Trends in junk food consumption among US children and adults, 2001–2018 (2024)

Data source, study population, and dietary assessment

NHANES is a series of cross-sectional surveys using a complex, multistage sampling design to enroll a nationally representative sample of the civilian, noninstitutionalized US population (11). The protocol was approved by the CDC/National Center for Health Statistics (NCHS) Ethics Review Board, and all participants provided written informed consent. Details on the NHANES study protocol and data collection methods are documented (12). Because the data are publicly available and deidentified, institutional review board approval was not required for the present study.

This investigation used data across the most recent 9 cycles of NHANES (2001–2002 through 2017–2018), including US children aged 2–19 y and adults aged≥20 y who completed at least 1 valid 24-h diet recall, as determined by NCHS criteria. All participants were eligible for dietary assessment with up to two 24-h recalls in which respondents reported all foods and beverages consumed during the previous 24 h (midnight to midnight). Through a computer-assisted dietary interview, trained NHANES interviewers used the 4-step Multiple-Pass Method in 2001 and the new 5-step Automated Multiple-Pass Method from 2002 and onward to collect dietary intakes. Both methods have been validated as research-based multiple-pass approaches to enhance complete and accurate food recall and reduce respondent burden (13). The 1-d value was used for individuals with single recalls and 2-d means for those with 2recalls. The protocol and data collection methods for the 24-h recall dietary assessment have been described in detail (14) (see Supplemental Methods). The same computer-assisted dietary interview system was used for both mobile examination centers in-person and phone follow-up dietary interviews. The food composition tables are updated with each cycle of NHANES to account for changes in food supply and are considered the best available by the USDA at the time of release. The flowchart for participants is shown in Supplemental Figure 1.

Assessment of junk food consumption

The primary outcome variables were the mean consumption of junk food (excluding SSBs) and its subtypes in percent energy (%E). Absolute energy intake, which varies with age, sex, physical activity, and metabolism, was assessed in secondary analyses. Junk food was characterized based on the most common specified categories used for junk food taxes across US localities, Mexico, and Hungary (Supplemental Table 1). Main subgroups were based on the USDA Food and Nutrient Database for Dietary Studies food grouping codes in combination with the USDA What We Eat in America food categories, including sweet bakery products, savory snacks, crackers, snack/meal bars, candy, and other desserts (15). To exclude potentially healthy options in these categories, we excluded any products with higher-quality fats (for products with ≥5% energy from fat), defined as an unsaturated to saturated fatty acid ratio of 4:1 or higher; higher-quality carbohydrates (for products with ≥10% energy from carbohydrate), defined as a carbohydrate to fiber ratio of 10:1 or lower; and with sodium content at or below the 2-y category-specific draft FDA voluntary sodium guidelines (16). In sensitivity analysis, we applied alternative nutrient thresholds and criteria from the Chilean Food Regulation used to define “unhealthy food” and jointly mandate front-of-package warning labels, restrict child-directed marketing, and ban sales in schools (Supplemental Methods).

Sources of junk food

Secondary outcome variables were sources of junk food, obtained through the question, “Where did you get (this/most of the ingredients for this) {FOODNAME}?” “Was it from a supermarket, a convenience-type store, a fast food place, a restaurant, from someone else, or another source?” For this investigation, food sources were grouped according to the NHANES Dietary Interviewers Procedures Manual (17) into 1) grocery stores (grocery or supermarket), 2) restaurants (restaurant fast food/pizza, restaurant with waiter/waitress, restaurant no additional information), 3) school/childcare for children (cafeteria in a K–12 school and childcare center; hereafter referred to as schools) and worksites for adults (cafeteria not in a K–12 school, vending machine, common coffee pot or snack tray), and 4) other sources (e.g., from someone else/gift; sport, recreation, or entertainment facility; food truck).

Population subgroups

To evaluate population subgroups at risk for diet-related health disparities, findings were stratified by age (2–5 y, 6–11 y, 12–19 y, 20–34 y, 35–49 y, 50–64 y, ≥65 y), sex, race/ethnicity (non-Hispanic white, non-Hispanic black, Mexican American), educational level (less than high school graduate, high school graduate or general equivalency diploma, some college, college or higher), and household income (ratio of family income to the federal poverty level<1.30, 1.30–3.49, ≥3.50). For children, parental education attainment was used. Information on race/ethnicity was collected by trained NHANES interviewers according to the fixed categories provided by the NCHS using the Computer-Assisted Personal Interview system, with classification identified by a household proxy for participants younger than 16 y and by the respondents directly for those aged 16 y and older. Race/ethnicity-specific results were presented for Mexican American individuals as opposed to all Hispanic individuals due to participant sampling changes over time (18). Trends in junk food consumption for other Hispanic and other race or mixed-race groups were not presented separately due to their small sample sizes.

Statistical analyses

The nationally representative population mean intake of junk food and its component food subgroups (%E and absolute intake) was estimated for each NHANES cycle among children and adults. Proportions of junk food from major sources were similarly assessed. The statistical significance of trends was assessed by treating survey year as a continuous variable in a survey-weighted linear regression model. Mean differences and uncertainty (95% CIs) in junk food consumption and proportions of junk food from major sources between the earliest (NHANES 2001–2002) and latest (2017–2018) cycles were calculated. Sensitivity analyses evaluated findings using data for each participant from their 24-h recalls for day 1 or 2 only. To understand whether observed trends could be driven by population demographic shifts, sensitivity analyses were adjusted for age, sex, race/ethnicity, education, and income levels within each cycle.

To evaluate potential differences in trends by population subgroups (age, sex, race/ethnicity, educational level, and income level), a survey-weighted Wald F test was used to test an interaction term between the 2-y cycle and subgroups.

Analyses were performed using Stata version 14 (StataCorp), with a 2-sided α = 0.05. Survey weights, accounting for sampling procedures, stratification, and clustering, were applied to all analyses to generate nationally representative estimates. Alpha values were not corrected for the possibility of type I error from multiple comparisons, so nominal P values should be interpreted accordingly.

Population characteristics

A total of 74,471 participants were analyzed, including 29,970 children aged 2–19 y (49.2% female) and 44,501 adults aged 20y and older (51.9% female). Participant characteristics in 2017–2018 are shown in Supplemental Table 2. In total, 92.7% of US children and 86.0% of US adults reported at least some junk food consumption on any given day. Higher percentages of junk food consumption among children compared with adults were observed across all population subgroups.

Junk food consumption and trends

From 2001–2002 to 2017–2018, the %E from junk food among children was high (nearly 1 in 5 calories) and remained stable (from 18.1% to 17.5%, P-trend=0.25) (Table 1). The %E was modestly lower (nearly 1 in 7 calories) and decreased over this period among adults (14.1% to 13.0%, P-trend=0.002) (Table 2). The major subtype in both children and adults was sweet bakery products, followed by savory snacks, other desserts, candy, crackers, and snack/energy bars. Evaluating trends in subtypes of junk food among children, increases were observed in %E from crackers (1.19% to 1.35%; P-trend=0.006) and snack/meal bars (0.38% to 0.69%; P-trend<0.001) and decreases from candy (2.58% to 1.96%; P-trend<0.001) and other desserts (3.11% to 2.48%; P-trend<0.001) (Figure 1). Trends were similar among adults, with increases in %E from snack/meal bars (0.31% to 0.78%) and decreases in %E from candy (1.97% to 1.55%; P-trend<0.001), sweet bakery products (5.52% to 4.98%; P-trend=0.04), and other desserts (2.19% to 1.86%; P-trend=0.001).

Findings for absolute energy intake from junk food are shown in Supplemental Tables 3–4. In sensitivity analyses using only the first day 24-h recall per person, the mean intake of junk food in each NHANES cycle was about 0.8–1.0% energy units higher in children and 0.3–0.4% energy units higher in adults, compared with using the average of both 24-h recalls per person (Supplemental Tables 5–7). These absolute differences corresponded to about 5% higher junk food intake in children and 2% higher junk food intake in adults, using only the first day compared with both days of 24-h recalls. However, inferences on subtypes of junk food and trends over time in children and adults were not materially altered. Results were also not materially altered by adjustment for sociodemographic characteristics over time (Supplemental Tables 8–11). Results were also largely consistent in sensitivity analyses including only products in these categories that also met the nutrient threshold/criteria from the Chilean food regulation (Supplemental Tables 12–15).

Junk food intake by sources

Among different venues, most junk food (>70%) was obtained from grocery stores among both children and adults throughout the study period, followed by other sources, restaurants, and then schools or worksites (Table 3,Supplemental Figure 2). From 2003–2004 (the first cycle of available data) to 2017–2018 among children, the proportion of junk food from grocery stores decreased from 80.6% to 72.7% (P-trend<0.001) but increased from restaurants (4.70% to 8.05%; P-trend<0.001) and other sources (10.0% to 13.1%; P-trend=0.001) and remained stable from schools (4.89% to 4.86%; P-trend=0.94). Among adults, proportions of junk food from grocery stores modestly decreased (78.7% to 77.1%; P-trend=0.002) but increased from restaurants (8.11% to 9.11%; P-trend=0.002) and other sources (from 10.2% to 12.1%; P-trend=0.001) and decreased from worksites (3.69% to 1.99%; P-trend<0.001). Details of sources of subtypes of junk food consumption are provided in Supplemental Tables 16 and 17.

Junk food intake and trends in population subgroups

Junk food consumption (%E) during this period by education and income is shown inFigure 2 and Supplemental Figure 3 and by these factors as well as age, sex, and income (%E, absolute energy) in Supplemental Tables 18–21. Among children, junk food consumption was higher throughout this period among non-Hispanic white and black Americans compared with Mexican Americans, as well as among children with heads of households with higher education and income levels compared with less than high school education and an income/poverty ratio<1.3. Junk food intake levels in children (%E) tended to be higher among 6- to 11-y-olds compared with 12- to 19-y-olds but were more similar by sex. In adults, similar differences in junk food consumption were seen as in children by race/ethnicity, education, and income. In addition, women tended to have higher intake levels than men (e.g., in 2017–2018, 13.8%E compared with 12.1%E, respectively), and surprisingly, older adults tended to have higher intake levels than younger adults (e.g., in 2017–2018, 11.1%, 12.1%, 13.3%, and 16.2% among ages 20–34 y, 35–49 y, 50–64 y, and≥65 y, respectively).

The overall national trends in junk food consumption seen between 2001–2002 and 2017–2018 were generally similar in each population subgroup (Supplemental Tables 18–21). One exception was differences in trends over time by age among adults (P-interaction=0.007): among younger adults (ages 20–34 and 35–49 y), %E from junk food decreased by ∼2%E during this period, whereas among older adults (ages 50–64 and 65+y), no significant decreases were evident. Findings for subtypes of junk food consumption in different population subgroups are shown in Supplemental Tables 22–26.

nqab129_Supplemental_File

The authors’ contributions were as follows—JL and DM: designed the research; YLee, RM, and YLi: conducted the research; JL: conducted data analyses and drafted the manuscript; DM: contributed to the draft of the manuscript; JL had full access to the data and takes responsibility for the integrity of the data and accuracy of the data analyses. All authors: read and approved the final manuscript.

Author disclosures: All authors report support from NIH grants during the conduct of the study. In addition, RM reports research funding from Unilever and personal fees from the World Bank and Bunge and from Development Initiatives for serving as chair of the Global Nutrition Report, all outside the submitted work. DM reports research funding from the NIH, the Gates Foundation, and the Rockefeller Foundation; personal fees from Acasti Pharma, America's Test Kitchen, Barilla, Cleveland Clinic Foundation, Danone, GOED, and Motif FoodWorks; scientific advisory board, Beren Therapeutics, Brightseed, Calibrate, DayTwo (ended June 2020), Elysium Health, Filtricine, Foodome, HumanCo, January Inc., Season, and Tiny Organics; and chapter royalties from UpToDate, all outside the submitted work.All other authors report no conflicts of interest.

1. Luscher TF. Nutrition, obesity, diabetes, and cardiovascular outcomes: a deadly association. Eur Heart J. 2020;41(28):2603–7. [PubMed] [Google Scholar]

2. Liu J, Rehm CD, Onopa J, Mozaffarian D. Trends in diet quality among youth in the United States, 1999–2016. JAMA. 2020;323(12):1161–74. [PMC free article] [PubMed] [Google Scholar]

3. Rehm CD, Penalvo JL, Afshin A, Mozaffarian D. Dietary intake among US Adults, 1999–2012. JAMA. 2016;315(23):2542–53. [PMC free article] [PubMed] [Google Scholar]

4. Birch L, Savage JS, Ventura A. Influences on the development of children's eating behaviours: from infancy to adolescence. Can J Diet Pract Res. 2007;68(1):s1–s56. [PMC free article] [PubMed] [Google Scholar]

5. Pomeranz JL, Wilde P, Huang Y, Micha R, Mozaffarian D. Legal and administrative feasibility of a federal junk food and sugar-sweetened beverage tax to improve diet. Am J Public Health. 2018;108(2):203–9. [PMC free article] [PubMed] [Google Scholar]

6. Taillie LS, Rivera JA, Popkin BM, Batis C. Do high vs. low purchasers respond differently to a nonessential energy-dense food tax? Two-year evaluation of Mexico's 8% nonessential food tax. Prev Med. 2017;105:S37–42. [PMC free article] [PubMed] [Google Scholar]

7. Bíró A. Did the junk food tax make the Hungarians eat healthier?. Food Policy. 2015;54:107–15. [Google Scholar]

8. Scheelbeek PFD, Cornelsen L, Marteau TM, Jebb SA, Smith RD. Potential impact on prevalence of obesity in the UK of a 20% price increase in high sugar snacks: modelling study. BMJ. 2019;366:l4786. [PMC free article] [PubMed] [Google Scholar]

9. Mozaffarian D, Liu J, Sy S, Huang Y, Rehm C, Lee Y, Wilde P, Abrahams-Gessel S, deSouzaVeigaJardim T, Gaziano Tet al.. Cost-effectiveness of financial incentives and disincentives for improving food purchases and health through the US Supplemental Nutrition Assistance Program (SNAP): a microsimulation study. PLoS Med. 2018;15(10):e1002661. [PMC free article] [PubMed] [Google Scholar]

10. Dunford EK, Popkin BM, Ng SW. Recent trends in junk food intake in U.S. children and adolescents, 2003–2016. Am J Prev Med. 2020;59(1):49–58. [PMC free article] [PubMed] [Google Scholar]

13. Blanton CA, Moshfegh AJ, Baer DJ, Kretsch MJ. The USDA Automated Multiple-Pass Method accurately estimates group total energy and nutrient intake. J Nutr. 2006;136(10):2594–9. [PubMed] [Google Scholar]

15. US Department of Agriculture, Agricultural Research Service . USDA food and nutrient database for dietary studies 2015–2016. [Internet]. 2018; [cited 2020 Jun 26]. Available from: www.ars.usda.gov/nea/bhnrc/fsrg. [Google Scholar]

16. US Food and Drug Administration . Draft guidance for industry: voluntary sodium reduction goals: target mean and upper bound concentrations for sodium in commercially processed, packaged, and prepared foods. Silver Spring (MD): Food and Drug Administration; 2016. [Google Scholar]

18. Curtin LR, Mohadjer LK, Dohrmann SM, Kruszon-Moran D, Mirel LB, Carroll MD, Hirsch R, Burt VL, Johnson CL. National Health and Nutrition Examination Survey: sample design, 2007–2010. Vital Health Statistics 2. 2013; (160):1–23. [PubMed] [Google Scholar]

19. Vercammen KA, Moran AJ, Soto MJ, Kennedy-Shaffer L, Bleich SN. Decreasing trends in heavy sugar-sweetened beverage consumption in the United States, 2003 to 2016. J Acad Nutr Diet. 2020;120(12):1974–85 e5. [PubMed] [Google Scholar]

20. Bleich SN, Vercammen KA, Koma JW, Li Z. Trends in beverage consumption among children and adults, 2003–2014. Obesity (Silver Spring). 2018;26(2):432–41. [PubMed] [Google Scholar]

21. Koma JW, Vercammen KA, Jarlenski MP, Frelier JM, Bleich SN. Sugary drink consumption among children by supplemental nutrition assistance program status. Am J Prev Med. 2020;58(1):69–78. [PubMed] [Google Scholar]

23. Child Nutrition . WIC Reauthorization Act of 2004. Public Law. 2004;108(265):118. [Google Scholar]

24. Gidding SS, Dennison BA, Birch LL, Daniels SR, Gillman MW, Lichtenstein AH, Rattay KT, Steinberger J, Stettler N, VanHorn Let al.. Dietary recommendations for children and adolescents: a guide for practitioners. Pediatrics. 2006;117(2):544–59. [PubMed] [Google Scholar]

25. Food and Drug Administration . Food labeling: revision of the nutrition and supplement facts labels. Final rule. Fed Regist. 2016;81(103):33741. [PubMed] [Google Scholar]

26. Huang Y, Kypridemos C, Liu J, Lee Y, Pearson-Stuttard J, Collins B, Bandosz P, Capewell S, Whitsel L, Wilde Pet al.. Cost-effectiveness of the US Food and Drug Administration added sugar labeling policy for improving diet and health. Circulation. 2019;139(23):2613–24. [PMC free article] [PubMed] [Google Scholar]

28. Liu J, Mozaffarian D, Sy S, Lee Y, Wilde PE, Abrahams-Gessel S, Gaziano T, Micha R, Project F-P. Health and economic impacts of the national menu calorie labeling law in the United States: a microsimulation study. Circ Cardiovasc Qual Outcomes. 2020;13(6):e006313. [PMC free article] [PubMed] [Google Scholar]

29. Lee Y, Mozaffarian D, Sy S, Liu J, Wilde PE, Marklund M, Abrahams-Gessel S, Gaziano TA, Micha R. Health impact and cost-effectiveness of volume, tiered, and absolute sugar content sugar-sweetened beverage tax policies in the United States: a microsimulation study. Circulation. 2020;142(6):523–34. [PMC free article] [PubMed] [Google Scholar]

30. US Department of Health and Human Services, Food and Drug Adminstration . Voluntary sodium reduction goals: target mean and upper bound concentrations for sodium in commercially processed, packaged, and prepared foods: guidance for industry. Draft guidance. [Internet] [cited 2020 Jan 12]. Available from: https://www.fda.gov/media/98264/download. [Google Scholar]

31. Pearson-Stuttard J, Kypridemos C, Collins B, Mozaffarian D, Huang Y, Bandosz P, Capewell S, Whitsel L, Wilde P, O'Flaherty Met al.. Estimating the health and economic effects of the proposed US Food and Drug Administration voluntary sodium reformulation: microsimulation cost-effectiveness analysis. PLoS Med. 2018;15(4):e1002551. [PMC free article] [PubMed] [Google Scholar]

32. US Department of Agriculture, US Department of Health and Human Services . Dietary Guidelines for Americans, 2020–2025. [Internet]. 9th ed. 2020; [cited 2021 Mar 22]. Available from: DietaryGuidelines.gov. [Google Scholar]

34. Bleich SN, Wolfson JA. Trends in SSBs and snack consumption among children by age, body weight, and race/ethnicity. Obesity. 2015;23(5):1039–46. [PMC free article] [PubMed] [Google Scholar]

35. Dunford EK, Popkin BM. 37 Year snacking trends for US children 1977–2014. Pediatr Obes. 2018;13(4):247–55. [PMC free article] [PubMed] [Google Scholar]

36. Piernas C, Popkin BM. Snacking increased among U.S. adults between 1977 and 2006. J Nutr. 2010;140(2):325–32. [PMC free article] [PubMed] [Google Scholar]

37. Dunford EK, Popkin BM. Disparities in snacking trends in US adults over a 35 year period from 1977 to 2012. Nutrients. 2017;9(8):809. [PMC free article] [PubMed] [Google Scholar]

38. Marriott BP, Hunt KJ, Malek AM, Newman JC. Trends in intake of energy and total sugar from sugar-sweetened beverages in the United States among children and adults, NHANES 2003–2016. Nutrients. 2019;11(9):2004. [PMC free article] [PubMed] [Google Scholar]