The nutritional environment plays a key role in determining eating behaviour and overall health(Reference Glanz, Sallis and Saelens1). Fast-food (FF) restaurants are abundant in industrialised countries as nearly a third of US adults consume FF on a daily basis(Reference Fryar, Hughes and Herrick2). FF and full-service restaurants account for roughly 20 % of daily calories in the USA(Reference Mazidi and Speakman3), while FF customers tend to underestimate how many calories they have consumed(Reference Block, Condon and Kleinman4). Over a recent 30-year span, FF restaurants have increased the variety of their menus along with the portion size, calories served and sodium content(Reference McCrory, Harbaugh and Appeadu5). Greater access to FF and FF consumption have been observed as associated factors in childhood(Reference Jia, Shi and Jiang6,Reference Jakobsen, Brader and Bruun7) and adult(Reference Reitzel, Regan and Nguyen8) obesity, and an increase in the density of FF restaurants is associated with an increase in BMI(Reference Acciai, DeWeese and Yedidia9). Overall, FF restaurants supply a major source of calories in the nutritional environment and thus are implicated in the current US obesity epidemic(Reference Temple10).
Over half of the calories in the US diet come from ultra-processed foods (UPF)(Reference Marino, Puppo and Del Bo’11), and consumption of UPF is correlated with an increased risk of various mental disorders, cardiometabolic disease and mortality outcomes(Reference Lane, Gamage and Du12). Per the NOVA classification, UPF are defined as foods that include the fractioning of whole foods into substances, chemical modifications of these substances, assembly of unmodified and modified food substances, frequent use of cosmetic additives and sophisticated packaging(Reference Monteiro, Cannon and Levy13). The presence of a single ultra-processed ingredient warrants the UPF designation, and thus, it is likely that many foods in FF restaurants will be ultra-processed; however, this has not been demonstrated via a menu analysis using the NOVA classification.
Eating outside of the house is associated with increased consumption of UPF(Reference Andrade, Gombi-Vaca and da Louzada14), and among people eating at FF restaurants, it has been observed that 88 % of their calories were from UPF(Reference Souza, Andrade and Rauber15), which suggests that FF restaurants serve UPF. Of note, the availability of energy-dense(Reference Gupta, Hawk and Aggarwal16) and highly palatable(Reference Fardet17) UPF in the nutritional environment has been associated with obesity(Reference de Araújo, de Moraes and Magalhães18,Reference Vitale, Costabile and Testa19) and clinical trials using UPF as the independent variable have shown increased weight gain with a fully UPF diet over a 2-week period(Reference Hall, Ayuketah and Brychta20). Therefore, both FF restaurants and UPF are implicated as contributing to the current US obesity epidemic by supplying energy-dense calories into the nutritional environment.
To determine the level of food processing of FF restaurant menus, the NOVA classification(Reference Monteiro, Cannon and Levy13) was used to evaluate menus from six different FF restaurants. These restaurants were those that had the highest sales within their respective restaurant category as reported by Quick Service Restaurants Magazine 2020 Top 50 Report: (Sandwich, Pizza, Burger, Snacks, Global and Chicken). There were three research questions: (1) What percentage of FF menu items are ultra-processed, and (2) what are the most common ingredients found in FF menus and what type of food processing group would the ingredient belong to? It was hypothesised that the majority of the menu items would be UPF and that some of the most common ingredients would belong to UPF.
Methods
Data collection
The top six highest grossing FF restaurants for each restaurant category were identified from QSR Magazine 2020 Top 50 Chart (www.qsrmagazine.com/content/qsr50-2020-top-50-chart; see online supplementary material, Supplemental Table 1). Nutritional information was gathered from each restaurant’s website (Burger: McDonald’s, Chicken: Chik-Fil-A, Global: Taco Bell, Pizza: Dominos, Sandwich: Subway and Snack: Starbucks; see online supplementary material, Supplemental Table 2). Nutritional information was presented as either purchasable menu items, individual ingredients found within food, or both (see online supplementary material, Supplemental Table 2), so the term ‘menu items’ is used to capture each of these. When ingredient information was not available for a brand-named item, the company’s website was reviewed (e.g. Starbucks sold ‘KIND® Salted Caramel & Dark Chocolate Nut Bar’ and the ingredient information was collected from www.kindsnacks.com). Alcoholic beverages, items from a specific version of restaurants (e.g. Cantina menu for Taco Bell) and carry-home items (e.g. the ‘at-home items’ from Starbucks) were not included in the analysis. In total, 872 menu items were collected across restaurants, and after removing duplicate items and items without ingredient information available (see online supplementary material, Supplemental Table 2), the remaining 748 items were then included in the analysis.
Analysis of menu items
To answer the research questions, the proportion of menu items for each NOVA group was determined with a mean calculated for all restaurants. Based on ingredient information, menu items were coded into the four NOVA classification groups: minimally processed (MPF), culinary processed (CPF), processed (PRF) and UPF(Reference Marino, Puppo and Del Bo’11). Coding was performed independently to agreement by NKN and AJB. Coders initially agreed on 93 % of all menu items (Chicken: 93 %, Pizza: 85 %, Burger: 96 %, Snack: 92 %, Sandwich: 93 % and Global: 100 %). Data were presented in a 100 % stacked column using Microsoft Excel. SPSS was used to conduct non-parametric single χ 2 tests to determine if there was a difference in the proportion of menu items in the processing groups for each menu and for all menu items (SPSS 29; IBM).
Ingredient analysis
Using Monkeylearn.com, word/phrase frequency clouds were created to visualise the top fifty words/phrases within and across menu items, where the larger the word/phrase appears, the more frequently it appeared in the ingredient list. From there, the top ingredients in each menu (words/phrases identified from monkeylearn.com) were ranked (i.e. the most and least common appearing ingredient were ranked from 1 to 50, respectively). Ingredient lists were then merged, the number of times an ingredient was listed was identified, and the mean rank and sd were calculated. Ingredients that appeared in half or more of the food menus were presented, and functional classes were identified from Codex General Standard for Food Additives (GSFA) from the FAO of the UN and WHO(21). Ingredients were then coded independently and to agreement by NKN and AJB (77 % initial agreement) into five groups: MPF, CPF, PRF, UPF and NTR (vitamins, minerals and water).
Results
Food processing percentage
Figure 1 shows the percentage of menu items for each NOVA classification group for each restaurant menu and all menus. The majority of menu items in each restaurant (range: 70–94 %), and across restaurants (85 %), were ultra-processed (P < 0·001 for all menus and for all menu items across menus). In addition, MPF comprised only 11 % of items on average across menus.
Figure 1. Proportion of menu items from each food processing group for six fast-food restaurant menus. *P < 0·001 for all; single sample χ 2 test.
Ingredient analysis
Figure 2 depicts the top fifty ingredients across the six chosen restaurants, where the larger the word appears the more frequently it appears in the menus. Thirty-nine ingredients appeared in half or more of the menus (Table 1). Salt and sugar, both CPF, were the two highest-ranked ingredients. Only three UPF ingredients appeared in all six menus: natural flavour, citric acid and xanthan gum (listed in descending mean rank). Of the thirty-nine ingredients, 15 % were CPF, 18 % were NTR, 21 % were MPF and 46 % were UPF, with zero PRF ingredients. Fifteen of the ingredients were listed in the Codex General Standard for Food Additives, and most (80 %) were UPF ingredients. The top three common UPF functional classes were emulsifier (n 8 appearances), thickener (n 6), and sequestrant and stabiliser (n 5).
Figure 2. Ingredient word cloud from six fast-food restaurant menu items. Word/phrase frequency cloud produced from the top fifty words from all menu items.
Table 1 Ingredients appearing in three or more of analysed fast-food restaurant menus
CPF, culinary processed; UPF, ultra-processed foods; NTR, nutrient and water; MPF, minimally processed.
Data developed from the top fifty appearing words/phrases in each menu; functional class determined via the Codex General Standard for Food Additives (GSFA) from the FAO of the UN and WHO.
Discussion
This was the first study to explore the level of food processing among menus for several popular FF restaurants in the USA using the NOVA classification. The results show that, on average, FF restaurant menu items are highly processed with 85 % of menu items being ultra-processed foods and only 11 % of menu items consisting of minimally processed foods. Thus, there are very few non-UPF options available at these six FF restaurants. Given the role of the food environment in health, FF restaurants may be contributing to the rise in obesity in the USA by providing predominantly energy-dense, ultra-processed foods to their customers. In addition, across ingredients that appeared in three or more of the menus, 46 % were considered ultra-processed, with the most common functional classes being emulsifier, thickener, and sequestrant and stabiliser. The three most frequently appearing ultra-processed ingredients in the FF restaurant menu items were natural flavours, citric acid and xanthan gum.
Considering that nearly a third of US adults consume FF on a daily basis(Reference Fryar, Hughes and Herrick2), and the association between FF restaurant proximity, FF consumption and UPF consumption with obesity(Reference Jia, Shi and Jiang6–Reference Temple10,Reference de Araújo, de Moraes and Magalhães18,Reference Vitale, Costabile and Testa19) , these results are of concern for public health. This prevalence of UPF across these six FF restaurants is higher than what has been observed in grocery stores, where the majority of items are also UPF(Reference Gupta, Hawk and Aggarwal16). Together, between the abundance of UPF at grocery stores and FF restaurants, consumers may need to go out of their way to seek non-UPF. Future research examining the prevalence of UPF in non-FF restaurants is warranted to better describe additional sources of food within the US nutritional environment. Public health efforts to decrease consumption of UPF and increase consumption of MPF, at every level of the nutritional environment, are warranted to combat the obesity epidemic.
Natural flavours were the most common UPF ingredient, which appeared in every FF restaurant that was analysed. According to the US Food and Drug Administration (FDA), natural flavours are anything (e.g. essential oil, extract and protein) collected from foods (e.g. spice, fruit, vegetables, herbs, etc.) whose function is flavouring, rather than nutritional(22). While there is debate whether natural flavours are ‘natural’(Reference Goodman23), the presence of them within a food item warrants the classification of an UPF. The next two most popular UPF ingredients were citric acid and xanthan gum. Citric acid (an organic acid) is currently the single largest chemical obtained from chemical biosynthesis, and its popularity as a food additive is due to its chemical nature (multiple functions within food)(Reference Ciriminna, Meneguzzo and Delisi24). In an analysis of ingredients used in culinary preparations from institutional food services (e.g. private cafeterias and universities), 8·4–12·6 % of ingredients were UPF and were mainly used in protein dishes and desserts(Reference Padovan, Thimoteo da Cunha and Adriano Martins25). This further demonstrates the invasiveness of ultra-processed ingredients used in food preparation.
A critique of the NOVA classification is the inclusion of specific ingredients as a method to identify UPF because, while citric acid is a food additive, citric acid is also found naturally in foods(Reference Gibney and Forde26). Xanthan gum is a naturally occurring microbial exopolysaccharide, and it is considered safe by the FDA but is not digestible by humans(Reference Abu Elella, Goda and Gab-Allah27). Emulsifiers form a uniform texture consistency and are abundant in UPF. This, there is concern about the impact of emulsifiers on the pathogenesis of certain diseases(Reference Sandall, Smith and Svensen28). While natural flavours, citric acid and xanthan gum are all ambiguous in their effects on human health, collectively, these three ingredients make UPF shelf stable and more flavourful, which may increase palatability and promote increased consumption.
This analysis has some strengths and limitations. First, the sample size was limited to six restaurants; however, these six represented the highest grossing restaurant for each restaurant category and thus represented the most frequented FF restaurants. In addition, selecting one from each of the restaurant types allowed for a broad menu analysis capturing the variety of FF options available within the USA. However, because we only chose one restaurant per food category, our findings cannot be applied to other FF restaurants. Another limitation is that some menu items were removed from analysis because the restaurants or Internet did not provide ingredient information; thus, the menu and ingredient analyses were produced with only available data. While the NOVA classification is one of the most common food processing classifications, the definition of ultra-processed foods has changed considerably over time(Reference Gibney29); thus, as new food processing definitions and categorisations are developed, the results of this study may differ.
This study illustrates the invasive nature of ultra-processed foods in the US nutritional environment. The high intake UPF in the US could also be due to the low availability of healthful, nutrient-dense, minimally processed food choices among some of the highest-grossing FF restaurants in the USA. Therefore, to improve the nutritional quality of their menu items, FF companies should consider reformulation or the addition of MPF to the menu to increase healthful food options.
Supplementary material
For supplementary material accompanying this paper visit https://doi.org/10.1017/S1368980025000060
Acknowledgements
None.
Financial support
N.K.N. was supported through the LEAP (Learning, Entering, Advising and Producing research) Scholar Program for transfer students through Arizona State University which is funded by a Scholarship in STEM (S-STEM) grant (1644236) from the US National Science Foundation.
Competing interests
There are no conflicts of interest.
Ethics of human subject participation
Not applicable.
Authorship
Conceptualisation: A.J.B.; Data curation: A.J.B., K.L.S. and N.K.N.; Formal analysis: A.J.B. and N.K.N.; Funding acquisition: n/a; Investigation: A.J.B., N.K.N. and K.L.S.; Methodology: A.J.B. and K.L.S.; Project administration: A.J.B. and K.L.S.; Resources: n/a; Software: A.J.B. and K.L.S.; Supervision: K.L.S.; Validation: A.J.B. and K.L.S.; Visualisation: A.J.B.; Writing – Original draft: A.J.B.; Writing – Review and Editing: A.J.B. and K.L.S.
Open access
