Sugar: the two WHO numbers are not one rule

The two sugar numbers you have heard are not the same kind of advice. Under 10 percent of your daily energy from free sugars is the strong WHO recommendation. Under 5 percent, roughly 25 g or six teaspoons, is conditional, which means the evidence behind it is weaker and WHO says so itself. Almost everyone who repeats the two repeats them as one rule. Free sugars also reach wider than the sugar bowl, because honey, syrups and fruit juice all count, and juice counts when the label says one hundred per cent. None of this makes sugar poison, or you careless for eating it. It tells you which number is worth arguing about.

The numbers, per 100 g

How much is actually in it

• What health bodies count as sugar is wider than the sugar bowl. Honey, syrups and fruit juice all count, and juice counts even when the label says one hundred per cent. — general population, as a definition used to set intake limits Evidence E source
  Free sugars defined as mono- and disaccharides added by manufacturer, cook or consumer, plus the sugars naturally present in honey, syrups, fruit juices and fruit juice concentrates

What a real portion looks like

• The headline number is under ten per cent of daily energy from free sugars. The stricter five per cent, about 25 g or six teaspoons, is a weaker recommendation and not the same kind of advice. — adults and children worldwide Evidence E source
  Less than 10% of total energy intake as the strong recommendation, with a further reduction below 5%, roughly 25 g a day, offered as conditional

What your body absorbs

• The same juice did different things to different people in the same trial, and the split showed up in their blood chemistry, not just in their glucose curve. — healthy young men, split after the fact into high and low responders Evidence B source
  High responders showed larger differences in glucose peak between full juice and matrix-free drink, alongside different 60-minute plasma metabolomes, notably in juice-derived metabolites such as dihydroferulic acid glucuronide
• People who feel ill after fructose are not necessarily failing to absorb it. In a blind trial the amount reaching their blood was the same as in people with no symptoms. — 30 patients with functional gastrointestinal disorders, sorted afterwards into fructose intolerant and tolerant by their symptoms (n = 30) Evidence B source
  Plasma fructose area under the curve over two hours 578 against 564 micromol per litre per hour (P = 0.39), with no correlation between fructose in the blood and symptoms, while symptoms did correlate with breath hydrogen and methane
• Among people sent to hospital clinics with irritable bowel syndrome, a breath test flags fructose malabsorption in about four in ten, which is far above the rate in healthy people. — patients referred to secondary care with irritable bowel syndrome, compared with healthy controls (n = 13 studies, 3 415 patients for fructose) Evidence A source
  Pooled prevalence of fructose malabsorption 41% (95% CI 21 to 60) or 43% (95% CI 23 to 62) depending on the fructose dose used for the breath test, with lactose malabsorption at 54% (95% CI 44 to 64), all significantly above healthy controls

What cooking and storage change

• Stripping the fruit matrix out of orange juice did not change how much glucose reached the blood in two hours, but it did make the peak higher and the rise faster. — healthy young men Evidence B source
  No significant difference in incremental area under the curve, but 100 per cent juice lowered the glucose peak against the matrix-free drink and lowered glucose at 15 minutes against both the half-matrix and matrix-free drinks

What people believe that the data does not support

• Honey is not a free pass. Pooling 69 trials, ten grams a day nudged long-term blood sugar down while pushing blood pressure, triglycerides, fasting glucose and an inflammation marker the wrong way. — 3 544 participants across 69 randomised controlled trials, searched to October 2024 (n = 69 trials, 3 544 participants) Evidence A source
  Haemoglobin A1c down by 0.25 percentage points (95% CI -0.45 to -0.05) on about 10 g of honey a day, against adverse movement in systolic blood pressure, aspartate transferase, triglycerides, fasting blood glucose and high-sensitivity C-reactive protein
• The old claim that sugar makes children hyperactive is no longer simply wrong: of sixteen studies looking at added sugar and attention problems, fifteen found a link. — children, adolescents and adults with a clinical ADHD diagnosis or validated symptom assessment, across 38 observational and 10 interventional studies (n = 48 studies from 1 174 records) Evidence C source
  15 of 16 studies of added sugars, sugar-sweetened beverages, sweets or sweet dietary patterns reported positive associations with ADHD diagnosis or symptom severity; 6 of 10 intervention studies reported improvement after dietary change

What is still unknown

• The same review that found the link says the arrow may point the other way: most of the studies measured diet and symptoms on the same day, so restless children may simply be eating more sweets. — the 38 observational studies in the review, dominated by cross-sectional designs (n = 48 studies) Evidence C source
  Reverse causation named as a live possibility, with severe ADHD symptoms potentially driving dietary choices rather than following from them

The bottom line

Honey gets treated as the safe sugar. Across 69 trials about 10 g of honey a day moved long term blood sugar down by a quarter of a percentage point and moved blood pressure, triglycerides, fasting glucose and an inflammation marker the wrong way at the same time. Juice is the other swap that looks better than it behaves. Taking the fruit matrix out of orange juice left the same amount of glucose in the blood over two hours, while the peak went higher and the rise came faster, and some people in that trial reacted far more than others. Sugar and hyperactivity is no longer a clean myth. Fifteen of sixteen studies of added sugar found a link with attention problems, and the same review warns that most of them measured diet and symptoms on the same day, so restless children eating more sweets would produce exactly this picture. If sugar leaves you bloated or in pain, that is a separate question. Among people referred to hospital with irritable bowel syndrome roughly four in ten failed a fructose breath test, and in another trial the people with symptoms had just as much fructose reaching their blood as the people without. Take that to a doctor rather than to a longer list of forbidden foods.

Sources

1. sug-e2-01 · expert body statement
   monosaccharides and disaccharides (e.g. sucrose, fructose, glucose) added to foods and beverages by the manufacturer, cook or consumer, and sugars naturally present in honey, syrups, fruit juices and fruit juice concentrates
   Healthy diet, fact sheet, World Health Organization · checked 2026-09-18
2. sug-e2-02 · expert body statement
   A further reduction to below 5% or roughly 25 grams (6 teaspoons) per day would provide additional health benefits.
   WHO calls on countries to reduce sugars intake among adults and children, World Health Organization, 2015 · checked 2026-09-18
3. sug-e2-04 · randomised crossover trial, post hoc analysis
   The occurrence of responder groups with differentiated plasma metabolomes underscores the importance of considering both the food matrix and individual physiological differences when assessing the glycemic impact of fruit juices.
   Food & Function, 2026 · checked 2026-09-18
4. sug-e2-05 · randomised double-blind crossover trial
   Fructose intolerance in FGID is not related to post-ingestion plasma concentrations of fructose and its metabolites.
   Neurogastroenterology & Motility, 2021 · checked 2026-09-18
5. sug-e2-06 · meta-analysis of prevalence studies
   Using random effects models, the estimated pooled rates were 41% (95% CI 21–60%) and 43% (95% CI 23–62%), respectively (Fig. 3).
   Scientific Reports, 2022 · checked 2026-09-18
6. sug-e2-03 · randomised crossover trial
   Differences in incremental area under the curve (iAUC) with the 100% OJ and 50% OJ against the 0% OJ did not reach significance, but the 100% OJ significantly lowered the glucose peak (Cmax) compared to 0% OJ and reduced glucose levels at 15 minutes (C15 min) compared to 50% OJ and 0% OJ.
   Food & Function, 2026 · checked 2026-09-18
7. sug-e2-07 · umbrella review with meta-analysis of randomised controlled trials
   Honey supplementation was associated with a reduction in HbA1C (WMD: −0.25, 95% CI: −0.45 to −0.05; I²: 58.6%, CoE: Very Low), while RJ lowered INS levels (WMD: −3.63 µU/mL, 95% CI: −6.77 to −0.49; I²: 27.8%, CoE: Very Low).
   Nutrition & Diabetes, 2025 · checked 2026-09-18
8. sug-e2-08 · systematic review
   Sugar-related exposures showed the most consistent pattern: 15 of 16 studies examining added sugars, sugar-sweetened beverages, sweets, candy, or sweet dietary patterns reported positive associations with ADHD diagnosis, symptom severity, hyperactivity, or less favorable ADHD-related outcomes.
   Nutrients, 2026 · checked 2026-09-18
9. sug-e2-09 · systematic review, risk of bias assessment
   In many studies, dietary intake and symptom burden were assessed at the same time point, increasing the possibility of reverse causation, whereby more severe ADHD symptoms may influence dietary choices rather than result from them.
   Nutrients, 2026 · checked 2026-09-18