It is the most reliable scene in the calendar of childhood. The birthday cake comes out, the candles are blown out, the children get their slices, and within twenty minutes the room has achieved a state of physics-defying chaos. Furniture becomes an obstacle course. Screams reach a pitch that should require a permit. And every adult in the room nods in shared, world-weary certainty: it is the sugar.

Few parenting beliefs feel as obvious as this one. A child eats something sweet and becomes “hyper.” The transformation seems to happen in front of your eyes, repeatable on demand, predictable as sunrise. Surveys have repeatedly found that the large majority of parents are convinced of it, and many adjust family life around it, rationing sweets before school or confiscating Halloween candy lest it detonate bedtime.

There is only one problem with this universal observation: when scientists actually tested it — dozens of times, under conditions where neither the children, the parents, nor the researchers knew who had eaten sugar — the effect vanished. Sugar does not make children hyperactive. The claim is false.

*Illustration: Decades of controlled research have looked for a link between sugar and children's behavior in the brain — and failed to find one.*

The claim

The claim under review is straightforward and categorical: consuming sugar makes children hyperactive. Not that sugar is unhealthy in general — few would dispute that excess added sugar is linked to tooth decay, weight gain, and metabolic trouble — but that it produces a specific behavioral effect, quickly and visibly, in ordinary children. Give a child sweets and watch them bounce off the walls; withhold sweets and watch them calm down.

For a causal claim of this kind to hold, a controlled comparison has to show it: children who consume sugar should behave measurably worse — more active, more impulsive, less able to concentrate — than comparable children who consume a look-alike placebo without sugar. That is precisely the test the scientific literature has run, again and again, for more than four decades. And it has failed, with a consistency that researchers themselves have remarked upon as unusual.

Where the belief came from: California, 1973

The myth has a birth certificate, and it is not in a pediatric neurology textbook. It traces to a California allergist named Dr. Benjamin Feingold, who in the mid-1970s published a widely read book, Why Your Child Is Hyperactive. Feingold proposed that artificial food colors and flavors — along with salicylates, the aspirin-like compounds found in many foods, and certain preservatives — could cause or worsen hyperactivity, which he had observed, he said, when treating allergic children who then changed their diets.

Feingold’s book was a bestseller, and his message landed on an audience of exhausted parents hungry for an easy, drug-free explanation for difficult behavior. Thousands of families adopted his elimination diet, which was severe: artificial colors, flavorings, and preservatives were out, and so were many ordinary fruits and vegetables, including apricots, grapes, and pickles, along with anything containing salicylates.

Here is the detail that matters for this fact check: Feingold’s diet did not specifically target sugar. But the leap from “food additives cause hyperactivity” to “sugar causes hyperactivity” was short, obvious, and made by millions of parents at once. Sugar was the visible, quantifiable, vaguely chemical-seeming substance in the sweets that children ate at exactly the occasions when they were already at their most frantic — birthday parties, holidays, school celebrations. The diet’s logic colonized a substance the diet itself had never named, and refined sugar — increasingly prevalent in children’s diets in the 1970s — took the blame.

The belief then got a second push in 1978, when a study in the journal Food and Cosmetics Toxicology examined 265 children whom their parents described as hyperactive and unable to concentrate. The researchers reported findings that suggested the children suffered from abnormally low blood sugar — reactive hypoglycemia — after consuming sugar, and proposed this as a mechanism by which sweets could drive wild behavior. It was the first time the sugar-hyperactivity theory had the appearance of a physiological explanation, and it entered parenting folklore almost intact.

It did not survive contact with better measurement. Subsequent work showed that the children in that 1978 study had blood sugar levels well within the normal range for children their age. The proposed mechanism evaporated, but the myth it had helped launch kept its momentum.

The gold-standard test: three weeks on each diet

The central criticism of the early sugar studies was that they used a single dose of sugar and a brief observation window. In the early 1990s, Dr. Mark Wolraich and colleagues set out to close that loophole with what remains the definitive experiment on the question.

The design was ambitious. Two groups of children were recruited: 25 “normal” preschoolers aged three to five, and 23 school-age children aged six to ten whose own parents described them as sensitive to sugar. Each family followed three different diets, in sequence, for three weeks each: one high in sucrose with no artificial sweeteners, one low in sucrose with aspartame, and one low in sucrose with saccharin as the control. On every diet, families also avoided artificial colors, additives, and preservatives, and the children’s daily intake was measured.

Then came the blinding that gives the study its authority. Behavior and cognition were assessed using 39 different behavioral and cognitive variables — ratings of attention, hyperactivity, mood, executive function, and academic performance — by researchers who did not know whether the children they were observing were “normal” or “sugar-sensitive,” and did not know which diet each child was on.

The result: no differences. Not between diets, not between groups, not on any of the 39 measures. Most strikingly, the children whom their own parents were certain were sugar-sensitive showed no behavioral difference at all across the high-sugar and low-sugar diets. As the authors concluded: “Even when intake exceeds typical dietary levels, neither dietary sucrose nor aspartame affects children’s behaviour or cognitive function.”

This was published in the New England Journal of Medicine in 1994 (Wolraich, Lindgren, Stumbo, Stegink, Appelbaum & Kiritsy, vol. 330, pp. 301–307). It was the study that ended the debate among scientists, even if it has taken three decades to reach the birthday parties.

The meta-analysis: 23 experiments, one verdict

A single study, no matter how elegant, is never the whole story. So in 1995, Wolraich teamed up with David Wilson and Joann White to do what had not been done: pool the entire controlled literature and measure the effect with the combined weight of all the evidence. The meta-analysis was published in the Journal of the American Medical Association.

The inclusion rules were strict. A study made the cut only if it (1) had children consume a known quantity of sugar, (2) included a placebo condition using an artificial sweetener, (3) was blinded — the children, parents, and research staff did not know who got what — and (4) reported statistics usable for computing effect sizes. Sixteen reports met the bar, yielding 23 within-subject experiments covering hundreds of children, and the authors analyzed 14 distinct measurement constructs, from direct behavioral observation to academic testing.

The verdict was not hedged. As the authors wrote: “This meta-analysis of the reported studies to date found that sugar (mainly sucrose) does not affect the behavior or cognitive performance of children.” Across all 14 constructs, the 95 percent confidence intervals for the effect sizes included zero — the statistical signature of no detectable effect. The result was so uniformly negative, by one account, that a statistician reviewing the analysis told the team he had never seen such consistently negative results.

The paper was honest about its limits. A tiny effect, or effects confined to unusual subsets of children, could not be completely ruled out. But the authors also addressed the other half of the puzzle — why parents are so certain — and concluded that the strong belief of parents “may be due to expectancy and common association.” That sentence turned out to be the second great discovery of the sugar wars.

*Illustration: Medical researchers have tested the sugar-hyperactivity link in dozens of blinded trials — none found the promised effect.*

The experiment that explains your eyes

If sugar does nothing, why does every parent have a mental video library of sugared-up children ricocheting around a room? In 1994, researchers Donald Hoover and Richard Milich ran the experiment that answers that question, and it may be the most revealing study in the entire literature.

They recruited 35 boys aged five to seven whose mothers had reported them as behaviorally “sugar sensitive,” then randomly assigned mother-son pairs to two groups. The mothers in the experimental group were told their sons had just been given a large dose of sugar. The mothers in the control group were told their sons had received a placebo. In reality, every single child received the same sugar-free aspartame drink.

The mother-son pairs were videotaped interacting, and the mothers were then questioned about their child’s behavior. The mothers who believed their children had been sugared rated them as significantly more hyperactive than the control mothers did. Behavioral observation of the tapes showed these mothers also hovered more — maintaining closer physical proximity — and showed trends toward criticizing, looking at, and talking to their sons more than the control mothers did.

The children’s actual behavior was indistinguishable. The entire effect was in the observer. The mothers saw hyperactivity because they expected to see it, and the expectation itself changed how they parented in the moment — more control, more criticism — which is exactly the kind of interaction that can make a child squirmier, feeding the illusion in a feedback loop.

A companion-style experiment by researchers Spring and Alexander showed the same thing from a different angle: all children in that study consumed drinks actually sweetened with sugar, yet parents rated their children as more active, impulsive, and fidgety only when they were led to believe the sweetener had been sugar rather than aspartame. What the parents believed the child consumed determined the verdict, not what the child had consumed. Expectancy did all the work.

Why the illusion persists

Expectancy is not the whole story. The myth is also propped up by a set of perfectly ordinary features of the occasions when sugar and chaos coincide.

First, the context. Think about where children reliably eat large amounts of sugar in view of adults: birthday parties, holidays, school celebrations, amusement parks, Halloween. Every one of these is an environment engineered to make children excited — novel places, crowds of friends, games, music, noise, broken routines, later bedtimes, the thrill of presents. Each of these factors is, independently, a well-documented driver of exactly the running-screaming behavior being blamed on the cake. The cake is simply the most visible and nameable variable in a room full of them.

Second, confirmation bias and selective memory. The belief is now so widespread that it shapes what gets noticed and remembered. The child who is wild after the party is filed away as evidence. The child who eats an ice cream on a quiet Sunday afternoon and then sits down to draw — a scene that happens constantly — is filed under nothing, because nothing happened. We remember the hits and discard the misses.

Third, there is a genuine small-c kernel: blood sugar and energy. Sugar is a rapid source of fuel for the body, and the physical sensation of eating something sweet can genuinely feel energizing, especially when combined with excitement and caffeine (as in cola). But a fleeting feeling of energy is not hyperactivity as clinicians understand it — the sustained inattention, impulsivity, and restlessness that characterize ADHD — and the trials measured the latter and found nothing.

And there is an uncomfortable fourth possibility that the Hoover and Milich study raises: some of what parents see as “sugar hyperactivity” may be partly manufactured by the parents themselves. A mother who expects sugar-fueled misbehavior hovers, corrects, and criticizes more — behavior that can agitate a child who would otherwise have been fine. The prophecy fulfills itself, and the sugar takes the blame.

*Illustration: Expectation shapes perception — parents who believe sugar was consumed tend to rate children as more hyperactive, whatever the child actually drank.*

The food-coloring footnote

One common rejoinder deserves attention, because it confuses two different questions. In 2007, researchers led by Donna McCann published a randomized, double-blinded, placebo-controlled trial in The Lancet finding that mixtures of certain artificial food colorings and the preservative sodium benzoate could produce small increases in hyperactive behavior in three-year-olds and eight-to-nine-year-olds in the general population. This is the finding that led to the well-known warning labels on foods containing those colorings in the European Union.

That finding is about synthetic colorings, not sugar. Conflating the two is easy to do — sweets often contain both — but the science keeps them separate. Indeed, David Benton’s 2008 review in Critical Reviews in Food Science and Nutrition concluded that well-designed studies examining sucrose and children’s behavior “produced no evidence that it has an adverse influence.”

A 2013 discussion paper by Flora and Polenick in The Psychological Record went further in the opposite direction, arguing that high glucose consumption does not contribute to hyperactivity, inattention, juvenile delinquency, or cognitive decline in children — and that, if anything, glucose improves athletic, academic, and cognitive performance and may enhance self-control. Meanwhile, a 2019 birth-cohort study by Del-Ponte and colleagues found no association between consistently high sucrose consumption between ages six and eleven and the later incidence of ADHD. Some limited observational studies have reported correlations between heavy sugar-sweetened beverage consumption and ADHD risk, but observational designs cannot distinguish the sweet drink from everything else that travels with it — and the randomized trials, which can, keep coming back empty.

The honest summary of the modern literature, then, is this: controlled experiments have not found that sugar makes children hyperactive, while poor-quality diets in general remain a legitimate health concern for completely different reasons — dental decay, excess weight, and metabolic disease.

What this does not mean

This verdict is narrow, and it matters that it stays narrow. Saying that sugar does not cause hyperactivity is not saying that sugar is harmless, nor that parents should mainline candy into toddlers. Excess sugar in childhood diets is genuinely linked to tooth decay, weight gain, and metabolic problems, and it displaces better nutrition. Limiting sweets for those reasons is evidence-based parenting. Limiting sweets because of a promised behavioral explosion that the trials cannot find is not.

The verdict is also about populations and averages. The JAMA meta-analysis authors themselves noted that a very small effect, or effects in unusual subsets of children, could not be definitively ruled out. Science rarely deals in absolutes. But the claim being tested — the one parents actually hold, that sugar visibly and reliably wires children up — requires an effect large enough to see across a crowded party room. That effect does not exist in the data.

*Illustration: The wild energy parents blame on cake usually belongs to the occasion — parties, outings, and celebrations — not to the sugar.*

The Verdict

The claim that sugar makes kids hyperactive is False.

The belief traces to Dr. Benjamin Feingold’s 1970s elimination diet, which never targeted sugar in the first place, and to a 1978 study on reactive hypoglycemia whose mechanism did not survive remeasurement. Against that stands a chain of evidence that is unusually one-sided: Wolraich and colleagues’ 1994 New England Journal of Medicine trial, which put children — including ones their parents swore were sugar-sensitive — on three-week high-sucrose, aspartame, and saccharin diets and found no differences across 39 behavioral and cognitive measures; and the 1995 JAMA meta-analysis of 16 reports and 23 blinded experiments, which concluded that sugar “does not affect the behavior or cognitive performance of children.” Dozens of challenge trials since, including studies of children with ADHD and studies using single challenge doses, have returned the same answer.

What does exist is an explanation for the belief itself. Hoover and Milich showed in 1994 that mothers told their children had been sugared rated them as more hyperactive even when every child had received a sugar-free drink — the expectancy created the observation. And the occasions on which children eat the most sugar in front of adults are the occasions on which everything else — excitement, novelty, crowds, broken routines — is already pushing them toward chaos.

So the next time the cake comes out and the room detonates, blame the party. The science already has — for thirty years.