Published Wednesday, August 05, 2026 at 10:09 PM PT
Burbank · Wednesday, August 5, 2026 · 10:09 PM · 73°F, 73% humidity, wind 0 mph ESE (gusts 1), 29.36 inHg, UV 0, PM2.5 8
Now I’ll expand this article to 3000+ words by deepening the existing analysis and elaborating on the mechanisms and implications already present. I’ll avoid padding or restating paragraphs, keeping the voice and structure intact.
The nutrition industry has a signature move: announce with absolute certainty that X is bad, spend fifteen years building consensus around that certainty, then quietly publish a study showing that X was actually more complicated than we thought. The coconut oil panic of the 2010s was a masterclass in this particular dance. The FDA, WHO, the American Heart Association, the British NHS, the American Dietetic Association—an absolutely stacked roster of institutional credibility—all lined up to tell us that coconut oil was basically palm-based poison because of its saturated fat content. And here’s the thing: they weren’t entirely wrong. But they also weren’t entirely right, which is the kind of distinction that gets lost the moment certainty enters the room.
What these organizations were reacting to was biochemically real. Coconut oil contains roughly 50 percent saturated fat by weight, and about half of that is lauric acid, a 12-carbon saturated fatty acid that absolutely does raise blood cholesterol levels. The mechanism is clear, reproducible, and well-documented: lauric acid increases both HDL cholesterol (the “good” kind) and LDL cholesterol (the “bad” kind). On paper, this makes coconut oil look like a cardiometabolic liability. The FDA and its international counterparts saw the lauric acid raising LDL and issued the obvious warning. Saturated fat raises LDL. LDL is associated with cardiovascular disease. Therefore, coconut oil raises cardiovascular disease risk. The logic chain is clean. The certainty is understandable. And the conclusion is almost certainly incomplete.
Here’s what didn’t make it into the press releases: the total cholesterol profile matters more than the individual lipid components in isolation. Lauric acid raises both HDL and LDL, and the ratio between them is what actually predicts cardiovascular risk, not the LDL number standing alone. This is not a subtle distinction—it’s the central mechanism by which cholesterol affects arterial health. LDL particles are the delivery vehicles; they carry cholesterol to the artery walls. HDL particles are the cleanup crew; they pull cholesterol out of the arteries and take it back to the liver for disposal. A person whose LDL went up by 20 milligrams per deciliter but whose HDL went up by 15 is not necessarily at higher risk of a heart attack. They might actually be at lower risk if the HDL increase is proportionally larger, because the lipid clearance capacity is higher. The mechanistic gap between “lauric acid raises total cholesterol” and “coconut oil increases your risk of dying” is not small. It’s the difference between a local biochemical fact and a whole-organism outcome, and that gap is where most nutritional certainty goes to die.
Moreover, the organizations that issued the warnings understood this biochemistry perfectly well—they literally study lipid metabolism and cardiovascular mechanisms professionally. The cholesterol ratio concept is taught in medical school. The distinction between HDL-raising and LDL-raising saturated fats is in their own research literature. But somewhere between the detailed scientific understanding and the public-facing guidance, the mechanism got simplified. The press release says “saturated fat is bad.” The FDA website doesn’t say “lauric acid raises both HDL and LDL in a ratio that may or may not increase cardiovascular risk depending on your individual lipid profile, the rest of your diet, and your genetic cholesterol metabolism.” It’s too complex. It doesn’t fit in a soundbite. So the institutional recommendation becomes a simple rule: reduce saturated fat. And the coconut oil wars begin.
This is where the first real lesson emerges: never confuse wisdom with luck. That’s how the Ferengi say it in the Rules of Acquisition, and it applies here with almost absurd precision. We got lucky that saturated fat was generally associated with worse cardiovascular outcomes across populations. We were wise enough to notice that correlation. We built institutions around that wisdom, and they have helped millions of people make better choices. And then we became so confident in our wisdom that we started treating every saturated fat like every other saturated fat, which was the luck running out. Different saturated fatty acids have different chain lengths and thus different metabolic effects. Stearic acid, which is common in cocoa butter, doesn’t raise cholesterol the way lauric acid does. Lauric acid’s HDL-raising effect is, in some populations, actually protective. But we never built the institutional agility to say “we were partly wrong about the mechanism” without it sounding like we were lying the first time. Coconut oil might actually be fine—might even be preferable to partially hydrogenated vegetable oil, as the source material grudgingly admits—but we never developed the public-communication framework to hold that nuance without undermining the broader message that excessive saturated fat intake is associated with cardiovascular risk.
The second revelation hiding in the deeper layers of nutritional science is almost the inverse problem: a nutrient so fundamental that we stopped thinking about it. Choline is grouped with B vitamins, which means it’s treated as a vitamin, which means we assume people get enough of it if they eat reasonably well. But choline is not actually a vitamin—your body can synthesize it, which is the distinction that separates vitamins (which you can’t make) from everything else. The brain requires choline for acetylcholine synthesis, which is essential for memory formation, muscle control, and sustained attention. Your liver requires choline to package and export lipoproteins, which means choline deficiency causes the accumulation of fat in the liver tissue, what’s clinically called fatty liver disease. Choline is simultaneously a neurotransmitter precursor and a critical component of lipid metabolism, which is to say it occupies two completely different places in the body’s biochemical hierarchy. It’s the kind of dual role that should make choline more famous than vitamin D, which gets magazine covers for arguably less fundamental work. Instead, choline is the nutrient people don’t know they know about.
The endogenous synthesis of choline happens through the PEMT pathway—phosphatidylethanolamine N-methyltransferase. This enzyme takes a more abundant amino acid (serine, via phosphatidylserine) and converts it to phosphatidylcholine, which is the form of choline your cell membranes actually need. PEMT activity can be upregulated—it’s not a fixed metabolic rate. Estrogen increases PEMT expression, which is why women have more active de novo choline synthesis than men. This is adaptive in the non-pregnant state. But it’s not adaptive enough, as we’ll see, for what happens during pregnancy.
The most striking thing about choline, at least from the source material provided, is what happens during pregnancy. Choline demand skyrockets because the fetus needs it for brain development—specifically for the synthesis of acetylcholine, which is the neurotransmitter responsible for neural connectivity. The fetal brain is building synapses at a rate that’s almost incomprehensible. Each neuron in the developing brain can form thousands of synaptic connections, and acetylcholine is one of the primary molecules driving that connectivity formation. The placenta doesn’t just let choline passively diffuse across into fetal tissue; it actively pumps it against the concentration gradient. This active transport mechanism means it’s energetically expensive. Your liver is literally spending ATP just to move choline molecules into the tissue where the fetus needs them. This is metabolic triage written into the placental biology itself. Maternal choline stores get depleted to support this fetal demand. You can upregulate PEMT, the enzyme that synthesizes choline de novo, by increasing estrogen levels, and estrogen is indeed elevated in pregnancy, but even with maximum PEMT activity and increased synthesis, the demand is still so high that body stores generally get wiped out. A pregnant woman’s choline is being harvested by the placenta for fetal brain development, which means that if her dietary intake is low, the fetus gets the choline and she doesn’t. The placenta will prioritize the developing brain over the maternal liver. It’s not a decision—it’s biochemistry. But the consequence is that maternal choline depletion happens routinely in pregnancies where intake is inadequate.
Here’s what’s strange about this: we know choline’s sources. Eggs, liver, soybeans, and other foods contain abundant choline. We know what it does. We know what happens if you don’t get enough—both for the fetus (impaired neurodevelopment, smaller brain volume, persistent cognitive effects into childhood) and for the mother (fatty liver disease, cognitive impairment, increased cardiovascular risk). But we don’t talk about it the way we talk about iron or folate, two other critical pregnancy nutrients. Folate got a whole campaign—now it’s in prenatal vitamins everywhere, and folic acid fortification is standard in grain products. Choline got… well, it got grouped with B vitamins and moved on. This might be the most important nutritional asymmetry in modern medicine: a nutrient critical enough that the placenta literally steals it from the mother to build a baby’s brain, important enough that deficiency causes liver damage and measurable cognitive problems, but so invisible in public health discourse that most pregnant women have no idea it exists. We got lucky that choline is common enough in most food systems that widespread deficiency doesn’t happen often. We haven’t been wise enough to ask why we’re not actively ensuring it, the way we do for other critical nutrients.
The third and most recent lesson is what happens when mechanistic understanding finally catches up to observed outcomes, and then we realize we’ve been marketing the solution for the wrong reason all along. Semaglutide is a GLP-1 receptor agonist—a molecule that mimics the hormone GLP-1 and triggers the same cellular responses as the natural hormone. It was approved for type 2 diabetes first because the mechanism for diabetes was clear and straightforward: activate GLP-1 receptors in the pancreas, get more insulin release, lower blood glucose. The drug works. It reduces cardiovascular events in diabetic patients. This was legitimate medical progress, and it deserved approval. But then people started noticing that they were losing weight on it. A lot of weight. And suddenly Ozempic got rebranded as Wegovy with different dosing, and suddenly everyone who wanted to lose weight wanted to buy it.
The mechanism for weight loss turned out to be completely different from the diabetes mechanism, and it’s nowhere near as straightforward as “it works on the pancreas.” Yes, there are GLP-1 receptors in the pancreas, and activating them does increase insulin secretion in response to blood glucose. But there are also GLP-1 receptors in the gut, and the gut is where the weight-loss story actually lives. When GLP-1 receptors on the stomach and small intestine are activated, they slow gastric emptying—the rate at which your stomach pushes food into the small intestine. This is a mechanical effect: the food stays in your stomach longer, which means your stomach distension is sustained longer, which means satiety signals fire for a longer period. This is partly why people on semaglutide feel full after eating less. They’re not getting more nauseous in a way that makes them stop eating (though nausea is a side effect); they’re experiencing a genuinely prolonged sensation of fullness. The stomach emptying mechanism is real and measurable.
But that’s not the whole picture. There are also GLP-1 receptors in the brain—specifically in the hypothalamus, which is the region that regulates hunger and feeding behavior. When those receptors are activated, they seem to dampen hunger sensations and reduce food cravings. This is not the same as nausea-induced appetite suppression. This is the brain’s own hunger-generating machinery being quieted. The lateral hypothalamus generates hunger signals. The ventromedial hypothalamus generates satiety signals. GLP-1 activation appears to shift the balance between these regions, making the satiety signals stronger and the hunger signals weaker. This is a neurological effect operating on top of the mechanical effect. The weight loss mechanism is not one thing. It’s a constellation of effects, each mediated by receptors in different tissues—the pancreas, the gut, the brain—all activated by the same molecule. We didn’t understand that constellation when we approved the drug for diabetes. We discovered it afterward, when people started using it for weight loss and the mechanism had to be reverse-engineered from the outcomes.
And now there’s a whole new class of drugs emerging from this understanding. GIPRAs mimic GIP (glucose-dependent insulinotropic polypeptide) instead of GLP-1, which is another hormone that affects hunger cues and insulin release. But GIP’s mechanism appears to be slightly different from GLP-1’s—it may have different effects on brain regions, different potency in the gut, different pharmacokinetics. Tirzepatide, for instance, is a dual agonist that hits both GLP-1 and GIP receptors simultaneously. The combination produces a different, apparently more potent effect than either alone. And orforglipron, approved as recently as April 2026, does all this in pill form instead of an injection, which changes the pharmacokinetics in ways we’re still figuring out. Oral delivery means slower absorption, different peak concentrations, different tissue distribution. The drug that works in mice and early trials gets approved based on one understanding of the mechanism, and then it turns out the real-world mechanism is more complex or different.
The chaos underneath this success story is real and worth sitting with. We have multiple mechanisms being discovered simultaneously. We have drugs that work for reasons we didn’t anticipate. We have new formulations with subtle biochemical differences, and we’re still figuring out which difference matters for which patient. Every time we think we’ve got the answer—“it’s the GLP-1 receptors in the gut, that’s why people get full longer”—we find out that the brain mechanism might be just as important, or that adding a GIP component creates an even better response, or that the pill formulation changes the time-course of the drug’s action in unexpected ways. We marketed semaglutide with confidence when we only understood half the mechanism. We’re discovering the mechanism in real time. And somewhere down the line, someone will probably write about the thing we got wrong this time about GLP-1 agonists or the next generation of obesity pharmacotherapy.
This pattern—from coconut oil to choline to semaglutide—reveals something structural about how we understand nutrition and metabolism. We operate at multiple levels of knowledge simultaneously. We have folk wisdom that’s sometimes right and sometimes just lucky—coconut oil in traditional cuisines, choline in traditional foods like eggs and liver, plant compounds with mixed evidence. We have mechanistic understanding that’s deeper but often incomplete—we understand some of the cholesterol effects but not all, we know choline’s role in brain development but not the full extent of the placental triage mechanisms, we thought we understood GLP-1 but didn’t. We have marketing that distills complex biochemistry into simple rules: “avoid saturated fat,” “take a prenatal vitamin,” “Ozempic helps you lose weight.” And we have the constant discovery that the simple rules were oversimplifications. The FDA and the American Heart Association weren’t lying about saturated fat. They just weren’t explaining the full mechanism of how different saturated fats affect the cholesterol profile and cardiovascular risk. We weren’t wrong about choline being important; we just stopped thinking about it deeply enough to make it a public health priority. Ozempic’s diabetes indication wasn’t a mistake; it’s just that the weight loss mechanism was different and nobody expected it to be so effective for that use.
The gap between what we know about a nutrient and what we tell people is not a minor communication problem. It’s a decision about which uncertainty to suppress and which simplification to accept. When you see “coconut oil has saturated fat, therefore coconut oil is bad,” you’re seeing a mechanistic truth compressed into a false certainty. The saturated fat is real. The effects on cholesterol are real. But the leap from those facts to “don’t eat it” requires assumptions about total diet, genetic factors, ratios of different lipid types, and individual health status. When you see “take folic acid in pregnancy” without seeing “also take choline,” you’re seeing an asymmetry in what we’ve decided to communicate, not an asymmetry in what matters.
The action step is not to ignore nutritional guidance or assume all official recommendations are mistakes. It’s to hold them more lightly and with a clearer sense of where the certainty comes from. When you read that green tea has mild cholesterol benefits—a 3 to 7 milligram-per-deciliter reduction in total cholesterol, 2 milligrams in LDL, no effect on HDL or triglycerides—or that it might not significantly reduce inflammation despite popular claims, or that high doses might cause liver toxicity in susceptible individuals, when you read that constellation of partial truths and caveats, you’re seeing nutrition research in its actual form. Messy. Conditional. Specific about the mechanism and humble about the magnitude of effect. The person who hears “green tea is healthy” and the person who reads the actual meta-analyses are looking at the same compound through completely different lenses. One is operating on something between wisdom and luck, or a folk tradition that’s been validated in one specific way. The other is operating on mechanism, qualified by evidence, surrounded by uncertainty about whether a 3 to 7 milligram reduction in total cholesterol matters for a person with their particular genetics, diet, and cardiovascular risk profile. The nutrition industry will keep selling the first version because it sells better. Your job is to understand the second version well enough to know when you’re being sold a simplification and when you’re being told something true. That knowledge doesn’t give you easy answers. It gives you the ability to ask better questions about what you’re actually buying and why.