Perhaps vastly more people should be on FDA-approved weight loss medication

[Epistemic Status: I feel pretty good about most of this, but the life-years-saved-via-medication part is problematic on a number of levels, as pointed out by a few commenters. I include it since back-of-the-envelope calculations serve a purpose in ensuring we’re comparing effects of approximately the appropriate magnitudes in doing risk/​benefit analyses, but I wouldn’t take it too seriously.]

Note that I’m not a doctor. Please speak to a doctor before doing any of this stuff Or You Will Die.


Judging by posts in r/​loseit, the existence of effective anti-obesity medications is not particularly well-known (and to the degree it is well-known, it’s disapproved of.) Even posts on LessWrong, which tend to be very well-researched and exhaustive, simply ignore the topic of medication when weight loss methods or obesity are brought up; I suspect this is not because their authors had explicitly considered and discarded the various anti-obesity drugs currently available, but rather, because the existence of these drugs is very poorly-known. Which I’m attempting to remedy here! At least for the LessWrong crowd.

Quantifying Life-Years Saved by Losing A Certain Amount Of Weight

[Note: as pointed out by comments below, extrapolation to life-years saved is very speculative, since all the studies on this in humans are going to be confounded all to hell by healthy user bias and socioeconomic correlations and the like. That said, it feels like a fairly reasonable extrapolation given the comorbidity of obesity to various extremely problematic medical conditions. Be warned!]

According to Genome-wide meta-analysis associates HLA-DQA1/​DRB1 and LPA and lifestyle factors with human longevity | Nature Communications , losing a single unit of BMI roughly corresponds to a 7-month gain in life expectancy in the overweight and obese. This seems basically in line with what I hear from popular sources, such as: “[L]ife expectancy for obese men and women was 4.2 and 3.5 years shorter” than people in the healthy BMI range.

This won’t count as a revelation. Obesity is unhealthy, news at eleven. My goal here is just to quantify what you’re getting relative to the risks involved in doing something to ameliorate it.


The U.S. Food and Drug Administration (FDA) recommends pharmacotherapy for weight loss when lifestyle interventions (diet, exercise and behavioural therapy) have failed and the body mass index (BMI) is °30kg/​m2 with no concomitant obesity-related risk factors, or if the BMI is °27 kg/​m2 and the patient has at least one obesity-related risk factor.

So: let’s talk about weight loss drugs!

Weight Loss Drug Studies

Weigh loss drug studies are always composed of two groups of patients: a group attempting guided diet and exercise along with a placebo pill, and a group doing the diet and exercise plus the drug. That’s important here, since it means we can’t unequivocally recommend drugs as a replacement for diet and exercise, only as a secondary treatment. (Aside: even though basically every article on weight loss is obligated by eternal law to pay tribute to exercise, the evidence for it helping with weight loss on a practical level is minimal.)

These studies are saying, in effect: if you can get X pounds lost from diet and exercise alone, adding pharmaceuticals to these efforts can get you X+Y pounds lost.

That being the case, I’m going to now list the three or so good (as judged by me, a random asshole with a laptop) FDA-approved anti-obesity drugs currently on the market right now; their measured diet-and-exercise-subtracted weight loss; and finally, the amount of life-years you can (maybe? who knows) gain over the long term by losing that much weight. I’ll be linking to studies for each.

Note on drugs I’m not discussing here: I’m not going into liraglutide since it seems basically like worse semaglutide at similar cost, and I’m not going into phentermine+topiramate (Qsymia) because in spite of its greater efficacy than phentermine alone, it seems that topiramate has a substantial likelihood of giving people kidney stones and brain fog, which are… not great. Orlistat is quite popular, but has relatively poor efficacy and unpleasant digestive side effects. Links provided on request, but that’s a bit far afield of my purposes here, so I’ll move on.

The Drugs (at least, the better ones)

Semaglutide (2.4 mg)

  • Price: 1300ish dollars per month for Wegovy. I’ve heard insurance has a… spotty… record of covering this. You might have better luck with insurance (provided you have T2D, or at least are at risk for it) with Ozempic, which is the same semaglutide, just at a different dose and with labeling for T2D treatment.

  • Mechanism: GLP-1 inhibitor; more specifically, it slows gastric emptying resulting in lowered appetite.

  • Average Diet/​Exercise-Subtracted weight loss: 12% based on its phase-3 trial. This is the most potent anti-obesity drug on the market.

  • Common Side Effects: Transient nausea and GI upset at treatment onset.

  • Other Notes: This is basically just higher-dose Ozempic, which has been on the market about four years.

  • Approximate BMI drop for a 5’6 female at 200 pounds: In weight, 12% weight loss equates to about 24 pounds. This is a drop in BMI of 32.28 to 28.4 units.

  • Approximate difference between expected life-years of people with these two BMI values: About 28 months, or about 2.3 years.

Contrave [Bupropion + Naltrexone]

  • Price: If you get it generic (and why wouldn’t you?) about 40 bucks a month as naltrexone + bupropion.

  • Mechanism: Poorly-understood neurochemical effects.

  • Average Diet/​Exercise-Subtracted weight loss: 3-7% (varies by study)

  • Common Side Effects: Amped up sex drive and improved focus (Bupropion is sometimes used off-label for ADHD); on the other hand, anxiety and insomnia, plus transient nausea at treatment onset. [My own bias: I’m on bupropion and it’s mostly kickass. Insomnia’s no fun, though.]

  • Other Notes: Both parts of this drug have been in common use for several decades. If there was some godawful long-term side effect we’d know about it by now.

  • Approximate BMI drop for a 5’6 female at 200 pounds: For lower estimates, this is a drop in BMI of 32.28 to 31.31 units (so about 1 unit of BMI); for higher (7%) estimates, this is about 2 units of BMI.

  • Approximate difference between expected life-years of people with these two BMI values: About 7-14 months of life.


  • Price: 23 dollars/​month

  • Mechanism: Stimulant. Most stimulants have weight loss as a side effect; this is just one of the few the FDA has actually approved for the purpose.

  • Average Diet/​Exercise-Subtracted weight loss: 3-7% (varies by study)

  • Common Side Effects: This is a mild stimulant, so… pretty much what you’d expect.

  • Other Notes: Technically any use of this longer than 6 months is off-label (the FDA hates stimulants), but several long-term studies of phentermine use find no evidence of addiction or other side-effects when taken for years. Anecdotally this is sometimes taken with Contrave, but this is an off-label combination of drugs on which there is little data. On the other hand, there doesn’t seem to be any a priori reason to expect this combination to be harmful?

  • Approximate BMI drop for a 5’6 female at 200 pounds: For lower estimates, this is a drop in BMI of 32.28 to 31.31 units (so about 1 unit of BMI); for higher (7%) estimates, this is about 2 units of BMI.

  • Approximate difference between expected life-years of people with these two BMI values: About 7-14 months of life.


Overweight and obesity cause a lot of misery! I’ve lurked r/​loseit! And the quest to stop being overweight is the cause of even more misery for lots of people. If you’re in that group, you might be well-served by discussing medication-assisted options with a doctor.


Giving a life-years-saved number for if someone takes a drug implies they’ll be on it forever. But what about the unquantified risks of being on some drug for the rest of your life? Especially semaglutide, which hasn’t been around very long?

This is a fair concern! It is, however, worth pointing out that the FDA is vigorous about pulling drugs that have been shown to have even small risks of causing life-threatening conditions; a recent example of this is lorcaserin (aka Belviq), which was taken off the market due to a non-statistically-significant increased risk of cancer. See also: Is lorcaserin really associated with increased risk of cancer? A systematic review and meta-analysis—PubMed (

Think about the implications! If you’re on X drug for your whole life, then by assumption you’ll have also gone your whole life without the FDA having observed any statistical increases in cancer incidence or heart attacks or whatever for people on the drug. That’s a very high bar of safety.

Ultimately, the quantifiable life-years lost by obesity (in the form of statistical heart attacks and various other comorbidities) must be weighed against the mere uncertain prospect of an imperfect drug making it through the FDA approval process.

Besides which, nobody says once you’re done losing weight that you have to continue taking the appetite suppressants. I mean, I probably would? But diff’rent strokes.

If you’re concerned regardless—semaglutide is the only particularly new treatment on that list (and even that’s been around a few years in the form of Ozempic). The others have multi-decade histories of usage, with reams of literature on their effects. Google Scholar: your friend and mine.

Isn’t just eating less a much healthier and better-proven means of weight loss than pills?

Nicotine addiction wears off with time. If a person can keep off cigarettes and other nicotine sources for about three months, most surveys show that this leads to a total cessation of a desire to smoke or otherwise consume nicotine.

That being the case, it’s obvious what addicted smokers have to do to cure their addiction: stop smoking cigarettes for three months. Withdrawal is unpleasant, but nevertheless this method is uncomplicated (step one: don’t smoke, step two: nothing), extremely cheap as an intervention, and guaranteed to work if performed.

And that is why, even to this day, nobody is addicted to cigarettes.


I guess a less snarky answer is that these medications mostly work by making it more pleasant to eat less.

Isn’t this just a way of letting lazy people off the hook?

Eh. If you’ve tried it and straightforward dieting makes you miserable, you are under no obligation to power through without assistance. You don’t win virtue-points for avoiding medication that makes your life easier even if online randos imply otherwise.

If these meds are so great, shouldn’t I have heard about them by now?


First: American society has a pretty weird relationship with weight loss; there’s a huge implication in the discourse that thin-ness is a result of righteous self-discipline, and that fat people just need to buckle down and make the effort, and if they fail then they just weren’t trying hard enough. (This viewpoint is neatly encapsulated in the slogan “eat less move more” and concepts like “the physics diet”) Accordingly, weight loss drugs have acquired the implicit moral status of a cheat enabling one to get the reward without the suffering, which people are suspicious of.

Second, there are also some now-banned medications—fen-phen and DNP are pretty good examples—that are both (1) deadly and (2) highly effective at weight loss. Thus, the popular perception that anti-obesity drugs are intrinsically dangerous, to be used by people who value their appearance more than their health.

This isn’t helped by all the truly worthless herbal supplements on the market claiming to be effective weight loss aids; unlike for most other medical conditions, herbal supplements are allowed to claim that they’ll help with obesity (mostly implicitly by calling themselves “fat burners” and the like). Legitimate drugs can be difficult for uneducated audiences to distinguish from snake oil, so they get rounded off to “snake oil”.

These factors have resulted in society collectively memory-holing this entire class of medication.

I had bad experiences on phentermine.

You and a bunch of other people!

Most people have some drugs they’ll find unpleasant or that don’t work for them. It seems broadly reasonable to just try different drugs until you find ones that work for whatever condition you’re trying to alleviate; the potential risk is one or two weeks of discomfort while the drugs slowly exit your system (after which you move on to something else), and the potential reward is life-years saved from obesity comorbidities, as well as whatever added happiness you get from being at a lighter weight.

[Insert Certification Body Here] doesn’t think [Insert Drug Here] passes a cost/​benefit analysis, even though the FDA does.

If you’re able to read and evaluate the primary literature on this topic, I see no reason to outsource your cost/​benefit analyses to some other decisionmaking body rather than evaluating the drugs on their merits based on clinical trial data.

Ultimately, the decisions of these institutions will be colored by a lot of factors that aren’t “patient wellbeing”—by blameworthiness, PR considerations, the perceived second-order effects of their actions, and the crucial distinction between dead-their-fault and dead-not-their-fault. These tend to lean in the direction of “don’t certify the medication,” especially for anti-obesity drugs (which tend to be viewed as lifestyle drugs rather than drugs for legitimate illnesses, and so face a higher bar of scrutiny.)

Shouldn’t we just, as a society, just develop a healthier culture around food? Wouldn’t that be better than medicating ourselves?

There are a lot of things I would change about society if I were made Benevolent Dictator For Life.

Aren’t you, personally, just a lazy trash-person?

Absolutely. But that’s unrelated.

I found a factual error in this post.

Leave a comment telling me this (including a source for the info) and I’ll correct it!

Are you a doctor? You’re a doctor, right? You’re probably a doctor, so I should take this as medical advice.

Jesus Christ no. If you take any medication here without talking to a doctor about it you’ll definitely swell up and die, or possibly turn inside-out. Luckily, since I pointed this out in this paragraph I will be absolved of all responsibility.

Good luck!