With an endorsement from Derek Lowe, my opinion is superfluous, but yes, this study looks legit. Semaglutide causes weight loss in a large, rigorous controlled trial. 🧵follows.
1961 participants, randomized 2:1 to semaglutide or placebo, plus a 500-calorie deficit and exercise program. Mostly white and Asian, mean BMI 38.
Over 68 weeks, the treatment group lost a mean of 16% of their body weight, or 37 pounds. The control group only lost a mean of 3% or 7 pounds.
This is about half a pound a week of weight loss. That's comparable to the effect size of topiramate, the drug that causes the largest amount of weight loss per week I've previously found in the literature. onlinelibrary.wiley.com/doi/pdf/10.111…
(Nobody takes topiramate just for weight loss, though; it's an epilepsy drug with some funky side effects, including seriously impaired cognition. The hope of obesity-drug researchers is to find something that healthy people can take safely.)
Back to semaglutide. It does some other nice things: mean reduction of waist circumference by 5 inches, and significant reductions in fasting glucose, blood pressure, cholesterol, triglycerides, and CRP.
Side effects are common but mild. 74% of semaglutide patients had some kind of gastrointestinal problem, like nausea, diarrhea, constipation, or vomiting. Increased risk of gallstones in the treatment group, in 1.8% of participants.
So what is semaglutide, anyway? It's a diabetes drug that stimulates the production of insulin by binding to the glucagon-like peptide (GLP1) receptors.
GLP-1 also reduces appetite, possibly by slowing gastric emptying (food remains in the stomach longer) and maybe also by doing something to the hypothalamus, where it's expressed. en.wikipedia.org/wiki/Glucagon-…
Animal experiments on GLP-1 show that the central nervous system effect works to reduce appetite even in the absence of the peripheral/gastric effect, and lesions to the hypothalamus-brainstem connections reduce the satiety effect. ncbi.nlm.nih.gov/pmc/articles/P…
(In other words, the hypothalamus effect is both necessary and sufficient to the appetite-suppression effect of GLP-1. God, I love the hypothalamus.)
Anyhow. Somebody once asked me to research DNP, the supposedly "super extreme" weight loss drug that makes ATP production less efficient and can make you go blind or die. en.wikipedia.org/wiki/2,4-Dinit…
Well, semaglutide causes more weight loss than DNP, and doesn't kill you.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
I could not find one research study using any of the peptides in the RADVAC white paper that found they inhibited SARS-CoV-2 infection in cells, let alone animals or humans.
All those peptides come from in silico studies: “the computer said they ought to bind to various viral proteins.” Plus a lot of theory/mechanism argument.
You have the legal, and IMO, the moral, right to experiment on yourself. But I don’t think this is very likely to work.
The most irritating thing about smartphones, to me, is that it’s more difficult to switch from one webpage to another on mobile than in a desktop browser. It makes it hard to do something like “take notes including links from a variety of websites.”
The second most annoying thing is that I can’t log into sites on my phone if their passwords are the long forgettable strings of letters and numbers that my laptop password manager remembers for me.
The third most annoying thing is that i’m slightly slower at typing on a phone than a keyboard.
The trend in deep learning for a lot of applications, for most of the past decade, seems to have been “you get out what you put in” — performance gains are proportional to increases in computing power.
I haven’t found anything I’m confident is a exception to that trend, in the direction of “performance grows faster than compute”. I’d be willing to bet that there aren’t any.
As long as that continues, it seems to me that the main question is how fast the cost of flops drops and how long it will continue to be profitable to keep buying more & better hardware.
This post on Roman-era peasants acoup.blog/2020/07/24/col… has made me question how I think about technology and economic growth.
My model had been something like “humanity used to be poorer because nobody had yet invented the technologies that could make more food and other goods with less labor.”
But this post is saying that a Roman peasant probably wouldn’t *want* a technique that made his yields 10% larger even if a time traveler magically handed it to him.
Keep in mind this is still 8 years from “idea for research program” to “drug in preclinical testing”. Biology is slow even if you have exceptional talent & drive.
The drug candidate is an “antisense oligonucleotide”, which is actually a fully general strategy for treating genetic diseases. So far seven such therapies have been approved.