Is Sugar Toxic?

Photo Illustration by Kenji Aoki for The New York Times

By GARY TAUBES

Published: April 13, 2011

On May 26, 2009, Robert Lustig gave a lecture called “Sugar: The Bitter Truth,” which was posted on YouTubethe following July. Since then, it has been viewed well over 800,000times, gaining new viewers at a rate of about 50,000 per month, fairlyremarkable numbers for a 90-minute discussion of the nuances of fructosebiochemistry and human physiology.

What the average American consumes in added sugars:

Graphic

High-Fructose Corn Syrup Consumption

Graphic

Kenji Aoki for The New York Times

Lustig is a specialist on pediatric hormone disorders and the leadingexpert in childhood obesity at the University of California, SanFrancisco, School of Medicine, which is one of the best medical schoolsin the country. He published his first paper on childhood obesity adozen years ago, and he has been treating patients and doing research onthe disorder ever since.

The viral success of his lecture, though, has little to do with Lustig’simpressive credentials and far more with the persuasive case he makesthat sugar is a “toxin” or a “poison,” terms he uses together 13 timesthrough the course of the lecture, in addition to the five references tosugar as merely “evil.” And by “sugar,” Lustig means not only the whitegranulated stuff that we put in coffee and sprinkle on cereal —technically known as sucrose — but also high-fructose corn syrup, whichhas already become without Lustig’s help what he calls “the mostdemonized additive known to man.”

It doesn’t hurt Lustig’s cause that he is a compelling public speaker.His critics argue that what makes him compelling is his practice oftaking suggestive evidence and insisting that it’s incontrovertible.Lustig certainly doesn’t dabble in shades of gray. Sugar is not just anempty calorie, he says; its effect on us is much more insidious. “It’snot about the calories,” he says. “It has nothing to do with thecalories. It’s a poison by itself.”

If Lustig is right, then our excessive consumption of sugar is theprimary reason that the numbers of obese and diabetic Americans haveskyrocketed in the past 30 years. But his argument implies more thanthat. If Lustig is right, it would mean that sugar is also the likelydietary cause of several other chronic ailments widely considered to bediseases of Western lifestyles — heart disease, hypertension and manycommon cancers among them.

The number of viewers Lustig has attracted suggests that people arepaying attention to his argument. When I set out to interview publichealth authorities and researchers for this article, they would ofteninitiate the interview with some variation of the comment “surely you’vespoken to Robert Lustig,” not because Lustig has done any of the keyresearch on sugar himself, which he hasn’t, but because he’s willing toinsist publicly and unambiguously, when most researchers are not, thatsugar is a toxic substance that people abuse. In Lustig’s view, sugarshould be thought of, like cigarettes and alcohol, as something that’skilling us.

This brings us to the salient question: Can sugar possibly be as bad as Lustig says it is?

It’s one thing to suggest, as most nutritionists will, that a healthfuldiet includes more fruits and vegetables, and maybe less fat, red meatand salt, or less of everything. It’s entirely different to claim thatone particularly cherished aspect of our diet might not just be anunhealthful indulgence but actually be toxic, that when you bake yourchildren a birthday cake or give them lemonade on a hot summer day, youmay be doing them more harm than good, despite all the love that goeswith it. Suggesting that sugar might kill us is what zealots do. ButLustig, who has genuine expertise, has accumulated and synthesized amass of evidence, which he finds compelling enough to convict sugar. Hiscritics consider that evidence insufficient, but there’s no way to knowwho might be right, or what must be done to find out, withoutdiscussing it.

If I didn’t buy this argument myself, I wouldn’t be writing about ithere. And I also have a disclaimer to acknowledge. I’ve spent much ofthe last decade doing journalistic research on diet and chronic disease —some of the more contrarian findings, on dietary fat, appeared in this magazine —– and I have come to conclusions similar to Lustig’s.

The history of the debate over the health effects of sugar has gone onfar longer than you might imagine. It is littered with erroneousstatements and conclusions because even the supposed authorities had notrue understanding of what they were talking about. They didn’t know,quite literally, what they meant by the word “sugar” and therefore whatthe implications were.

So let’s start by clarifying a few issues, beginning with Lustig’s useof the word “sugar” to mean both sucrose — beet and cane sugar, whetherwhite or brown — and high-fructose corn syrup. This is acritical point, particularly because high-fructose corn syrup has indeedbecome “the flashpoint for everybody’s distrust of processed foods,”says Marion Nestle, a New York University nutritionist and the author of“Food Politics.”

This development is recent and borders on humorous. In the early 1980s,high-fructose corn syrup replaced sugar in sodas and other products inpart because refined sugar then had the reputation as a generallynoxious nutrient. (“Villain in Disguise?” asked a headline in this paperin 1977, before answering in the affirmative.) High-fructose corn syrupwas portrayed by the food industry as a healthful alternative, andthat’s how the public perceived it. It was also cheaper than sugar,which didn’t hurt its commercial prospects. Now the tide is rolling theother way, and refined sugar is making a commercial comeback as thesupposedly healthful alternative to this noxious corn-syrup stuff.“Industry after industry is replacing their product with sucrose andadvertising it as such — ‘No High-Fructose Corn Syrup,’ ” Nestle notes.

But marketing aside, the two sweeteners are effectively identical intheir biological effects. “High-fructose corn syrup, sugar — nodifference,” is how Lustig put it in a lecture that I attended in SanFrancisco last December. “The point is they’re each bad — equally bad,equally poisonous.”

Refined sugar (that is, sucrose) is made up of a molecule of thecarbohydrate glucose, bonded to a molecule of the carbohydrate fructose —a 50-50 mixture of the two. The fructose, which is almost twice assweet as glucose, is what distinguishes sugar from othercarbohydrate-rich foods like bread or potatoes that break down upondigestion to glucose alone. The more fructose in a substance, thesweeter it will be. High-fructose corn syrup, as it is most commonlyconsumed, is 55 percent fructose, and the remaining 45 percent is nearlyall glucose. It was first marketed in the late 1970s and was created tobe indistinguishable from refined sugar when used in soft drinks.Because each of these sugars ends up as glucose and fructose in ourguts, our bodies react the same way to both, and the physiologicaleffects are identical. In a 2010 review of the relevant science, LucTappy, a researcher at the University of Lausanne in Switzerland who isconsidered by biochemists who study fructose to be the world’s foremostauthority on the subject, said there was “not the single hint” thatH.F.C.S. was more deleterious than other sources of sugar.

The question, then, isn’t whether high-fructose corn syrup is worse thansugar; it’s what do they do to us, and how do they do it? Theconventional wisdom has long been that the worst that can be said aboutsugars of any kind is that they cause tooth decay and represent “emptycalories” that we eat in excess because they taste so good.

By this logic, sugar-sweetened beverages (or H.F.C.S.-sweetenedbeverages, as the Sugar Association prefers they are called) are bad forus not because there’s anything particularly toxic about the sugar theycontain but just because people consume too many of them.

Those organizations that now advise us to cut down on our sugarconsumption — the Department of Agriculture, for instance, in its recentDietary Guidelines for Americans, or the American Heart Association inguidelines released in September 2009 (of which Lustig was a co-author) —do so for this reason. Refined sugar and H.F.C.S. don’t come with anyprotein, vitamins, minerals, antioxidants or fiber, and so they eitherdisplace other more nutritious elements of our diet or are eaten overand above what we need to sustain our weight, and this is why we getfatter.

Whether the empty-calories argument is true, it’s certainly convenient.It allows everyone to assign blame for obesity and, by extension,diabetes — two conditions so intimately linked that some authoritieshave taken to calling them “diabesity” — to overeating of all foods, orunderexercising, because a calorie is a calorie. “This isn’t aboutdemonizing any industry,” as Michelle Obama said about her Let’s Moveprogram to combat the epidemic of childhood obesity. Instead it’s aboutgetting us — or our children — to move more and eat less, reduce ourportion sizes, cut back on snacks.

Lustig’s argument, however, is not about the consumption of emptycalories — and biochemists have made the same case previously, thoughnot so publicly. It is that sugar has unique characteristics,specifically in the way the human body metabolizes the fructose in it,that may make it singularly harmful, at least if consumed in sufficientquantities.

The phrase Lustig uses when he describes this concept is “isocaloric butnot isometabolic.” This means we can eat 100 calories of glucose (from apotato or bread or other starch) or 100 calories of sugar (half glucoseand half fructose), and they will be metabolized differently and have adifferent effect on the body. The calories are the same, but themetabolic consequences are quite different.

The fructose component of sugar and H.F.C.S. is metabolized primarily bythe liver, while the glucose from sugar and starches is metabolized byevery cell in the body. Consuming sugar (fructose and glucose) meansmore work for the liver than if you consumed the same number of caloriesof starch (glucose). And if you take that sugar in liquid form — sodaor fruit juices — the fructose and glucose will hit the liver morequickly than if you consume them, say, in an apple (or several apples,to get what researchers would call the equivalent dose of sugar). Thespeed with which the liver has to do its work will also affect how itmetabolizes the fructose and glucose.

In animals, or at least in laboratory rats and mice, it’s clear that ifthe fructose hits the liver in sufficient quantity and with sufficientspeed, the liver will convert much of it to fat. This apparently inducesa condition known as insulin resistance, which is now considered thefundamental problem in obesity, and the underlying defect in heartdisease and in the type of diabetes, type 2, that is common to obese andoverweight individuals. It might also be the underlying defect in manycancers.

If what happens in laboratory rodents also happens in humans, and if weare eating enough sugar to make it happen, then we are in trouble.

The last time an agency of the federal governmentlooked into the question of sugar and health in any detail was in 2005,in a report by the Institute of Medicine, a branch of the NationalAcademies. The authors of the report acknowledged that plenty ofevidence suggested that sugar could increase the risk of heart diseaseand diabetes — even raising LDL cholesterol, known as the “badcholesterol”—– but did not consider the research to be definitive. Therewas enough ambiguity, they concluded, that they couldn’t even set anupper limit on how much sugar constitutes too much. Referring back tothe 2005 report, an Institute of Medicine report released last fallreiterated, “There is a lack of scientific agreement about the amount ofsugars that can be consumed in a healthy diet.” This was the sameconclusion that the Food and Drug Administration came to when it lastassessed the sugar question, back in 1986. The F.D.A. reportwas perceived as an exoneration of sugar, and that perceptioninfluenced the treatment of sugar in the landmark reports on diet andhealth that came after.

The Sugar Association and the Corn Refiners Association have also portrayed the 1986 F.D.A.report as clearing sugar of nutritional crimes, but what it concludedwas actually something else entirely. To be precise, the F.D.A.reviewers said that other than its contribution to calories, “noconclusive evidence on sugars demonstrates a hazard to the generalpublic when sugars are consumed at the levels that are now current.”This is another way of saying that the evidence by no means refuted thekinds of claims that Lustig is making now and other researchers weremaking then, just that it wasn’t definitive or unambiguous.

What we have to keep in mind, says Walter Glinsmann, the F.D.A.administrator who was the primary author on the 1986 report and who nowis an adviser to the Corn Refiners Association, is that sugar andhigh-fructose corn syrup might be toxic, as Lustig argues, but so mightany substance if it’s consumed in ways or in quantities that areunnatural for humans. The question is always at what dose does asubstance go from being harmless to harmful? How much do we have toconsume before this happens?

When Glinsmann and his F.D.A. co-authors decided no conclusive evidencedemonstrated harm at the levels of sugar then being consumed, theyestimated those levels at 40 pounds per person per year beyond what wemight get naturally in fruits and vegetables — 40 pounds per person peryear of “added sugars” as nutritionists now call them. This is 200calories per day of sugar, which is less than the amount in a can and ahalf of Coca-Cola or two cups of apple juice. If that’s indeed all weconsume, most nutritionists today would be delighted, including Lustig.

But 40 pounds per year happened to be 35 pounds less than whatDepartment of Agriculture analysts said we were consuming at the time —75 pounds per person per year — and the U.S.D.A. estimates are typicallyconsidered to be the most reliable. By the early 2000s, according tothe U.S.D.A., we had increased our consumption to more than 90 poundsper person per year.

That this increase happened to coincide with the current epidemics ofobesity and diabetes is one reason that it’s tempting to blame sugars —sucrose and high-fructose corn syrup — for the problem. In 1980, roughlyone in seven Americans was obese, and almost six million were diabetic,and the obesity rates, at least, hadn’t changed significantly in the 20years previously. By the early 2000s, when sugar consumption peaked,one in every three Americans was obese, and 14 million were diabetic.

This correlation between sugar consumption and diabetesis what defense attorneys call circumstantial evidence. It’s morecompelling than it otherwise might be, though, because the last timesugar consumption jumped markedly in this country, it was alsoassociated with a diabetes epidemic.

In the early 20th century, many of the leading authorities on diabetesin North America and Europe (including Frederick Banting, who shared the1923 Nobel Prize for the discovery of insulin) suspected that sugarcauses diabetes based on the observation that the disease was rare inpopulations that didn’t consume refined sugar and widespread in thosethat did. In 1924, Haven Emerson, director of the institute of publichealth at Columbia University, reported that diabetes deaths in New YorkCity had increased as much as 15-fold since the Civil War years, andthat deaths increased as much as fourfold in some U.S. cities between1900 and 1920 alone. This coincided, he noted, with an equallysignificant increase in sugar consumption — almost doubling from 1890 tothe early 1920s — with the birth and subsequent growth of the candy andsoft-drink industries.

Emerson’s argument was countered by Elliott Joslin, a leading authorityon diabetes, and Joslin won out. But his argument was fundamentallyflawed. Simply put, it went like this: The Japanese eat lots of rice,and Japanese diabetics are few and far between; rice is mostlycarbohydrate, which suggests that sugar, also a carbohydrate, does notcause diabetes. But sugar and rice are not identical merely becausethey’re both carbohydrates. Joslin could not know at the time that thefructose content of sugar affects how we metabolize it.

Joslin was also unaware that the Japanese ate little sugar. In the early1960s, the Japanese were eating as little sugar as Americans were acentury earlier, maybe less, which means that the Japanese experiencecould have been used to support the idea that sugar causes diabetes.Still, with Joslin arguing in edition after edition of his seminaltextbook that sugar played no role in diabetes, it eventually took onthe aura of undisputed truth.

Until Lustig came along, the last time an academic forcefully putforward the sugar-as-toxin thesis was in the 1970s, when John Yudkin, aleading authority on nutrition in the United Kingdom, published apolemic on sugar called “Sweet and Dangerous.” Through the 1960s Yudkindid a series of experiments feeding sugar and starch to rodents,chickens, rabbits, pigs and college students. He found that the sugarinvariably raised blood levels of triglycerides (a technical term forfat), which was then, as now, considered a risk factor for heartdisease. Sugar also raised insulin levels in Yudkin’s experiments, whichlinked sugar directly to type 2 diabetes. Few in the medical communitytook Yudkin’s ideas seriously, largely because he was also arguing thatdietary fat and saturated fat were harmless. This set Yudkin’s sugarhypothesis directly against the growing acceptance of the idea,prominent to this day, that dietary fat was the cause of heart disease, anotion championed by the University of Minnesota nutritionist AncelKeys.

A common assumption at the time was that if one hypothesis was right,then the other was most likely wrong. Either fat caused heart disease byraising cholesterol, or sugar did by raising triglycerides. “The theorythat diets high in sugar are an important cause of atherosclerosis andheart disease does not have wide support among experts in the field, whosay that fats and cholesterol are the more likely culprits,” as Jane E.Brody wrote in The Times in 1977.

At the time, many of the key observations cited to argue that dietaryfat caused heart disease actually support the sugar theory as well.During the Korean War, pathologists doing autopsies on American soldierskilled in battle noticed that many had significant plaques in theirarteries, even those who were still teenagers, while the Koreans killedin battle did not. The atherosclerotic plaques in the Americans wereattributed to the fact that they ate high-fat diets and the Koreans atelow-fat. But the Americans were also eating high-sugar diets, while theKoreans, like the Japanese, were not.

In 1970, Keys published the results of a landmark study in nutritionknown as the Seven Countries Study. Its results were perceived by themedical community and the wider public as compelling evidence thatsaturated-fat consumption is the best dietary predictor of heartdisease. But sugar consumption in the seven countries studied was almostequally predictive. So it was possible that Yudkin was right, and Keyswas wrong, or that they could both be right. The evidence has alwaysbeen able to go either way.

European clinicians tended to side with Yudkin; Americans with Keys. Thesituation wasn’t helped, as one of Yudkin’s colleagues later told me,by the fact that “there was quite a bit of loathing” between the twonutritionists themselves. In 1971, Keys published an article attackingYudkin and describing his evidence against sugar as “flimsy indeed.” Hetreated Yudkin as a figure of scorn, and Yudkin never managed to shakethe portrayal.

By the end of the 1970s, any scientist who studied the potentiallydeleterious effects of sugar in the diet, according to Sheldon Reiser,who did just that at the U.S.D.A.’s Carbohydrate Nutrition Laboratory inBeltsville, Md., and talked about it publicly, was endangering hisreputation. “Yudkin was so discredited,” Reiser said to me. “He wasridiculed in a way. And anybody else who said something bad aboutsucrose, they’d say, ‘He’s just like Yudkin.’ ”

What has changed since then, other than Americansgetting fatter and more diabetic? It wasn’t so much that researcherslearned anything particularly new about the effects of sugar orhigh-fructose corn syrup in the human body. Rather the context of thescience changed: physicians and medical authorities came to accept theidea that a condition known as metabolic syndrome is a major, if not the major, risk factor for heart disease and diabetes. The Centers for Disease Control and Prevention now estimatethat some 75 million Americans have metabolic syndrome. For those whohave heart attacks, metabolic syndrome will very likely be the reason.

The first symptom doctors are told to look for in diagnosing metabolicsyndrome is an expanding waistline. This means that if you’reoverweight, there’s a good chance you have metabolic syndrome, and thisis why you’re more likely to have a heart attack or become diabetic (orboth) than someone who’s not. Although lean individuals, too, can havemetabolic syndrome, and they are at greater risk of heart disease anddiabetes than lean individuals without it.

Having metabolic syndrome is another way of saying that the cells inyour body are actively ignoring the action of the hormone insulin — acondition known technically as being insulin-resistant. Because insulinresistance and metabolic syndrome still get remarkably little attentionin the press (certainly compared with cholesterol), let me explain thebasics.

You secrete insulin in response to the foods you eat — particularly thecarbohydrates — to keep blood sugar in control after a meal. When yourcells are resistant to insulin, your body (your pancreas, to be precise)responds to rising blood sugar by pumping out more and more insulin.Eventually the pancreas can no longer keep up with the demand or itgives in to what diabetologists call “pancreatic exhaustion.” Now yourblood sugar will rise out of control, and you’ve got diabetes.

Not everyone with insulin resistance becomes diabetic; some continue tosecrete enough insulin to overcome their cells’ resistance to thehormone. But having chronically elevated insulin levels has harmfuleffects of its own — heart disease, for one. A result is highertriglyceride levels and blood pressure, lower levels of HDL cholesterol(the “good cholesterol”), further worsening the insulin resistance —this is metabolic syndrome.

When physicians assess your risk of heart disease these days, they willtake into consideration your LDL cholesterol (the bad kind), but alsothese symptoms of metabolic syndrome. The idea, according to ScottGrundy, a University of Texas Southwestern Medical Center nutritionistand the chairman of the panel that produced the last edition of theNational Cholesterol Education Program guidelines, is that heart attacks50 years ago might have been caused by high cholesterol — particularlyhigh LDL cholesterol — but since then we’ve all gotten fatter and morediabetic, and now it’s metabolic syndrome that’s the more conspicuousproblem.

This raises two obvious questions. The first is what sets off metabolicsyndrome to begin with, which is another way of asking, What causes theinitial insulin resistance? There are several hypotheses, butresearchers who study the mechanisms of insulin resistance now thinkthat a likely cause is the accumulation of fat in the liver. Whenstudies have been done trying to answer this question in humans, saysVarman Samuel, who studies insulin resistance at Yale School ofMedicine, the correlation between liver fat and insulin resistance inpatients, lean or obese, is “remarkably strong.” What it looks like,Samuel says, is that “when you deposit fat in the liver, that’s when youbecome insulin-resistant.”

That raises the other obvious question: What causes the liver toaccumulate fat in humans? A common assumption is that simply gettingfatter leads to a fatty liver, but this does not explain fatty liver inlean people. Some of it could be attributed to genetic predisposition.But harking back to Lustig, there’s also the very real possibility thatit is caused by sugar.

As it happens, metabolic syndrome and insulinresistance are the reasons that many of the researchers today studyingfructose became interested in the subject to begin with. If you want tocause insulin resistance in laboratory rats, says Gerald Reaven, theStanford University diabetologist who did much of the pioneering work onthe subject, feeding them diets that are mostly fructose is an easy wayto do it. It’s a “very obvious, very dramatic” effect, Reaven says.

By the early 2000s, researchers studying fructose metabolism hadestablished certain findings unambiguously and had well-establishedbiochemical explanations for what was happening. Feed animals enoughpure fructose or enough sugar, and their livers convert the fructoseinto fat — the saturated fatty acid, palmitate, to be precise, thatsupposedly gives us heart disease when we eat it, by raising LDLcholesterol. The fat accumulates in the liver, and insulin resistanceand metabolic syndrome follow.

Michael Pagliassotti, a Colorado State University biochemist who didmany of the relevant animal studies in the late 1990s, says thesechanges can happen in as little as a week if the animals are fed sugaror fructose in huge amounts — 60 or 70 percent of the calories in theirdiets. They can take several months if the animals are fed somethingcloser to what humans (in America) actually consume — around 20 percentof the calories in their diet. Stop feeding them the sugar, in eithercase, and the fatty liver promptly goes away, and with it the insulinresistance.

Similar effects can be shown in humans, although the researchers doingthis work typically did the studies with only fructose — as Luc Tappydid in Switzerland or Peter Havel and Kimber Stanhope did at theUniversity of California, Davis — and pure fructose is not the samething as sugar or high-fructose corn syrup. When Tappy fed his humansubjects the equivalent of the fructose in 8 to 10 cans of Coke or Pepsia day — a “pretty high dose,” he says —– their livers would start tobecome insulin-resistant, and their triglycerides would go up in just afew days. With lower doses, Tappy says, just as in the animal research,the same effects would appear, but it would take longer, a month ormore.

Despite the steady accumulation of research, the evidence can still becriticized as falling far short of conclusive. The studies in rodentsaren’t necessarily applicable to humans. And the kinds of studies thatTappy, Havel and Stanhope did — having real people drink beveragessweetened with fructose and comparing the effect with what happens whenthe same people or others drink beverages sweetened with glucose —aren’t applicable to real human experience, because we never naturallyconsume pure fructose. We always take it with glucose, in the nearly50-50 combinations of sugar or high-fructose corn syrup. And then theamount of fructose or sucrose being fed in these studies, to the rodentsor the human subjects, has typically been enormous.

This is why the research reviews on the subject invariably conclude thatmore research is necessary to establish at what dose sugar andhigh-fructose corn syrup start becoming what Lustig calls toxic. “Thereis clearly a need for intervention studies,” as Tappy recently phrasedit in the technical jargon of the field, “in which the fructose intakeof high-fructose consumers is reduced to better delineate the possiblepathogenic role of fructose. At present, short-term-interventionstudies, however, suggest that a high-fructose intake consisting of softdrinks, sweetened juices or bakery products can increase the risk ofmetabolic and cardiovascular diseases.”

In simpler language, how much of this stuff do we have to eat or drink,and for how long, before it does to us what it does to laboratory rats?And is that amount more than we’re already consuming?

Unfortunately, we’re unlikely to learn anything conclusive in the nearfuture. As Lustig points out, sugar and high-fructose corn syrup arecertainly not “acute toxins” of the kind the F.D.A. typically regulatesand the effects of which can be studied over the course of days ormonths. The question is whether they’re “chronic toxins,” which means“not toxic after one meal, but after 1,000 meals.” This means that whatTappy calls “intervention studies” have to go on for significantlylonger than 1,000 meals to be meaningful.

At the moment, the National Institutes of Health are supportingsurprisingly few clinical trials related to sugar and high-fructose cornsyrup in the U.S. All are small, and none will last more than a fewmonths. Lustig and his colleagues at U.C.S.F. — including Jean-MarcSchwarz, whom Tappy describes as one of the three best fructosebiochemists in the world — are doing one of these studies. It will lookat what happens when obese teenagers consume no sugar other than whatthey might get in fruits and vegetables. Another study will do the samewith pregnant women to see if their babies are born healthier andleaner.

Only one study in this country, by Havel and Stanhope at the Universityof California, Davis, is directly addressing the question of how muchsugar is required to trigger the symptoms of insulin resistance andmetabolic syndrome. Havel and Stanhope are having healthy people drinkthree sugar- or H.F.C.S.-sweetened beverages a day and then seeing whathappens. The catch is that their study subjects go through thisthree-beverage-a-day routine for only two weeks. That doesn’t seem like avery long time — only 42 meals, not 1,000 — but Havel and Stanhope havebeen studying fructose since the mid-1990s, and they seem confidentthat two weeks is sufficient to see if these sugars cause at least someof the symptoms of metabolic syndrome.

So the answer to the question of whether sugar is as bad as Lustigclaims is that it certainly could be. It very well may be true thatsugar and high-fructose corn syrup, because of the unique way in whichwe metabolize fructose and at the levels we now consume it, cause fat toaccumulate in our livers followed by insulin resistance and metabolicsyndrome, and so trigger the process that leads to heart disease,diabetes and obesity. They could indeed be toxic, but they take years todo their damage. It doesn’t happen overnight. Until long-term studiesare done, we won’t know for sure.

One more question still needs to be asked, and this iswhat my wife, who has had to live with my journalistic obsession on thissubject, calls the Grinch-trying-to-steal-Christmas problem. What arethe chances that sugar is actually worse than Lustig says it is?

One of the diseases that increases in incidence with obesity, diabetesand metabolic syndrome is cancer. This is why I said earlier thatinsulin resistance may be a fundamental underlying defect in manycancers, as it is in type 2 diabetes and heart disease. The connectionbetween obesity, diabetes and cancer was first reported in 2004 in largepopulation studies by researchers from the World Health Organization’sInternational Agency for Research on Cancer. It is not controversial.What it means is that you are more likely to get cancer if you’re obeseor diabetic than if you’re not, and you’re more likely to get cancer ifyou have metabolic syndrome than if you don’t.

This goes along with two other observations that have led to thewell-accepted idea that some large percentage of cancers are caused byour Western diets and lifestyles. This means they could actually beprevented if we could pinpoint exactly what the problem is and preventor avoid that.

One observation is that death rates from cancer, like those fromdiabetes, increased significantly in the second half of the 19th centuryand the early decades of the 20th. As with diabetes, this observationwas accompanied by a vigorous debate about whether those increases couldbe explained solely by the aging of the population and the use of newdiagnostic techniques or whether it was really the incidence of canceritself that was increasing. “By the 1930s,” as a 1997 report by theWorld Cancer Research Fund International and the American Institute forCancer Research explained, “it was apparent that age-adjusted deathrates from cancer were rising in the U.S.A.,” which meant that thelikelihood of any particular 60-year-old, for instance, dying fromcancer was increasing, even if there were indeed more 60-years-olds witheach passing year.

The second observation was that malignant cancer, like diabetes, was arelatively rare disease in populations that didn’t eat Western diets,and in some of these populations it appeared to be virtuallynonexistent. In the 1950s, malignant cancer among the Inuit, forinstance, was still deemed sufficiently rare that physicians working innorthern Canada would publish case reports in medical journals when theydid diagnose a case.

In 1984, Canadian physicians published an analysis of 30 years of cancerincidence among Inuit in the western and central Arctic. While therehad been a “striking increase in the incidence of cancers of modernsocieties” including lung and cervical cancer, they reported, there werestill “conspicuous deficits” in breast-cancer rates. They could notfind a single case in an Inuit patient before 1966; they could find onlytwo cases between 1967 and 1980. Since then, as their diet became morelike ours, breast cancer incidence has steadily increased among theInuit, although it’s still significantly lower than it is in other NorthAmerican ethnic groups. Diabetes rates in the Inuit have also gone fromvanishingly low in the mid-20th century to high today.

Now most researchers will agree that the link between Western diet orlifestyle and cancer manifests itself through this association withobesity, diabetes and metabolic syndrome — i.e., insulin resistance.This was the conclusion, for instance, of a 2007 report published by theWorld Cancer Research Fund and the American Institute for CancerResearch — “Food, Nutrition, Physical Activity and the Prevention ofCancer.”

So how does it work? Cancer researchers now consider that the problemwith insulin resistance is that it leads us to secrete more insulin, andinsulin (as well as a related hormone known as insulin-like growthfactor) actually promotes tumor growth.

As it was explained to me by Craig Thompson, who has done much of thisresearch and is now president of Memorial Sloan-Kettering Cancer Centerin New York, the cells of many human cancers come to depend on insulinto provide the fuel (blood sugar) and materials they need to grow andmultiply. Insulin and insulin-like growth factor (and related growthfactors) also provide the signal, in effect, to do it. The more insulin,the better they do. Some cancers develop mutations that serve thepurpose of increasing the influence of insulin on the cell; others takeadvantage of the elevated insulin levels that are common to metabolicsyndrome, obesity and type 2 diabetes. Some do both. Thompson believesthat many pre-cancerous cells would never acquire the mutations thatturn them into malignant tumors if they weren’t being driven by insulinto take up more and more blood sugar and metabolize it.

What these researchers call elevated insulin (or insulin-like growthfactor) signaling appears to be a necessary step in many human cancers,particularly cancers like breast and colon cancer. Lewis Cantley,director of the Cancer Center at Beth Israel Deaconess Medical Center atHarvard Medical School, says that up to 80 percent of all human cancersare driven by either mutations or environmental factors that work toenhance or mimic the effect of insulin on the incipient tumor cells.Cantley is now the leader of one of five scientific “dream teams,”financed by a national coalition called Stand Up to Cancer, to study, inthe case of Cantley’s team, precisely this link between a specificinsulin-signaling gene (known technically as PI3K) and tumor developmentin breast and other cancers common to women.

Most of the researchers studying this insulin/cancer link seem concernedprimarily with finding a drug that might work to suppress insulinsignaling in incipient cancer cells and so, they hope, inhibit orprevent their growth entirely. Many of the experts writing about theinsulin/cancer link from a public health perspective — as in the 2007report from the World Cancer Research Fund and the American Institutefor Cancer Research — work from the assumption that chronically elevatedinsulin levels and insulin resistance are both caused by being fat orby getting fatter. They recommend, as the 2007 report did, that weshould all work to be lean and more physically active, and that in turnwill help us prevent cancer.

But some researchers will make the case, as Cantley and Thompson do,that if something other than just being fatter is causing insulinresistance to begin with, that’s quite likely the dietary cause of manycancers. If it’s sugar that causes insulin resistance, they say, thenthe conclusion is hard to avoid that sugar causes cancer — some cancers,at least — radical as this may seem and despite the fact that thissuggestion has rarely if ever been voiced before publicly. For just thisreason, neither of these men will eat sugar or high-fructose cornsyrup, if they can avoid it.

“I have eliminated refined sugar from my diet and eat as little as Ipossibly can,” Thompson told me, “because I believe ultimately it’ssomething I can do to decrease my risk of cancer.” Cantley put it thisway: “Sugar scares me.”

Sugar scares me too, obviously. I’d like to eat it in moderation. I’dcertainly like my two sons to be able to eat it in moderation, to notoverconsume it, but I don’t actually know what that means, and I’ve beenreporting on this subject and studying it for more than a decade. Ifsugar just makes us fatter, that’s one thing. We start gaining weight,we eat less of it. But we are also talking about things we can’t see —fatty liver, insulin resistance and all that follows. Officially I’m notsupposed to worry because the evidence isn’t conclusive, but I do.

Gary Taubes (gataubes@gmail.com) is a Robert Wood JohnsonFoundation independent investigator in health policy and the author of“Why We Get Fat.” Editor: Vera Titunik (v.titunik-MagGroup@nytimes.com).

A version of this article appeared in print on April 17, 2011, on page MM47 of the Sunday Magazine.