Is allulose safe? What regulators and human trials actually found

Yes, allulose is safe on the evidence that exists. Regulators in the US, Australia and New Zealand, and China have cleared it, the longest human trial (48 weeks, up to 15 g a day) found no clinical problems, and it lowers the post-meal glucose rise. What's thin: long-term data, and studies in pregnancy.

So is allulose safe for you, in your chai, every day? That's really three questions: what regulators decided, what the human trials found, and who those trials left out. They have different answers, and every source below is named so you can check them.

What have regulators actually decided about allulose?

Allulose is cleared for food use in the US, Australia and New Zealand, China, Japan, South Korea, Singapore and Mexico. The three most detailed decisions are the FDA's GRAS letters from 2017 onwards, FSANZ's novel food approval in August 2024, and China's approval in July 2025. The EU, UK and Canada haven't authorised it.

The FDA's position gets misdescribed the most. Under the GRAS route, a company concludes an ingredient is safe and the FDA reviews its file. For Samyang's GRAS notice 693, the agency replied on 28 August 2017 that it had no questions, and later notices got the same reply.

That's a regulator accepting someone else's safety case. It's narrower than an approval, and it still counts. The FDA's October 2020 labelling guidance then said it would use enforcement discretion to keep allulose off the Total Sugars and Added Sugars lines, which is why allulose skips the sugars line on an American label.

FSANZ did the fullest job of the three. It ran its own assessment of Application A1247, found no public health and safety concerns, and approved allulose on 7 August 2024.

China's National Health Commission approved it on 2 July 2025, with conditions: a consumption level of up to 20 g a day, and labels saying it isn't for infants and young children or pregnant and breastfeeding women. That's the most cautious approval on the list, and we'd rather you hear it from us.

Does not being authorised in the EU mean allulose is unsafe?

The European Food Safety Authority concluded in June 2025 that allulose's safety "cannot be established" because the applicant stopped answering its requests for data on identity, production, use levels, genotoxicity and human safety. EFSA's opinion reports no finding of harm. An application with missing data can't be authorised, whatever the ingredient.

"Allulose is banned in Europe" is probably the most repeated line about it online, and it's wrong twice over. The EU has never banned allulose. It has never authorised it either, which is the default position for any novel food until someone completes an application.

The opinion is in the EFSA Journal, 23(6), e9468, and it's worth reading the reasons section yourself. EFSA asked for the data, followed up, and got nothing back.

The UK now runs its own process. The Food Standards Agency lists an allulose application from Savanna Ingredients, still at risk assessment when we checked on 2 October 2026. Canada treats allulose as a novel food needing a pre-market assessment and hasn't authorised it for ordinary foods.

So the pattern is simple. Every regulator that finished an assessment with full data in front of it cleared allulose. The holdouts are one application that stalled and two still in the queue.

What do long-term human studies on allulose show?

The longest human trial ran 48 weeks. Tanaka and colleagues, in Fundamental Toxicological Sciences in 2020, gave 90 adults with high LDL cholesterol 5 g, 15 g or no allulose daily and found no clinical problems and no change in cardiovascular risk factors. Most other human trials on allulose run 12 weeks or less.

Read that 48-week study properly and two details matter. It was randomised, double-blind and placebo-controlled, which is the design you want.

The shorter trials point the same way. Han et al. in Nutrients (2018) raised doses in healthy young adults to find where digestive comfort ran out, and landed on a maximum daily intake of 0.9 g per kg of body weight.

A 2022 systematic review of rare sugar trials by Ahmed et al. in Nutrition Reviews found short and longer-term benefits for glycaemic control, and said plainly that most studies were small and large trials were missing.

Is allulose safe for people with diabetes?

Allulose has human trial evidence in type 2 diabetes specifically. A 2024 meta-analysis by Ayesh, Suhail and Ayesh in Metabolism Open pooled 6 trials and 126 people with type 2 diabetes and found a smaller post-meal glucose rise and less time above range. Fasting glucose and time-in-range didn't move.

In healthy adults, the effect has a number on it. Tani et al. in PLOS ONE (2023) pooled 8 experiments and found 5 g to 10 g of allulose cut the post-meal glucose curve by roughly 13 to 14%, in a review that left people with diabetes out on purpose.

Two limits, stated as precision. Those diabetes trials were short and built to measure glucose, which is a different job from a long-term safety study.

And CSPI, the US consumer watchdog, rates allulose "cut back", partly because people with diabetes and pregnant or breastfeeding women were generally excluded from the safety studies. It's a cautious rating, and it's a fair description of who has been tested.

Your meter settles the rest. Check it two hours after a chai sweetened with allulose and again after one sweetened with sugar, same cup, same milk.

That's your own data, and it beats any average. If you want the wider comparison, here's how allulose stacks up against other sweeteners for a diabetic in India.

What are the benefits of allulose?

Allulose's main benefit is on blood glucose: in human trials it lowers the post-meal rise instead of simply not adding to it. It also carries 0.2 to 0.4 calories per gram against sugar's 4, and it browns, dissolves and stays smooth in syrups the way sugar does, which most zero-calorie sweeteners can't.

The glucose effect is the one that sets allulose apart. Tani et al. (2023) put it at roughly 13 to 14% off the post-meal glucose curve at 5 g to 10 g in healthy adults, and Ayesh et al. (2024) found a smaller post-meal rise in people with type 2 diabetes.

Compare that with erythritol, which had no effect on glucose or insulin in a 2016 trial by Wölnerhanssen et al. in the American Journal of Physiology. Neutral is good. A reduction is better.

In the kitchen, allulose behaves like a sugar because it is one. Ertugrul et al. (International Journal of Food Science & Technology, 2021) found it browned faster than fructose or glucose under the same heat, which is why a cake made with it takes colour and a chaashni stays glossy.

It also won't recrystallise out of a cooled syrup, so kheer and halwa don't turn grainy on day two. The same property is why barfi and kaju katli won't set firm with it, and we've written up what allulose does in an Indian kitchen, and where it fails.

And it sits outside the WHO's 2023 advice against non-sugar sweeteners. That guideline covers stevia, sucralose and aspartame, and states that low-calorie sugars and sugar alcohols aren't non-sugar sweeteners, so allulose is out of its scope (and so is erythritol).

Is allulose better than other sweeteners?

For cooking and blood glucose, allulose does more than stevia, sucralose or erythritol. It has meta-analyses of human trials showing a lower post-meal glucose rise, and it's the only one of the four that browns and bulks like sugar in a pan. It costs more, and it's about 70% as sweet as sugar.

Where allulose loses: price and strength. Stevia and sucralose cost far less per cup because so little goes in, and allulose's 70% sweetness means you use more of it. That's why our monk fruit and allulose blend pairs it with monk fruit, which carries the sweetness while allulose does the sugar's other jobs.

What's the catch with erythritol, sucralose and stevia?

Each has a documented catch. Erythritol has an EFSA daily limit most people exceed and a disputed 2023 clotting finding. Sucralose cut insulin sensitivity by about 23% in one small 2013 trial. Stevia carries an EFSA daily limit, a bitter edge for some people, and usually arrives bulked with maltodextrin or erythritol.

Erythritol is the bulking agent in many monk fruit sweeteners on an Indian shelf, so it's the fair comparison. EFSA set its daily intake at 0.5 g per kg in December 2023 and estimated that intake runs above it in every population group it assessed.

Nature Medicine also published a 2023 study by Witkowski et al. linking high blood erythritol with heart attack and stroke and showing it made platelets clot faster. It's an association plus a mechanism, and the FDA has noted the people studied were already at high heart risk.

Sucralose's issue is subtler. In Pepino et al.'s 2013 trial in Diabetes Care, 17 people with obesity who didn't usually use sweeteners drank sucralose before a glucose drink, and their insulin sensitivity fell by about 23%. One small trial, and a result worth knowing.

Stevia's catch is taste and quantity. Steviol glycosides switch on bitter taste receptors as well as sweet ones (Hellfritsch et al., Journal of Agricultural and Food Chemistry, 2012), and since stevia is used in milligrams, sachets are usually bulked out with maltodextrin or erythritol.

How much allulose is safe in a day?

Up to 0.9 g of allulose per kg of body weight a day, which is about 54 g for a 60 kg adult. Han et al. set that figure in Nutrients in 2018, and FSANZ used the same daily number in its 2024 approval. Most people sweetening chai and coffee use a fraction of it.

That number marks where digestive comfort ran out in the trial. It's a comfort ceiling, set by tolerance. FSANZ, which assessed the toxicology in 2024, didn't set an acceptable daily intake for allulose.

China's approval carries the more cautious figure: up to 20 g a day. If you sweeten everything with pure allulose, that's the number to know: allulose is about 70% as sweet as sugar, so 2 well-sweetened cups of chai made with it alone can reach 20 g.

Whereas Japan has no restrictions on its dosage, and has been used approximately for 20 years, without adverse effects being reported. Similarly there is no usage restrictions/limit on allulose in countries like Singapore and South Korea. This is very encouraging as these countries have some of the most stringent food regulators in the world.

The reason tolerance is high comes down to where allulose goes. It's a rare sugar, absorbed mostly in the small intestine and passed out intact in urine, because, as Daniel et al. put it in the British Journal of Nutrition in 2022, the human genome has no enzyme that breaks it down.

Sugar alcohols like sorbitol and lactitol travel further and get fermented in the colon, which is why their tolerance thresholds sit lower (Livesey, British Journal of Nutrition, 2001). Individual tolerance varies, so start with your usual amount and see how you feel. For what allulose is, from the molecule up, the pillar guide has the chemistry.

What don't we know about allulose yet?

Three things are still open on allulose. Most human trials run 12 weeks or less, at research doses. Mouse work points to allulose shifting gut bacteria, which hasn't been properly tested in people. And pregnant and breastfeeding women, infants and people with diabetes have mostly been left out of the safety studies.

The sharpest version of the gut question comes from Daniel, Hauner, Hornef and Clavel, four researchers at German universities who declared no funding and no conflicts. They note that Klebsiella pneumoniae, a bacterium that causes infections, can use allulose as fuel.

In lab dishes and a mouse infection model, bacteria able to take up allulose grew slightly better. The authors ask whether high allulose intake could give harmful bacteria an advantage, and call for human studies.

That's a hypothesis from lab and mouse work. No study has shown it in a human. It's still the right question to ask, and it's exactly what a 48-week cholesterol trial wouldn't be designed to catch.

The same review lists other gaps: trials at realistic everyday doses, whether rodent findings on gut hormones hold in people, and what regular high intakes do over years. We'd rather you knew the list than found it later.

How should you read an allulose safety claim?

Check four things before trusting an allulose headline: whether the study was in humans or animals, how long it ran, what dose it used against what you actually eat, and who ran or paid for it. Most alarming and most glowing allulose claims fail at least one. Apply the same test to this page.

Animal or human first. A lot of the striking allulose research, on liver fat and body weight, is in rodents. Interesting, and it tells you very little about a teaspoon in your coffee.

Then duration and dose. A 12-week trial at 10 g a day answers a different question from 10 years of daily use, and a rat study at a dose scaled far above a human diet answers neither.

Then the people behind it. The 48-week trial came from a producer's R&D team, and the gut-bacteria concern came from independent academics.

CSPI's rating comes from a watchdog that rates on caution.

And this page comes from a brand that sells a monk fruit and allulose sweetener. Weigh it accordingly, and click the sources.

Is allulose safe to use every day?

On the current evidence, yes, at normal sweetening amounts. The longest trial dosed 15 g every day for 48 weeks without clinical problems, and daily use in chai and coffee sits well inside the 0.9 g/kg figure. If you're pregnant, breastfeeding, sweetening for a toddler or managing kidney disease, ask your doctor first.

That's the honest shape of it. Allulose has a 48-week human trial behind it, a measured benefit on post-meal glucose, and a short list of open questions worth watching.

It also browns, dissolves and caramelises like sugar, which is why it does the heavy lifting in a blend. Our blend uses it with no erythritol and no maltodextrin.

A small number of people find the blend a touch less sweet than they expected, so start with your usual spoon and adjust. And keep reading the nutrition panel on every pack you buy, ours included.

Frequently asked questions

Is allulose safe for your kidneys?

Allulose leaves the body mostly intact in urine, because the human body has no enzyme to break it down. No human trial or regulator review we found reports kidney harm in people with healthy kidneys. We couldn't find a trial in people with chronic kidney disease, so if that's you, ask your nephrologist before using it daily.

Is allulose safe during pregnancy or for young children?

We found no trials of allulose in pregnancy, and pregnant and breastfeeding women were generally excluded from the safety studies. China's 2025 approval says allulose isn't for infants and young children or pregnant and breastfeeding women. Until better data exists, skip it in pregnancy and for toddlers, and check with your doctor.

Does allulose cause cancer?

No human study or regulator assessment has found a cancer signal for allulose. FSANZ reviewed the safety evidence in 2024 and found no public health and safety concerns. EFSA's 2025 opinion asked for genotoxicity data the applicant never supplied, which leaves a gap in that one file and reports no finding of harm.

Is allulose banned in Europe?

Allulose isn't banned in Europe; it isn't authorised there yet, which is the default for any novel food. EFSA concluded in June 2025 that its safety couldn't be established because the applicant stopped supplying the data EFSA requested. The opinion reports no finding of harm, and a fresh application could change the answer.

Is allulose healthier than sugar?

On calories and blood glucose, yes. Allulose carries 0.2 to 0.4 calories per gram against sugar's 4, and meta-analyses in PLOS ONE (2023) and Metabolism Open (2024) found it lowers the post-meal glucose rise. Allulose also has a much shorter research record than sugar, so there are fewer years of human data behind it either way.

Is allulose better than stevia?

For cooking and post-meal glucose, allulose has the edge: it browns and bulks like sugar, and meta-analyses show it lowers the post-meal glucose rise. Stevia is far cheaper per cup because it's used in milligrams, and it carries an EFSA daily limit of 4 mg per kg. Stevia is also covered by the WHO's 2023 sweetener advice; allulose isn't.

Is allulose natural or artificial?

Allulose is a rare sugar that occurs in nature in tiny amounts; Daniel et al. estimate a typical Japanese diet supplies about 0.2 g a day from ordinary food. Commercial allulose is made from fructose using an enzyme, or by fermentation, the two routes described in China's 2025 approval.

Is allulose zero calories?

Allulose carries between 0.2 and 0.4 calories per gram, against sugar's 4. The FDA tells US labels to use 0.4, the top of that range. A spoon-sized serving rounds down to zero on a nutrition panel, which is how every zero-calorie sweetener pack reaches its zero, ours included.

Start with the 150 g pouch of the monk fruit and allulose blend, put it in your usual chai, and check your meter two hours later. Your own numbers are the study that counts.