World Diabetes Day falls on Saturday, 14 November, and the International Diabetes Federation's theme this year is early action on diabetes. In India that covers two groups: 101 million adults living with diabetes and 136 million with prediabetes, according to the ICMR-INDIAB study.
If you're in either one, you've heard the lectures. So, one narrow question: what does a spoonful of allulose do to your blood sugar, measured in actual people?
The short answer. On its own, a spoonful of allulose doesn't raise blood sugar, because your body can't break it down for fuel. Taken with a meal, 5 to 10 g, roughly one to two and a half teaspoons, trimmed the rise in blood glucose after eating in human trials, including in people with type 2 diabetes. The trim is real, and it's modest.
Does allulose raise blood sugar?
No. Allulose is a rare sugar, a near-twin of fructose with one part of the molecule flipped, and the human body has no enzyme that turns it into energy. Roughly 50 to 70% of what you eat is absorbed, circulates, and leaves in urine unchanged, and trials that gave allulose on its own found blood glucose stayed where it was.
A 2022 review in the British Journal of Nutrition put the reason plainly: the human genome "does not encode for enzymes that are able to metabolise allulose." So it travels through you the way a tourist passes through a city, seen everywhere and spending nothing.
That's also why the calorie count is so low. Published estimates put allulose 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 the range, and you can read how regulators count allulose on a label if you want the detail.
What happens when you eat allulose with a meal?
It trims the rise. A 2023 meta-analysis of 8 small experiments in healthy adults found that 5 g or 10 g of allulose, taken with a carbohydrate drink or meal, lowered the post-meal glucose curve against a control without it. Translated into one trial's numbers, that's a curve roughly 13 to 14% smaller (Tani et al., PLOS ONE, 2023).
The clearest early trial is from 2008. Twenty Japanese adults drank 75 g of maltodextrin, a starch that digests almost straight into glucose, with 0, 2.5, 5 or 7.5 g of allulose. From 5 g upward their blood glucose rose less, and the bigger the dose, the smaller the rise (Iida et al., 2008).
The leading explanation is traffic. Allulose uses the same doors in the gut wall as glucose and fructose, so when it's present, glucose gets in more slowly. Rat studies also suggest it slows the enzymes that break starch down, though neither mechanism has been fully pinned down in people yet.
One detail matters for your kitchen. Six of the 8 experiments gave allulose dissolved in a drink, and one of the two that used it in solid form came out the wrong way, which the authors suggested might come down to how well it dispersed. Most of the evidence is for allulose taken dissolved in a drink.
What about prediabetes?
The closest trial to an Indian breakfast table is Japanese. Twenty-six adults, 15 of them with borderline diabetes, ate a standard meal with tea that held either 5 g of allulose or none. Blood glucose was lower at 30 and 60 minutes with allulose, and the effect showed up mainly in the borderline group (Hayashi et al., 2010).
This year's World Diabetes Day puts its focus on early diagnosis and lowering risk, which is exactly where this group sits. India's 136 million adults with prediabetes outnumber the 101 million already living with diabetes, and early action for them is mostly daily stuff: what's in the cup, what's on the plate.
The same paper ran a second, longer test. Seventeen healthy adults took either 5 g of allulose or 5 g of glucose with each of three meals a day for 12 weeks, and the researchers found no abnormal effects and no clinical problems from allulose.
Worth keeping in proportion: it's one study of 26 people, and the drink was tea in a Japanese lab, a long way from a cup of masala chai. It points in a useful direction. It doesn't settle anything on its own.
What happens in people with type 2 diabetes?
Allulose has been tested directly in type 2 diabetes, and the post-meal trim holds. A 2024 meta-analysis pooled 6 trials and 126 people: allulose lowered the glucose curve after meals and cut the time spent above the target range. Fasting glucose and overall time in range didn't change significantly (Ayesh et al., Metabolism Open, 2024).
The most detailed of those studies came out of Kagawa University Hospital in 2023. Twenty inpatients with type 2 diabetes (24 enrolled) ate the hospital's diabetic diet for two days as usual and two days with 8.5 g of allulose added to each meal, in random order, wearing a FreeStyle Libre glucose sensor throughout. Nobody's medication was changed (Fukunaga et al., Nutrients, 2023).
With allulose, the average post-meal peak was 173 mg/dL against 191 mg/dL, about 18 points lower. Time spent above 180 mg/dL fell. C-peptide, a marker of the body's own insulin output, was lower too, and the study recorded no safety problems over the four test days.
An earlier trial from a Toronto group found a smaller effect in the same direction. In people with type 2 diabetes drinking 75 g of glucose, 10 g of allulose cut the area under the glucose curve by 8% compared with none (Noronha et al., 2018).
How big is the effect, honestly?
Modest. Where trials put a percentage on it, the trim comes out around 8 to 14% of the post-meal glucose curve, or about 18 mg/dL off the peak in the hospital study, and the 2023 meta-analysis's own authors call the effect small. It shaves the spike from a meal, and the roti and rice on the plate still count.
The same goes for sweets. A gulab jamun soaked in allulose syrup has dropped the sugar, and it still has khoya and maida in it.
Four more things are worth knowing before you quote any of this to your family WhatsApp group.
Not every trial found it. A sister trial from the same Toronto group, in healthy adults, found no significant effect from 5 or 10 g of allulose against a 75 g glucose drink (Braunstein et al., Nutrients, 2018).
The healthy-adult evidence is thin. The 2023 meta-analysis's own sensitivity check found that removing one small study, the 2008 trial above, changed the result. Its authors say plainly that a single new trial of similar size could move the conclusion.
Much of the research has commercial ties. The 2023 meta-analysis's lead author works for Matsutani Chemical, a Japanese allulose producer, and two co-authors hold patents. The hospital study's senior author holds intellectual property rights to allulose, and the 2018 Toronto trial in healthy adults was funded through a Tate & Lyle research fund, with Tate & Lyle, another allulose maker, supplying the drinks.
Each paper declares its ties, and the results can still be right. Larger independent trials are what to watch for, and we'll update this page when they land.
The trials measure the hours after a meal. The 2024 meta-analysis found no significant change in fasting glucose. One 8-week study with no control group reported a small HbA1c drop, but it used a diabetes nutrition drink with plenty else in it, so it can't isolate allulose.
Where does monk fruit fit?
Monk fruit is the sweetness, and its blood sugar story is about what it leaves out. In a 2017 Singapore trial, 30 men tried four drinks, and the one sweetened with monk fruit caused no glucose spike in the hour after it, where the sugar drink caused a sharp one (Tey et al., International Journal of Obesity, 2017).
The same trial had a twist. After lunch, glucose ran higher for the monk fruit, stevia and aspartame drinks, people ate more at that lunch, and over three hours the totals came out level with sugar.
Monk fruit leaves the glucose curve where it was.
That's the reason the two sit in one pouch. Monk fruit carries most of the sweetness in a tiny amount, and allulose carries the bulk and the post-meal evidence above. How much allulose is in a spoon of any blend, ours included, is printed on its nutrition panel, and the trials used pure allulose, so read the panel before you compare.
How can you check this on your own meter?
Run two mornings that differ in one thing: day one your usual cup, day two the same cup sweetened with allulose, with the same breakfast, the same time and the same walk (or lack of one) on both days. Check your glucose before eating and at 30, 60 and 120 minutes, or read the curve off a CGM.
One pair of readings is a hint. Repeat the pair two or three times before you trust the pattern, because fingerstick meters wobble and so do mornings.
Keep your medication exactly as prescribed while you test. If you take insulin or a sulfonylurea, both of which can push blood sugar too low, mention the swap to your doctor first. Nothing here replaces what your doctor has you on.
If chai is the cup in question, there's a full guide to sweetening chai without sugar, spoon by spoon.
Where does a spoonful make the most difference?
In the cup you drink with a meal. That's the closest match to how most trials gave allulose, dissolved in a drink with carbohydrate, and for most of us in India it's the morning chai with breakfast. The evidence is thinnest for sweets with allulose cooked into them, so treat those as sugar you've removed and enjoy them for that.
Our Super Sweetener is monk fruit and allulose with no erythritol, made to dissolve into chai and coffee and cook into halwa and kheer. Start with a teaspoon where you'd use one of sugar and adjust up.
It costs more than sugar, and the reason is in the pouch: allulose costs far more to make than cane sugar. A 5 g sachet is the easiest way to try it in a cup, and like any blend it holds less than 5 g of allulose, because the monk fruit is in there too.
For the full picture on the ingredient itself, there's the full allulose explainer, and for how it stacks up against stevia, sucralose and jaggery, the sweetener guide for diabetics in India, ingredient by ingredient.
On World Diabetes Day, early action can be small and daily. The same cup of chai you were going to drink anyway, minus the sugar.
Frequently asked questions
Can people with diabetes use allulose?
Allulose has been tested directly in people with type 2 diabetes, including a 2024 analysis of 6 trials and 126 participants and a hospital study where nobody's medication changed. Those trials found a trim on post-meal glucose. Check your own response on a meter, and mention the swap to your doctor if you take insulin.
Does allulose spike insulin?
The trials so far lean the other way. In a 2023 hospital study of 20 people with type 2 diabetes, C-peptide, a marker of the body's own insulin output, was lower after meals containing 8.5 g of allulose, and a 2024 pooled analysis saw the same direction without reaching statistical significance.
How much allulose did the trials use?
Between 5 and 10 g with a carbohydrate drink or meal, roughly one to two and a half teaspoons of pure allulose. The type 2 hospital study used 8.5 g per meal, three meals a day. A spoon of a blended sweetener holds less allulose than a spoon of the pure ingredient, and its nutrition panel gives the figure.
Is allulose zero GI?
A GI of 0 on a sweetener pouch, ours included, is arithmetic. ISO 26642 says foods with under 10 g of glycaemic carbohydrate per serving shouldn't be GI-tested at all, which is why the zero on every sweetener pouch was never measured. The post-meal curve is the more useful number, and it's what these trials measured.
Does monk fruit raise blood sugar?
In a 2017 trial, a monk fruit drink caused no glucose spike in the hour after it, where a sugar drink did. Glucose ran a little higher after the next meal, and over three hours the totals evened out with sugar. The post-meal trim in the trials above comes from allulose.
Does allulose lower fasting blood sugar or HbA1c?
The evidence so far covers the hours after a meal. The 2024 meta-analysis in type 2 diabetes found no significant change in fasting glucose, and one 8-week study with no control group that reported a small HbA1c drop used a diabetes nutrition drink with other ingredients, so it can't isolate allulose.