11/2/2010

1

Eating, Obesity, and Food Choice in Americans and Others

Or Eating in the 21st Century

Developed WorldDeveloped World

Paul Rozin, PhDProfessor of Psychology

University of PennsylvaniaSeptember 20, 2010

Two questions

How much do people eat?What do they eat?

The obesity epidemic

• Since 2000• Not an epidemic• Not contagious• Not contagious• Not a growth curve like an epidemic• For Americans in last 20 years

– About 1.5 pound gain per year

Selling obesity as public health disaster # 1

• Switch from obesity to overweight (BMI>=25)

• Attributing all negatives associated with g gobesity to obesity

• Cost estimates for USA per year vary from from 48,000,000,000 to 1,000,000,000,000

• Biggest effect may be on female self-esteem and quality of life

Body Mass Index (BMI)

• BMI = (weight in kg)/(height in m2)• BMI >18.5, < 25 is “normal• BMI >=25 < 30 is “overweight”• BMI >=25, < 30 is “overweight”• BMI >= 30 is “obese”

http://www.surgeongeneral.gov/news/testimony/obesity07162003.htmOffice of surgeon generalStatement of Richard H. Carmona, M.D., M.P.H., F.A.C.S.Surgeon GeneralU.S. Public Health ServiceWednesday, July 16, 2003y, y ,“The crisis is obesity. It’s the fastest-growing cause of disease and death in America. And it’s completely preventable.

•Nearly two out of every three Americans are overweight or obese. •One out of every eight deaths in America is caused by an illness directly related to overweight and obesity.

But the fact is that we have an epidemic of childhood obesity.

11/2/2010

2

Flegal et al., 2005Main Outcome Measures Number of excess deaths

in 2000 associated with given BMI levels

100

150

S (1

000S

)

-100

-50

0

50

BMI<18.5 BMI >=18.5,<25

BMI >=25, <30 BMI>=30

EX

CES

S D

EATH

S

BMI

• (weight[kg]) /(height[meters])2

• Obesity BMI>=30• The gold standard• The gold standard

Take out your piece of paper and pencil

• Write your discipline or major– (e.g., Nutrition, Psychology, Marketing)

• BMI = (weight[kg]) /(height[meters])2• BMI (weight[kg]) /(height[meters])• Why is HEIGHT SQUARED in the

denominator?

BMI: Why is height squared?Quetelet: 19 century

Hgt 1.3mDiam 1mH/D = 1.3

Assume 1 square meter = 100kg

Hgt 2.0mDiam 1.54mH/D = 1.3

Volume = hgt * (pi*r2)

Vol = .785 m2

Wgt = 102kgBMI = 60.35Bm3 = 46.4

Vol = 3.71 m2

Wgt = 371kgBMI = 92.7Bm3 = 46.4

Adaptations to our ancestral environment that don’t work now

• 1. Eating predisposition

2. The magical law of similarity

• Appearance = reality• The image = the object

11/2/2010

3

• 3. The univariate mind

A

Height4 ft 11 in

3

Height5 ft 9 in

3

11/2/2010

4

• 4. The monotonic mind

A diet totally free of salt is healthier than a diet of the same number of calories that

includes a pinch of salt every dayGroup % Agree

College students 19

National Sample 27

Physical Plant workers 37

Overall 28

Adaptations to our ancestral environment that don’t work now

• 5. The categorical/dichotomous mind

A pint of cottage cheese has more calories than one teaspoon of ice cream.

Group % Disagree

College students 30

National Sample 25

Physical Plant workers 38

Overall 31

6. Optimal foraging

11/2/2010

5

Mismatch: Health information and lay ability to interpret it

• 7. Lack of understanding of the scientific enterpriseenterprise

Medicalh

Foodindustry

Govern- Mediaresearchment

Public

Late 20th Century developed world

1. Epidemiological revolution: longer life and death from degenerative diseases

2. Food surplus3. Development of super-foods (hi sugar, hi fat)4. Extraordinary variety5. No work needed to attain choices6. Massive amounts of risk information7. No training in dealing with risks/benefits

The combination of health and beauty normsy

“Concerned about being overweight”

• % responding “often“ or “almost always”• 57% females 21% males• 57% females, 21% males• US college students from 6 universities

across the country

Rozin, Bauer & Catanese, 2003

11/2/2010

6

“I am embarrassed to buy a chocolate bar in the store”

• American college students from six campuses across the USAp

• % Females: 13.5

• % Males: 4

Weighing sensitivityMean (s.d) (Discomfort 0-100)

FemaleN=285-287

MaleN = 183

Alone 19.4 (30.0) 12.5 (29.9)( ) ( )

Male acquaint

59.8 (35.0) 21.1 (31.1)

FemaleAcquaint

53.7 (33.1) 26.5 (31.8)

Geier and Rozin

What is your current figure?____What is your ideal figure?_____What is the ideal female figure selected by

American women? ______

slide body Image USA StudentsLate 1980s

A CI OA2004

General female ideal: pink = female estimate, blue = male

Late 1980s

2004C I=COA

Little change from 1980s

A

11/2/2010

7

France versus USA

• Claude Fischler• Rebecca Bauer, Dana Catanese, Kim

Kabnick Estelle Masson Erin PeteKabnick, Estelle Masson, Erin Pete,• Alison Sarubin, Christy Shields, • Amy Wrzesniewski

Life expectancy at birthUN Demographic Yearbook (1993

u country years rank country years1 Japan 79.2 9/10 Israel,Italy 76.82 Sweden 78.1 11 Canada 76.43 Switzerland 77.8 12 U. K. 76.24 Australia 77.4 13 Austria 76.15/6 Norway,

Netherlands77.2 14 Belgium 75.8

7 France 77.0 15 USA 75.48 Spain 76.9 16 WGermany 75.1

Overweight:France vs USA (late 1990s)

• % BMI >= 25• % BMI >= 25• France: 39%• USA: 61%

Age-standardized annual mortality from CHD and related risk factors

(males 35-64)WHO/MONICA Renaud & de Logeril, 1992

Location Mortality / 100,000

Serum chol-esterol (mg/dl)( g )

Toulouse, France 78 230

Lille, France 105 252

Stanford, USA 182 209

Attitudes to food and eatingPercent of subjects preferring luxury

hotel to gourmet hotel at the same price

Female students Male students

France 13% 8%

USA 83% 71%

11/2/2010

8

Percent of subjects saying “unhealthy” for choice:

Heavy cream: whipped or unhealthy

Female students Male students

France 28% 23%

USA 67% 48%

Percent of subjects agreeing that they eat a “healthy diet”

Females Males

France 76% 72%

USA 28% 38%

Factors in four country study• Concern about healthiness of food• Worry about weight/fat• Pleasure/Importance• Culinary associations• Fat/salt reduced diet• Healthy eater

Rozin, P., Fischler, C., Imada, S., Sarubin, A., & Wrzesniewski, A. (1999). Attitudes to food and the role offood in life: Comparisons of Flemish Belgium, France, Japanand the United States. Appetite, 33, 163-180.

Chernoff Faces

Metaphor: Food and the body are like:

USA France

Tree 26 66

Car or factory 43 26

Temple 32 10

Representative national samplesFischler, Rozin et al., 2004

11/2/2010

9

The food environment

Restaurant portion sizeRestaurant France USA

McDonald’s (7) 189g 256g

Quick/Bking(5) 207g 322g Quick/Bking(5) 207g 322g

Chinese (6) 244g 418g

Rozin, P., Kabnick, K., Pete, E., Fischler, C., & Shields, C. (2003). The ecology of eating: Part of the French paradox results from lower food intake in French than Americans, because of smaller portion sizes. Psychological Science, 14, 450-454.

Supermarket food portions

ITEM Carrefour Acme

Yogurt (modal) 125g 227g g ( ) g g

Fresh fruit (mean,4 types)

431g 553g

Coca cola (modal)

330ml 500ml

Supermarket non-food portions

ITEM Carre-four

Acme

toothpaste 75 170 toothpaste (modal, ml)

75 170

toilet paper (mean, sq cm)

121 117

Cat food (modal, g)

100 85

Social norms and eating

11/2/2010

10

Unit Bias

• Norm for eating one entity• M&Ms free in bowl• Small spoon or 4X spoonSmall spoon or 4X spoon• 70% more consumed with 4X spoon• 60% more with double vs single pretzels

• Geier, A. B., Rozin, P., & Doros, G. (2006). Unit bias: A new heuristic that helps explain the effect of portion size on food intake. Psychological Science, 17, 521-525.

EnergyIntake

EnergyExpenditure

EnergyStorage

Energy

EnergyExpenditure

Intake

EnergyStorage

car

The French Garage

street

11/2/2010

11

MANHATTAN

NEW YORKSUBURB

PARIS

French vs American differences(with Abigail Remick & Claude Fischler)

1. Quality vs quantity2. Moderation vs abundance3 Collective values vs individualization3. Collective values vs individualization4. Joys vs comforts5. Food more associated with conviviality6. Environment limits modest amounts of

food to mealtimes and smaller portions

Obesity as individual responsibility

• Dieting looks easy: less food, lower weight• It rarely works for the long term

Incremental changes

• One apple day less, all else equal: loss of 8 pounds

• About one regular coke switched to dietAbout one regular coke switched to diet coke a WEEK: eliminate obesity epidemic

• ALL ELSE EQUAL

Working on the environment

Summing small differences ideaKelly Brownell, James Hill, David Levitsky, Marion Nestle, Barbara

Rolls, Andrew Geier and Paul Rozin

11/2/2010

12

Working on the environment

Cafeteria StudiesWith Sydney Scott, Megan Dingley,

Andrew Geier, Jenn Rineer, Ian Schwartz, Hong Jiang, Kalina

Urbanek, Mark Kaltenbach.Also work by Barbara Rolls

12 vs 16 oz plastic cups2 soda machines, no charge

4 beverages with highest intake

1

1.2

1.4

rson

0

0.2

0.4

0.6

0.8

Diet Pepsi Pepsi Fruit Punch Ginger Ale

Oun

ces/

per

12 oz16 oz

Net .48 oz less in the meal, or for one such meal a day, annualWeight loss of 0.6 pounds

Soup

Sushi

Sandwiches/Coffee

Exit

Salad Bar

Entrance

Salad Bar ArrangementsABRAMSON’S SALAD BAR (Chicken in the middle: Mon, Wed, Fri)

GG

CKE

N

TOFU

SSI

NG

BR

EAD

C

RO

UTO

NS

WR

S N

DRY

RE

D

PEPP

ER

BREA

D

CRU

MB

S

BA

NZO

BE

ANS

OLI

VES

KID

NEY

BE

ENS

CH

OK

E

O P

EP

PE

R

A F

ISH

AR C

UBE

S

CA

RR

OT

LETT

UC

E LE

TTU

CE

MIX

OU

TS

OU

TS

SE

AFO

OD

SLD

BLA

CK

OLI

VE

S

PEPP

ER

CA

RR

OT

MU

SH

RO

OM

MU

SH

RO

OM

SO

Y B

EA

NS

ION

S

PEPP

ER

OR

N

KID

NEY

BE

ENS

EG

CH

IC

DR

ES

FLA

X

SEED

S

PUM

P-K

IN

SEED

S

SUN

-FL

OW

SEED

S

CR

AN

B

ER-

RIE

S

RED

BE

ETS

RED

BE

ETS

TAN

GER

INES

CU

CU

MBE

RS

CU

CU

MBE

RS

AR

TIC

JALA

PE

N

TUN

A

BLA

CK

BEA

NS

FETA

CH

EE

SE

PRO

VOLO

NE

CH

ED

DA

OLI

VES

GR

EEN

PEP

PER

CA

ULI

FLO

WER

SPIN

ACH

LETT

UC

E

SP

RO

SP

RO

OLI

VE

OIL

/ V

INE-

GA

R

PO

TATO

E S

LD

GR

EEN

PEP

PER

PAST

A SL

D

RED

P

ON

I

RED

P

CO

TO

MAT

OE

S

TO

MAT

OE

S

BLU

E C

HE

ES

E

SHR

ED

. CH

EDD

AR

BR

OC

OLL

I

BR

OC

CO

LI

Access by ease of serving

11/2/2010

13

Schematics: Spoons/Tongs, Manipulation A

Dressing and T

Oranges Artichoke

Dressing and Condiments Lettuces

• Isolates edge v. middle

Tomato

Cheddar

Schematics: Spoons/Tongs, Manipulation B

Dressing and T

Oranges Artichoke

Dressing and Condiments Lettuces

• Isolates edge v. middle

Tomato

Cheddar

Utensil Study Results (% of Total Intake)

10%

12%

14%

16%

ood

In

take

SpoonTongs

0%

2%

4%

6%

8%

Artichoke Oranges Cheddar TomatosIngredient

% o

f Tota

l Fo

Net difference of 5.53 g/day, or 1,476 g/year or 0.43 pounds/yr

Salad Bar Position. Study 3. Middle versus Single Edge

Dressing and

C

Dressing and Condiments GreensB

• Isolates edge v. middle

A

Edge v. Middle

0.0150

0.0200

0.0250

0.0300

0.0350

0.0400

EdgeMiddle

0.0000

0.0050

0.0100

Chick

en Egg

Salm

onTu

na

Tomatoe

s

Carro

ts

Broc

coli

Mushr

ooms

IngredientNet difference of 5.6 g/day, or 1,899 g/year or 0.54 pounds/yr

• Accessibility in terms of height: eye or floor level. Alternated once a week

11/2/2010

14

Soda Cabinets, Manipulation A

Sobe Lifewater

Pepsi

Aquafina Flavored Water

Mountain Dew

Soda Cabinets, Manipulation B

Aquafina Flavored WaterMountain Dew

Sobe Lifewater

Pepsi

Soda Purchases by Display Level

6.00

8.00

10.00

12.00

ag

e #

Ite

ms

ase

d P

er

Da

y

Eye-LevelBottom

0.00

2.00

4.00

Pepsi Mtn Dew FlavoredAq

Sobe

Soda

Av

era

Pu

rch

a

Average reduction of 19.1% when soda is on bottom level

Segmentation cues or consumption interruptsp p

Geier, Wansink & Rozin

Lay’s stacked potato chipsAd lib potato chips for each student while watching a 35

minute movie

• Controls: All Lay’s Original• Lo Segment: All Lay’s original except• Lo Segment: All Lay s original except

every 10th chip is Red Basil/tomato chip• Hi Segment: All Lay’s original except every

5th chip is Red Basil/tomato chip

11/2/2010

15

Mean # chips eaten as a function of segmentation cues

P<.001 in both studies.: Study 1 at Cornell, Study 2 at Penn

Compensation problemsAll else equal: ceteris parebis

1. Increase calories at subsequent meals– Via regulation of weight system– Via personal food rewards for weight loss orVia personal food rewards for weight loss or

healthy eating2. Metabolic adaptation3. Increase in energy output

David Levitsky

• One of few studies controlling for compensation

• 18 young adults with ad-lib feeding except18 young adults, with ad lib feeding except for daily (weekday) lunch restriction to 250 calories

• All intake monitored and weighed

Figure 7

-0.2-0.10.00.10.20.30.40.50.60.7

Meal Replacement

ght C

hang

e (k

g)

Normal Lunch

-1.2-1.1-1.0-0.9-0.8-0.7-0.6-0.5-0.4-0.3

Wei

g

Levitsky, 2010. Weight change over 2 weeks by replacing anormal lunch (500-600 kcal) with a lunch of 250 kcal for 10 week days

Doctors and lay people:The power of culture

• 50-70 doctors and 50-70 lay people from each of five countries

• France, Germany, Italy, U.K., U.S.A• 20 items relating diet and eating to health

– Value of vitamin pills– Healthiness of dairy products, wine, meat– Importance of food, exercise, moderation for health– (Leeman, Fischler & Rozin, 2006)

11/2/2010

16

Sample of correlation technique

0 5

1

1.5

2

2.5

e A

-2

-1.5

-1

-0.5

0

0.5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

z sc

ore A

BC

Lay-Doctor similarities on food-health (Pearson rs across 20 variables)

• France doctor with – France Lay .45– US doctor -.53US doctor .53

• US doctor with– US Lay .48

Lay-Doctor similarities on food-health (Pearson rs across 20 variables)

• France doctor with – France Lay .45– German doctor -.38German doctor .38

• German doctor with– German Lay .69

-0 1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Cor

rela

tion

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

Mea

n C

Physicians--Lay Same Country

Physicians--Physicans Different

Countries

Physicians--Lay Dif ferent Countries

Lay--Lay Different Countries

Categories of Comparison

END