Mark ở vũ trụ khácVIMark học AI
Mark ở vũ trụ khác · Another Universe Common Myths · Children · Expectations and evaluation

Sugar makes kids hyperactive

?

"No candy for him, or he'll be bouncing off the walls."

The mythsugar makes kids hyper, a "sugar high"
Spread fromthe 1970s, around the Feingold diet movement
The evidenceblind trials: no different from placebo
Key sourcesWolraich et al., JAMA 1995; Hoover and Milich, 1994
One-minute recap

Essence

Blind the test, the effect vanishesWhen nobody knows which cup has sugar, kids on sugar and kids on placebo behave the same.
23 studies, one answerA 1995 meta-analysis found no effect on behaviour or on thinking and learning.
The adults' eyes changeMothers told "your son just had sugar" rated him more hyperactive, though he only had placebo.
The setting is the culpritParties, crowds of friends, late nights, excitement. Sugar is just the most visible thing on the table.
But not harmlessWHO still advises limiting free sugars, for tooth decay and weight gain. That is about health, not about "hyper".

Sources: Wolraich, Wilson and White, JAMA 274(20), 1995; Hoover and Milich, Journal of Abnormal Child Psychology 22(4), 1994; WHO, guideline on sugars intake for adults and children, 2015.

Part 1

The question behind the myth

Millions of parents and teachers have "seen with their own eyes" kids go wild after candy. If blind trials find nothing, what are they seeing? The answer of this whole series fits in one line: belief comes first, and the eyes follow.

1970sA hypothesis about foodDyes and flavourings were suspected of causing hyperactivity. Sugar got pulled in too.
→
1994–1995Blind testsKids get sugar or placebo, nobody knows which. No difference.
→
TodayStill believedBecause what changes is how the adults look, not the child.
Part 2

Where the myth came from

There is no single "original quote". The myth grew through a chain of real events, each retold a little off.

Timeline. 1973: Feingold suspects dyes and flavourings, not sugar. Mid-1970s: some doctors shift the blame to sugar. 1979: the press coins the "Twinkie defense". 1980: a correlational study is read as causal. 1994 to 1995: blind trials find no effect. 2012: most teachers surveyed in the UK and the Netherlands still believe it. 1973Feingold:dyes, flavourings mid-1970s"cane sugar"pulled in too 1979"Twinkie defense"in the papers 1980correlationread as cause 1994–1995blind trials:no difference 201257% of UK teachersstill believe a variant the myth "drifted" from food dyes to sugar, then stayed after the evidence was in
The myth did not start with sugar. Feingold talked about dyes, artificial flavourings and salicylates. Because heavily dyed foods were often sugary too, the two slowly merged into one.Sources: National Research Council, Understanding and Preventing Violence, vol. 2 (1994), on Feingold and the 1980 study (secondary source); Prinz, Roberts and Hantman, J. Consulting and Clinical Psychology 48(6), 1980; Dekker et al., Frontiers in Psychology 3:429, 2012, table 1.
1973 · Paediatric allergist Benjamin FeingoldPresents at the American Medical Association meeting: dyes, artificial flavourings and salicylates cause hyperactivity. Later publishes the bestseller Why Your Child Is Hyperactive.
Mid-1970s · Sugar gets namedPaediatrician William Crook writes that cane sugar is a leading culprit. (Secondary source: a US nutrition review citing "Crook, 1975"; the original was not available.)
1979 · The "Twinkie defense"The Dan White trial in San Francisco. The defence argued depression; junk food was cited as a symptom. The press named it as if snack cakes made a man kill.
1980 · A correlational studyPrinz and colleagues: in hyperactive children, more sugar went with more destructive behaviour. The authors called it "suggestive"; many read it as "sugar causes".
1994–1995 · Blind trials and a meta-analysisWolraich (NEJM 1994), Hoover and Milich (1994), then a meta-analysis of 23 studies (JAMA 1995).
2012 · Still here"Children are less attentive after sugary drinks and snacks": 57% of UK teachers and 55% of Dutch teachers surveyed said true.

Sources: National Research Council (1994), vol. 2 (Feingold, the 1980 study); Snopes, "The Twinkie Defense" (secondary source); Prinz, Roberts and Hantman (1980); Dekker, Lee, Howard-Jones and Jolles, Frontiers in Psychology 3:429 (2012), table 1, sample of 242 teachers.

Why this myth is so easy to believe

Sugar shows up exactly when kids are already wild: birthday parties, Halloween, holidays, days off. There are also crowds of friends, games, late nights, presents. Adults remember the cake and forget the rest.

The story has a plausible mechanism: "sugar is energy, energy has to go somewhere". It sounds tidy, but the body does not work that simply, and trials do not show that result.

It gives adults a comfortable explanation: the child is not wild because he is tired, bored, or because of how we handled things, but because of the candy. The authors of the 1995 meta-analysis themselves wrote that parents' belief may come from "expectancy and common association".

Part 1

Where the myth came from

Feingold 1973food dyesTwinkie57% of teachers
Part 2

What the evidence says

blind tests23 studies35 mothersWHO
Part 3

Work lesson

criteria written firstevidencehide cuesscore independently
Part 3

What the evidence says

Double-blind trial: nobody knows which cup has sugar

double-blind, placebo-controlledWolraich et al., NEJM 1994
48 children; each family followed three diets in turn, three weeks each: high sugar, low sugar with aspartame, low sugar with saccharin. The research team supplied the food. Children, parents and assessors did not know which diet was which. In the 23 children whose parents called them sugar-sensitive: no difference on 39 behavioural and cognitive measures. 3 weeks · high sugarsucrose 3 weeks · aspartamelow sugar 3 weeks · saccharinplacebo everyone "blindfolded"everyone "blindfolded"everyone "blindfolded" = "sugar-sensitive" kids: 0 differences across 39 behavioural and cognitive measures
The team deliberately picked children whose parents believed they were "sugar-sensitive", 23 children aged 6–10, along with 25 typical preschoolers. In the preschoolers, only 4 of 31 measures differed, with no consistent pattern.Source: Wolraich et al., New England Journal of Medicine 330(5):301–307, 1994 (paper abstract).

Meta-analysis: 23 studies, every confidence interval touches zero

meta-analysisWolraich, Wilson and White, JAMA 1995
An effect-size axis from minus 0.4 to plus 0.6 with a zero line. The means of 14 measures range from minus 0.14, direct observation of behaviour, to plus 0.30, academic tests. All 14 confidence intervals include zero. −0.4−0.20+0.2+0.4+0.6 no effect −0.14+0.30 observed behaviouracademic tests band: spread of the 14 mean values · all 14 confidence intervals include zero
Only studies that blinded the children, the parents and the assessors were included, with an artificial sweetener as the control. 16 reports, 23 studies. The authors concluded sugar does not affect behaviour or cognition, but noted that a small effect, or an effect in a subset of children, cannot be ruled out.Source: Wolraich, Wilson and White, JAMA 274(20):1617–1621, 1995 (abstract: 16 reports, 23 studies, effect sizes −0.14 to +0.30 across 14 measures). The diagram shows only the two ends of the range, not each measure.

The key experiment: change what the mother hears, change nothing in the child

expectancy effectHoover and Milich, 1994
35 pairs of mothers and sons aged 5 to 7, mothers who believed their sons were sugar-sensitive. All boys drank the placebo aspartame. One group of mothers was told their son had a large dose of sugar, the other was told placebo. Mothers in the "expecting sugar" group rated their sons significantly more hyperactive, stayed physically closer, and tended to criticise, look at and talk to them more. 35 boysall drank placebo mother hears: "he justhad lots of sugar" mother hears: "hehad a placebo" rates him more hyperactivestays physically closertends to criticise more rates him as usualintervenes less
The children were the same; what differed was the sentence said to the adult. Mothers and sons played together on video. The effect was stronger in mothers with more rigid thinking styles.Source: Hoover and Milich, Journal of Abnormal Child Psychology 22(4):501–515, 1994 (paper abstract: 35 mother–son pairs, boys aged 5–7, all given aspartame).

Eyes see the belief

An ordinary burst of running around gets logged as "see, sugar high".

Hands act on the belief

The mother hovers and corrects more. A tightly shadowed child may really push back.

A self-confirming loop

Every "sighting" makes the belief stronger, even though the cup held no sugar.

Correlation is not causation: the arrow may point the other way

reverse causationDel-Ponte et al., 2019
Three explanations for why high sugar intake and hyperactivity go together. One: sugar causes hyperactivity, not supported by blind trials. Two: hyperactive, impulsive kids eat more sweets. Three: a third factor such as parties, late nights or family routine causes both. 1 · sugar → hypernot supported byblind trials 2 · hyper → sugarimpulsive kidssnack more 3 · a third factorparties, late nights,home routine → both seeing two things together does not tell you which one causes which
Observational studies give mixed results. A Brazilian cohort following 2,924 children from age 6 to 11 found high sugar intake did not go with new ADHD diagnoses, and suggested the reverse: ADHD leading to more sugar. A 2020 meta-analysis of observational studies did find an association, with high heterogeneity. No observational study can replace a blind trial.Sources: Del-Ponte et al., Journal of Affective Disorders 243:290–296, 2019 (Pelotas 2004 cohort, 2,924 children); Farsad-Naeimi et al., Complementary Therapies in Medicine 53:102512, 2020.

Don't fix one myth with another: sugar is still not harmless

free sugars · WHO 2015WHO, 4 March 2015
Two separate questions. Behaviour: blind trials do not find that sugar makes kids hyperactive. Health: WHO strongly recommends keeping free sugars below 10% of daily energy, and conditionally below 5%, about 25 grams or 6 teaspoons, because of tooth decay and unhealthy weight gain. Behaviour blind trials: no sign that sugar makes kids "hyper" → don't blame the candy Health (WHO 2015) below 10% of energy: strong below 5% (~25 g, ~6 tsp): conditional why: tooth decay, unhealthy weight gain
Free sugars are sugars added to foods and drinks, plus sugars naturally present in honey, syrups and fruit juices; not sugars in fresh fruit, vegetables and milk. This article is about behaviour; it is not nutrition advice, and it does not cover children with medical conditions that need special diets.Source: WHO, Guideline: Sugars intake for adults and children (2015) and WHO press release, 4 March 2015.
Part 4

Fact check: Solid, Caution, False

10 things people say about sugar and kids, checked against the original research. Tap a row for the reasoning.

SOLIDIn blind trials, kids given sugar are no more hyperactive than kids given placebo

The meta-analysis of 23 studies (JAMA 1995) and the three-weeks-per-diet trial (NEJM 1994) agree. The researchers deliberately included children whose parents believed they were "sugar-sensitive".

SOLIDExpectations make adults rate and behave differently

Hoover and Milich (1994): changing only what the mother was told made her rate her son more hyperactive and shadow him more closely. The 1995 meta-analysis also points to expectancy as a likely explanation for parents' belief.

SOLIDIt is still wise to limit sugar for children

For teeth and weight, not behaviour. WHO (2015): free sugars below 10% of daily energy is a strong recommendation; below 5% is conditional.

CAUTION"Some children are especially sensitive to sugar"

The grain of truth: the JAMA 1995 authors say an effect in a small subset cannot be ruled out, and a critical letter argued averaging could hide children who react in opposite directions. But when the very children parents called sensitive were tested (NEJM 1994), there was still no difference.

CAUTION"Food can make kids hyperactive"

For artificial food colours and the preservative sodium benzoate, there is one notable double-blind trial: McCann et al. (Lancet 2007), 153 three-year-olds and 144 eight-to-nine-year-olds, found higher hyperactivity scores in some groups. Germany's Federal Institute for Risk Assessment (BfR) called the effects "limited". This is about dyes, not sugar.

CAUTION"Kids with ADHD eat more sugar"

Some observational studies find an association (Farsad-Naeimi et al., 2020), others do not (Del-Ponte et al., 2019). Even where there is one, the causal arrow may run backwards: impulsive kids snack more.

CAUTION"A killer got a lighter sentence by blaming Twinkies"

In 1979 Dan White was convicted of voluntary manslaughter rather than murder. But the defence argued diminished capacity from depression; the switch to junk food was presented as a symptom. "Twinkie defense" was a press label. A psychiatrist did say junk food could worsen depression, so the rumour did not come from nowhere. (Secondary sources: Snopes, The Straight Dope.)

FALSE"Eating sugar causes ADHD"

ADHD is a neurodevelopmental disorder; no blind trial shows sugar causes it. The Pelotas cohort (2,924 children) found high sugar intake did not precede new ADHD cases. If your child's behaviour worries you, ask a doctor rather than just cutting candy.

FALSE"Just look at a birthday party, sugar makes kids wild"

Parties have crowds of friends, games, presents, late nights, and adults expecting the kids to go wild. Watching like this cannot separate sugar from everything else. That is why blind trials are needed.

FALSE"Sugar doesn't make kids hyper, so any amount is fine"

This "fixes" one myth with another. Sugar does not cause hyperactivity, but high free-sugar intake is linked to tooth decay and unhealthy weight gain (WHO 2015).

Sources: Wolraich et al. (JAMA 1995; NEJM 1994); Hoover and Milich (1994); McCann et al. (Lancet 370(9598), 2007); BfR, "Hyperactivity and additives" (2007); Del-Ponte et al. (2019); Farsad-Naeimi et al. (2020); WHO (2015); Snopes (secondary source).

Part 5

Verdict on the myth

Scores are Mark's assessment based on the sources in Part 8, not figures from any study.

How wrong
4/5
How persistent
5/5
How harmful
2/5
Why not 5/5 wrong
  • A small effect in a small subset cannot be ruled out
  • Food dyes and preservatives have some evidence
  • Limiting sugar is still good advice, just for a different reason
Where it does harm
  • Blames the candy and overlooks tiredness, boredom, overload
  • Adults hover and scold more because "he's on a sugar high"
  • May delay looking for the real cause

A myth that is almost completely wrong about behaviour, yet "accidentally right" about the advice. Its best lesson is about the observer.

More: why the 1995 evidence still hasn't won

A 2012 survey in the UK and the Netherlands found more than half of teachers surveyed still believed children are less attentive after sugary snacks (57% and 55%). The myth lives on because every party produces a fresh "eyewitness proof", while nobody gets to watch a blind trial.

This is also the common thread of the whole 10-part series: the 10% brain, left brain and right brain, learning styles, 7-38-55, 21 days, 10,000 hours, the goldfish, brainstorming, memory as a video camera. Every myth survives on a story that is easier to remember than a number.

Source: Dekker et al., Frontiers in Psychology 3:429 (2012), table 1.

Part 6

Work lesson: we see what we expect

The mothers in the experiment were not lying. They really saw their sons as wilder. Interviewers, KPI raters and CV screeners are the same: hear "this candidate is from a big-name company" or "this employee was weak last year", and the eyes follow.

Four steps to rate people more fairly: write criteria and a rating scale first, score specific evidence, hide expectation-setting information in the first round, score independently before discussing. 1 · criteria first1–5 scale, describedbefore meeting anyone 2 · score evidence"what, and what result"not "a feeling" 3 · hide the cuesphoto, school, referrerin the first round 4 · score alonewrite your scorethen discuss "I can tell just by looking" is what the mother told "he just had sugar" would say.
Each step is a "blindfold", like in a blind trial: remove what sets expectations, keep what needs judging.
Validity for predicting job performance: structured interviews 0.42, unstructured interviews 0.19. structured interview0.42 unstructured interview0.19 validity coefficient (0 = no prediction, 1 = perfect prediction)
The same questions for every candidate, scored on a ready-made scale, predict performance much better than a free-flowing chat.Source: Sackett, Zhang, Berry and Lievens, Journal of Applied Psychology 107(11), 2022 (figures via SIOP and HumRRO summaries, secondary sources; not yet checked against the original table).
Interviews: same questions, same scaleWrite 4–6 questions on the skills you need, with what a 1, 3 and 5 answer looks like. Score right after each question.
CVs: hide the expectation-setters in round onePhoto, age, school name, referrer. Score the work sample or test first, open the profile after.
Performance reviews: an evidence logOne line a week: what was done, what came of it. At review time, score from the log, not from impressions or last year's rating.
Panels: write scores before anyone speaksThe first speaker, especially the boss, sets the "expectation" for the whole room.
The question to ask before rating anyone: if I didn't know this about them, would I score them differently?Drawn from the Hoover and Milich experiment (1994).
Part 7

Read next

Wolraich, Wilson and White, "The Effect of Sugar on Behavior or Cognition in Children", JAMA (1995)Short, clear, and upfront about its own limits. A good model of how to write a scientific conclusion.
Famous Experiments: The Pygmalion effectThe same lesson from the classroom side: teachers' expectations change how they treat students.
Common Myths #01: We only use 10% of our brainsThis is the last of the 10-myth series. Read it again from the start to see how alike they are. The full set is at markhocai.com/vu-tru-khac.
Part 8

Sources

Original papers first. During this write-up the original pages were blocked from direct access, so details come from paper abstracts; secondary-only items are marked.

  1. M. L. Wolraich, D. B. Wilson, J. W. White, "The Effect of Sugar on Behavior or Cognition in Children: A Meta-analysis", JAMA 274(20):1617–1621, 1995. jamanetwork.com
  2. M. L. Wolraich et al., "Effects of Diets High in Sucrose or Aspartame on the Behavior and Cognitive Performance of Children", New England Journal of Medicine 330(5):301–307, 1994. nejm.org
  3. D. W. Hoover, R. Milich, "Effects of Sugar Ingestion Expectancies on Mother-Child Interactions", Journal of Abnormal Child Psychology 22(4):501–515, 1994. link.springer.com
  4. R. J. Prinz, W. A. Roberts, E. Hantman, "Dietary Correlates of Hyperactive Behavior in Children", Journal of Consulting and Clinical Psychology 48(6):760–769, 1980.
  5. National Research Council, Understanding and Preventing Violence, vol. 2, 1994 (section on diet, Feingold). nationalacademies.org
  6. S. Dekker, N. C. Lee, P. Howard-Jones, J. Jolles, "Neuromyths in Education: Prevalence and Predictors of Misconceptions among Teachers", Frontiers in Psychology 3:429, 2012, table 1. frontiersin.org
  7. C. Del-Ponte et al., "Sugar consumption and attention-deficit/hyperactivity disorder (ADHD): a birth cohort study", Journal of Affective Disorders 243:290–296, 2019.
  8. S. Farsad-Naeimi et al., "Sugar consumption, sugar sweetened beverages and Attention Deficit Hyperactivity Disorder: A systematic review and meta-analysis", Complementary Therapies in Medicine 53:102512, 2020.
  9. D. McCann et al., "Food additives and hyperactive behaviour in 3-year-old and 8/9-year-old children in the community", The Lancet 370(9598):1560–1567, 2007; and BfR, "Hyperactivity and additives: is there an association?", 2007.
  10. WHO, Guideline: Sugars intake for adults and children, 2015; press release, 4 March 2015. who.int
  11. P. R. Sackett, C. Zhang, C. M. Berry, F. Lievens, "Revisiting meta-analytic estimates of validity in personnel selection", Journal of Applied Psychology 107(11), 2022 (figures via secondary sources).
  12. Snopes, "The Twinkie Defense" (secondary source). snopes.com