Erythritol: What It Is, Sweetness, Digestion, and Safety
Author: Ukrainian Psychological Hub · Published: September 29, 2026 · Editorial Policy
Erythritol is a four-carbon sugar alcohol, or polyol, used to provide sweetness and bulk with far less usable energy than sucrose. It occurs naturally in small amounts and is also produced by the human body, while commercial food-grade erythritol is typically made by fermenting carbohydrate sources and then purifying and drying the product. The European Food Safety Authority classifies erythritol (E 968) as a polyol sweetener and describes commercial production by fermentation.
For a person choosing or evaluating erythritol, four facts matter most. It is less sweet than table sugar, crystalline erythritol can create a distinctive cooling sensation as it dissolves, most ingested erythritol is absorbed rather than extensively fermented in the colon, and large doses can still cause gastrointestinal symptoms. The safety picture also has two different layers: the long-established dose-related digestive effects, and a newer cardiovascular research debate in which circulating erythritol, dietary exposure, platelet biology, and clinical outcomes must be kept separate.
This article owns the direct erythritol definition, sweetness, digestion, and safety intent. It does not treat erythritol as interchangeable with every sugar alcohol, every “artificial sweetener,” or every non-sugar sweetener. For the broader class comparison, see Sugar vs Sugar Alcohol: Sweetness, Calories, and Digestion.
Quick answer: what is erythritol?
Erythritol is a sugar alcohol rather than table sugar and rather than a high-intensity sweetener such as aspartame or sucralose. Chemically, “alcohol” describes part of the molecule’s structure; it does not mean beverage alcohol and erythritol does not contain ethanol in the ordinary sense. JECFA lists erythritol as 1,2,3,4-butanetetrol and identifies sweetener, humectant, flavor-enhancer, and carrier functions.
In sensory testing, erythritol is substantially less sweet than sucrose. A controlled psychophysical study found that its relative sweetness varied with concentration, ranging from about 0.50 to 0.78 relative to sucrose across the tested sweetness range. That is why “erythritol is 70% as sweet as sugar” is a useful shorthand, but not a universal constant. The concentration-dependent data come from Fujimaru, Park, and Lim.
For U.S. nutrition-label calculations, federal rules provide a general factor of 0 calories per gram for erythritol. That is a labeling factor, not a claim that every biochemical interaction is literally energy-free. It is also distinct from sucrose, which is ordinarily counted at about 4 calories per gram. The current eCFR nutrition-labeling rule also treats sugar alcohols separately from Total Sugars and Added Sugars.
Is erythritol natural, artificial, or something in between?
The natural-versus-artificial binary is a poor description of erythritol. Erythritol can occur naturally, can be synthesized endogenously in human metabolism, and can be manufactured at food scale by microbial fermentation. A review of mammalian erythritol metabolism describes endogenous production from glucose through the pentose phosphate pathway, which is crucial for interpreting studies that measure erythritol in blood rather than measured dietary intake. Ortiz and Field summarize this endogenous pathway.
Commercial erythritol is therefore not usually “extracted from fruit” in the way a simple naturalness story might imply. EFSA describes food-additive erythritol as being obtained by fermenting carbohydrate sources with suitable food-grade yeasts, followed by purification and drying. In the United States, FDA’s GRAS notice inventory includes several erythritol manufacturing processes; the most recent listed notice, GRN 1296, concerned erythritol produced from glucose by fermentation with Yarrowia lipolytica and closed in April 2026 with an FDA “has no questions” response.
That phrase should be read precisely. A GRAS “no questions” response is not the same thing as saying that FDA performed a premarket drug-style approval of every erythritol product. It means FDA did not question the notifier’s GRAS conclusion under the conditions described in that notice. The regulatory language matters because “FDA approved” and “FDA has no questions regarding this GRAS notice” are different claims.
How sweet is erythritol compared with sugar?
Erythritol is a bulk sweetener: gram-for-gram it contributes physical mass as well as sweetness. Sucrose is sweeter at the same mass in most ordinary formulations, so replacing sugar with erythritol by weight often produces a less sweet result unless the recipe is redesigned or erythritol is blended with a more intense sweetener. Human sensory data show that the exact relative sweetness changes with concentration rather than remaining fixed.
Sweetness also depends on the food matrix. A 2020 sensory study compared sweeteners in black tea, chocolate milk, and plain yogurt and found that erythritol was among the sweeteners whose overall taste profiles were relatively similar to sucrose across those foods, but the match was not identical and matrix effects mattered. The study found larger sweetener differences in some foods than others, illustrating why a sweetener that works well in one product can taste different in another.
At equisweet concentrations, temporal profiles can also differ. In a 2019 study of 16 sweeteners, erythritol retained a mainly sweet profile but could be associated with bitter, metallic, or chemical side-taste citations in some participants and time points. Tan and colleagues’ temporal sensory study is a useful reminder that “clean taste” is an average sensory description, not a promise that every person will perceive it identically.
Why does erythritol feel cooling in the mouth?
The characteristic cooling sensation is physical as well as perceptual. Crystalline erythritol absorbs heat from its surroundings as it dissolves, so dissolution on the tongue can lower local temperature and create a cool sensation without mint or menthol. A peer-reviewed review of erythritol production and properties describes this strong cooling effect as a consequence of its heat of solution. Regnat, Mach, and Mach-Aigner review the property here.
The effect is most noticeable when solid erythritol is dissolving during eating. If erythritol is already fully dissolved in a beverage, there is less dissolution left to happen in the mouth, so the cooling cue can be less prominent. In low-moisture foods, frostings, confectionery, powdered mixes, and crystalline tabletop products, the sensation can be much more obvious.
Psychologically, that cooling cue becomes part of the flavor object. Some people learn to associate it with “sugar-free” products, freshness, or a particular sweetener brand. Others interpret the same cue as unfamiliar or artificial. The sensation is therefore produced by chemistry, while liking is shaped by perception, context, familiarity, and learned expectation.
How erythritol differs from table sugar
Sucrose and erythritol can occupy the same culinary role—making food sweet—but they are not functionally interchangeable. Sucrose is a disaccharide that is hydrolyzed into glucose and fructose and provides about 4 calories per gram. Erythritol is a small polyol, is less sweet at comparable mass, and is handled very differently after ingestion. For a broader class-level explanation of calories, absorption, and digestive tolerance, see Sugar vs Sugar Alcohol.
Food technology differences matter too. Sugar contributes not only sweetness but also bulk, browning chemistry, crystallization behavior, water activity, texture, freezing-point effects, and structure. Erythritol contributes bulk and has useful crystallization and moisture properties of its own, but it does not reproduce every sucrose function. A recipe can therefore taste less sweet, feel cooler, crystallize differently, or brown differently even when the two ingredients look similarly white and granular.
How erythritol is absorbed and digested
Erythritol’s digestive behavior is unusual among common polyols because a large fraction is absorbed in the small intestine. Classic human metabolic data and later reviews indicate that most absorbed erythritol is excreted in urine, historically reported at roughly 90% at high oral doses, with relatively little appearing in feces. The comprehensive human and toxicology review by Munro and colleagues documents rapid gastrointestinal absorption and high urinary recovery.
That helps explain why erythritol often produces fewer fermentation-related symptoms than less completely absorbed sugar alcohols at comparable amounts. If less reaches the colon, there is less substrate available for colonic fermentation. But “better tolerated than another polyol” does not mean “impossible to cause symptoms,” and absorption itself does not settle every safety question.
The older description that erythritol passes through the body completely unchanged is also too simple. Modern work indicates that humans can produce erythritol endogenously and that a small portion of dietary erythritol may participate in metabolism before excretion. Ortiz and Field review evidence for endogenous synthesis and limited conversion of dietary erythritol to related metabolites. The central practical point remains that most ingested erythritol is absorbed and eliminated rather than extensively metabolized for energy.
Can erythritol cause gas, bloating, nausea, or diarrhea?
Yes, especially at higher doses, although the pattern differs from person to person and from other polyols. In a randomized double-blind study of 64 young adults who completed testing, single doses of 20, 35, or 50 g erythritol in water were compared with sucrose and xylitol. The 20- and 35-g erythritol doses did not significantly increase the measured gastrointestinal symptoms relative to the control condition, while 50 g increased reports of nausea and borborygmi. Xylitol produced more watery stools at comparable test doses. Storey and colleagues reported the full dose comparison.
Those numbers are study conditions, not universal personal thresholds. Gastrointestinal tolerance depends on dose, body size, how quickly the amount is consumed, whether the sweetener is taken with other foods, the rest of the formulation, and individual susceptibility. A large serving of several polyols together can behave differently from the same amount of erythritol alone.
EFSA’s 2023 re-evaluation focused heavily on the laxative endpoint. It identified 0.5 g/kg body weight as the lowest no-observed-adverse-effect level for diarrhea in the human evidence it assessed and set an acceptable daily intake of 0.5 g/kg body weight per day to protect against immediate laxative effects and possible secondary effects of repeated diarrhea. EFSA’s plain-language summary also concluded that the warning that excessive consumption may produce laxative effects remains valid.
This ADI should not be repackaged as a personalized “safe dose” for every person, and it is not a cardiovascular-risk threshold. It is a regulatory intake value derived from a gastrointestinal endpoint. For scale only, 0.5 g/kg corresponds to 35 g/day for a 70-kg adult, but individual tolerance can be lower or higher and a product may contain other ingredients that affect the gut.
Does erythritol affect blood glucose or insulin?
Acute human studies generally find little or no rise in glucose or insulin after erythritol alone, and EFSA described the evidence as limited but consistent for no meaningful effect on blood sugar levels. EFSA’s 2023 review and FDA’s sweetener overview both distinguish sugar alcohols from ordinary sugar in this respect.
That fact is about the ingredient under study; it does not turn every food containing erythritol into a low-carbohydrate food and it does not provide a blood-glucose target or diabetes-treatment instruction. A “sugar-free” cookie can still contain starch, fat, calories, and other carbohydrates. Blood glucose readings, A1C targets, continuous glucose monitoring, hypoglycemia, and personalized diabetes management belong to clinical care rather than to this ingredient article.
Does erythritol affect appetite or satiety?
Small mechanistic studies suggest that erythritol can influence gastrointestinal signaling. In a randomized crossover pilot study of 12 healthy participants, intragastric erythritol produced dose-dependent increases in CCK, active GLP-1, and PYY and slowed gastric emptying, without increasing glucose or insulin. Wölnerhanssen and colleagues reported these acute findings.
This is preliminary mechanistic evidence, not proof that erythritol causes meaningful long-term weight loss or reliably suppresses eating in everyday life. Acute hormone changes, subjective fullness, later food intake, body weight, and long-term health outcomes are different endpoints. The practical evidence status is therefore: plausible acute gut signaling, limited evidence for behavioral consequences, and no basis for treating erythritol itself as a weight-loss therapy.
Sensory psychology: why erythritol does not taste exactly like sugar
A sweetener is not experienced as a molecule in isolation. The brain integrates receptor activation with temperature, aroma, texture, aftertaste, visual expectation, prior experience, and the food in which the sweetener appears. Erythritol therefore differs from sucrose in at least three simultaneous channels: sweetness intensity, the cooling cue produced during dissolution, and its temporal/side-taste profile.
The primary oral sweet signal is still part of human sweet-taste biology. For the receptor-level mechanism and why sugars and structurally different sweeteners can converge on a sweet percept, see Sweet Taste Receptors: How Humans Detect Sugar and Sweeteners. Erythritol’s perceptual result then depends on concentration and the rest of the food matrix, as sensory studies show.
Expectations can matter in food perception, but they should not be assumed to override chemistry. A 2025 fMRI study in 30 healthy men matched erythritol and sucrose for perceived sweetness intensity and manipulated “low-calorie” versus “high-calorie” labels. Participants liked erythritol less than sucrose, but the calorie label itself did not significantly change liking or the main neural responses. Budzinska and colleagues reported this null labeling effect. That is useful evidence against the simplistic claim that a health or calorie label automatically makes a sweetener taste better.
Learned preference, reward, and the meaning of “sweetness without sugar”
Sweetness can become linked to flavors through learning. In a 2026 randomized double-blind crossover study of 20 healthy adults, novel flavors were paired with erythritol, sucrose, or sucralose. Wanting and liking increased after conditioning, with no significant difference among the three sweeteners. The measured gastrointestinal hormones, glucose, and insulin did not explain the reward response. Flad and colleagues interpreted the pattern as consistent with flavor-flavor learning rather than a simple nutrient-reward mechanism.
This finding supports a narrow psychological conclusion: a low-calorie sweetener can participate in learned flavor liking because sweetness itself is a salient sensory event. It does not establish “erythritol addiction,” withdrawal, or a clinical disorder. Preference learning, cue-triggered wanting, habit, and substance addiction are different constructs and should not be collapsed into one vocabulary.
Is erythritol safe? The evidence needs two separate answers
For the long-established toxicology and gastrointestinal question, erythritol has a substantial regulatory and experimental history. JECFA evaluated erythritol in 1999 and listed its acceptable daily intake as “not specified,” while EFSA’s much newer 2023 re-evaluation established an ADI of 0.5 g/kg/day based on the diarrhea endpoint. These positions reflect different assessment histories, datasets, and regulatory approaches rather than a simple contradiction. JECFA’s current database entry and EFSA’s current re-evaluation should be read in their own contexts.
In the United States, FDA permits sugar alcohols including erythritol as sugar substitutes and continues to monitor new safety information. FDA specifically states that its review of the 2023 cardiovascular paper found that the observational studies cited there did not establish a causal link between consuming erythritol and the observed cardiovascular outcomes. FDA’s current sweetener page makes that distinction explicitly.
The cardiovascular question is newer and genuinely unresolved. It deserves more than either “erythritol is proven dangerous” or “regulators allow it, therefore every cardiovascular concern is settled.” The studies measure different exposures and endpoints, so they answer different questions.
What the 2023 cardiovascular study found
A 2023 Nature Medicine study found that higher circulating erythritol concentrations were associated with a higher three-year risk of major adverse cardiovascular events in cohorts of patients undergoing cardiac risk assessment. In two validation cohorts, the highest versus lowest quartile of circulating erythritol was associated with higher adjusted event risk. The study also included platelet and thrombosis experiments and a small eight-person ingestion component. The original study is available through PubMed.
The association is important, but circulating erythritol is not the same variable as measured dietary erythritol intake. Humans synthesize erythritol endogenously, and metabolic dysfunction can influence circulating levels. The cohorts were also enriched for people already being evaluated for cardiovascular risk. These features create a major causal-interpretation problem: high blood erythritol might partly mark underlying metabolism, might partly reflect recent dietary exposure, or might participate biologically in risk through mechanisms that still require clarification.
A critical review by Mazi and Stanhope emphasized exactly this problem and concluded that the available evidence at that time could not establish that dietary erythritol causes thrombosis or cardiometabolic disease, while also calling for long-term clinical trials. Their 2023 review is useful because it separates the biomarker question from the food-exposure question.
What the 2024 platelet study added
A 2024 brief interventional report gave 30 g erythritol to 10 healthy volunteers and 30 g glucose to another 10. Erythritol ingestion produced a very large acute rise in plasma erythritol and increased several laboratory measures of stimulus-dependent platelet responsiveness; the glucose group did not show the same pattern. Witkowski and colleagues reported the study in Arteriosclerosis, Thrombosis, and Vascular Biology.
This strengthens the biological plausibility of an acute platelet effect, but it still does not measure heart attacks, strokes, or long-term clinical outcomes. The sample was very small, exposure was acute, and platelet-reactivity endpoints are not interchangeable with event rates. The result is therefore meaningful mechanistic evidence and a reason for further research, not a finished estimate of long-term dietary risk.
What longer and newer human evidence shows
A small five-week randomized pilot trial in 42 adults with obesity compared daily erythritol, xylitol, or no added study substance. Thirty-six grams of erythritol per day produced no statistically significant changes in the measured vascular-function, abdominal-fat, glucose-tolerance, uric-acid, liver-enzyme, or creatinine outcomes, and gastrointestinal tolerance was generally good apart from some diarrhea-related symptoms. Bordier and colleagues reported the pilot trial. Its short duration and small sample mean it cannot rule out uncommon or long-latency cardiovascular outcomes.
A 2025 Cardiovascular Research review integrated the emerging erythritol and xylitol literature. It noted temporary platelet-aggregation findings from recent pilot work while also emphasizing endogenous polyol production, evidence from other clinical contexts, and the absence of definitive long-term causal data. Wölnerhanssen and colleagues’ review therefore supports a position of active uncertainty rather than a binary safety verdict.
The newest relevant study located for this review was published online on September 18, 2026. In an observational analysis within the POUNDS Lost randomized weight-loss trial, higher baseline plasma erythritol was associated with higher estimated ASCVD risk, and declines in plasma erythritol during the intervention were associated with improvements in estimated risk and some atherogenic lipid measures. Heianza and colleagues reported these associations in The American Journal of Clinical Nutrition.
That 2026 analysis still does not randomize people to dietary erythritol exposure. It measures plasma erythritol, a metabolite that can be made by the body, and relates changes in that biomarker to cardiometabolic changes during weight-loss interventions. It adds evidence that circulating erythritol tracks metabolic and cardiovascular-risk phenotypes; it does not by itself show that eating erythritol caused those phenotypes.
What can responsibly be concluded about cardiovascular safety?
Established evidence: eating erythritol can acutely raise circulating erythritol; humans also produce erythritol endogenously; high circulating erythritol has been associated with cardiovascular-risk measures and events in several observational settings; and a very small acute intervention found increased platelet reactivity after a 30-g dose. These statements are supported by direct human data.
Unresolved evidence: whether habitual dietary erythritol consumption causes a clinically meaningful increase in heart attack, stroke, thrombosis, or cardiovascular mortality; whether any such effect is dose-dependent in free-living populations; whether specific risk groups are more susceptible; and what long-term intake threshold, if any, would map to cardiovascular outcomes. Current research does not provide a validated cardiovascular “safe limit.”
Practical meaning: cardiovascular concerns should not be dismissed, and they should not be inflated into a settled causal claim. People who use large amounts of erythritol regularly—especially those with established cardiovascular disease or major risk factors—may reasonably discuss that pattern with a clinician while the evidence develops. That is a risk-management conversation, not a diagnosis or an instruction to substitute one specific sweetener for everyone.
Erythritol in sugar-free foods and sweetener blends
Erythritol commonly appears in products marketed as sugar-free or with reduced sugar because it provides bulk and sweetness without being counted as Total Sugars or Added Sugars in the same way sucrose is counted on U.S. labels. A sugar-free claim, however, describes the sugar content under regulatory criteria; it does not mean the food contains no carbohydrate, no calories, no sugar alcohols, or no other sweeteners. For the exact U.S. claim rules and the psychology of the label, see Sugar-Free: What the Label Means and What Sweeteners May Replace Sugar.
Erythritol is also frequently used as the bulk component in blends with high-intensity sweeteners. This solves a practical problem: a tiny amount of a very intense sweetener can provide sweetness but not the spoon-for-spoon mass, texture, or handling properties consumers expect. The ingredient list therefore matters. A package marketed around stevia or monk fruit may or may not contain erythritol; the front label alone does not establish the full formulation.
The WHO 2023 recommendation on non-sugar sweeteners is often misapplied here. WHO explicitly states that its recommendation does not apply to low-calorie sugars and sugar alcohols (polyols), because those substances are outside the guideline’s NSS definition. WHO’s own summary states this exclusion. That does not mean WHO declared erythritol risk-free; it means that particular NSS guideline is not the correct evidence document for judging erythritol.
Does erythritol count as added sugar on a U.S. Nutrition Facts label?
No. Erythritol is a sugar alcohol, not an Added Sugar for U.S. Nutrition Facts purposes. It can contribute to Total Carbohydrate while being handled separately from Total Sugars and Added Sugars, and “Sugar Alcohol” may appear as its own declaration depending on the product and claims. The current federal Nutrition Facts regulation contains the operative definitions and labeling rules.
This is another reason not to infer ingredient quantity from the Added Sugars line alone. Two products can both display 0 g Added Sugars and still contain very different amounts and types of polyols, starches, fibers, fats, proteins, and high-intensity sweeteners.
Erythritol versus xylitol and other sugar alcohols
Erythritol belongs to the same broad polyol class as xylitol, sorbitol, maltitol, mannitol, lactitol, and isomalt, but class membership does not make their properties identical. They differ in molecular size, absorption, fermentation, caloric factors, relative sweetness, cooling, dental effects, and gastrointestinal tolerance. The Sugar vs Sugar Alcohol article provides the class-level comparison without turning any one polyol into the template for all others.
For digestion specifically, erythritol is absorbed more completely than many larger polyols, which helps explain its generally lower fermentation burden in the colon. The 2007 dose study also found fewer measured gastrointestinal effects from erythritol than xylitol at comparable test doses. That evidence does not justify saying erythritol is universally “gentler”; tolerance remains dose- and person-dependent.
Using erythritol in cooking, baking, and drinks
If a recipe was developed for sucrose, erythritol is not automatically a one-for-one sensory or functional replacement. Because it is less sweet, a weight-for-weight swap can reduce sweetness. Increasing erythritol to compensate can change bulk and crystallization and can make the cooling sensation more noticeable. In baking, sugar’s browning and structural roles also mean that a technically successful substitution may require more than matching sweetness.
In beverages, dissolved erythritol may contribute sweetness with less obvious cooling than crystalline erythritol dissolving directly on the tongue. In dairy, tea, and other matrices, sensory work shows that the surrounding food changes how close a sweetener comes to sucrose. That is why tasting the complete product is more informative than judging a sweetener from a spoonful of crystals alone.
Blends are common because they can solve multiple formulation problems at once: erythritol can provide bulk while a high-intensity sweetener supplies additional sweetness. But blends also complicate causal interpretation. If a product causes an aftertaste or gastrointestinal symptom, the responsible factor could be erythritol, another polyol, fiber, flavoring, fat composition, or the combination rather than the front-of-pack sweetener name.
Evidence status: what is established, preliminary, and unresolved
Established
Erythritol is a sugar alcohol used as a sweetener and food ingredient; it is commercially produced by fermentation; it is less sweet than sucrose; crystalline erythritol can create a cooling sensation as it dissolves; most ingested erythritol is absorbed and largely excreted; U.S. labeling assigns it 0 calories per gram; and sufficiently high intake can produce gastrointestinal or laxative effects. These points are supported by regulatory assessments, human metabolism studies, and sensory research.
Preliminary or mechanistic
Small human studies indicate acute effects on gut hormones and gastric emptying, laboratory measures of platelet reactivity, neural responses to sweet taste, and learned flavor preference. These findings help explain possible mechanisms but do not establish long-term clinical outcomes or universal behavioral effects.
Contested or unresolved
The central unresolved question is whether long-term dietary erythritol exposure itself causes cardiovascular events. Observational blood-metabolite associations, endogenous erythritol production, short-term ingestion studies, acute platelet findings, and limited longer interventions do not yet form a single causal estimate. The correct evidence label is active uncertainty with biologically plausible concern—not proven harmlessness and not proven dietary causation.
Frequently asked questions
Is erythritol the same as sugar?
No. Erythritol is a sugar alcohol, while ordinary table sugar is sucrose. They differ in chemistry, sweetness per gram, energy value, digestion, labeling, browning, crystallization, and sensory properties.
Is erythritol an artificial sweetener?
It is more precise to call erythritol a sugar alcohol or polyol sweetener. It occurs naturally and is produced by human metabolism, while food-grade erythritol is commonly manufactured by fermentation. It is not the same class as high-intensity sweeteners such as aspartame or sucralose.
How sweet is erythritol?
Human sensory testing has placed erythritol at roughly 50% to 78% of sucrose sweetness across tested concentrations rather than at one fixed ratio. The underlying psychophysical study is here.
Why does erythritol taste cold?
Crystalline erythritol absorbs heat as it dissolves. That physical cooling changes oral sensation and can become part of the perceived flavor. It is strongest when crystals are dissolving in the mouth and can be less obvious when erythritol is already dissolved.
Does erythritol have calories?
For U.S. nutrition labeling, the general caloric factor is 0 calories per gram. The current rule is in 21 CFR 101.9. This is a regulatory labeling value and should not be confused with a claim that erythritol is biologically inert.
Can erythritol cause diarrhea?
Yes. Digestive tolerance is dose-dependent and varies among individuals. EFSA’s 2023 re-evaluation set an ADI of 0.5 g/kg body weight per day based on the diarrhea endpoint and retained the warning that excessive consumption may have laxative effects. EFSA explains the basis here.
Is erythritol safe for the heart?
A causal long-term answer has not been established. High circulating erythritol has been associated with cardiovascular risk and acute ingestion has increased platelet reactivity in a very small study, but blood erythritol is also produced endogenously and existing studies do not yet show that habitual dietary erythritol causes cardiovascular events. Long-term exposure trials with clinical endpoints are still needed.
Does the WHO non-sugar-sweetener recommendation apply to erythritol?
No. WHO explicitly excludes low-calorie sugars and sugar alcohols, or polyols, from that 2023 recommendation. The WHO summary states the exclusion directly.
Does erythritol count as Added Sugars?
No under U.S. Nutrition Facts rules. Erythritol is a sugar alcohol and is not counted as Added Sugars in the way sucrose, syrups, or other added caloric sugars are. It can still contribute to Total Carbohydrate and may be separately declared as Sugar Alcohol.
Can erythritol still influence preference even though it has little energy?
Yes, at least in experimental flavor-learning settings. A 2026 randomized crossover study found increased wanting and liking after flavors were paired with erythritol, sucrose, or sucralose, with no significant differences among sweeteners. That evidence supports learned flavor preference, not an addiction diagnosis.