Documentation

USDA, Economic Research Service (ERS) provides descriptive data on U.S. consumers’ diet quality, including intake amounts, distribution percentages (shares), and density of nutrients and food groups by food source from 1977 to 2018. There are two main food sources—food at home and food away from home—and the latter is further separated into restaurants (with waiter service), fast food places (self-service), schools, and other food-away-from-home places.

The descriptive statistics cover the mean values and their associated standard errors for the U.S. population aged 2 and older for each of the 11 nationally-representative food consumption surveys conducted by the Federal Government since 1977. Children under the age of 2 are excluded because the Dietary Guidelines for Americans, 2020–2025 (DGA) recommendations for healthy dietary patterns are intended for those aged 2 and older (DGA, 2020–2025).

About the Estimates: Data Sources

The descriptive statistics in this data product are based on 11 nationally-representative food consumption surveys conducted by the Federal Government since 1977. The 11 surveys include:

  • 1977–1978 Nationwide Food Consumption Survey (NFCS);
  • 1989–1991, 1994–1996, and 1998 Continuing Survey of Food Intakes by Individuals (CSFII); and
  • 2003–2004, 2005–2006, 2007–2008, 2009–2010, 2011–2012, 2013–2014, 2015–2016, and 2017–2018 What We Eat in America/National Health and Nutrition Examination Surveys (WWEIA/NHANES).

The 1977–1978 NFCS and 1989–1991 CSFII collected food intakes by individuals of all ages for 3 consecutive days, and the 1994–1996 and 1998 (1994–1998) CSFII and 2003–2018 WWEIA/NHANES recorded 2 nonconsecutive days of food intakes using a 24-hour dietary recall methodology. In all surveys, respondents were interviewed in person to collect intake data for the first day, whereas intakes on subsequent days were collected using different methodologies (i.e., a self-reported food diary in 1977–1978 and 1989–1991, and a telephone interview for the 1994–1998 and 2003–2018 surveys). For this data product, only the first-day data was analyzed to minimize the potential bias that may occur due to differences in timing (consecutive versus nonconsecutive days) and data collection method. Children under the age of 2 are excluded because the Dietary Guidelines for Americans, 2020–2025 (DGA) recommendations for healthy dietary patterns are intended for those aged 2 and older (DGA, 2020–2025).

USDA, Agricultural Research Service (ARS) has developed the Food and Nutrient Database for Dietary Studies (FNDDS) to translate food intakes into nutrient intakes, which can be compared with the nutrient intake recommendations established by the National Academy of Sciences, Engineering and Medicine, known as Dietary Reference Intakes. FNDDS has been updated over time so there are some key nutrients in current FNDDS that are not available in the older survey data, such as saturated fats, cholesterol, sodium, and dietary fiber in the 1977–1978 NFCS. USDA, ARS has also developed food group equivalent databases (MyPyramid Equivalents Database and Food Patterns Equivalents Database) to translate foods recorded in surveys into food equivalent amounts, which can be compared with the Healthy U.S.-Style Dietary Pattern specified in the Dietary Guidelines for Americans. The food group equivalent databases are developed for the 1994–1998 CSFII and 2003–2018 WWEIA/NHANES but they are not available for 1977–1978 NFCS and 1989–1991 CSFII surveys. USDA, ERS researchers developed consistent FNDDS and food group equivalent databases for the 11 surveys. The USDA, ERS methodology for developing FNDDS and food group equivalent databases is documented in the following Economic Information Bulletin (EIB):

Dietary Quality by Food Source and Demographics in the United States, 1977–2018

Food consumption survey respondents reported whether a food was eaten at home and where it was acquired. Food source coding differs between the NFCS, CSFII, and different waves of WWEIA/NHANES. However, major food sources—such as supermarkets, restaurants with servers, fast food establishments, and school cafeterias—are common to all surveys (for more information, see appendix A in the EIB). There are two main categories of food sources: “food at home” (FAH) and “food away from home” (FAFH), with the latter separated into four subcategories including full-service restaurant, fast food establishment, school (K–12 school and daycare), and other FAFH. The coding of food sources in the 11 surveys is documented in the EIB. Food at home is generally obtained at a retail store such as a supermarket, grocery store, or convenience store. Food away from home is generally purchased from foodservice establishments such as full-service restaurants with waitstaff (restaurant), fast food restaurants with limited menus and no waitstaff, and carryout places (fast food), school cafeterias, daycare centers, and summer camps for children aged 2–19 (school) as well as from other food-away-from-home places, such as community food programs, street vendors, vending machines, etc.

The surveys collect demographic information, such as household income, race and ethnicity, educational achievement, age, and sex. The demographic data are useful to monitor the nutritional status of the U.S. population subgroups. Data on the nutrient and food group intakes, distribution, and density by food source and demographics are available in tables 2–7. The sample sizes for the U.S. population in total and by demographics are reported in table 1.

About the Estimates: Method

The descriptive statistics are generated using the statistical software SAS/STAT Survey procedures to incorporate complex survey design effects and sample weights to generate national estimates of averages (means) and standard errors. In the literature there are two methods identified for generating population estimates of dietary intakes: (1) the mean approach, and (2) the population approach (Krebs-Smith et al., 1989). In the EIB, USDA, ERS used the mean approach, in which the food or nutrient density was calculated for everyone who reported eating the food or nutrient and then the weighted average density was calculated using the survey sample weight (the number of U.S. consumers a survey respondent represents).

The mean and population methods often yield similar results, although they may produce different and equally valid results (Lin et al., 2023). In essence, the mean approach estimates the dietary quality of a representative U.S. consumer—the average diet of U.S. consumers. The population approach generates the estimate of nutritional quality of foods consumed by all U.S. consumers. An illustration of the main difference between the mean and population approaches can be found at Lin et al. (2016).

Nutrient Intake Estimates

Using USDA's nutrient database FNDDS, food consumption data in WWEIA/NHANES were converted to nutrient consumption data. The mean value (average) and its standard error of daily intakes of these nutrients of concern and other nutrients for the U.S. population aged 2 and above are reported by food source in table 2. The distribution (percentage) of nutrient intakes by food source is summarized in table 3. Nutrient density (the amount of a nutrient for each 1,000 calories of food consumed) by food source is reported in table 4.

Food Consumption Estimates

Since foods provide an array of nutrients and other components that have health benefits, the Dietary Guidelines for Americans (DGA) translate the Dietary Reference Intakes into food and beverage recommendations—namely the Healthy U.S.-Style Dietary Pattern—to provide advice on what and how much to eat and drink to promote health, reduce the risk of chronic disease, and meet nutrient needs.

Food consumption data from WWEIA/NHANES were combined with USDA's food group equivalent databases (My Pyramid Equivalent Database and Food Patterns Equivalents Database) to estimate food consumption by food groups as specified in the 2020–2025 DGA. For example, a respondent in WWEIA/NHANES may report having eaten a specific amount of apple pie then those data are translated into cups of fruit, ounces of grain, grams of oils and solid fats, and teaspoons of added sugar.

Food consumption in terms of the food groups is reported for all sources as well as by food source for the total U.S. population aged 2 and older. Averages and their standard errors of daily food-group intakes are reported in table 5. The food-group intake distribution (percentage) by food source is reported in table 6. Table 7 has the averages and their associated standard errors of food-group density, the amount of food group for each 1,000 calories in a person’s diet.

U.S. Diets in 2017–2018 Compared With Dietary Recommendations

Table 8 shows the 2020–2025 DGA recommended density measures for a 2,000-calorie diet and the 2017–2018 actual reported densities for total foods, food-at-home, and various food-away-from-home sources. Table 8 also reports the ratios of the actual 2017–2018 densities to the recommended densities to show the distance between actual diet and the recommended diet.

Demographic Data

USDA, ERS researchers also examined differences in dietary quality across several demographics, including age, gender, race and ethnicity, household income, and education attainment for adults aged 20 and older. For each of tables 2–7 (intake, distribution, and density of nutrients and food groups) in this data product, there are corresponding tabs on demographics.

Cautions in Data Interpretation

The Federal food consumption survey methodology has improved over the 1977–2018 period. Some of the methodological changes could contribute to the differences reported in this data product. USDA, ARS has used the five-step Automated Multiple-Pass Method (AMPM) to improve the completeness of data collection since the 1994–1998 CSFII (Raper et al., 2004). The adoption of AMPM is likely to lessen the extent of under-reporting, but the extent of reduction in under-reporting is unknown. Although the surveys’ methodology has been validated, dietary recalls may be limited by an individual’s ability to describe consumed foods accurately. Other changes in survey methodology and implementation, as well as the updated nutrient and food equivalent databases, need to be considered when comparing food intake and dietary changes as reported in the surveys conducted between 1977 and 2018. For example, intakes of tap water were not collected until 2003–2004. Tap water has no calories but contains minerals (e.g., calcium) such that this change in data collection was found to result in outliers in nutrient density for certain food sources, as discussed in the EIB.

References

Lin, B.H., Guthrie, J., & Smith, T. (2023). Dietary quality by food source and demographics in the United States, 1977–2018 (Report No. EIB-249). U.S. Department of Agriculture, Economic Research Service.

Lin, B.H., Anekwe, T. D., Buzby, J. C., & Bentley, J. (2016). U.S. food commodity availability by food source, 1994–2008 (Report No. ERR-221). U.S. Department of Agriculture, Economic Research Service.

Krebs-Smith S. M., Kott, P. S., & Guenther, P. M. (1989). Mean proportion and population proportion: Two answers to the same question? Journal of American Dietetic Association, 1989(89), 671–676.

Raper, N., Perloff, B., Ingwersen, L., Steinfeldt, L., & Anand, J. (2004). An overview of USDA's dietary intake data system. Journal of Food Composition and Analysis, 17(3–4), 545–555.

U.S. Department of Agriculture & U.S. Department of Health and Human Services. (2020). Dietary guidelines for Americans, 2020–2025, 9th Edition. U.S. Department of Agriculture and U.S. Department of Health and Human Services.

Recommended citation

U.S. Department of Agriculture (USDA), Economic Research Service (ERS). (2024). Food Consumption, Nutrient Intakes, and Diet Quality.