I’ve been doing my own seed oil / obesity investigation for several months now, and I must commend this post for covering all of the major points. My only gripe is that I believe most meta-analyses are wrong because they don’t weed out the bad studies (e.g. ones that are poorly designed from the outset or that mistakenly confound their analysis by, say, lumping omega-3 and omega-6 PUFAs together). I imagine the meta-analysts would want to remove these problematic studies so I understand that there are limits to what can be measured.
I posted a comment similar to the following on the Substack version of this post but was hoping to get some more feedback from the LW community:
One thing that puzzles me is that calories in the food supply, as well as calories in food diary data, have basically flattened in the US since 2000 but obesity has doubled since then. I’ve put together data from several US government sources in this writeup (all data + R code is available in this github repo).
From a raw data standpoint, per-capita supply of vegetable oil much more closely tracks obesity than does per-capita supply of calories or caloric sweeteners. People are also exercising a lot more than they used to (which one would expect given the public health advice to “eat less, move more”).
While this is prima facie evidence in favor of seed oils, we also know that seed oils are combined with lots of other potential toxins in processed foods. If you haven’t seen this study yet, it definitively shows that “ultra processed food” has a very low omega-3:omega-6 ratio, causes weight gain, messes with satiety and hunger hormones, and affects metabolism.
The question then becomes “what exactly is it about ‘ultra processed food’ that is so bad?”
Seed oils or some form of omega-3/omega-6 imbalance?
MSG or other flavor enhancements?
Modified starches? [which are basically alternative forms of MSG]
Lecithins and other emulsifiers? [mono and diglycerides]
Artificial colors?
To answer this question we’d somehow have to make an “ultra processed diet” that was identical in content to the status quo “ultra processed diet” but didn’t include the above substances. Which is likely infeasible.
The other major question I’m grappling with is why there is an obesity-elevation gradient. This has been litigated on LW in the past and my preferred theory at the moment is that there is some sort of relationship between hypoxic environment and cell metabolism that favors metabolic efficiency. However, it’s not like high-elevation places are immune from obesity; they just have lower rates and more or less increase in parallel with low-elevation places. However, elevation isn’t the whole story because there’s plenty of low-elevation places that also have low-obesity. I don’t think seed oils fit into this at all unless it is through how they influence cell metabolism.
...which leads me into my recent reading of Anthony Hulbert’s Omega Balance. I’ve found it useful for understanding some basic biochemistry of cell membranes and how cell membrane fatty acid composition might interfere with basic metabolic processes. He probably oversteps the evidence with some of his arguments. But as a result of reading it, I’ve spent the last week eating high-omega-3 by focusing on consuming lots of flaxseed and chia seed. (It turns out a basic-sized salmon filet has <1g of omega-3 so fish isn’t the most concentrated source of omega-3 even though it is for “long-chain” omega-3.) So far I’ve been pleased with the results: I feel like I’ve slept better and I’ve dropped a few pounds. Tonight for dinner we had people over and I ate a slice of Marie Calendar’s cherry pie that has immediately given me a stomach ache. What does the pie have a lot of? Modified starch and seed oil.
I’m surprised that the AHA and similar continue to push high-omega-6 oils. Yes, they reduce cholesterol, but they seem to increase a lot of other long-term disease risks. I view the focus on cholesterol to be short-sighted and mistaken. But I’m no biochemist or physician.
The other major question I’m grappling with is why there is an obesity-elevation gradient.
A guy is going alone through the wilderness, with a solar powered icebox on his back. He crosses a raging river by swimming. He slashes his way through a jungle. He is blasted by sun on an endless desert. It’s been weeks and he has no company at all, save the bleached bones (and ice boxes) of those who did not make it. He climbs a mountain until he finally comes to a cave in the snow. Inside is a man with beard like silvery horsehair, eyes like fire. Very old but fit as a mountain goat.
“O wise sensei, I brought your pizza and ice cream with me. Now tell me the secret of perfect health!”
“It is vely simple”, says the old man while greedily unwrapping a stick of icecream. “Only meet with people who have made a journey such as you youlself just did.”
More prosaically: “It” does not run uphill.
~90% of Earth’s mammalian excreta is produced by humans and their livestock. The livestock especially are immobilized in close quarters and their manure is spread to fields by mechanical means. The manure is often in a fresh condition with viable gut flora present. This means that the fitness of the gut microbiome is independent of the host’s mobility and fitness. They can find plenty of new hosts nearby, and have a good chance of spreading far and wide with tractors. If it can make the host eat x% more, it gains roughly that big an advantage: that much more manure along which it can spread. In state of nature that tactic obviously would not work, it would be much better to have a slim host with good legs.
Selection for larger animals, antibiotic feeding and selection by market forces could also contribute towards more hunger-inducing gut microbes.
Maybe overweight is not the only hit on the host’s well-being. There could be other pathologies by which the mechanised microbes improve their fitness. If this is the case, exposure to bad manure – or biome that derives from it – would be a common cause for both obesity and some other diseases associated with it.
I’ve been doing my own seed oil / obesity investigation for several months now, and I must commend this post for covering all of the major points. My only gripe is that I believe most meta-analyses are wrong because they don’t weed out the bad studies (e.g. ones that are poorly designed from the outset or that mistakenly confound their analysis by, say, lumping omega-3 and omega-6 PUFAs together). I imagine the meta-analysts would want to remove these problematic studies so I understand that there are limits to what can be measured.
I posted a comment similar to the following on the Substack version of this post but was hoping to get some more feedback from the LW community:
One thing that puzzles me is that calories in the food supply, as well as calories in food diary data, have basically flattened in the US since 2000 but obesity has doubled since then. I’ve put together data from several US government sources in this writeup (all data + R code is available in this github repo).
From a raw data standpoint, per-capita supply of vegetable oil much more closely tracks obesity than does per-capita supply of calories or caloric sweeteners. People are also exercising a lot more than they used to (which one would expect given the public health advice to “eat less, move more”).
While this is prima facie evidence in favor of seed oils, we also know that seed oils are combined with lots of other potential toxins in processed foods. If you haven’t seen this study yet, it definitively shows that “ultra processed food” has a very low omega-3:omega-6 ratio, causes weight gain, messes with satiety and hunger hormones, and affects metabolism.
The question then becomes “what exactly is it about ‘ultra processed food’ that is so bad?”
Seed oils or some form of omega-3/omega-6 imbalance?
MSG or other flavor enhancements?
Modified starches? [which are basically alternative forms of MSG]
Lecithins and other emulsifiers? [mono and diglycerides]
Artificial colors?
To answer this question we’d somehow have to make an “ultra processed diet” that was identical in content to the status quo “ultra processed diet” but didn’t include the above substances. Which is likely infeasible.
The other major question I’m grappling with is why there is an obesity-elevation gradient. This has been litigated on LW in the past and my preferred theory at the moment is that there is some sort of relationship between hypoxic environment and cell metabolism that favors metabolic efficiency. However, it’s not like high-elevation places are immune from obesity; they just have lower rates and more or less increase in parallel with low-elevation places. However, elevation isn’t the whole story because there’s plenty of low-elevation places that also have low-obesity. I don’t think seed oils fit into this at all unless it is through how they influence cell metabolism.
...which leads me into my recent reading of Anthony Hulbert’s Omega Balance. I’ve found it useful for understanding some basic biochemistry of cell membranes and how cell membrane fatty acid composition might interfere with basic metabolic processes. He probably oversteps the evidence with some of his arguments. But as a result of reading it, I’ve spent the last week eating high-omega-3 by focusing on consuming lots of flaxseed and chia seed. (It turns out a basic-sized salmon filet has <1g of omega-3 so fish isn’t the most concentrated source of omega-3 even though it is for “long-chain” omega-3.) So far I’ve been pleased with the results: I feel like I’ve slept better and I’ve dropped a few pounds. Tonight for dinner we had people over and I ate a slice of Marie Calendar’s cherry pie that has immediately given me a stomach ache. What does the pie have a lot of? Modified starch and seed oil.
I’m surprised that the AHA and similar continue to push high-omega-6 oils. Yes, they reduce cholesterol, but they seem to increase a lot of other long-term disease risks. I view the focus on cholesterol to be short-sighted and mistaken. But I’m no biochemist or physician.
A guy is going alone through the wilderness, with a solar powered icebox on his back. He crosses a raging river by swimming. He slashes his way through a jungle. He is blasted by sun on an endless desert. It’s been weeks and he has no company at all, save the bleached bones (and ice boxes) of those who did not make it. He climbs a mountain until he finally comes to a cave in the snow. Inside is a man with beard like silvery horsehair, eyes like fire. Very old but fit as a mountain goat.
“O wise sensei, I brought your pizza and ice cream with me. Now tell me the secret of perfect health!”
“It is vely simple”, says the old man while greedily unwrapping a stick of icecream. “Only meet with people who have made a journey such as you youlself just did.”
More prosaically: “It” does not run uphill.
~90% of Earth’s mammalian excreta is produced by humans and their livestock. The livestock especially are immobilized in close quarters and their manure is spread to fields by mechanical means. The manure is often in a fresh condition with viable gut flora present. This means that the fitness of the gut microbiome is independent of the host’s mobility and fitness. They can find plenty of new hosts nearby, and have a good chance of spreading far and wide with tractors. If it can make the host eat x% more, it gains roughly that big an advantage: that much more manure along which it can spread. In state of nature that tactic obviously would not work, it would be much better to have a slim host with good legs.
Selection for larger animals, antibiotic feeding and selection by market forces could also contribute towards more hunger-inducing gut microbes.
Maybe overweight is not the only hit on the host’s well-being. There could be other pathologies by which the mechanised microbes improve their fitness. If this is the case, exposure to bad manure – or biome that derives from it – would be a common cause for both obesity and some other diseases associated with it.