The David Milkshake and the Fat Built Not to Digest
David Protein’s first drink launched Monday with 30 grams of protein, 150 calories and a billboard campaign. Its fat system is one ingredient: a molecule engineered so your enzymes cannot take it apart. The only two human trials of that fat measured exactly what you would guess — at doses a bottle does not come close to.
David Protein's new chocolate and vanilla milkshakes (announced September 8, 2026; 30 grams of protein, 140 to 150 calories, under 1 gram of sugar, $49 per dozen) get their fat from a single ingredient, "modified plant fat (EPG)": esterified propoxylated glycerol, made by Epogee, which David acquired in 2025. Small propylene-glycol bridges stop digestive lipases from breaking it down, so, in Epogee's words, it "moves through your body much like dietary fiber, largely intact and without being absorbed," delivering about 0.7 calories per gram instead of 9. The published human evidence is two trials: 139 healthy adults given 10, 25 or 40 grams a day for eight weeks, in which "gas with discharge; diarrhea; oily spotting; oily evacuation; oily stool; liquid stool; soft stool" rose with dose at 25 and 40 grams while 10 grams was "reasonably well tolerated," and 16 men given single doses of 30 to 150 grams with no effect on bowel function except in one subject. A milkshake lists 1.5 to 2 grams of total fat and a bar 2 grams, so even two bars plus a shake stays under the lowest tested dose; at labelled servings the fat is an unlikely cause of symptoms, and the bar's maltitol, the protein load and the shake's gums are more plausible. The precedent is olestra, whose 1996 "may cause abdominal cramping and loose stools" label FDA removed in 2003 after a 1,123-person study found no difference in gut complaints. A January 2026 calorie lawsuit was voluntarily dismissed in March; FDA's GRAS letter on EPG states the agency did not evaluate labeling claims, so the 0.7 figure is the manufacturer's. Educational, not medical advice.
On Monday, David Protein announced its first drink: a chocolate or vanilla milkshake with 30 grams of protein, 140 to 150 calories and less than a gram of sugar in an 11-to-12-ounce bottle, sold at $49 a dozen on the company’s site and in a handful of New York City stores, with a “Drink David” billboard campaign starring the model Stella Maxwell. If you have spent any time on the protein side of the internet in the past two years, you already know the bar this comes from. David’s 28-gram, 150-calorie bar is the object that made a protein company a TIME100 most-influential company with about $131 million in first-year sales, and the object that made a certain kind of Reddit thread inevitable.
Those threads are the reason this article exists. Search the bar’s name next to almost any bathroom word and you get a steady stream of anecdotes — not a viral moment, just a persistent hum. The threads all point at one ingredient, and they are half right about it. So here is the full label read: what the milkshake is made of, what the fat in it was engineered to do, what the only human trials of that fat actually measured, and the arithmetic that separates the ingredient’s reputation from the dose in the bottle.
The ingredientA fat that is designed to leave
The chocolate milkshake’s ingredient statement is organised into “systems”: a protein system (ultrafiltered skim milk, water, lactase enzyme), a flavour system (cocoa, natural and artificial flavour, salt, sucralose, acesulfame potassium), a structure system (dipotassium phosphate, gum arabic, cellulose gum, gellan gum) — and a fat system consisting of exactly one thing: “modified plant fat (EPG).” The same three letters sit in the bar’s fat system, next to coconut oil. EPG is esterified propoxylated glycerol, made by Epogee, a company David bought outright in 2025 after becoming, per the TIME profile, its dominant customer.
Epogee’s own description is the clearest statement of the idea: EPG delivers 0.7 calories per gram against the 9 of ordinary fat, “up to 92% fewer calories absorbed,” because it “moves through your body much like dietary fiber — largely intact and without being absorbed.” Chemically, it is a fat molecule with small propylene-glycol bridges inserted between the glycerol backbone and the fatty acids, which is enough to stop the enzymes that normally take fat apart from getting a grip. It tastes and behaves like fat in the mouth, in a bar, and in a milkshake. It then does something ordinary fat never does: it leaves, mostly as it arrived.
A fat engineered so that digestive enzymes cannot take it apart is, by definition, a fat that has to go somewhere. The entire question is how much of it.
The precedentOlestra, and the label that came off
If “a fat that passes through undigested” rings a bell, that bell is olestra, the Procter & Gamble fat substitute FDA approved for savoury snacks in 1996 with a mandatory label that has become food-history folklore: “Olestra may cause abdominal cramping and loose stools.” What is less remembered is the sequel. In a 1998 JAMA study, researchers handed 1,123 cinema-goers aged 13 to 88 unlabelled bags of either olestra or regular potato chips and asked about symptoms afterwards: 15.8% of the olestra group reported a gastrointestinal complaint versus 17.6% of the regular-chip group, a difference that did not reach significance. By 2003, FDA removed the warning label, writing that olestra eaters were “no more likely to report having had loose stools, abdominal cramps, or any other GI symptom,” and citing roughly one complaint per million servings sold.
The lesson of olestra was never that the mechanism was fake. It was that the mechanism has a dose, and that a snack-sized serving sat below it for almost everyone. Hold that thought.
The trialsTwo studies, 155 people, one threshold
EPG has been through FDA’s GRAS process three times — notices 583, 640 and 761, each closed with a “no questions” letter — and the human tolerance data behind those filings is published. The main study, Davidson and Bechtel 2014, gave 139 healthy volunteers 10, 25 or 40 grams of EPG a day for eight weeks against a margarine control. The gastrointestinal adverse events it lists are specific and unglamorous: “gas with discharge; diarrhea; oily spotting; oily evacuation; oily stool; liquid stool; soft stool.” They occurred more often at 25 and 40 grams, their “incidence and duration” tracked the dose directly, and the authors’ conclusion was that 10 grams a day “was reasonably well tolerated.” The same study saw lower blood levels of beta-carotene and vitamin K markers in the EPG groups — the olestra pattern again, without clinical deficiency — while vitamins A, E and D were unaffected.
A second study went the other direction: 16 men, single-day doses climbing from 30 to 150 grams. “EPG had no effect on bowel function, except in a single subject,” the abstract reports, and it puts expected real-world intake at “approximately 20 g/day or less.” So the picture from the only two human trials that exist is coherent with olestra’s: the undigested fat produces exactly the effects you would predict, and it starts producing them somewhere around the 25-gram-a-day mark, sustained.
The arithmeticWhere one bottle lands
| Dose of EPG per day | Where it comes from | What the trials found |
|---|---|---|
| ~2 g | One David milkshake (total fat 1.5–2 g, all of it the “fat system”) or one bar (2 g, EPG plus coconut oil) | Below any dose ever tested for tolerance |
| ~6 g | Two bars and a shake — the company’s own bar ceiling is two a day | Still under the dose the 8-week trial called “reasonably well tolerated” |
| 10 g | Lowest arm of the 139-person trial | “Reasonably well tolerated” |
| 25–40 g | Middle and top arms of the same trial | Gas with discharge, diarrhea, oily spotting, oily and liquid stool — rising with dose |
| 30–150 g | Single-day rising doses, 16 men | “No effect on bowel function, except in a single subject” |
Now the number that the threads skip. A David milkshake lists 1.5 to 2 grams of total fat. A bar lists 2 grams, and part of that is coconut oil. The company’s own FAQ recommends no more than two bars a day and says plainly that going past it “may cause gastrointestinal discomfort” — but even a maximal day of two bars and a shake tops out near 6 grams of EPG, which is under the lowest dose the eight-week trial tested and well under the 25 grams where the oily-stool column started to fill. Nobody eating David products as labelled is anywhere near the doses that produced the effects in the literature. If EPG were the whole story, the Reddit threads would be empty.
They are not empty, which means the honest read of the bar is that EPG has been taking the blame for its neighbours. The bar’s binding system opens with maltitol and allulose, then soluble corn fiber; the panel shows 12 grams of carbohydrate and 0 grams of sugar, and does not break out how much of that is sugar alcohol. Maltitol is a well-characterised sugar alcohol with a well-characterised laxative threshold — in one Japanese dose-escalation study, 22 of 27 women developed osmotic diarrhea as maltitol intake was stepped up toward 45 grams — and it is the ingredient the bar shares with every sugar-free candy that ever ruined an afternoon. Then there is the reason the two products exist at all: 28 to 30 grams of protein in a single small serving, twice or three times a day, which is a genuine digestive workload of its own, as anyone who has tried a high-protein reset can confirm.
The milkshake, notably, drops the maltitol entirely. Its sweetness comes from sucralose and acesulfame potassium, its milk has been run through lactase so it is labelled lactose-free, and its texture comes from gellan, cellulose gum and gum arabic. That makes it, on paper, a gentler product than the bar for a sensitive gut — with the caveat that gum arabic is itself a fermentable fiber, in an amount the label does not state. Whether the bottle produces its own threads is a question the internet will answer in roughly a month.
View this post on Instagram
The fine printWhat the 0.7 is, and is not
One more thing worth reading correctly, because it was the centre of a lawsuit. In January, three customers filed a class action claiming lab tests showed the bar contained far more fat and calories than its label — NBC reported the figure as roughly 80% more calories and 400% more fat, measured by bomb calorimetry, which counts all the energy in a food regardless of whether a body can extract it. The founder, Peter Rahal, called the method wrong for an unabsorbed fat, and in late March the plaintiffs withdrew the case “voluntarily, without prejudice,” for reasons that were never made public. The company says it remains confident in its labelling.
Both sides were describing the same fact from different ends. EPG contains about as much raw energy as any fat; the label counts only what the body absorbs, which Epogee puts at 0.7 calories per gram. That is a legitimate approach — the federal labelling rule permits several calorie-counting methods, including ingredient-specific factors — but it is worth knowing whose number it is. FDA’s most recent GRAS letter on EPG states that the agency “did not” consult its labelling office or “evaluate any information in terms of labeling claims.” The 0.7 is the manufacturer’s figure, resting on the same digestibility that produces the trial effects. It is also worth noting that the three GRAS notices cover confectionery coatings, bars, baked goods, candy, frozen desserts and fry oil; a milk-based drink is not among the listed uses, which under the GRAS system a company is free to determine for itself without filing.
So: the fat in the David milkshake is built not to digest, the science of what that does to a bowel is real and dose-dependent, and the dose in a bottle is small enough that the fat is probably not what a sensitive gut is reacting to — look first at the bar’s maltitol, the protein load and, in the shake, the gums. For what an oily or floating stool means when it does show up, that has its own explainer, and the enzymes-in-a-bowl demo that went viral this summer is a useful companion read on the difference between chemistry in a glass and chemistry in a person. David Protein had no involvement in this article. It is educational, describes what the cited trials, filings and ingredient statements say, and is not medical advice.