That's all good stuff, but you're leaving out one key detail in your analysis. GLPs reduce death (not to mention all the other factors you brought up) relative to poorly controlled diabetes. So if someone has poorly controlled diabetes then your analysis is absolutely correct, since that is what we currently have long-term data on. If someone doesn't have diabetes, it's quite possible GLPs will be found to reduce death, but that's still speculative at this point.The thing that really stands out to me from the research is convincing and repeatable evidence that semaglutide and tirzepatide reduce overall death rates. The younger and healthier the population involved in the studies the harder it will be to see this signal, and it requires lots of patients over a long time. The longest term studies show reduced chances of heart attack, stroke and death, by around 25 to 30%. They reduce the chances of a lot of other things, the development of diabetes in those with impaired glucose tolerance is reduced by about 90%. The results on cancers is more complex but there is pretty good evidence of reduced chances of developing many of the most common cancers. For Alzheimer's the evidence so far is mixed, population studies show significantly reduced risks, but prospective trials ( which are more accurate ) so far have not shown any benefit. They improve quality of life, mobility and arthritis mainly due to weight loss. This post would be pages long if I tried to include them all. The number of studies being published on these sort of related consequences is enormous ( many hundreds to a 1000 per year in 2025 ), and over the next few years the picture will slowly get clearer.
Shorter term negative effects are very common and cause a fairly high percentage of people to stop taking them, mainly nausea and vomiting diarrhoea and constipation, as well as skin sensory symptoms. Most of these type side effects peak at months 2 to 4 after starting and then improve, but not for everyone. Most of these effects are not severe, but some people develop dehydration and subsequent renal failure from gi side effects, and some die, most at risk are older physically and medically frail persons and diabetics with multimorbidity.
Are there negative effects in the long term - yes Gallstones are a consequence of any type of weight loss, sudden improvements in diabetic control can cause neuropathy, pancreatitis, nerve compression syndromes due to weight loss, NAION, plus many more, but I think the important issue is that the sum of all of the negative long term consequences is quite small, and the sum of the positive ones is very large, in terms of order of magnitude effects positive effects outweigh negative ones by more than 10 to 1. The fact that overall chances of death go down, essentially means that there are no hidden or unknown negative effects that increase your chances of death. A lot of the studies I have seen that emphasise the long term negative effects or concern about unknown long term effects, as far as I am concerned seem to be mostly ignoring the large amount of evidence that already exists, and mostly seem to be coming from areas of medicine where their whole job is questioned by glp's mainly dietitians ( whose treatments for obesity do not work over the long term as far as I am concerned ) and weight loss surgeons ( whose treatments do work and very well, but are expensive, have limited access, have significant adverse effects, and whose treatment method is being questioned by increasingly effective glp medications )
TLDR These medicines make you significantly less likely to die while taking them, this alone massively outweighs any long term known or unknown harms.
A similar logical error was made a few years back when it was announced that diabetics on metformin monotherapy lived longer on average than the general population. This was a correct observation, but people erroneously took it to mean that metformin more generally had life-extension properties. When metformin use in non-diabetics was studied, it was found that it didn't appear to lead to life-extension in non-diabetics.
Now, I'll grant you that one key difference with GLP1s is that they also serve to reduce other intermediate markers (e.g. inflammation, lipids, BMI, etc.) in non-diabetics, which supports the idea that they'll be beneficial in non-diabetics. I want it to be true just as much as you do, but unfortunately our desires are not a replacement for long-term data in non-diabetics.