The protein bar sat in its pristine wrapper, promising clean fuel for a low-carb lifestyle. One ingredient near the top of the ingredient list was erythritol, a sugar alcohol that has become a staple of the "keto" and "healthy" food aisles. Millions of consumers reach for products like this every day, trusting that erythritol's FDA-approved status means it is safe. A study published in July 2025 suggests that assumption deserves a harder look.
Researchers at the University of Colorado Boulder exposed brain endothelial cells, the cells that form the inner lining of blood vessels in the brain, to erythritol at concentrations matching what would be found in a person's bloodstream after a single sugar-free beverage. After three hours of exposure, the cells showed significant dysfunction. The findings, published in the Journal of Applied Physiology, add a new dimension to a growing body of research connecting erythritol not just to heart risks, but now to the delicate machinery that keeps the brain protected from blood clots and stroke [1].
Why the Blood-Brain Barrier Changes the Conversation
The blood-brain barrier is not a wall in the simple sense. It is a dynamic interface of endothelial cells joined by tight junctions, controlling what substances circulating in the bloodstream can reach brain tissue. These cells regulate vessel dilation, control inflammation, and produce the enzymes and signaling compounds that keep clotting in check. When those cells misfire, the consequences can be neurological.
Christopher DeSouza, a professor of integrative physiology at the University of Colorado Boulder and the study's senior author, framed the stakes directly: damage to these cells impairs the brain's ability to regulate blood flow and defend itself against clot formation, which is precisely what happens in ischemic stroke [1].
Before this study, most of the alarm around erythritol came from research on the cardiovascular system, not the brain. The 2023 Cleveland Clinic study led by Wittenborn and colleagues published in Nature Medicine found that individuals with the highest measured blood erythritol levels were approximately twice as likely to experience a major cardiac event such as a heart attack or stroke within three years [2]. That was the first large-scale signal that erythritol in the blood was not physiologically inert. The new data from DeSouza's laboratory provides a mechanistic hypothesis: if erythritol impairs brain endothelial cell function in the same way it appears to affect cardiovascular tissue, that could represent a plausible biological pathway linking consumption to stroke risk. The hypothesis is biologically coherent, but it has not yet been demonstrated in living human brains.
What the Cells Showed
The Colorado study used an in vitro model, applying erythritol to cultured human brain endothelial cells and observing what happened over a three-hour window. That is a relatively short exposure, and it matters. The doses used corresponded to what a person would absorb from a single serving of an erythritol-sweetened drink or food [1].
The results across several measured endpoints pointed in the same concerning direction:
Reduced nitric oxide production. Nitric oxide (NO) is a vasodilator produced by endothelial cells. It relaxes and expands blood vessels, improving blood flow. Cells treated with erythritol produced significantly less NO than untreated controls [1]. Lower NO availability means vessels are less able to dilate when the brain needs increased blood supply, and it also means less protective signaling against platelet clumping.
Increased endothelin-1 output. Endothelin-1 is a potent vasoconstrictor. Its levels rose in erythritol-treated cells [1]. More constriction means tighter, less flexible vessels. Combined with reduced NO, the net physiological environment shifts toward reduced flow and increased pressure at the vessel wall.
Suppressed t-PA release. Tissue plasminogen activator (t-PA) is the body's natural clot-busting compound. When endothelial cells were challenged with thrombin, a molecule that triggers clotting, the erythritol-exposed cells produced markedly less t-PA than normal [1]. This is the direct mechanism that links endothelial dysfunction to stroke risk: if the cells that line brain vessels cannot release adequate t-PA in response to a clotting signal, a clot that might otherwise be dissolved instead travels to or forms in the cerebral vasculature.
Elevated reactive oxygen species. Erythritol-treated cells generated significantly more reactive oxygen species (ROS), causing oxidative stress and measurable damage to the cells themselves [1]. Oxidative stress damages the endothelial lining over time, impeding the barrier's structural and functional integrity.
These four endpoints each play a documented role in stroke pathophysiology. The findings collectively warrant careful further study.
The Cardiovascular Evidence, Contextualized
The new study does not exist in isolation. The 2023 Nature Medicine paper from Wittenborn and colleagues, analyzing over 4,000 individuals in the United States and Europe, established that elevated erythritol in the blood correlated with a roughly doubled risk of major cardiac events within three years [2]. That study measured endogenous erythritol levels in participant blood and tracked outcomes prospectively, a design that carries more weight than a single case observation. The same Cleveland Clinic group had previously demonstrated that adding erythritol to isolated blood and platelets accelerated clot formation, establishing a mechanistic plausibility [2].
A follow-up human intervention study from the same Cleveland Clinic team, published in 2024 in Arteriosclerosis, Thrombosis and Vascular Biology, moved from correlation to causation. In twenty healthy volunteers, consuming an erythritol-sweetened food or drink produced a more than 1,000-fold increase in blood erythritol levels within hours [3]. In the same subjects, erythritol made platelets measurably more active, a condition that increases thrombotic risk. Glucose, tested under identical conditions, produced no such effect [3]. The senior author of that study was Stanley Hazen, MD, PhD.
These two studies, spanning a large observational cohort and a controlled intervention, give the 2025 brain-cell findings a backdrop. The mechanism proposed by the Colorado team, involving nitric oxide suppression, t-PA impairment, and oxidative stress in brain endothelial cells, does not contradict the platelet and cardiovascular data. It extends it. The blood-brain barrier endpoint is new, but the direction is consistent with a compound that affects vascular function broadly.
What This Means for Daily Choices
It is worth being precise about what we do and do not know.
The July 2025 University of Colorado findings demonstrate harm to brain endothelial cells in a controlled laboratory setting. No trial has yet measured erythritol's effect on stroke incidence in living human brains. In vitro studies cannot fully replicate the complexity of living physiology, and dietary erythritol consumption at realistic levels has not been proven to cause stroke in humans through this mechanism.
What the research does establish is that erythritol consumption raises blood erythritol concentrations by a factor of 1,000 or more within hours of intake [3]. Higher baseline erythritol in blood correlates with elevated cardiovascular event risk [2]. Erythritol impairs endothelial cell function in ways that directly involve nitric oxide regulation and the t-PA clotting pathway [1]. The blood-brain barrier's endothelial cells are the same cell type that lines blood vessels throughout the body, which is why cardiovascular findings and brain-cell findings are mechanistically related rather than separate concerns.
The population with highest exposure includes those who consume multiple servings daily of erythritol-sweetened products such as protein powders, energy drinks, keto snacks, and sugar-free beverages. The ScienceAlert coverage of the Colorado study included DeSouza's explicit call for regulatory review of whether erythritol's approval status should be revisited [4].
The Reporting Gap That Needs Closing
The evidence now is substantial enough that DeSouza specifically called for a reassessment of whether erythritol's regulatory status adequately reflects current science [4]. Food standards and regulations vary by jurisdiction. Food Standards Australia New Zealand maintains its own approval process for erythritol as a food additive, and readers should consult current TGA or FSANZ status for the most up-to-date Australian regulatory position.
Science journalism has a tendency to treat metabolic research as a parade of conflicting studies where no conclusion ever solidifies. That framing does a disservice to readers when the body of evidence has actually built a fairly consistent signal. Three independent research groups spanning the Cleveland Clinic and the University of Colorado have produced data across different methodologies that point in the same direction. That is not noise.
Whether regulators act, and on what timeline, remains to be seen. What is available now is the evidence itself. Readers with health concerns should consult a qualified healthcare professional before making dietary changes based on this research.