25 Comments
User's avatar
Frankenheimer Graff's avatar

Nothing about serious adverse side effects?

Please read: https://www.midwesterndoctor.com/p/the-great-ozempic-scam-and-safer and https://jamanetwork.com/journals/jama/fullarticle/2810542.

GLP-1 users were found in the latter study to have:

9.09 times greater risk of pancreatitis

4.22 times greater risk of bowel obstruction

3.67 times greater risk of gastroparesis (which means you can barely eat because the stomach is constantly full—and in many cases after Ozempic, ends up being permanent)

1.48 times greater risk of biliary disease (e.g., painful gallstones).

Expand full comment
Realist's avatar

I am very leery of the 'A Midwestern Doctor'; he seems to be against all modern medicine, a kind of medical Luddite, a home remedy advocate.

Expand full comment
The Gray Man's avatar

These are not good studies that don't take into account other issues. I hope you're not one of those who go around trying to stoke fear about this with obscure studies that don't actually represent the reality: these are some of the safest drugs you can possibly take

Expand full comment
Realist's avatar

Agreed, see my comment.

Expand full comment
Frankenheimer Graff's avatar

You probably said the same thing about the Covid shots. The JAMA study covered 16 million patients. It doesn't get much better.

Expand full comment
Cremieux's avatar

Having a large sample size doesn't obviate methodological concerns, like with insufficient matching.

Try reading something like https://www.nature.com/articles/s41586-021-04198-4

Expand full comment
The Gray Man's avatar

I did not, which is why you're showing you don't read studies and instead are anti medications as a personal identity trait.

Can you comprehend that? Anti vaxxxx (for better reasons than you, I can tell), and pro GLP-1?

Expand full comment
Cremieux's avatar

"Please read:"

Your first source is ignorant schizopoasting. The person's post shows they're a conspiracy theorists who doesn't even have a grasp on what discontinuation means and makes The Worst Argument Against Ozempic: https://www.cremieux.xyz/p/the-worst-argument-against-ozempic.

Your second source shows you missed the memo on residual confounding: https://www.cremieux.xyz/i/168430028/matching-is-usually-insufficient. You can see the error by looking at some of their findings like their absurd result for pancreatitis. Compare their result to condition-matched samples (https://x.com/cremieuxrecueil/status/1948127978195386629) or trials (https://x.com/cremieuxrecueil/status/1948127991352983754).

Think more about methods before getting overly concerned.

Expand full comment
Unirt's avatar

Pancreatitis seems the most prevalent of these conditions, at 33 cases per 100,000 globally. So these drugs may raise it to almost 3 per 1000. 3 per 1000 is still a low-ish number, and most people would take that risk to get rid of obesity, which causes higher risks of several other conditions. Diseases with a prevalence of 3 per 1000 tend to be those that we laypeople haven't heard of much, as they're so rare.

Expand full comment
Ted Hyman's avatar

Apart from you drawing conclusions from one retrospective study, you need to actually read the papers you cite:

“This study found that use of GLP-1 agonists for weight loss compared with use of bupropion-naltrexone was associated with increased risk of pancreatitis, gastroparesis, and bowel obstruction but NOT biliary disease.”

(Emphasis mine)

Expand full comment
Epaminondas's avatar

Interesting article. However, I thought there were around 20-25% of people who simply don't respond to GLP-1s at all. This is just more depressing news for them, but also scientifically intriguing. If all of these mechanisms that you walked through don't prevent GLP-1s from working, I wonder what is going on with that population.

Expand full comment
Cremieux's avatar

Over a meaningful timeframe, there are incredibly few people who do not respond to modern GLP-1RAs. I've looked in EHR data and found <1% of those prescribed fail to lose weight if you have measurements of them after six or more months. The belief that there are so many failures is based on studies like this (https://pmc.ncbi.nlm.nih.gov/articles/PMC11751938/), which involve subsetting away large shares of the population and failing to note things like that *by far* their most powerful moderator of treatment success was maximum drug dosage used.

In that study, relative to a dosage of ≤0.5 mg, the chance of being a <5% responder was about 66% lower at 1 mg (RR ≈ 0.34), 84% lower at 2mg (RR ≈ 0.16), and 77% lower at 2.4mg (RR ≈ 0.23). The fraction of nonresponse attributable to low maximum dosage (≤ 1mg) in this sample comes out to ~40-43% of all nonresponders if you compare the people taking those doses with those taking 2 or 2.4mg. If you look at the raw composition, the 'nonresponders' (who still might be responders if we follow up for longer) were 71.3% users of such low dosages. Noticing the nonlinearity when we know the dosage effect is linear in trials, we can also see there's confounding to consider, which these authors practically did not.

Expand full comment
Unirt's avatar
5hEdited

I understand that the "no excuses" in the title is an artistic exaggeration. Still, unfortunately, even though these drugs usually lead to weight loss, it doesn't mean weight loss until you're perfectly lean. Fat people who lose 6 lbs are still fat people, and may feel like they need an excuse for that. Of course, it's better than nothing, and the good effects of these drugs seem to outweigh any bad ones heavily.

Expand full comment
Kayla's avatar
5hEdited

GLPs won’t eliminate obesity. They lead to typically 5-25% of body weight lost. That’s GREAT, it’s more effective than almost anything else we know, it’ll improve health…but if you start out at 400 lb and go to 300 lb, you’re still gonna be pretty fat. Even 50% loss isn’t enough to take some people from morbid obesity to normal weight.

Expand full comment
John Wilkins's avatar

Thank you for sharing this. I’m on semaglutide for Type II diabetes, and this clarifies things for me. One point, though: why is it an “excuse”? Is weight a moral or legal, rather than a medical, issue? That seems to me (philosophy doctorate) to conflate value systems.

Expand full comment
awarre1's avatar

"they have taken away every last excuse to be fat." Except poverty.$1000/month is more than many can do.Might be cost effective for Medicare/Medicaid though-less strokes heart disease diabetes etc etc.

Expand full comment
PdC's avatar

The pre-existing conditions 'excuse' is not insignificant. I, like many, many millions of people, have a condition that causes life-threatening arrhythmias. As far as I've understood, GLPs can make these worse. Hopefully they'll develop GLPs that aren't problematic in this way but until then, lots of people are stuck.

Expand full comment
Cremieux's avatar

Yeah, the currently-approved batch is likely *helpful*. But I would still exercise caution until we have more data about the new ones that contain glucagon.

Expand full comment
Ted Hyman's avatar

Caution, absolutely. But avoiding them entirely is likely detrimental, as you said. Even if they DID slightly raise the risk of arrhythmia, most people don’t die from arrhythmias (we’re good at treating that). Conversely, people die often from complications of diabetes, obesity, heart disease, kidney disease, and other things that GLP-1As help with. So, like everything in medicine, patients and doctors need to weigh the risks against the (substantial) benefits.

Expand full comment
Cremieux's avatar

I agree that even the GCGN-containing GLP-1RAs are probably beneficial for even obese people with prior afib on net.

Expand full comment
Ted Hyman's avatar

Good article that summarizes the data well. My one issue is the use of the term “schizophrenics”. It’s generally accepted now to use person-first language in medicine (e.g. “persons with schizophrenia”), since labeling someone as “a schizophrenic” defines them by their disease.

https://www.nih.gov/nih-style-guide/person-first-destigmatizing-language

Expand full comment
Realist's avatar

Great article.

Expand full comment
Amanda From Bethlehem's avatar

Now if only the FDA would lighten up about GLP-1RA use during (and just before) pregnancy.

The current FDA guideline is to stop semaglutide two months before trying to get pregnant. Two months! Except when you're trying, those "two" months can easily turn into three, or four, or even longer. Before you know it you've binge eaten 20 pounds back, with nothing to show for it.

Oh, and the best part? Semaglutide might require two months, but tirzepatide doesn't actually have a stoppage timeline from the FDA. Doctors just repeat "two months" to you, like for semaglutide, even though tirzepatide leaves your system in only one month. (The Canadian safety doc says one month.)

Novo Nordisk has an ongoing "semaglutide exposure" registry/study for women who accidently get pregnant while taking it (and then immediately stop). They haven't read out any preliminary results yet. (Unfortunately, afaik Eli Lilly isn't doing a similar study for tirzepatide.)

Is anyone working on a preemptive safety study for pregnancy outcomes in women who have differences in the GLP-1RA gene? That would put all this paranoia to rest.

Expand full comment
jbnn's avatar

If it eats like a hippo it should look like a hippo.

Expand full comment