Someone asked me this week what the deal is with EPG. There are fights breaking out online over a protein bar, and I wanted to know what its makers know that made them comfortable building their whole product around an ingredient most people had never heard of a year ago. I said I was legitimately on the fence, asking to be swayed.
I am still on the fence. But I have read enough now to know why it is the right place to stand, and I would rather walk you to the fence than push you off either side of it. This is not a piece about whether David is good or evil, or whether Peter Rahal, who built RxBar before this, is a visionary or a salesman. He has called EPG a "holy grail for public health." That is a marketing sentence. I want the science sentence. But I owe his sentence a fair hearing first, because it is better than its critics admit.
What EPG actually is
EPG stands for esterified propoxylated glycerol. Start with a molecule of glycerol, the same three-carbon backbone at the center of every fat you eat. Attach fatty acids to it the way nature does, and you get an ordinary triglyceride that your gut digests for nine calories per gram. EPG does one thing differently: before the fatty acids go on, short propylene-oxide bridges are inserted between the glycerol backbone and the fatty-acid tails. Those bridges are the entire trick. Your digestive lipases, the enzymes that snip fat into absorbable pieces, evolved to recognize the natural bond. They do not recognize the bridged one, so they cannot cut it, so most of the molecule never gets absorbed. It travels the length of your gut and leaves.
The payoff is striking on paper. Conventional fat delivers nine calories per gram. EPG, by its maker's accounting, delivers about 0.7, because almost none of it is taken up. It behaves like fat in the mouth and in the pan, carrying flavor and giving food the richness fat gives, while contributing nearly none of fat's energy. That is what makes a 150-calorie bar with 28 grams of protein arithmetically possible, and it is why a fat that passes through you is suddenly the foundation of a company.
Peter Rahal's case, in his words
Peter Rahal is not an idiot, and an honest piece has to take his argument seriously before it pokes at it. He co-founded RxBar, sold it to Kellogg for a reported $600 million, and has spent the years since arguing that the next move in food is not branding but reformulation. He now talks about David less as a snack company than as a food-technology platform, and he bought Epogee, the maker of EPG, outright so the platform would own the molecule it runs on.
His thesis is a public-health thesis, and he states it plainly: "Our mission is simple," he has said, "to remove unnecessary calories and sugar from the American diet and replace them with what the body actually needs, which is high-quality protein." Privately, in the message that sent me down this whole rabbit hole, he put it more bluntly and called EPG a "holy grail for public health." Read uncharitably, that is a founder talking his own book. Read charitably, it is a real argument. In a country where excess calories and the metabolic disease that trails them are arguably the dominant health problem, a fat that delivers fat's pleasure without fat's energy is not a gimmick. It is a lever on the one variable that matters most, arriving in the middle of the GLP-1 era, when the whole culture has suddenly decided that eating fewer calories with less suffering is worth engineering for. I take that seriously. I have spent my career arguing that the composition of the food supply is upstream of metabolic health, and a tool that quietly removes calories from foods people will eat anyway is, in principle, exactly the kind of upstream lever I respect.
On the lawsuit, Rahal has not blinked. He calls the claims "meritless" and "based on a fundamental misunderstanding of basic, well-established scientific principles regarding how calories are determined," and David has published a formal response standing behind its labels as accurate and FDA-compliant. His core scientific point, which I think is correct as far as it goes, is this: "If you burn ingredients like EPG in a calorimeter, [they] would appear to deliver far more calories than the body actually metabolizes." He is right that bomb calorimetry is the wrong instrument for this question. Where I keep my hand on my wallet is the phrase "the body actually metabolizes," because that is precisely the number nobody has nailed down, and it is a question about your gut rather than about his integrity.
It is not Olestra, and that matters
Everyone who knows food history reaches for the same comparison, and they are right to reach for it. In the 1990s, Procter & Gamble launched Olestra, a sucrose molecule loaded with fatty acids, also non-absorbed, also a calorie-free fat. It became a cautionary tale. It pulled fat-soluble vitamins out with it on the way through, which forced a fortification requirement, and at the doses people ate in snack chips it produced the gastrointestinal effects that launched a thousand jokes. Olestra is the ghost in every EPG conversation.
EPG is a genuinely different molecule, and the difference is not just spin. The published safety work, a multi-paper series in Food and Chemical Toxicology, reports that EPG does not strip fat-soluble vitamins the way Olestra did and is better tolerated in the gut. Epogee has filed multiple GRAS notices with the FDA and received "no questions" letters in response, the agency's way of saying it does not dispute the safety conclusion for the stated uses. None of that is nothing. A molecule with decades of toxicology behind it and a cleaner vitamin profile than its infamous predecessor is not Olestra wearing a new hat.
So if you are looking for a reason to panic, the obvious one does not hold up. The honest concerns are quieter and more interesting.
What the human data actually show
Here is the study that should anchor the conversation, because it is the one that put EPG in free-living humans for a meaningful stretch. In a double-blind, randomized, controlled trial, 139 healthy volunteers ate a controlled diet for eight weeks with EPG at 10, 25, or 40 grams a day, against a margarine control. The reassuring part: circulating retinol (vitamin A), alpha-tocopherol (vitamin E), and vitamin D held steady. The part that deserves your attention: beta-carotene and phylloquinone (vitamin K1) came down in the EPG groups, and PIVKA-II, a functional marker that rises when vitamin K is undersupplied, went up.
Granted, the shifts were modest, the trial was eight weeks, and the participants were healthy adults. But it tells you the mechanism is real. A non-absorbed lipid moving through your intestine acts as a "sink," a passing oily phase that some fat-soluble compounds partition into and ride out with. The data say that sink is gentle for most fat-soluble vitamins and not quite negligible for vitamin K and carotenoids. Eight weeks is not a life. Nobody has run the decades-long, eat-it-every-day version of this study, because nobody has had a reason to until a bar built an entire macro profile on the molecule and people started eating it daily. The absence of that data is not evidence of harm. It is just an absence, and an honest person names it.
The thing almost no one is measuring
This is where my own field walks into the room. I have spent years arguing that what passes through the gut is never inert, because the gut is not a pipe but a dense, crowded ecosystem.
EPG is a fat-like substance that, by design, reaches the colon largely intact. That is precisely the territory where the microbiome lives and where bile acids, the body's fat-handling detergents, get recycled and transformed by bacteria. A lipid phase traveling all the way down has at least three places it could matter that the vitamin studies were not built to see: it could alter how bile acids are reabsorbed and signal, it could change the lipid environment that shapes which microbes thrive, and it could be partially fermented by those microbes into something the body does absorb. I went looking for studies on EPG and the human gut microbiome. I did not find them. For a molecule whose entire selling point is that it visits the colon and leaves, that is a conspicuous gap, and it is the gap I would fund first.
It also feeds directly into the fight everyone is actually having, which is about calories.
The calorie fight is a real scientific dispute, not a gotcha
David has been sued over its calorie counts. Third-party testing reported far more energy in the bars than the label claims. Rahal's defense is technical, and it is also correct as far as it goes: the testing used a bomb calorimeter, which incinerates a sample and measures every joule of heat, including the energy locked in EPG that your body never absorbs. Food labels are not supposed to count that. They count metabolizable energy, the calories you actually extract. By that standard, EPG's calories do not belong on the label, because you do not get them.
Both sides have a real point, and the place they collide is exactly the microbiome question above. The label is correct if EPG is truly almost entirely non-absorbed and almost entirely non-fermented in real people eating it every day. The lawsuit's number is what you get if you (wrongly) count combustion energy, but it gestures at a fair question: how confident are we in that "almost entirely," across millions of different guts, over years? Dietary fiber went through this same reckoning and settled at roughly two calories per gram precisely because gut bacteria ferment some of it into absorbable short-chain fats. Whether EPG has its own small fermentation number is, as far as I can tell, unanswered. That single unanswered number is what separates "the label is honest" from "the label is optimistic."
What GRAS does and does not mean
A word on the regulatory bar, because it gets oversold in both directions. "GRAS" means Generally Recognized As Safe, and for ingredients like this it is frequently self-determined: a company assembles the evidence, convenes its own expert panel, and concludes safety, sometimes notifying the FDA and sometimes not. Epogee did notify, and the FDA's "no questions" letters are real and worth something. But "no questions" is not the same as the agency independently affirming safety, and GRAS is a lower, faster bar than full food-additive approval. That is not a scandal; most of the modern ingredient supply runs on this pathway. It is simply a reason to hold the certainty loosely. The system was not designed to catch a slow, decades-out, microbiome-mediated effect, because that is not what it tests for.
Follow the money on both sides
The reason this turned into online warfare is that there is money on every side of it, and that should make you skeptical of everyone, not just the people you were already inclined to doubt. David reportedly raised tens of millions and used much of it to acquire Epogee outright, which is why competitors are now fighting in court over access to the ingredient. A celebrated physician was brought on as a science voice and later stepped back. None of that tells you whether EPG is safe. It tells you that the loudest reassurances and the loudest alarms are both being made by people with a stake.
Read the trials, not the cap table.
Where I land
After all of this investigation, I am still on the fence. Here's my take.
EPG is not the villain the Olestra reflex wants it to be. The chemistry is clever, the acute safety record is real, and the panic is mostly recycled from a different molecule. But I think it's far too soon to call EPG a holy grail, because I need the long-term, free-living microbiome-inclusive human data to come to any sort of conclusion. The genuine open questions are narrow and specific: the small vitamin K and carotenoid signal over time, the unmeasured colonic and microbial story, and the honest size of the absorbed-calorie number that the entire label depends on.
That points to a position, and it is a health position, not an ideological one. I would not treat EPG as a free pass to eat without consequence, and I would not treat it as a toxin to be feared. I would eat it the way a careful person eats anything novel: in moderation, paying attention to my own gut, not as a daily staple stacked three-deep, and with more trust in the brands that label conservatively and fund the missing studies than in the ones that call it a holy grail. If the colon-and-microbiome work gets done and comes back clean, I will happily move off the fence toward "useful tool." If it never gets done, the fence is the most accurate place to stand.
I told the person who asked me that I wanted to be swayed. The science swayed me off the panic. It has not yet swayed me onto the grail. That is not the same as having no opinion. It is, in my opinion, the only opinion the evidence currently supports.