The Case Against Sugar

By Gary Taubes

10 min read

Why does it matter? Because "eat it in moderation" was the sugar industry's most successful invention.

The comfortable version of the sugar debate goes like this: it's empty calories, it's portion control, it's common sense. You know you eat too much, you know you should cut back, end of story. That framing is not a scientific conclusion. It's the argument the sugar industry spent fifty years and millions of dollars engineering — funding the scientists, buying the nutrition chairs, shaping the FDA reviews, and quietly destroying the careers of the researchers who disagreed. The actual case against sugar is a mechanism: fructose driving insulin resistance, insulin resistance triggering obesity, diabetes, heart disease, hypertension, cancer, and Alzheimer's through a single hormonal cascade. That case was available in the 1970s. The evidence was there. The people making the argument were there. That's the story this summary traces — the evidence, the people who buried it, and what it cost.

Sugar Hijacks the Same Brain Circuits as Cocaine — That's Not Metaphor, It's Mechanism

Imagine designing a drug with one requirement: it has to hide. From regulators. From the person consuming it. It should produce no staggering, no slurred speech, no flushing — none of the signals we've learned to associate with intoxication. It should taste like a reward and feel like one, while doing to the brain exactly what heroin does.

Sugar does this.

The hardware is the same. When you eat sugar, it triggers dopamine release in the nucleus accumbens — the brain's reward center, the same region activated by nicotine, cocaine, heroin, and alcohol. Repeated exposure produces dopamine down-regulation. The brain generates less dopamine naturally, receptors grow less sensitive, and you need progressively more sugar to register the same pleasure while ordinary satisfactions quietly flatten.

Serge Ahmed, a French researcher, demonstrated what this looks like. He took rats with months-long intravenous cocaine addictions and offered them a choice: their daily cocaine fix or a sweet solution. Within two days, the rats had abandoned the cocaine. The reason, since replicated in monkeys, is that neurons in the reward circuitry tuned to sweet taste outnumber those tuned to cocaine fourteen to one. The rat's brain was following the stronger signal.

This is why moderation advice has been failing people for centuries. Once you understand the mechanism, the way sugar worked its way into every packaged food and every celebration stops looking like a marketing success story and starts looking like something more familiar: a dependency spreading through a population that had no name for what it was developing.

The Link Between Sugar and Disease Was Named in 1857 — and Then Buried for a Century

In 1924, Haven Emerson took the podium at the American Medical Association's annual meeting with a decades-long dataset and one clear pattern: when sugar consumption rose, diabetes death rates rose with it. When sugar fell, as it did during WWI rationing, diabetes mortality fell too. He and his Columbia colleague Louise Larimore had tracked it across American cities and found the same pattern in each. A wartime shortage had produced, accidentally, the kind of controlled experiment no researcher could design. The data was not fringe science. It was mainstream epidemiology, presented to the country's largest medical gathering.

What happened next took thirty years and two men.

Elliott Joslin was America's dominant diabetes authority, the author of the textbook every physician learned from. Harold Himsworth was his British counterpart, eventually running the Medical Research Council for two decades. Through the 1930s and 1940s, each cited the other as proof that sugar didn't cause diabetes. Joslin called Himsworth's work meticulous and authoritative; Himsworth treated Joslin as the definitive reference. Both pointed to the same evidence: Japan had a carbohydrate-heavy diet and almost no diabetes, which they took to mean carbohydrates, including sugar, were benign.

Neither man checked Japanese sugar consumption specifically.

When you do, the argument collapses. As late as 1963, Japan's per capita sugar intake matched levels in England and the United States from a century earlier — the same era when diabetes was rare in those countries. The Japanese data, read correctly, supported the sugar hypothesis. Joslin and Himsworth had used it for forty years to support the opposite conclusion.

By the 1971 edition of Joslin's textbook, the question of whether sugar caused diabetes had simply vanished. Not refuted. Dropped.

Between the New York Times naming sugar as a disease driver in 1857 and a British chancellor announcing a sugar tax in 2016 lies 159 years. That gap is not the time it took science to reach a conclusion. It's the time it took a closed loop of authority — two men, each citing the other, neither checking the data — to be dismantled.

The Scientists Who Cleared Sugar Were on Sugar's Payroll

The science showing sugar causes disease was strong enough by the 1970s that the sugar industry needed to do something about the scientists. What it did instead was buy the regulators.

When the FDA launched its first comprehensive safety review of sugar in the early 1970s, the process that would officially decide whether sugar was "generally recognized as safe," it subcontracted the work to an independent committee of scientists. The man chosen to chair that committee was George W. Irving, a biochemist who had spent years on the scientific advisory board of the sugar industry's own research foundation, including two years as its chairman. Irving's committee, working under a mandate that required avoiding even the appearance of conflicts of interest, produced a report that leaned heavily on a single document: an 88-page white paper called "Sugar in the Diet of Man," written by researchers whose careers the sugar industry had been funding for decades and distributed to 25,000 recipients, including the press packets handed to newspaper food editors at their annual conference. The Sugar Association's role in commissioning and paying for that white paper appeared nowhere in the document itself.

The committee's 1977 conclusion: sugar was safe, the evidence linking it to diabetes was "circumstantial," and its connection to heart disease was "less than clear." The Sugar Association immediately translated those hedges into flat safety claims. What those claims never mentioned: the industry had spent the previous decade shaping which evidence existed to cite.

The dietary-fat hypothesis was the perfect alibi: if saturated fat caused heart disease, sugar was cleared by default, and the burden of proof fell on anyone who said otherwise. That hypothesis hadn't arrived by accident. Ancel Keys, whose work built that idea into nutritional orthodoxy, had been receiving sugar-industry funding since 1944. In 1970, he published a widely circulated letter eviscerating John Yudkin, the British researcher whose lab work had most directly linked sugar to elevated triglycerides and heart disease risk, dismissing Yudkin's entire body of evidence as beneath serious consideration. Keys's demolition of Yudkin was as vicious as it was influential. Shortly after, Yudkin retired, and his university broke its promise to let him keep his laboratory. The research program that might have answered the sugar question definitively was shuttered.

The result held for decades. The 1988 Surgeon General's report, the 1989 National Academies Diet and Health report, and the USDA's Dietary Guidelines all cited the GRAS review as authoritative, each one recycling a verdict written under a conflict of interest the public was never told about.

Sugar Isn't Dangerous Because It's Empty Calories — It's Dangerous Because of What It Does to Your Hormones

Why do people with metabolic problems tend to develop all of them at once? Get heavy enough, and you'll likely arrive with high blood pressure, elevated triglycerides, low HDL cholesterol, and glucose intolerance as a package. Cardiologists catalogued these as a cluster of bad luck. They were wrong. It's one problem with one mechanism. Sugar is the most likely trigger.

In 1960, Rosalyn Yalow, a medical physicist, and Solomon Berson, a physician, built the first technology capable of measuring hormone levels in living blood: the radioimmunoassay, which later earned Yalow a Nobel Prize. What they found overturned a century of assumptions. Virtually every hormone in the body mobilizes fat from fat cells, releasing stored energy when the body needs it. Insulin does the opposite. It drives fat into cells and holds it there. Releasing stored fat requires what Yalow and Berson called "the negative stimulus of insulin deficiency" — insulin has to fall before fat can escape. Chronically elevated insulin means fat cells are perpetually in hoarding mode.

Then came their second finding. Both obese people and type 2 diabetics had elevated blood sugar and elevated insulin simultaneously, a combination that makes no sense if insulin is doing its job. Their cells had become resistant to insulin's effect. The pancreas compensated by secreting more, compounding the problem. And the implication nobody wanted to follow: if elevated insulin causes fat accumulation, and the obese have elevated insulin, their obesity might not be causing their insulin resistance. The insulin resistance might be causing their obesity.

Gerald Reaven, an endocrinologist at Stanford, spent the 1980s assembling the evidence that insulin resistance was the common engine behind conditions medicine had been treating as separate diseases. His 1988 address to the American Diabetes Association named it Syndrome X, now called metabolic syndrome. Insulin resistance, he showed, doesn't produce one problem. It produces high triglycerides, low HDL cholesterol, hypertension, glucose intolerance, and an expanding waistline simultaneously. The cardiovascular risk that doctors had been attributing to cholesterol is better predicted by this cluster. Elevated LDL cholesterol isn't even a criterion for it. NIH clinicians, by one account, wished the finding would go away. Nobody knew how to deal with it.

What causes insulin resistance? Fructose — the half of sugar the liver converts to fat alone, with no shutoff. Fat accumulates, insulin resistance follows, the cluster assembles. Obesity, type 2 diabetes, heart disease, hypertension: one disruption, one substance.

Gout, Hypertension, Cancer, Alzheimer's — the Same Mechanism, Different Organs

Gout, hypertension, cancer, Alzheimer's are one upstream mechanism, expressing itself differently depending on which organ system it reaches. Medicine assigns each its own specialist and its own dietary villain: gout blames red meat, hypertension blames salt, cancer blames carcinogens. But when a cluster of conditions appears in every population that adopts Western eating, in the same rough order, Occam's Razor demands a single explanation.

Gout makes the argument most clearly, because the history is unusually precise. For centuries it was almost exclusively a disease of the British aristocracy — the rich could afford sugar, the rich got gout. Then in the 18th century, as sugar prices fell and sugar spread through British society, gout spread with it in exact parallel. Historians call this the "gout wave." The conventional explanation (too much meat and port) was already strained: gout incidence among largely vegetarian populations in India ran around 7 percent, and eating more protein actually increases uric acid excretion from the kidneys, which should be protective.

The biochemical explanation came from Finnish researchers in the late 1960s: fructose metabolism in the liver accelerates the breakdown of ATP, the cell's primary energy molecule, which is loaded with purines that the body converts into uric acid. More ATP breakdown means more uric acid. Alcohol raises uric acid the same way, through the same pathway to the same endpoint. Gout is a fructose disease.

That mechanism connects upward to cancer and Alzheimer's through insulin resistance. Elevated insulin and its close relative IGF-1 (a hormone that amplifies those same growth signals) are what cancer cells become, as researchers put it, "addicted to." Lewis Cantley, who leads the cancer research program at Weill Cornell and has traced how insulin's signaling pathway drives tumor growth in breast and colon cancer, says this is why sugar scares him. In Alzheimer's, the route is one step further: insulin resistance drives both diabetes and hypertension, which accelerate vascular damage in the brain and lower the threshold at which amyloid plaques tip into dementia. Type 2 diabetics face one and a half to two times the Alzheimer's risk of nondiabetics.

One molecule in the liver. One cascade of metabolic disruption. Every organ system it touches, a different diagnosis.

The Damage Doesn't Stop With You — It Programs the Next Generation in the Womb

In 1940, a survey of Arizona's Native Americans turned up 73 diabetes cases — comparable to any other population, nothing alarming. Twenty-three years later, Peter Bennett and Tom Burch arrived at the Gila River Reservation to study rheumatoid arthritis. They tested 900 Pima and found 30 percent with diabetic blood-sugar levels; among those over 30, every other person was an undiagnosed diabetic.

The hospitals weren't full because the Pima hadn't had the disease long enough to accumulate complications. Bennett and Burch moved permanently to Arizona. Over the next three decades, what they documented wasn't just one epidemic. It was a mechanism for producing the next one.

Children born to diabetic Pima mothers became diabetic at 30 times the rate of children born to healthy mothers, not because of what those children ate, but because of what crossed the placenta before they were born.

The mechanism: a pregnant woman with elevated blood sugar delivers glucose to the fetus in direct proportion to her own levels. The developing pancreas, flooded with glucose, overproduces insulin-secreting cells. The baby arrives metabolically primed, a fat infant with a pancreas already calibrated to a sugar-saturated world. By the time that child reaches adulthood, the insulin resistance that took their mother decades to develop arrives faster, harder, younger.

The researchers described a self-reinforcing loop each generation made worse, and wrote in 1988 that whether it could be broken was unknown. Each generation of diabetic mothers produces a generation more susceptible than the last, not through shared meals or learned habits, but through the chemistry of the womb itself.

This is why, for populations already deep in the epidemic, individual moderation may be too late.

The threshold wasn't crossed when the disease became visible. It was crossed one generation earlier, before anyone was born.

159 Years Is a Long Time to Know Something and Do Nothing

The gap between 1857 and 2016 isn't a story of delayed discovery. The connection was named, documented, buried, and renamed — while the industry that profited kept funding the scientists who cleared it. What sits between those two dates is a record of deliberate uncertainty, manufactured just long enough to matter.

The honest question the book leaves you with isn't how much sugar is safe. It's whether that question was ever answerable on the terms it was asked. Populations consuming a fraction of what any modern authority calls moderate were already developing diabetes, heart disease, and hypertension. The threshold, if it exists, sits below the amount you were told was fine. And for some of us — born to mothers already metabolically changed — the question was decided before we ever sat down to eat.