Are sugar alcohols safe? Erythritol, xylitol, and the clot study that spooked the “zero-sugar” aisle

What sugar alcohols actually are

Despite the name, sugar alcohols contain neither sugar nor the alcohol you drink. The term is a chemistry label: polyols are carbohydrate molecules with a structure that sits partway between a sugar and an alcohol — hence the “-ol” endings you see on the ingredients panel. The common ones are erythritol, xylitol, sorbitol, maltitol, mannitol, and isomalt. They occur naturally in small amounts in fruits and vegetables, and are manufactured at scale (erythritol typically by fermenting glucose with yeast) for use as sweeteners.

What makes them useful is a specific combination: they taste sweet — erythritol lands around 70% as sweet as table sugar, xylitol roughly as sweet, one-for-one — while being poorly absorbed or poorly metabolized, so they deliver far fewer usable calories and little to no glucose load.⁵ Erythritol is the extreme case: it’s absorbed in the small intestine but then largely excreted unchanged in urine, so it contributes almost no calories and passes through you mostly intact. That’s exactly why it became the darling of the low-carb world — and, as we’ll see, why measuring “blood erythritol” is trickier to interpret than it sounds.

The category matters because these aren’t interchangeable. Sorbitol and maltitol carry a meaningful glycemic load and are notorious for gut distress; erythritol is the gentlest on both counts. So when a headline says “sugar alcohols,” the honest question is always which one — and most of the current controversy is about erythritol and xylitol specifically.

The 2023 erythritol study that started it all

The paper that changed the conversation was Witkowski, Nemet, Hazen and colleagues’ 2023 study in Nature Medicine, and it’s worth describing precisely because the coverage rarely was.¹ It wasn’t one experiment; it was a stack of them, which is what gave it weight. First, in large groups of patients undergoing cardiac risk assessment — thousands of people across US and European cohorts — the researchers found that those with the highest blood levels of erythritol had a roughly two-fold higher risk of major adverse cardiovascular events (MACE: heart attack, stroke, or death) over the following three years, compared with those in the lowest group.

Then they went looking for a mechanism, and found a plausible one. In lab work, adding erythritol to human platelets — the cell fragments that form clots — made them more reactive, meaning they clumped and activated more readily in response to a trigger. In a mouse model, erythritol accelerated clot formation in an injured artery. And in a small human experiment, eight healthy volunteers drank a beverage containing 30 grams of erythritol; their blood erythritol shot up more than a thousand-fold and stayed elevated for days, and measures of platelet reactivity rose in parallel. That combination — an association in people, a dose-response in cells, an effect in animals, and a signal in a human challenge — is why the study landed so hard. It wasn’t “a correlation”; it was a coherent, mechanistically-supported hypothesis. The 2024 follow-up brief report in Arteriosclerosis, Thrombosis, and Vascular Biology reinforced the human piece: ingesting erythritol, but not an equivalent glucose load, enhanced platelet reactivity and clotting potential in healthy volunteers.²

Taken at face value, that’s alarming. But “taken at face value” is exactly where careful reading has to slow down — because the study has a structural feature that the reassuring and the panicked takes both tend to miss.

The catch: your body makes erythritol too

Here is the fact that reframes everything, and it’s the reason this claim earns an emerging grade rather than a strong one: erythritol is not only a sweetener you eat — it’s a molecule your body produces. Human cells make erythritol from glucose through the pentose phosphate pathway, and that endogenous production increases in states of high blood sugar and metabolic stress. This is the crux of the interpretation problem, and Mazi and Stanhope laid it out squarely in a 2023 Nutrients review titled, pointedly, “Elevated Erythritol: A Marker of Metabolic Dysregulation or Contributor to the Pathogenesis of Cardiometabolic Disease?”³

Consider what that does to the association. The people in the observational cohorts with the highest blood erythritol were patients already being assessed for cardiovascular risk — many with diabetes, obesity, and metabolic syndrome, all conditions that raise the body’s own erythritol output. So a high blood-erythritol reading could simply be a marker of the underlying metabolic dysfunction that independently drives heart attacks and strokes — a smoke detector, not the fire. In that reading, erythritol level is doing the same job as a high HbA1c: flagging the disease process, not causing it. The observational data alone genuinely cannot tell those two stories apart, and the study authors acknowledged the endogenous-production complication.

This is the honest tension. The mechanistic and interventional work — platelets getting stickier when you add erythritol, clots forming faster in mice, reactivity rising after a drink — does argue for some causal contribution, and shouldn’t be waved away. But the human challenge experiments used a single, large bolus dose (30 grams at once), producing blood spikes far beyond what a packet in your coffee or a square of keto chocolate would. They tell us what a big one-time load does to platelet biology over hours; they don’t tell us what modest, habitual dietary use does to hard outcomes over years. And no randomized trial has yet assigned people to eat erythritol or not and counted heart attacks — which is the study that would actually settle causation. Until then, the grade is emerging: a real, biologically plausible signal, confounded by the body’s own chemistry, and not proven to mean your sweetener is hurting you.

The clot signal isn’t nothing, and it isn’t proof. The body makes erythritol when metabolism is already in trouble — so a high blood level may be the alarm, not the arsonist.

Xylitol’s parallel 2024 study

If erythritol were a one-off, it might be easier to dismiss. It isn’t. In 2024 the same Cleveland Clinic group published a study in the European Heart Journal reporting that xylitol is prothrombotic and associated with cardiovascular risk — essentially the same playbook applied to a second polyol.⁴ Higher blood xylitol tracked with elevated three-year MACE risk in over 3,000 patients; xylitol enhanced platelet responsiveness in lab assays; and in a small human intervention, drinking a xylitol-sweetened beverage raised platelet activation markers.

The consistency cuts both ways, and it’s worth being even-handed about. On one hand, finding the same pattern for a chemically distinct polyol makes a shared class effect on platelets look more credible — two independent signals pointing the same direction is harder to dismiss as a fluke. On the other hand, xylitol carries some of the same interpretive baggage: it too can be produced endogenously in small amounts, and the human outcome data are again observational, in cardiovascular-risk populations, so confounding by underlying metabolic disease is once more in play. And a key difference matters: unlike erythritol, xylitol is meaningfully metabolized and doesn’t reach the extreme thousand-fold blood spikes, so the exposure story isn’t identical. The fair summary is that xylitol’s cardiovascular data sit in the same “emerging, plausible, unproven” box as erythritol’s — a signal worth watching, not a verdict. It does not erase xylitol’s well-established benefit in a completely different tissue, which we’ll get to.

Blood-sugar neutrality: the real advantage

Step back from the clot headlines and there’s a benefit here that is genuinely well supported, and it’s the whole reason these sweeteners exist: near-total blood-sugar neutrality. This is the moderate-to-strong claim in this article’s Evidence Radar. Because erythritol is absorbed and then excreted largely unchanged, and because xylitol is absorbed slowly and metabolized by an insulin-independent route, both produce little to no rise in blood glucose or insulin when they replace sugar. Livesey’s review of polyols as low-glycaemic sugar replacers documents exactly this: polyols have low glycaemic indices, with erythritol effectively zero, making them useful tools for reducing the glycaemic and insulin load of foods.⁵

This is the advantage that actually holds up under scrutiny, and it’s not trivial. For someone managing blood sugar — a person with diabetes, prediabetes, or anyone deliberately blunting glucose swings — swapping sugar for erythritol in a recipe removes the glucose spike that sugar would cause. That’s a real, measurable metabolic win, and it’s why these sweeteners earned their place in diabetic and low-carb products in the first place. It’s also worth noting the irony baked into the controversy: the very population most likely to reach for erythritol — people with metabolic disease — is the same population whose bodies produce more erythritol endogenously, which is part of why the cardiovascular association is so tangled. If you’re interested in the broader question of when chasing a flat glucose curve is worth it, we go deep on that in our read on glucose spikes and anti-spike hacks.

The dentist’s sweetener: xylitol and cavities

Xylitol has a second, older claim to fame that has nothing to do with metabolism: it fights cavities. This is the other moderate-graded benefit, and it’s genuinely established, if not as bulletproof as toothpaste marketing implies. The mechanism is clean: the bacteria most responsible for tooth decay, chiefly Streptococcus mutans, can’t ferment xylitol into the acid that erodes enamel. Feed them xylitol instead of sugar and they don’t produce acid; some evidence suggests they’re also suppressed over time. Replacing dietary sugar with a sweetener the decay bacteria literally cannot eat is a real dental positive.

The evidence backs a modest but genuinely contested effect. Some meta-analyses find xylitol reduces decay when it replaces sugar and is delivered regularly, yet a 2024 systematic review of nine human trials could not conclusively confirm the preventive effect, citing wide variation in doses and protocols and calling for standardized research.⁶ The honest counterweight is the 2015 Cochrane review, the most rigorous look at the question, which found the overall evidence was low-quality and that much of it came from xylitol chewing gum rather than the sweetener as a general sugar substitute — with only limited certainty about the size of the benefit.⁷ So the fair read is: xylitol is a legitimately tooth-friendly sweetener, and xylitol gum after meals is a defensible habit, but it’s a helpful adjunct, not a replacement for brushing and flossing. One hard caveat that isn’t about humans at all: xylitol is acutely toxic to dogs, causing dangerous blood-sugar crashes and liver damage even in small amounts. If you keep dogs, xylitol-containing gum and baked goods need to be stored well out of reach.

The reliable downside: your gut

Forget the clots for a moment — the side effect you’re actually most likely to experience is gastrointestinal, and it’s the most reliably documented downside of the whole category. This is the other moderate-graded claim, and anyone who’s eaten too many sugar-free gummies already knows the finding. Polyols are, by design, poorly absorbed — that’s what makes them low-calorie. But what isn’t absorbed in the small intestine draws water into the bowel (an osmotic effect) and gets fermented by gut bacteria in the colon (producing gas). Past an individual threshold, the result is bloating, cramping, flatulence, and a frank laxative effect.⁵

The dose-dependence is the practical key. Erythritol is the best tolerated of the polyols because it’s absorbed in the small intestine before it can reach the colon to ferment — so it causes less gas than most, though large amounts can still trigger the osmotic laxative effect. Sorbitol and maltitol are the usual culprits behind the “excess consumption may have a laxative effect” warning on candy packaging. Xylitol sits in between. Everyone has a personal tolerance ceiling, and it climbs somewhat with regular exposure as the gut adapts. The takeaway is simple and non-alarming: these are dose-limited by your own digestive system long before you’d reach anything resembling the bolus doses used in the clot experiments. In practice, most people’s upper limit on erythritol is set by their gut, not by a lab study — and that’s a useful natural brake.

The overreaction: “avoid entirely”

Now the claim that earns the hype pill: sugar alcohols are dangerous and you should throw them all out. That conclusion outruns the evidence in a way worth spelling out, because the reasoning error is instructive. The clot studies are real and deserve respect — but leaping from “high blood erythritol is associated with cardiovascular events, and a big bolus dose stirs up platelets” to “the erythritol in your coffee will give you a heart attack” skips over every caveat that matters: the association-versus-causation gap, the endogenous-production confounder, and the difference between a one-time 30-gram challenge and a habitual pinch.

There’s also a comparison the panic conveniently drops: what’s the alternative? The reason people use these sweeteners is to avoid sugar, whose harms across cardiovascular and metabolic health are extensively documented and not in dispute. If the choice is a keto dessert sweetened with erythritol versus the same dessert made with sugar, the sugar version’s risks are far better established than erythritol’s hypothetical ones. Fear-driven avoidance that sends someone back to sugar — or to ultra-processed “natural” syrups that behave like sugar metabolically — may trade a speculative risk for a well-proven one. Regulatory bodies including the FDA and EFSA have reviewed erythritol and continue to classify it as safe at typical intakes, and a 2023 Nutrients review even catalogued its potential upsides as a dietary component.⁸ None of that disproves the clot hypothesis — it just means the balanced position is caution and moderation, not blanket avoidance. “Throw it all out” is a bigger claim than the data can carry.

The verdict

Sugar alcohols are a case where the honest answer refuses to be a headline. On the reassuring side, the benefits are real and well supported: erythritol and xylitol are close to blood-sugar neutral, which is a genuine metabolic advantage over sugar; xylitol has a legitimate anti-cavity effect; and both deliver sweetness at a fraction of sugar’s calories.⁵⁶ The reliable cost is gastrointestinal — overdo it and your gut will tell you — which is annoying but self-limiting and not dangerous.⁵

The cardiovascular question is the one to hold with genuine uncertainty rather than false confidence in either direction. The clot signal from the 2023 erythritol study and the 2024 xylitol study is real, biologically coherent, and backed by human, animal, and mechanistic data — it should not be dismissed.¹⁴ But it is not proven to be causal: the human outcome data are associations in people who were already metabolically ill, the body makes erythritol itself so a high blood level may be a symptom rather than a cause, and the interventional experiments used single large bolus doses far beyond everyday use.²³ The randomized trial that would settle it hasn’t been run. So the clean way to hold all of it: occasional, moderate use is likely fine; the clot signal is real but unproven; and endogenous production means the scary number might be measuring your metabolism, not your sweetener. If you have established cardiovascular disease or a clotting disorder, it’s a reasonable conversation to have with your doctor. For everyone else, the move isn’t panic or dismissal — it’s moderation, a preference for erythritol over the harsher polyols, and keeping an eye on the research as the causal question gets properly tested.