PART C: METHODOLOGY
Beginning with the 1985 edition, the Department of Health
and Human Services (HHS) and the Department of Agriculture
(USDA) have appointed a Dietary Guidelines Advisory Committee
(DGAC) of prominent experts in nutrition and health to assist
in the preparation of the Dietary Guidelines for Americans.
This Committee has been an effective mechanism to obtain a
competent review of the science, recommendations from experts,
and broad public acceptance of the Dietary Guidelines.
The 2005 DGAC was established for the single, time-limited
task of reviewing the 2000 edition of Nutrition and Your
Health: Dietary Guidelines for Americans and determining
if, on the basis of current scientific and medical knowledge,
revision was warranted. The Committee determined that a revision
was warranted and developed nutrition and health recommendations
in this report to the Secretaries of HHS and USDA. The Committee
was dissolved upon delivery of this report.
Nominations were sought from the public through a Federal
Register notice published on May 15, 2003, and from Federal
agencies. Prospective members of the DGAC were expected to
be knowledgeable of current scientific research in human nutrition
and be respected and published experts in their fields. They
would be familiar with the purpose, communication, and application
of the Dietary Guidelines and have demonstrated interest
in the public's health and well-being through their research
and/or educational endeavors. Expertise was sought in specific
specialty areas, including, but not limited to, cardiovascular
disease, cancer, pediatrics, gerontology, epidemiology, general
medicine, overweight and obesity, physical activity, public
health, nutritional biochemistry, nutrient bioavailability,
nutrition education, and food safety and technology.
The Secretaries of HHS and USDA jointly selected individuals
for membership to the DGAC. The individuals selected are highly
respected by their peers for their depth and breadth of scientific
knowledge of the relationship between dietary intake and health.
The expertise of these individuals addresses all the relevant
areas of the current Dietary Guidelines for Americans.
To ensure that recommendations of the Committee took into
account the needs of the diverse groups served by HHS and
USDA, membership included, to the extent practicable, individuals
with demonstrated ability to represent minorities, women,
and persons with disabilities. Efforts were made to ensure
equitable geographic distribution and racial, ethnic, and
gender representation. Appointments were made without discrimination
on the basis of age; race and ethnicity; gender; sexual orientation;
disability; or cultural, religious, or socioeconomic status.
Equal opportunity practices, in line with HHS and USDA policies,
were followed in all membership appointments to the Committee.
Back to Top
CHARGE TO THE 2005 DGAC
The Dietary Guidelines for Americans provides science-based
eating and physical activity advice for healthy Americans over
age 2 years. The DGAC shall advise the Secretaries of HHS and
USDA whether revisions to the 2000 edition of Nutrition
and Your Health: Dietary Guidelines for Americans are warranted
on the basis of the preponderance of the scientific and medical
knowledge currently available.
The Committee, whose duties are solely advisory and time-limited,
will perform the following functions:
If the Committee decides that no changes are necessary,
it will so inform the Secretaries of the Departments. This
action will terminate the DGAC.
If the Committee decides that changes are warranted
on the basis of the preponderance of the scientific and medical
knowledge, the Committee will determine what issues for change
need to be addressed.
The focus of the Committee should be on the review
of the new scientific evidence.
The Committee shall make and submit its technical
recommendations and the rationale for these recommendations
in a report to the Secretaries. The Committee's focus should
be its recommendations and the supporting science rather than
translating the recommendations into a communication document.
Upon the submittal of the Committee's recommendations,
the DGAC will be terminated.
Back to Top
THE COMMITTEE PROCESS
The Committee served without pay and worked under the regulations
of the Federal Advisory Committee Act. It held public meetings,
announced in the Federal Register, in Washington, DC,
in September 2003 and in January, March, May, and August 2004.
Meeting summaries are available at www.health.gov/dietaryguidelines.
Written public comments were received throughout the Committee's
deliberations. Those received before August 10, 2004, were
shared with all Committee members. Comments with recommendations
for the Committee received before May 12, 2004, are summarized
in Appendix G-6. In response to a solicitation for oral comments,
31 organizations or individuals presented oral testimony during
the January 28–29, 2004, meeting of the Committee. These comments
are summarized in the January Public Meeting Summary (http://www.health.gov/dietaryguidelines).
Comments are available for examination at the Office of Disease
Prevention and Health Promotion, 1101 Wootton Parkway, Suite
LL100, Rockville, MD, 20852.
To promote a fresh examination of the science base for dietary
guidance, the content areas to be addressed differed somewhat
from the topics of the 10 guidelines in the 2000 Dietary
Guidelines. In particular, the workload was divided and
managed by subcommittees on nutrient adequacy, carbohydrates,
fats, fluid and electrolytes, energy, ethanol, and food safety.
Midway through the effort, a macronutrient subcommittee was
appointed to address some crosscutting topics, and a subcommittee
was formed to address fruits and vegetables, grains, milk,
and milk products. To aid in coordination and communication,
a lead Committee member was appointed for each subcommittee,
but the conclusions reached reflected the consensus of the
entire group. One or more designated staff members from HHS
or USDA assisted each subcommittee.
The Science Review Subcommittee was formed to help maintain
consistent standards for the reviews across subcommittees.
The Subcommittee also addressed quality standards for the
entire process, including consideration of the format of the
report to the Secretaries, integration of the various subcommittees'
work into a cohesive document, and meeting plans.
The subcommittees communicated by conference call, e-mail,
and face-to-face meetings. Each subcommittee was responsible
for presenting the basis for its conclusions and recommendations
to the full Committee, responding to questions, and making
changes if indicated. To gain perspectives for interpreting
the science, some subcommittees invited experts to respond
to specific questions during conference calls. The full Committee
heard presentations from 12 invited experts, who addressed
questions posed by the Committee in advance and responded
to additional questions during the meeting. The conclusions
in this report reflect the consensus of the entire Committee.
Back to Top
Each subcommittee generated an initial list of research questions
that might be relevant to setting dietary guidelines. The
subcommittee then set priorities based on the perceived level
of importance and availability of literature. This process
was iterative. Throughout the deliberations, the wording and
intent of the research questions evolved, as did the need
for additional questions. Available time, expertise, and resources
precluded an examination of all issues related to the effects
of diet on chronic disease.
Back to Top
SYSTEMATIC REVIEW OF THE SCIENTIFIC EVIDENCE
The DGAC relied on the published literature and, in a few
instances, preprints of articles that had been accepted for
publication and provided to the Committee by individual members
and invited experts. Major sources of evidence were the Dietary
Reference Intake reports prepared by expert committees
convened by the Institute of Medicine (IOM). Other sources
were systematic evidence-based reports such as the Agency
for Healthcare Research and Quality report on omega-3 fatty
acids and the World Health Organization International Agency
for Research on Cancer (IARC) report on the relation between
fruit and vegetable intake and cancer. In addition to these
comprehensive documents, the subcommittees relied on literature
searches to identify pertinent articles on research questions
not addressed in any evidence-based report and to update previously
published evidence reports.
Back to Top
TYPES OF EVIDENCE
The Committee focused on studies conducted in humans. The
primary types of studies used were observational studies and
clinical trials. Specific types of observational studies were
cross-sectional studies, case-control studies, and cohort
studies. The Committee placed greatest emphasis on results
from cohort studies and trials with well-accepted, clinically
relevant outcomes. Such outcomes included clinical diseases
(e.g., incident cancer and myocardial infarction) and well-accepted
risk factors (e.g., systolic blood pressure, low-density lipoprotein
cholesterol, and weight). Meta-analyses also were considered.
The majority of studies evaluated were based on adults; there
were limited studies on children.
Back to Top
Staff developed the search strategy in consultation with each
subcommittee chair to meet the needs of that subcommittee.
The search strategy included search parameters, search terms,
search databases, and exclusion criteria (including years
Typical exclusion criteria included the following: in vitro
studies, animal studies, articles before "X" date, and drug
studies. The specific exclusion criteria varied by question
(e.g., questions involving cancer as an endpoint may not exclude
animal studies). In some cases, additional references were
identified by checking the reference lists of review articles.
The years covered were influenced by the availability of evidence-based
reviews that addressed the same topic. For example, the literature
search regarding fiber covered only 1999 and later years because
a prior IOM report covered the earlier years. Some searches
were expanded if results from the initial research were meager.
Back to Top
Summaries of Results
The Science Review Subcommittee developed a prototype table
to be used for summarizing information obtained from relevant
articles for priority questions. Content included in the tables
was to be concise, factual, and descriptive and to provide
a basis for formulating tentative conclusions. Staff worked
with the respective subcommittee chair to examine the search
results and eliminate articles that were not relevant to the
subcommittee's topic. They then extracted the key information
and, by using the prototype, produced a table to cover key
information about each question for which relevant articles
were identified. See Appendix G-3 for working summary tables.
Back to Top
Critical Review of Studies
Subcommittee members read the tables and requested key articles.
They then critically assessed study quality and relevance
to the overall question being addressed. The subcommittee
members, not the staff, made the decisions on study quality
and on the relative value of clinical trials and observational
studies. They considered these factors, along with the data
summarized in the tables, in reaching tentative conclusions
for consideration by the full Committee.
Back to Top
Preparation of Conclusive Statements
For each research question, subcommittees prepared a brief
document that included a conclusion that specifically addressed
the research question, a list of key sources, and a summary
of key studies and findings. The subcommittee presented draft
summary statements to the DGAC for consideration. Members
of the Committee who were not members of the subcommittee
were also assigned to review the statements and provide in-depth
critical review. For especially controversial topics, the
entire Committee examined the key published evidence on which
a conclusion was based. At the May and August meetings, the
whole Committee voted on the wording of each conclusion.
Back to Top
USE OF THE USDA FOOD INTAKE PATTERN AND SPECIAL ANALYSES
The Committee had access to the food pattern proposed by
the USDA (Federal Register notice, vol. 68, no. 176,
September 11, 2003, p. 53536) and to technical support data
related to the pattern. This information included the following:
A proposed daily food intake pattern that lists
the daily amounts of food from each food group and subgroup
for 12 age/energy groups.
Energy levels for the proposed food intake pattern.
Nutritional goals for the daily food intake pattern
covering vitamins, minerals, and macronutrients.
Nutrient profiles of the basic food groups and their
subgroups and for additional fats, oils, soft margarines,
and added sugars. The food groups and subgroups are composites
that reflect the types and amounts of foods commonly consumed
Nutrients provided by the proposed food pattern.
At the request of three subcommittees, USDA staff used its
food modeling system to conduct several types of analyses.
Most of these analyses involved the modeling of the food pattern
intended to meet selected specifications for nutrient intake.
For example, the subcommittees requested analyses to obtain
information relevant to flexibility in the choice of food
to meet nutrient needs, the effects of different fat intakes
on the nutrients provided by the food pattern, and the approximate
number of calories needed to meet recommended nutrient intakes.
See Appendix G-2 for the descriptions of these analyses and
The USDA food modeling process used in these analyses was
developed originally for deriving the Food Guide Pyramid.
It was updated for these analyses to include nutrient goals
from the IOM Dietary Reference Intakes report that
was released in 2004 (after the Federal Register
notice regarding the proposed food pattern) and the most recent
National Health and Nutrition Examination Survey (NHANES)
1999–2000 food consumption data. The USDA food modeling process
involves the following steps:
Establishing nutritional goals. Goals were obtained
from the Dietary Reference Intakes reports for various
vitamins, minerals, macronutrients, and electrolytes released
by the IOM between 1997 and 2004.
Establishing energy levels. The food pattern
was developed for caloric levels from 1,000 to 3,200 calories
per day in 200-calorie increments. The pattern was created
for each age/gender group and was deemed applicable, whether
individuals were sedentary or physically active.
Assigning nutritional goals to each specific food
intake pattern. The specific nutritional goals assigned
to each food intake pattern were the goals of age/gender
groups with sedentary energy requirements that most closely
matched the caloric level. For example, the goals of females
age 31 to 50 years, males/females age 9 to 13 years, and
females age 14 to 18 years matched the 1,800 calories
per day level. In some cases the nutrient levels in a
food pattern were compared to nutritional goals for several
age/gender groups. For example, at the 1,800-calorie level,
three goals were specified for each nutrient: those for
females age 31 to 50 years, for males/females age 9 to
13 years, and for females age 14 to 18 years.
Assigning a nutrient content to each food group and
subgroup. Foods included in each of the commodity
food groups or subgroups (fruits, milk, meat and beans,
whole grains, enriched grains, dark green vegetables,
orange vegetables, legumes, starchy vegetables, and other
vegetables) are based on the food consumption of Americans,
with any food that represents 1 percent or more of the
consumption from that group or subgroup included in the
development of a nutrient profile. Other foods (less than
1 percent of group or subgroup consumption) are grouped
with a similar food item for analysis. The nutrient profiles
of each commodity group are the weighted averages of the
nutrient content of foods in each food group based on
consumption. The USDA Continuing Survey of Food Intakes
by Individuals (CSFII) 1994–1996 was the source of food
consumption data in the Federal Register notice,
but the NHANES 1999–2000 food consumption data were used
to determine new nutrient profiles for this analysis.
Two-day food intakes from 14,262 individuals over age
2 years were weighted to represent the nationwide population.
For example, the nutritional composition of dark green
leafy vegetables reflected the nationwide consumption
of foods falling into that group, which were about 53
percent broccoli and 20 percent spinach. Therefore, the
nutritional value of the dark green leafy vegetable group
was 0.53 of the nutritional value of broccoli, 0.20 of
the value of spinach, and 0.27 other. Foods in their lowest
fat form were selected for determining the nutrient profile
of the milk and meat groups. For the milk group, fat-free
milk was the single food item used to represent this food
group. For the meat group, only the leanest cuts of meat,
fish, and poultry prepared with all fat or skin removed
were used. Eggs and nuts were also included in this group.
Determining the daily intake amounts for each food
group or subgroup. Starting from the original Pyramid
food pattern, the amounts of each food group or subgroup
were increased or decreased in an iterative manner until
the pattern for each calorie level achieved its nutritional
goal or came within a reasonable range. A reasonable progression
from pattern to pattern of the amounts recommended in
each food group was maintained to make the pattern logical
from an educational standpoint.
Because 12 different levels of energy intake ranging from
1,000 to 3,200 calories per day have been used, a person can
select a food pattern according to his or her level of physical
activity. The pattern was developed for individuals with low,
moderate, or active levels of physical activity.
There are advantages to the approach used in developing this
food intake pattern. One advantage is that it provides continuity
with previous food guidance and allows evolution of the guidance
over time to build on what consumers already understand while
updating the science base. Also, the approach provides an
educational tool that integrates the gamut of IOM recommendations
into a food intake pattern. That is, the approach assists
in converting the full set of nutrient recommendations to
food-based recommendations suitable for males and females
of different ages and activity levels. The process has resulted
in a food pattern that meets IOM recommendations for almost
all nutrients at all calorie-intake levels.
There are disadvantages to the approach, however. First,
the nutrient profile of each food group and subgroup is based
on Americans' current consumption of foods within that group.
Because Americans may not select rich sources of certain nutrients,
the nutrient profiles for a group or subgroup also may be
low in that nutrient. This makes it more difficult to develop
models that meet the appropriate Dietary Reference Intakes
for some nutrients. For example, Americans eat very few nuts
relative to other choices in the meat, poultry, fish, dry
beans, eggs, and nuts group; and the nuts they tend to eat
are not especially rich in vitamin E. Therefore, the nutrient
profile for "nuts" and for the entire meat, poultry, fish,
dry beans, eggs, and nuts group is relatively low in vitamin
E. This also is true for the types of oils that Americans
tend to select; relatively few individuals use oils that are
especially rich in vitamin E. When using the nutrient profiles
for these food groups, it is difficult to develop a food intake
pattern that meets the Recommended Dietary Allowance for vitamin
E. This holds true even if only lacto-ovo vegetarian choices
are made from the meat, poultry, fish, dry beans, eggs, and
nuts group, by including only eggs, nuts, and legumes in the
nutrient profile for this group. The same problem exists in
trying to use these nutrient profiles to meet IOM recommendations
for sodium and potassium because the profiles rely on current
consumption and the food supply, both of which are high in
sodium and low in potassium.
Second, the five basic food groups used in the modeling stemmed
from historical nutritional concerns: vitamin C (fruit), vitamin
A (vegetables), calcium (milk), protein (meat), and energy
(grains). The original 1992 Pyramid pattern considered and
evaluated 21 different diet components (i.e., vitamins, minerals,
different fats, and energy). The new dietary reference intakes
include standards for a total of 27 diet components (vitamins,
minerals, electrolytes, essential fats, all the macronutrients,
and fiber). As with the original Pyramid development, this
requires the use of more than the five basic food groups to
meet the dietary reference intakes. Consequently for these
analyses, vegetables were broken down into the following subgroups:
dark green, deep yellow, legumes, starchy, and other. Grains
were divided into whole and enriched grains. Meats and legumes
were not divided into subgroups, however. The amounts from
several of the various subgroups increased (e.g., dark green
vegetables, legumes) to meet the new nutrient recommendations.
Third, persons using this pattern need to take great care
to account for the (1) fat contained in milk products and
meats, (2) fats and added sugars that are a part of processed
foods (such as muffins or soft drinks) and that are added
when preparing or serving food, and (3) calories provided
by alcoholic beverages. Otherwise, their intakes of calories
and of saturated fats are likely to be too high.
Although the food modeling program was not perfect, it was
a valuable tool for the Committee in determining how the food
pattern could be developed that met science-based criteria
for a healthful diet.
Back to Top
SOURCES OF NUTRIENTS IN AMERICAN DIET
Several tables found in Parts D and E of this report present
food sources of nutrients consumed by Americans. These tables,
adapted from tables published by Cotton et al. (2004), draw
upon CSFII 1994–1996 data. To confirm that the CSFII data
are still representative, a prototype analysis of 1999–2000
NHANES food consumption data was run on one key nutrient—potassium.
The analytical methodology and comparison follow.
The NHANES analysis included individual consumption records
that were considered reliable and met the daily minimum (acceptable)
number of foods consumed. The Cotton analysis used similar
data from CSFII 1994–1996. The nutrient content of the individual
foods were drawn from the USDA Nutrient Composition Database
Standard Reference 16.1. The two analyses incorporated similar
food groupings. However, because time was short for completing
the Committee's work, the prototype NHANES analysis did not
disaggregate food mixtures to their basic ingredients, as
the Cotton analysis did. Rather, in most cases, the most prominent
ingredient in the food mixture dictated the category in which
the food was placed. For example, the coffee category includes
coffee with milk or other combinations and does not break
down into the coffee and milk categories, as in the Cotton
article. Similarly, the tomatoes from pizza do not appear
in the tomato category because they are picked up in the multi-ingredient
category called egg rolls, pizza, etc. An examination of the
top 10 food contributors (shown in Table C1) indicates that
there may be about a 5 percent difference when using the different
A comparison of the results from the two types of analysis
indicates that the percent contribution by food category to
the total potassium intake did not differ substantially. The
reasons for differences may be due to the placement of multi-ingredient
foods or real changes in intake. In terms of potassium intake,
the top 12 foods identified by Cotton et al. (2004) are found
within the top 14 foods on the NHANES list, and the order
does not change radically. The decision was made to use the
tables from Cotton and colleagues (2004) since they were from
a peer-reviewed, published article.
Table C1. Comparison of Potassium
Consumption Estimated Using Data from
the 1994–1996 Continuing Survey of Food
Intake by Individuals and from the 1999–2000
National Health and Nutrition Examination
||Egg rolls, pizza, other mixtures
|Corn/potato chips, popcorn
*USDA National Nutrient Database for Standard
Reference, Release 16.1.
**Followed by consommé (include
soups), then corn/potato chips and popcorn.
Back to Top
Using results from the systematic review of the scientific
literature and the food modeling exercises, the Committee
evaluated and integrated the evidence into a set of conclusive
statements and major conclusions regarding the components
of the diet and physical activity that promote the health
and well-being of Americans over age 2 years. These statements
provide the basis for a set of straightforward guidelines
for diet and physical activity.
Back to Top