Allulose, stevia and monk fruit are often sold together as the "natural" sugar substitutes and promoted as healthier than sucralose and aspartame. But they are three very different things. Allulose is a rare sugar your body absorbs but does not burn. Stevia and monk fruit are plant extracts hundreds of times sweeter than sugar, so they are used in tiny amounts. And many products sold under their names are mostly something else.
This guide covers what each one is, what the trials show for blood sugar, weight and side effects, how US and European regulators treat them, what the World Health Organization advises, and how to read the labels on sweeteners and sweetened supplements. For erythritol, xylitol and the other sugar alcohols, see our companion guide to sugar alcohols.
The short version
- None of the three raises blood sugar much on its own. Allulose has the most evidence for modestly lowering the rise after a carbohydrate meal, but a 2026 meta-analysis (a study that pools the results of several trials) found no effect on HbA1c (long-term blood sugar), fasting glucose or body composition.
- Allulose can upset your stomach at larger doses. In one study, single doses above about 0.4 grams per kilogram of body weight (about 28 grams for a 70 kilogram adult, our arithmetic) caused diarrhea. It is not authorized in the EU, where regulators said in 2025 its safety could not be established with the data provided.
- Stevia sweeteners are purified extracts. The FDA has not objected to high-purity (at least 95 percent) steviol glycosides, the sweet compounds in the leaf, while whole stevia leaf is not permitted as a sweetener in ordinary foods. Claims that stevia lowers blood pressure or blood sugar rest on doses far above what sweetens food.
- Monk fruit has very little human research as a sweetener: a few single-day studies, no long-term trials, and European regulators said in 2019 they did not have enough safety data.
- Read the ingredient list. Most "stevia" and "monk fruit" packets are mainly dextrose, erythritol, allulose, maltodextrin or inulin, and that filler can matter more than the sweetener.
- For weight, the WHO advises against relying on non-sugar sweeteners, including stevia, though newer year-long trials found small benefits for keeping weight off.
Allulose, stevia and monk fruit at a glance
| Allulose | Stevia | Monk fruit | |
|---|---|---|---|
| What it is | A rare sugar | Purified leaf extract | Fruit extract |
| Sweetness vs sugar | About 70% | 200 to 400 times | 100 to 250 times |
| Calories (US label) | 0.4 per gram | Near zero | Near zero |
| US status | GRAS, FDA has no questions | GRAS, FDA has no questions (at least 95% pure) | GRAS, FDA has no questions |
| EU status | Not authorized | Authorized (E 960) | Not authorized as a sweetener additive |
| Human evidence | Many short trials | Many trials | Very few, single-day |
GRAS means "generally recognized as safe." For these sweeteners, companies submitted their own safety conclusions and the FDA replied that it had no questions; the FDA's letters note that this is not its own determination, so "FDA approved" is not accurate.
Is allulose safe? Side effects and blood sugar
What it is. Allulose (also called D-psicose) is a rare sugar found in small amounts in foods such as raisins and figs; commercial allulose is made from fructose using enzymes. It tastes and bakes much like sugar at about 70 percent of the sweetness, so it is used gram for gram. The body absorbs most of it but cannot use it for energy: in a small study, about 70 percent left in the urine. Since 2020, the FDA has let companies leave allulose out of "Total Sugars" and "Added Sugars" on Nutrition and Supplement Facts labels and count it at 0.4 calories per gram, pending a formal rule. It still counts toward Total Carbohydrate.
Blood sugar. Taken with a carbohydrate meal or drink, 5 to 10 grams of allulose modestly lowered the blood sugar rise in several small trials, though not all of them. A 2026 meta-analysis of 12 allulose trials found a moderate-certainty reduction in the post-meal glucose rise, but no effect on HbA1c, fasting glucose or insulin, blood fats, or body composition. Many of the individual trials were funded, supplied or co-written by allulose makers. Allulose is a useful sugar swap for people watching blood sugar, but it is not a diabetes treatment.
Weight. One 12-week Korean trial of 121 adults, using drinks supplied by an allulose maker, reported less body fat at 4 to 7 grams twice a day, and its authors called it preliminary. Pooled trials found no effect on body composition.
Side effects. Large amounts cause gas, bloating, stomach pain and diarrhea. In a non-randomized study of young Korean adults, severe diarrhea appeared at a single dose of 0.5 grams per kilogram of body weight. The researchers suggested maximum amounts of 0.4 grams per kilogram in one dose and 0.9 grams per kilogram per day, roughly 28 and 63 grams for a 70 kilogram adult (our arithmetic). In a 30-day study of 24 to 36 grams a day (with no placebo group), about half of participants had mild, short-lived stomach symptoms, mostly in the first 3 days. Two studies reported unexplained lab changes that need confirming. In an independent 12-week crossover trial of 16 people with type 2 diabetes, HDL ("good") cholesterol fell and an inflammation marker rose while they took 7 grams twice a day, with no change in blood sugar control or body composition. In the 30-day study, bone density scores fell at the higher dose in participants 35 and older, though the authors said they stayed within normal ranges. We found no allulose trial longer than 12 weeks in PubMed searches through November 2026, so long-term safety has not been well studied.
Why it is not sold in Europe. In 2025, Europe's food safety agency concluded that the safety of allulose "cannot be established," because the applicant did not answer its requests for more data on production, uses, genotoxicity (potential to damage DNA) and human studies. That is a statement about an incomplete application, not a finding that allulose is harmful.
Is stevia safe? Side effects and blood sugar
What it is. Stevia sweeteners are steviol glycosides, such as rebaudioside A ("Reb A") and Reb M, extracted from the leaves of the Stevia rebaudiana plant and purified. Some newer forms are made by converting other stevia compounds with enzymes or by fermentation. Because they are 200 to 400 times sweeter than sugar, a packet contains only milligrams.
Regulation. The FDA has not objected to GRAS notices for steviol glycosides that are at least 95 percent pure. Whole stevia leaf and crude extracts are not considered GRAS and may not be imported for use as sweeteners in ordinary foods. They can still be sold as a dietary ingredient in supplements, such as liquid stevia leaf extracts. But when the label presents them as a sweetener or flavoring, the FDA treats them as unsafe food additives. The acceptable daily intake (the amount considered safe to eat every day) set by international experts is 4 milligrams per kilogram of body weight, expressed as steviol, and WHO and FAO experts reconfirmed it in 2026. In 2024, Europe's food safety agency declined a request to raise it. By the FDA's arithmetic, a 60 kilogram adult would need about 27 packets of a Reb A tabletop sweetener a day to reach that limit. Stevia is authorized in the EU as E 960.
Blood sugar. Purified stevia has no meaningful effect on blood sugar. In short trials, stevia drinks moved glucose and insulin no more than water, and in a 16-week trial of 1,000 milligrams of Reb A a day in people with type 2 diabetes, HbA1c changed no more than with a placebo. A 2024 pooled analysis found a small drop in blood glucose, rated low certainty, mainly at about 3.3 grams of stevia a day or more, far above sweetening amounts. It found no change in HbA1c. Some "stevia" packets contain dextrose or maltodextrin, which do add a little sugar.
Blood pressure. Two trials in Taiwan found lower blood pressure with 750 to 1,500 milligrams a day of stevioside capsules for 1 to 2 years. But a Brazilian trial in people with mild high blood pressure found no effect beyond placebo, and Reb A did not change blood pressure in trials of people with normal blood pressure or type 2 diabetes. Those capsule doses are far above what you would eat as a sweetener, so stevia should not be counted on to lower blood pressure.
Gut bacteria and allergy. In a two-week trial, stevia changed the makeup of gut and mouth bacteria but, unlike saccharin and sucralose, did not significantly worsen glucose tolerance; two 4- and 12-week trials found no significant changes. Stevia belongs to the daisy family, which includes ragweed, but a 2015 review found that the few reported allergic reactions predate today's high-purity extracts. In our PubMed searches through November 2026, we found no later case reports.
Monk fruit sweetener: is it safe?
What it is. Monk fruit, or luo han guo, is a small gourd native to southern China. Its sweetness comes from compounds called mogrosides, especially mogroside V. Gut bacteria break much of it down, but Europe's food safety agency noted that mogroside V is absorbed to some extent and reaches the bloodstream. Extracts are 100 to 250 times sweeter than sugar.
Regulation. The FDA has not questioned GRAS notices for monk fruit extract since 2010, and it has not set an acceptable daily intake, noting that one may not be needed. In 2019, Europe's food safety agency said the toxicity data on a monk fruit extract were "insufficient to conclude on the safety" of using it as a food additive. It cited effects on the testes in one 90-day rat study that "cannot be dismissed." A later 90-day rat study, written by employees of a biotechnology company, found no such effects. But it tested a high-purity mogroside preparation made by microbial fermentation, not the fruit extract Europe assessed. We could not confirm claims online that the EU has since authorized a monk fruit extract. A 2025 review describing monk fruit extract as partially approved in the EU was retracted in 2026, after the journal found a significant number of non-existent references in it.
Human evidence. It is very thin. In two small single-day studies of healthy men in Singapore, total blood sugar over 3 hours and over 24 hours did not differ between a monk fruit drink and a sugar drink. In the 3-hour study, which also measured insulin, the sugar drink caused a sharp early spike in blood sugar and insulin, while responses to the monk fruit drink were higher after the next meal, so total insulin did not differ either. People in that study also ate more at later meals, making up for all the calories the sweetener saved. In PubMed searches through November 2026, we found no trial of monk fruit as a sweetener lasting longer than a single day, and none on weight, HbA1c or blood pressure. Claims about mogrosides and diabetes or antioxidant effects come from cell and animal studies.
What is really in the packet
Stevia and monk fruit are so sweet that a packet needs only a speck, so products add a bulking agent to make them pour and measure like sugar. Monk Fruit In The Raw, for example, describes its products as blends of monk fruit extract with dextrose, erythritol or allulose, depending on the version. Supplement labels in the NIH label database show carriers such as maltodextrin, inulin and erythritol, and formulas change: one electrolyte mix listed erythritol in 2022 and no longer did in 2025.
That filler matters. Dextrose and maltodextrin are sugars, in small amounts. Erythritol can cause nausea and diarrhea in large amounts and is the subject of a heart-risk debate (see our sugar alcohols guide). Inulin and similar fibers can cause gas, and a 2026 study of monk fruit blends containing fructooligosaccharides reported more diarrhea. A "zero calorie" label is allowed when a serving has fewer than 5 calories, which a mostly dextrose packet can meet.
Do sweeteners help with weight? The WHO advice
In 2023, the World Health Organization issued a conditional recommendation against using non-sugar sweeteners "as a means of achieving weight control or reducing the risk of noncommunicable diseases" (long-term conditions such as type 2 diabetes and heart disease). It names stevia and stevia derivatives. Monk fruit is not named but fits its definition. Sugar alcohols and low-calorie sugars are excluded; the guideline does not name allulose, but as a low-calorie sugar it falls outside the definition. The WHO says the advice may not be relevant for people who already have diabetes, and that it is not based on safety assessments of individual sweeteners.
The evidence behind it is low certainty. In the WHO-commissioned review, short trials showed a small weight loss of about 0.7 kilograms, which was no longer significant once industry-funded trials were removed, while long-term observational studies linked sweetener use with higher risks of type 2 diabetes and heart disease. Those links may partly reflect reverse causation, since people already gaining weight or developing diabetes often switch to sweeteners. Two newer year-long trials point the other way. In the UK and in the European SWEET project, people who used sweetened products kept off about 1.4 to 1.6 kilograms more weight than comparison groups. The UK trial's authors called the difference not clinically significant, both trials had industry ties, and SWEET tested a mix of sweeteners, including some sugar alcohols, not stevia alone.
The practical reading: swapping sugar for a sweetener cuts sugar and calories, but it does not guarantee weight loss, partly because people may eat the saved calories back later. In the Singapore study of monk fruit, stevia and aspartame drinks, people made up all the saved calories at later meals; in a US study of stevia and aspartame preloads, they did not.
Are natural sweeteners better than artificial ones?
Not clearly. The FDA has no formal definition of "natural," only a longstanding policy that nothing artificial or synthetic has been added. High-purity stevia and monk fruit extracts are heavily processed, some are now made by fermentation, and allulose is made from fructose with enzymes. The least processed form, whole stevia leaf, is the one the FDA does not allow as a sweetener in ordinary foods.
Some artificial sweeteners carry their own questions. In 2023 the WHO's cancer agency classified aspartame as "possibly carcinogenic to humans," while the WHO and FAO expert committee kept its acceptable daily intake unchanged and the FDA disagreed with the classification. A lab study found that sucralose-6-acetate, an impurity in sucralose, damaged DNA in cell tests (see our sucralose page). In the two-week microbiome trial, saccharin and sucralose significantly worsened glucose responses while stevia and aspartame did not. But we found no trial comparing natural and artificial sweeteners on disease outcomes, and in the Singapore study, aspartame, stevia and monk fruit drinks had similar effects.
Sweeteners in supplements and protein powders
Stevia, monk fruit, sucralose and allulose are common in protein powders, electrolyte mixes, chewables and gummies. In supplements they usually appear in the "other ingredients" list, though some stevia and monk fruit products list them in the Supplement Facts panel. Allulose follows the same label rule as in food: it counts in Total Carbohydrate but not in Total or Added Sugars. If you are sensitive to sugar alcohols or follow a low-FODMAP diet, check whether a "monk fruit" or "stevia" product also contains erythritol, inulin or other fibers. If you are fasting, see do supplements break a fast?
Which sugar substitute is best, including for diabetes?
- For baking and texture, allulose behaves most like sugar and browns well, but large servings can upset your stomach.
- For drinks and small amounts, purified stevia or monk fruit add sweetness with essentially no calories; choose by taste, since stevia can taste bitter or lingering to some people.
- For blood sugar or diabetes, all three are reasonable swaps for sugar; check that the product is not mainly dextrose or maltodextrin, and if you take diabetes medicines, tell your doctor about changes to your diet.
- For a sensitive gut, watch the filler, especially erythritol, inulin and large amounts of allulose.
- Overall, the WHO's point stands: sweeteners are a way to cut sugar, not a health food, and drinking water or gradually getting used to less sweetness are simple alternatives.
Frequently asked questions
Is allulose safe?
In the US, the FDA has not questioned several companies' conclusions that allulose is generally recognized as safe, and short trials of up to 12 weeks reported mainly digestive side effects, plus a few unexplained lab changes that need confirming. Long-term safety has not been well studied. Allulose is not authorized in the EU, where regulators said in 2025 that its safety could not be established because the application was incomplete.
What are the side effects of allulose?
Gas, bloating, stomach pain and diarrhea, mainly with larger amounts. In one study, the researchers suggested maximum amounts of about 0.4 grams per kilogram of body weight in a single dose and 0.9 grams per kilogram per day, roughly 28 and 63 grams for a 70 kilogram adult. Starting with small amounts helps.
Does allulose raise blood sugar?
No. Allulose is absorbed but not used for energy, and taken with a carbohydrate meal it modestly lowered the blood sugar rise in several small trials. A 2026 meta-analysis found no effect on HbA1c (long-term blood sugar) or fasting glucose, so it is a sugar swap, not a diabetes treatment.
Is allulose better than erythritol?
They suit different uses. Allulose bakes and browns more like sugar, while erythritol has zero calories on US labels and is very well absorbed. Both can cause digestive symptoms in large amounts. Erythritol is the subject of a debated heart-risk signal from blood-level studies; in PubMed searches through November 2026, we found no comparable studies of allulose.
Is allulose keto-friendly?
Allulose counts at 0.4 calories per gram on US labels, is left out of Total Sugars and Added Sugars, and does not raise blood sugar much on its own, which is why it is popular for keto and low-carb baking. It still counts toward Total Carbohydrate on the label, so check how your plan counts it, and keep servings moderate, since large amounts can upset the stomach.
Is stevia safe?
Purified stevia extracts (the leaf's sweet compounds, called steviol glycosides, at least 95 percent pure) are widely used, and the FDA has not questioned their GRAS (generally recognized as safe) status. International experts set an acceptable daily intake (the amount considered safe to eat every day) of 4 milligrams per kilogram of body weight. Typical use stays well below it; conservative estimates for young children in Europe who consume a lot can slightly exceed it, which WHO and FAO experts judged not likely to be a health concern. Whole stevia leaf is not permitted as a sweetener in ordinary US foods.
Does stevia raise blood sugar or insulin?
Purified stevia has no meaningful effect on blood sugar or insulin; in short trials it behaved like water. Some stevia packets contain dextrose or maltodextrin, which add a small amount of sugar, so check the ingredients.
What are the side effects of stevia?
Some people find stevia bitter or lingering in taste. In a two-week trial, stevia changed the makeup of gut and mouth bacteria but did not significantly worsen glucose tolerance, and 4- and 12-week trials found no significant changes. Stevia belongs to the daisy family, which includes ragweed, but a 2015 review found that the few reported allergic reactions predate today's high-purity extracts, and we found no later case reports in PubMed searches through November 2026. Gas or diarrhea from a stevia product can come from added erythritol, inulin or other fibers.
Is monk fruit safe?
The FDA has not questioned companies' GRAS (generally recognized as safe) notices for monk fruit extract since 2010. Human research on monk fruit as a sweetener is very limited, with no trials longer than a single day that we found, and European regulators said in 2019 there were not enough safety data to approve it as a food additive.
Monk fruit vs stevia: which is better?
Neither has been shown to be healthier. Both add sweetness with essentially no calories and little effect on blood sugar. Stevia has far more research and a set intake limit; monk fruit has very little human data. Taste and the filler in the product, such as erythritol or dextrose, often matter more.
Is allulose banned in Europe?
Not banned, but not authorized. In the EU, allulose counts as a novel food, a category that needs approval before it can be sold, and it has not been approved. In 2025, Europe's food safety agency said its safety could not be established because the applicant did not supply requested data. That is not a finding that allulose is harmful. Stevia, by contrast, is authorized in the EU as E 960.
Are natural sweeteners better than artificial ones?
Not clearly. Stevia, monk fruit and allulose are processed ingredients, and the FDA has no formal definition of natural. Some artificial sweeteners carry their own questions, but we found no trial comparing natural and artificial sweeteners on disease outcomes.
What is the best sugar substitute for people with diabetes?
There is no single answer. Allulose, purified stevia and monk fruit all have little effect on blood sugar on their own, so all three can be reasonable swaps for sugar, but none is a diabetes treatment. Allulose modestly lowered the blood sugar rise after a carbohydrate meal in several small trials, though not all of them. Check the ingredient list for dextrose or maltodextrin, and if you have diabetes or take diabetes medicines, talk with your doctor or dietitian before changing your diet.
Do sugar substitutes help you lose weight?
Not reliably. The WHO advises against relying on non-sugar sweeteners for weight control, based on low-certainty evidence, while two newer year-long trials found sweetened products helped people keep off slightly more weight. The evidence on eating the saved calories back is mixed: in one single-day study, people made up all of them at later meals, and in another they did not.
The bottom line
Allulose, stevia and monk fruit are reasonable ways to cut sugar, and none raises blood sugar much on its own. Allulose has the best evidence for modestly blunting the rise after a carbohydrate meal, but no long-term blood sugar or weight benefit has been shown, and larger doses upset the stomach. Purified stevia has the longest safety record and a set intake limit. Monk fruit has the least human research and unresolved safety questions in Europe.
"Natural" does not mean better: these are processed ingredients, and many products sold under their names are mostly dextrose, erythritol or other fillers, so read the ingredient list. For weight control, sweeteners help only if you do not eat the saved calories back.