Family Economics and Nutrition Review [Volume 10, Number 3]

              Feature Articles
2 The U.S. Food Supply Series, 1970 to 1994:
Nutrient Availability and Policy Implications
Shirley Gerrior and Lisa Bente
20 Nutrient Intakes and Eating Patterns of Teenagers
Elyse Levine and Joanne F. Guthrie
36 Lower Fat Food Choices Identified Through a Novel
Sorting Procedure for CSFII 1989-91 Data
Sharon Peterson, Madeleine Sigman-Grant, and Cheryl Achterberg
Research Brief
SO Role of Breakfast in the American Diet by Income Group
Shanthy A. Bowman
Research Summaries
55 Social Security and Earnings of Male Immigrants
59 An Experimental Consumer Price Index for the Poor
Regular Items
62 Charts From Federal Data Sources
64 Recent Legislation Affecting Families
65 Journal Abstracts
66 Research and Evaluation Activities in USDA
SERIALS DEPARTMENT
68 Estimated Annual Expenditures on Children by Families, 1996
75 Poverty Thresholds
76 Cost of Food at Home
77 Consumer Prices
UNITED STATES DEPARTMENT OF AGRICULTURE
Volume 10, Number 3
1997
Guidelines for Authors
Family Economics and Nutrition Review is a peer-reviewed quarterly journal
published by the Center for Nutrition Policy and Promotion; Food, Nutrition, and
Consumer Services; U.S. Department of Agriculture.
Family Economics and Nutrition Review follows the guidelines of the Publication
Manual of the American Psychological Association, Fourth Edition.
I. Manuscripts may not be submitted elsewhere while under consideration by the
Review.
2. To ensure anonymity, include a separate title page with author's full name, title,
affiliation, full address, and telephone number. There should be no reference to
the authors in the text or footnotes.
3. Manuscripts should be double-spaced, on 8-1/2" x 11" paper, have 1-1/4" margins,
and use I 0 or 12 pitch type size.
4. Manuscripts should include an abstract of no more than 150 words.
5. Abstract, text, references, tables, and figures should not exceed 30 pages.
6. Each table or figure should be placed on a separate page.
7. Upon acceptance, authors must send a 3-112" computer disk copy of the final
accepted version of the manuscript in WordPerfect 7.0 or lower. Disks and
manuscripts will not be returned.
8. All authors and coauthors must sign a release acknowledging that contents of
the Review are in the public domain.
9. Please send four copies to:
Joan C. Courtless, Editor
Family Economics and Nutrition Review
USDA, Center for Nutrition Policy and Promotion
1120 20th Street, NW, Suite 200 North Lobby
Washington, DC 20036
(202) 606-4816
Editor
Joan C. Courtless
Managing Editor
Jane W. Fleming
Family Economics and Nutrition Review is
written and published each quarter by the
Center for Nutrition Policy and Promotion,
U.S. Department of Agriculture, Washington, DC.
The Secretary of Agriculture has determined
that publication of this periodical is neces­sary
in the transaction of the public busiless
required by law of the Department.
This publication is not copyrighted. Contents
may be reprinted without permission, but
credit to Family Economics and Nutrition
Review would be appreciated. Use of
commercial or trade names does not imply
approval or constitute endorsement by USDA.
Family Economics and Nutrition Review is
indexed in the following databases: AGRICOLA,
Ageline, Economic Literature Index, ERIC,
Family Studies, PAIS, and Sociological
Abstracts.
Family Economics and Nutrition Review is
for sale by the Superinten:lent of Documents.
Subscription price is $10.00 per year ($12.50
for foreign addresses). Send subscription
orders and change of address to Superinten­dent
of Documents, P.O. Box 371954,
Pittsburgh, PA 15250-7954. (See subscription
form on p. 80.)
Original manuscripts are accepted for
publication. (See "guidelines for authorS' on
back inside cover). Suggestions or comments
concerTing this publication should be
addressed to: Joan C. Courtless, Editor,
Family Economics and Nutrition Review,
Center for Nutrition Policy and Prorrdion, USDA,
112020th St., NW, Suite 200 North Lobby,
Washin~on. DC 20036. Phone (202) 606-4816.
USDA prohibits discrimination in its programs
on the basis of race, color, national origin,
sex, religion, age, disability, political beliefs,
and marital or familial status. (Not all prohibited
bases apply to all programs.) Persons with
disabilities who require alternative means for
communication of program information (Braille,
large print, audiotape, etc.) should contact
USDA's TARGET Center at (202) 720-2600
(voice and TOO). To file a complaint, write
the Secretary of Agricutture,U.S. Department
of Agricuture, Washington, DC 20250, or call
1 ~-245-6340 (voice) or (202) 720.1127
(TOO). USDA is an equal employment
opportunity employer.
Center for Nutrition Policy and Promotion
PROPERTY OF TiiE
LIBRARY
DEC 3 11S~7
University of Nor h Carolina
at Greensburo
Feature Articles
2
20
36
The U.S. Food Supply Series, 1970 to 1994:
Nutrient Availability and Policy Implications
Shirley Gerrior and Lisa Bente
Nutrient Intakes and Eating Patterns of Teenagers
Elyse Levine and Joanne F. Guthrie
Lower Fat Food Choices Identified Through a Novel
Sorting Procedure for CSFII 1989-91 Data
Sharon Peterson, Madeleine Sigman-Grant, and Cheryl Achterberg
Research Brief
50 Role of Breakfast in the American Diet by Income Group
Shanthy A. Bowman
Research Summaries
55 Social Security and Earnings of Male Immigrants
59 An Experimental Consumer Price Index for the Poor
Regular Items
62 Charts From Federal Data Sources
64
65
66
68
75
76
77
Recent Legislation Affecting Families
Journal Abstracts
Research and Evaluation Activities in USDA
Estimated Annual Expenditures on Children by Families, 1996
Poverty Thresholds
Cost of Food at Home
Consumer Prices
Volume I 0, Number 3
1997
2
Feature Articles
The U.S. Food Supply Series,
1970 to 1994:
Nutrient A vai labi I ity
and Policy Implications
Shirley Gerrior
Nutritionist
Center for Nutrition Policy and Promotion
Lisa Bente
Nutritionist
Center for Nutrition Policy and Promotion
The U.S. food supply series measures the amount of nutrients available per
capita per day. This paper reviews trends in food supply consumption and
nutrient levels for 1970 to 1994. Substantial change in the U.S. food supply
during this period led to per capita availability increases tor most nutrients,
except tor saturated tatty acids, cholesterol, and vitamin B12. Food and nutrient
level trends are examined in terms of Federal dietary guidance, nutrition
monitoring activities, and fortification policy. Future research directions are
suggested to improve the quality of food supply estimates in terms of these
activities.
ietary guidance to improve
the public health and well­being
of Americans has
been in place since early
this century (25)-a period similar to
that covered by the U.S. food supply
series. The translation of this dietary
guidance into recommendations for a
healthful diet and the successful imple­mentation
of these recommendations
require that the consumer has greater
access to affordable health-promoting
foods. Thus, the quantity and nutrient
composition of the foods available for
consumption are vital to increasing the
prevalence of healthy eating. Techno­logical
alterations, the designing of
foods, and enrichment and fortification
policy during this century have increased
the variety of foods in the food supply
as well as enhanced the health benefits
associated with these foods. For example,
the removal of nutrients or dietary com­ponents
from foods has increased the
variety of lower fat dairy and leaner meat
products in the food supply, while the
addition of nutrients to food through
enrichment and fortification has improved
the nutritional quality of the food supply.
This article discusses trends in U.S.
food and nutrient supplies since 1970,
compares data for 1970 and 1994, and
considers the significant implications of
Family Economics and Nutrition Review
these data for food and nutrition policy.
The U.S. food supply series measures
the amount of food available for con­sumption
per capita per year and the
amounts of nutrients per capita per day
( 5, I 4 ). It is the only continuous source
of data on food and nutrient availability
in the United States. Food supply nutri­ent
estimates, extended back to 1909,
were calculated for the first time during
World War II to assess the nutritive
value of the food supply for civilian
use in the United States and to provide a
basis for international comparisons with
the food supplies of our allies.
Per capita food supply estimates provide
unique and essential information on the
amount of food and nutrients available
for consumption. They are useful to
assess trends in food and nutrient con­sumption
over time, for monitoring the
potential of the food supply to meet the
nutritional needs of Americans, and for
examining relationships between food
availability and diet-health risk. There­fore,
the U.S. food supply series is one
of the five major components of the
National Nutrition Monitoring and
Related Research Program (NNMRRP),
mandated by the National Nutrition and
Related Research Program Act of 1990
(4). In addition, because the food supply
series has measured food and nutrients
over time using the same conceptual
methodology, it is useful for evaluating
the effects of technological alterations,
advertising and promotion efforts,
changes in marketing practices, and
nutrition education policies. It also helps
agricultural policymakers translate
nutrient recommendations into goals
for food production and supply levels
to ensure that adequate nutrients are
available to Americans (see box,
"About the Data").
1997 Vol.J0No.3
Food Consumption Trends
Are Driving Major Changes
in Nutrient Availability
During 1970-94, substantial change in
the quantity and mix of foods in the U.S.
food supply produced marked change
in nutrient availability. Much of the
change is due to advances in technology
and alterations in marketing practices,
as well as to Federal dietary guidance 1 to
make healthy food choices that promote
health and prevent disease. The intro­duction
of many foods into the food
supply since 1970 reflects industry's
response to this dietary guidance and
consumer demand for a variety of low­fat,
flavorful, and nutritious food choices.
Production techniques and marketing
changes over the last two decades have
been particularly responsive to and
reflective of dietary recommendations
for fat, saturated fat, and cholesterol.
The creation of newly formulated foods
or the modification of foods or ingredi­ents
used in foods has enhanced the
health benefits of the food supply.
1Current Federal dietary guidance as used in this
paper considers the following: Recommended
Dietary Allowances (tenth edition) ( 11 ), Nutrition
and Your Health: Dietary Guidelines for Americans
(4th ed.) (20), The Food Guide Pyramid (18), and
the Diet and Health 1mplicationsfor Redt~eing
Chronic Disease Risk ( 10). lbese serve as the basis
for nutrition policies of the Federal Government.
Earlier Federal dietary guidance, which influenced
production and consumption practices in the late
1970's and 1980' s, includes Dietary Goals for
the United States ( 1977) (24), Healthy People: The
Surgeon General's Report on Health Promotion
and Disease Prevention (1979) (2 1 ), Nutrition and
Your Health: Dietary Guidelines for Americans
(1980) ( 19), and Diet, Nutrition and Cancer
(1982)(9).
Per capita food
supply estimates
provide unique and
essential information
on the amount of
food and nutrients
available for
consumption.
3
About the Data
USDA's Center for Nutrition Policy and Promotion uses data
on the amount of food available for consumption from USDA's
Economic Research Service (ERS) and information on the nutrient
composition of foods from USDA's Agricultural Research Service
(ARS)1 to calculate the nutrients available in the U.S. food supply.
Nutrients reported include food-energy, energy-yielding compo­nents
(carbohydrate, protein, and fat), cholesterol, 10 vitamins,
and 7 minerals.
Food supply data, collected and published annually by ERS,
estimate the amount of food available for consumption in the
United States by measuring commodity flows from production
to end uses. Food available for consumption is calculated as the
difference between available commodity supplies (the sum of
production, imports, and beginning-of-the-year inventories) and the
sum of exports, year-end inventories, and nonfood uses. Foods are
measured as primary commodities before they are combined with
other foods or processed into final products seen in the market ·
place.
The food supply consumption estimates reflect amounts of food
available prior to moving through marketing channels-not the
amounts actually consumed. Therefore, the supplies are greater
than what individuals ingest due to losses in trimming, cooking,
plate waste, and spoilage that are not accounted for in the esti­mates.
For example, the food supply estimates overstate human
consumption of fats and oils, since large amounts are used for fry­ing
by fastfood establishments and later discarded as waste. Also,
increasing proportions of the total turkey supply go into pet foods,
but such use is included in per capita turkey estimates and thus
overstate turkey consumption ( 14 ).
The food supply nutrient estimates are calculated by multiplying the
per capita amount of each food by the nutrient composition of
that food. The results from approximately 400 foods are then totaled
for each nutrient and presented on a per day basis. As with the food
supply estimates, the nutrient estimates presented here do not
account for losses during processing, marketing, or home use and
are better indicators of trends in consumption rather than actual
amounts ingested. For example, vegetables generally lose nutri­ents
when cooked in water, particularly water-soluble nutrients like
vitamin C and thiamin.
Nutrients not included in these values are those from phosphorus
contained in carbonated soft drinks, vitamin and mineral
supplements, alcoholic bevemges (or the grains and sugar used
to make alcoholic beverages), baking powder, yeast, and certain
vitamins and minerals used for functional or flavoring agents in
foods. Nutrients added to foods commercially through enrich­ment
and fortification are included in the nutrient values.
4 ·
Food supply data differ from dietary survey data: food supply data
measure food and nutrient availability as national totals whereas
dietary survey data (such as USDA's Continuing Survey of Food
Intakes by Individuals) provide data on food and nutrient intakes
reported by individuals and households. Dietary or food intake
surveys record food intake data over a specific time period and
combine it with demographic information. These data are used to
assess food consumption behavior and the nutritional content of
diets for policy implications relating to food production and
marketing, food safety, food assistance, and nutrition education.
Both the food supply and food intake data are major components
of the NNMRRP, a set of related Federal activities intended to
provide regular information on the nutritional status of the U.S.
population. However, both have strengths and limitations that
affect their ability to measure food consumption and their useful­ness
in dietary assessment. For example, food supply estimates reflect
the amount available prior to moving through marketing channels,
not the amount actually consumed. Thus, supplies are greater than
what individuals ingest due to losses in processing, marketing,
home use, and spoilage that are not accounted for in the estimates.
Likewise, nutrient estimates determined for food supplies do not
account for these losses and need to exceed those recommended for
good health by a generous amount to account for such losses and
to ensure adequate nutrients are available to the U.S. population.
Also, these levels represent averages for the entire population.
As a result, food supply data typically overestimate food and nutrient
availability and are better indicators of trends in consumption over
time rather than actual amounts ingested.
On the other hand, the quality of dietary or food intake surveys
depends on the appropriateness of the methodology used for data
collection and on the accuracy and completeness of the recall or
recording of the individual. Underreporting of total diet or different
food groups is common with these surveys. Because of such limita­tions,
food intake surveys may underrepresent actual food intakes.
Due to the current limitations of these two types of data, their
joint application is best directed toward trend analysis of food
and nutrient consumption. The U.S. food supply series serves this
purpose well since the conceptual basis for measuring food and
nutrient estimates has remained the same over the series. However,
apparent trends from food intake surveys may be misleading due
to changes in survey design, sampling strategy, and interview
methodology from one survey to another.
1Nutrient data used for the most recent update are from the Primary Nutrient
Data Set developed for use with the 1994 Continuing Survey of Food Intakes
by lndividuals(CSFli) (17).
Family Economics and Nutrition Review
Table 1. The U.S. food supply nutrient per capita per day, 1970 and 1994
Nutrient (unit) 1970 1994 Percent change
Food energy (Kcal) 3,300 3,800 15
Carbohydrate (g) 386 491 27
Protein (g) 95 110 16
Total fat (g) 154 159 3
Saturated fatty acids (g) 54 52 -4
Monounsaturated fatty acids (g) 63 65 2
Polyunsaturated fatty acids (g) 26 31 19
Cholesterol (mg) 470 410 -13
Vitamin A (RE) 1,500 1,520 I
Carotenes (RE) 510 660 29
Vitamin E (mg) 13.7 16.9 23
Vitamin C (mg) 107 124 16
Thiamin (mg) 2.0 2.7 35
Riboflavin (mg) 2.3 2.6 13
Niacin (mg) 22 29 32
Vitamin B6 (mg) 2.0 2.3 15
Folate (j.l.g) 279 331 19
Vitamin B 12 (j.l.g) 9.5 8.1 -15
Calcium (mg) 890 960 8
Phosphorus (mg) 1,460 1,680 15
Magnesium (mg) 320 380 16
Iron (mg) 15.4 21.2 38
Zinc (mg) 12.2 13.2
Copper (mg) 1.6 1.9 19
Potassium (mg) 3,510 3,780 8
1997 Vol.JO No.3
The food supply provides a wide variety
of healthy food choices, but compliance
with dietary recommendations is often
slow and not easily achieved by the
general population. For example, although
Americans have made some positive
dietary changes in terms of consumption
of grain products, vegetables, and fruits
and the use of lower fat animal foods
from the dairy and meat groups, they
are doing less well with overall con­sumption
of sugars and sweeteners and
total fat. This finding agrees with recent
studies that found, on average, most
Americans are not meeting recommended
servings for most of the Food Guide
Pyramid's five major food groups,
particularly whole grains, fiber-rich
fruit, and dark-green, deep-yellow
vegetables, while consuming excess
calories from fats and sugars ( 1,7) from
foods such as cheese, salad and cooking
oils and shortening, and regular carbon­ated
soft drinks, respectively.
Availability of Nutrients,
1970-94
Per capita nutrient estimates show
higher levels for most nutrients in 1994
than in 1970, except for saturated fatty
acids, cholesterol, and vitamin B12
(table 1). The level of food energy
reached an all-time high of 3,800 calories
per capita per day in 1994, in part due
to a higher level of carbohydrate. The
considerable increase in carbohydrate
reflects the increased use of both grain
products and total sugar and sweeteners
(specifically, corn syrup sweeteners)
(fig. 1, p. 6). During the late 1980's,
grains replaced sugars and sweeteners
as the major source of carbohydrate. By
1994, grains provided 41 percent of the
total carbohydrate in the food supply
(fig. 2, p. 6).
5
Between 1970 and 1994, the percent
kilocalories from complex carbohydrate
in the total food supply increased from
22 to 26 percent, and that from sugar
decreased slightly from 25 to 24 percent.
While these trends are consistent with
dietary guidance to choose a diet with
plenty of grain products and high in
complex carbohydrates, Americans still
need to work harder to consume less
foods high in sugar. In 1994, Americans
consumed a record high amount of
caloric sweeteners. Much of this carne
from the increased consumption of
carbonated soft drinks. Between 1970
and 1994, per capita consumption of
these regular drinks increased from 22
gallons to 40 gallons, while that from
diet drinks increased from 2 gallons to
12 gallons per capita.
As with carbohydrate, the levels of
several vitamins and minerals rose
because of the increased consumption
of grain products. For example, levels
of thiamin, riboflavin, and niacin,2
which rose due to the increase in the
proportion of enriched flour since 1970
( 15 ), are higher still in the 1990's because
of increased grain consumption in more
recent years (table 1). In 1994, riboflavin
contributions from grain products were
similar to those from dairy products,
the leading source of riboflavin in 1970.
Likewise, niacin contributions from grain
products in 1994 were similar to those
from meat, poultry, and fish products,
the leading source of niacin in 1970.
Since 1970, the higher levels of vitamin
B6, folate, phosphorus, magnesium,
iron, zinc, copper, and potassium are,
for the most part, related to increases
2Food composition data give only the amount of
preformed niacin in food. Thus, per capita niacin
estimates for the food supply refer to the availability
of preformed niacin from foods-not that formed
in the metabolism of tryptophan.
6
Figure 1. Trends in grains and sugars in the U.S. food supply,
1970-94
Pounds
per capita
per year
250
200
150
100 Sugar
Total grains
~~
~~ ___ ...
:::::===-~--- · - · - - - - - · - ·- - -· - · - · ·
50
Q LL~~L_L-i_~-L-L~~L_L_i_~_L_L~ _ L_L_~_L_L~~--L_~
1970 74 78 82 86 90
Year
Figure 2. Sources of carbohydrate in the U.S. food supply,
1970 and 1994
Other
Sugars and
sweeteners
39%
1970
Fruits and
vegetables
1994
94
Family Economics and Nutrition Review
Table 2. Percent contribution from grains for selected nutrients in the
U.S. food supply, 1970 and 1994
Nutrient 1970 1994
Carbohydrate 35 41
Thiamin 40 55
Riboflavin 19 31
Niacin 28 40
Vitamin B6 8 13
Folate 13 22
Phosphorus 14 21
Magnesium 18 26
Iron 37 51
Zinc 12 18
Copper 17 23
Potassium 6 10
Figure 3. Trends in fruits and green/yellow vegetables in the
U.S. food supply, 1970-94
Pounds
per capita
per year
160
140
120
100
80
60
40
20
Non-citrus fruit __ ... -······ ·····
Dark-green/deep-yellow vegetables
0 LL~~~L-L-~~~_L_L_L~~~L-~~~~_L_L-L
1970 94 74 78 82 86 90
Year
1997 Vol./0 No.3
in grain consumption. Grain consump­tion
contributed over one-fifth of the
folate, phosphorus, and copper, over
one-fourth of the magnesium, and over
one-half of the iron in the food supply
in 1994 (table 2).
Another trend seen between 1970 and
1994 that is consistent with dietary
guidance is the increase in vegetable
and fruit availability.ln 1994, vegetables
and fruits made important nutrient
contributions to levels of vitamin A,
carotene, vitamin C, folate, and potas­sium
in the food supply. Despite the
fact that the vitamin A level increased
only slightly due to a decrease in egg
and organ meat consumption, larger
nutrient contributions than in previous
years came from the dark-green and
deep-yellow vegetables in the vegetable
group-the largest contributor of vitamin
A to the food supply. Consumption of
dark-green and deep-yellow vegetables
increased from 23 to 28 pounds between
1970 and 1994, contributing to higher
levels of both vitamin A and carotene
in 1994 (fig. 3).
Vitamin C reached a peak in 1994 due
to increases in variety and year-round
availability of fresh citrus fruits. Since
1970, fruits and vegetables have been
responsible for about 90 percent of the
vitamin C in the food supply. Also,
higher levels of folate and potassium
are associated with the trend in increased
fruit consumption (as well as grain
consumption). Although not contributing
to an increase in potassium levels, the
vegetable group continued to be the
primary source of this nutrient from
1970 to 1994.
7
Trends in consumption of the dairy and
meat, fish, poultry, and egg groups in
the food supply are closely associated
with levels of protein, fats, fatty acids,
and cholesterol and to levels of selected
nutrients, such as calcium and phosphorus
in dairy foods, and iron, zinc, and vitamin
B12 in meat and egg products. The
increase in protein from 1970 to 1994
is due mostly to higher consumption
of poultry, and, to a lesser extent, grain
products, cheese, and lowfat milks (figs.
1, 4, and 5).
There was an overall shift in sources
of fat from animal to vegetable (fig. 6)
from 1970 to 1994. This was primarily
caused by the increased use of salad
dressings, cooking oils, and vegetable
shortening3 but is also reflected in changes
in levels of fatty acids for animal foods.
The lower level of saturated fatty acids
shows the decreased use of whole milk
and increased use of leaner red meats
and poultry. Despite lower whole milk
consumption, saturated fatty acids from
dairy foods increased somewhat between
1970 and 1994 because of consistent
year-to-year increases in per capita
consumption of cheese. The higher level
of polyunsaturated fatty acids reflects
the increased use of vegetables oils.
Higher vitamin E levels are related to
this trend as well, with the fats and oils
group providing over three-fifths of the
vitami n E to the food supply between
1970 and 1994. The level of cholesterol,
a dietary component of the lipid family
and found only in animal products,
declined because of the lower consump­tion
of eggs and red meat and the shift
from fluid whole milk to lowfat and
skim milks (figs. 4 and 5). The lower
level of vitamin B 12 in 1994 is also due
to decreased consumption of eggs as
well as organ meats:
3See box "About the Data" for limitations of fat
data.
8
Figure 4. Trends in meats and eggs in the U.S. food supply, 1970-94
Pounds
per capita
per year
200
150
100
50
Red meat
1
Poultry
Eggs
Fish
0 LL~~~~~~~L-L_L_L_L_~~~~~_L_L_L_L~~~~
1970 74 78 82 86 90 94
Year
11ncludes beef, pork, lamb, veal, and game.
Figure 5. Trends in dairy products in the U.S. food supply, 1970-94
Pounds
per cap~a
per year
250
200
150
100
50
Lowfat and skim milks
____ ,
Cheese
0 ~~~~~~~L-L-L-L-~~~~~-L-L-L-L~~~~--L-L
1970 74 78 82 86 90 94
Year
Family Economics and Nutrition Review
Figure 6. Fat sources in the U.S. food supply, 1970 and 1994
Percent
100
80
60
40
20
0
1970
Year
1994
D Vegetable
• Animal
Figure 7. Sources of calcium in the U.S. food supply, 1970 and 1994
Meat, fish,
poultry, and eggs
Vegetables
1970
Other dairy
1997 Vol.10No.3
Lowlat and
skim milk
1994
With the decline
in consumption
of whole milk and
beverage milks
overall, increased
calcium levels are
due ... to increases
in cheese
consumption ....
9
The dairy group is the primary source
of both calcium and phosphorus, con­tributing
about three-fourths of the
calcium and one-third of the phosphorus
in the food supply between I970 and
I994 (fig. 7, p. 9 and fig. 8). An increased
consumption of cheese was principally
responsible for higher levels of both
of these nutrients, with calcium and
phosphorus contributions from cheese
doubled from II to 23 percent and 5 to
I 0 percent, respectively, between 1970
and I994.
With the decline in whole milk con­sumption
and the shift to lowfat and
skim milk between I970 and I994,
calcium contributions more than doubled
from lowfat and skim milk, but were
only one-third those from whole milk
in 1970 (fig. 7). Overall, calcium contri­butions
from beverage milks were less
in 1994 than in 1970 due to the decreased
consumption of these milks. Although
lowfat and skim milk contributions to
total magnesium in the food supply
increased in 1994 over those in 1970,
this increase was not enough to prevent
the shift from the dairy group to the
grain group as the main contributor
of magnesium in 1994 (fig. 9).
The meat, poultry, and fish group
continued to provide important and
relatively stable contributions of both
magnesium and phosphorus and to be
the primary contributor of zinc in the
food supply between I970 and 1994
(figs. 8-I 0). However, the higher level
of zinc in I994 was due to increased
grain consumption as the zinc contribu­tions
from the meat, poultry, and fish
group declined and those from dairy
products remained stable during this
period (fig. I 0).
10
Figure 8. Sources of phosphorus in the U.S. food supply, 1970
and 1994
1970 1994
Vegetables
Figure 9. Sources of magnesium in the U.S. food supply, 1970
and 1994
Vegetables
1970
Meat, poultry,
and fish
Grains
1994
Family Economics and Nutrition Review
Figure 10. Sources of zinc in the U.S. food supply, 1970 and 1994
1970
Grains
Significant Implications of
U.S. Food Supply Data for
Food and Nutrition Policy
Public Health Issues: Nutrient
Availability Compared With
Federal Dietary Guidance
Whereas early dietary guidance was
directed toward the avoidance of defi­ciency
diseases, current guidance is
focused on the role of the diet in the
etiology and prevention of chronic
diseases (see box, p. 12, "Federal
Dietary Guidance"). The Recommended
Dietary Allowances (RDAs) ( 11 ), estab­lished
in 1943, have served as the basis
for the nutritional content of foods and
diets, and over the years, a nutritionally
adequate food supply was linked to
providing sufficient energy, macro­nutrients,
and micronutrients to meet
the needs of consumers ( 16). With
expanded scientific knowledge of the
roles of nutrients since the inception
of the RDAs, the purpose of the food
supply has become more complex,
ranging from the prevention of classical
nutritional deficiency diseases, such as
1997 Vol.JO No.3
1994
Eggs
rickets, to one of reduction of the risk of
such chronic diseases as cardiovascular
disease, cancer, and osteoporosis.
To better serve this purpose, food supply­related
research is based on activities
specified in the Ten-Year Comprehensive
Plan (23 ), the basis of the planning and
coordination of the NNMRRP. Per
capita data indicate that the U.S. food
supply is capable of providing recom­mended
nutrient levels for the total
population; however, dietary selection
is quite variable and may not be adequate
for some individuals because of social,
cultural, and economic factors ( 13 ). The
Third Report on Nutrition Monitoring
in the United States (TRONM), prepared
for the Interagency Board of the
NNMRRP, has classified several food
components as current or potential
public health issues (table 3, p. 13) (4).
Food components classified as current
public health issues are associated with
dietary intake disease risk, such as
obseity, coronary heart disease, anemia,
or osteoporosis. Those food components
classified as potential public health issues
require additional research to develop
... the purpose of
the food supply
has become more
complex, ranging
from the prevention
of classical nutri­tional
deficiency
diseases ... to one
of reduction of the
risk of ... chronic
diseases ....
11
Federal Dietary Guidance
Federal dietary guidance is outlined in Nutrition and Your Health: Dietary Guidelines for Americans and illustrated graphically in
the Food Guide Pyramid (figs. a and b) ( 18,20). Both are designed to help Americans, 2 years of age and older, choose diets
that meet their nutritional needs and improve health by reducing chronic disease risks. The fourth edition of the Dietary
Guidelines bulletin, published in 1995. outlines seven dietary recommendations that consumers should adopt for better heal th:
• Eat a variety of foods
• Balance the food you eat with physical activity-maintain or improve your weight
• Choose a diet with plenty of grain products, vegetables, and fruits
• Choose a diet low in fat, saturated fat, and cholesterol
• Choose a diet moderate in sugars
• Choose a diet moderate in salt and sodium
• If you drink alcoholic beverages, do so in moderation.
The Dietary Guidelines bulletin recommends that people choose a diet that provides no more than 30 percent of total calories
from fat and less than 10 percent of calories from saturated fat. Additionally, it discusses the role of enriched and fortified
foods in the diet and highlights good sources of several nutrients of concern to public health.
Figure a. Nutrition and Your Health:
Dietary Guidelines for Americans
12
Figure b. Food Guide Pyramid
Fats, Oils, & Sweets
USE SPARINGLY
Milk, Yogurt,
& Cheese
Group
Z·3 SERVINGS
Vegetable
Group
3-5 SERVINGS
Meat, Poultry. Fish.
Dry Beans, Eggs.
& Nuts Group
Z·3 SERVINGS
Fru1t
Group
z.4 SERVINGS
Bread, Cereal.
Rice, & Pasta
Group
6·11
SEJMHGS
Family Economics and Nutrition Review
Table 3. Classification of food components as public health issues in
the Third Scientific Report on Nutrition Monitoring1
Current public
health issues
Food energy
Total fat
Saturated fatty acids
Cholesterol
Alcohol
Iron
Calcium
Sodium
Potential public
health issues
Total carbohydrate
Dietary fiber
Sugar
Polyunsaturated fatty acids
Monounsaturated fatty acids
Trans fatty acids
Fat substitutes
Protein
Vitamin A
Vitamin C
VitaminE
Carotenes
Folate2
Vitamin B6
Vitamin B12
Magnesium
Potassium
Zinc
Copper
Selenium
Phosphorus
Fluoride
Modified from: Th!rd Report on Nutrition Monitoring in the United States (4 ).
Not current public
health issues
Thiamin
Riboflavin
Niacin
Iodine
1This classification of nutrients should be regarded as provisional. As new data become available, future
assessments of pub!ic health significance and of the levels of monitoring needed will result in changes in
the categorization of some food components.
2Since the publication of this classification table, the relationship of folic acid to neural tube defects has
been much publicized. In 1996, FDA announced that effective January I, 1998, folic acid must be added
to most enriched flour, breads, com meals, rice, noodles, macaroni, and other grain products. This action
heightens the folate role in national nutrition monitoring, and reclassification of this nutrient may be
necessary.
1997 Vol.JONo.3
... food supply
nutrient estimates
must be as accurate
as possible to
support policy
decisions that
improve the
nutritional status
of Americans.
13
interpretive criteria to link monitoring
data to functional or health outcomes
and/or improved food composition data.
Currently, the intake recommendations
for many nutrients and dietary compo­nents
are under scientific evaluation
by the Food and Nutrition Board of the
Institute of Medicine, National Academy
of Sciences as part of the Dietary
Reference Intake project. To comple­ment
this evaluation and because of the
important roles played by iron, folate,
and calcium in public health and in
nutrition and fortification policy, the
intake and availability levels of each
of these nutrients are compared with
Federal dietary guidance.
Iron deficiency is the most prevalent
nutrient deficiency in America, with
some of the most serious concerns
related to its effects on the health and
development of infants and children (6).
Infants, adolescents, and women of
childbearing age are those who are
most at risk of developing anemia.
Their greater iron needs, due to rapid
growth or excessive blood loss during
menstruation, usually cannot be met by
dietary intake alone.
In 1994, the level of iron in the food
supply was about 6 mg higher than in
1970 (table 1). The trend towards in­creased
iron reflects increased iron
fortification of breakfast cereals and
increased consumption of grain products
between 1970 and 1994 (fig. 11 ). At
21.2 mg per capita per day, the average
content of the food supply exceeded
recommendations for all sex-age groups
with the exception of pregnant women.
This does not mean, however, that all
Americans ingest the recommended
amounts of iron, since food supply
estimates can overestimate what is
actually consumed.
14
Figure 11 . Sources of iron in the U.S. food supply, 1970 and 1994
1970 1994
Meat, fish,
Legumes,
nuts, and soy
Vegetables
According to data from USDA's 1994
Continuing Survey of Food Intakes by
Individuals (CSFII), Americans as a
whole consumed an average of 136
percent of their Recommended Dietary
Allowance (RDA) for iron from food
sources (intake data do not include
dietary supplements). However, whereas
male adults 20 to 49 years old were
getting 182 percent of their RDA, females
of the same age were consuming 82 to
88 percent of their RDA, on average.
Epidemiological evidence indicates that
low serum folate levels are associated
with elevated serum homocysteine, an
independent risk factor for vascular
disease, and that the use of dietary sup­plements
containing folate by females
before they become pregnant and during
early pregnancy is associated with a
decreased incidence of some types of
neural-tube defects ( 4 ). In response to
this, the Food and Drug Administration
(FDA) recently directed that a folate
fortification policy of 140 J..Lg per 100 gm
for cereal-grain products, along with
fortification of ready-to-eat breakfast
cereals up to 100 J..Lg per serving and 400 J..Lg
per unit or serving for supplements, be
implemented in early 1998 (22). FDA's
goal is to increase folate intakes for the
target population (women of child­bearing
age) as close as possible to
recommended intakes while maintaining
safe levels for all other people (26).
In 1994, the level of folate available in
the food supply was about one-fifth
higher than the level in 1970 (table 1 ).
This is due to an increase use of grain
products and citrus fruit. However, the
vegetable and legumes, nuts, and soy
groups continue to be the major con­tributors
of folate (fig. 12). At 331 j.l.g
per capita per day, the average content
Family Economics and Nutrition Review
Figure 12. Sources of folate in the U.S. food supply, 1970 and 1994
1970 1994
Other
Meat, fish,
Legumes,
nuts, and soy
Vegetables
of folate in the food supply exceeded
recommendations for all sex-age groups
except pregnant women. According to
estimates from USDA's 1994 CSFII,
Americans as a whole consumed an
average of 169 percent of their RDA
for folate with no population subgroup
falling below 100 percent of the RDA;
however, the 1989 RDA does not account
for the safety factor needs of folate for
women of childbearing age.
Calcium is essential for the formation of
bones and teeth, and requirements increase
significantly during adolescence, early
adulthood, pregnancy, and lactation.
Calcium is very important from a public
health perspective because current
calcium intakes may be insufficient to
attain optimal peak adult bone mass
and to prevent age-related loss of bone
mass. Inadequate intake of calcium
may increase the risk of osteoporosis,
a condition in which decreased bone
mass weakens bone. In 1994, a National
1997 Vol.J0No.3
Institutes of Health Consensus Develop­ment
Conference on Calcium Intake
recommended that the current RDA for
calcium be increased to between 1,000
and I ,500 mg, depending on age and
other health factors, to help reduce the
risk of this disease (8).
The calcium level in 1994 was higher
than in 1970 (table 1 ). With the decline
in consumption of whole milk and
beverage milks overall, increased
calcium levels are due primarily to
increases in cheese consumption (fig. 7).
At 960 mg per capita per day, the average
content of the food supply failed to meet
calcium recommendations for all males
and females 11 to 24 years of age and
for pregnant and lactating women.4
4-rhe food supply calcium level of 960 mg in 1994
was adequate to provide the RDA by sex-age group
when the RDA was based on the population distribu­tion
as detennined by Census. More of the popula­tion
require 800 mg than I ,200 mg; thus 960 mg
was sufficient on a national basis.
... the addition of
nutrients to foods
through enrichment
and fortification has
been an effective
way to maintain and
improve the overall
nutritional quality
of the U.S. food
supply ....
15
According to estimates from USDA's
1994 CSFII, Americans as a whole
consumed an average of 92 percent of
their RDA for calcium. The intake by
males ranged from 91 to 118 percent
of the RDA and men 20 years and over
averaged 107 percent of the RDA. The
intake by females ranged from 67 to
82 percent of the RDA and women 20
years and over averaged 75 percent of
the RDA. These calcium intakes indicate
that individuals are not consuming
enough of the foods rich in calcium
available to them in the food supply,
much less the amounts that are currently
being suggested by many health experts.
This finding has policy implications
for nutrition educators as they design
nutrition education programs to increase
calcium consumption, especially for
adolescent girls and women, and for
policymakers as they consider strategies
for improving calcium intakes of
Americans.
The Role of Food Fortification in
Meeting Public Health Objectives
for Chronic Disease Prevention
Historically, especially during the 1930's
and 1940's, the addition of nutrients to
foods through enrichment and fortifica­tion
has been an effective way to main­tain
and improve the overall nutritional
quality of the U.S. food supply and to
solve the public health problems seen
earlier in this century ( 1 6,26). For example,
enrichment of cereal-grain products has
been an extremely effective means of
enhancing the nutrient quality of the
food supply and serves as a practical
and effective public health strategy for
providing required nutrients. Cereal
grains are eaten by essentially all popu­lation
groups. Other nutrient-deficiency
diseases were dealt with by fortifying
various foodstuffs with specific nutrients.
16
By the 1950's, four specific programs­fortification
of salt with iodine; fortifica­tion
of milk with vitamin D; enrichment
of flour and grains with thiamin, ribo­flavin,
niacin, and iron; and fortification
of margarine with vitamin A-were in
effect for the addition of nutrients to the
U.S. food supply.
Over time, guidelines for food fortifica­tion
have evolved, as interest in adding
nutrients to foods has shifted from
prevention of deficiency diseases to
broader issues of improving overall
health. Food fortification policy has also
been established for other commodities
and additional nutrients. Although grain
products have proven to be the most
suitable vehicles for fortification in the
food supply, fortified foods also include
(but are not limited to) ready-to-eat
cereals, fruit drinks and juice, meal
replacement bars and beverages, infant
formulas, margarine, and milk-based
products. Many juice drinks are fortified
with vitamin C, which is the key vitamin
provided by most juices and juice drinks.
As a result of fortification, some juices
and juice drinks are also a good source
of calcium and vitamin A as beta­carotene.
Nutrient additions supplement the
amount of thiamin, riboflavin, niacin,
iron, iodine, and vitamins A, C, and D
in the food supply substantially and the
prevalence of specific nutrient deficiencies
in the population has been greatly reduced.
In particular, the gradual rise in the per­centage
of white flour that is enriched
has contributed to increased levels of
thiamin, riboflavin, niacin, and iron
( 15 ), accounting for over half of the
thiamin and iron, two-fifths of the
niacin, and almost one-third of the
riboflavin in the food supply (table 2)
in 1994. Also, the fortification of marga­rine
with vitamin A accounts for about
9 percent of the total vitamin A in the
food supply.
Because the food supply series measures
foods and nutrients over time, the impact
of added nutrients for purposes of
enrichment and fortification of basic
food commodities can be gauged. This
important association of the.food supply
with fortification policy continues to
be monitored with research activities
directed towards the evaluation and
determination of nutrients added to the
food supply for enrichment, fortification,
and functional purposes.
If nutrient requirements can be met readily
by the food supply without nutrient
addition, then nutrition educators can
develop programs to help individuals
meet nutrient requirements or improve
nutrient intakes. However, a dietary
recommendation to increase intake of a
particular nutrient may hinder the food
supply's ability to provide an adequate
level of that nutrient. In this case, policy­makers
need to consider fortification
options or other strategies to improve
the food supply. Most recently, the
public health goal of increasing folate
intakes for women of childbearing age
to reduce the occurrence of neural-tube
defects in infants led FDA to rule that
cereal-grain products be fortified with
folic acid by January 1, 1998 (22).
Family Economics and Nutrition Review
Future Directions
Scientific evidence on the relationship
between diet and health and the impor­tance
of fortification in public health
require accurate and updated information
on the quality and composition of the
American diet. Such information is used
to determine the extent to which diets
differ from Federal dietary guidance
and to monitor the dietary and nutri­tional
status of Americans. In order for
policy makers to translate dietary guid­ance
into goals for food production and
supply levels and to make appropriate
decisions regarding food fortification,
the link between food supply and food
intake data needs to be better understood.
This requires further research and analysis
of these two types of data to recognize
their differences and similarities. Addi­tionally,
to minimize the limitations of
food supply data and to better translate
per capita food consumption and nutri­ent
estimates into intake data, improved
food supply estimates of fats and oils
and discard and processing factors are
needed.
Also, research is needed to update the
fortification data base that supports the
U.S. food supply series. Several studies
have looked at how the nutrient content
of the food supply has been affected by
enrichment and fortification (2,3,12).
The earlier study (2) showed that fortifi­cation
and enrichment of grain products
provide a significant proportion of the
thiamin, niacin, iron, and riboflavin
supply for the average American. More
recent studies (3,12) have examined
the effect of food fortification in terms
of distribution of nutrient intakes by
population subgroups as well as over-
1997 Vol.10No.3
fortification of the food supply. Both
studies acknowledge the importance of
fortifying the food supply as a public
health intervention to improve nutritional
intake of a particular nutrient by a target
population.
Nutrient data bases that account for
added nutrients are limited and the
U.S. Department of Agriculture (USDA)
nutrient data bases do not routinely
identify levels of added nutrients. As
a result, the food supply fortification
data base has not been updated since
1970, except for the percentage of flour,
enriched. Since 1970, enormous changes
in food industry fortification practices
have occurred and both the range of
fortified foods and the number of added
nutrients have expanded. An updated
version of USDA's food composition
data base, designed to include nutrients
added to foods commercially through
enrichment and fortification, is needed
to generate more accurate food supply
nutrient estimates for policy work on
fortification issues and nutrient recom­mendations.
As a major component
of the NMRRP, food supply nutrient
estimates must be as accurate as possible
to support policy decisions that improve
the nutritional status of Americans.
17
18
References
I. Cleveland, L.E., Cook, D.A., Krebs-Smith, S.M., and Friday, J. 1997. Method
for assessing food intakes in terms of servings based on food guidance. American
Journal of Clinical Nutrition 65(suppl): 1254S-1263S.
2. Cook, D.A. and Welsh, S.O. 1987. The effect of enriched and fortified grain
products on nutrient intake. Cereal Foods World 32(2): 191-196.
3. Crane, N.T., Wilson, D.B., Cook, D.A., Lewis, C.J., Yetley, E.A., and Rader, J.l.
1995. Evaluating food fortification options: General principles revisited with
folic acid. American Journal of Public Health 85(5):660-666.
4. Federation of American Societies for Experimental Biology, Life Sciences
Research Office. Prepared for the Interagency Board for Nutrition Monitoring and
Related Research. 1995. Third Report on Nutrition Monitoring in the United States.
Volume 1.
5. Gerrior, S. and Bente, L. 1997. Nutrient Content of the U.S. Food Supply, 1909-94.
U.S. Department of Agriculture, Center for Nutrition Policy and Promotion. (In Press).
6. Guthrie, J.F. and Schwenk, N.E. 1996. Current issues related to iron status: Implication
for nutrition education and policy. Family Economics and Nutrition Review 9( 4 ):2-19.
7. Kantor, L.S. 1996. Many Americans are not meeting food guide pyramid dietary
recommendations. F oodReview 19( 1 ):7 -15.
8. National Institutes of Health, Office of the Director. 1994. Optimal Calcium
Intake, NIH Consensus Statement 12(4):1-31.
9. National Research Council, Committee on Diet, Nutrition and Cancer. 1982.
Diet, Nutrition and Cancer. National Academy Press, Washington, DC.
10. National Research Council, Committee on Diet and Health, Food and Nutrition
Board. 1989. Diet and Health Implications for Reducing Chronic Disease Risk.
National Academy Press, Washington, DC.
11. National Research Council, Subcommittee on the Tenth Edition of the
RDAs, Food and Nutrition Board. 1989. Recommended Dietary Allowances,
l01
h ed. National Academy Press, Washington, DC.
12. Popkin, B.M., Siega-Riz, A.M., and Haines, P.S. 1996. The nutritional
impact of food fortification in the United States during the 1970's. Family Economics
and Nutrition Review 9(4):20-30.
13. Position of the American Dietetic Association: Enrichment and fortification
of foods and dietary supplements. 1994. Journal of the American Dietetic Association
94(6):661-63.
Family Economics and Nutrition Review
14. Putnam, J. and Allshouse, J. 1997. Food Consumption, Prices, and Expenditures,
I997 Annual Data. U.S. Department of Agriculture, Economic Research Service.
(In Press).
15. Ranum, P.M. 1980. Notes on levels of nutrients to add under expanded wheat
flour fortification/enrichment programs. Cereal Chemistry 57( I ):70-72.
16. Thomas, P.R. and Earl, R. (eds.) 1994. Opportunities in the Nutrition and
Food Sciences: Research Challenges and the Next Generation of Investigators.
Food and Nutrition Board, Institute of Medicine. National Academy Press,
Washington, DC.
17. U.S. Department of Agriculture, Agricultural Research Service. 1996. I994
Continuing Survey of Food Intakes by Individuals and the Diet and Health
Knowledge Survey, on CD-ROM.
18. U.S. Department of Agriculture, Human Nutrition Information Service. 1992.
The Food Guide Pyramid. Home and Garden Bulletin No. 252.
19. U.S. Department of Agriculture and U.S. Department of Health and Human
Services. 1980. Nutrition and Your Health: Dietary Guidelines for Americans.
U.S. Department of Agriculture. Home and Garden Bulletin No. 232.
20. U.S. Department of Agriculture and U.S. Department of Health and Human Services.
1995. Nutrition and Your Health: Dietary Guidelines for Americans. (4th ed.).
U.S. Department of Agriculture. Home and Garden Bulletin No. 232.
21. U.S. Department of Health, Education, and Welfare, Public Health Service. 1979.
Healthy People: The Surgeon General's Report on Health Promotion and Disease
Prevention. DHEW (PHS) Publication No. 79-55071.
22. U.S. Department of Health and Human Services, Food and Drug Administration.
1996. Food standards: Amendment of standards of identity for enriched grain products
to require addition of folic acid. Final Rule. Federal Register 61( 44 ):8781-8797.
23. U.S. Department of Health and Human Services and U.S. Department of Agriculture.
1993. Ten-Year Comprehensive Plan for the National Nutrition Monitoring and
Related Research Program. Federal Register 58( I I I ):32752-32806.
24. U.S. Senate, 95th Cong., 1st sess., Select Committee on Nutrition and Human
Needs. 1977. Dietary Goals for the United States, 2nd ed.
25. Welsh, S. 1994. Atwater to the present: Evolution of nutrition education.
Journal of Nutrition 124: 1799S-1807S.
26. Yetley, E.A. and Rader, J.l. 1996. The challenge of regulating health claims and
food fortification. Journal of Nutrition 126:765S-772S.
1997 Vol.IO No.3 19
20
Nutrient Intakes and Eating
Patterns of Teenagers
Elyse Levine 1
Department of Marketing
George Washington University
Joanne F. Guthrie
Nutritionist
Center for Nutrition Policy and Promotion
When compared with current dietary recommendations, diets of teenagers
are marked by underconsumption of several vitamins and minerals, as well
as dietary fiber, and overconsumption of fat and saturated fat. This paper
examines nutrient intakes of adolescents 13-18 years of age by gender and
race using a nationally representative sample from the 1989-91 Continuing
Survey of Food Intakes. It also explores food consumption and meal patterns in
order to gain insights into dietary behavior that may be useful in developing
nutrition promotion messages and strategies. Average intakes of female
adolescents were below recommendations for vitamin E, calcium, iron, and
zinc. Average intakes of male adolescents were lower than recommended
for vitamin E, calcium, and zinc. Mean vitamin A intakes of Black teenagers
were also below recommendations. Dietary fiber intakes were below
recommendations for all subgroups. Overall, teens averaged 35 percent of
calories from fat and 13 percent from saturated fat, with no significant
differences between subgroups. Examining food group intakes, male
teenagers did not meet recommendations for consumption of foods from
the fruit, vegetable, and grain groups. Black male teens and female teens
of both races did not meet recommendations for the fruit, vegetable, grains,
or milk groups. Examination of meal patterns revealed that females, especially
Black females, tended to skip breakfast more than males. Black teenagers
were significantly more likely to skip lunch than Whites. The results suggest
potential strategies for tailoring nutrition promotion efforts to the dietary
problems and eating patterns of teens.
[QJ ietary surveys of adolescents
between the ages of 13 and
18 years have revealed two
disturbing trends: under­consumption
of vitamins, minerals, and
dietary fiber, especially by females; and
higher than recommended intakes of fat
1Currently with Prospect Associates, Rockville,
MD.
and saturated fat. During adolescence,
the demands of physical maturation
require more calcium, iron, and energy
than at any other age (26). Because
teenage males eat more food than
females, they are less likely to have
intakes below recommendations, with
the exception of calcium and dietary
fiber (9,40). However, findings from
Family Economics and Nutrition Review
The School Nutrition Dietary Assess­ment
Study, the 1987-88 Nationwide
Food Consumption Survey (NFCS),
and the second National Health and
Nutrition Examination Survey (NHANES
II) confirm that many teenage females
fall short of meeting the recommended
dietary allowances (RDAs) established
by the Food and Nutrition Board of the
National Academy of Sciences (25) for
calcium, iron, magnesium, and zinc
(12,13,15,18).
In addition to gender differences in
nutrient intake, racial differences have
also been noted. Analyses of the 1987-
88 NFCS and the NHANES II found
consumption of calcium by Black
adolescents was lower than that of
White teens ( 12,13,21 ). Iron deficiency
in adolescent females has also been
found to be associated with race.
Results from NHANES II indicated
8 percent of non-Hispanic Black females
12-15 years of age and 14 percent of
those 16-19 years old were iron-deficient
compared with fewer than 4 percent of
White females in those age categories (20).
Overconsumption of fat and saturated
fat is also characteristic of teens' diets.
This dietary imbalance may increase
risk of chronic health problems later
in life (38), and may contribute to the
growing problem of adolescent obesity.
According to findings from NHANES
II and NHANES III, the prevalence of
obesity between the ages of 12 and 17
jumped from 16 percent to 21 percent
between the 1970's and the 1980's (33).
Several studies find the trend to obesity
more prevalent among low-income and
racial minority children and adolescents
( 3,33 ). Diets high in calories and fat
have been associated with obesity in
children and adolescents (28) as have
1997 Vol.JO No.3
a lack of physical activity ( 5) and an
abundance of hours spent watching
television ( 10,28,29).
Although these findings indicate several
nutritional problems occurring in this
age group, few studies using national
survey data have reported on eating
patterns that contribute to these problems.
In this study, data from the U.S. Depart­ment
of Agriculture's 1989-91 Continuing
Survey of Food Intakes by Individuals
(CSFII) were used to assess nutrient
consumption and eating patterns of a
national sample of White non-Hispanic
and Black non-Hispanicl teenagers. A
previous analysis of CSFII data found
that only 11.6 percent of boys and 7.2
percent of girls between the ages of
12 and 18 consume the number of daily
servings of fruits and vegetables recom­mended
in the USDA/DHHS Food
Guide Pyramid ( 19 ). The current study
expands on previous work by examining
teenagers' consumption of all five of
the major food groups defined in the
Food Guide Pyramid (35) and the meals
at which they are eaten. This information
can be used to help design more effective
nutrition promotion interventions by
tailoring messages to actual eating
behaviors.
Methods
Sample
The USDA Continuing Survey of Food
Intakes by Individuals (CSFII) is an
ongoing national survey. In 1989-91,
households were selected for the survey
using a stratified area probability sampling
method that over-samples for low-income
households. Data collected between
1989 and 1991 included consumption
information from each household member
2These groups will be referred to as White and
Black for the remainder of the article.
for 3 consecutive days, using the dietary
recall method for the first day and the
food record method for the second and
third days. The survey was designed to
collect dietary information on all days
of the week and all seasons of the year
to avoid temporal biases. In this study,
only the first day's data were used
because many subjects did not complete
the second and third days of dietary
data. Dietary recalls were conducted
by a trained interviewer who probed as
necessary to ensure complete descriptions
of foods and beverages and to check for
omissions. If a household member was
absent at the time of the interview, the
main meal planner/preparer was asked
to report what that person ate.
The 1989-91 CSFII included 1,087
adolescents between the ages of 13 and
18 years who provided 1-day dietary
recalls. For this study, the sample was
screened for teenagers between the ages
of 13 and 18 who were living at home
with at least one parental figure and
who were not pregnant or nursing.
Further screening removed outliers for
self-reported consumption above 7,000
calories per day. Finally, 36 individuals
whose dietary records included a high
proportion of foods (greater than I 0 per­cent
of total calories) with incomplete
information on meal time or name of
eating occasion were removed from
the data base. The resulting sample
comprised 926 adolescents of whom
735 were White and 191 were Black.
Information on racial and cultural
origins was self-reported.3
3The sample included teens of Hispanic origin,
American Indian/Eskimo, and Asian/Pacific Islander
races; however, these groups had insufficient sample
size for statistical tests conducted in this article
and were therefore omitted from the study sample.
21
22
White females
consumed more
milk, and Black
females ate more
foods from the meat
group, particularly
poultry.
In this sample, about 36 percent of the
responses were provided primarily by
the main meal planner on behalf of the
teenager. When these responses were
compared with information on consump­tion
of major food groups that was
provided primarily by the teenagers
themselves, no significant differences
were found. Therefore, information from
parent-respondents was used for the
study.
Definition of Variables
Dietary intakes of protein, carbohydrates,
and fat were calculated as percent of
total energy (calories). Consumption of
vitamins and minerals was calculated
as a percent of the individual's RDA.4
Nutrient intakes were computed from
reported food consumption using a data
base developed for this survey (36).
Reported values represent intake from
food only. CSFII respondents were
queried on use of vitamin-mineral sup­plements,
but the information obtained
was not precise enough to translate into
measured intake. Frequency of use of
vitamin-mineral supplements is reported,
however.
To analyze eating patterns, eating occa­sions
were defined by the respondent
as breakfast, lunch, supper, or a snack.
Self-reported snacks remained classified
as snacks when they were reported in
addition to meals, regardless of energy
value. In some cases individuals reported
not eating a meal, but did report eating a
snack around those meal times. Following
a protocol used to assess eating occasions
in a previous study (22), snacks eaten
in lieu of meals that provided more than
50 calories were recoded as breakfast if
they were consumed between 4:30 and
11:00 a.m.; lunch if between 11:01 a.m.
and 3:00p.m.; and supper between
4RDAs are age- and gender-specific.
3:01 p.m. and 10:00 p.m. Eating occasions
that totaled 50 calories or less remained
classified as snacks, even if no meal
was reported.
Grams of foods consumed were converted
to servings consistent with the Food Guide
Pyramid using a data base developed by
USDA's Center for Nutrition Policy
and Promotion (CNPP) for this purpose.
Mixed foods were disaggregated into
their basic Pyramid components (for
example, a ham and cheese sandwich
was disaggregated into ham, cheese, and
bread, and each component placed in the
appropriate group). Categorization of
foods and food components into food
groups was consistent with Food Guide
Pyramid definitions ( 35 ). For example,
french-fried potatoes were defined as
vegetables in the Food Guide Pyramid
and therefore were placed in the vegetable
group; ice cream was considered part of
the milk, yogurt, and cheese group and
was placed in that group.
Survey data report food intake by weight
(grams); however, the weight of a serving
size may vary from item to item. For
example, 3/4 cup tomato juice and 1/2 cup
cooked tomato are each considered a
serving of vegetable but weigh slightly
different amounts. To translate intake
into serving amounts, serving sizes for
all foods consumed by survey respondents
were identified using USDA information
(31,36) and the gram weight equivalent
for each serving size was identified using
the USDA Survey Codebook (37). In
some cases, food ingredients that pro­vided
only small amounts (less than 1/4
of a serving) of a Pyramid food group
were not counted toward the group total.
Baby foods, fats and oils, sugars, soft
drinks, coffee, tea, or other foods not
counted in at least one of the five major
food groups were not included in the
data base.
Family Economics and Nutrition Review
Statistical Analysis
The Statistical Package for the Social
Sciences (SPSS) ( 32) was used to calcu­late
average nutrient intake and consump­tion
of foods. Weights developed by the
USDA to adjust for variable probabilities
of selection, differential nonresponse
rates, and sampling frame considerations
were applied to the sample in order to
obtain results more generalizable to the
U.S. population. SUDAAN, a statistical
analysis program designed for use with
survey data collected using a complex,
stratified sampling design ( 30 ), was
used to derive standard errors and to
conduct student t-tests and chi-square
analyses for assessing differences
between groups.
Results
Description of Sample
As shown in table 1, the sample was
distributed almost equally between
males and females. The average age
was 16 years. The majority were White
and lived in households with two parents.
In this sample, Black teenagers were
more likely than Whites to live in single­parent
households (fig. 1, p. 24 ). House­hold
income as a percentage of the Federal
poverty level was used as an indicator
of subjects' socioeconomic status. Black
teenagers, on average, came from house­holds
at 184 percent of the poverty level,
compared with White teenagers at 381
percent.
The average body mass index (BMI)5
for the sample was 21.6, and 63 percent
of the respondents had BMis between
19 and 26. A BMI of 26 is a recom­mended
cut-off for the definition of
5 BMI was calculated based on self-reported
heights and weights.
1997 Vol.JO No.3
obesity in adolescents (1 8). Of those
teens who did not fall into this range, 23
percent had BMis below 19 and 14 per­cent
had BMis above 26. Fewer than 4
percent of the sample had BMis greater
than 30, a marker for severe obesity.
Nutrient Consumption
Tables 2, p. 25 and 3a, p. 26 confirm
findings from previous national surveys
on deficiencies and excesses in diets of
teenagers. For the total sample, the pro­portion
of energy (calories) from fat
(35 percent) and saturated fat (13 percent)
exceeded current recommendations of
no more than 30 percent and less than
10 percent, respectively. Average choles­terol
intake approached, but did not
surpass, the recommended ceiling of
300 mg per day, except for White males.
Cholesterol consumption by Black females
was significantly higher than that by
White females. Among all groups,
dietary fiber consumption ranged from
11 to 16 grams, below the "age plus
five" levels recommended by the
American Health Foundation for children
and adolescents (39). Intakes by females
were particularly low, with White females
averaging 11.2 grams of dietary fiber
per day and Black females averaging
11.0 grams/day.
Examining nutrient intakes from food,
average intakes by all age-sex groups
exceeded their RDAs for vitamin C and
folate, but consumption of other vitamins
and minerals fell short of recommenda­tions
for some subgroups. Among females,
calcium consumption averaged only 67
percent of their RDA. Females' average
intakes of vitamin E, iron, and zinc were
about 80 percent of their RDAs for these
nutrients. Consumption of all vitamins
and minerals tended to be lower among
Table 1. Demographic
characteristics of teenagers1
Characteristics
Gender
Male
Female
Race
White
Black
Age (years)
13
14
15
16
17
18
Region
Northeast
Midwest
South
West
Household income
Total
White
Black
1 Weighted data.
Percent
51.1
48.9
82.5
17.5
16.1
16.9
17.7
15.5
16.6
17.2
19.6
26.4
37.4
16.6
Percent
poverty
347
381
184
23
Black females than among White females,
and the difference was significant for
vitamin A. Mean vitamin A intakes of
Black females were below recommenda­tions,
but intakes of White females were
not.
On average, males met or exceeded 90
percent of their RDAs for most vitamins
and minerals with the exception of
vitamin E, for which they averaged
88 percent of their RDA. Consumption
by Black males tended to be less than
consumption by White males for all
vitamins and minerals, and these differ­ences
were significant for vitamin E,
calcium, and zinc.
As shown in table 3b, about one-quarter
of teens use vitamin-mineral supplements
at least occasionally. Usage was signifi­cantly
higher among Whites (28.2 percent)
compared with Blacks (8.7 percent). It
was slightly higher among males than
females (26.7 percent compared with
22.7 percent), but the difference was
not significant.
Consumption of Foods
The Food Guide Pyramid provides
recommended serving ranges for each
of the five major food groups. Within
those ranges, individuals are advised
to choose more or less servings based
on their caloric needs. The USDA's
Healthy Eating Index (34) has identified
specific serving recommendations for
the major age-sex groups, based on their
usual caloric requirements (fig. 2, p. 27).
These serving recommendations can
be used as a basis for comparison with
actual food group servings consumed
by adolescents. Table 4, p. 28 summarizes
the average daily number of Food Guide
Pyramid servings consumed by teenagers
in the sample. Consumption of most
major food groups tended to be slightly
24
Figure 1. Household structure of White and Black teenagers
Percent
• Single parent ffil Two parent
100
80
60
40
20
0
White
higher for White teens, but the differ­ences
were not significant for most
foods. Blacks' intake of the meat, fish,
poultry, dry beans, eggs, and nut group
(hereafter called "meat group" for
simplicity) was higher than that of White
teens, and this difference was significant
for females. For all teens in the sample,
consumption of whole-grain products
accounted for a small proportion of
foods eaten from the bread, cereal, rice,
and pasta group (hereafter called "grain
group" for simplicity). The majority of
grain products consumed were white
bread or rolls, which accounted for 26.5
percent of grains eaten. Starchy vegetables,
including french fries, comprised 39.3
percent to 58.4 percent of total vegetable
intake for the teens in the sample, with
the proportions tending to be slightly
higher for Black teens. Black teens also
tended to consume more dark-green
vegetables compared with Whites, but
all teens, on average, ate much less than
one serving of dark-green or deep-yellow
vegetables daily.
Black
Consumption by Males
Males, on average, consumed less than
the recommended number of servings
for plant-based foods at the base of the
Pyramid. The average daily consump­tion
of foods in the grain group, 7.75
servings, was low compared with the
recommended levels of 9.9 servings for
males 11-14 years old, and 11 servings
for males 15-18. Consumption of
vegetables, 2.3 servings, was about
half of the recommended amount (4.5-
5 servings). White males ate significantly
more fruits than Blacks (1.05 and 0.37
servings, respectively), but consumption
for both groups was far less than the
recommendation of 3.5-4 servings.
Differences between White and Black
males were also significant for the average
consumption from the milk, yogurt, and
cheese group (2.53 and 1.66 servings),
and only White males met recommenda­tions
(3 servings). Servings of the meat
group, averaging 4.45, exceeded the
recommendations (2.6-2.8 servings).
Family Economics and Nutrition Review
Table 2. Nutrient consumption: Macronutrients by teenagers, 1 day, CSFII 1989-911
Total sample Males Females
Total White Black Total White Black Total White Black
Nutrient (n = 926) (n = 735) (n = 191) (n = 460) (n = 381) (n = 79) (n =466) (n = 354) (n = 112)
Calories 2109 2146
Protein
Grams 79 80
% kcal 15.0 14.8
Carbohydrates
Grams 268 274
% kcal 50.9 51.1
Fat
Grams 83 84
% kcal 35.4 35.2
Saturated fat
Grams 31 32
% kcal 13.2 13.4
Cholesterol
Milligrams 259 254
Fiber
Grams 13.6
1Weighteddata.
* p S .05 between White and Black.
Consumption by Females
Females' average daily consumption
of vegetables ( 1.71 servings) and fruits
(0.77 servings) was less than half of
their recommendations for these food
groups (4 and 3 servings, respectively).
Their average consumption from the
grain group (5.32 servings) and milk
group (1.70 servings) was comparatively
closer to recommended amounts,
but still fell short (9 and 3 servings,
respectively). Differences between the
average consumption of foods from the
milk and meat groups by White and
1997 Vol.JO No.3
13.8
1934 2455 2502 2149 1747 1732 1800
76 92 94
15.6 15.0 15.0
242 311 316
50.1 50.6 50.5
76 97 99
35.4 35.6 35.6
27 36 37
12.6 13.2 13.3
278 296 302
12.7 16.0 16.0
Black females were significant. White
females consumed more milk, and
Black females ate more foods from the
meat group, particularly poultry. Total
consumption of foods from the meat
group by Black females was slightly
higher than the recommended 2.4 servings.
Meal Patterns
Table 5, p. 29 and figure 3, p. 30 provide
more detailed information on when teen­agers
were most likely to eat different
food types. In tables 5a-5c, consumption
of food groups is broken out by eating
occasion. Foods from the grain and
80 66 64 73
14.9 15.0 14.7 16:2
279 224 226 219
51.9 51.3 52.1 48.8
82 68 67 72
34.3 34.9 34.6 36.0
29 25 25 26
12.3 13.0 13.0 12.8
255 219* 198 293
15.5 11.2 11.2 11.0
milk groups were consumed throughout
the day, but other food groups appear to
be favored at specific eating occasions.
Following patterns of typical American
diets, foods from the vegetable and meat
groups were eaten mostly at lunch and
supper. White males are more likely
to eat fruit at breakfast or as snacks,
whereas Black males had no discernable
meal patterns for consuming the low
amount of fruit observed with this group.
Consumption by White and Black females
tended to be more constant throughout
the day.
25
Table 3a. Nutrient consumption: Vitamins and minerals1 by teenagers, 1 day, CSFII 1989-912
Total sample Males Females
Total White Black Total White Black Total White Black
Nutrient (n = 926) (n = 735) (n = 191) (n = 460) (n=381) (n = 79) (n = 466) (n = 354) (n=ll2)
Percent RDA
Vitamin A 99* 104 74 105 108 85 93* 100 68
Vitamin C 168 172 149 184 191 142 151 150 153
Vitamin E 85 87 75 88* 91 65 81 81 82
Calcium 82* 86 62 96* 99 72 67 70 56
Iron 115 119 101 148 149 145 81 83 73
Folate 152 156 135 171 171 171 132 138 113
Zinc 86 88 78 93* 96 73 79 78 81
1Intakes are from food only.
2Weighted data.
* p ~ .05 between White and Black.
Table 3b. Use of vitamin and mineral supplements by teenagers, 1 day, CSFII 1989-911
Usage
Everyday
Every so often
Not at all
1Weighted data.
Total*
(n = 926)
14.0
10.8
75.2
* p ~ .05 between White and Black.
Meal Skipping
Total sample
White
(n = 735)
15.9
12.3
71.8
As shown in figure 3, all subgroups of
teens in this sample generally reported
similar meal-skipping behavior. Females
tended to skip breakfast more than males,
and this was especially pronounced for
Black females, of whom only two-thirds
reported eating breakfast on the survey
day. On average, the great majority of
males reported eating breakfast, lunch,
and supper. However, Black teens in
26
Males Females
Black
(n = 191)
Total
(n = 460)
White
(n = 381)
Black
(n = 79)
Total
(n = 466)
White
(n = 354)
Black
(n=ll2)
4.8
3.9
91.3
16.5
10.2
73.3
Percent
18.1
11.4
70.5
5.7
2.7
91.6
this sample were significantly less
likely to eat lunch than were White
teens. Only 57 percent of Black males
and 63 percent of Black females reported
eating lunch on the survey day. The
majority of teens ate snacks; the most
common snacking times were between
2:00p.m. and 4:30p.m. and between
8:00p.m. and 10:00 p.m.
11.3
11.4
77.3
Discussion
13.3
13.3
73.4
4.2
4.6
91.2
Before considering the implications of
this study, its limitations must be noted.
Dietary intake was assessed using
24-hour recall data. Such data are poor
estimators of a given individual's usual
diet. They are, however, useful in
characterizing the intake of groups,
given a sufficient sample size.
Family Economics and Nutrition Review
Figure 2. Recommended numbers of food group servings per
day for male and female adolescents, 13-18 years of age
Males Females
9
Source: U.S. Department of Agriculture, Center for Nutrition Policy and Promotion, 1995, The
Healthy Eating Index, CNPP-1 (34).
Comparison groups were created on the
basis of gender and race; however, the
groups may also have differed in numerous
other characteristics that may have influ­enced
dietary differences reported in
this study. In particular, the discrepancy
in income between the two racial groups
indicates the need for examination of
other factors, such as income, when
interpreting comparisons between races.
A prior analysis of CSFII 1989-91 data
found household income to have only a
minor influence on nutrient intakes of
children and adolescents (2). However,
income may affect purchases of specific
foods . In a multivariate analysis of
women's diets, higher income was
found to be significantly associated
with increased fruit and vegetable
intake ( 16 ). Further research examining
the effect of household income on
adolescents' food and nutrient intakes
is indicated.
1997 Vol.JO No.3
A documented limitation of dietary
recalls is the tendency for respondents
to underreport consumption of foods,
especially foods high in fat and calories
(23 ). Recommendations from the
National Academy of Sciences (25)
for average energy intake for males and
females at median height and weight
exceeded those reported in the study
by 12 percent for males and 26 percent
for females, indicating that some under­reporting
is likely.
A further limitation of this study was
the use of secondary respondents
(usually parents) to report some teens'
diets. Although we found no significant
differences between the average con­sumption
of major food groups or nutri­ents
by teens when " parent-reported
diets" were compared with self-reported
diets, it would have been preferable to
have obtained all dietary data from the
teens themselves.
Black teens in
this sample were
significantly less
likely to eat lunch
than were White
teens.
27
Table 4. Consumption of foods by teenagers (Food Guide Pyramid servings), 1 day, CSFII 1989-911
Total sample Males Females
Mean intake Mean intake Mean intake
Total White Black Total White Black Total White Black
Food group (n = 926) (n = 735) (n=l91) (n = 460) (n=381) (n = 79) (n =466) (n = 354) (n= 112)
Total bread, cereal,
n.e e, and pasta group2 6.57 6.75 5.72 7.75 7.86 7.03 5.32 5.45 4.90
Whole grain 1.01 1.04 0.85 1.27 1.31 0.97 0.74 0.73 0.78
Other grain 5.56 5.71 4.86 6.49 6.55 6.06 4.59 4.72 4.12
Grain chips3 0.19 0.17 0.24 0.15 0.15 0.16 0.22 0.20 0.28
Total vegetable group4 2.01 2.04 1.88 2.30 2.26 2.55 1.71 1.78 1.47
Starchy vegetables 0.85 0.83 0.94 1.01 0.94 1.49 0.68 0.70 0.60
Vegetable chjps5 0.25 0.21 0.40 0.25 0.28 0.09 0.24 0.14 0.59
Dark-green vegetables 0.10 0.08 0.18 0.11 0.09 0.27 0.09 0.07 0.13
Deep-yellow vegetables 0.07 0.08 0.04 0.08 0.09 0.04 0.06 0.07 0.04
Other vegetables 0.99* 1.05 0.72 1.09 1.14 0.74 0.89 0.94 0.71
Total fruit group 6 0.87* 0.94 0.56 0.96* 1.05 0.37 0.77 0.80 0.69
Citrus fruit 0.48* 0.53 0.23 0.53* 0.58 0.15 0.43 0.48 0.29
Other fruit 0.39 0.40 0.33 0.43 0.47 0.22 0.34 0.32 0.40
Total milk, yo9urt, 2.06* 2.20 1.45 2.41 * 2.53 1.66 1.70 1.81 1.31
cheese group
Fluid mjJk 1.56* 1.66 1.10 1.87 1.93 1.49 1.23* 1.34 0.86
Yogurt 0.02* 0.03 0 0.02 0.02 0 0.02* 0.03 0
Cheese 0.48 0.51 0.34 0.52* 0.57 0.18 0.44 0.44 0.45
Total meat, fish, poultry,
dry beans, eggs, nuts 3.80 3.72 4.21 4.45 4.48 4.22 3.13* 2.82 4.21
group8
Meat 2.62* 2.79 1.83 3.18* 3.43 1.53 2.04 2.05 2.01
Poultry 0.96* 0.75 1.97 1.04* 0.85 2.30 0.87* 0.62 1.77
Fish 0.22 0.18 0.41 0.23 0.20 0.39 0.21 0.14 0.43
Legumes 0.16 0.16 0.15 0.21 0.22 0.18 0.10 0.09 0.14
1Weighteddata.
2Eisewhere called " Grains" for simplicity.
3Includes tortilla chips.
4EJsewhere called "Vegetables" for simplicity.
5Includes potato chips.
6EJsewhere called "Fruits" for simplicity.
7 Elsewhere called "Milk" for simplicity.
8EJsewhere called "Meat" for simplicity.
• p s; .05 between White and Black.
28 Family Economics and Nutrition Review
Table Sa. Consumption of food groups by all teenagers for meals and snacks (Food Guide Pyramid servings),
1 day, CSFII 1989-911
Breakfast Lunch Su1212er Snacks
Total White Black Total White Black Total White Black Total White Black
Food group (n=926) (n=735) (n=l91) (n=926) (n=735) (n=l91) (n=926) (n=735) (n=191) (n=926) (n=735) (n=l91)
Grains 1.82 1.81 1.85 1.70* 1.79 1.26 2.23 2.26 2.06 0.61 * 0.68 0.27
Vegetables 0.06 0.07 0.01 0.57 0.60 0.44 1.31 1.29 1.38 0.05 0.05 0.02
Fruits 0.34* 0.37 0.21 0.17 0.18 0.15 0.12 0.12 0.14 0.20* 0.23 0.07
Milk 0.67 0.68 0.63 0.52* 0.56 0.37 0.53* 0.61 0.17 0.26 0.27 0.20
Meat 0.17 0.17 0.20 1.09 1.10 1.05 2.35 2.24 2.83 0.10 0.10 0.11
Table Sb. Consumption of food groups by male teenagers for meals and snacks (Food Guide Pyramid servings),
1 day, CSFII 1989-911
Breakfast Lunch Supper Snacks
Total Total Total Total
males White Black males White Black males White Black males White Black
Food group (n=460) (n=381) (n=79) (n=460) (n=381) (n=79) (n=460) (n=381) (n=79) (n=460) (n=381) (n=79)
Grains 2.24 2.22 2.34 2.00 2.05 1.63 2.53 2.54 2.44 0.93* 0.99 0.53
Vegetables 0.10 0.11 0.02 0.62* 0.67 0.26 1.49 1.40 2.14 0.06 0.07 0.04
Fruits 0.43* 0.47 0.15 0.17 0.17 0.11 0.12 0.13 0.04 0.25* 0.28 0.06
Milk 0.79 0.77 0.93 0.58* 0.62 0.32 0.61 * 0.67 0.23 0.39* 0.43 0.12
Meat 0.21 0.20 0.27 1.30 1.39 0.76 2.67 2.63 2.95 0.19 0.19 0.21
Table Sc. Consumption of food groups by female teenagers for meals and snacks (Food Guide Pyramid servings),
1 day, CSFII 1989-911
Breakfast Lunch Supper Snacks
Total Total Total Total
females White Black females White Black females White Black females White Black
Food group (n=466) (n=354) (n=112) (n=466) (n=354) (n=ll2) (n=466) (n=354) (n=112) (n=466) (n=354) (n=112)
Grains 1.38 1.34 1.54 1.38* 1.48 1.04 1.91 1.93 1.83 0.56 0.59 0.49
Vegetables 0.02 0.03 0 0.53 0.52 0.55 1.11 1.17 0.91 0.05* 0.06 0.01
Fruits 0.25 0.26 0.24 0.18 0.18 0.17 0.13 0.11 0.19 0.17* 0.20 0.07
Milk 0.54 0.57 0.45 0.47 0.49 0.40 0.44* 0.53 0.13 0.21 0.18 0.33
Meat 0.14 0.13 0.15 0.87 0.77 1.24 2.00* 1.79 2.76 0.09 0.10 0.07
1Weighted data.
* P S .05 between White and Black.
1997 Vol.JO No.3 29
Nevertheless, the results of this study
correspond with previous research. The
findings on nutrient consumption are
generally consistent with those obtained
from the Third National Health and
Nutrition Examination Survey (NHANES
III), collected between 1988-91. In this
study, females were also more likely to
have lower-than-recommended intakes
of essential vitamins and minerals than
males ( 1 ). Low intakes of calcium and
iron were especially notable. Although
nutrient intakes are calculated based
on food intake only, given that only
one-quarter of teens take vitamin and
mineral supplements at least occasionally,
it is unlikely that supplementation com­pensates
for low nutrient intakes from
food in the majority of teens.
Nutrient intakes below the RDAs do not
necessarily mean that intakes are below
requirements or that physiological
nutrient deficiencies necessarily exist.
The RDAs are recommended allowance
levels for population groups, not individual
requirements, and are set high intention­ally
to cover the needs of almost everyone
in a given sex-age group. Nevertheless,
when average intakes are below recom­mendations,
concern that diets are less
than optimal for at least some individuals
is reasonable. This is particularly true
for children and adolescents because
nutrient inadequacies during this period
of rapid growth may have far-reaching
effects. Most notably, low calcium
intakes by young females have been
associated with lower bone density, a
condition predisposing older women to
osteoporosis (25 ).
Figure 3. Meals skipped by teenagers
Percent
100
80
60
40
20
0
O white
• Black
Total Males Females
sample
Breakfast
• p .s .05 between White and Black.
30
Total Males Females Total Males Females
sample* sample
Lunch Supper
Deficiencies in trace minerals, including
iron and zinc, may impair physical
growth and/or cognitive performance.
Very low levels of iron in the blood can
result in iron deficiency anemia, which
is characterized by fatigue and weakness.
Although anemia is not common,
many adolescent females are at risk
for symptoms linked to moderate iron
deficiencies, including fatigue ( 8) and
impaired school performance ( 4 ).
Deficiencies in zinc in developed
countries are also likely to be mild,
but some research suggests that even
moderate deficiencies may affect
growth and/or immune function (7).
Teens' high proportions of energy
(calories) from fat, saturated fat, and
cholesterol may portend adult problems.
Diets high in fat and cholesterol are
strongly associated with heart disease
in adults, the leading cause of death
in the United States. The build-up of
atherosclerotic plaque, a precursor of
heart disease, has been reported in children
and adolescents, suggesting that early
diet patterns can set the stage for heart
attacks or strokes in later years. High
fat intakes may also be associated with
obesity. Research tracking health records
of young children into adulthood suggests
that obesity during adolescence elevates
health risks in adulthood, even if the over­weight
teen slims down in later years.
According to the study, teenaged boys
who were overweight were twice as
likely to have died or developed heart
disease by age 70, compared with their
slender peers. The heavier boys were
also more likely to develop colon and
rectal cancer and gout as adults. Teen­aged
girls who were heavy had greater
problems walking, climbing stairs, and
lifting heavy objects as older adults, and
they were more likely to develop arthritis,
compared with their slimmer cohorts
(24).
Family Economics and Nutrition Review
The eating patterns of teens provide
insight into possible explanations for
the low nutrient intakes identified in
this study. For example, in teenage girls'
diets the lack of choices from the milk
group and dark-green leafy vegetables
parallels their low consumption of
calcium. Similarly, females' low con­sumption
of grains may be one reason
that average iron intakes are below
recommendations since grain products
contributed 49 percent of iron in the
U.S. food supply in 1990 (14). Diets
lacking in foods from the bread group,
fruits, and vegetables also explain low
consumption of dietary fiber by both
males and females.
That teens had adequate consumption
of vitamin C despite the lack of citrus
fruits in their diets may seem surprising.
Citrus fruits and juices contributed
approximately 24 percent of vitamin C
to the U.S. food supply in 1990 (14).
However, it appears that teenagers are
more likely to consume vitamin C-rich
vegetables than fruits. Specifically,
pot;:ttoes (often in the form of french
fries) and tomatoes used in sauces (for
example, on pizza) are vitamin C sources
that feature prominently in the diets of
teenagers.
A review of foods consumed (table 4)
illustrates the need for promotions geared
to encouraging greater consumption of
fruits, vegetables, and foods from the
milk group. Fruits stand out as the least
favored food group among teens, a finding
recently replicated by Cleveland et al.
(6). All subgroups in this sample con­sumed,
on average, less than half the
servings recommended, and consumption
was lowest among Black male youths.
Interventions that encourage more
consumption of fruits and vegetables,
such as "Gimme 5," a curriculum-based
program, appear to be successful for
1997 Vol.JO No.3
children and adolescents ( 11,27). Among
the successful elements of this and other
campaigns, the use of social modeling
and development of increased behavioral
self-efficacy (i .e. , increased belief in
one's ability to perform a behavior) are
techniques likely to translate well to
campaigns aimed at teens. Social modeling
techniques could associate eating fruits
and vegetables with peer or celebrity
models. Behavioral self-efficacy could
be increased by suggesting easy ways to
eat more fruits and vegetables that fit
teens' eating patterns-for example,
choosing fruit for a snack, since most
teens eat snacks daily.
Table 5 and figure 3 also highlight
opportunities for intervention. For
example, the majority of Black males
eat breakfast, but their choices generally
leave out fruits. Campaigns directed to
this group can focus on making healthy
food choices more desirable and easy to
accommodate for breakfast.
In contrast, one-third of the Black females
in this sample skipped breakfast, a meal
that can easily incorporate milk, fruit,
and whole grains-good sources of
vitamins and minerals often lacking in
this population. Breakfast was also the
meal most likely to be skipped among
White teenaged girls. Reasons for break­fast
skipping may include concerns
about controlling body weight and time
constraints. Messages targeted to Black
and White females could address such
reasons; for example, emphasizing the
benefits of having breakfast and its role
in healthy weight control, and providing
tips on how to fit a good breakfast into
a busy morning.
Similarly, opportunities for increasing
consumption of fruits, vegetables, and
calcium-rich foods among Black teens
could be met by encouraging them to
eat lunch. Effective campaigns would
require further research with this group
to discern reasons for their tendency to
skip lunch and appropriate strategies to
make healthy lunches appealing.
Another intervention suggested by the
findings could focus on introducing
teens to lowfat yogurt. Black teens may
have a higher incidence of lactose intol­erance
than other teens, which positions
yogurt as a particularly appropriate
source of calcium for this group.
However, in this study, yogurt was
consumed by only 3 percent of White
teens on the survey day and was virtually
untouched by Black teens. Campaigns
that encourage eating yogurt as well as
other good sources of calcium can help
bridge the gap between recommended
and reported consumption of this nutrient.
Conclusions
Frequently, dietary assessment of popu­lation
groups has focused on intakes of
vitamins, minerals, and other dietary
components. This level of assessment
is important for the identification of
nutritional problems; however, it offers
little insight into potential solutions.
Individuals eat foods, not nutrients;
therefore, nutrient intake levels are
best considered outcomes of a complex
sequence of behaviors including food
selection and meal patterns. This study
not only identified nutritional problems
with teenagers' diets but also identified
associated food choices and meal pattern
behaviors. Combining this information
can enhance our ability to improve
dietary status by identifying behaviors
associated with less-than-optimal
nutrient intakes and suggesting possible
strategies for change. It is hoped that
these findings will aid in the development
of effective interventions to improve the
nutritional well-being of adolescents.
31
32
References
1. Alaimo, K., McDowell, M.A., Briefel, R.R., Bischof, A.M., Caughman, C.R.,
Loria, C.M., and Johnson, C.L. 1994. Dietary intake of vitamins, minerals, and fiber
of persons ages 2 months and over in the United States: Third National Health and
Nutrition Examination Survey, Phase I , 1988-91. Advance Data, Number 258.
2. Biing-Hwan, L., Guthrie, J., and Blaylock, J.R. 1996. The Diets of America's Children:
Influence of Dining Out, Household Characteristics, and Nutrition Knowledge.
Agricultural Economic Report No. 746. U.S. Department of Agriculture, Economic
Research Service.
3. Bronner, Y.L. 1996. Nutritional status outcomes for children: Ethnic, cultural,
and environmental contexts. Journal of the American Dietetic Association 96:891-
900,903.
4. Bruner, A.B., Joffe, A., Duggan, A.K., Casella, J.F., and Brandt, J. 1996.
Randomised study of cognitive effects of iron supplementation in non-anaemic
iron-deficient adolescent girls. The Lancet 348:992-996.
5. Centers for Disease Control (CDC). 1994. Prevalence of overweight among
adolescents-United States, 1988-1991 . Morbidity and Mortality Weekly Report
43:818-821.
6. Cleveland, L.E., Cook, A.J., Wilson, J.W., Friday, J.E., Ho, J.W. , and Chahil, P.S.
1997. Pyramid Servings Data: Results from USDA's 1994 Continuing Survey of
Food Intakes by Individuals [Online]. ARS Food Surveys Research Group. Available
(under ''Releases''): http://www.barc.usda.gov/bhnrc/foodsurvey!home.htm.
7. Cousins, R.J. and Hempe, J.M. 1990. Zinc. In M.H. Brown (Ed.), Present Knowledge of
Nutrition, (pp. 251-260). International Life Sciences Institute, Nutrition Foundation,
Washington, DC.
8. Dallman, P.R., Buetler, E., and Finch, C.A. 1978. Effect of iron deficiency exclusive
of anemia. British Journal of Heamatology 40:179-184.
9. Devaney, B.L., Gordon, A.R., and Burghardt, J.A. 1995. Dietary intakes of students.
The American Journal of Clinical Nutrition 61 (suppi):205S-212S.
10. Dietz, W.H. Jr. and Gortmaker, S.L. 1985. Do we fatten our children at the
television set? Obesity and television viewing in children and adolescents. Pediatrics
75:807-812.
Family Economics and Nutrition Review
II. Dome!, S.B., Baranowski, T., Davis, H., Thompson, W.O., Leonard, S.B., Riley, P.,
Baranowski, J., Dudovitz, B., and Smyth, M. 1993. Development and evaluation of
a school intervention to increase fruit and vegetable consumption among 41h and 51h
grade students. The Journal of Nutrition Education 25(6):345-349.
12. Eck, L.H. and Hackett-Renner, C. 1992. Calcium intake in youth: Sex, age and
racial differences in NHANES II. Preventive Medicine 21:473-482.
13. Fleming, K.H. and Heimbach, J.T. 1994. Consumption of calcium in the U.S.:
Food sources and intake levels. Journal of Nutrition 124(8 suppl): 1426S-1430S.
14. Gerrior, S.A. and Zizza, C. 1994. Nutrient Content of the U.S. Food Supply 1909-90.
Home Economics Research Report No. 52. U.S. Department of Agriculture.
15. Gordon, A.R. and McKinney, P. 1995. Sources of nutrients in students' diets.
American Journal of Clinical Nutrition 61 (suppl):232S-240S.
16. Guthrie, J.F. and Fulton, L. 1995. Relationship of knowledge offood group servings
recommendations to food group consumption. Family Economics and Nutrition
Review 8(4):2-17.
17. Himes, J .H. and Dietz, W .H. 1994. Guidelines for overweight in adolescent
preventive services: Recommendations from an expert committee. The American
Journal of Clinical Nutrition 59(2):307-316.
18. Johnson, R.K., Johnson, D.G., Wang, M.Q., Smicklas-Wright, H., and Guthrie,
H. A. 1994. Characterizing nutrient intakes of adolescents by sociodemographic
factors. Journal of Adolescent Health 15:149-154.
19. Krebs-Smith, S.M., Cook, A., Subar, A.F., Cleveland, L., Friday, J., and Kahle,
L.L. 1996. Fruit and vegetable intakes of children and adolescents in the United
States. Archives of Pediatric Adolescent Medicine 150:81-86.
20. Life Sciences Research Office (LSRO), Federation of American Societies for
Experimental Bwlogy. 1989. Nutrition Monitoring in the United States-An Update
Report on Nutrition Monitoring. Prepared for the U.S. Department of Agriculture
and the U.S. Department of Health and Human Services. DHHS Publication No.
(PHS) 89-1255.
21. Looker, A.C., Loria, C.M., Carroll, M.D., McDowell, M.A., and Johnson, C.L.
1993. Calcium intakes of Mexican Americans, Cubans, Puerto Ricans, non-Hispanic
whites, and non-Hispanic blacks in the United States. Journal of the American
Dietetic Association 93(1 J): 1274-1279.
1997 Vol.l0No.3 33
34
22. Mathematica Policy Research, Inc. 1989. Documentation for a SAS analysis file
containing merged 4-day data for the core and low-income samples of the 1985
CSFII. Contract no. FNS-53-3198-5-41 under subcontract to the Urban Institute.
U.S. Department of Agriculture, Food and Nutrition Service, Alexandria, VA.
23. Mertz, W., Tsui, J.C., Judd, J.T., Reiser, S., Hallfrisch, J., Morris, E.R., Steele,
P.D., and Lashley, E. 1991. What are people really eating? The relation between
energy intake derived from estimated diet records and intake determined to maintain
body weight. American Journal of Clinical Nutrition 54:291-295.
24. Must, A., Jacques, P.F., Dallal, G.E., Bajema, C.J., and Dietz, W.H. 1992. Long­term
morbidity and mortality of overweight adolescents. A follow-up of the Harvard
Growth Study of 1922 to 1935. The New England Journal of Medicine
327( 19): 1350-1355.
25. National Academy of Sciences, National Research Council, Food and Nutrition
Board. 1989. Recommended Dietary Allowances (lOth ed.). National Academy
Press, Washington, DC.
26. National Academy of Sciences, National Research Council, Food and Nutrition
Board. 1989. Diet and Health: Implications for Reducing Chronic Disease Risk.
National Academy Press, Washington, DC.
27. Nicklas, T.A., Johnson, C.C., Farris, R., Rice, R., Lyon, L., and Shi, R.
(in press.) Development of a school-based nutrition intervention for high school
students: Gimme 5. American Journal of Health Promotion.
28. Obarzanek, E., Schreiber, G.B., Crawford, P.B., Goldman, S.R., Barrier, P.M.,
Frederick, M.M., and Lakatos, E. 1994. Energy intake and physical activity in
relation to indexes of body fat: The National Heart, Lung, and Blood Institute
Growth and Health Study. American Journal of Clinical Nutrition 60{1): 15-22.
29. Robinson, T.N., Hammer, L.D., Killen, J.D., Kraemer, H.C., Wilson, D.M.,
Hayward, C., and Taylor, C.B. 1993. Does television viewing increase obesity
and reduce physical activity? Cross-sectional and longitudinal analyses among
adolescent girls. Pediatrics 91:273-280.
30. Shah, B.V., Barnwell, B.G., Hunt, P.N., and LaVange,L.M. 1991. SUDMN
User's Manual: Professional Software for Survey Data Analysis for Multi-Stage
Sample Designs. Research Triangle Institute, Research Triangle Park, NC.
31. Shaw, A., Fulton, L., Davis, C., and Hogbin, M. 1996. Using the Food Guide
Pyramid: A Resource for Nutrition Educators [Online]. U.S. Department of
Agriculture, Center for Nutrition Policy and Promotion. Available:
http :1/www. usda. gov /fcs/cn pp.h tm.
Family Economics and Nutrition Review
32. SPSS-X User's Guide (3rd ed.). 1988. SPSS, Inc., Chicago.
33. Troiano, R.P., Flegal, K.M., Kuczmarski, R.J., Campbell, S.M., and Johnson, C.L.
1995. Overweight prevalence and trends for children and adolescents. Archives of
Pediatric and Adolescent Medicine 149:1085-1091.
34. U.S. Department of Agriculture, Center for Nutrition Policy and Promotion.
1995. The Healthy Eating Index. CNPP-1.
35 . U.S. Department of Agriculture, Human Nutrition Information Service. 1992.
The Food c;uide Pyramid. Home and Garden Bulletin No. 252.
36. U.S. Department of Agriculture, Human Nutrition Information Service. 1992.
USDA Nutrient Data Base for Individual Intake Surveys, release 7. Computer tape,
accession no. PB94-504552GEI.
37. U.S. Department of Agriculture, Human Nutrition Information Service. 1994.
Manual of food codes for individual intake. On 1991 Continuing Survey of Food
Intakes by Individuals and 1991 Diet and Health Knowledge Survey. Computer
tape, accession no. PB94;500063.
38. U.S. Department of Agriculture and U.S. Department of Health and Human
Services. 1995. Nutrition and Your Health: Dietary Guidelines for Americans. (4th ed.).
U.S. Department of Agriculture. Home and Garden Bulletin No. 232.
39. Williams, C.L. 1995. Importance of dietary fiber in childhood. Journal of the
American Dietetic Association 95( 10 ): 1140-1146, 1149.
40. Wright, H.S., Guthrie, H.A., Wang, M.Q., and Bernardo, V. 1991. The 1987-88
Nationwide Food Consumption Survey: An update on the nutrient intake of respondents.
Nutrition Today 26(3):21-27.
1997 Vol.l0No.3 35
36
Lower Fat Food Choices
Identified Through a Novel
Sorting Procedure for CSFII
1989-91 Data
Sharon Peterson
Madeleine Sigman-Grant
Cheryl Achterberg
The Pennsylvania State University
A unique sorting procedure for the 1989-91 Continuing Survey of Food
Intakes by Individuals is described. Following sorting, the extent of adoption
of skim milk, lean meats, and fat-modified products and the demographic
characteristics of users and nonusers of these fat-reduction strategies were
examined using the "Diffusion of Innovations" framework. Americans were in
the early stages of adopting these strategies. Exclusive use of skim milk and
fat-modified products increased over time while exclusive use of lean meats
decreased. Multivariate logistic regression indicated skim milk users were
more likely to be older, Caucasian, and have greater incomes than other
milk users. Lean meat users were more likely to be female and older than
other meat users. Fat-modified product users and multiple strategy users
were more likely to be older. To increase use of fat-reduction strategies,
nutrition educators should target those most likely to be early adopters and
encourage successful users to influence peer food choices.
ertain dietary patterns are
associated with increased
risk of chronic disease. To
reduce the risk, several
national nutrition education campaigns
promoted lower fat diets to American
consumers during the late 1980's and
early 1990's (5,16,19,25,26). These
projects shared the goal of reducing
total fat intake of Americans to 30 percent
of total energy or less per day, with the
assumption that by doing so, incidence
of chronic disease would decrease.
Health professionals translated dietary
recommendations into food-based fat­reduction
strategies that, in turn, were
disseminated to the public. A common
approach was to target major food
sources of total fat and saturated fat,
such as whole milk, high-fat meats, and
added fats and oils, with messages to
use lower fat versions (2,6,8,11,21,22,
25,27). Skim can be used to replace
whole, 2-percent, and !-percent milks.
Lean cuts of meat, prepared with little
or no fat, can be used to replace higher
Family Economics and Nutrition Review
fat meats. Commercial fat-modified and
fat-free products, recent additions to the
marketplace, can replace the added fats
and oils in high-fat foods such as salad
dressing, mayonnaise, margarine, ice
cream, cheese, and baked goods. In
order for any of these choices to impact
the incidence of chronic disease, they
must be used consistently.
The development and dissemination
of innovative messages about lower fat
food choices create the potential for
their adoption within the population. The
impact of these messages on the food
choices of Americans can be analyzed
using the Diffusion of Innovations
concept ( 18 ). According to Rogers, an
innovation is an idea, practice, or object
that is perceived as new by an individual
or group, and diffusion is the planned
or spontaneous process by which an
innovation is communicated through
certain channels over time among the
members of a social system.
Americans have had the opportunity
to interpret and implement common
fat-reduction strategies without profes­sional
guidance. The extent of exclusive
adoption of common fat-reduction
strategies by free-living Americans­and
who is most likely to use them-is
unknown. The Continuing Survey of
Food Intakes by Individuals (CSFII) is
an ideal data base to use for determining
these food choice trends. Traditional
analytical procedures for examining
nationwide consumption studies sort by
single nutrient (e.g., total fat or vitamin
C) or by food grouping (e.g., all beef
products). In order to determine the
impact of using specific lower fat food
choices, a different sorting procedure
is required. The purpose of this report
is (I) to describe the steps involved in
1997 Vol.JO No.3
developing data subsets using the 1989-
91 CSFII food codes (24) that represent
specific lower fat and higher fat food
choices; (2) to describe the extent of
adoption of skim milk, lean meats,
and fat-modified products and change
in adoption rates from 1989-91; and
(3) to compare the demographics of
those who have and have not adopted
these fat-reduction strategies.
Methods
Sorting Procedures
The specific comparison groups used
in this study were: Users of skim milk
versus users of !-percent, 2-percent,
or whole milk (hereafter referred to as
higher fat milks); users of lean meats
versus users of medium-fat and high-fat
meats (including poultry, beef, pork,
lamb, fish, wild game, eggs, and peanut
butter as categorized by the American
Diabetes and American Dietetics Asso­ciations
exchange system and hereafter
referred to as higher fat meats); and
users of fat-modified cheeses, salad
dressings, yogurts, cakes, and puddings
(hereafter referred to as fat-modified
products) versus users of full-fat cheeses,
salad dressings, yogurts, cakes, and
puddings (hereafter referred to as full­fat
products). These products were the
only specific food types in the 1989-91
CSFII that included both fat-modified
and full-fat versions.
It should be noted that during the years
when these surveys were undertaken,
there was much confusion regarding
label terminology. Standardization of
the definition of "lowfat" did not occur
until 1994 when final implementation
of the Nutrition Labeling and Education
Act was required for all processed foods.1
1 Fluid 2-percent milk was included among lowfat
foods until 1996.
In addition, the availability and con­sumption
of !-percent milk during the
survey years were limited. Therefore,
!-percent milk is included in the higher
fat milk group. The inclusion of lowfat
cheese as a fat-modified product was
predicated on the theoretical computer
model in which milks were separated
from other dairy products and all other
commercially available fat-modified
foods were placed into one category.
The initial step was to write three hierar­chical
sets of SAS (Statistical Analysis
System) programs to read the 1989-91
CSFII data tapes and create new data
subsets. All respondents who completed
3 days of intake data were retained as
use of multiple days of intake is pre­ferred
over a single record to reduce
within-person variability and to allow
for more accurate assessment of between­person
variability (7). Each year was
then searched for food codes representing
type of milk used, type of meat used,
and type of products used. Cooking
methods were considered and mixed
meat dishes were included when the
description of the food clearly identified
the respondents' food choice. Only food
codes that clearly identified users and
nonusers of the targeted fat-reduction
strategies were used; thus, descriptions
that included "non-specified type" or
"non-specified cooking method" were
not included in any group. Those who
reported use of some lower fat food
choices and some higher fat food
choices formed a third comparison
group for each analysis and were called
"mixed" users.
37
38
Use of skim milk
increased from
6.7 percent in 1989
to 9.7 percent in
1991 ....
When using food codes to sort data, a
person has a new record for each food
eaten with key demographic variables
repeated for each food record. There­fore,
many lines of data represent any
individual person. After sorting by food
choices, data sets were converted from
food code level records to person level
records by retaining only one copy of
each person's record. Each data set
included exclusive users, mixed users,
and nonusers for each fat-reduction
strategy. A total of nine data sets were
created, representing each of the three
fat-reduction strategies for each survey
year. Demographic data then were
added to the data sets by matching
ID numbers across record types.
In addition to comparing pure users,
mixed users, and nonusers of each fat­reduction
strategy, a measure of overall
strategy use was developed. This required
examination of each person's food choice
pattern across strategies. First, the three
data sets (milk use, meat use, and product
use) were combined into one set for
each survey year. A total of 63 possible
combinations of food choices for milk,
meat, and products existed. These
combinations were used to define single
strategy users, multiple strategy users,
and nonusers.
Single strategy users included: Exclusive
use of skim milk but not exclusive use
of lean meats or fat-modified products;
exclusive use of lean meats but not ex­clusive
use of skim milk or fat-modified
products; or exclusive use of fat-modified
products but not exclusive use of skim
milk or lean meats. Multiple strategy
users included: Exclusive use of skim
milk and exclusive use of lean meats;
exclusive use of skim milk and exclusive
use of fat-modified products; exclusive
use of lean meats and exclusive use
of fat-modified products; or exclusive
use of skim milk, lean meats, and fat­modified
products. Users of no strategies
included those who used !-percent,
2-percent, or whole milk; those who
used medium-fat or high-fat meats;
and/or those who used full-fat versions
of cheese, yogurt, salad dressing, cake,
or pudding.
Demographic Variables
Independent variables were chosen
based on their inclusion in the CSFII
data base and frequent measurement in
Diffusion of Innovations studies ( 18).
Thus, gender, race (White, non-White),
education (head of household), house­hold
income (percent of poverty), and
age served as independent variables
for bivariate analyses. Education was
categorized as: 11th grade or less, 12th
grade (i.e., high school graduate), 13-15
years (i.e., some college), 16 years
(i.e., college graduate), and 17 years
(i.e., formal schooling beyond college
graduation). Income as a percent of
poverty was categorized as: 130 percent
of poverty or less (130 percent represents
the cutoff for food stamps), 131-300 per­cent
of poverty (300 percent represents
a conventional cutoff used by USDA
for dividing middle income and upper
middle income ( 4 )), 301-600 percent
of poverty (representing upper middle
income), and 601 percent or greater
(representing upper income). Age
was categorized as: 2- to 1 0-year-olds,
11- to 19-year-olds, 20- to 25-year-olds,
26- to 35-year-olds, 36- to 45-year-olds,
46- to 65-year-olds, and 66- to 1 00-year­olds.
Family Economics and Nutrition Review
Statistical Analyses
Chi square analysis was used to deter­mine
statistically significant differences
in rates of adoption among strategies
and in strategy adoption rates across
the 3 years. USDA-calculated sampling
weights were used to correct for over­sampling
of low-income groups in the
1989-91 CSFII. Resultant values after
weighting then were adjusted to reflect
the actual number of cases by dividing
the respondents' sampling weight by
the mean sampling weight for the group
( 13).
Bivariate analyses of the relationship
of each independent variable to milk
use, meat use, product use, and strategy
use were completed. Due to insufficient
expected frequencies, the age categories
2 to l 0 years and 11 to 19 years were
combined for 1989 and 1990 analyses
of age and strategy use, and the 1990
analysis of age and product use.
Multivariate logistic regression was used
to determine the demographic profile
most strongly associated with milk use,
meat use, product use, and overall strategy
use. Interrelationships among demo­graphic
characteristics were checked
for multicollinearity, resulting in the
elimination of education of head of
household in the multivariate analyses.
Each multivariate analysis compared
pure users with mixed and nonusers.
For overall strategy use, multiple strategy
users were compared with users of one
strategy or none.
Unweighted data were used for the
logistic regression analyses ( 10). A
p-value of .01, more conservative than
standard practice, was used to determine
significant relationships in an attempt
to compensate for the design effect of
this large, complex survey.
1997 Vol.JONo.3
Results
Although the separation into specific
groupings based on fat level of milk
and the above-mentioned food products
was relatively easy, division of meat
was very complex. Consideration for
cut of meat as well as preparation method
have to be taken into account when
categorizing food codes for meat by fat
level. A representative list of food codes
used to determine users and nonusers of
each fat-reduction strategy is presented
in table 1, pp. 40-41. The final data
set was composed of all adults who
completed 3 days of intake and reported
use of some type of milk, some type of
meat, or some type of cake, pudding,
salad dressing, yogurt, or cheese. The
total number of respondents for each
strategy and for overall number of
strategies by year of survey is presented
in figure 1, p. 42.
Demographic profiles of all strategy
users and nonusers for each survey year
are presented in table 2, p. 43. Signifi­cant
relationships between year and milk
use (p<O.OOl), year and product use
(p<O.OOI ), year and meat use (p<O.Ol),
and year and overall strategy use
(p<0.001) were identified (figs. 2-5,
pp. 44, 45). Use of skim milk increased
from 6.7 percent in 1989 to 9.7 percent
in 1991, while use of higher fat milks
decreased from 69.6 percent to 61.2
percent (fig. 2). Use of fat-modified
products increased from 2.1 percent
to 2.8 percent. Use of full-fat products
increased from 55 .8 percent to 59.1 per­cent
and use of no products decreased
when comparing the first year with the
third year (fig. 3). Use of lean meats
decreased from 3.3 percent to 2.9 percent
and use of higher fat meats increased
from 64.6 percent to 67.4 percent when
comparing the first year of the survey
with the third year (fig. 4). Overall use
of any one strategy increased from
10.5 percent to 12.6 percent. Use of
more than one strategy increased slightly
from 0.8 percent to 1.4 percent, while
use of no fat-reduction strategies
decreased from 88.7 percent in 1989
to 86.0 percent in 1991 (fig. 5).
All bivariate relationships of milk use,
meat use, product use, or strategy use
and each of the independent variables
were significant for all 3 years at the
p<O.Ol level except for gender and milk
use in the 1990 survey (data not shown).
Results of the multivariate logistic
regressions indicate that exclusive users
of skim milk were significantly more
likely to be older, Caucasian, and have
greater incomes when compared with
higher fat milk and mixed milk users
for all 3 years (table 3, p. 46). For 2 of
the 3 years, skim milk users were also
significantly more likely to be female.
Exclusive users of lean meats were
significantly more likely to be older
and female when compared with users
of higher fat and mixed meats for all 3
years (table 3). For 2 of the 3 years, lean
meat users were also significantly more
likely to have greater incomes. Users of
only fat-modified products were signifi­cantly
more likely to be older when
compared with users of full-fat and
mixed products for all 3 years (table 3).
Users of multiple strategies were signifi­cantly
more likely to be older than users
of one or no strategies for all 3 years
(table 4, p. 47). For 2 of the 3 years,
users of multiple strategies were also
significantly more likely to be female
when compared with users of one or no
strategies. Income expressed as dollars
instead of percent of poverty had no
impact on the results of the multivariate
model.
39
Table 1. Foods used to sort respondents of the Continuing Survey of Food Intakes by Individuals, 1989-91,
into users and nonusers of fat-reduction strategies
Selected foods to define use of
Skim milk (includes all foods used):
Milk, cow, fluid, skim or nonfat
Milk, low lactose, nonfat
Milk, dry reconstituted, nonfat
Fat-modified products (43 food codes actually used):
Yogurt, plain, nonfat milk
Yogurt, frozen, not chocolate, nonfat milk
Cheese, Swiss, lowfat
Cheese, cottage, lowfat
Cheese, cream, lowfat
Cheese, process~d, American, lowfat
Cake, pound, very lowfat, no cholesterol,
including Entemann's
Pudding, chocolate, reduced fat, including Jell-0 light
Salad dressing, French type, reduced calorie, fat-free,
cholesterol-free
Salad dressing, low calorie, oil-free
40
Selected foods to define use of
Higher fat milks (includes all foods used):
Milk, cow, fluid, whole
Milk, cow, fluid, whole, low sodium
Milk, cow, fluid, whole, fortified with calcium
Milk, cow, fluid, !-percent fat, fortified with calcium
Milk, cow, fluid, lowfat, nonspecific percent fat
Milk, cow, fluid, 2-percent fat
Milk, cow, fluid, acidophil us, !-percent fat
Milk, cow, fluid, acidophilus, 2-percent fat
Milk, cow, fluid, 1-percent fat
Milk, low lactose, !-percent fat
Milk, low lactose, !-percent fat, fortified with calcium
Milk, cow, fluid, lactose-reduced, 2-percent fat, lactaid
Milk, dry, reconstituted, whole
Milk, dry, reconstituted, lowfat
Milk, cow, fluid, filled with vegetable oil, whole
Milk, cow, fluid, filled with vegetable oil, lowfat
Full-fat products (226 food codes actually used):
Yogurt, vanilla, lemon, coffee, whole milk
Yogurt, frozen, nonspecific flavor, lowfat milk
Cheese, cheddar or American
Cheese, Swiss
Cheese, cottage, dry curd
Cheese, cream
Cheese, processed, Swiss
Cake, black forest
Cheesecake
Cake, German chocolate
Cake, marble, with icing
Pudding, canned, chocolate
Salad dressing, French type
Salad dressing, bacon and tomato
Family Economics and Nutrition Review
Table 1. Foods used to sort respondents of the Continuing Survey of Food Intakes by Individuals, 1989-91,
into users and nonusers of fat-reduction strategies (cont'd)
Selected foods to define use of
Lean meat ( 185 food codes actually used):
Ham, smoked/cured, cooked, lean only
Pork tenderloin, baked
Venison, roasted
Chicken breast, broiled, without skin
Turkey, light meat, cooked without skin
Cornish game hen, roasted without skin
Selected foods to define use of
Higher fat meat (709 food codes actually used):
Beef steak, fried, lean and fat
Beef steak, battered, fried, lean only
Ground beef or patty
Pork chop, fried, lean and fat
Turkey ham, sliced, extra lean, prepackaged or deli-sliced
Carp, broiled, baked
Pork spareribs, cooked, lean and fat
Lamb roast, cooked, lean and fat
Veal chop, fried, lean and fat
Chicken, boneless, breaded, baked/fried, with skin
Cod, steamed, poached Chicken leg, roasted, with skin
Tuna, fresh, steamed, poached
Clams, steamed, boiled
Discussion
Recent surveys have reported similar,
albeit higher, usage rates of lower fat
food choices. Of all respondents to the
ADA/International Food Information
Council survey who said they had made
changes with regard to fat in the diet,
36 percent said they were using lowfat
foods/less fat, 23 percent reported no/
less meat in the diet, 9 percent reported
fat-free foods/no fats, 5 percent said
they were eating more poultry, and
4 percent reported using skimllowfat
milk (3). Nearly 9 of 10 adult Americans
1997 Vol.JO No.3
Chicken thigh, battered, fried, with skin
Turkey, light and dark meat, stewed, with skin
Turkey nuggets
Beef liver, breaded, fried
Frankfurter, beef and pork
Fish sticks, battered, fried
Cod, floured, breaded, fried
Egg salad
Peanut butter sandwich
Hamburger, plain, on bun
Sausage and peppers, no sauce
(88 percent) reported regular use of fat­modified
food and beverages, including
use of Iowfat or skim milk by 66 percent,
according to a report by the Calorie
Control Council where "regular use"
was defined as use of a product at least
once every 2 weeks ( 14). It should be
emphasized that in this, as well as other
surveys, 2-percent and ]-percent milks
are combined with skim milk and
reported as "low fat" milk despite the
large differences in fat content between
each type.
Hence, since the objective of this study
was to describe exclusive use of fat­reduction
strategies, the observed
differences in reported usage rates are
expected. The sorting procedure developed
for this study was rather restrictive.
Qualifying as a strategy user required
consistent use of the strategy over 3 days.
For example, in order to qualify as a
lean meat user, a respondent must have
reported only lean meats across all 3
days of the survey and no use of medium­fat
or higher fat meats. Thus, any record
of ground beef across the 3 days auto­matically
removed a respondent from
41
~
N
=e"'.l
-<
M n
0 = 0 e.
12
=
5.
z -= :::!.
:::;
0 = ~
~ -<
~·
Figure 1. Total number of respondents in the 1989-91 Continuing Survey of Food Intakes by Individuals who
were identified for each type of fat-reduction strategy and overall strategy use by year of survey 1,2,3
5,746 respondents to
1989 CSFII survey
4,215 (73%) completed
3 days' intake
Milk Users: I Meat Users:
Skim=275 Lean=133
Mixed=135 Mixed=1 ,109
Higher fat=2,833 Higher fat=2,629
Product Users:
Fat-modified=86
Mixed=129
Full-fat=2,273
Overall strategy use:
None=3,611
One=429
Two/three=32
5,806 respondents to
1990 CSFII survey
3,601 (62%) completed
3 days' intake
Milk Users: I Meat Users: Product Users:
Skim=331 Lean=101 Fat-modified=30
Mixed=120 Mixed=1 ,004 Mixed=133
Higher fat=2, 182 Higher fat=2, 190 Full-fat=1 ,827
Overall strategy use:
None=3,024
One=407
Two/three=28
6,182 respondents to
1991 CSFII survey
4,096 (67%) completed
3 days' intake
Milk Users: Meat Users: Product Users:
Skim=385 Lean=116 Fat-modified=112
Mixed=191 Mixed=990 Mixed=216
Higher fat=2,425 Higher fat=2,668 Full-fat=2,341
Overall strategy use:
None=3,406
One=498
Two/three=57
1U.S. Department of Agriculture, Human Nutrition Information Service. 1992-94. Nationwide Food Consumption Survey, Continuing Survey of Food Intakes by Individuals:
1989-91 . Study No. 09010-073.
2Frequencies were weighted to correct for oversampling of low-income households.
3Numbers reflect usable 3-day intake records.
Table 2. Demo~raphic profile of all strategy users and nonusers by year of Continuing Survey of Food Intakes
by Individuals
1989 1990
Characteristic N Percent N
Gender
Male 1,961 48.2 1,834
Female 2,110 51.8 1,625
Age (years)
2- 10 612 15.0 521
II - 19 509 12.5 455
20-25 374 9.2 292
26-35 735 18.1 637
36-45 588 14.4 504
46 - 65 774 19.1 662
66 - 100 480 11.8 387
Education (head of household)
11th grade or less 883 21.7 668
12th grade 1,320 32.5 1,078
Some college 825 20.3 718
College graduate 497 12.2 529
Graduate school 542 13.3 461
Income (percent of poverty)
130 or less 697 17.3 590
131- 300 1,279 31.7 1,077
301 - 600 1,432 35.5 1,365
601 or more 630 15.6 409
Race
White 3,137 77.1 2,698
Non-White 934 22.9 760
1 Frequencies were weighted to correct for oversampling of low-income households.
the lean meats group. A single report
of skim milk during the 3 survey days
would classify a person as a skim milk
user as would consumption of skim
milk at every meal for 3 days; however,
use of skim milk at most meals with one
report of lower fat milk would classify a
person as a mixed-milk user.
1997 Vol.10No.3
This restrictive sorting procedure may
provide insight into recent questions
regarding the use of fat-modified products
and their impact on chronic disease, in
particular obesity (I). In order for an
innovation to begin rapidly diffusing
throughout a population and to affect
incidence rates, according to Rogers, it
must reach an adoption rate between 10
1991
Percent N Percent
53.0 1,853 46.8
47.0 2,108 53.2
15.1 580 14.6
13.2 505 12.7
8.5 363 9.2
18.4 658 16.6
14.6 615 15.5
19.1 788 19.9
11.2 452 11.4
19.3 765 19.6
31.2 1,102 28.2
20.8 895 22.9
15.3 578 14.8
13.4 562 14.4
17.1 703 17.7
31.3 1,276 32.2
39.7 1,474 37.2
11.9 507 12.8
78.0 3,094 78.1
22.0 867 21.9
and 25 percent (18 ). Demographic
characteristics of early adopters of these
fat-reduction strategies are congruent
with those predicted by Diffusion of
Innovations theory that consistently
predicts high levels of education and
income for the earliest adopters of an
innovation ( 18). Education and income
were significantly greater for exclusive
43
users of each fat-reduction strategy and
for multiple strategy users. Previous
studies have reported a relationship
between education or income and use
of lower fat food choices ( 12,17,20,23 ).
Income was a significant factor in pre­dicting
use of skim milk and use of lean
meats in the multivariate models.
There has been inconsistent evidence
about the relationship of age and innova­tiveness
across Diffusion of Innovations
studies ( 18 ). In this study, multivariate
analyses identified age as a positive
predictive variable for each fat-reduction
strategy as well as for users of multiple
fat-reduction strategies. Previous studies
have reported increased use of lower fat
diets with advancing age, suggesting
older Americans may be a highly
motivated group to target.
Generalizations about race and innova­tiveness
or gender and innovativeness
have not been made by Rogers ( 18).
Societal pressures on American women
for thinness suggest that females may
be more likely than males to use fat­reduction
strategies. In fact, results of a
1993 survey showed women were more
likely than men to report making a
change in their diet with regard to fat
(3). In this study, being female was a
positive predictor of use of skim milk
and use of lean meats when included in
multivariate models.
Differences in adoption rates can occur
for different innovations within a popu­lation,
which was the case for the various
fat-reduction strategies investigated in
this study. Use of skim milk in place
of higher fat milks consistently increased
across survey years, as use of fat -modified
products increased slightly. It will be
interesting to follow changes in the type
of milk consumed as the availability of
!-percent milk increases and consumers
44
Figure 2. Comparison of milk use by year using 3-day records
of participants in the 1989, 1990, and 1991 Continuing Survey of
Food Intakes by Individuals 1
Percent
100
80
60
40
20
0
No milk used
0 1989·90
D 1990·91
• 1991-92
Sklnmilkonty
Milk use
Mixedmtlk2 1%, 2%, or
whole milk
1Weighted Chi-square