I apologize, I misread your initial argument. I thought you were defending the canavore diet, the subject of the article went through (there are definitely comments in this comments section making that claim).

The research behind ultra processing has legs from what I've read. My none expert take is that it has more to do with the fact that portion sizes are larger and caloric density is higher.

It doesn't take long to find videos of people from all over the world visiting the US and commenting on how much food is served.

That also jives with why drugs like semaglutide work so well at reducing weight.

Caloric density is also pretty high in traditional diets like those of the Hadza, the Inuit, the French, the Argentines, and the Italians, but those produce low levels of obesity. So that can't be it. Those diets also don't include semaglutide.

Large portion sizes are real, but they are downstream of whatever is unnaturally increasing US appetites. Restaurants, housewives, and househusbands in the US don't serve 1500-kcal meals because they're in some kind of conspiracy to fatten your kids up for slaughter. They serve them because people complain if they don't.

The "ultra-processed" meme is clearly pointing at something critical, but it can't itself be the answer, because it's ontologically incoherent. As the post I linked above says:

> I’ve previously found [this meme] frustrating – it reeks of a sort of unreflective technophobia. What part of processing makes food bad? How does mere contact with a machine turn food from healthy to unhealthy? What food counts as “processed” or “not processed”? Is ground beef processed, since you grind it? Are scrambled eggs processed, since you scramble them? Is bread processed, since wheat doesn’t grow in loaves? Is water processed, since it goes through water processing facilities? Is the Eucharist processed, even though the processing only changes its metaphysical essence and not its physical properties? Everybody I ask acts like the answers to these questions are obvious, but everyone has different answers, and nobody can tell me their decision procedure.

So, is it corn? Missing micronutrients? Industrial polysaccharide thickeners? Fungicides like propionate? Nanoparticles of titanium dioxide? Antinutrients like phytate and citrate reducing bioavailability of polyvalent-cation micronutrients? Some combination? Lots of people have promising hypotheses but nobody knows. Anybody who claims to know is either lying or insane, often both.

Agree that there are a lot of unknowns. Kevin Hall is carrying out some illuminating studies on the issue but, by his own admission, there’s simply not enough funding to do this at pace or scale.

I think we have some experimental evidence to suggest that at least some of the UPF associations are due to texture - those foods that you can basically inhale without chewing seem to drive higher caloric intake. We also have an incredibly strong signal of mass gain from sugar sweetened beverages which are also easy to consume with minimal effort.

I think it’s likely that cultural and economic factors play into this too. Japan has the same proliferation of UPFs, yet does not suffer the same obesity effects as the West does. There are probably a number of things that play into this: Japan has an incredible children’s food education policy, and they presumably have a different culture around food (many western cultures emphasise a “finish your plate or you’re ungrateful” approach).

I also have an intuition that the rise of two income households raising children is a factor in this. It’s a lot to ask of a couple that they both hold down full time jobs while cooking the kind of meals that used to be the domain of a stay-at-home parent.

Really interesting area to think about in terms of research and policies, but as you say we just don’t have strong evidence at the moment.

In the meantime it’s great that GLP-1 agonists are available to reduce the damage done, but it’s a shame we have to resort to such things in the first place.

We've had effective weight loss drugs for quite a long time; smoking tobacco and various phenethylamines are well known to work, for example, methamphetamine being the oldest. (I suspect chewing coca may also work.) Unfortunately, those also turned out to have harmful effects which weren't noticed until they had been in use for decades or centuries. It wouldn't be surprising if GLP-1 agonists also turned out to do more harm than good.

It’s certainly possible that there are long term side effects from GLP-1 agonists, but we can only go with the data we have. The five year data (which definitely has limitations in terms of the kind of inferences we can draw from it) is suggesting around a 4x reduction in mortality at the 5 year mark in non-diabetics IIRC (https://pubmed.ncbi.nlm.nih.gov/39171569/).

The changes in behaviour are also really interesting. Even after cessation of the drug, it seems like many patients keep healthier eating patterns and while they do put on weight, they tend to settle on a lower weight than baseline before intervention.

So we should absolutely keep an eye on outcomes as time goes on, but so far the results are overwhelmingly positive. These drugs are saving lives and the new ones in the pipeline look significantly more effective. It’s an exciting time to be alive!