https://orcid.org/0000-0003-0241-2378
Accessible Data
Abstract
Objective
This study assessed the impact of nutrient warnings on product selection and ability to identify food products high in nutrients of concern in Colombia.
Methods
In an online experiment (May-June 2023), Colombian adults were randomized to a nutrient warning, guideline daily amounts (GDA), Nutri-Score, or no-label condition (n = 8,004). Participants completed selection tasks between two fruit drinks labeled according to their condition, one high in sugar and one not. Next, participants answered questions about products high in sugar, sodium, and/or saturated fat (“high-in” product). Finally, they selected which label would most discourage them from consuming a high-in product.
Results
Fewer participants (17%) exposed to the nutrient warning indicated they would purchase the high-sugar fruit drink compared to Nutri-Score (27%, Holm-adjusted (adj) p<0.001) and no label conditions (31%, adj p<0.001); there were no differences between the nutrient warning and GDA label (14%, adj p = 0.087). Compared to the nutrient warning, the GDA label was slightly more effective at helping consumers identify which drink was high in sugar (89% versus 92%, adj p<0.001), while the Nutri-Score and no-label conditions were less effective. Compared to all other conditions, nutrient warnings were more effective at helping participants identify that products were high in nutrients of concern, were more effective at decreasing intentions to purchase these high-in products and were perceived as more effective. Nutrient warnings were most often selected as the label that most discouraged consumption.
Figures
Citation: Mora-Plazas M, Higgins ICA, Gomez LF, Hall MG, Parra MF, Bercholz M, et al. (2024) Impact of nutrient warning labels on Colombian consumers’ selection and identification of food and drinks high in sugar, sodium, and saturated fat: A randomized controlled trial. PLoS ONE 19(6): e0303514. https://doi.org/10.1371/journal.pone.0303514
Editor: Rosely Sichieri, State University of Rio de janeiro, BRAZIL
Received: November 14, 2023; Accepted: April 9, 2024; Published: June 10, 2024
Copyright: © 2024 Mora-Plazas et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data and do files are available from Open Source Framework (https://osf.io/5vbzm).
Funding: Funding support was provided by Bloomberg Philanthropies. The funder did not play any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Background
Obesity and diet-related non-communicable diseases have become great health challenges, posing risks to the health and lives of individuals, the well-being of families, and economic development [1, 2]. Colombia is not immune to such health challenges; according to the Colombian National Nutritional Health Surveys (ENSIN) conducted in 2010 and 2015, the prevalence of overweight and obesity increased 5.6 percentage points in school-aged children (5–12 years old), 2.4 percentage points in adolescents (13–17 years old), and 5.2 percentage points in adults (18–64 years old) [3, 4]. Furthermore, a recent study assessing the diseases for which obesity and overweight are risk factors found that in Colombia, obesity and overweight contribute to approximately 9.4% of disability-adjusted life years, 17% of years lived with a disability, and 3% of years of life lost [5].
There is compelling evidence about the link between the shift from consumption of unprocessed foods to ultra-processed foods and the increase in obesity and diet-related non-communicable diseases [6–9]. Ultra-processed foods are generally low in beneficial nutrients like fiber, protein, micronutrients, and bioactive compounds [10–12] and tend to be high in nutrients related to chronic diseases such as sugar, sodium, and saturated fat (“high-in” products) [13]. In Colombia, from 2009 to 2014, per capita sales of ultra-processed foods and beverages increased by 7.7% [14] and an analysis of responses to the ENSIN 2005 and ENSIN 2015 found that both children and adults had worsening diets with increased consumption of “high in” products including sugar-sweetened beverages and processed meats [15].
To address the rising prevalence in overweight, obesity and diet-related non-communicable diseases, scholars, advocates and policymakers are increasingly calling for policies to communicate the health risks of consuming “high in” products and to discourage their consumption [16, 17]. Front-of-package labels have emerged as one promising policy to guide and influence consumers to make healthier food choices and purchasing decisions [18, 19]. Many countries in the world have applied different voluntary or mandatory front-of-package labels such as guideline daily amounts (GDA) labels, Nutri-Score labels, and nutrient warning labels (hereinafter referred to as nutrient warnings). Currently, in Latin America, nutrient warnings are the most common labeling system and are required in Peru, Uruguay, Chile, Mexico, Argentina, Brazil, and Colombia [20]. Emerging evidence suggests that of the different front-of-package labels, nutrient warnings may be most effective at helping consumers to identify “high in” products and discourage them from selecting such products [21, 22].
Colombia is one of the most recent countries to adopt mandatory front-of-package nutrient warnings. On July 30, 2021, the president signed a nutrient warning bill into law [23], and on December 13, 2022, a resolution was passed to implement the law [24]. The law requires nutrient warnings on packages for excess sodium, sugar, saturated fat, trans fat, and if they contain artificial sweeteners. The law is currently being implemented in Colombia, with a final enforcement date of June 2024. Before this law, there was public debate between advocacy coalitions, the Colombian government, and the food industry with regards to which type of front-of-package label should be implemented, with the food and health coalition advocating for a nutrient warning, and the food industry advocating for alternative front-of-package labels, like the GDA label [25, 26].
Thus, data on which type of front-of-package label is most impactful at shifting consumers’ selections, perceptions, and intentions to purchase is timely and critical. While a recent online randomized controlled trial in Colombia, which assessed perceptions of and reactions to different nutrient warning designs, concluded that the octagonal nutrient warnings performed best, compared to circular and triangular nutrient warnings [27], there is a dearth of evidence demonstrating which type of front-of-package label performs best in Colombia. There is also no information as to whether the impact of the front-of-package labels on food and drink selections varies by education among Colombians. This is important to know given concerns that front-of-package labeling systems may have less benefit for people with lower socio-economic status [28].
To inform this discussion, we originally published results from a randomized experiment testing front-of-package labels in Colombia in PLOS ONE in 2022 (doi: 10.1371/journal.pone.0263324) [29], and after publication, the study authors and a reader, independently noticed and then notified PLOS ONE about errors in the calculation of the nutritional information used in the mock front-of-package labels used in this study. We decided to retract our original article [cite retraction once published] and replicated the original study following an identical study protocol and identical sampling procedures, using corrected nutrition labels. This replication study included an additional quality control process in which all nutrition labels were independently created by two dietitians and then compared to ensure all steps of the nutrition label creation process were correct.
As in the original study, the objectives of this study were to identify the impact of nutrient warnings on participants’ selection of “high in” products and ability to identify them, compared to GDA labels, Nutri-Score labels, and a no-label condition in Colombia. Specifically, the primary outcomes were 1) selection of the product high in sugar as the product the participant would rather buy, and 2) correct identification of the product higher in sugar. Secondary outcomes included perceived message effectiveness of the labels, likelihood of purchasing the product in the next week, ability to identify the less healthy product, ability to identify the product with excess of a nutrient of concern, and which label was perceived as most discouraging. We also investigated whether the primary outcomes varied by education level.
Methods
Prior to launching data collection, we pre-registered the design, hypotheses, and analytic plan on ClinicalTrials.gov (#NCT05783726). Procedures and analyses in this study were identical to the previously published and retracted study, with two exceptions: 1) nutritional profiles and corresponding front-of-package labels were updated to correct values, and serving sizes for two products were updated to more closely reflect typical servings in Colombia; and 2) we measured and reported one additional survey item (intentions to purchase breakfast cereals) that was not reported in the previous study. We report this study according to the CONSORT statement (S1 Checklist).
Ethics statement
The online randomized study was approved by the institutional review board at the University of North Carolina at Chapel Hill (#20–0401) and designated as exempt from review at Universidad Nacional de Colombia. Prior to participating in the study, participants read the consent form and provided online written informed consent by proceeding onto the study.
Study design and procedures
We selected labels to test in this study based on which front-of-package labeling systems were most relevant in the public discussion relating to the passing and implementation of Colombia’s labeling law: nutrient warnings, which were proposed by advocacy coalitions and governmental groups, the GDA label (which was promoted by the food industry), and the Nutri-Score label (also promoted by the industry and used in Europe). We hypothesized that nutrient warnings would perform the best at helping consumers identify and reduce selection of foods high in nutrients of concern, as previous research has found nutrient warnings to perform best on such outcomes [30]. The labels tested are shown in Fig 1.
Note. Labels listed above represent the version of each label used on the yogurt (excess sugar).
We selected the octagonal nutrient warnings because they performed best in our previous randomized experiment (compared to circular and triangular nutrient warnings and a control) which investigated front-of-package nutrient warnings [27]. The octagonal nutrient warnings were the label type that most participants selected as discouraging them from purchasing foods and sugary drinks high in nutrients of concern and were scored with the highest perceived message effectiveness (PME) [27]. The nutrient warning tested in this study was a black octagon that contained a statement about the product containing excess of a nutrient of concern (sugar, sodium, or saturated fat). For example, “EXCESO DE AZÚCARES” (Excess sugar). The octagon also contained “MINSALUD” indicating the message was authorized by the Colombian Ministry of Health and the text “EVITAR SU ALTO CONSUMO” (Avoid high consumption). To determine if a product would receive a nutrient warning, we used the final stage nutrient thresholds for sugar, sodium, and saturated fat from Chile’s nutrient warning label law [31].
The GDA label included Spanish text above the GDA figure stating the product serving size. Below the serving size, a row of light blue blocks listed the Calories, total fat, saturated fat, sugar, and sodium per serving, as well as percentages indicating what percent of the GDA the serving contained. Underneath the light blue blocks, Spanish text explained the percentages were based on the guideline daily amounts for a 2,000-Calorie diet [32]. The GDA label was the only label condition that displayed the nutrition information of the product.
The Nutri-Score label system, which is currently used voluntarily in some European countries [33], is a color coded, letter rated (A-E) system. A dark green “A” indicates the healthiest nutritional value and a dark red “E” indicates the least healthy nutritional value. A product’s letter rating is determined based on a point system. A higher point value indicates a less healthy product. The more calories, sugar, sodium, and saturated fat a product contains, the more points it receives. However, a product can also receive negative points for containing fiber, protein, and fruits and vegetables, which can decrease a product’s total points [34].
Finally, we included a no-label condition. Our previous experiment of nutrient warnings in Colombia used a neutral barcode as a control in order to measure perceptions of and reactions to front-of-package labels [27]. However, in this study, we wanted to test actual policies that could be implemented by the Colombian government. It was possible that the government could decide to not implement a front-of-package labelling system (status quo), so we also tested a no-label condition to measure the outcomes of maintaining the status quo compared to implementing the nutrient warnings. We used the Peruvian nutrient warning guidelines to design the size and placement of the label conditions [35].
Product development and applied labels
We selected food and drink products from categories that make up some of the most commonly consumed ultra-processed foods in Colombia [17], and created nutrient profiles for the products modeled after real Colombian ultra-processed products that are high in nutrients of concern (sugar, sodium, and saturated fat). We used three products we had previously tested (fruit drink, oatmeal cookies, and sliced bread) [27], and a graphic designer developed three new products: a no-sugar-added fruit drink, breakfast cereal, and strawberry yogurt. The breakfast cereal contained excess amounts of both sugar and sodium, whereas the other products contained excess amounts of only one nutrient of concern. Therefore, the breakfast cereal contained two nutrient warnings, while the other products only had one. All products contained fictional brand names to avoid the influence of consumer brand loyalty.
For each labelling system, the presence or absence of the label (nutrient warning) or content of the label (Nutri-Score, GDA) depended on the nutritional composition of the product. Table 1 provides each product’s nutritional profile and the corresponding label applied. The back panels of the mock products were not visible; as such, there was no nutrition facts panel available for participants to review.
Prior to launching the replication study, our team implemented a quality control checklist to ensure accuracy of all stimuli and study procedures (S1 Table).
Participants
From May 12, 2023 to June 22, 2023, we recruited adults in Colombia to participate in an experiment. We recruited participants through Offerwise, a market research company with over 300,000 panel participants in Colombia. Inclusion criteria included presently residing in Colombia and being between 18 and 65 years of age. We excluded panel members that participated in our previous study of front-of-package nutrient warnings in Colombia [27]. We set sample quotas for gender to reflect the Colombian population and for education level (half high school graduate or less, half college degree or higher) to ensure our sample was powered to detect differences in the primary outcomes by education level.
Procedures
Participants completed an online survey programmed in Spanish using Qualtrics survey software. After providing informed consent, Qualtrics randomized participants using a simple 1:1:1:1 allocation ratio to one of the four front-of-package label conditions: nutrient warning, Nutri-Score, GDA, or a no-label condition. Participants then completed an online survey (measures described below). After finishing the survey, participants earned a pre-determined amount of points from Offerwise for completing the study. Participants are able to convert points into money once they accumulate a specified amount.
Measures
Our study had two primary outcomes: 1) selection of the less healthy fruit drink as the fruit drink the participant would rather buy and 2) correctly identifying which fruit drink was higher in sugar. These outcomes were selected as primary outcomes because they are key steps on the pathway from labels to discouraging consumption of less healthy foods [30]. Secondary outcomes included the ability to identify the less healthy fruit drink, objective understanding, PME, intentions to purchase the products, and the most discouraging label. All measures were cognitively tested with Colombians of different education levels to make sure the measures were properly adapted to the Colombian context and accessible to all education levels. Measures can be found in S2 Table.
Participants first completed a fruit drink selection task, where they were asked a series of questions about two fruit drinks, one of which was healthier (contained 3 grams of naturally occurring sugar in the fruit and no added sugar) and one of which was less healthy (contained 18 grams of total sugars, including added sugar). We only included fruit drinks in the selection task due to survey space constraints. We used fruit drinks because of the increasing consumption of sugar-sweetened beverages in Colombia [15]. In the selection task, participants were asked to select which fruit drink was higher in sugar (“Which of these products is higher in sugar?”), which they would rather buy (“Which of these products would you rather buy?”), and which was most unhealthy (“Which of these products is MOST unhealthy?”). The fruit drinks were displayed with the labels corresponding to the participant’s randomly assigned label condition. Both the order of the three questions and the order of the two fruit drinks (left or right) were randomized.
Next, participants completed single product assessment tasks. They viewed a prompt that read: “The next questions are about food products. You will look at a few different products and answer questions about each one. Please keep in mind that this study seeks to evaluate your survey responses and not the sale of the product.” Then, they answered a series of questions about the yogurt, cookies, sliced bread, and breakfast cereals, shown according to the participant’s randomly assigned label condition. In this section of the survey, participants answered all questions about one product at a time (with products displayed in random order). The breakfast cereal was always displayed last as the nutrient warning condition contained two labels.
The questions measured whether participants could correctly identify if the product contained excess of the nutrient of concern with the following item: “Do you think this product has excess [sugar, sodium, or saturated fat]?” (yes/no). We measured perceived message effectiveness (PME) of the labels using three items from the UNC PME scale [36, 37] which read: “How much does the label…” “make you worried about the health consequences of consuming this product?” (range from “not at all” (coded as 1) to “very much” (coded as 5)), “make consuming this product seem unpleasant to you?” (range from “not at all” (coded as 1) to “very much” (coded as 5)), and “discourage you from wanting to consume this product?” (range from “not at all” (coded as 1) to “very much” (coded as 5)). Because PME is specifically about labels, we did not measure PME for the no-label condition. The survey then assessed participants’ likelihood of wanting to purchase the product in the next week if it were available, which read “How likely is it for you to want to purchase this product next week, if it were available?” (range from “not at all” (coded as 1) to “very much” (coded as 5)).
Finally, participants were randomly assigned by Qualtrics using a simple allocation ratio to see the yogurt, cookies, or sliced bread again (one product only). However, this time, the product did not include a front-of-package label. Instead, the three label types (nutrient warnings, GDA, and Nutri-Score) were displayed underneath the product and the participant was asked to select the most discouraging label. Participants were asked to select which label would most discourage them from wanting to consume the product (“Which of these labels would discourage you most from wanting to consume this product?”). At the end of the survey, participants answered standard demographic questions.
Analyses
All analyses were conducted in STATA version 17.0. De-identified data used in this study and do files are available at https://osf.io/5vbzm. A two-sided critical alpha of 0.05 was used to assess statistical significance. Prior to the study, we used G.Power 3.1.9.4 to estimate that with a sample of 8,000 (2,000 per arm), alpha of 0.05, and 80% power, we would be able to detect a small effect of d = 0.09 for both primary outcomes (two-sided two-sample t-test for means).
We randomized 10,160 participants with a final analytical sample size of 8,004 (Fig 2). After randomization, 1,999 participants did not complete the study; we used per protocol analysis, so these participants were not included in our analytic sample. Additionally, 157 participants were excluded from analyses for any of the following reasons: missing all outcome data (n = 2), invalid age (n = 6), completing the study in less than two minutes (2), and repeat or overlapping responses (n = 147). In these cases, we dropped repeat responses (keeping each participant’s first response) if they did not overlap and dropped all responses (including the first) if they did overlap. In the analytic sample, analyses were conducted on a complete-case basis. In the single product assessment task where outcomes were analyzed at the participant-product level, we used all available data from any complete participant-product observation.
We calculated unadjusted means (and standard deviations) and percentages for the primary and secondary outcomes. For the secondary outcome of PME, we calculated it at the participant-product level as the average of the three items for each product type (Cronbach’s alpha for each product type >.70) if all items were non-missing. PME was not calculated for participant-product observations with any PME item missing. We then assessed whether primary and secondary outcomes differed by condition compared to the nutrient warnings. We used linear regression for Likert-scale outcomes and PME (treated as continuous) and logistic regression for binary outcomes. For outcomes that were assessed using repeated measures for multiple product types, we used mixed models treating the intercept as random at the respondent level to adjust for repeated measures across multiple products. These models included the between-subjects factor (i.e., label type), the within-subjects factor (i.e., product type), and their interaction.
To evaluate the most discouraging label, we compared the proportions of participants who selected the Nutri-Score and the GDA labels as the one that most discouraged them from consuming products high in sugar, sodium, or saturated fat relative to the proportion who selected the nutrient warning label (pairwise proportions tests).
We applied Holm’s sequentially rejective procedure [38] to all comparisons with the nutrient warnings across the primary and secondary outcomes to account for multiple comparisons.
Finally, to assess whether the effect of label type on the primary outcomes differed by education, we tested for an interaction of nutrient warnings with education level specified as low (high school diploma or less) vs. high (college degree or higher) and used a Wald chunk test to determine the joint interaction. We conducted pairwise comparisons to predict percentages by label type and education level.
Results
Descriptive statistics
Participant characteristics are listed in Table 2.
Fruit drink selection tasks
Selection of the less healthy fruit drink did not differ between the nutrient warnings and GDA labels; however, the nutrient warnings led to lower selection of the less healthy fruit drink compared to the Nutri-Score and no label conditions (Fig 3). Only 17% in the nutrient warning condition and 14% in the GDA label condition selected the less healthy fruit drink for purchase (adj p = 0.087), whereas 27% and 31% in the Nutri-Score and no-label conditions selected the less healthy fruit drink (both adj p<0.001 compared to the nutrient warning condition).
Note. * p<0.001 compared to nutrient warning (adjusted for 19 comparisons using Holm’s sequentially rejective method). Data were analyzed by logistic regressions of the outcomes on indicator variables for the arm. Inference on the predicted percent is based on the delta method. Missing data at the participant level were as follows: 40 (0.5%) for ‘Wanted to purchase the less healthy fruit drink’ (11 in nutrient warning, 12 in no label, 9 in Nutri-score, and 8 in GDA), 16 (0.2%) for ‘Correctly identified the less healthy fruit drink as higher in sugar’ (4 in nutrient warning, 3 in no label, 8 in Nutri-score, and 1 in GDA), and 26 (0.3%) for ‘Correctly identified the less healthy fruit drink as less healthy’ (4 in nutrient warning, 15 in no label, 4 in Nutri-score, and 3 in GDA).
Relative to the nutrient warning condition, the GDA label was more effective at helping consumers identify which fruit drink was higher in sugar, although this difference was small (89% versus 92%, adj p<0.001). In contrast, the Nutri-Score condition and no-label conditions were less effective than the nutrient warnings. Only 67% in the Nutri-Score condition and 69% in the no-label condition correctly identified the fruit drink higher in sugar (both adj p<0.001 compared to the nutrient warnings).
When asked which fruit drink was less healthy, 89% in the nutrient warnings condition made the correct identification, compared to 72% in the Nutri-Score condition and 68% in the no-label condition (adj p<0.001 for both conditions compared to the nutrient warnings). There was no statistically significant difference between the nutrient warning and GDA label conditions.
Interaction of label type and education, primary outcomes
There was no significant interaction between label condition and education level on selection of the fruit drink higher in sugar as the product the participant would rather buy (Wald test of joint significance for the condition-education level interactions: unadj p = 0.091) (S3 Table). There was also no significant interaction between label condition and education level on the likelihood of correctly identifying the fruit drink higher in sugar (Wald test of joint significance for the condition-education level interactions: unadj p = 0.267). Logistic regression estimates for the fruit drink selection task outcomes (with and without interactions with education level) can be found in S4 Table.
Single product assessment of yogurt, bread, cookies, and cereal high in sugar, sodium, saturated fat, or sugar and sodium
In the single product assessment tasks, the nutrient warnings were more effective than the no-label, Nutri-Score, and GDA conditions in helping participants to correctly identify that the products contained excess of a nutrient of concern and more effective in decreasing the participants’ likelihood of wanting to purchase the product if it were available (Table 3; adj p<0.001 for each condition compared to the nutrient warnings). While 79% of participants in the nutrient warnings condition correctly identified that the product contained excess of the nutrient of concern, only 33% in the no-label condition, 37% in the Nutri-Score condition, and 50% in the GDA condition were able to do so (Table 3; adj p<0.001 for each condition compared to the nutrient warnings). The nutrient warnings also led to greater perceived message effectiveness (PME) compared to both the Nutri-Score and GDA (adj p<0.001 for each condition compared to the nutrient warnings; PME not assessed in the no-label condition). Unadjusted results on individual products can be found in S5 Table. Multilevel mixed-effects regression estimates for the single product assessment task outcomes can be found in S6 Table.
Other outcomes
Participants were most likely to select the nutrient warnings as the label that most discouraged them from wanting to consume a product high in sugar, saturated fat, or sodium (Fig 4). Seventy percent of participants selected the nutrient warnings as most discouraging compared to only 19% selecting the GDA label and 11% selecting the Nutri-Score label (adj p<0.001 for both conditions compared to the nutrient warnings).
Note. * p<0.001 compared to nutrient warning (adjusted for 19 comparisons using Holm’s sequentially rejective method). Data are estimated proportions and their logit-transformed 95% confidence intervals. Missing data at the participant level were as follows: 131 (1.6%) (22 in nutrient warning, 27 in no label, 39 in Nutri-score, and 43 in GDA).
Discussion
This online experiment aimed to assess the impact of nutrient warnings on product selection and identification of less healthy products, compared to GDA, Nutri-Score, and no label conditions, among Colombian adults. We found that the nutrient warning performed better than the Nutri-Score and no label condition across outcomes. The nutrient warning performed similarly to or slightly worse than the GDA label on outcomes related to selection between two drinks, including selecting the less healthy drink, correctly identifying the high-sugar drink, and identifying which drink was less healthy. However, the nutrient warning label outperformed the GDA label as well as the Nutri-Score and no label conditions on outcomes assessed for multiple products (assessing one product at a time), including correctly identifying those products contained excess nutrients of concern, intentions to purchase products, and perceived message effectiveness. The nutrient warnings were also most frequently selected as the label that most discouraged consumption of nutrients of concern. The pattern of results illustrates the promise of nutrient warnings as a policy to help Colombian consumers identify and reduce consumption of products high in nutrients of concern.
While this study assessed only consumer reactions, not behaviors, the results are broadly consistent with the literature on nutrient warnings’ impact on behavior, including results from a systematic review and meta-analysis of experiments [39], which found that nutrient warnings are effective at reducing purchases of sugar-sweetened beverages. The results are also consistent with evaluations of real-world nutrient warning label policies, which have found that such policies, in conjunction with other policies, have led to reductions in purchases of foods and drinks high in nutrients of concern [27, 40].
Prior studies have found that GDA labels have performed worse than nutrient warnings [41, 42]. Indeed, our hypothesis was that GDA would perform worse than nutrient warnings in this study. However, the GDA and nutrient warnings performed similarly on outcomes for the drink product selection tasks, and the GDA was slightly better than nutrient warnings at helping consumers identify which of the two fruit drinks was high in sugar (92% vs. 89% of consumers, respectively). One possible reason for the relatively strong performance of the GDA could be due to the design of the selection task. In comparing only two products using the GDA in an online setting, consumers could see that sugar in one drink (20% daily value) was higher than sugar in the other drink (3% daily value). However, in a real life shopping situation, consumers’ ability to use the GDA to identify “high-in” products could be limited, as people have little time to examine labels in detail, need to examine many products at once, and the numerical information provided in the GDA can be difficult to understand quickly [43].
On the other hand, when assessing one product at a time, for multiple products, we found that nutrient warnings performed better than the GDA label across outcomes, including identifying whether a product was high in nutrients of concern. One possibility is that consumers’ ability to use the GDA could be limited if other products are not available for comparison; in other words, there are no other products to compare numerical values to. In contrast, nutrient warnings, which provide a binary signal about nutrient content, do not require other products to be available as a frame of reference for understanding the nutritional profile of a product. In short, the ability to use a given label type may be affected by the number of products being compared. Future research that investigates the performance of front-of-package labels in more realistic food shopping environments with a variable number of products would shed light on this question.
Our study found that nutrient warnings consistently performed better than Nutri-Score labels across all outcomes. The findings of this study are in contrast to some previous studies that have found more favorable outcomes for Nutri-Score [44, 45]. However, such studies have typically involved ranking tasks, while this study was concerned with identifying the less healthy products and the product that was “high in” the nutrient of concern. It is intuitive that Nutri-Score would perform worse than nutrient warnings on this outcome because the Nutri-Score provides a single summary measure or grade on the overall nutritional profile, rather than information about specific nutrients. We selected identification of “high in” products as primary outcomes, rather than ability to rank products, to align with the overarching public health goal of labeling laws in Colombia, which is to help consumers understand and reduce purchases of unhealthy foods.
This study found no evidence that nutrient warnings had differential effects by education level. In other words, it seems unlikely that Colombian consumers with high educational levels would benefit more from a nutrient warning policy than low-educated consumers. However, in the real world, differences in other factors like price or consumer preferences for certain foods may interact with the labeling law to create differences in response by socio-economic status. As nutrient warnings are implemented in Colombia, evaluation research should monitor whether the law differentially impacts consumer understanding and food purchases for high- vs. low-educated consumers.
Strengths and limitations
Strengths of this study include between-subjects randomization allowing for causal inference, the use of standardized questions from previous studies, which have shown appropriate psychometric characteristics [27, 36, 37], as well as the assessment of multiple products for several of our outcomes.
However, one important limitation is that this experiment was conducted during the process of implementation of nutrient warnings in Colombia. Participants may have already been exposed to circular warnings (implemented in a preliminary stage by the Colombian Ministry of Health) as well as octagonal warnings (similar to those used in this study and implemented under the new law) at the time of the study. It is possible that this prior exposure may have amplified or weakened the impact of the warning. Participants may have also been exposed to the GDA warnings, which are voluntarily implemented by the food industry and have been present in the Colombian food supply for many years. However, prior exposure is unlikely to have affected experimental findings due to randomization. More broadly, our results showing the effectiveness of nutrient warnings relative to other label types are consistent with experimental results in countries prior to implementation of nutrient warnings [43, 46–48] lending confidence to these findings. Secondly, there may be limitations to generalizability of study results to the general Colombian population due to our use of an online sample; we sought to mitigate this by including socioeconomically diverse participants from different Colombian regions.
Conclusion
Colombian advocacy groups working on the labeling law in Colombia, as well as international scholars, have emphasized reducing excess consumption of unhealthy foods high in nutrients of concern as a key first step towards obesity prevention [49–51]. In this experiment, nutrient warnings consistently performed better than Nutri-Score and no label conditions at helping consumers to identify products high-in nutrients of concern and discourage their purchases. Nutrient warnings performed similar or slightly worse than the GDA label on outcomes when participants picked between two products and better than the GDA label on outcomes when participants saw one product at a time. Compared to all three label types, nutrient warnings were perceived as most effective and were most frequently selected as the label that most discouraged participants from wanting to consume a less healthy product. The overall results suggest that nutrient warnings are a promising policy for helping Colombian consumers identify and reduce the consumption of unhealthy foods. Given that nutrient warnings are now being implemented in Colombia, evaluation research is needed to understand the impact of nutrient labels on actual “high in” food and beverage purchases, as well as to evaluate the impact across populations with differing socioeconomic status.
Supporting information
S1 Checklist. CONSORT 2010 checklist of information to include when reporting a randomised trial*.
https://doi.org/10.1371/journal.pone.0303514.s001
(DOC)
S3 Table. Predicted probabilities between high and low education by outcome and condition.
*P-values (except for the joint significance of the interaction terms) are for the difference in the contrast with nutrient warning (the reference) by education level (e.g., for no label, the difference between the contrast for secondary or lower, 30.4–16.2, and the contrast for tertiary, 31.3–17). Thus, the effects of no label, Nutri-score, and GDA relative to nutrient warning did not significantly differ by education level. Data analyzed by logistic regressions of the outcomes on indicator variables for the arm, education level, and their interactions. Inference on the contrasts is based on the delta method. Missing data were as follows: 40 (0.5%) for ‘Wanted to purchase the less healthy drink’ (11 in nutrient warning, 12 in no label, 9 in Nutri-score, and 8 in GDA) and 16 (0.2%) for ‘Correctly identified the less healthy fruit drink as higher in sugar’ (4 in nutrient warning, 3 in No label, 8 in Nutri-score, and 1 in GDA).
https://doi.org/10.1371/journal.pone.0303514.s004
(DOCX)
S4 Table. Logistic regression estimates for the fruit drink selection task outcomes (with and without interactions with education level for the primary outcomes).
Standard errors in parentheses. Data analyzed at the participant level.
https://doi.org/10.1371/journal.pone.0303514.s005
(DOCX)
S5 Table. Unadjusted means and proportions by label type for single product assessment task.
https://doi.org/10.1371/journal.pone.0303514.s006
(DOCX)
S6 Table. Multilevel mixed-effects regression estimates for the single product assessment task outcomes.
Standard errors in parentheses. PME not measured in No label. No residual variance for ‘Correctly identified product as having excess nutrients’ (multilevel mixed-effects logistic regression). Data analyzed at the participant-product level.
https://doi.org/10.1371/journal.pone.0303514.s007
(DOCX)
Acknowledgments
The authors would like to thank the UNC Global Food Research Program and Javeriana team. In particular, we thank Dr. Barry Popkin for his insight on the global food policy landscape, Dr. Yazmin Cadena for advising and providing insight into the cognitive interview process, Emily Busey for her assistance developing the images used in this study, Carmen E. Prestemon for programming the survey and assisting with formatting the manuscript, and Cindy P. Evans for assisting in the development of nutritional profiles for all products.
References
- 1. Patterson D, Buse K, Magnusson R, Toebes B. Identifying a human rights-based approach to obesity for States and civil society. Obes Rev. 2019;20:45–56. pmid:31297936
- 2.
GBD 2015 Obesity Collaborators, Afshin A, Forouzanfar MH, Reitsma MB, Sur P, Estep K, et al. Health Effects of Overweight and Obesity in 195 Countries over 25 Years. N Engl J Med. 2017;377(1):13–27. https://doi.org/10.1056/NEJMoa1614362
- 3.
Encuesta Nacional de la Situación Nutricional en Colombia 2015 Bogotá, Colombia: Instituto Colombiano de Bienestar Familiar (ICBF); 2015.
- 4.
Encuesta Nacional de la Situación Nutricional en Colombia 2010 Bogotá, Colombia: Instituto Colombiano de Bienestar Familiar (ICBF); 2010.
- 5. Gil-Rojas Y, Garzon A, Hernandez F, Pacheco B, Gonzalez D, Campos J, et al. Burden of Disease Attributable to Obesity and Overweight in Colombia. Value Health Reg Issues. 2019;20:66–72. pmid:31035116
- 6. Popkin BM, Corvalan C, Grummer-Strawn LM. Dynamics of the double burden of malnutrition and the changing nutrition reality. Lancet. 2020;395(10217):65–74. pmid:31852602
- 7. Pagliai G, Dinu M, Madarena MP, Bonaccio M, Iacoviello L, Sofi F. Consumption of ultra-processed foods and health status: a systematic review and meta-analysis. Br J Nutr. 2021;125(3):308–18. pmid:32792031
- 8. Harb AA, Shechter A, Koch PA, St-Onge M-P. Ultra-processed foods and the development of obesity in adults. Eur J Clin Nutr. 2023;77(6):619–27. pmid:36280730
- 9. Lane MM, Davis JA, Beattie S, Gómez-Donoso C, Loughman A, O’Neil A, et al. Ultraprocessed food and chronic noncommunicable diseases: A systematic review and meta-analysis of 43 observational studies. Obes Rev. 2021;22(3):e13146. pmid:33167080
- 10. Monteiro CA, Cannon G, Moubarac JC, Levy RB, Louzada MLC, Jaime PC. The UN Decade of Nutrition, the NOVA food classification and the trouble with ultra-processing. Public Health Nutr. 2018;21(1):5–17. pmid:28322183
- 11. Monteiro CA, Cannon G, Levy RB, Moubarac JC, Louzada ML, Rauber F, et al. Ultra-processed foods: what they are and how to identify them. Public Health Nutr. 2019;22(5):936–41. pmid:30744710
- 12.
Popkin BM. Ultra-processed foods’ impacts on health. Santiago de Chile: FAO, 2019.
- 13.
PAHO. Pan American Health Organization Nutrient Profile Model. 2016.
- 14.
Alimentos y bebida ultraprocesados en América Latina: tendencias, efecto sobre la obesidad e implicaciones para las políticas públicas. Washington, D.C.: Organización Panamericana de la Salud, Departmento de Enfermedades no Transmisibles y Salud Mental, 2015.
- 15. Mora-Garcia G, Ruiz-Diaz MS, Villegas R, Garcia-Larsen V. Changes in diet quality over 10 years of nutrition transition in Colombia: analysis of the 2005 and 2015 nationally representative cross-sectional surveys. Int J Public Health. 2020;65(5):547–58. pmid:32632458
- 16. Scrinis G, Monteiro CA. Ultra-processed foods and the limits of product reformulation. Public Health Nutr. 2018;21(1):247–52. pmid:28703086
- 17. Khandpur N, Cediel G, Obando DA, Jaime PC, Parra DC. Sociodemographic factors associated with the consumption of ultra-processed foods in Colombia. Rev Saude Publica. 2020;54:19. pmid:32049210
- 18. Kanter R, Vanderlee L, Vandevijvere S. Front-of-package nutrition labelling policy: global progress and future directions. Public Health Nutr. 2018;21(8):1399–408. pmid:29559017
- 19. 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. pmid:36844016
- 20.
Global Food Research Program at UNC Chapel Hill. Front-of-package labeling Chapel Hill, NC2023 [cited 2023 October 30]. Available from: https://www.globalfoodresearchprogram.org/resource/front-of-package-label-maps/.
- 21.
Nohlen H, Bakogianni I, Grammatikaki E, Ciriolo E, Pantazi M, Alves Dias J, et al. Front-of-pack nutrition labelling schemes: An update of the evidence. Luxembourg: JRC Publications Repository, 2022 Contract No.: EUR 31153 EN.
- 22. Song J, Brown MK, Tan M, MacGregor GA, Webster J, Campbell NRC, et al. Impact of color-coded and warning nutrition labelling schemes: A systematic review and network meta-analysis. PLoS Med. 2021;18(10):e1003765. pmid:34610024
- 23.
Ley 2120: Por medio de la cual se adoptan medidas para fomentar entornos alimentarios saludables y prevenir enfermedades no transmisibles y se adoptan otras disposiciones: El Congreso de Colombia; 2021. Available from: http://www.andi.com.co/Uploads/LEY%202120%20DEL%2030%20DE%20JULIO%20DE%202021.pdf.
- 24.
Ministerio de salud y protección social. Resolución Número 2492: Por la cual se modifican los artículos 2, 3, 16, 25, 32, 37 y 40 de la Resolución 810 de 2021 que establece el reglamento técnico sobre los requisitos de etiquetado nutricional y frontal que deben cumplir los alimentos envasados y empacados para consumo human: El Congreso de Colombia; 2022 [cited 2023 October 30]. Available from: https://www.minsalud.gov.co/Normatividad_Nuevo/Resoluci%C3%B3n%20No.%202492de%202022.pdf.
- 25.
Análisis de impacto normativo en la temática de etiquetado nutricional y frontal de los alimentos envasados en Colombia. Ministerio de Salud y Protección Social de Colombia, 2020.
- 26. Mialon M, Gaitan Charry DA, Cediel G, Crosbie E, Scagliusi FB, Perez Tamayo EM. ’I had never seen so many lobbyists’: food industry political practices during the development of a new nutrition front-of-pack labelling system in Colombia. Public Health Nutr. 2021;24(9):2737–45. pmid:32819452
- 27. Taillie LS, Hall MG, Gomez 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(10). pmid:33066130
- 28. Shrestha A, Cullerton K, White KM, Mays J, Sendall M. Impact of front-of-pack nutrition labelling in consumer understanding and use across socio-economic status: A systematic review. Appetite. 2023;187:106587. pmid:37169260
- 29. Mora-Plazas M, Aida Higgins IC, Gomez LF, Hall M, Parra MF, Bercholz M, et al. RETRACTED: Impact of nutrient warning labels on choice of ultra-processed food and drinks high in sugar, sodium, and saturated fat in Colombia: A randomized controlled trial. PLOS ONE. 2022;17(2):e0263324. pmid:35143553
- 30. Taillie LS, Hall MG, Popkin BM, Ng SW, Murukutla N. Experimental Studies of Front-of-Package Nutrient Warning Labels on Sugar-Sweetened Beverages and Ultra-Processed Foods: A Scoping Review. Nutrients. 2020;12(2). pmid:32098363
- 31.
Ley de alimentos: manual de etiquetado nutricional: Ministerio de Salud ‐ Gobierno de Chile; 2019. Available from: https://www.minsal.cl/ley-de-alimentos-manual-etiquetado-nutricional/.
- 32.
GDA Labelling: The Food and Drink Federation: Delivering Sustainable Growth; 2019. Available from: https://www.gdalabel.org.uk/gda/gdalabel.html.
- 33. Deschasaux M, Huybrechts I, Julia C, Hercberg S, Egnell M, Srour B, et al. Association between nutritional profiles of foods underlying Nutri-Score front-of-pack labels and mortality: EPIC cohort study in 10 European countries. BMJ. 2020;370:m3173. pmid:32938660
- 34.
Nutri-Score, a simple labelling system for nutritional value: COLRUYTGROUP; 2020. Available from: https://nutriscore.colruytgroup.com/colruytgroup/en/about-nutri-score/.
- 35.
Manual on Advertising Warnings approved pursuant to the provisions of Law No. 30021, Law to Promote Healthy Eating for Children and Adolescents, and its Regulations approved by Supreme Decree No. 017-2017-SA, (2018).
- 36. Baig SA, Noar SM, Gottfredson NC, Boynton MH, Ribisl KM, Brewer NT. UNC Perceived Message Effectiveness: Validation of a Brief Scale. Ann Behav Med. 2019;53(8):732–42. pmid:30321252
- 37. Grummon AH, Hall MG, Taillie LS, Brewer NT. How should sugar-sweetened beverage health warnings be designed? A randomized experiment. Prev Med. 2019;121:158–66. pmid:30772370
- 38. Holm S. A Simple Sequentially Rejective Multiple Test Procedure. Scand J Statist. 1979;6(2):65–70.
- 39. Grummon AH, Hall MG. Sugary drink warnings: A meta-analysis of experimental studies. PLoS Med. 2020;17(5):e1003120. pmid:32433660
- 40. Taillie LS, Reyes M, Colchero MA, Popkin B, Corvalan 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(2):e1003015. pmid:32045424
- 41. Hammond D, Acton RB, Rynard VL, White CM, Vanderlee L, Bhawra J, et al. Awareness, use and understanding of nutrition labels among children and youth from six countries: findings from the 2019–2020 International Food Policy Study. Int J Behav Nutr Phys Act. 2023;20(1):55. pmid:37143053
- 42. Kroker-Lobos MF, Morales-Juárez A, Pérez W, Kanda T, Gomes FS, Ramírez-Zea M, et al. Efficacy of front-of-pack warning label system versus guideline for daily amount on healthfulness perception, purchase intention and objective understanding of nutrient content of food products in Guatemala: a cross-over cluster randomized controlled experiment. Arch Pub Health. 2023;81(1):108. pmid:37328782
- 43. Jauregui A, Vargas-Meza J, Nieto C, Contreras-Manzano A, Alejandro NZ, Tolentino-Mayo L, et al. Impact of front-of-pack nutrition labels on consumer purchasing intentions: a randomized experiment in low- and middle-income Mexican adults. BMC Public Health. 2020;20(1):463. pmid:32252716
- 44. Egnell M, Talati Z, Galan P, Andreeva VA, Vandevijvere S, Gombaud M, et al. Objective understanding of the Nutri-score front-of-pack label by European consumers and its effect on food choices: an online experimental study. Int J Behav Nutr Phys Act. 2020;17(1):146. pmid:33213459
- 45. Egnell M, Galan P, Farpour-Lambert NJ, Talati Z, Pettigrew S, Hercberg S, et al. Compared to other front-of-pack nutrition labels, the Nutri-Score emerged as the most efficient to inform Swiss consumers on the nutritional quality of food products. PLoS One. 2020;15(2):e0228179. pmid:32107489
- 46. 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. pmid:36027994
- 47. Singh S, Taillie LS, Gupta A, Bercholz M, Popkin B, Murukutla N. Front-of-package labels on unhealthy packaged foods in India: Evidence from a randomized field experiment. Nutrients. 2022;14(15):3128. pmid:35956305
- 48. Khandpur N, Mais LA, Sato PD, 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. pmid:31108818
- 49. Popkin BM, Barquera S, Corvalan C, Hofman KJ, Monteiro C, Ng SW, et al. Towards unified and impactful policies to reduce ultra-processed food consumption and promote healthier eating. Lancet Diabetes Endocrinol. 2021;9(7):462–70. pmid:33865500
- 50. Palacio Y. Congreso aplaza nuevamente la sesión para estudiar la Ley de Comida Chatarra. W Radio. 2021 May 11 [Cited 2021 Aug 24]. Available from: https://www.wradio.com.co/noticias/actualidad/congreso-aplaza-nuevamente-la-sesion-para-estudiar-la-ley-de-comida-chatarra/20210511/nota/4133974.aspx.
- 51.
Red Papaz. 2021: Paquete: No Comas Más Mentiras. 2021 April [Cited 2021 Aug 24]. Available from: https://www.nocomasmasmentiras.org/2021-paquete/.
