What Science Says About Monk Fruit: The Evidence, Labelled
Every study on this page carries a label. Clinical means the work was done in people, preclinical means cells or animals, and most of the exciting monk fruit research is preclinical.
Monk fruit tastes sweet because of mogrosides, a family of cucurbitane triterpene glycosides. Purified extracts standardised to a quarter or half mogroside V run roughly a hundred and fifty to two hundred and fifty times as sweet as sugar. Your small intestine absorbs little of them, gut bacteria strip the sugar groups off, and your body does not use them for energy. The human evidence is thin. A 2025 systematic review found five randomised trials, and the metabolic ones enrolled thirty to fifty-five people for a single sitting. Most of the promising work sits in mice and in test tubes. In June 2026 JECFA set a temporary acceptable daily intake and flagged child exposure at proposed maximum use levels. No long-term human trial of monk fruit exists.
Most of what you read about monk fruit online rests on studies in mice. That is not a scandal. Mouse work is how research starts, and mogrosides have been studied that way for two decades. The problem is that the mouse disappears somewhere between the journal and the blog post, and the reader ends up with a health claim built on an animal that weighs thirty grams.
So this page labels every study before it describes it. Clinical means the work was done in people. Preclinical means cells, tissue or animals. Where the evidence is thin, we say so, and by the end you will have a list of the things nobody has measured.
How this page labels the evidence
Nutrition research runs on a ladder, and the rungs are far apart. A compound that kills free radicals in a test tube may never reach the tissue where it would matter. A mouse fed a dose no human could swallow may improve in ways a person never would. The ladder matters more in sweetener research than almost anywhere else, because the field has a long history of animal findings that shrank to nothing in people.
Clinical, meaning in people
A clinical study enrols human beings. The useful ones randomise, blind participants where possible, and report how many took part. Sample size decides how much a result is worth. Thirty people in a single sitting can show you a glucose curve. They cannot tell you what happens after five years.
Preclinical, meaning cells or animals
A preclinical study uses cell culture, tissue or laboratory animals. It generates a hypothesis and a mechanism. It does not establish a benefit in humans, and no honest writer should present it as though it did. When you see a phrase like reduced oxidative stress attached to monk fruit, check the species. In most cases it is a rat.
The chemistry, and the numbers under it
Monk fruit is the fruit of Siraitia grosvenorii, a gourd vine from southern China. Its sweetness comes from mogrosides, which chemists class as cucurbitane triterpene glycosides. Each one has a rigid steroid-like core with sugar units hung off it, and the number and position of those sugars decide how sweet the molecule tastes.
Mogroside V carries five glucose units and does most of the work. On its own it measures somewhere between two hundred and fifty and four hundred times as sweet as sucrose, a figure Food Standards Australia New Zealand used when it assessed the extract as a food additive. Siamenoside I is sweeter still but present in much smaller amounts. Mogroside IV, mogroside III and the aglycone mogrol fill out the family.
Commercial extracts are blends, so their sweetness tracks their mogroside V content. The tables in the American GRAS notices show the pattern clearly: an extract at seven per cent mogroside V lands near forty-five times sugar, one at ninety per cent or above passes four hundred, and the grades sold as tabletop sweeteners sit between them at roughly a hundred and fifty to two hundred and fifty times.
Why a decoction is a different product
A purified extract and a decoction share a plant and very little else. A decoction is fresh fruit infused in water and boiled down, which concentrates the mogrosides along with everything else water pulls out of the fruit. Ours measures ten to fifteen times as sweet as sugar, so a serving carries a small fraction of the mogroside V in an equivalent serving of a fifty per cent extract. Six drops replace a teaspoon of sugar. The process is set out in how monk fruit decoction is made, and the distinction matters again when we reach the regulators.
What happens after you swallow them
Mogrosides are large, heavily sugared molecules, and your small intestine absorbs very little of them intact. Published pharmacokinetic estimates put oral bioavailability at roughly three and a half to ten per cent, falling as the sugar chain gets longer. The rest travels on to the colon, where bacteria remove the glucose units one at a time. Mogroside V becomes mogroside III, then mogrol, the sugar-free core. In vitro fermentation with human faecal bacteria converts most of it within twenty-four hours.
Two things follow. Your body does not convert mogrosides into energy, which is why the sweetness arrives without the calories of sugar. And the compound your gut actually produces is not the compound you swallowed, which is one of the reasons regulators want more toxicology on the metabolites. A 2026 review of mogroside metabolism sets out the pathway in detail and concludes that human pharmacokinetic data remain incompletely characterised.
Here is the whole evidence base on one screen, labelled.
| The study | Design and size | What it showed |
|---|---|---|
| Tey and colleagues, 2017 | Clinical, randomised crossover, 30 adults, single session | Glucose and insulin area under the curve lower than sucrose |
| Tey and colleagues, second 2017 report | Clinical, randomised crossover, 30 adults, single session | More eaten at the next meal, total daily intake unchanged |
| Epstein and colleagues, 2024 | Clinical, double blind crossover, 50 adults, single session | Fasting glucose about 6 per cent lower, sugar less reinforcing |
| Glucose tolerance trial in type 2 diabetes | Clinical, randomised double blind, 55 adults, 26 with type 2 diabetes | Flat glucose curve where sucrose peaked at 247 mg/dL |
| Tan and colleagues, 2019 | Clinical, placebo controlled, 203 patients, 48 hours | Less throat pain after intubation. Nothing about metabolism |
| Wu and colleagues, 2024 | Clinical, double blind, 103 patients, 15 days | Fewer pharyngitis symptoms. Longest trial on the list |
| Mogroside-rich extract, 2019 | Preclinical, diabetic mice, 5 weeks, up to 300 mg/kg | Lower fasting glucose, AMPK switched on in the liver |
| Faecal fermentation assays | Preclinical, human gut bacteria in vitro | Mogroside V broken down to mogrol within 24 hours |
Six clinical rows, none longer than fifteen days, and the two largest measured sore throats rather than blood sugar.
Nothing here is alarming. Almost none of it is long term either.
What the human trials found
A systematic review published in 2025 searched the literature for randomised controlled trials of monk fruit extract and found five. Two of those measured throat symptoms. That leaves a handful of small metabolic studies, and they agree with each other on the one thing they were built to test.
Swap sucrose for monk fruit in a drink and the glucose and insulin responses fall. That result is consistent, it is exactly what the chemistry predicts, and it is measured over hours. Nobody has run the same comparison for a year.
Thirty adults, one sitting
Tey and colleagues gave healthy adults beverages sweetened with monk fruit, stevia, aspartame or sucrose in randomised order. Glucose area under the curve came out about 18 per cent lower than sucrose and insulin about 22 per cent lower. Thirty people, one session each.
Fifty adults, one sitting
Epstein and colleagues ran a double blind crossover in 2024 with fifty adults, some insulin sensitive and some insulin resistant. Fasting glucose was around 6 per cent lower and sugar became less reinforcing. Again a single session.
Fifty-five, half with diabetes
A randomised double blind glucose tolerance trial tested 26 adults with type 2 diabetes and 29 without. Sucrose pushed the diabetes group to 247 mg/dL at ninety minutes. Monk fruit left the curve close to flat in both groups.
What we do not claim
None of those trials tested our decoction, and we have not run one. We say the liquid is low calorie, about 1 kcal per serving, with no glycemic index per portion, and we stop there. Anything past that would be borrowed evidence.
One finding deserves a mention because it cuts against the marketing. In the second Tey report, participants ate more at the following meal after the low calorie drink, and their intake across the day came out much the same. Replacing sugar in a glass does not automatically remove the energy from the day.
The animal and laboratory work
This is where the volume is, and where the caution belongs.
From about 2018 onward, teams in China published a run of studies on mogrosides in rodent models of diabetes. The best known gave mogroside-rich extract to mice made diabetic with a high fat diet and streptozotocin, at up to three hundred milligrams per kilogram of body weight for five weeks. Fasting glucose fell, insulin resistance improved, fatty change in the liver reversed, and the authors traced the effect to activation of AMP-activated protein kinase. Later work reported similar things for mogroside III in models of gestational diabetes and for mogrol in adipocytes.
Read those doses against a human. Three hundred milligrams per kilogram in a mouse is not three hundred milligrams in a person, but even after the usual scaling it lands far above anything a spoonful of sweetener delivers. And the field has form. Sweetener research has repeatedly produced rodent results that failed to reproduce in people, which is why the 2026 review cited above states plainly that most of the antidiabetic evidence for mogrosides comes from in vitro studies or animal models, and that well powered human trials remain limited or absent.
Antioxidant and anti-inflammatory assays
Mogrosides scavenge free radicals in laboratory assays. In rodents they raise superoxide dismutase and glutathione peroxidase activity and lower malondialdehyde, the usual markers of oxidative stress. Anti-inflammatory work points at the NLRP3 inflammasome and the TLR4 and NF-kB pathways, all of it in cell culture or animals.
Not one of these results has been shown to matter in a human being. Antioxidant capacity measured in a tube is a chemical property, not a health outcome, and the EU does not authorise antioxidant health claims on that basis. We go through what can and cannot fairly be said in monk fruit risks and benefits.
The 2026 survey of the fruit's other compounds
In January 2026, Liu, Wu and colleagues published a study in the Journal of the Science of Food and Agriculture comparing the peel and pulp of four monk fruit varieties. Using metabolomics with network pharmacology and molecular docking, they mapped terpenoids, flavonoids and amino acids across the tissues and picked out candidate compounds behind the fruit's antioxidant activity. It is compositional and computational analysis, not a health study, and the docking step predicts binding rather than measuring an effect. We cover it properly in monk fruit's hidden compounds.
What the regulators concluded
Regulatory assessments are the most useful documents in this whole field, because the panels writing them have no product to sell and no headline to win. Four of them matter here, and they do not all say the same thing.
EFSA, 2019: unable to conclude
Europe's food safety authority assessed monk fruit extract as a proposed new food additive and declined to clear it. The panel found the toxicity database insufficient. Three things were missing: genotoxicity testing on the metabolites that gut bacteria produce, an explanation for testicular effects seen in a ninety day study, and any chronic or carcinogenicity data at all, which the panel considered warranted because mogroside V reaches the bloodstream. No acceptable daily intake was set. Purified monk fruit extract is still not an authorised sweetener in the EU as a result.
JECFA, June 2026: a temporary limit, and a flag
The Joint FAO and WHO Expert Committee on Food Additives reviewed monk fruit extract at its 102nd meeting and set a temporary acceptable daily intake of 0 to 10 milligrams per kilogram of body weight per day, expressed as mogroside V. The figure came from a fifty-two week rat study with a no observed adverse effect level of 2144 mg/kg, divided by a two hundred fold uncertainty factor, which the committee applied because data are still missing. The committee then estimated the highest dietary exposure for children at 38 mg/kg per day under the maximum use levels the industry proposed, roughly four times the temporary limit, and called that a potential safety concern. It asked for reproductive toxicity work on the second generation and revised use levels by the end of 2029.
Read that carefully, because it is easy to misread in both directions. The concern attaches to purified extract used at the highest levels proposed across the global food supply, in the group with the lowest body weight. The committee was capping how much concentrate the food industry may spread across the whole shelf. It did not warn anyone off a spoonful in their tea. A decoction at ten to fifteen times sugar carries far less mogroside per serving than a purified extract at two hundred and fifty times. Even so, the flag is real and a reader deserves to see it here rather than nowhere.
The United States, Australia and New Zealand
American regulators took a different route. Monk fruit extract entered the food supply through GRAS notices, where a company submits a safety dossier and the FDA replies. Notice 522, covering Siraitia grosvenorii fruit extract as a tabletop and general purpose sweetener, closed in December 2014 with no questions from the agency. Food Standards Australia New Zealand assessed the same ingredient as a food additive, found the genotoxicity assays negative and the ninety day no effect levels high, and concluded that an acceptable daily intake of not specified was appropriate.
The decoction, which is a separate question
Novel food law asks something narrower than safety. It asks whether a food was eaten to a significant degree in the EU before 15 May 1997. In October 2024 the Food Safety Authority of Ireland confirmed that monk fruit decoctions were not novel, on evidence of consumption before that date assembled over seven years. The purified extract has no such history and remains novel.
Hold the two findings side by side. EFSA could not conclude on a concentrate made by modern industrial fractionation. The FSAI accepted a water infusion with a documented past. Both are correct, they answer different questions, and anyone who quotes one as though it settles the other is misleading you.
WHO, 2023: the guideline people misquote
The World Health Organization published a guideline on non-sugar sweeteners in May 2023. The wording is this: WHO suggests that non-sugar sweeteners not be used as a means of achieving weight control or reducing the risk of noncommunicable diseases. It is labelled a conditional recommendation. Certainty of evidence for body weight in short term trials was rated low, and for long term outcomes such as type 2 diabetes, cardiovascular disease and mortality it ran from very low to low. WHO chose conditional partly because reverse causation may explain some of the observational associations, meaning people already gaining weight may be the ones who reach for sweeteners.
Three details usually get dropped. The named sweeteners are acesulfame K, aspartame, advantame, cyclamates, neotame, saccharin, sucralose, stevia and stevia derivatives. Monk fruit is not among them. The recommendation applies to everyone except people with pre-existing diabetes. And it is a statement about weight control, not a safety warning.
We would rather show you the evidence than a slogan.Europe's first monk fruit brand. One ingredient in the liquid, about 1 kcal per serving, with no glycemic index per portion.
What nobody has studied yet
A page that lists its own gaps is worth more than one that hides them, so here they are.
No long-term human trial. The longest randomised study of monk fruit in people ran fifteen days, and it measured throat symptoms. Every metabolic trial was a single sitting. Nothing tells you what daily use does over a year.
No cardiovascular outcome data. No trial has followed people taking monk fruit and counted heart attacks, strokes or deaths. The 2025 systematic review named cardiovascular assessment as an outstanding need.
Nothing in pregnancy. No human study has tested monk fruit in pregnant women. Animal reproductive screening exists, and JECFA has asked for more of it by 2029. That is not the same as evidence in people.
Small samples throughout. Thirty, fifty and fifty-five participants. Trials this size can detect a glucose curve and little else. None was powered to find a rare effect.
Unfinished toxicology on the concentrate. EFSA's questions about metabolite genotoxicity and the testicular finding have not been answered publicly. JECFA applied a two hundred fold uncertainty factor for the same reason.
Nothing on children specifically. Every trial enrolled adults, which is awkward given that children are the group JECFA's exposure estimate flagged.
No trial of a decoction. Including ours. The clinical work used purified extract, so we cannot claim its results and we do not.
How to read a monk fruit study yourself
Five checks that will settle most arguments in under a minute.
-
Find the species in the first paragraph
Look for mice, rats, cells or participants. Abstracts state it plainly, and press releases almost never do. If the word human does not appear anywhere, you are reading a hypothesis.
-
Find the number of people
A trial with thirty participants and a trial with three thousand are different kinds of object. The metabolic monk fruit studies sit at the low end, so treat any single result from them as a starting point.
-
Ask how long it ran
Single session, two weeks, two years. Almost all monk fruit trials are single session, which means they measured a response and not a consequence. Duration is the fastest way to spot an overreaching headline.
-
Scale the dose to a human
Rodent studies use milligrams per kilogram of body weight, often hundreds of them. Multiply that by seventy and compare the result with the amount in your cup. Usually the study dose is an order of magnitude higher.
-
Check which product was tested
Ninety per cent mogroside V, a fifty per cent tabletop grade and a water decoction are three different ingredients. A finding about one does not transfer to the others, and this is where most confusion about monk fruit starts.
One honest tip: if a claim about monk fruit arrives without a species and a sample size attached, do not argue with it. Go and find those two numbers first, then decide whether the claim survives them.
Frequently asked questions
Is monk fruit backed by science?
Partly. The chemistry is well described and the short-term glucose effect is consistent across small trials. The health claims that circulate online mostly rest on cell and rodent work, and a 2025 systematic review found only five randomised human trials in total.
How many human trials of monk fruit exist?
A 2025 PRISMA systematic review identified five randomised controlled trials. Two of them measured throat symptoms rather than metabolism. The metabolic trials enrolled 30, 50 and 55 participants, each for a single session.
Does monk fruit raise blood sugar?
In the trials run so far, no. Swapping sucrose for monk fruit lowered glucose area under the curve by about 18 per cent in a crossover study of 30 adults, and in a trial of 55 adults the curve stayed close to flat where sucrose peaked at 247 mg/dL. These were single sittings, not long-term studies.
Does monk fruit affect insulin?
Insulin response in the published crossover trials came out roughly 22 per cent lower than with sucrose. Mogrosides are barely absorbed, so there is little to trigger a release. The evidence is short-term and covers no more than about fifty people per study.
What are mogrosides?
They are the sweet compounds in monk fruit, classed as cucurbitane triterpene glycosides. Mogroside V is the main one and measures roughly 250 to 400 times as sweet as sugar on its own. Siamenoside I, mogroside IV and mogroside III make up the rest of the family.
Does the body absorb mogrosides?
Very little of them. Published estimates put oral bioavailability at about 3.5 to 10 per cent. The remainder reaches the colon, where bacteria strip off the glucose units to leave mogrol, and the body does not convert any of it into energy.
Is monk fruit safe long term?
No long-term human trial has been run, so nobody can answer that from human data. The toxicology published so far shows no safety signals, and a 52 week rat study found no adverse effects at 2144 mg/kg per day. Long-term human evidence remains a genuine gap.
Why is monk fruit extract not approved in the EU?
EFSA assessed purified monk fruit extract as a food additive in 2019 and found the toxicity database insufficient to conclude. Missing pieces included genotoxicity data on gut metabolites, an explanation for testicular effects in a 90 day study, and chronic toxicity data. Water decoctions are a separate case and were confirmed as not novel in October 2024.
Is monk fruit approved by the FDA?
It reached the American market through GRAS notices rather than an additive approval. Notice 522, for Siraitia grosvenorii fruit extract as a tabletop and general purpose sweetener, closed in December 2014 with the FDA raising no questions. Several other monk fruit notices have closed the same way.
What did JECFA decide about monk fruit in 2026?
At its 102nd meeting in June 2026 the committee set a temporary acceptable daily intake of 0 to 10 mg per kilogram of body weight, expressed as mogroside V. It estimated the highest child exposure at 38 mg/kg under the maximum use levels industry proposed, about four times the limit, and asked for further reproductive toxicity data by 2029.
Does the WHO recommend against monk fruit?
Monk fruit is not named in the 2023 WHO guideline. That guideline suggests non-sugar sweeteners not be used for weight control or to reduce disease risk, as a conditional recommendation based on low certainty evidence, and it excludes people with pre-existing diabetes. It addresses weight control rather than safety.
Is monk fruit safe in pregnancy?
No human study has tested it in pregnancy, so there is no evidence either way. Animal reproductive screening found no adverse effects at up to 4000 mg/kg per day, and JECFA has asked for further work by 2029. Ask your midwife or doctor rather than a blog.
Final thoughts
The fair summary of monk fruit research is quieter than either side of the argument would like. The chemistry is solid, the short-term glucose data are consistent, the toxicology has thrown up no safety signals so far, and the long-term human evidence does not exist. Everything beyond that comes from mice, and mice have misled this field before.
We sell the stuff, so treat our reading with the scepticism it deserves and go to the sources below. They are all public. If you come back with a question we have handled badly on this page, write to us and we will fix it, because a page that gets quoted has to be a page that holds up.
One ingredient, and no claim we cannot show you
Europe's first registered monk fruit brand. Fresh fruit infused in water and boiled down to a decoction, with nothing else added. Low calorie, about 1 kcal per serving, with no glycemic index per portion.
Buy on Amazon1st
monk fruit
References
- EFSA ANS Panel, safety of Monk fruit extract as a food additive, EFSA Journal 2019: toxicity database insufficient to conclude
- JECFA, 102nd meeting summary report, June 2026: temporary ADI of 0 to 10 mg/kg bw as mogroside V and child exposure estimate
- WHO guideline on the use of non-sugar sweeteners, 2023: conditional recommendation and certainty of evidence ratings
- Monk fruit extract and sustainable health, a PRISMA-guided systematic review of randomised controlled trials, Nutrients 2025
- Food Safety Authority of Ireland, monk fruit decoctions confirmed as not novel, October 2024
Related reading
Happy Monkfruit is a food, not a medicine. Information here is general and not medical advice. Monk fruit is low calorie, about 1 kcal per serving, with no glycemic index per portion. If you manage a health condition, speak to your doctor.
