Two numbers circulate everywhere: 25 grams of sugar per day, or 50. Both are official. Both come from the same organization, published on the same day, in the same document. And neither was calculated with your waistline in mind.
That last part almost never makes it into articles about sugar. Before you start counting spoonfuls, it is worth knowing where the counter came from.
Two thresholds, two levels of certainty
In 2015, the World Health Organization published its guideline on sugars intake. It contains two sentences, and everything hinges on the difference between them.
The first is a strong recommendation: reduce free sugars to less than 10% of total energy intake. For an adult eating around 2,000 calories a day, that works out to roughly 50 grams.
The second is a conditional recommendation: dropping below 5% would bring additional benefits. That is roughly 25 grams, or six teaspoons.
In WHO vocabulary, those two adjectives are not polite synonyms. "Strong" means the desirable effects clearly outweigh the undesirable ones, and the recommendation can be adopted as public policy as it stands. "Conditional" means the expert group judged the balance less clear-cut, and that debate is needed before it becomes a rule.
Put plainly: 50 g is a solid safety floor. 25 g is a desirable target, openly flagged as less well established. Presenting them as a single non-negotiable number already misrepresents the document.
The number was calibrated on tooth decay
Here is the detail that surprises people most. Read the WHO guideline itself and you find that both thresholds, 10% and 5%, do not derive from the data on body weight or diabetes. They derive from the data on dental caries.
The 10% threshold rests on moderate-quality evidence from observational studies of tooth decay. The 5% threshold rests on very low-quality evidence from ecological studies in which a dose-response relationship was observed below 5%.
The body-weight evidence was reviewed too. It shows that increasing free sugars increases body weight, and that reducing them lowers it. But it could not pin down a numeric cutoff. Tooth decay could: it accumulates across a lifetime, it starts in childhood, and its relationship with sugar is regular enough to place a marker on.
None of this makes the recommendation absurd. It means that "25 g" is not the point where your metabolism tips over. It is the point where caries incidence keeps declining in the populations studied. A public health benchmark, not a personal physiological threshold.
"Free sugars" does not mean "sugar"
The second misunderstanding throws off every kitchen calculation. The recommendation is not about sugar in the everyday sense, but about a precise category.
The WHO defines free sugars as monosaccharides and disaccharides added to foods by manufacturers, cooks or consumers, plus the sugars naturally present in honey, syrups, fruit juices and fruit juice concentrates.
What falls outside that definition is worth remembering:
- sugars in whole, fresh fruit and vegetables;
- lactose in milk and unsweetened dairy products.
The WHO's reasoning is blunt: there is no evidence that those sugars have harmful effects. An apple does not count. A glass of apple juice does, because pressing the fruit releases the sugars from the matrix that held them.
The scale of the quiet sources says more than any list of banned foods. A tablespoon of ketchup contains around 4 grams of free sugars; a can of soda contains up to 40. You can spend the entire 25 g allowance without a single recognizably sweet item appearing on your plate.
National agencies use different yardsticks
If you have run into a figure of 100 grams somewhere and assumed a contradiction, there isn't one. It is a different counter.
In 2016, France's food safety agency recommended that adults consume no more than 100 g of total sugars per day, excluding lactose and galactose, and no more than one sweetened drink. The agency justified that scope with an honest admission: it could not distinguish the health effects of naturally occurring sugars from those of added sugars.
So the two recommendations do not contradict each other. They measure different things. One hundred grams of total sugars includes fruit. Twenty-five or fifty grams of free sugars excludes it. Comparing the two numbers head-to-head is meaningless, and yet that is exactly what most comparison tables online do.
What cardiovascular risk adds
If the WHO thresholds come from tooth decay, where does that leave everything else? One landmark study contributes a hard number.
Published in 2014 in JAMA Internal Medicine, the analysis by Quanhe Yang and colleagues followed 11,733 US adults for a median of 14.6 years. Compared with people who drew less than 10% of their calories from added sugar, participants in the 10 to 24.9% band showed an elevated risk of cardiovascular mortality (hazard ratio 1.30; 95% confidence interval, 1.09-1.55). Above 25%, the hazard ratio rose to 2.75 (1.40-5.42).
Two caveats matter. This is an observational study: it establishes an association, not causation. And the confidence interval for the most exposed group is wide, a sign of small numbers in that band. Still, the 10% threshold, the one behind the strong recommendation, is also where the risk curve starts bending upward. Two independent lines of reasoning converge on the same place.
An average describes nobody
There is one more limitation running under all these numbers, and it may be the most useful one to know.
In 2015, a team at the Weizmann Institute published a study in Cell that has become a reference point: Zeevi and colleagues measured 46,898 post-meal glucose responses across 800 participants. The central finding is that those responses vary enormously between individuals, including after standardized meals that were strictly identical. Two people, the same food, two different curves.
This does not invalidate population-level recommendations. They exist to steer food policy, and they do that job well. It means no universal threshold can tell you what 25 g does to you, on a Tuesday, after a short night.
Observe instead of counting
Counting your grams of free sugars daily means reading labels that often do not carry the information, and separating added sugars from naturally occurring ones, which European labelling does not do for you. It is a discouraging exercise, and discouragement is the surest route to giving up.
A more sustainable approach shifts the question. Instead of "how much did I eat?", try "what did it produce?".
- Identify your dominant sources. A single week of observation is usually enough to reveal the two or three inputs that carry the most weight: drinks, breakfast, late-afternoon snacking.
- Record the context, not just the quantity. A sugar craving that shows up at the same hour every day usually tells a story about sleep or a skipped meal, not about willpower.
- Change one thing at a time. Swapping out a single daily sweetened drink already removes several dozen grams without banning anything else.
- Watch what follows. Mid-afternoon energy, sleep quality, mood: those are your real output measures.
This is precisely the terrain where a logbook beats a reference table. In Kantise you can create a custom tracker for whatever you actually want to observe, then cross that series with sleep or activity. Our page on understanding sugar cravings and energy stability walks through the approach. The logic is the same one behind how much water you should actually drink: the general recommendation gives you a frame, your own series gives you the answer.
And since sugar also feeds your gut flora, the question connects directly to the gut-brain axis and its influence on mood.
Why this is coming up now
Nutri-Score, the front-of-pack label used across much of Europe, is changing its formula, and sugar sits at the center of the revision. The new algorithm tightens how sweetened products are scored, and manufacturers have a two-year window to bring their packaging into line. Sweetened dairy drinks, flavored drinking yogurts and sweetened plant-based beverages are now classified as beverages, which worsens their grade. Plain semi-skimmed milk, meanwhile, drops from A to B.
The practical consequence: during the transition, two comparable products can display letters calculated under two different formulas. Understanding what "free sugars" actually covers stays more reliable than trusting a letter printed on a package.
What to take away
- 50 g of free sugars per day is the strong recommendation; 25 g is the conditional target. They are not two versions of the same number.
- Both thresholds were calibrated on dental caries data, not on body weight or diabetes.
- "Free sugars" excludes whole fruit and lactose, but includes fruit juice and honey.
- France's 100 g figure covers total sugars: a different scope, not a contradiction.
- Glycemic response varies sharply between individuals, so your own series beats any average.
FAQ
25 g or 50 g: which number should I aim for?
The 50 g figure matches the WHO's strong recommendation, the one resting on the firmest evidence. The 25 g figure is an additional target, explicitly labelled conditional. Reaching 50 g is already a substantial change for most people; getting down toward 25 g is a bonus, not a failure if you do not manage it.
Does fruit count toward the 25 g?
Not for whole, fresh fruit: those sugars are not free sugars. Yes for fruit juice, even home-squeezed with nothing added, because the WHO explicitly classifies it as a free sugar.
Why does France's agency say 100 g?
Because it measures total sugars, excluding lactose and galactose, so fruit is included. The agency explained in 2016 that it could not separate the health effects of naturally occurring sugars from added ones. The 100 g and the 25 g are not directly comparable.
Is honey a better option than white sugar?
As far as the recommendation is concerned, no. The WHO explicitly places honey and syrups among free sugars, on the same footing as added sugar. They draw on the same budget.
Is it worth counting grams every day?
Rarely, over the long run. European labelling does not separate added sugars from naturally occurring ones, which makes the arithmetic approximate at best. Spotting your two or three dominant sources and watching their effect on energy and sleep produces more usable results than daily accounting.
Does a Nutri-Score A guarantee low sugar?
The score aggregates several criteria, and sugar is only one of them. The new algorithm tightens exactly that component, but manufacturers have two years to update packaging: during the transition, two products may carry letters from different formulas. Reading the "of which sugars" line remains safer.
Going further
The 25 g figure is neither a myth nor a law of physiology. It is a public health benchmark, tied to one specific kind of evidence, and issued with caveats that media coverage almost always strips away. Knowing it for what it is leaves you free to use it intelligently: a frame first, your own observations second.
Ready to see what sugar actually changes in your days? Start with one week of observation, changing nothing.
Sources
- World Health Organization (2015). Guideline: Sugars Intake for Adults and Children, WHO Guidelines / NCBI Bookshelf.
- World Health Organization (2015). WHO calls on countries to reduce sugars intake among adults and children.
- Yang Q. et al. (2014). Added Sugar Intake and Cardiovascular Diseases Mortality Among US Adults, JAMA Internal Medicine.
- Zeevi D. et al. (2015). Personalized Nutrition by Prediction of Glycemic Responses, Cell.
- ANSES (2016). Sugar in food.
- CLCV. The new Nutri-Score finally approved.
Kantise is a tool for observing and connecting your personal data, not a medical device. The information in this article is provided for general information only and does not replace professional medical advice. If you have diabetes, an eating disorder, or questions about your diet, talk to your doctor or a registered dietitian.