Fairer competition? Warning labels narrow the gap between the emotional reactions evoked by ultra-processed and unprocessed/minimally processed foods.

Fairer competition? Warning labels narrow the gap between the emotional reactions evoked by ultra-processed and unprocessed/minimally processed foods - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice BMC Nutr . 2026 Mar 7;12:70. doi: 10.1186/s40795-026-01296-2 Search in PMC Search in PubMed View in NLM Catalog Add to search Fairer competition? Warning labels narrow the gap between the emotional reactions evoked by ultra-processed and unprocessed/minimally processed foods Laiz A A Silva Laiz A A Silva 1 Department of Physiology and Pharmacology, Biomedical Institute, Universidade Federal Fluminense, Niteroi, RJ Brazil 2 Institute of Food and Nutrition, Universidade Federal do Rio de Janeiro, Macaé, RJ Brazil Find articles by Laiz A A Silva 1, 2 , Thayane C Lemos Thayane C Lemos 1 Department of Physiology and Pharmacology, Biomedical Institute, Universidade Federal Fluminense, Niteroi, RJ Brazil 3 Social and Cognitive Neuroscience Laboratory, Universidade Presbiteriana Mackenzie, São Paulo, SP Brazil 4 National Institute of Science and Technology on Social and Affective Neuroscience (INCT-SANI), SP São Paulo, Brazil Find articles by Thayane C Lemos 1, 3, 4 , Sara D J Gaspar Sara D J Gaspar 2 Institute of Food and Nutrition, Universidade Federal do Rio de Janeiro, Macaé, RJ Brazil Find articles by Sara D J Gaspar 2 , Guilherme M Coutinho Guilherme M Coutinho 1 Department of Physiology and Pharmacology, Biomedical Institute, Universidade Federal Fluminense, Niteroi, RJ Brazil Find articles by Guilherme M Coutinho 1 , Leticia Oliveira Leticia Oliveira 1 Department of Physiology and Pharmacology, Biomedical Institute, Universidade Federal Fluminense, Niteroi, RJ Brazil Find articles by Leticia Oliveira 1 , Mirtes G Pereira Mirtes G Pereira 1 Department of Physiology and Pharmacology, Biomedical Institute, Universidade Federal Fluminense, Niteroi, RJ Brazil Find articles by Mirtes G Pereira 1 , Carlos F U Rojas Carlos F U Rojas 5 Design Postgraduate Program, Department of Design, Universidade Federal do Paraná, Curitiba, Brazil 6 Department of Design, Universidade da Região de Joinville, Joinville, Brazil Find articles by Carlos F U Rojas 5, 6 , Carla G Spinillo Carla G Spinillo 5 Design Postgraduate Program, Department of Design, Universidade Federal do Paraná, Curitiba, Brazil Find articles by Carla G Spinillo 5 , Isabel A David Isabel A David 1 Department of Physiology and Pharmacology, Biomedical Institute, Universidade Federal Fluminense, Niteroi, RJ Brazil Find articles by Isabel A David 1, ✉ Author information Article notes Copyright and License information 1 Department of Physiology and Pharmacology, Biomedical Institute, Universidade Federal Fluminense, Niteroi, RJ Brazil 2 Institute of Food and Nutrition, Universidade Federal do Rio de Janeiro, Macaé, RJ Brazil 3 Social and Cognitive Neuroscience Laboratory, Universidade Presbiteriana Mackenzie, São Paulo, SP Brazil 4 National Institute of Science and Technology on Social and Affective Neuroscience (INCT-SANI), SP São Paulo, Brazil 5 Design Postgraduate Program, Department of Design, Universidade Federal do Paraná, Curitiba, Brazil 6 Department of Design, Universidade da Região de Joinville, Joinville, Brazil ✉ Corresponding author. Received 2025 May 9; Accepted 2026 Feb 28; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13081361  PMID: 41795099 Abstract Background Compared with unprocessed/minimally processed foods (UMPF), ultraprocessed foods (UPF) elicit stronger emotional reactions. We aimed to investigate whether octagon and/or triangle warning labels (WL) reduce the difference between the emotional reactions evoked by UPF and those evoked by UMPF. Methods University students ( n = 247) participated in a supervised remote experiment in which normative ratings for the International Affective Pictures System (IAPS) methodology were applied. The participants were randomly allocated into three experimental study arms: (1) the barcode label (control), (2) the triangle label, and (3) the octagon label study arms. In all study arms, the participants were shown 22 food pictures (11 UPF and 11 UMPF) interspersed with 70 pictures from other affective categories in the IAPS catalog. The UPF contained one of the labels (barcode, triangle or octagon) depending on the study arm. After viewing each picture, the participants rated it according to two dimensions of emotion, namely, hedonic valence (pleasantness) and arousal. Results The participants in study arm 1 (the control group) rated UPF pictures as more pleasant and more arousing than UMPF pictures did. When UPF contained a WL (octagon or triangle label), the participants rated the UPF pictures as being equally pleasant and arousing as the UMPF pictures were. WL (triangle or octagon) can potentially reduce the greater hedonic valence and arousal evoked by UPF relative to UMPF. Conclusion The current study expands the understanding of how triangle and octagon warning labels (WL) may influence consumer perceptions and behaviors. Supplementary Information The online version contains supplementary material available at 10.1186/s40795-026-01296-2. Keywords: Front-of-pack nutrition labels, Emotions, Ultra-processed foods, Unprocessed foods, NOVA classification, Public health Background In accordance with the Nova classification system [ 1 – 3 ], ultra-processed foods (UPF; category IV) are defined as industrially made products that have undergone various industrial processes. The Nova classification system supports the recommendations of the Brazilian Dietary Guidelines and has played a critical role in informing public policies in Brazil, the country in which the present study was conducted [ 4 ]. The consumption of UPF has been linked to health risks such as obesity, heart disease, and cancer [ 5 ]. In contrast, unprocessed/minimally processed foods (UMPF; category I) are composed of health-promoting chemicals and fibers found in fruits and vegetables [ 6 ]. Overall, low UPF and high UMPF intakes provide the foundation for healthy and sustainable food systems [ 7 ]. In many nations, the production, distribution, and consumption of UMPF that made up of traditional diets have decreased as a result of the globalization of the diet, which is typified by an abundance of UPF [ 8 , 9 ]. The Pan American Health Organization (PAHO) noted that taxation, limitations on advertisements, and the implementation of front-of-package nutrition labels (FoPNLs) are viable strategies that can be used to regulate unhealthy food products and promote healthier food environments [ 10 , 11 ]. FoPNLs inform consumers about the nutritional properties of foods, especially nutrients and ingredients that might be present in excess quantities in a product. For this purpose, graphic design and nutritional information summaries are included on the fronts of packages in a visible and accessible way to facilitate consumers’ food purchase decisions. FoPNLs: Warning Labels Many different FoPNLs are used globally. Interpretive nutrient-specific systems, such as warning labels (WL), multiple traffic light and Nutri-Score, may use symbols, colors, and qualitative descriptions to inform consumers about the high amount of nutrients important for nutrition and health [ 12 ]. Although there is no consistency among countries regarding the elements and layout of FoPNLs, in Latin America, there seems to be a preference for the use of WL [ 13 ]. This may be because warnings are intended to alert individuals of a hazard or risk, which is a communication strategy that combines text and symbols. In line with the warning literature [ 14 , 15 ], WL are expected to present the following information and graphic characteristics: the use of a signal word (High in, Excess) followed by the information the warning refers to (nutrient in excess), standard color and a graphic symbol/shape [ 16 ]. Accordingly, a black octagon (the symbol for ‘stop’ and security notices) was initially proposed and adopted by Chile in 2016 [ 17 ]. Among the ten Latin American countries that have implemented FoPNLs, seven have adopted WLs that display black octagons: Chile, Peru, Uruguay, Mexico, Argentina, Colombia and Venezuela [ 13 , 18 ]. A WL that displays a black triangle, an internationally accepted shape for alerts [ 19 ], was adapted from the black octagons used in Chile by Brazilian information design researchers [ 20 ]. In South Africa, research on triangle WL has also been carried out [ 21 , 22 ]. Despite studies suggesting that triangle WL would be a reliable option for the Brazilian population [ 20 , 23 ], FoPNLs in the form of a magnifying glass were ultimately adopted in Brazil instead of a WL (see [ 24 , 25 ] for a discussion about FoPNL regulation in Brazil). The design elements of WL, such as familiar warning signs and the black color, serve to make the labels more visible to consumers, capture their attention, and clearly inform them about the nutritional content of foods [ 20 , 26 , 27 ]. Examples of the triangle and octagon WL are shown in Fig. 1 . Fig. 1. Open in a new tab Examples of warning labels proposed for Latin American countries. Top panel. Triangle warning label (TRI). Bottom panel. Octagon warning label (OCT). The text is in Portuguese, as presented to the participants in this study. From left to right: ‘alto em sódio’ = high in sodium, ‘alto em açúcar’ = high in sugar, ‘alto em gorduras saturadas’ = high in saturated fats. Min. Saúde = Ministry of Health FoPNLs should assist in the selection of products. Compared with other interpretive labels, such as the Nutri-Score, the multiple traffic light label, and the Health Star Rating, the published literature largely supports the idea that WL, such as octagon and triangle WL, are more effective at reducing perceptions of the healthfulness of unhealthy products and the intention to purchase them [ 23 , 26 , 28 ]. Octagon WL has been shown to decrease both the intention to purchase and the actual purchase of products that are high in nutrients of concern, including sugars, saturated fats, and sodium [ 29 ]. It has been demonstrated that in Chile, following FoPNLs implementation, the consumption of high-sugar foods has decreased [ 29 ], and children and adolescents have reduced their consumption of nutrients of concern at school [ 30 ]. Octagon WL is considered to have the best visualization, comprehension, and influence on the purchase intentions of low-income individuals in Chile [ 31 ]. Khandpur et al. [ 20 ] showed that in Brazil, triangle and octagon WL outperformed the ingredient list and nutrition facts panel in enhancing participants’ understanding and perception of a product’s nutrient profile, particularly by helping them identify nutrients present in excessive amounts. In comparison to a multiple traffic light label, the triangle WL significantly decreased participants’ desire to purchase unhealthy food in a sample from South Africa [ 32 ]. Role of emotions in food purchase decisions Although previous studies have examined the effectiveness of WL in conveying simplified information that is understandable at a glance concerning nutrient content and in shaping purchase intentions [ 20 , 33 ], the role of emotions in this process has been somehow overlooked. Emotions play an important role in purchase decisions, and understanding and harnessing emotional triggers is critical for marketers aiming to connect with consumers and increase sales [ 34 ]. Consumers often base their purchasing decisions on how a product makes them feel rather than a completely rational evaluation. For instance, the dual processing theory of decision-making [ 35 ] states that consumers make decisions based on the interaction of instinctive (automatic, heuristic-based) and deliberate (controlled) thought processes. Heuristic-based decisions are more likely to be made by consumers who are pressed for time, who are tired, or who are highly motivated to eat hedonistically [ 36 ]. Therefore, taste and brand reputation are likely to be more important to hedonically motivated consumers than nutritional information is, which may affect how well FoPNLs work [ 37 ]. Importantly, hedonic motivation is considered the driving force of emotional responses [ 38 , 39 ]. When FoPNLs are implemented, it is important to consider that consumers’ purchase decisions are influenced by food-evoked emotions [ 40 ]. Consumers’ decision to buy UPF may be prompted by the powerful positive feelings that UPF trigger [ 41 ]. It is often explored by the food industry to attract and persuade consumers [ 42 ]. Given the importance of emotions in purchasing decisions, research on the performance of FoPNLs should include, among other things, measuring emotions. However, as emotions are difficult to describe, a significant challenge when assessing emotions induced by food experiences is obtaining an accurate and reliable evaluation of food-evoked emotions [ 43 ]. Although there are many theories of emotions (e.g., discrete emotions [ 44 , 45 ]), dimensional theories of emotion have conceptualized emotions primarily along two main dimensions: (i) hedonic valence (pleasantness), which varies from negative to positive, and (ii) arousal (intensity of motivational mobilization), which varies from low to high [ 46 – 48 ]. Assessment of the hedonic valence and arousal dimensions of emotions significantly improves the prediction of food choices [ 49 , 50 ]. The Self-Assessment Manikin (SAM) scale is a widely used nonverbal tool for assessing arousal and pleasantness in response to emotional pictures [ 47 ]. Furthermore, brain event-related potentials, sweating, bradycardia or tachycardia, and facial muscle tension or relaxation are examples of physiological responses to emotional stimuli that correlate with SAM scale ratings of affective experiences [ 51 , 52 ]. These physiological reactions can provide insight into latent motivational behaviors. In the context of eating behavior, arousal and pleasantness ratings provide additional information beyond typical liking ratings, aiding in food choice prediction [ 50 ]. For instance, the study by Gutjar et al. [ 50 ] provides direct behavioral evidence that emotional modulation can influence, though not fully determine, consumer decisions. After rating their emotional reactions on hedonic valence and arousal dimensions, participants in this study made food choices in two realistic choice contexts that mimicked real-world scenarios. Notably, ratings of arousal and hedonic valence were found to be important indicators of food choice [ 50 ]. These findings, together with previous results from Dalenberg et al. [ 49 ], demonstrate that emotions, in particular their hedonic valence and arousal components, play a measurable and predictive role in real food choice and consumption, beyond what is accounted for by hedonic liking alone. In fact, other previous studies using the SAM scale have also shown that valence and arousal ratings to food stimuli are strongly correlated with the intention to consume them [ 41 , 53 ]. The hedonic valence and arousal dimensions of emotion are closely linked to action dispositions, i.e., the bidirectional motivational tendencies of approach and withdrawal, reflecting how positive and negative affective states drive behavioral engagement or avoidance [ 51 ]. UPF typically elicit high positive hedonic valence and arousal, reflecting reward-related motivational salience [ 53 ]. By introducing salient visual warnings, such as textual messages or symbols to highlight excessive sugar, sodium, or fat, WL may shift these responses by reducing positive valence and enhancing arousal linked to avoidance motivation. Studies employing event-related potentials, derived from electroencephalography, support this assumption [ 54 ]. For instance, Rosenblatt et al. [ 55 ] examined how text-only and image-based health warnings influence emotional reactivity to images of palatable snacks, as measured by the late positive potential (LPP), a brain marker of emotional arousal. Following exposure to the warnings, participants exhibited reduced LPP amplitudes associated with tastiness ratings and were more likely to choose healthy foods over tasty but unhealthy options. These findings indicate that health warnings may diminish the emotional appeal of unhealthy foods and encourage healthier choices. Other studies using event-related potentials have also examined how cognitive emotional regulation strategies [ 56 ] influence emotional responses to food. Instructing participants to consider the long-term health consequences of consuming high-energy foods increases LPP amplitude, indicating greater motivational engagement [ 57 – 59 ]. Specifically, Fernandes et al. [ 57 ] found that text warnings about the health risks of UPF enhanced LPP responses and led participants to rate UPF images as less positive. These findings suggest that text warnings may effectively reduce UPF appeal. These effects align with the Neural Model of Front-of-Package Label Processing [ 60 ], which proposes that front-of-pack labels downregulate reward-related regions (e.g. amygdala, insula, orbitofrontal cortex) while enhancing top-down control via prefrontal and parietal areas, thereby reducing reward value and promoting healthier decisions. The present study Previous studies have demonstrated that warning labels are useful for modulating consumption intentions [ 33 ]. However, exploring aspects associated with consumer intention, such as food-evoked emotions, which may be influenced by WL, is important. In the current study, we combined perspectives from public health nutrition and psychology to contribute to a better understanding of the emotional mechanisms through which WL exerts its influence. Importantly, the present study was not designed to assess behavioral outcomes such as food choice or consumption, but rather to examine implicit emotional responses to food cues. In this context, implicit is defined operationally, in line with affective neuroscience and the International Affective Picture System (IAPS) tradition, as referring to emotional responses elicited in the absence of explicit choice, verbal report, or goal-directed decision-making tasks [ 46 , 47 ]. Emotional responses were assessed using the SAM scale, a non-verbal and intuitive measure designed to capture core affective dimensions without requiring reflective or analytical judgment. Responses were obtained after passive picture viewing, with warning labels embedded incidentally within a broader affective picture set. Importantly, implicit processing in this context does not imply a lack of awareness of the labels, but rather that emotional responses were elicited without instructing participants to deliberately evaluate nutritional or health-related attributes [ 61 ]. Accordingly, the findings should be interpreted as preliminary evidence of affective modulation rather than direct behavioral change. Specifically, understanding whether UPF, when presented with a WL, elicit emotional responses lower than or similar to those elicited by UMPF is very important. A previous study using the SAM scale revealed that pictures of UPF elicit higher levels of approach motivation than do pictures of UMPF, followed by a greater intention to consume the former instead of the latter [ 53 ]. Individuals typically select tasty, high-energy foods [ 62 ]. In addition to food marketing [ 63 ], UPF typically have an advantage over UMPF in terms of food choice [ 53 , 64 ]. The UMPF might have a better chance of being chosen if the WL can make the UPF appear less appealing, thus making the competition between UPF and UMPF more even. In the present experimental study, we aimed to provide new evidence concerning the impact of WL (octagon and triangle) on consumers’ instinctive and automatic emotional judgments of UPF. We predicted that, for participants exposed to WL, the emotional reactions to UPF would better resemble those elicited by UMPF, indicating that WL may reduce the greater emotional appeal of UPF than UMPF. In summary, we intend to provide new insights into the role of emotions in WL efficacy, thereby contributing to the implementation and refinement of WL capable of regulating this critical aspect of consumer decision-making. Materials and methods Participants The study consisted of a supervised remote (online) behavioral psychology experiment [ 65 ]. Participants were recruited through email or social media or in person (classrooms). Three hundred and sixty-four university students from the Federal University of Rio de Janeiro (Brazil) volunteered to participate in the study. Students who met the following criteria were included in the experimental sample: were 18–30 years old, were native Portuguese speakers, had access to a computer or notebook with an internet connection, were omnivorous, and reported having normal or corrected visual acuity. The exclusion criteria were as follows: incomplete evaluation of the pictures presented (answering less than 50% of the Self-Assessment Manikin Scale, SAM scale) [ 47 ]), texting the researcher during the experiment (which demonstrated that the participant was not focused on it), reporting being pregnant, and reporting an eating disorder diagnosis. The final sample comprised 247 university students (median age 22 years, interquartile range (IQR) = 4.00) who were enrolled in different fields of study and who were naive to the purpose of the study. Before the experimental session, the participants were instructed to complete a questionnaire composed of items related to sociodemographic, gender and anthropometric data (Supplementary Material 1). Visual stimuli The main experimental procedure was to elicit emotions in the participants by exposing them to a series of pictures with different emotional contents via the normative rating approach of the International Affective Picture System (IAPS), a standard technique suggested by Lang, Bradley & Cuthbert [ 66 ]. The procedures proposed by Lang, Bradley & Cuthbert [ 66 ] were adapted and validated for the remote format in a previous study [ 65 ]. In accordance with Lang, Bradley & Cuthbert [ 66 ], to determine the emotions evoked by each picture, a group of participants (between 50 and 100) were shown a series of pictures that might include the pictures of interest (in our case, the food pictures) as well as background pictures from other emotional categories. To provide an emotive basis for comparison when evaluating target pictures (food stimuli), pictures from the IAPS were used as the background. Pictures of food (UPF or UMPF) and other emotional content (e.g., erotic images, images of nature, puppies, sports, adventure, mutilated bodies, disease, and loss) were used for this purpose. This technique yields equivalent ratings between groups of participants [ 67 ]. Set of pictures from the International Affective Picture System The participants viewed 70 pictures from the IAPS with positive, negative, and neutral valences and evoked varying levels of arousal. A list of the IAPS picture codes applied during the study experiments can be found in (Supplementary Material 1). Set of food pictures The seventy pictures from the IAPS were interspersed with twenty-two food pictures (eleven UPF pictures and eleven UMPF pictures). When the food pictures (the pictures of interest in the study) were evaluated, the background pictures from the IAPS provided an emotive basis for comparison. Consequently, the valence and arousal ratings given to the target pictures were validated via this procedure [ 53 ]. Interestingly, when this procedure was employed, the participants were completely unaware of which categories of pictures were of interest in the study. Thus, this weakened or eliminated the social desirability effect [ 68 ]. In fact, at the end of the experiment in the present study, the participants answered the question “Did you think that any of the picture categories were of interest in the study?” after they were presented with a list of picture categories used in the study (e.g., food, loss, puppies, family, sports, and mutilation). Only six of the 247 participants responded “food”. The UMPF and UPF pictures were selected from a standardized picture dataset formulated by the research team [ 53 ]. The standardization method used in this previous study [ 53 ] enabled us to select UPF and UMPF pictures with the greatest (UPF) and lowest (UMPF) affective reactions to subsequently test them in the present study. The goal of selecting pictures with strong emotional responses for UPF and low emotional responses for UMPF was to determine the difference between these two categories of pictures in the control study arm and to test whether this difference would be minimized in the study arms that contained the WL. The food pictures from this dataset were classified into UPF and UMPF pictures according to the Nova classification system [ 2 ]. In every food picture (UMPF and UPF pictures), the food item was presented as ready to eat, prepared and unpackaged where appropriate. This process increased the similarity among the UPF, UMPF and IAPS pictures [ 67 ]. Since a product name or brand was absent from the pictures of UPF, we concentrated the study on the intrinsic qualities of UPF. The following is a list of the 22 food pictures used in the experiment. UMPF- watermelon, apple, mandarin juice, salad, corn, egg, lettuce, pear, mango, banana and bean; UPF- gums, potato chips, chocolate bar, ready-to-eat lasagne, ice cream, margarine, cookies, soft drinks, sausages, cookies stuffed with vanilla and Brazilian cheese bread. See Supplementary Material 1 to visualize the food pictures. We paired the pictures of UPF and UMPF with respect to their perceptual properties, complexity, and aesthetic attributes. Brightness, contrast, and spatial frequency were computed following Bradley et al. [ 69 ]. Picture complexity was indexed by JPEG file compression size, which correlates with subjective complexity ratings [ 70 ]. Aesthetic quality was estimated through the ACQUINE system [ 71 ], a machine learning tool that predicts aesthetic value based on visual feature extraction and classification. The pictures of UPF and UMPF did not differ in brightness (t(20) = 0.41, p value = 0.68), contrast (t(20) = 0.41, p value = 0.68), spatial frequency (t(20) = 0.30, p value = 0.77), complexity (t(20) = -1.90, p = 0.07) or aesthetics (t(20) = 0.41, p value = 0.07). The nutritional composition of foods in the pictures was assessed by Lemos et al. [ 53 ] via the Food Standard Agency (FSA) guidelines, which measure nutrients per 100 g. Foods scoring four or more are unhealthy. The UPF in the present study had high FSA scores (FSA score = 15.7; standard deviation (SD) = 7.32), indicating low nutritional quality. In contrast, UMPF scored well below the cutoff and were considered healthy (FSA score = − 4.82; SD = 2.96). The nutritional composition of UPF is usually worse than that of UMPF [ 72 ]. Although FoPNLs are not directly based on the Nova classification system, many UPF in Brazil contain FoPNLs that indicate high levels of sodium, saturated fat, or added sugar [ 73 ]. The pictures of UPF and UMPF were paired in terms of sweet and salty tastes. Experimental study arms The participants were randomly allocated into three experimental study arms: (1) the barcode label (Control), (2) the triangle label and (3) the octagon label study arms. In all study arms, the pictures presented were the same except for the label presented along with the UPF pictures. Study arm 1 was the control condition, where UPF pictures were presented along with a barcode label ( n = 83). A barcode label was used as a control label, as in previous labeling research (e.g. [ 74 ]), . In study arm 2 , the participants ( n = 84) were shown UPF pictures with triangle WL. In study arm 3 , the participants ( n = 80) were shown UPF pictures with octagon WL. Each picture had only one WL. The distribution of the participants across the three experimental study arms is shown in Fig. 2 . Fig. 2. Open in a new tab Distribution of the participants across the three experimental study arms. Study arm 1 was the control condition, where UPF pictures were presented (mixed with UMPF and IAPS pictures) along with a barcode label. In study arm 2 , participants were shown UPF pictures with a triangle WL. In study arm 3 , participants were shown UPF pictures with an octagon WL The criteria used to determine whether the UPF items depicted in the pictures would receive a WL were based on the PAHO nutrient profile model [ 75 ]. All the UPF depicted in the pictures presented an excess of one of the following critical nutrients: sugar, sodium or saturated fat. Although the PAHO nutrient profile model includes other nutrients (total fat, trans fat and other sweeteners), we focused on the effects of WL with information on sugar, sodium and saturated fat because these nutrients are considered important to include in WL in most Latin American countries. Evaluative report: hedonic valence and arousal We applied the SAM scale [ 47 ] to obtain the hedonic valence and arousal ratings for each presented picture. The SAM scale uses nonverbal pictographic scales that are easy and quick to apply (see Supplementary Material 1). Valence is assessed via a pictorial scale with manikins and interspersed spaces between them that range from “smiling-happy” (score = 9) to “frowning-unhappy” (score = 1). Arousal is measured via manikins and interspaces between them and reflects a range from an “excited wide-eyed” figure (score = 9) to a “relaxed-sleepy” figure (score = 1). We collected data remotely by utilizing an adapted version of the normative rating procedure of the IAPS for which the remote format was previously validated [ 65 ]. The instructions used can be accessed at http://labnec.sites.uff.br/2023/03/02/experimental_tools/ . The digitized SAM scale (size [height, width] ≅ 13.9°, 35.4°) was positioned centrally on the participant’s computer screen. The method of rating involved selecting the manikin (or the spaces between them) option in each dimension that most accurately represented the individual’s emotional perception of each picture. The participants were limited to selecting a single response option for each picture from each dimension and were required to click on their choice by using a mouse or touchpad. Assessment of hunger Prior to the experimental session, we measured the participants’ subjective levels of hunger via a hunger scale [ 76 ]. The components of the hunger scale included items such as the amount of time since their last meal, subjective hunger, the amount of their favorite food that they could consume, and the estimated time until their next meal. We combined the items of the hunger scale as proposed by Tapper & Turner [ 77 ], which resulted in a hunger score. Experimental procedure Remote behavioral psychology experiments necessitate several experimental controls to ensure internal validity [ 78 ]. In the current experiment: (i) each participant was addressed by a member of our research team who was responsible for carrying out the experiment with them individually; (ii) the participants were informed ahead about the basic technological conditions for completing the experiment, which included access to the task via a computer or notebook, an appropriate internet connection, and a quiet room; (iii) The experimenter briefed the participants not to talk, eat, or do anything else during the experiment; (iv) The experimenter informed the participant, using text and graphics, on how to estimate the size of the computer screen and to maintain a set distance from the computer of approximately 47 cm to allow for the estimation of stimuli size in terms of visual angle; (v) the stimuli were presented through the PsychoPy 3.0 ® version 2021.2.0 [ 79 ] software, which enables a wide range of behavioral science research to be conducted remotely, while maintaining the control settings required for the IAPS’s normative rating technique [ 65 , 66 ]; (vi) Perceptual properties of pictures, such as brightness, contrast, spatial frequency, and complexity, which may influence brain responses and subjective assessments to emotional pictures, were controlled for both UPF and UMPF pictures. Additional details of the experimental protocols are described in Lemos et al. [ 65 ], which can be consulted for further information if desired. Tools for remote data collection The software used to present the stimuli and collect responses in the remote environment was Psychopy 3.0v ® , version 2021.2.0, which is a free tool used to conduct a wide variety of studies in the behavioral sciences [ 79 ]. Visual stimuli were shown in a browser via HTML and JavaScript. The experimental scripts were exported to a webpage ( https://pavlovia.org/ ) for access from anywhere. To collect sociodemographic data, as well as data on eating habits (including hunger scale scores) [ 76 ]), we used Google Forms. The experimenters conducted the remote sessions individually with each participant, a procedure that proved necessary to replicate the findings of the in-person normative IAPS protocols in a remote format [ 65 ]. Sequence of events The experimenter instructed the participant via WhatsApp on the day before the experiment’s scheduled date to measure the screen size of the computer or notebook that would be used for the experiment and to provide information about their height and weight so that the body mass index (BMI) could be calculated. To conduct the experiment, the researcher then contacted the subject via WhatsApp on the appointed day. The entire process was carried out remotely by the participants using a notebook or personal computer. Initially, a link to Google Forms was emailed to the participants, which began with the ethics consent form. Sociodemographic information was gathered following an electronic signature on the ethics consent form. A text message including general directions on how to complete the SAM scale was then sent to the participants. This text was based on the “instructions for adults participants” found in the IAPS instruction manual [ 66 ]. The participants were shown a didactic video created by our research team that detailed the steps for completing the SAM scale after viewing each picture. The video can be seen at labnec.sites.uff.br/2023/03/02/experimental_tools/. After completing this step, the experimenter provided the participant with instructional pictures and another didactic video (labnec.sites.uff.br/2023/03/02/experimental_tools/), emphasizing the importance of using a computer or notebook for the experiment and positioning it in a quiet room to avoid distractions. They were also reminded to use eyeglasses or contact lenses to adapt their vision if needed. Afterward, the participants were instructed to report the computer screen’s size (diagonal length) previously measured and to stand 47 cm away from the computer. The mean reported computer screen size was 15.9” (SD = 3.43). The pictures (UPF, UMF, and IAPS) took up around 47% of the computer screen area, with all labels (WL and barcode) taking up 4.3% of the upper right corner of the UPF picture area (Fig. 3 ). To control for visual variability, all food images were standardized so that only the upper 4.3% of each image contained either the warning label or a barcode of identical size, position, and luminance (Fig. 3 ). Participants were instructed to evaluate the pictures , not the labels, ensuring that any modulation of hedonic valence or arousal reflected incidental processing of label cues under tightly controlled visual conditions [ 80 ]. Coulthard et al. [ 80 ] showed that geometric front-of-pack labels influence food choices even when barely fixated upon, indicating implicit, peripheral processing. Similarly, Storcksdieck genannt Bonsmann et al. [ 81 ] found that simplified, salient formats remain effective despite limited attention, guiding behavior through rapid, low-effort perception. Despite this consideration, to assess explicit label perception, participants were asked at the end of each experiment: “Did you notice a label in the top right corner of the food pictures?” Overall, 96.36% of participants reported noticing the labels. The size of the pictures (UPF, UMF, and IAPS) and the visual angle of the labels were approximately as follows: pictures (height: 18.1°, width: 26.3°); labels (height: 4.9°, width: 4.9°). The visual angle of the pictures was large enough to evoke emotional responses [ 82 ]. Fig. 3. Open in a new tab Example of UPF pictures for each study arm Finally, after answering any participant questions, the experimenter provided a link to the pavlovia.org page containing the behavioral (affective rating task) experiment. Affective rating task: trial structure Nine additional IAPS pictures were used for training (3 positive, 3 negative, and 3 neutral pictures) prior to the start of the experimental trial. This procedure gave the participants an overview of the emotional content of the pictures that would be presented during the experiment. The participants were shown the following instructions after the training: “Click the mouse or touchpad to begin the experiment when you’re ready”. The experimental session began with a screen showing the following instructions for 5 s: “Get ready, the experiment is about to begin”. This message appeared only at the beginning of the experimental session and served to prepare the participant for the start of the experiment. The sequence of events for each trial is described in Fig. 4 (top panel). Each trial began with the sentence “Observe the next picture”, which was shown for 2 s. Next, a picture was presented for 6 s, during which the participants simply looked at the picture with careful attention. The picture was a picture of interest (UPF or UMPF) or a picture from another emotional category of the IAPS, and the participants were not informed about what pictures were of interest in the study. After the picture disappeared, the “Rate the picture” screen appeared. During the 10-second presentation of this last screen, the participants reported how they felt when looking at the picture, classifying its valence and arousal of emotion via the SAM scale [ 47 ]. Fig. 4. Open in a new tab Behavioral experimental design. A Schematic representation of the sequence of events in a trial. The participants clicked with a mouse or touchpad to indicate how they felt after observing each of the 92 pictures (22 food pictures, 11 UPF pictures and 11 UMPF pictures, plus 70 pictures from the IAPS) via the Self-Assessment Manikin scale. B Left: Example of a picture from the IAPS. A picture that resembles the puppy category from the IAPS catalog is displayed, but it is not a part of the IAPS catalog because of permissions for the use of pictures from the catalog. The puppy picture displayed is attributed to kitty.green66, CC BY-SA 2.0 < https://creativecommons.org/licenses/bysa/2.0%3E , via Wikimedia Commons. Middle: Example of a UPF picture. UPF pictures were presented with a barcode label, a triangle label, or an octagon label depending on the study arm. Right: Example of a UMPF picture The participants were randomly assigned to one of the three study arms (barcode/control, triangle WL or octagon WL). The three study arms involved the same pictures. The only difference among the study arms was the label (barcode, triangle WL or octagon WL) that is shown in the UPF picture (Fig. 4 , bottom panel). Data analysis Sample characteristics The demographic and anthropometric characteristics of the sample were summarized via descriptive statistics. Categorical variables (gender, body mass index (BMI), race, area of study and income) are reported as frequencies. Chi-square analysis was applied to the categorical variables to compare the observed frequencies across participants in each study arm (barcode label, triangle label or octagon label). The hunger score was described as a continuous variable and is reported as the median and IQR because it did not present a normal distribution according to the Shapiro‒Wilk test (w = 0.84, p < 0.05). The hunger scores across the study arms were compared through Kruskal‒Wallis ANOVA. A p value < 0.05 was considered to indicate statistical significance for all analyses. Distribution of hedonic valence and arousal ratings in the two-dimensional affective space The hedonic valence and arousal ratings of a heterogeneous set of emotion-laden pictures plotted on a Cartesian plane are arranged in a “boomerang” shape as vectors pointing in two directions [ 51 ]. The upper arm of the boomerang indicates appetitive motivation (approach-like behavior), and the lower arm indicates defensive motivation (avoidance behavior) [ 69 ]. We evaluated whether the distribution of the mean valence and arousal ratings per picture in the two-dimensional affective space fit the typical boomerang shape reported by Lang, Bradley & Cuthbert [ 66 ] in their study on successive North American versions of the IAPS. To this end, the hedonic valence and arousal ratings of each picture were obtained by calculating the average of the rating values attributed to each picture by all the participants and then plotting the averages on a Cartesian plane, with the “y” axis representing the hedonic valence and the “x” axis representing arousal. Spearman’s correlations between the mean valence and arousal ratings were conducted for the upper (appetitive motivation, valence ratings ≥ 5) and lower parts (defensive motivation, valence ratings ≤ 5) of the obtained boomerang to ascertain that the expected vectors pointing in two directions were obtained [ 67 ]. Comparison between UPF and UMPF ratings within each study arm The food pictures were selected based on the method used by Lemos et al. [ 53 ] so that the standardized mean valence and arousal ratings for the UPF pictures were higher than those for the UMPF pictures. We predicted that the valence and arousal values would be greater for the UPF pictures than for the UMPF pictures in the control study arm (replicating the findings of Lemos et al. [ 53 ]) and that the presence of a WL would reduce this difference in the other study arms. To test this hypothesis, we conducted planned comparison Wilcoxon signed-rank tests to compare UPF and UMPF picture ratings within each study arm. Thus, we compared the affective ratings from UPF and UMPF within the same individual. This analysis, which focused on comparing picture ratings for the same individual within each arm of the study, limited the impact of individual variability across the study arms. For these Wilcoxon signed-rank test analyses, a Bonferroni-corrected p value < 0.017 (0.05/3) was considered significant. The rank-biserial correlation was utilized as an indicator of effect size (< 0.10 very small; 0.10–0.29 small; 0.30–0.49 moderate; ≥ 0.5 large) [ 83 ]). The normality assumption was violated for both the arousal and valence ratings ( p < 0.05, Shapiro‒Wilk test). We also calculated the delta between UPF and UMPF ratings (delta = UPF minus UMPF) and conducted a Kruskal-Wallis ANOVA to compare the UPF-UMPF delta directly across study arms. Post-hoc Dunn’s test was conducted when applicable. For the delta analysis, the threshold p-value considered to indicate significance was p < 0.05. Results Sample characteristics The descriptive characteristics of the study sample are shown in Table 1 . Most of the participants were female (67.6%), most were eutrophic (59.9%), and most were white (54.7%), although black/brown individuals accounted for 43.3% of the sample. The majority of the participants were enrolled in graduation courses in health science (54.1%) and reported a household income of less than five times the minimum wage per month (74.1%). As expected, there were no significant differences in terms of gender, body mass index (BMI), household income level, race/ethnicity, or area of study across the three study arms (Table 1 , comparison among study arms). In addition, the hunger scores (median = 0.8, IQR = 3.53) did not differ across study arms (Chi-square = 5.22, p > 0.05), nor did age (mean = 22 years old, IQR = 4.0; Chi-square = 0.14; p > 0.05). Table 1. Sample characteristics Indicator Total n = 247 Study arm Comparison among study arms Barcode label n = 83 Triangle label n = 84 Octagon label n = 80 Chi-Square/ p value Gender, n (%) Female 167 (67.6) 56 (67.5) 56 (66.7) 55 (68.8) 0.41/0.982 Male 72 (29.1) 25 (30.1) 25 (29.8) 22 (27.5) Nonbinary or fluid 8 (3.2) 2 (2.4) 3 (3.6) 3 (3.8) Body Mass Index (kg/m²), n (%) Underweight (< 18.5) 18 (7.3) 3 (3.6) 8 (9.5) 7 (8.8) 4.39/0.625 Eutrophic (18.5–24.9) 148 (59.9) 51 (61.4) 52 (61.9) 45 (56.3) Overweight (25-29.9) 55 (22.3) 21 (25.3) 17 (20.2) 17 (21.3) Obesity (≥ 30) 26 (10.5) 8 (9.6) 7 (8.3) 11 (13.8) Race/Ethnicity, n (%) White 135 (54.7) 52 (62.7) 44 (52.4) 39 (48.8) 4.18/0.653 Black/Brown 107 (43.3) 30 (36.1) 38 (45.2) 39 (48.8) Asian descent 3 (1.2) 1 (1.2) 1 (1.2) 1 (1.3) Indigenous 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Not reported 2 (0.8) 0 (0.0) 1 (1.2) 1 (1.3) Area of Study, n (%) Health Sciences 133 (54.1) 44 (53.0) 42 (50.6) 47 (58.8) 6.73/0.875 Biological Sciences 36 (14.6) 14 (16.9) 11 (13.3) 11 (13.8) Exact and Earth Sciences 23 (9.3) 6 (7.2) 10 (12.0) 7 (8.8) Engineering 20 (8.1) 9 (10.8) 6 (7.2) 5 (6.3) Applied Social Sciences 17 (6.9) 4 (4.8) 8 (9.6) 5 (6.3) Linguistics, Letters and Arts 12 (4.9) 3 (3.6) 5 (6.0) 4 (5.0) Human Sciences 5 (2.0) 3 (3.6) 1 (1.2) 1 (1.3) Income, n (%) minimum wages 1–2 93 (37.7) 28 (33.7) 33 (39.3) 32 (40.0) 8.94/0.177 2–5 90 (36.4) 26 (31.3) 30 (35.7) 34 (42.5) 5–10 38 (15.4) 15 (18.1) 15 (17.9) 8 (10.0) > 10 26 (10.5) 14 (16.9) 6 (7.1) 6 (7.5) Open in a new tab The body mass index (BMI) cutoffs were based on the WHO cutoff (2000). In 2022, when the data were collected, the minimum wage in Brazil was 1,212 Brazilian reais per month (approximately 250 US dollars). Participants who indicated that they did not want to report their race were not included in the race analysis Two-dimensional affective space The results (Fig. 5 ) demonstrated that the IAPS pictures (open dots) were appropriately dispersed throughout the affective space, exhibiting a standard boomerang-shaped distribution [ 41 , 51 , 53 ]. The upper half of the chart featured pleasant pictures from the IAPS (e.g., erotic images and images of sports, families, nature, and puppies), whereas the lower half featured unpleasant pictures (e.g., pictures of mutilation, illness, pollution, accidents, and disgusting subjects). In the middle of the graph are the neutral pictures from the IAPS (objects). There was a positive correlation between valence and arousal ratings in the upper part of the boomerang and a negative correlation in the lower part of the boomerang for the three study arms ( p < 0.05), confirming the typical boomerang arrangement of pictures with two vectors pointing up and down. The rho value for the Spearman’s correlation analyses (between mean valence and arousal ratings) was run separately for the upper (valence ratings ≥ 5) and lower parts (valence ratings ≤ 5) of the boomerang are described in Fig. 5 , along with the correlation fit lines. Fig. 5. Open in a new tab “Boomerang” pattern of the valence (pleasantness) vs. arousal affective space. The pleasantness and arousal values for each picture were derived by computing the average of the valence and arousal values indicated by the participants for each picture. Each IAPS picture is represented by an open dot in the plot. The square blue symbol represents the mean affective ratings of all UMPF pictures, whereas the square brown symbol represents the mean of all UPF pictures. Regression lines, which are thought to represent the underlying motivational systems of appetite and defense, are shown separately in each plot for pictures with a neutral to pleasant valence (upper arm of the boomerang) and pictures with a neutral to unpleasant valence (bottom arm of the boomerang). Left panel. Barcode label (control) study arm. Middle panel. Triangle WL study arm. Right panel. octagon WL study arm. UPF, ultra-processed foods; UMPF, unprocessed/minimally processed foods; IAPS, International Affective Picture System In addition to the pleasant pictures from the IAPS, the food pictures (Fig. 5 , UMPF and UPF pictures, blue square symbol and brown square symbol, respectively) were situated in the upper part of the graph (appetitive arm of the boomerang-shaped affective space), in agreement with previous studies [ 41 , 51 , 53 , 67 ]. As expected, during the barcode label (control) study arm, the UPF pictures seemed to evoke more intense emotions (greater valence and arousal ratings) than the UMPF pictures did (Fig. 5 left panel), as previously shown [ 53 ]. Notably, UPF and UMPF pictures evoked similar emotional responses in the triangle WL and octagon WL study arms (Fig. 5 , middle and right panels, respectively). UPF vs. UMPF The results of the comparison between UPF and UMPF affective ratings within each study arm are displayed in Table 2 . The participants in the barcode label (control) study arm rated the UPF pictures as more positive and arousing than the UMPF pictures did (valence: W = 763, p < 0.001; effect size = 0.50; arousal: W = 463, p < 0.001; effect size = 0.68). In the triangle and octagon label study arms, the participants rated the UPF pictures as being as positive as the UMPF pictures (valence, triangle label: W = 1452, p = 0.53, effect size = 0.09; octagon label: W = 1115, p = 0.27, effect size = 0.15). The participants also rated the UPF pictures as being as arousing as the UMPF pictures (arousal, triangle label: W = 1108, p = 0.02, effect size = 0.30; octagon label: W = 804, p = 0.04, effect size = 0.30). A p value less than 0.017 was considered to indicate statistical significance, as previously mentioned. Notably, the reduction in the emotional gap between UPF and UMPF across the warning-label study arms was primarily driven by decreased pleasantness ratings for UPF. For arousal, the reduced gap reflected a convergence of UPF and UMPF ratings rather than a consistent decrease in arousal elicited by UPF (Table 2 ). Table 2. Median pleasantness and arousal ratings for the pictures depicting UMPF and UPF by study arm Pleasantness (hedonic valence) - Median (IQR) Study arm UMPF UPF Bar code 6.27 a (1.2) 6.55 b (1.5) Triangle 6.09 a (1.5) 6.27 a (1.5) Octagon 6.27 a (1.5) 6.41 a (1.6) Arousal - Median (IQR) Study arm UMPF UPF Bar code 3.09 a (3.0) 3.45 b (3.0) Triangle 4.18 a (2.4) 4.22 a (2.6) Octagon 3.23 a (3.2) 3.36 a (3.3) Open in a new tab Pleasantness ratings ranged from 1 (extremely unpleasant) to 9 (extremely pleasant), with 5 indicating neutrality. Arousal ratings ranged from 1 (very low arousal) to 9 (extremely arousing). The median rating scores within a row from the same study arm group with different superscript letters ( a, b ) were significantly different ( p < 0.017) according to the Wilcoxon signed-rank test. IQR, interquartile range; UMPF, unprocessed/minimally processed foods; UPF, ultra-processed foods The Kruskal-Wallis’s ANOVA showed that the presence of a WL (triangle label or octagon label) had a significant effect on how participants rated the food pictures (Delta = UPF minus UMPF); χ 2 = 6.52, p < 0.05 (for hedonic valence/pleasantness) and χ 2 = 6.32, p < 0.05 (for arousal), Table 3 . The post-hoc test for hedonic valence delta using Dunn’s test showed significant differences between the barcode label study arm and both the octagon and triangle study arms, p < 0.05 for both comparisons. The octagon and triangles study arms did not differ p = 0.96. The post-hoc test for arousal delta using Dunn’s test showed a significant difference between the barcode label study arm and the octagon label study arm ( p = 0.02), but no significant difference between the barcode label and the triangle label study arm ( p = 0.06). The octagon and triangular study arms did not differ p = 0.59. Table 3. Median of the emotional reactivity index (delta = UPF minus UMPF) obtained from pleasantness and arousal ratings by study arms Barcode label Triangular label Octagon label ∆ Pleasantness, median (IQR) 0.45 a (0.79) 0.14 b (1.18) 0.00 b (0.84) ∆ Arousal, median (IQR) 0.32 a (0.91) 0.10 ab (0.97) 0.00 b (0.72) Open in a new tab The index was calculated by subtracting the mean ratings assigned for the UMPF from the mean ratings assigned for the UPF (delta = UPF minus UMPF). The median rating scores within a row from the same study arm group with different superscript letters ( a, b ) were significantly different ( p < 0.05) UMPF, unprocessed/minimally processed foods; UPF, ultra-processed foods. IQR, interquartile range Discussion The present study used the IAPS methodology and the SAM scale [ 66 ] to investigate whether octagon and/or triangle WL could reduce the difference in emotional responses elicited by UPF pictures versus UMPF pictures. The participants from the barcode study arm (control) rated the UPF pictures as more pleasant and arousing than the UMPF pictures did, as expected [ 53 ]. In agreement with our hypothesis, in the study arms in which the participants viewed the UPF pictures with a WL (triangle label or octagon label), the pleasantness and arousal ratings assigned to the UPF pictures were similar to those assigned to the UMPF pictures. Food is essential to an organism’s survival and is fundamental to its day-to-day existence. Thus, food stimuli in general (UMPF or UPF) are expected to act as drivers of approach behavior through the pleasure they elicit. In contrast to UMPF, UPF may prompt exaggerated approach motivation, which may predispose individuals to consume UPF instead of UMPF [ 53 ]. UPF are made extremely tasty, which increases their hedonic value and the pleasure of eating them [ 84 ]. Both types of WL were successful in modifying UPF to appear as pleasant as UMPFs, reducing the exaggerated positive and arousing emotions they usually evoke [ 41 , 53 ]. These findings may provide support for countries that use or are considering using WL as a FoPNL system to reduce the positive emotional appeal and consumption of UPF [ 17 , 22 ]. WL reduced the difference in pleasantness and arousal between UPF and UMPF pictures, although the reduction seemed to be more pronounced for pleasantness than for arousal. Although both warning label shapes (triangle and octagon) reduced the positive emotional responses to UPF relative to UMPF, their effects were statistically equivalent. Emotional arousal can be defined as a state of physiological activation that induces feelings of alertness or excitement, thereby preparing the individual for action [ 51 ]. The slightly higher arousal observed in the triangle WL study arm may reflect the intrinsic alerting quality of angular shapes, which are perceptually associated with warnings and vigilance [ 85 , 86 ]. This pattern aligns with the widespread cultural use of triangles to denote caution or hazard, whereas the octagon conveys a learned stop signal that has become standardized in Latin-American front-of-package labeling policies [ 13 ]. The comparable reduction in pleasantness across both shapes may suggest that, once they are integrated into the food context, their warning connotation could override subtle perceptual distinctions. This interpretation is consistent with evidence that the meaning attributed to visual design elements, such as shape, is often shaped by associative learning and contextual conventions rather than by purely perceptual differences [ 87 ]. In this sense, consumers might interpret angular or rounded forms according to culturally learned associations that, when embedded in a meaningful context such as a warning label, have the potential to outweigh lower-level perceptual effects. Although arousal levels did not differ significantly across WL conditions, the overall pattern of emotional modulation offers insight into underlying mechanisms. Warning labels reduced the positive valence associated with UPF while maintaining moderate arousal, indicating a shift from approach-related affect to a more neutral evaluation rather than a direct induction of avoidance motivation [ 51 ]. This interpretation is consistent with affective neuroscience models linking decreased pleasantness with diminished reward appraisal. Neurophysiological evidence supports this view: Fernandes et al. [ 57 ] showed that exposure to textual warnings decreased pleasantness ratings while increasing late positive potential amplitudes (a neural marker of emotional arousal), suggesting enhanced evaluative attention to health risk cues. Thus, WLs may capture attention and reduce the affective appeal of UPF without necessarily amplifying avoidance or risk perception, underscoring the need for future research integrating behavioral and neurophysiological measures to clarify these mechanisms. To our knowledge, this is the first study focused on the effect of WL on the emotional responses evoked by UPF pictures relative to UMPF pictures. Consumer decision-making is largely impacted by emotions, and UPF can generate strong emotional responses in individuals [ 41 ], making it difficult for them to resist. Using WL to make UPF at least as emotionally charged as UMPF would be a remarkable accomplishment. Here, we provide novel evidence that using WL as a public health tool effectively reduces the heightened emotional reactions elicited by UPF. Notably, it was feasible to capture the implicit influence of WL on the emotional responses elicited by UPF pictures by using the IAPS methodology [ 65 , 66 ]. Because the food pictures were mixed with background pictures from the IAPS and each study arm was assigned different experimental conditions (barcode label, triangle WL, and octagon WL), the participants were unaware that the food pictures were the images of interest. Here, implicit should be understood as referring to the absence of explicit evaluative or decision-making demands, rather than to unconscious perception, as most participants reported noticing the presence of the labels. Thus, by using the IAPS methodology, we were able to capture individuals’ instinctive and automatic emotional judgments toward UPF and their modulation by WL. The IAPS methodology applied here is very useful for providing insights into the factors affecting consumer behavior that occur implicitly and cannot be reported by consumers [ 88 ]. Overt consumer responses may be affected by social desirability, a type of response bias that leads the consumers to answer questions in a manner that they judge will be viewed favorably by the experimenters [ 89 ]. Interpretation of the present findings should also be situated within the broader scientific debate surrounding the NOVA food classification system. There are criticisms of NOVA, particularly concerns about the distinction between industrial processing and nutritional quality. As highlighted by Astrup and Monteiro [ 90 ], the current scientific debate includes divergent perspectives about the Nova classification system, one emphasizing that the concept of UPF is well-founded and consistently associated with adverse health outcomes and another questioning its conceptual clarity and mechanistic basis. Importantly, recent evidence suggests that these dimensions frequently converge in practice. A recent study by Canella et al. [ 91 ] provided empirical support for the Nova framework, showing that, contrary to what some studies have suggested [ 90 , 92 , 93 ], the NOVA classification is robust and functional. The authors demonstrated that the characteristics of foods and beverages (considering both nutrient and non-nutrient profiles, such as the presence of cosmetic additives and excessive critical nutrients) effectively contribute to the identification of UPF and align with the classification made by trained researchers. According to this study, 97% of UPF include an excessive amount of critical ingredients. The study looked at 10,000 UPF products available in supermarkets and found that almost all contained high sodium, fat, sugar, or other critical ingredients [ 91 ]. Within this context, our findings do not rely on resolving the NOVA debate but rather show that foods classified as UPF consistently evoke stronger affective responses than UMPF, underscoring emotional reactivity as a complementary pathway influencing food choice. Nevertheless, some limitations of this study should be highlighted. The sample of participants was composed of university students only. The reason for this was to facilitate comparisons with earlier studies on emotions that used the normative rating procedure for the IAPS with the paper-and-pencil version of the SAM scale [ 41 , 53 , 67 , 94 , 95 ]. In addition, women made up most of the sample. This is explained by the fact that in Brazil, a greater proportion of undergraduate students in health and biomedical courses are female [ 96 ]. To overcome the fact that the sample was primarily composed of women and students from health and biomedical courses, a similar number of participants fitting this description were assigned to each of the three study arms. Furthermore, UPF stimuli were presented without packaging, allowing us to focus on intrinsic sensorial aspects of the foods and to ensure compatibility with the IAPS methodology, which requires standardized visual backgrounds. Because the study examined the effects of warning labels on intrinsic food attributes, caution is warranted when extrapolating these findings to UPF presented within full packaging contexts. These findings call for future studies on the effect of package elements on emotional reactivity to UPF. The ecological validity of the present study is inherently limited because it was conducted as a supervised psychological experiment under controlled conditions. However, such designs offer strong internal validity, allowing for the precise examination of specific processes that are difficult to isolate in real-world settings. In the context of FoPNL, understanding whether labels influence food choices is not equivalent to understanding why they do so. Controlled psychological and psychophysiological experiments are particularly well suited to addressing mechanistic questions, including affective responses, that are challenging to quantify in ecological or population-based studies. Evidence from controlled studies with high internal validity should therefore be considered alongside findings from larger ecological and behavioral studies. We also acknowledge that the present study focuses on subjective emotional responses (pleasantness and arousal) and does not directly assess all psychological mechanisms that may contribute to the effects of warning labels, such as visual attention, risk perception, or automatic processing. Research on nutrition labeling demonstrates that these mechanisms are often investigated using complementary but methodologically distinct paradigms (e.g., eye-tracking studies for attention), and no single experimental approach can capture all relevant processes simultaneously. Emotional responses, nonetheless, represent a recognized and theoretically meaningful component of food-related decision making, as affective reactions can shape evaluations and influence downstream judgments and intentions. Thus, our findings should be interpreted as preliminary mechanistic evidence that warning labels modulate the emotional appeal of UPF cues, rather than as a comprehensive account of all cognitive pathways involved. Finally, although luminance and contrast were carefully controlled, the background design differed across warning labels, as the control and triangle labels shared a white rounded-rectangle background, whereas the octagon label did not. This difference may have influenced the visual salience of the control and triangle labels and therefore warrants caution when interpreting comparisons across label formats. Importantly, however, nearly all participants reported noticing the labels regardless of whether they were triangle, octagon, or control, and all stimuli occupied the same total size within the pictures, supporting comparable exposure across conditions. Policy implications The current food environment is designed in ways that make it easier to select UPF over UMPF [ 97 ]. Advertising, health and nutritional claims, and enticing packaging strategies provide consumers with multiple cues favoring UPF. For example, in supermarkets, UPF are frequently promoted through prominent displays, whereas UMPF are less visible [ 98 ]. Moreover, UPF are known to elicit stronger affective and reward-related responses than UMPF, contributing to habitual or quasi-addictive consumption patterns [ 53 , 99 , 100 ]. Here, we provide new evidence that warning labels (WL) may alter important features of food environments to make healthier eating more manageable. By reducing the difference between the pleasantness and arousal evoked by UPF and UMPF, WLs have the potential to implicitly modify the emotional aspects that contribute to UPF preference. Although no behavioral outcomes were measured, this affective equalization suggests that WL may help counteract one of the psychological mechanisms sustaining UPF consumption, namely their heightened emotional appeal. Thus, while our results do not demonstrate direct behavioral change, they provide preliminary evidence that WL can modify the implicit affective drivers of food choice, supporting future interventions aimed at creating healthier and more balanced food environments. Conclusion The present investigation broadens the body of evidence regarding the potential mechanisms through which triangle and octagon WL may affect consumer attitudes and actions. This is the first study to demonstrate the utility of WL (triangle label and octagon label) in making UPF appear as pleasant and arousing as UMPF. The results of the present study may aid in the development of public policies that promote sustainable and healthy food environments involving the use of WL among FoPNLs. Supplementary Information Supplementary Material 1. (223.2KB, docx) Acknowledgements Not applicable. Abbreviations FoPNLs Front-of-package nutrition labels IAPS International Affective Picture System PAHO Pan American Health Organization SAM Self-Assessment Manikin UMPF Unprocessed/ minimally processed foods UPF Ultra-processed foods WL Warning labels Authors’ contributions **Silva, Laiz A. A.** : Conceptualization, Methodology, Formal analysis, Investigation, Writing – Original Draft. **Lemos, Thayane C** : Investigation, Formal analysis, Writing – Original Draft. **Gaspar, Sara D. J.** : Conceptualization, Methodology. **Coutinho, Guilherme M.** : Methodology, Formal analysis, Investigation, Writing - Review & Editing. **Oliveira, Leticia** : Writing - Review & Editing. **Pereira, Mirtes G.** : Writing - Review & Editing. [**Rojas**]( https://pubmed.ncbi.nlm.nih.gov/?term=Urquizar+Rojas+CF&cauthor_id=29843449 ) **, Carlos F. U.** : Resources, Writing - Review & Editing. [**Spinillo**]( https://pubmed.ncbi.nlm.nih.gov/?term=Spinillo+CG&cauthor_id=29843449 ) **, Carla G** .: Resources, Writing - Review & Editing. **David, Isabel A.** : Conceptualization, Supervision, Funding acquisition, Formal analysis, Writing – Original Draft. Funding This study was supported by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPQ), and the Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ). Data availability The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate This research was conducted in accordance with the Declaration of Helsinki and was approved by the Research Ethics Committee of the Federal University of Rio de Janeiro - Multidisciplinary Center, Macaé (UFRJ-Macaé, CAAE: 29357820.1.0000.5699). All participants in the study provided informed consent by signing a consent form approved by the local ethics committee. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References 1. Martinez-Steele E, Khandpur N, Batis C, Bes-Rastrollo M, Bonaccio M, Cediel G, et al. Best practices for applying the Nova food classification system. Nat Food. 2023;4:445–8. 10.1038/s43016-023-00779-w. [ DOI ] [ PubMed ] [ Google Scholar ] 2. Monteiro CA, Cannon G, Levy R, Moubarac J-C, Jaime P, Martins AP, et al. NOVA. The star shines bright. World Nutr. 2016;7:28–38. [ Google Scholar ] 3. Louzada ML, da Baraldi C, Steele LG, Martins EM, Canella APB, Moubarac DS. Consumption of ultra-processed foods and obesity in Brazilian adolescents and adults. Prev Med (Baltim). 2015;81:9–15. 10.1016/j.ypmed.2015.07.018. [ DOI ] [ PubMed ] [ Google Scholar ] 4. Louzada ML, da Gabe C. Nova food classification system: a contribution from Brazilian epidemiology. Rev Bras Epidemiol. 2025;28. 10.1590/1980-549720250027. [ DOI ] [ PMC free article ] [ PubMed ] 5. Lane MM, Gamage E, Du S, Ashtree DN, McGuinness AJ, Gauci S, et al. Ultra-processed food exposure and adverse health outcomes: umbrella review of epidemiological meta-analyses. BMJ. 2024;e077310. 10.1136/bmj-2023-077310. [ DOI ] [ PMC free article ] [ PubMed ] 6. Slavin JL, Lloyd B. Health Benefits of Fruits and Vegetables. Adv Nutr. 2012;3:506–16. 10.3945/an.112.002154. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Northcott T, Lawrence M, Parker C, Baker P. Ecological regulation for healthy and sustainable food systems: responding to the global rise of ultra-processed foods. Agric Hum Values. 2023;40:1333–58. 10.1007/s10460-022-10412-4. [ Google Scholar ] 8. Dicken SJ, Batterham RL. Ultra-processed food: a global problem requiring a global solution. Lancet Diabetes Endocrinol. 2022;10:691–4. 10.1016/S2213-8587(22)00248-0. [ DOI ] [ PubMed ] [ Google Scholar ] 9. The Lancet. Ultra-processed foods: time to put health before profit. Lancet. 2025;406:2601. 10.1016/S0140-6736(25)02322-0. [ DOI ] [ PubMed ] [ Google Scholar ] 10. PAHO. Pan American Health Organization. Front-of-package labeling as a policy tool for the prevention of noncommunicable diseases in the Americas 2020. https://iris.paho.org/handle/10665.2/51094 . Accessed 28 Apr 2025. 11. PAHO. Ultra-processed food and drink products in Latin America: Sales, sources, nutrient profiles, and policy implications. Washington, DC: PAHO 2019. https://iris.paho.org/handle/10665.2/51094 . Accessed 21 Aug 2025. 12. Guo Z, Ning Y, Mustafa M. Impact of Five Types of Front-of-Package Nutrition Labels on Consumer Behavior among Young Adults: A Systematic Review. Nutrients. 2024;16. 10.3390/nu16172819. [ DOI ] [ PMC free article ] [ PubMed ] 13. Mota BEF, de Alcântara IC, de Souza PM, Souza GGL, Mais LA, Menezes CC, et al. Commentary: Editorial: Strengthening food labeling policies in Brazil. Front Nutr. 2023;10. 10.3389/fnut.2023.1331250. [ DOI ] [ PMC free article ] [ PubMed ] 14. Rousseau GK, Wogalter MS. Research on warning signs. In: Wogalter MS, editor. Handbook of Warnings. 1st ed. Boca Raton: CRC; 2006. pp. 147–58. 10.1201/9781482289688. [ Google Scholar ] 15. Laughery KR, Wogalter MS. A three-stage model summarizes product warning and environmental sign research. Saf Sci. 2014;61:3–10. 10.1016/j.ssci.2011.02.012. [ Google Scholar ] 16. Rojas CFU, Spinillo CG. Avaliação de advertências: Contribuições do design da informação para avaliação de eficácia comunicacional de rotulagem nutricional frontal. Rev Bras Des Inf. 2021;1:1–28. [ Google Scholar ] 17. Crosbie E, Gomes FS, Olvera J, Rincón-Gallardo Patiño S, Hoeper S, Carriedo A. A policy study on front–of–pack nutrition labeling in the Americas: emerging developments and outcomes. Lancet Reg Health - Am. 2023;18:100400. 10.1016/j.lana.2022.100400. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. FAO, PAHO, UNICEF., Front-of-pack nutrition labelling in Latin America and the Caribbean 2023. 10.4060/cc1545en 19. ISO. 3864 Graphical symbols - Safety colours and safety signs. Part 1: Design principles for safety signs and safety markings. 2nd ed. 2011. 20. Khandpur N, Mais LA, de Morais Sato P, Martins APB, Spinillo CG, Rojas CFU, et al. Choosing a front-of-package warning label for Brazil: A randomized, controlled comparison of three different label designs. Food Res Int. 2019;121:854–61. 10.1016/j.foodres.2019.01.008. [ DOI ] [ PubMed ] [ Google Scholar ] 21. Bopape M, Taillie LS, Frank T, Murukutla N, Cotter T, Majija L, et al. South African consumers’ perceptions of front-of-package warning labels on unhealthy foods and drinks. PLoS ONE. 2021;16:e0257626. 10.1371/journal.pone.0257626. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Bopape M, De Man J, Taillie LS, Ng SW, Murukutla N, Swart R. Effect of different front-of-package food labels on identification of unhealthy products and intention to purchase the products– A randomised controlled trial in South Africa. Appetite. 2022;179:106283. 10.1016/j.appet.2022.106283. [ DOI ] [ PubMed ] [ Google Scholar ] 23. Khandpur N, Amaral Mais L, Bortoletto Martins AP. A comparative assessment of two different front-of-package nutrition label designs: A randomized experiment in Brazil. PLoS ONE. 2022;17:e0265990. 10.1371/journal.pone.0265990. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Mais LA, Mialon M, Hassan BK, Peres JMD, Santos MG dos, Martins APB et al. Do they really support your freedom of choice? FoPNL and the food industry in Brazil. Front Nutr. 2023;9. 10.3389/fnut.2022.921498 [ DOI ] [ PMC free article ] [ PubMed ] 25. Giuberti Coutinho J, Feldenheimer da Silva AC, de Castro IRR, Recine EGIG, Makuta G, Rocha NC, et al. The challenges of front-of-package labeling in Brazil. Front Nutr. 2022;9. 10.3389/fnut.2022.921421. [ DOI ] [ PMC free article ] [ PubMed ] 26. Khandpur N, Swinburn B, Monteiro CA. Nutrient-Based Warning Labels May Help in the Pursuit of Healthy Diets. Obesity. 2018;26:1670–1. 10.1002/oby.22318. [ DOI ] [ PubMed ] [ Google Scholar ] 27. Deliza R, de Alcantara M, Pereira R, Ares G. How do different warning signs compare with the guideline daily amount and traffic-light system? Food Qual Prefer. 2020;80:103821. 10.1016/j.foodqual.2019.103821. [ Google Scholar ] 28. PAHO. Pan American Health Organization. Superior Efficacy of Front-of-Package Warning Labels in Jamaica 2021. https://iris.paho.org/handle/10665.2/53328 (accessed April 28, 2025). 29. Taillie LS, Reyes M, Colchero MA, Popkin B, Corvalán C. An evaluation of Chile’s law of food labeling and advertising on sugar-sweetened beverage purchases from 2015 to 2017: A before-and-after study. PLoS Med. 2020;17. 10.1371/JOURNAL.PMED.1003015. [ DOI ] [ PMC free article ] [ PubMed ] 30. Fretes G, Corvalán C, Reyes M, Taillie LS, Economos CD, Wilson NLW, et al. Changes in children’s and adolescents’ dietary intake after the implementation of Chile’s law of food labeling, advertising and sales in schools: a longitudinal study. Int J Behav Nutr Phys Activity. 2023;20. 10.1186/s12966-023-01445-x. [ DOI ] [ PMC free article ] [ PubMed ] 31. Reyes M, Garmendia ML, Olivares S, Aqueveque C, Zacarías I, Corvalán C. Development of the Chilean front-of-package food warning label. BMC Public Health. 2019;19. 10.1186/s12889-019-7118-1. [ DOI ] [ PMC free article ] [ PubMed ] 32. Bopape M, De Man J, Taillie LS, Ng SW, Murukutla N, Swart R. Effect of different front-of-package food labels on identification of unhealthy products and intention to purchase the products– A randomised controlled trial in South Africa. Appetite 2022;179. 10.1016/j.appet.2022.106283 [ DOI ] [ PubMed ] 33. Jakhar R, Singh DS. Front-of-Pack-Labelling Impact on Purchase Intention: A Systematic Literature Review. Int J Res Publ Rev. 2024;5:3677–83. [ Google Scholar ] 34. Pluta-Olearnik M, Szulga P. The importance of emotions in consumer purchase decisions - a neuromarketing approach. Mark Sci Res Organ. 2022;44:87–104. 10.2478/minib-2022-0010. [ Google Scholar ] 35. Kahneman D. Maps of Bounded Rationality: Psychology for Behavioral Economics. Am Econ Rev. 2003;93:1449–75. 10.1257/000282803322655392. [ Google Scholar ] 36. Köster M, Buabang EK, Ivančir T, Moors A. A value accumulation account of unhealthy food choices: testing the influence of outcome salience under varying time constraints. Cogn Res Princ Implic. 2023;8:4. 10.1186/s41235-022-00459-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Sanjari SS, Jahn S, Boztug Y. Dual-process theory and consumer response to front-of-package nutrition label formats. Nutr Rev. 2017;75:871–82. 10.1093/nutrit/nux043. [ DOI ] [ PubMed ] [ Google Scholar ] 38. Becker S, Bräscher A-K, Bannister S, Bensafi M, Calma-Birling D, Chan RCK, et al. The role of hedonics in the Human Affectome. Neurosci Biobehav Rev. 2019;102:221–41. 10.1016/j.neubiorev.2019.05.003. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 39. Hirschman EC, Holbrook MB. Hedonic Consumption: Emerging Concepts, Methods and Propositions. J Mark. 1982;46:92–101. 10.1177/002224298204600314. [ Google Scholar ] 40. Jiang Y, King JM, Prinyawiwatkul W. A review of measurement and relationships between food, eating behavior and emotion. Trends Food Sci Technol. 2014;36:15–28. 10.1016/j.tifs.2013.12.005. [ Google Scholar ] 41. David IA, Krutman L, Fernández-Santaella MC, Andrade JR, Andrade EB, Oliveira L, et al. Appetitive drives for ultra-processed food products and the ability of text warnings to counteract consumption predispositions. Public Health Nutr. 2018;21:543–57. 10.1017/S1368980017003263. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. David IA, Krutman L, De Andrade JR, Araújo RL, Braga F, Gomes FDS et al. Pistas implícitas e obesidade: estratégias de proteção contra o marketing de alimentos. DEMETRA: Aliment Nutr & Saúde 2016;11. 10.12957/demetra.2016.21777 43. Kaneko D, Toet A, Brouwer AM, Kallen V, van Erp JBF. Methods for evaluating emotions evoked by food experiences: A literature review. Front Psychol 2018;9. 10.3389/fpsyg.2018.00911 [ DOI ] [ PMC free article ] [ PubMed ] 44. Ekman P, Friesen WV. Constants across cultures in the face and emotion. J Pers Soc Psychol. 1971;17:124–9. 10.1037/h0030377. [ DOI ] [ PubMed ] [ Google Scholar ] 45. Ekman P, Levenson RW, Friesen WV. Autonomic Nervous System Activity Distinguishes Among Emotions. Science (1979) 1983;221:1208–10. 10.1126/science.6612338 [ DOI ] [ PubMed ] 46. Lang PJ, Bradley MM, Cuthbert BN. Emotion, attention, and the startle reflex. Psychol Rev. 1990;97:377–95. 10.1037/0033-295X.97.3.377. [ PubMed ] [ Google Scholar ] 47. Bradley MM, Lang PJ. Measuring emotion: The self-assessment manikin and the semantic differential. J Behav Ther Exp Psychiatry. 1994;25:49–59. 10.1016/0005-7916(94)90063-9. [ DOI ] [ PubMed ] [ Google Scholar ] 48. Russell JA, Clement J, Jiwani D, Ridgeway K, Schroeder M. A Circumplex Model of Affect. Volume 39. Ryman; 1980. 49. Dalenberg JR, Gutjar S, ter Horst GJ, de Graaf K, Renken RJ, Jager G. Evoked Emotions Predict Food Choice. PLoS ONE. 2014;9:e115388. 10.1371/journal.pone.0115388. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 50. Gutjar S, Dalenberg JR, de Graaf C, de Wijk RA, Palascha A, Renken RJ, et al. What reported food-evoked emotions may add: A model to predict consumer food choice. Food Qual Prefer. 2015;45:140–8. 10.1016/j.foodqual.2015.06.008. [ Google Scholar ] 51. Bradley MM, Codispoti M, Cuthbert BN, Lang PJ. Emotion and Motivation I: Defensive and Appetitive Reactions in Picture Processing. Emotion. 2001;1:276–98. 10.1037/1528-3542.1.3.276. [ PubMed ] [ Google Scholar ] 52. Sabatinelli D, Lang PJ, Keil A, Bradley MM. Emotional perception: Correlation of functional MRI and event-related potentials. Cereb Cortex. 2007;17:1085–91. 10.1093/cercor/bhl017. [ DOI ] [ PubMed ] [ Google Scholar ] 53. Lemos TC, Coutinho GMS, Silva LAA, Stariolo JB, Campagnoli RR, Oliveira L et al. Ultra-Processed Foods Elicit Higher Approach Motivation Than Unprocessed and Minimally Processed Foods. Front Public Health 2022;10. 10.3389/fpubh.2022.891546 [ DOI ] [ PMC free article ] [ PubMed ] 54. David IA, Gomes FS, Silva LAA, Coutinho GMS, Pacheco LB, Figueira JS, et al. Use of event-related potentials to measure the impact of front-of-package labels on food-evoked emotion. Food Qual Prefer. 2023;111:104995. 10.1016/j.foodqual.2023.104995. [ Google Scholar ] 55. Rosenblatt DH, Summerell P, Ng A, Dixon H, Murawski C, Wakefield M, et al. Food product health warnings promote dietary self-control through reductions in neural signals indexing food cue reactivity. Neuroimage Clin. 2018;18:702–12. 10.1016/j.nicl.2018.03.004. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 56. Gross JJ. Emotion regulation: Affective, cognitive, and social consequences. Psychophysiology. 2002;39:281–91. 10.1017/S0048577201393198. [ DOI ] [ PubMed ] [ Google Scholar ] 57. Fernandes TFdaC, Ferreira NB, Campagnoli RR, Gomes F da, Braga S, David F et al. IA, Impact of textual warnings on emotional brain responses to ultra-processed food products. Front Nutr. 2022;9. 10.3389/fnut.2022.895317 [ DOI ] [ PMC free article ] [ PubMed ] 58. Ma Q, Wang C, Wu Y, Wang X. Event-related potentials show taste and risk effects on food evaluation. NeuroReport. 2014;25:760–5. 10.1097/WNR.0000000000000171. [ DOI ] [ PubMed ] [ Google Scholar ] 59. Meule A, Kübler A, Blechert J. Time course of electrocortical food-cue responses during cognitive regulation of craving. Front Psychol. 2013;4:669. 10.3389/fpsyg.2013.00669. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 60. Chen J, Fan Y, Zhang M, Wu S, Li H. The neural model of front-of-package label processing. Nutr Rev. 2024;82:374–88. 10.1093/nutrit/nuad060. [ DOI ] [ PubMed ] [ Google Scholar ] 61. De Houwer J, Teige-Mocigemba S, Spruyt A, Moors A. Implicit measures: A normative analysis and review. Psychol Bull. 2009;135:347–68. 10.1037/a0014211. [ DOI ] [ PubMed ] [ Google Scholar ] 62. Ha O-R, Lim S-L. The role of emotion in eating behavior and decisions. Front Psychol. 2023;14. 10.3389/fpsyg.2023.1265074. [ DOI ] [ PMC free article ] [ PubMed ] 63. Coutinho GMS, Lemos TC, Almo A, Campagnoli RR, David IA. A cor vermelha nas embalagens e o consumo de alimentos ultraprocessados doces. Neurociências Sociedade. 2024;1:e024004. 10.22409/nes.v1i1.64278. [ Google Scholar ] 64. Folkvord F. The Psychology of Food Marketing and (Over)eating. London: Routledge; 2019. 10.4324/9780429274404. [ Google Scholar ] 65. Lemos TC, Silva LAA, Gaspar SDJ, Coutinho GMS, Stariolo JB, Oliveira PGMR, et al. Adaptation of the normative rating procedure for the International Affective Picture System to a remote format. Psicol Refl Crit. 2024;37:41. 10.1186/s41155-024-00326-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 66. Lang PJ, Bradley MM, Cuthbert BN. International affective picture system (IAPS): Affective ratings of pictures and instruction manual. Technical report A-8. Gainesville: 2008. 67. Miccoli L, Delgado R, Rodríguez-Ruiz S, Guerra P, García-Mármol E, Fernández-Santaella MC, Meet OLAF. a Good Friend of the IAPS! The Open Library of Affective Foods: A Tool to Investigate the Emotional Impact of Food in Adolescents. PLoS ONE. 2014;9:e114515. 10.1371/journal.pone.0114515. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 68. Podsakoff PM, MacKenzie SB, Podsakoff NP. Sources of method bias in social science research and recommendations on how to control it. Annu Rev Psychol. 2012;63:539–69. 10.1146/annurev-psych-120710-100452. [ DOI ] [ PubMed ] [ Google Scholar ] 69. Bradley MM, Hamby S, Löw A, Lang PJ. Brain potentials in perception: Picture complexity and emotional arousal. Psychophysiology. 2007;44:364–73. 10.1111/j.1469-8986.2007.00520.x. [ DOI ] [ PubMed ] [ Google Scholar ] 70. Marin MM, Leder H. Examining Complexity across Domains: Relating Subjective and Objective Measures of Affective Environmental Scenes, Paintings and Music. PLoS ONE. 2013;8:e72412. 10.1371/journal.pone.0072412. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 71. Datta R, Wang JZ. ACQUINE. Proceedings of the international conference on Multimedia information retrieval, New York, NY, USA: ACM; 2010, pp. 421–4. 10.1145/1743384.1743457 72. Louzada ML, da Martins C, Canella APB, Baraldi DS, Levy LG, Claro RB. Ultra-processed foods and the nutritional dietary profile in Brazil. Rev Saude Publica. 2015;49. 10.1590/S0034-8910.2015049006132. [ DOI ] [ PMC free article ] [ PubMed ] 73. Borges CA, Khandpur N, Neri D, Duran AC. Comparing Latin American nutrient profile models using data from packaged foods with child-directed marketing within the Brazilian food supply. Front Nutr. 2022;9. 10.3389/fnut.2022.920710. [ DOI ] [ PMC free article ] [ PubMed ] 74. Taillie LS, Hall MG, Gómez LF, Higgins I, Bercholz M, Murukutla N, et al. Designing an Effective Front-of-Package Warning Label for Food and Drinks High in Added Sugar, Sodium, or Saturated Fat in Colombia: An Online Experiment. Nutrients. 2020;12:3124. 10.3390/nu12103124. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 75. PAHO. Pan American Health Organization Nutrient Profile Model. 2016. https://iris.paho.org/handle/10665.2/18621 . Accessed 28 Apr 2025. 76. Grand S. Color-word interference: An investigation of the role of vocal conflict and hunger in associative priming. J Exp Psychol. 1968;77:31–40. 10.1037/h0025759. [ DOI ] [ PubMed ] [ Google Scholar ] 77. Tapper K, Turner A. The effect of a mindfulness-based decentering strategy on chocolate craving. Appetite. 2018;130:157–62. 10.1016/j.appet.2018.08.011. [ DOI ] [ PubMed ] [ Google Scholar ] 78. Rodd JM. Moving experimental psychology online: How to obtain high quality data when we can’t see our participants. J Mem Lang. 2024;134. 10.1016/j.jml.2023.104472. 79. Peirce J, Gray JR, Simpson S, MacAskill M, Höchenberger R, Sogo H, et al. PsychoPy2: Experiments in behavior made easy. Behav Res Methods. 2019;51:195–203. 10.3758/s13428-018-01193-y. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 80. Coulthard S, Hooge I, Smeets M, Zandstra E. Nudging food into a healthy direction: The effects of front-of-pack implicit visual cues on food choice. Int J Food Des. 2017;2:225–40. 10.1386/ijfd.2.2.225_1. [ Google Scholar ] 81. Storcksdieck genanntB, Stefan, Marandola G, Ciriolo E, Bavel R. van., Wollgast Jan. Front-of-pack nutrition labelling schemes : a comprehensive review. Luxembourg: 2020. 10.2760/436998 82. De Cesarei A, Codispoti M. When does size not matter? Effects of stimulus size on affective modulation. Psychophysiology. 2006;43:207–15. 10.1111/j.1469-8986.2006.00392.x. [ DOI ] [ PubMed ] [ Google Scholar ] 83. López-Martín E, Ardura D. El tamaño del efecto en la publicación científica. Educación XX1 2023;26. 10.5944/educxx1.36276 84. Gearhardt AN, Schulte EM. Is Food Addictive? A Review of the Science. Annu Rev Nutr. 2021;41:387–410. 10.1146/annurev-nutr-110420-111710. [ DOI ] [ PubMed ] [ Google Scholar ] 85. Bar M, Neta M. Humans Prefer Curved Visual Objects. Psychol Sci. 2006;17:645–8. 10.1111/j.1467-9280.2006.01759.x. [ DOI ] [ PubMed ] [ Google Scholar ] 86. Palumbo L, Ruta N, Bertamini M. Comparing Angular and Curved Shapes in Terms of Implicit Associations and Approach/Avoidance Responses. PLoS ONE. 2015;10:e0140043. 10.1371/journal.pone.0140043. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 87. Spence C, Van Doorn G. Visual communication via the design of food and beverage packaging. Cogn Res Princ Implic. 2022;7:42. 10.1186/s41235-022-00391-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 88. Stasi A, Songa G, Mauri M, Ciceri A, Diotallevi F, Nardone G, et al. Neuromarketing empirical approaches and food choice: A systematic review. Food Res Int. 2018;108. 10.1016/j.foodres.2017.11.049. [ DOI ] [ PubMed ] 89. Podsakoff PM, MacKenzie SB, Podsakoff NP. Sources of Method Bias in Social Science Research and Recommendations on How to Control It. Annu Rev Psychol. 2011;63:539–69. 10.1146/annurev-psych-120710-100452. [ DOI ] [ PubMed ] [ Google Scholar ] 90. Astrup A, Monteiro CA. Does the concept of ultra-processed foods help inform dietary guidelines, beyond conventional classification systems? Debate consensus. Am J Clin Nutr. 2022;116:1489–91. 10.1093/ajcn/nqac230. [ DOI ] [ PubMed ] [ Google Scholar ] 91. Canella DS, Pereira Montera Vdos, Oliveira S, Mais N, Andrade LA, Martins GC. Food additives and PAHO’s nutrient profile model as contributors’ elements to the identification of ultra-processed food products. Sci Rep. 2023;13:13698. 10.1038/s41598-023-40650-3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 92. Visioli F, Marangoni F, Fogliano V, Del Rio D, Martinez JA, Kuhnle G, et al. The ultra-processed foods hypothesis: a product processed well beyond the basic ingredients in the package. Nutr Res Rev. 2023;36:340–50. 10.1017/S0954422422000117. [ DOI ] [ PubMed ] [ Google Scholar ] 93. Braesco V, Souchon I, Sauvant P, Haurogné T, Maillot M, Féart C, et al. Ultra-processed foods: how functional is the NOVA system? Eur J Clin Nutr. 2022;76:1245–53. 10.1038/s41430-022-01099-1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 94. Paes J, de Oliveira L, Pereira MG, David I, Souza GGL, Sobral AP, et al. The Perception of Aversiveness of Surgical Procedure Pictures Is Modulated by Personal/Occupational Relevance. PLoS ONE. 2016;11:e0160582. 10.1371/journal.pone.0160582. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 95. Silva HD, Campagnoli RR, Mota BEF, Araújo CRV, Álvares RSR, Mocaiber I, et al. Bonding Pictures: Affective Ratings Are Specifically Associated to Loneliness But Not to Empathy. Front Psychol. 2017;8. 10.3389/fpsyg.2017.01136. [ DOI ] [ PMC free article ] [ PubMed ] 96. INEP. Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira. Censo da Educação Superior 2022: Principais Resultados. 2023. https://download.inep.gov.br/publicacoes/institucionais/estatisticas_e_indicadores/notas_estatisticas_censo_escolar_2022.pdf . Accessed 12 Ma 2024. 97. de Menezes RCE, Oliveira JS, Almendra R, de Lira PIC, Costa EC, Leal VS, et al. Influence of food environment on ultra-processed drinks consumption among an economically vulnerable population in a metropolitan area in Brazil: A multilevel analysis. Health Place. 2022;77:102869. 10.1016/j.healthplace.2022.102869. [ DOI ] [ PubMed ] [ Google Scholar ] 98. Borges CA, Gabe KT, Canella DS, Jaime PC. Caracterização das barreiras e facilitadores para alimentação adequada e saudável no ambiente alimentar do consumidor. Cad Saude Publica 2021;37. 10.1590/0102-311x00157020 [ DOI ] [ PubMed ] 99. Monteiro CA, Cannon G. The foods that are addictive. Addiction. 2023;118:603–4. 10.1111/add.16101. [ DOI ] [ PubMed ] [ Google Scholar ] 100. Gearhardt AN, Bueno NB, DiFeliceantonio AG, Roberto CA, Jiménez-Murcia S, Fernandez-Aranda F. Social, clinical, and policy implications of ultra-processed food addiction. BMJ. 2023;e075354. 10.1136/bmj-2023-075354. [ DOI ] [ PMC free article ] [ PubMed ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Material 1. (223.2KB, docx) Data Availability Statement The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. Articles from BMC Nutrition are provided here courtesy of BMC ACTIONS View on publisher site PDF (2.6 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top