EATINGMADE

How do you cope with a mountainof conflicting diet advice?

By Marion Nestle

KEY CONCEPTS

III Nutrition advice isconfus-ing.Scientistshavediffi-culty deriving clear guide-lines because a study of anindividual nutrient fails to

produce an understandingof what happens to itwhen mixed with other

nutrients in the body.

. The picture becomes moreclouded because industrygroups constantly presstheir message to both gov-ernment agencies andconsumers about the ben-

efits of eating particularfood products.

. The simplestmessage maybe the best: do not over-

eat, exercise more, con-

sume mostly fruits, vege-

tables and whole grains,and avoid junk foods.

-The Editors

42 SCIENTIFIC AMERICAN

As a nutrition pr~fessor, I am constantlyasked why nutrition advice seems tochange so much and why experts so

often disagree. Whose information, people ask,can we trust? I'm tempted to say, "Mine, ofcourse," but I understand the problem. Yes,nutrition advice seems endlessly mired in scien-tific argument, the self-interest of food compa-nies and compromises by government regula-tors. Nevertheless, basic dietary principles arenot in dispute: eat less; move more; eat fruits,vegetables and whole grains; and avoid toomuch junk food.

"Eat less" means consume fewer calories,which translates into eating smaller portionsand steering clear of frequent between-mealsnacks. "Move more" refers to the need to bal-

ance calorie intake with physical activity. Eat-ing fruits, vegetables and whole grains providesnutrients unavailable from other foods. Avoid-

ing junk food means to shun "foods of minimalnutritional value"-highly processed sweetsand snacks laden with salt, sugars and artificialadditives. Soft drinks are the prototypical junkfood; they contain sweeteners but few or nonutrients.

If you follow these precepts, other aspects of

the diet matter much less. Ironically, this advice '"has not changed in years. The noted cardiolo- ~gist Ancel Keys (who died in 2004 at the age of ~

I'

&.=. ~...-

I

IiI

OLDADVICESTILL HOLDSTRUEIn 1959 Anceland MargaretKeysoffered the following-familiarand still useful-precepts regard-

ing nutrition and activity:

III Do not get fat; if you are fat,reduce.

l1liRestrictsaturated fats: fats in

beef, pork, lamb, sausages,margarine and solid shortenings;fats in dairy products.

III Prefervegetable oils to solid fats

but keeptotal fats under30 percent of your diet calories.

. Favorfreshvegetables,fruitsandnonfatmilkproducts.

. Avoidheavyuseofsaltandrefinedsugar.

. Gooddiets do not depend on

drugs and fancy preparations.

. Getplentyofexerciseandoutdoorrecreation.

44 SCIENTIFIC AMERICAN

100) and his wife, Margaret, suggested similarprinciples for preventing coronary heart diseasenearly50 yearsago [seesidebarat left].

But I can see why dietary advice seems like a

moving target. Nutrition research is so difficultto conduct that it seldom produces unambigu-ous results. Ambiguity requires interpretation.And interpretation is influenced by the individ-ual's point of view, which can become thor-oughly entangled with the science.

Nutrition Science ChallengesThis scientific uncertainty is not overly surpris-

ing given that humans eat so many differentfoods. For any individual, the health effects ofdiets are modulated by genetics but also by edu-cation and income levels, job satisfaction, phys-ical fitness, and the use of cigarettes or alcohol.To simplify this situation, researchers typicallyexamine the effects of single dietary compo-nents one by one.

Studies focusing on one nutrient in isolationhave worked splendidly to explain symptomscaused by deficiencies of vitamins or minerals.But this approach is less useful for chronicconditions such as coronary heart disease anddiabetes that are caused by the interactionof dietary, genetic, behavioral and social fac-tors. If nutrition science seems puzzling, it is be-cause researchers typically examine single nu-trients detached from food itself, foods separatefrom diets, and risk factors apart from otherbehaviors. This kind of research is "reductive"in that it attributes health effects to the con-

sumption of one nutrient or food when it is the

ORGANICFOODShavebeenshownto leavepeoplewho eat them with fewer synthetic pesticidesin their bodies than are found in those whoconsume conventional foods. Proving thatorganics contain more vitamins or antioxidantsis more difficult, but preliminary studies clearlysuggest that they do.

overall dietary pattern that really counts most.For chronic diseases, single nutrients usually

alter risk by amounts too small to measure ex-cept through large, costly population studies. Asseen recently in the Women's Health Initiative, aclinical trial that examined the effects of low-fat

diets on heart disease and cancer, participantswere unable to stick with the restrictive dietary

protocols. Because humans cannot be caged andfed measured formulas, the diets of experimen-tal and control study groups tend to converge,making differences indistinguishable over thelong run-even with fancy statistics.

It's the CaloriesFood companies prefer studies of single nutri-ents because they can use the results to sell prod-ucts. Add vitamins to candies, and you can mar-ket them as health foods. Health claims on the

labels of junk foods distract consumers fromtheir caloric content. This practice mattersbecause when it comes to obesity-which dom-inates nutrition problems even in some of thepoorest countries of the world-it is the caloriesthat count. Obesity arises when people con-sume significantly more calories than theyexpend in physical activity.

America's obesity rates began to rise sharplyin the early 1980s. Sociologists often attributethe "calories in" side of this trend to the de-

mands of an overworked population for conve-nience foods-prepared, packaged productsand restaurant meals that usually contain morecalories than home-cooked meals.

But other social forces also promoted the cal-orie imbalance. The arrival of the Reagan ad-ministration in 1980 increased the pace of in-dustry deregulation, removing controls on agri-cultural production and encouraging farmers togrow more food. Calories available per capita inthe national food supply (that produced byAmerican farmers, plus imports, less exports)rose from 3,200 a day in 1980 to 3,900 a daytwo decades later [see box on opposite page].

The early 1980s also marked the advent ofthe "shareholder value movement" on Wall

Street. Stockholder demands for higher short-term returns on investments forced food com-

~0:z;;:;2«"z«g;:

September 2007

panies to expand sales in a marketplace that al-ready contained excessive calories. Food com-panies responded by seeking new sales andmarketing opportunities. They encouraged for-merly shunned practices that eventually changedsocial norms, such as frequent between-mealsnacking, eating in book and clothing stores,and serving larger portions. The industry con-tinued to sponsor organizations and journalsthat focus on nutrition-related subjects and in-tensified its efforts to lobby government for fa-vorable dietary advice. Then and now food lob-bies have promoted positive interpretations ofscientific studies, sponsored research that canbe used as a basis for health claims, and at-tacked critics, myself among them, as propo-nents of "junk science." If anything, such activ-ities only add to public confusion.

FOODFACTOIDSToreduceyour weight by a poundof fat aweek, eat 500 fewer calories

eachday.

Carbohydratesand proteins haveabout 4 caloriespergram. Foodfatscontain more than twice asmuch: 9

calories pergram. A teaspoon holdsabout 5 grams.

Alcohol is metabolized in awaythat promotes accumulation of fatin the liver, leading to the proverbialbeer belly.

An adult expendsabout 100 caloriesfor everymile walked or run. It takesnearly three miles to burn off the calo-ries in a20-ounce soft drink.

"jL

Supermarkets as "Ground Zero"No matter whom I speak to, I hear pleas for helpin dealing with supermarkets, considered byshoppers as "ground zero" for distinguishinghealth claims from scientific advice. So I spent ayear visiting supermarkets to help people thinkmore clearly about food choices. The result wasmy book What to Eat.

Supermarkets provide a vital public servicebut are not social services agencies. Their job isto sell as much food as possible. Every aspect ofstore design-from shelf position to backgroundmusic-is basedonmarketing research [seecen-ter item on page 50]. Because this researchshows that the more products customers see, themore they buy, a store's objective is to exposeshoppers to the maximum number of productsthey will tolerate viewing.

III

A substantialrise in U.S.obesity ratesduring the pastfew decadeswasparalleledby increasesin the availabilityof largerportion sizes,total calories,caloricsweetenersandsugarysoft drinks in the food supply.Theapparentdip in threeof thesemeasures(calories,sugarsandsugarysoft drinks)after 1998maybeexplainedbygreateruseof artificial sweetenersandthe partial replacementof sugarysoft drinkswith beveragesthat arenot sweetenedwith sugars.

U.S. OBESITYRATESON THERISEPercentoftotalpopulation(ages20-74)classifiedasobese

~

{§~e;~%~'"'0-"

I;,;~.~~ibiS>~

~.§~3~~ffi~~"wz;:;:8OOw

5~'g<iNOw"'.~"z~~g~z~c;;~:;;

~~~:=~J,Z<ivO~N

~::;;:]O~",OO<ii"z~.

~~~U~O

;:~>

~gg~~

;r:~~~;;>~~~0 >-0~.~"z""""."-'3;::gQ

:5~~C'Q",8:i!<;~~~~

U~~:5""'-0

~~~:1::J;s,,~o"'<i>3§bZ<i>"'z~

50..

01976-1980 1988-1994 1999-2000

.................................................................................................................................................................................................

63SUPER-SIZE PORTIONS GROWNumberoffooditemsintroducedin largersizesbyrestaurantsandmanufacturersintheU.S.

36

126

1975-1979 1980-1984 1985-1989

www.SciAm.com

2001-2002 2003-2004

47

1990-1994 1995-1999

CALORIESAVAILABLEPerperson per day in the U.S. food supply

4,000 """""""'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''',

3,500................................

3,000 -,1975

T1985

T1995

T2005

CALORICSWEETENERSAVAILABLEPoundsperpersonintheU.S.foodsupply

T

1975

T1985

T1995

T2005

SUGARYSOFTDRINKSAVAILABLEGallonsperpersonintheU.S.foodsupply

::::::_::::::::::::~25

1975 1995T

20051985

SCIENTIFICAMERICAN 45

If consumers are confused about which foods

to buy, it is surely because the choices requireknowledge of issues that are not easily resolvedby science and are strongly swayed by social andeconomic considerations. Such decisions playout every day in every store aisle.

Are Organics Healthier?Organic foods are the fastest-growing segmentof the industry, in part because people are will-ing to pay more for foods that they believe arehealthier and more nutritious. The U.S. Depart-ment of Agriculture forbids producers of "Cer-tified Organic" fruits and vegetables from usingsynthetic pesticides, herbicides, fertilizers,genetically modified seeds, irradiation or fertil-

--

izer derived from sewage sludge. It licensesinspectors to ensure that producers followthose rules. Although the USDAis responsiblefor organics, its principal mandate is to pro-mote conventional agriculture, which explainswhy the department asserts that it "makes noclaims that organically produced food is saferor more nutritious than conventionally pro-duced food. Organic food differs from conven-tionally grown food in the way it is grown, han-dled and processed."

This statement implies that such differencesare unimportant. Critics of organic foods wouldagree; they question the reliability of organiccertification and the productivity, safety andhealth benefits of organic production methods.

Fats,Oilsand Sweets JJJ&..

...

.

.

...

USESPARINGLY~--

Milk, Yogurt andCheeseGroup

2-3 SERVINGS

Meat, Poultry, Fish, Dry Beans,

Eggs and Nuts Group2-3 SERVINGS

Vegetable Group3...5 SERVINGS

FruitGroup2-4 SERVINGS

Bread, Cereal. Rice and

~ PastaGroup." 6"'11 SERVINGS

1992

Whether you found the food pyramid created by the U.S. Depart-ment of Agriculture in 1992 beneficial or not, it was at least

simple to use. The familiar triangular nutrition guide suggested howmuch of each food category-grains, dairy products, fruits and vege-tables, meats and fats, oils and sweets-one should eat every day.

But in my opinion, the USDA'S2005 replacement, MyPyramid, is a

disaster. The process the agriculture agency employed to replace the1992foodpyramid(left) has been keptsecret. It remainsa mystery,for example, just how the department came up with a design for anew food guide that emphasizes physical activity but is devoid offood (right).Accordingto the USDAstaff, peopleshouldkeepphysi-cally active, eat in moderation, make personalized food choices, eat avariety of foods in the recommended number of servings, and pursuegradual dietary improvement. The color and width of the verticalbands of MyPyramid are meant to denote food groups and servings,but the only way to know this in detail is to log on to a computer.Users must go to www.pyramid.gov and type in gender, age andactivity level to obtain a "personalized" dietary plan at one of a doz-en calorie levels.

People who seek advice from this site, and millions have, find diet

46 SCIENTIFIC AMERICAN

-1

....

...

2005

plans notable for the large amounts of food they seem to recommendand for the virtual absence of appeals to "eat less" or to "avoid" cer-tain foods. Critics, not surprisingly, discern the strong influence offood industry lobbyists here. I myself, for example, am expected toconsume four cups of fruits and vegetables, six ounces of grains, fiveounces of meat and, of course, three cups of milk a day, along with acouple of hundred "discretionary calories" that I can spend on junk

foods. For all its flaws, the 1992 pyramid was easier to understandand use.

What MyPyramid really lacks is any notion of a hierarchical rank-ing of the items in a single food group in terms of nutritional desir-ability. The preliminary design of MyPyramid in 2004 looked muchlike the final version with one critical exception: it illustrated a hierar-chy of desirable food choices. The grain band, for instance, placedwhole-grain bread at the bottom (a positive ranking), pasta abouthalfway up (a middle rank) and cinnamon buns at the top ("eat

les~").Inthe finalversion,the USDAeliminatedalltraces of hierarchy,presumably because food companies do not want federal agencies toadvise eating less of their products, useful as such recommendationsmight be to an overweight public. -M.N.

~":>~uZ

September 2007

SOMEOBESECHILDRENin the U.S.consume 1,000calories a day (equal to about 59 sugar cubes) insweetened beverages alone. That is equivalent tomore than three 20-ounce bottles of soft drinks.

-15

i"~

Meanwhile the organic food industry longs forresearch to address such criticisms, but studiesare expensive and difficult to conduct. Never-theless, existing research in this area has estab-lished that organic farms are nearly as produc-tive as conventional farms, use less energy andleave soils in better condition. People who eatfoods grown without synthetic pesticides oughtto have fewer such chemicals in their bodies,and they do. Because the organic rules requirepretreatment of manure and other steps to re-duce the amount of pathogens in soil treatments,organic foods should be just as safe-or safer-than conventional foods.

Similarly, organic foods ought to be at leastas nutritious as conventional foods. And prov-ing organics to be more nutritious could helpjustify their higher prices. For minerals, thistask is not difficult. The mineral content of

plants depends on the amounts present in thesoil in which they are grown. Organic foods arecultivated in richer soils, so their mineral con-tent is higher.

But differences are harder to demonstrate

for vitamins or antioxidants (plant substancesthat reduce tissue damage induced by free rad-icals); higher levels of these nutrients relatemore to a food plant's genetic strain or protec-tion from unfavorable conditions after harvest-

ing than to production methods. Still, prelimi-nary studies show benefits: organic peachesand pears contain greater quantities of vitaminsC and E, and organic berries and corn containmore antioxidants.

Further research will likely confirm that or-ganic foods contain higher nutrient levels, butit is unclear whether these nutrients would

make a measurable improvement in health. Allfruits and vegetables contain useful nutrients,albeit in different combinations and concentra-

tions. Eating a variety of food plants is surelymore important to health than small differenc-es in the nutrient content of anyone food. Or-ganics may be somewhat healthier to eat, butthey are far less likely to damage the environ-ment, and that is reason enough to choose themat the supermarket.

"0,g51z0:z0:1:<t"Z..~

Dairy and CalciumScientistscannot easilyresolvequestionsaboutthe healtheffectsof dairy foods.Milk hasmany

www.SciAm.com

P'" ~

~~~t;::. .Sa."fa" II

"

< / '- '-- ~,

IIII; , II II

'1"r

""I

II

I~

components, and the health of people who con-sume milk or dairy foods is influenced by every-thing else they eat and do. But this area ofresearch is especially controversial because itaffects an industry that vigorously promotesdairy products as beneficial and opposes sug-gestions to the contrary.

Dairy foods contribute about 70 percent ofthe calcium in American diets. This necessarymineral is a principal constituent of bones,which constantly lose and regain calcium duringnormal metabolism. Diets must contain enoughcalcium to replace losses, or else bones becomeprone to fracture. Experts advise consumptionof at least one gram of calcium a day to replaceeveryday losses. Only dairy foods provide thismuch calcium without supplementation.

But bones are not just made of calcium; theyrequire the full complement of essential nutri-ents to maintain strength. Bones are strongerin people who are physically active and who donot smoke cigarettes or drink much alcohol.Studies examining the effects of single nutri-ents in dairy foods show that some nutritionalfactors-magnesium, potassium, vitamin Dand lactose, for example-promote calcium re-

Marion Nestle is Paulette

Goddard Professor in the

department of nutrition, food

studies and public health and

professor of sociology at New

York University. She receiveda Ph.D. in molecular biology

and an M.P.H. in public health

nutrition from the Universityof California. Berkeley. Nestle'sresearch focuses on scientific and

social factors that influence food

choices and recommendations.

She is author of Food Politics

(2002, revised 2007), Safe Food

(2003) and What to Eat (2006).

SCIENTIFICAMERICAN 47

tention in bones. Others, such as protein, phos-

phorus and sodium, foster calcium excretion.So bone strength depends more on overall pat-terns of diet and behavior than simply on calci-um intake.

Populations that do not typically consumedairy products appear to exhibit lower rates ofbone fracture despite consuming far less calci-umthan recommended[seesidebaron oppositepage]. Why this is so is unclear. Perhaps theirdiets contain less protein from meat and dairyfoods, less sodium from processed foods andless phosphorus from soft drinks, so they retaincalcium more effectively. The fact that calciumbalance depends on multiple factors could ex-plain why rates of osteoporosis (bone densityloss) are highest in countries where people eat

........--

the most dairy foods. Further research mayclarify such counterintuitive observations.

In the meantime, dairy foods are fine to eatif you like them, but they are not a nutritionalrequirement. Think of cows: they do not drinkmilk after weaning, but their bones supportbodies weighing 800 pounds or more. Cowsfeed on grass, and grass contains calcium insmall amounts-but those amounts add up. If

you eat plenty of fruits, vegetables and wholegrains, you can have healthy bones withouthaving to consume dairy foods.

A Meaty DebateCritics point to meat as the culprit responsiblefor elevating blood cholesterol, along with rais-ing risks for heart disease, cancer and other

different interpretationofthe findings."Whathappenedin our studywas very modest weightloss in all four groups:' says ChristopherD. Gardner, a nutrition scientist at the Stanford Prevention

Research Center. Allgroups also showed improvementJn individu-als' levels of cholesterol, blood pressure and insulin, even though

none of them followed their diet plans exactly. And far from over-

turning established ideas aboutJow-fat diets, the Stanfordinvestigation provided resounding confirmation of anothergenerally held belief; most people who try to lose weight, on

any kind of diet, will succeed, even if many of them regainthe weight later.

Contrast those conclusions with the results of

another study published in the Aprilissue ofAmerican Psychologist by researchers at theUni-

versity of California, LosAngeles. They analyzed31 long-term diet studies and found, as Gardner said, thatmost participants did see results~losing about 5 to 10percentof their total body mass. And they didit while onall kinds of diets.. But mostalso regained all that weight

over the longer term, and some put on even.more than

they had lost. Onlya small minority of subjects in the 31studies keptthe extra pounds off. The researchers' con-clusion? Eat in moderation and exercise regularly. (This

statement parallels similar advice nutritionist Marion

Nestle presents in the accompanying article.)Gardner thinks the traditional exhortation to cut

dietary fat has turned out to be a bad message. The

public health experts gotH wrong, he says: "It totallybackfired on us." People who consumed less fat often

turned to soda and similar corn-syrup-sweetened

products, along with other refined,low-fiber, car-bohydrate-rich foods. As a result, "the obesityepidemic has continued to grow. Calories have

continued to creep up, and it's been predominantly inthe refined carbohydrates."

Q'~::J~0EO:i5:""~",-",0;-;;;i'jtie,"';;;§N",:;!«c:>0;;«

goo'~o~~.,,=00~>;c"'._.0

§~~'":i':;:§310>'

~~;:;:0;;95'-"'~u~'.."~~

~"00"'~~oo:i'~'~'"uO~~00o~

;§",10.t;,§1ai'"~"',zo~9~'"",u3;"ZZ

::;;;o~

Ii§1:;!~'"

8~z>

By Paul Raeburn

'T" his past March, Stanford Universityresearchers published the resultsof one of the longest and most persuasive comparisons

of weight-loss programs ever conducted. Three of the

four dietsin the study are heavily promoted regi-

mens that have made their originators famous;.the

Atkins diet and th.e Zone diet, which both empha-

size. high-protein foods, and the Ornish diet, a

plan that prohibits most fatty foods. The fourth ~W"'heno-f

.

"'ldOW. -fa,d;m'h atmost nutrition '.,.'.,."".

,,

"..'.' ',.,' .,expertsrecommend.", ..r"

The results, published in theJourna/ of the Ameri-

can Medica/Association, were a surprise because

they seemed to overturn the conventionalwis-

dom. The experts' low-fatdiet was beaten by

Atkins's steak dirmers and bacon-and-egg

breakfasts. A year after starting their diets,

people on the Atkins plan-which unapol-

ogetically endorses high-fat protein such

as meats and dairyproductsto keep diet-

ers sated---haddropped an average of 10

pounds. Subjects onthe other diets had

lost between three and sixpounds (graph

on oppositepage). And members ofthe

Atkins test groupshowed no jump in

blood cholesterol levels, despite the highlevels of cholesterol intheir diet.

Reporters jumped on the obvious head-lines: "Atkins Fares Best ..." stated the Wash-

ington Post. "Atkins Beats Zone, Ornish and U.S.Diet Advice," the Associated Press declared; It

was the same everywhere else: Atkins had bested

the competition.

The newspaper accounts were not

wrong. But the lead author of

th.e Stanford study suggests a

.............

48 SCIENTIFIC AMERICAN September 2007

!

conditions. Supporters cite the lack of compel-ling science to justify such allegations; theyemphasize the nutritional benefits of meat pro-tein, vitamins and minerals. Indeed, studies indeveloping countries demonstrate healthimprovements when growing children are fedeven small amounts of meat.

But because bacteria in a cow's rumen attach

hydrogen atoms to unsaturated fatty acids, beeffat is highly saturated-the kind of fat that in-creases the risk of coronary heart disease. Allfats and oils contain some saturated fatty acids,but animal fats, especially those from beef, havemore saturated fatty acids than vegetable fats.Nutritionists recommend eating no more thana heaping tablespoon (20 grams) of saturated

fatty aci.ds a day. Beef eaters easily meet or ex-

II:II!

ceed this limit. The smallest McDonald's

cheeseburger contains 6 grams of saturated fat-ty acids, but a Hardee's Monster Thickburgerhas 45 grams.

Why meat might boost cancer risks, however,is a matter of speculation. Scientists began tolink meat to cancer in the 1970s, but even afterdecades of subsequent research they remain un-sure if the relevant factor might be fat, saturat-ed fat, protein, carcinogens or something elserelated to meat. By the late 1990s experts couldconclude only that eating beef probably increas-es the risk of colon and rectal cancers and pos-sibly enhances the odds of acquiring breast,prostate and perhaps other cancers. Faced withthis uncertainty, the American Cancer Societysuggests selecting leaner cuts, smaller portions

TheAtkinsplan,which advisesdietersto be lesscon-

cernedaboutfat, steerspeopletoward vegetablesandpro-

tein andawayfromsugarsand refinedcarbohydrates."May.

below carbisa better simplemessageto the publicthan low

fat," Gardnersays."Wetell them low carb,and theyget it.

Theycut out a coupleof sodasor a coupleof cookies,andthat addsup."

JamesHill, a psychologistandauthorityon weight loss,agreesthat the Atkins approachhasvirtues. "TheAtkinsdiet

isa greatwayto loseweight," hesays.But it "is not a wayto

keepweight off," heasserts."There'sno wayyoucandoitforever."

Hill isnotterriblyinterestedincomparingdietsordevis-ing newones."I think the weight-losspart issomethingwedo pretty well," hesays.Oneof hisareasof researchcon-

BATTLEOF THE DIET PLANS

Mean weight change over time (in kilograms)

0

~"""""""" "....

-6 :&:Start

T6

Months

_Zone -Ornish -Standard _Atkins

12

cernsindividualswho havereducedtheir weightandsustainedit. HiUandRenaWingof BrownUni-versityhaveestablishedwhatthey call theNationalWeightControlRegistryto collectdataon peoplewho havecut at least30 poundsandkept them offfor ayear.Manyhavelostmuch more~the averageis a 70-poundweight lossmaintainedfor sixyears."If you look athow they lost weight, there'snocommonalityat all," Hillsays.But"ifyou look athow theykept it off, there'salot of commonality."

Thekey,hecontinues,isexercise."Activitybecomesthe driver;food restrictiondoesn'tdo it.

Theideathat for the restof yourUfeyou'regoingtobe hungryall the time~that's just silly." Peopleinthe registryget an averageof an hourofphysicalactivity everyday,with someexercisingfor asmuchas90 minutesa day.Theyalsokeepthe fat in theirdiet relativelylow, at about25 percentof theircalorieintake.Nearlyall of them eatbreakfasteveryday,andtheyweigh themselvesregularJy."Theytell ustwo things," Hill says."Thequalityof life is highehllfeis better than it wasbefore.."And "they get tothe pointwith physicalactivitywherethey don'tsay they loveit. but theysay'it'spart of mylife:"

Hilladmitsthat fitting anhour ormoreof exer-ciseinto the dayisdifficult, which iswhy healsofocuseson prevention.Manyofthese peoplemightneverhavebecomeobeseinitially if they hadexer-ciseda mere15to 20minutesa day."Ithink youpaya pricefor havingbeenobese."hestates, "andyouhaveto do alot of activity to makeupfor that." .PaulRaeburn writes about science. Dolicvand

NOT MILK?

Surprisingly, some populationsthat eat few calcium-rich milkproducts appear in some descrip-tive studies to have lower ratesof hip fractures than others thatconsume large quantities ofdairy foods, despite the fact thatdiets of the former group con-tain far less calcium than expertsrecommend. This observation

has not been fully explained.

Calcium intake andhip fractures

Incidenceof hip fracturesper 100,000 people

120 ..""."."..."""."""" " """ "".""..

. United States. NewZealand

80 "..."" """"".""...".".!"~.~.~~~~"".,,

. Jerusalem. UnitedKingdom

G Holland. Finland

40 ."".""".,,",,... Yugoslavia..."""""...". HongKong

. Singapore

0..., _!- !.. 1

and alternatives such as chicken, fish or beans-

steps consistent with today's basic advice aboutwhat to eat.

Fish and Heart DiseaseFatty fish are the most important sources oflong-chain omega-3 fatty acids. In the early1970s Danish investigators observed surpris-

ingly low frequencies of heart disease amongindigenous populations in Greenland that typi-cally ate fatty fish, seals and whales. Theresearchers attributed the protective effect tothe foods' content of omega-3 fatty acids. Somesubsequent studies-but by no means all-con-firm this idea.

Because large, fatty fish are likely to have ac-cumulated methylmercury and other toxinsthrough predation, however, eating them raisesquestions about the balance between benefitsand risks. Understandably, the fish industry iseager to prove that the health benefits of omega-3s outweigh any risks from eating fish.

Even independent studies on omega-3 fatscan be interpreted differently. In 2004 the Na-tional Oceanic and Atmospheric Administra-tion-for fish, the agency equivalent to theUSDA-asked the Institute of Medicine (10M)to review studies of the benefits and risks of

consuming seafood. The ensuing review of theresearch on heart disease risk illustrates the

challenge such work poses for interpretation.The 10M's October 2006 report concluded

that eating seafood reduces the risk of heart dis-ease but judged the studies too inconsistent todecide if omega-3 fats were responsible. In con-trast, investigators from the Harvard School ofPublic Health published a much more positivereport in the Journal of the American Medical

50 SCIENTIFICAMERICAN

DESIGNERSUPERMARKETS

Marketing experts design nearly

every feature of food stores-from product placement to moodmusic-to maximize sales.

When customers enter a grocery store,

the first thing they see is typically

something colorful, aromatic and

enticing-fresh produce, for example.

The long center aisles and aisle-end

displays are jam-packed with prod-

ucts, forcing shoppers to pass by

many items that they might purchase

on impulse.

Food companies pay supermarkets to

get their products-salty chips and

other junk foods-positioned promi-

nently in huge displays.

Checkout lines are plastered with can-

dy and other junk food items-the

last temptation.

OMEGA-3FATTYACIDS,which are thought bysomeresearchers to protect against heart disease, arefound in oily fish such as salmon and trout.

Association that same month. Even modest con-

sumption of fish omega-3s, they stated, wouldcut coronary deaths by 36 percent and totalmortality by 17 percent, meaning that not eat-ing fish would constitute a health risk.

Differences in interpretation explain how

distinguished scientists could arrive at such dif-ferent conclusions after considering the samestudies. The two groups, for example, had con-flicting views of earlier work published inMarch 2006 in the British Medical Journal.

That study found no overall effect of omega-3son heart disease risk or mortality, although asubset of the original studies displayed a 14 per-cent reduction in total mortality that did notreach statistical significance. The 10M team in-terpreted the "nonsignificant" result as evi-dence for the need for caution, whereas the Har-

vard group saw the data as consistent with stud-ies reporting the benefits of omega-3s. Whenstudies present inconsistent results, both inter-pretations are plausible. I favor caution in suchsituations, but not everyone agrees.

Because findings are inconsistent, so is di-etary advice about eating fish. The AmericanHeart Association recommends that adults eat

fatty fish at least twice a week, but U.S. dietaryguidelines say: "Limited evidence suggests anassociation between consumption of fatty acidsin fish and reduced risks of mortality from car-diovascular disease for the general population... however, more research is needed." Whetheror not fish uniquely protects against heart dis-ease, seafood is a delicious source of many nu-trients, and two small servings per week of theless predatory classes of fish are unlikely tocause harm.

Sodas and ObesitySugars and corn sweeteners account for a largefraction of the calories in many supermarketfoods, and virtually all the calories in drinks-soft, sports and juice-come from added sugars.

In a trend that correlates closely with risingrates of obesity, daily per capita consumption ofsweetened beverages has grown by about 200calories since the early 1980s. Although com-mon sense suggests that this increase mighthave something to do with weight gain, bever-age makers argue that studies cannot prove thatsugary drinks alone- independent of calories orother foods in the diet-boost the risk of obesi-

~a:2«2«~«'"2«

~

September 2007

ty. The evidence, they say correctly, is circum-stantial. But pediatricians often see obese chil-dren in their practices who consume more than1,000 calories a day from sweetened drinksalone, and several studies indicate that children

who habitually consume sugary beverages takein more calories and weigh more than thosewho do not.

Nevertheless, the effects of sweetened drinks

on obesity continue to be subject to interpreta-tion. In 2006, for example, a systematic reviewfunded by independent sources found sweet-ened drinks to promote obesity in both chil-dren and adults. But a review that same yearsponsored in part by a beverage trade associa-tion concluded that soft drinks have no specialrole in obesity. The industry-funded research-ers criticized existing studies as being short-term and inconclusive, and pointed to studiesfinding that people lose weight when they sub-stitute sweetened drinks for their usual meals.

These differences imply the need to scruti-nize food industry sponsorship of research it-self. Although many researchers are offended

by suggestions that funding support might af-fect the way they design or interpret studies,systematic analyses say otherwise. In 2007 in-

WHAT TO EAT: Fruits,vegetables and whole grainsconstitute the main components of a healthy diet.

vestigators classified studies of the effects ofsweetened and other beverages on health ac-cording to who had sponsored them. Industry-supported studies were more likely to yield re-sults favorable to the sponsor than those fundedby independent sources. Even though scientistsmay not be able to prove that sweetened drinkscause obesity, it makes sense for anyone inter-ested in losing weight to consume less of them.

The examples I have discussed illustrate whynutrition science seems so controversial. With-

out improved methods to ensure compliancewith dietary regimens, research debates arelikely to rage unabated. Opposing points ofview and the focus of studies and food advertis-

ing on single nutrients rather than on dietarypatterns continue to fuel these disputes. Whilewe wait for investigators to find better ways tostudy nutrition and health, my approach-eatless, move more, eat a largely plant-based diet,and avoid eating too much junk food-makessense and leaves you plenty of opportunity toenjoyyourdinner. .

Noted nutritionist Marion Nestle discussesgenetically modified and irradiated produceand the politics surrounding governmentdietary guidelines for fruits and vegetablesin an excerpt from her book What ToEat.Goto www.SciAm.comjontheweb

~a:z'";2<:'"z<:~

. MORETOEXPLORE

Low-Fat Dietary Pattern and Riskof Invasive Breast Cancer:

The Women's Health Initiative

Randomized Controlled DietaryModification Trial. RossL.Prenticeet al. inJournal of the American Medi-

cal Association, Vol. 295, No.6,

pages 629-642; February 8, 2006.

What to Eat. Marion Nestle.

North Point Press, 2006.

Relationship between FundingSource and Conclusion amongNutrition-Related Scientific

Articles. L. I. Lesser, C. B. Ebbeling,

M. Goozner, D. Wypij and D. S.

Ludwig in PLoS Medicine, Vol. 4,

No.1, article e5, pages 41-46;

January 9, 2007.

Effects of Soft Drink Consump-tion on Nutrition and Health:

A Systematic Review and Meta-

analysis. L.R.Vartanian, M. B.Schwartz and K. D. Brownell inAmeri-

canJournal of Public Health, Vol. 97,

No.4, pages 667-675; April 2007.

Food Politics: How the Food

Industry Influences Nutritionand Health. Revised edition.

Marion Nestle. University of Califor-nia Press, 2007.

SCIENTIFICAMERICAN 51