ultrarare metabolic condition.

This is good for the baby, but maybe for many others too 🧵Image The deficiency the baby was born with is carbamoyl-phosphate synthetase 1 or CPS1 deficiency.

CPS1 is a mitochondrial enzyme that is involved in detoxification during the urea cycle. If you lack it like this baby, you tend to see hyperammonemia (ammonia buildup) right away. Image
May 14 • 4 tweets • 2 min read
The association between LDL reduction and coronary heart disease (CHD) risk reduction is sizable and confirmed by

- Prospective cohort studies
- Mendelian randomization
- RCTs

There really is no room to deny it. You can't get clearer than this: Image The evidence that LDL causes ASCVD is totally unequivocal, and minimizing LDL is essentially a free lunch for cardiovascular health.

Same situation with lp(a), so thank god new therapeutics are here that make both things trivial. Image
May 14 • 25 tweets • 10 min read
People might be able to limit the side-effects of GLP-1 drugs by avoiding Ozempic/Wegovy and instead using Mounjaro/Zepbound.

The reason has to do with Zepbound's other ingredient besides GLP-1: GIP🧵Image GIP is short for either gastric inhibitory peptide or glucose-dependent insulinotropic polypeptide.

It was originally called gastric inhibitory peptide, but people now prefer the 'insulinotropic' name because the gastric stuff rarely happens in normal circumstances. Image
May 13 • 9 tweets • 3 min read
🇺🇸🇺🇸🇺🇸

America's GLP-1 for weight loss is superior to Europe's GLP-1 for weight loss!

🇺🇸🇺🇸🇺🇸 Image If you've been following along, you already knew this.

Tirzepatide (Zepbound) and Retatrutide (no name yet) add GIP and GIP + Glucagon, respectively, making them more effective products. Not only that, but they might even have milder side effects.

May 12 • 4 tweets • 2 min read
Pharmaceutical R&D is on death's doorstep and its current rebound is fragile and temporary.

Returns are already below the cost of capital, and any additional harms to profitability will drain the life-blood of the future, instantly snuffing out biomedical progress. Image I am talking about the engine of survival, the thing that explains so much of why so many of us are alive today, and the thing that will keep you alive in the future.

This is also the thing that might eventually bring you immortality.

Destroy it, and we all lose.
May 12 • 13 tweets • 4 min read
Does the U.S. overpay for drugs?

Common knowledge says "Yes".

Real data says "No", because the U.S. gets favorable prices for generic drugs and it primarily consumes generic drugs. In fact, they're 91% of prescriptions! Image But then, we have a mystery:

If the prices America tends to pay are, in fact, reasonable, then why does it account for such a large share of all global spending on drugs? Image
May 11 • 10 tweets • 4 min read
Where are these guys from?

They're the famous Terracotta Warriors from the mausoleum of the first Chinese Emperor Qin Shihuang, and researchers have recently begun scanning their faces to learn more about them, possibly including their ethnic backgrounds🧵 Image It's unknown if these lifelike statues were based on real people.

But, under the assumption that they were, we can get a lot of information about who they were, because people's faces vary from group to group.

If you've played the game "Ethnoguessr", you're familiar with this.Image
May 10 • 8 tweets • 3 min read
PLACE YOUR BETS.

Will this result remain significant after I trim-and-fill it?

Current effect size: 0.8374 (0.5508-1.1240).

r(g, SE) = 0.5908. Image WILL THE RESULT REMAIN SIGNIFICANT?
May 9 • 11 tweets • 4 min read
Conservatives love to attack the Great Society as if it's responsible for modern high divorce and low marriage rates as well as high Black crime rates.

But this is a narrative-based belief, not a statistically-justified one. High Black crime rates precede the Great Society.Image The Great Society gets a lot of senseless blame.

Some people say it caused people to work a lot less. Not clearly supported:

May 7 • 5 tweets • 2 min read
Chinese born during the Great Famine were no more or less intelligent than earlier or later cohorts.

Observed score trends over time are largely illusory, with the potential to reflect forces like improved education, but not things like nutrition or outbreeding. Image And this makes sense.

After all, even extreme, extended malnutrition at various ages doesn't tend to affect the size of the head.

It affects height much more, because the body preferentially preserves brain mass.

Although even height resists transient nutrition shocks:Image
May 7 • 16 tweets • 5 min read
Roughly one-third of all of Japan's urban building was done through a process of replotting land parcels and reconstructing homes to increase local density while making way for new infrastructure🧵

Conceptually, it's like this: Image In that diagram, you see an area of low-density homes that has undergone land rights conversion, where, when two-thirds of the area’s existing homeowners agree, everyone’s right to their land is converted to the rights to an equivalent part of a new building. Image
May 6 • 18 tweets • 7 min read
Chinese cities occasionally sell land to developers before buying out all the existing residents.

But sometimes existing residents refuse to be bought out, so developers are forced to build around them.

These are "nail houses"🧵 The most famous nail house is undoubtedly Wu Ping's home in Chongqing.

Wu Ping came to national acclaim when she and her husband refused to give up their property to make way for a luxury apartment complex.

They wanted more compensation, and they fought for it for three years. Image
May 6 • 27 tweets • 13 min read
I've seen a lot of people recently claim that the prevalence of vitiligo is 0.5-2%.

This is just not true. In the U.S. today, it's closer to a sixth of a percent, with some notable age- and race-related differences.

But where did the 0.5-2% claim come from?🧵 Image The claim of a 0.5-2% prevalence emerged on here because Google's Gemini cited a 2020 review in the journal Dermatology which proclaimed as much in the abstract.

Simple enough, right? They must have a source that supports this estimate in the review somewhere.Image
May 2 • 6 tweets • 3 min read
There's a myth that the Islamic world has figured out fertility, but it has not.

They show the same declining fertility rates that other places have. Barring Iraq, the Middle East has lower fertility rates than Israel now. Image Exceptions: Yemen and maybe Palestine, both of which have terrible data, so their comparative situation is unclear.

But, two things on that:

Firstly, Jewish fertility is ahead of Arab fertility in Israel. Image
May 2 • 6 tweets • 3 min read
Relationships between class and fertility and IQ and fertility used to routinely be negative in the not-so-distant past.

But across the developed world, they're increasingly positive, albeit only slightly. In this Swedish birth cohort (1951-67), the transition came early: Image In this example, there's also some interesting confounding: between families, IQ isn't monotonically associated with fertility, but within families, it is.

Something seems to suppress the IQ-fertility relationship between families!

See also:
May 2 • 10 tweets • 3 min read
One of the reasons people are so pessimistic about fertility policy is because they misjudge the counterfactual🧵

Consider this. We have a country with a given fertility level: Image The country intervenes with some fertility policy, and the fertility rate continues to fall.

The program is therefore dubbed a failure. Oh no! Image
Apr 30 • 7 tweets • 2 min read
This is a really strong claim based on really scant evidence.

Add in a control for family history or use Bonferroni instead of Benjamini-Hochberg and 5-aminovaleric acid betaine goes nonsignificant. Add in polygenic risk scores too and Cyclo(Leu-Pro) goes nonsignificant. Using a small number of the total tests (multiple comparison correction was too lax), the model with both metabolites in it alone led to p-values of 0.3512 for 5-AVAB and 0.0188 for Cyclo(Leu-Pro) and that's from a model without family history or genetic risk.
Apr 29 • 22 tweets • 9 min read
The Mafia is undoubtedly cool.

It makes for good TV and good movies, and some even argue that it makes for economic growth, that it 'greases the wheels'.

But I've never believed this theory, and I think there's considerable evidence against it🧵 Image Italy is the homeland of the Mafia, and though they've tried everything to get rid of them, they're still around.

Check this date out: They're still doing anti-Mafia stings in 2025!Image
Apr 25 • 16 tweets • 5 min read
Why have autism rates risen over time?🧵

I have just put out an article dealing with numerous misconceptions about this topic, and a complete explanation of why autism diagnoses have become more common.

It starts with acknowledging that more kids are diagnosed than in the past: Image But this is misleading for a few reasons.

One has to do with how this data was sourced. We didn't have a DSM with autism in it before 1980, so all the oldest people in this cohort were diagnosed as adults.

Adults are underdiagnosed. Go out of your way to diagnose? Same rates.Image
Apr 24 • 5 tweets • 2 min read
In 2016, researchers found that the minority-White wage gap was overestimated by about 10% because, at work, non-Whites tended to partake in more leisure, waiting around, etc.

They delayed releasing the study out of fear Trump would "use it as a propaganda piece." Image They explicitly admitted that they let their personal politics get in the way of releasing a study with contentious but correct findings.

That doesn't inspire trust, but at the same time, given the topic, it might!