Caballero B. (ed.) Encyclopaedia of Food Science, Food Technology and Nutrition. Ten-Volume Set

Подождите немного. Документ загружается.

dietary assessment and for characterization of nutri-

tional status in population studies. For a fuller discus-

sion of the use of biological markers, readers are

referred to Margetts and Nelson (1997).

See also: Dietary Surveys: Surveys of National Food

Intake; Surveys of Food Intakes in Groups and Individuals;

Food and Agriculture Organization of the United

Nations

Further Reading

Bingham SA (1987) The dietary assessment of individuals:

methods, accuracy, new techniques and recommenda-

tions. Nutrition Abstracts and Reviews 57: 705–742.

Cameron ME and Van Staveren WA (1988) Manual on

Methodology for Food Consumption Surveys. Oxford:

Oxford University Press.

Margetts B and Nelson M (1997) Design Concepts in

Nutritional Epidemiology, 2nd edn. Oxford: Oxford

University Press.

Nelson M (2000) Methods and validity of dietary assess-

ment. In: Garrow J, James WPT and Ralph A (eds)

Human Nutrition and Dietetics, 10th edn. London:

Churchill Livingstone.

Nelson M, Morris JA, Black AE and Cole TJ (1989)

Between- and within-subject variation in nutrient intake

from infancy to old age: estimating the number of days

required to rank dietary intakes with desired precision.

American Journal of Clinical Nutrition 50: 155–67.

Surveys of National Food Intake

J A T Pennington, Division of Nutrition Research

Coordination, Bethesda, MD, USA

Purpose/Objectives of National Food

Intake Surveys

0001 National food intake (consumption) surveys are

usually undertaken by government agencies res-

ponsible for the health and/or food supplies of a

population. Such agencies may include departments

or ministries of health, departments of agriculture, or

food administrations.

0002 National food intake surveys may be carried out to

assess and monitor the dietary status of the popula-

tion. Other objectives for food intake surveys are to

determine the adequacy and safety of the national

food supply, to determine the effects of dietary pat-

terns on nutrient intake and health status, to assess

changes in dietary patterns and their affects on

nutrient intake, and to identify needs for changes in

national food policies or regulations.

0003The dietary status of the population can be moni-

tored by comparing food and nutrient intakes with

recommended intake standards set by government,

academic, or private agencies. For example, the aver-

age number of servings of fruits, vegetables, dairy

products, grain products, and meats or meat substi-

tutes may be compared with the recommended stand-

ards, and the average daily intake of nutrients (e.g.,

protein, fat, carbohydrate, vitamins, and minerals)

may be compared with recommended dietary allow-

ances for these nutrients. The dietary status of a

nation can also be compared with measures of disease

incidence to determine potential relationships be-

tween nutrient intakes and dietary deficiency diseases

or chronic diseases such as obesity, heart disease,

diabetes, stroke, alcoholism, and various types of

cancer.

0004Results from food intake surveys help determine

the adequacy of the national food supply and identify

populations without adequate food that are under-

served by current food relief efforts. Assessments of

food and nutrient intake help determine the extent of

physical, mental, and emotional problems relating to

food insecurity (i.e., inadequate food intake). Dietary

surveys help determine which population subgroups

as defined by age, sex, race, ethnicity, income, urban-

ization, or education are at risk for energy and nutri-

ent deficiencies. Food intake surveys should include

questions about the use of public food assistance,

food banks, and community feeding programs be-

cause the responses to these questions may assist

government agencies in making decisions about the

need for continued or increased support of such

programs. Information from the survey may help

determine the effectiveness of public health policies

regarding food fortification, food stamps, and feeding

programs for children, the elderly, institutionalized

persons, and those below the poverty level. Survey

results may assist governments in revising current

food-related policies or developing new policies.

0005Results from food intake surveys also help deter-

mine the safety of the food supply. In addition to

providing estimates of nutrient intake, the informa-

tion from national food intake surveys may help pro-

vide dietary exposure estimates for pesticide residues,

food contaminants, and food toxins. Concentration

data on the levels of pesticide residues, contaminants,

and toxins in foods may be merged with the food

consumption data from the surveys to provide such

exposure estimates. Information on dietary exposure

to potentially harmful constituents in foods may be

useful for modifying or revising policies concerning

agricultural and food manufacturing practices.

1876 DIETARY SURVEYS/Surveys of National Food Intake

0006 Results from food intake surveys help determine

the effects of dietary patterns on nutrient intake and

on health status. Families perpetuate the dietary pat-

terns learned from their parents and other relatives,

and families within communities and larger geo-

graphic areas (e.g., countries) tend to have similar

dietary patterns. These learned dietary patterns may

be modified or changed with exposure to the foods

and eating patterns of other cultures via travel, immi-

gration, migration, or the presence of visitors. Media

food advertising and the offering of new and different

foods from grocery stores and restaurants may also

affect dietary patterns. Dietary surveys should include

questions about cultural and ethnic dietary patterns

so that the relationships of cultural and ethnic factors

to food and nutrient intake can be identified. This

may allow for the identification of relationships

between culturally or ethnically defined dietary pat-

terns on dietary and health status.

0007 Food intake surveys are useful for following

changes in dietary patterns, their effects on nutrient

intake, and potential implications for health. For this

purpose, the surveys are most useful if they are done

routinely (e.g., yearly or biennially) using the same or

similar methodology so that changes and trends in

food and nutrient intake can be identified. Changes

in food intake may reflect the impact of the availabil-

ity of ready-to-eat products from grocery stores, fast

food restaurants, and other food vendors. Changes in

agricultural and manufacturing practices may result

in changes in the composition of foods (e.g., the type

of fats and oils in baked products and the type of

sweeteners in soft drinks and desserts) so that nutrient

intake changes even without changes in food pat-

terns. It is important to have ongoing monitoring

of food and nutrient intake as well as monitoring of

disease incidence so that cause and effect might be

investigated.

0008 Food intake surveys are useful for identifying issues

that might require changes in policies or regulations

concerning food safety, food labeling, food fortifica-

tion, food additives, agriculture practices, or manu-

facturer practices. Ultimately, the food policies and

regulations developed by government agencies should

serve to maintain and improve the health of its citi-

zens. Food intake surveys may be done in conjunction

with health examination surveys that include physical

exams, health histories, clinical measures, and/or

the collection of biological samples (urine, blood, or

other tissue) that are analyzed to assess nutritional

or health status. When surveys of food intake and

health measures are done concurrently on the same

population, there are opportunities to relate food and

nutrient intake to health status measures such as

blood pressure, body weight, and blood levels of

cholesterol, triglycerides, and various nutrients.

Changes in the health status of people within a coun-

try could be the result of many factors (diet being one

of them). Dietary status data derived from food

intake surveys could be an important factor when

assessing the health of a population. Other factors

(physical activity patterns, environmental health, eco-

nomics, epidemics, etc.) are also very important with

regard to the health of a population. Dietary surveys

provide an important piece of information that will

help governments in making policy and regulatory

decisions concerning foods and agriculture.

Sample Population

0009A food intake survey should include survey partici-

pants that are representative of the population in

terms of the demographic features of interest to

the investigators. Usually, the demographic factors

include sex, age, region, urbanization, income,

education, race, and ethnicity. There may be other

important considerations with respect to population

sampling for a national survey. For example, a gov-

ernment agency may wish to oversample at-risk

populations or minority populations for whom few

dietary data are currently available.

0010Obtaining a representative sample population for a

food intake survey requires the expertise of statisti-

cians who are knowledgeable of the population and

its dynamics. The statisticians can assess the currently

available population databases (e.g., census database,

voters registration database) and determine whether

they might be useful to develop a sample population

for the dietary intake survey.

0011The agency responsible for the food intake survey

must consult with the statisticians and survey design-

ers regarding important facts such as:

0012If individuals in institutions (hospitals, prisons, rest

homes, retirement communities, convents, monas-

teries, etc), individuals without permanent resident

status (immigrants, aliens), and individuals with-

out homes are to be included in the survey. Survey

participants generally need to have a home address,

a telephone or other means of contact so they can

be requested to participate in a survey.

0013Which demographic features (age, sex, race/ethnic

group, income, region, urbanization, education)

are of importance to the survey results and how

many individuals with the demographic features

must be in the survey so that representative dietary

intakes can be calculated for them.

0014If minority or at-risk groups should be over-

sampled if they will not be present in sufficient

numbers in the overall sample.

DIETARY SURVEYS/Surveys of National Food Intake 1877

.0015 If the survey should be designed to correct for

sample bias (e.g., nonresponse rate, oversampling).

One method to correct for sample bias is to develop

weighting factors for each individual in the survey

that pertains to his/her demographic characteris-

tics.

0016 How to obtain dietary information for those not

able to respond for themselves (e.g., young chil-

dren, ill, elderly, at work, traveling, illiterate, etc.).

0017 How to maximize survey participation of the

selected participants, e.g., by sending a letter of

intent, following up with a phone call, minimizing

respondent burden, and providing remuneration

(if possible) for the subjects’ time and effort.

0018 How to recontact people if necessary (e.g., to clar-

ify information) without violating privacy issues.

Methods Used for National Food Intake

0019 Information about the foods consumed by a nation

can be obtained using a variety of methods such as

food disappearance (per capita intake), food balance

sheets, household inventories, 24-h dietary recalls

with a professional interviewer or computer assist-

ance, food diaries (records) kept by survey partici-

pants, weighed food intakes, diet histories, duplicate

portions, and food frequency questionnaires. There

are various advantages and disadvantages of each

method with regard to costs, respondent burden,

and accuracy and reliability of the information

obtained. Costs include the staff time, field survey

transport and housing, materials, data processing,

and participant renumeration. Respondent burden

refers to the time and effort required by the survey

participant to provide information. (See Dietary

Surveys: Measurement of Food Intake.)

0020 The preferred methods for national food intake

surveys are 24-h dietary recalls or food diaries, and

it is most useful if there are two or more days of

dietary information for each survey participant. The

methods of food disappearance, food balance sheets,

and household inventories are less accurate, tending

to overestimate food intake, and do not provide spe-

cific information based on the age and sex of partici-

pants. Weighed food intakes, diet histories, and

duplicate portions take much interviewer and partici-

pant time and are expensive (i.e., the cost per person

is high for the large sample population needed for

national data). The weighed intakes and duplicate

portion technique also tend to alter usual eating pat-

terns, usually by decreasing total food intake. Food

frequency questionnaires may provide useful infor-

mation but are generally not quantitative and are

not specific enough because they include a limited

number of foods (as low as 50).

0021Although the 24-h dietary recalls and food diaries

are likely the methods of choice for national surveys,

they too have disadvantages. The 24-h dietary recall

requires accurate memory and honesty of the survey

participant as well as knowledge of food names, de-

scriptions, and serving sizes. Food diaries have the

effect of altering usual food intake and are also sub-

ject to the honesty of the survey participant. The act

of writing down consumed foods in a diary tends to

decrease total food intake and may result in ‘health-

ier’ food choices that are not typical or usual in

countries where obesity prevalence is high. The

main problem with dietary recalls and diaries appears

to be underestimation of total food intake. Recent

studies, using the doubly labeled water technique,

have estimated that dietary recalls and diaries may

be underestimated by about 20% of energy. It is not

known if the underestimation applies equally to all

foods or if there is selective underestimation as for

example with alcoholic beverages, snack foods, or

desserts.

0022The information derived from dietary surveys is

particularly useful to identify potential health prob-

lems related to food and nutrient intake. The poten-

tial problems can be investigated more fully by

collecting and evaluating information obtained from

clinical and biochemical measures of nutritional

status. Because of the difficulty of obtaining accurate,

valid, and reliably dietary information from survey

participants, dietary data are usually not considered

sufficient for making changes in nutrition policy, i.e.,

the dietary results need to be verified with clinical and

biochemical results. Otherwise the benefits of policy

changes (which can be expensive and burdensome on

the food industry) may not be forthcoming.

Considerations with Dietary Recalls and

Records

0023The food intake survey should be designed with

considerations for the number of days of recalls or

records, the days of the week and seasons of the year,

memory cues and probes, the interviewing situation

and interviewing techniques, the food composition

database, default food codes, dietary supplements,

cultural sensitivity of the dietary tool, specific food

use questions, and the use of biochemical markers to

assess the validity of the dietary method.

0024For most nutrients, as many as 7–14 days of diet-

ary information may be necessary to characterize

usual intake for an individual. However, for popu-

lation distributions of nutrient intake, 2 days of

food recalls or diaries are usually sufficient. Col-

lecting and processing additional days (beyond

1878 DIETARY SURVEYS/Surveys of National Food Intake

two) of dietary information result in increased

financial costs as well as reduced response rates

as individuals tire of participating.

0025 Because people tend to eat differently on work

days and days off work, it is important to factor

days of the week into the sample design for food

recalls and diaries. Eating patterns are affected by

the seasonality of some foods (e.g., watermelon,

strawberries, plums, grapes, corn, spring lamb,

oysters, and crabs). The seasonality of outside

temperatures may affect food preferences and

eating patterns. People often prefer hot and hearty

soups and meals on cold days and salads or lighter

meals on hot days. Feasting and fasting patterns

around religious and national holidays are often

different than other days. It may be best to avoid

collecting data for such days or to factor these

days into the sampling design so that the overall

results will be representative.

0026 Cues and probes provided by the dietary inter-

viewer or the computer-assisted interview help

survey participants remember the details about

the foods they eat. Questions about snacks and

about added ingredients (sugar and cream in

coffee, catsup, salad dressing, sauces) are import-

ant to obtain complete food recalls and diaries.

The use of food models and food pictures also

helps to clearly identify foods and to improve the

accuracy of serving size estimates.

0027 The manner in which food-related questions are

asked affects responses. Certain characteristics of

the interviewing situation may affect responses

about foods that might be depicted as ‘good

foods’ or ‘bad foods’ by the survey participants.

For example, the presence of other family members

during the dietary interview may lead to underesti-

mation of the intake of alcoholic beverages,

snacks, and dessert items (especially for parents

and teenagers). An interview in a health setting

such as a clinic may elicit responses with under-

estimated amounts of fat and overestimated

amounts of foods with higher nutrient content

(e.g., fruits and vegetables). Interviewers should

attempt individual interviews, if possible, refrain

from either positive of negative feedback about

food choices, and encourage subjects to accurately

report all foods. Interviewers can emphasize the

confidentiality of the recall or diary and indicate

that only grouped results will be reported unless

sent in advance via mail to the participants.

0028 Dietary interviews have traditionally been done as

in-person (one-on-one or in family interviews), but

are now also done by computer and by telephone.

The computer-assisted interview helps standardize

the questions and probes because some potential

human errors are removed. Computers do not get

tired, sick, forgetful, or irritable, and computers

do not require training, quality control checks, or

intrainterviewer compatibility checks. Telephone

interviews are less expensive than in-person inter-

views as they do not require an examination site or

the travel costs of survey participants. However,

telephone interviews have the disadvantage of not

permitting the use of food pictures or models of

serving portions.

0029To derive accurate and reliable information from

the dietary interview, the food consumption data

must be merged with a food composition database

that contains accurate and relevant data on the

nutrient content of the foods consumed by the

population. The database should contain both

the generic and brand name products typically

eaten. To ensure accuracy, the nutrient informa-

tion on brand name products in the database

should be obtained from the manufacturers or

from a qualified laboratory. The foods in the data-

base should contain the current levels of nutrient

fortification for products usually fortified such as

wheat flour, white rice, cornmeal, ready-to-eat

breakfast cereals, milk, salt, etc. The database

should include foods from fast food and other

restaurants as well as manufactured and home-

made mixed dishes, ethnic foods, and desserts.

0030Survey participants often have difficulty identify-

ing foods, remembering (or knowing) the ingredi-

ents in mixed dishes, and properly identifying

serving portions. Food models, pictures, and

computer graphics may help participants identify

foods and serving sizes. However, default food

codes will be necessary when detailed information

cannot be elicited from the participants. Default

food codes are needed for mixed dishes, especially

homemade and restaurant where exact recipes and

ingredients are not known. The database can carry

mixed dishes with traditional ingredients (from

cookbooks) and calculated nutrient values as de-

faults. Defaults may also be necessary for sand-

wiches, soups, pizza, pie, cake, cereal, salad, and

many other general food categories where more

detail cannot be elicited.

0031Information about the intake of dietary supple-

ments is important because supplements can be

very potent sources of nutrients and far outweigh

the concentration of nutrients obtained from

foods. However, it is usually best to maintain

information about supplement intake separately

from food intake so that survey results can be

presented both ways (i.e., as nutrient intakes

excluding supplements and intakes with supple-

ments for those individuals who consume them).

DIETARY SURVEYS/Surveys of National Food Intake 1879

Of primary interest is the type of supplement

taken (the nutrients included and the brand

name), the potency for each nutrient, the fre-

quency of use of the supplement, and the dur-

ation of use. There are thousands of supplement

products available both over-the-counter and by

prescription. Information about supplement

intake is difficult to capture and code. A particu-

lar complexity in assessing supplement use is that

many individuals take supplements inconsist-

ently or in patterns that are difficult to character-

ize (e.g., when they do not feel good, on certain

days of the week, until the bottle is empty, when

they feel tired). If possible, the survey partici-

pants should be asked to bring their supplement

bottles to the interview so that the interviewer

can account for the types and amounts taken.

If supplement intake data are to be collected

during the survey, a supplement database will

be needed to convert the supplement intake

data into nutrient intake data.

003 2 The dietary tool developed for a food intake

survey may not work equally well for all groups

within a population. The method may not re-

trieve the same level and quality of information

from minority populations or population sub-

groups that differ by culture, urbanization,

income, education, or living situation. (In fact,

men and women may respond differently to the

same tool, and children may respond differently

than teenagers or adults.) The food intake tools

should be culturally competent, or there should

be different tools for different cultures. Infants,

children, adolescents, and elderly persons may

require surrogate reports or confirmation of

dietary intakes from parents or caretakers.

People who are institutionalized (i.e., in hos-

pitals, rest homes, retirement homes, prisons, or

psychiatric institutions) may require surrogate

dietary intake reports from their caretakers.

10.

0033 It may be necessary to include specific food use

questions in the survey to meet the identified

objectives of the agency conducting the survey.

For example, there may be questions about the

type of cooking and table fat typically used

(butter, margarine, lard, suet, shortening), the

use of discretionary salt during cooking and at

the table, whether the salt is iodized or not, the

use of a microwave oven, typical methods of food

preparation (frying, baking, grilling), the use of

artificial sweeteners, the use of foods with fat or

sugar replacers, the use of medical foods, if fat is

trimmed from meat cuts, if skin is removed from

poultry, or the use of low calorie products (e.g.,

salad dressings, soft drinks).

11.

0034Dietary surveys are much less expensive than

physical exams with clinical and biochemical

measurements to assess the nutritional status of

a population. However, as noted, the results of

dietary surveys are not completely accurate. It

would help to validate (or invalidate) the results

of dietary surveys by using a short-term biochem-

ical marker, if resources permit. Short-term bio-

chemical markers that have been used to attempt

to validate the previous day’s nutrient intake

include urinary nitrogen, potassium, sodium,

and carotenoids for dietary protein, potassium,

sodium, and carotene as well as the doubly

labeled water technique for energy intake.

Summarizing Data and Data Availability

0035The information collected during dietary surveys

must be stored, reviewed, aggregated, and evaluated

with special attention to weighting factors and to the

sample number when data are reported by demo-

graphic characteristics. Nutrient intakes may be

reported as means with standard deviations, medians,

and percentile distributions. The data derived from a

national food intake survey are generally valid for the

population as a whole and for population subgroups

for which the sample size is sufficient. Government

agencies should make the results of their food intake

surveys widely available through publications and on

the internet to assist other government agencies that

are attempting to undertake food intake surveys for

the first time or that wish to improve upon the

methods used in their previous surveys. Government

agencies should summarize the results of their food

intake surveys and explain and justify their conclu-

sions regarding the dietary status of their population.

They may also attempt to relate the dietary status

of the population or population subgroups to the

reported incidence of diet-related diseases in the

population or in population subgroups.

See also: Dietary Surveys: Measurement of Food Intake