Accurate and easy-to-understand nutrition labeling is a worthy public health goal that should be considered an important strategy among many to address obesity and poor diet. Updating the Nutrition Facts Panel on packaged foods, developing a uniform front-of-package labeling system and providing consumers with nutrition information on restaurant menus offer important opportunities to educate people about food’s nutritional content, increase awareness of reasonable portion sizes and motivate consumers to make healthier choices. The aims of this paper were to identify and discuss: (1) current concerns with nutrition label communication strategies; (2) opportunities to improve the communication of nutrition information via food labels, with a specific focus on serving size information; and (3) important future areas of research on nutrition labeling as a tool to improve diet. We suggest that research on nutrition labeling should focus on ways to improve food labels’ ability to capture consumer attention, reduce label complexity and convey numeric nutrition information in simpler and more meaningful ways, such as through interpretive food labels, the addition of simple text, reduced use of percentages and easy-to-understand presentation of serving size information.
In the past four decades, obesity in both adults and children has increased dramatically.1,2 The rapid rise is thought to be due largely to changes in the food and physical activity environments, given the relative stability of the population’s gene pool over this time. Energy-dense, nutrient-poor foods are conveniently available and heavily marketed.3, 4, 5, 6, 7 In addition, the past two decades have seen a proliferation of restaurants, increased snacking, decreased family meals and greater consumption of meals prepared outside the home.8, 9, 10, 11 The growth in portion sizes of packaged and restaurant food have been implicated in increasing obesity prevalence.12,13 Portions of French fries, hamburgers and sugar-sweetened beverages have more than doubled in size,12,14 and a robust body of research has found that people consume more when served larger portions.15, 16, 17
The USDA’s 2010 Dietary Guidelines advise Americans to control total caloric intake and reduce sodium, saturated fat, trans fat, cholesterol and added sugar consumption.18 The provision of clear and accurate nutrition information is one important way to help consumers adhere to these guidelines and make informed choices. Nutrition labels on food packaging and restaurant menus offer one of the best ways to disseminate and make salient such information at the point-of-purchase, when it is arguably most likely to influence purchasing behavior. In addition, required disclosure of nutrition information can incentivize food manufacturers to improve the nutrient profile of their products.19
Recent global food policy efforts have focused on providing consumers with greater access to easy-to-understand nutrition information. In the United States, the Food and Drug Administration (FDA) has expressed interest in updating the Nutrition Facts Panel (NFP) on packaged foods to improve its clarity20 and undertook an initiative21 to recommend a uniform, front-of-package (FOP) labeling system that could be adopted by the food and beverage industries.22,23 In addition, a menu labeling mandate, included as part of the 2010 Patient Protection and Affordable Care Act,24 will require chain restaurants with ⩾20 locations to provide calorie information on restaurant menus at the point-of-purchase.
Although a growing number of studies have examined effective ways to communicate nutrition information through the NFP and newer labeling initiatives, there is still much to learn. In addition, fewer studies have focused specifically on educating consumers about appropriate serving and/or portion sizes via nutrition labels. Therefore the aims of this paper were to identify and discuss: (1) current concerns with nutrition label communication strategies; (2) opportunities to improve the communication of nutrition information via food labels, with a specific focus on serving size information; and (3) important future areas of research on nutrition labeling as a tool to improve diet. In this paper, serving size refers to the amount of a food recommended for consumption in one sitting, while portion size refers to the actual amount of food a person portions out for consumption in one sitting.25 Portion size and serving size are related concepts, but they exert different influences on the amount of food consumed. In this paper, we discuss the ways in which serving size information can influence consumer perceptions of appropriate portion sizes, which in turn influence the amount consumed.13,14
The NFP on packaged foods
The passage of the Nutrition Labeling and Education Act of 1990 required the provision of standardized nutrition information through the NFP on most packaged foods in the United States.26 Although some nutrition information on the NFP can vary based on the food product, the standard label includes information about serving size, kilocalories (kcal; calories) and calories from fat, total, saturated and trans fat, cholesterol, sodium, total carbohydrates, dietary fiber, sugars and protein. The NFP also displays information for certain vitamins and minerals. Nutrient amounts are presented in grams and milligrams accompanied by percentages derived from recommended daily allowances or daily values (based on a single serving for a 2000 kcal diet).26,27
Consumer use of the NFP
Across studies, approximately half of American adults report using the NFP when making food-purchasing decisions, suggesting it is an important source of information for consumers.28, 29, 30, 31 More specifically, 54% of adult respondents in the 2008 Health and Diet Survey reported using the NFP ‘often’ when buying a product for the first time, and >60% reported they ‘often’ or ‘sometimes’ accessed information about calories and serving size.28 In a similar sample of adults, 53% reported using the NFP ‘always or almost always’ when making food-purchasing decisions.29 However, objectively measured viewing of the NFP with eye-tracking technology suggests these self-reported estimates of label usage may be inflated.32 Graham and Jeffery32 found that only 9% of 203 adult participants viewed the NFP calorie content during a food purchasing task, despite 33% self-reporting that they ‘almost always’ used it when food shopping.32 Similarly, although 31% reported ‘almost always’ looking at the total fat content on the NFP, the eye-tracking data revealed that <1% actually did.
Although intended for use by the entire population, nutrition labels are more likely to be used by those who are well-educated, Caucasian, female and/or young adults33 as well as by those with healthier eating habits, higher incomes and greater nutrition knowledge.33,34 A greater proportion of non-NFP users tend to be overweight, Black or Hispanic, unmarried and male.29 Unfortunately, it is not uncommon to find low NFP use among population groups who stand to benefit most from it.35 Design limitations of the current NFP might partially explain why it is an under-utilized source of nutrition information. There is, however, evidence that people with chronic disease (that is, hypertension, diabetes, heart disease) report greater nutrition awareness and food label use compared with those without chronic disease.36
Addressing concerns about the NFP
The problem of serving size label inconsistencies
All of the nutrient information presented on the NFP hinges upon the listed serving size. Serving size labels are created by food manufacturers based on Reference Amounts Customarily Consumed (RACC) Per Eating Occasion described in common household measurements appropriate to the type of food.26,37,38 The RACC were originally determined by the FDA based on Nationwide Food Consumption Surveys conducted in the late 1970s and 1980s. In instances when survey data were inadequate, other sources were considered, including dietary guidance recommendations and serving sizes used by manufacturers, grocers and other countries. One concern with continuing to use the RACC is that stated serving sizes of commonly consumed items, such as cereal and punch, have been found to be substantially less than what is realistically portioned out by consumers.39,40
Another concern is that serving sizes for packaged food can vary over a wide range. Current guidelines state that one unit of a food commodity can be considered a single serving if it weighs between 67% and 200% of the RACC.37 Usually the nutrition information for a food containing ⩾200% of the RACC is based on a single RACC serving, and the package indicates the number of servings it contains. However, a packaged food container weighing >200% of the RACC can also be considered a single serving if the food manufacturer believes that the entire container can reasonably be eaten during a single eating occasion.38 Although in 2004 the FDA encouraged food manufacturers to label foods usually consumed in one sitting as a single serving, there has not been a formal mandate to do so.26 Given the increase in portion sizes over time, it is unclear whether the RACC should also be overhauled to reflect what consumers are actually eating or if an increase in the RACC would inadvertently promote further overconsumption; these questions warrant further study.
The current FDA guidelines allow food manufacturers flexibility to define the amount of a single serving of their product.41 This means that two very similar products could appear to have different nutritional profiles depending on the serving size.41 For example, Mohr et al.41 identified that the RACC for a regular candy bar (typically consumed in one sitting), is 40 g. At the time, they found that this was the serving size listed for the Endangered Species Brand Milk Chocolate and Peanut Butter bar. However, one serving of a 3 Musketeers Bar and a Milky Way Bar was listed as 23 g (57.5% of the RACC), and thus a serving of those candy bars appears to be half the calories of a serving of the Endangered Species bar.41 These kinds of discrepancies in serving size within the same food category have been documented for products, such as granola bars, yogurt, soup and candy bars.41
Mohr et al.41 call this kind of serving size manipulation ‘health framing’, because consumers who view items with smaller serving sizes are prone to incorrectly perceive the product as healthier than a comparable product with a larger serving size. In one study, Mohr et al.41 randomized 151 participants from an Internet panel to view a pizza and soup product where the unit weight of the product and product serving sizes were manipulated. The study revealed that health framing (presenting smaller serving sizes) reduced the anticipated guilt of consuming the product and increased the intent to purchase the product. This effect was moderated by level of dietary concerns; health framing led those with high dietary concerns to experience significant reductions in anticipated guilt. This suggests that those most concerned with nutrition, and therefore more likely to read the NFP, might also be most vulnerable to the negative effects of health framing. The results from this study indicate that health framing might influence consumers at the point-of-purchase to buy a specific product or choose that product over similar ones. However, it is unknown whether such health framing impacts consumers at the point of consumption. It is possible that the smaller serving size advocated on the packaging influences consumers to eat less. Alternatively, the health frame might create an initial health halo that persists, which could translate into greater consumption;42 these are important questions for future research to address.
Concerns about consumer numeracy and literacy
National and international surveys have found that >90 million Americans have limited literacy skills,43 which raises concerns about the amount of numeric and technical information on the NFP. Several studies have documented consumer difficulty understanding quantitative information presented on food labels, especially with respect to serving size information44, 45, 46 and the percentages of recommended daily amounts.47, 48, 49, 50
Serving size calculations
In one study, portion size estimation skills of primary care patients were assessed by asking them to serve an amount of three foods and one beverage that represented what they thought a single standard serving was for each of the products. Then patients were told what the actual serving size was and were asked to serve that specific amount. The sample consisted primarily of women, half of whom reported having previous nutrition and portion size education. The results revealed that higher literacy (but not numeracy) was associated with greater accuracy when portioning out a single serving of the foods/beverage.44 In another study of 90 health center patients, 86% of the respondents assumed that a unit of packaged food was a single serving even if it contained multiple servings and incorrectly equated calories from a single serving with the caloric content of the entire package.45 After some assistance and prompting from research staff to re-evaluate incorrect answers, people improved only slightly, with 63% of the participants still confusing calories per serving with total calories in the package. Participants in this study who had low levels of education were more prone to incorrectly apply information from the NFP to estimate calories contained per package.45
A study conducted by Rothman et al.46 also examined patients’ ability to read and understand nutrition information on food labels. Only 32% of patients could accurately calculate the number of carbohydrates that would be consumed when drinking a 20-oz. bottled beverage containing 2.5 servings. Only 22% of patients could determine total carbohydrates when presented with nutrition information for two slices of low-carb bread.46 Across study tasks, people consistently made errors when trying to mathematically manipulate serving size information to draw a conclusion about a food’s nutritional profile. In this study, low numeracy and literacy skills were also significantly associated with poor understanding of nutrition labels. Finally, another study found that as little as 10.5% of college students could correctly describe serving size from the NFP after viewing different food labels.51 The results from these studies highlight the difficulty people have manipulating and using the numeric information presented on labels with respect to serving size, particularly for foods containing multiple servings.
Addressing serving size inconsistencies through labeling
One proposed way to address the confusion around serving size is to change the NFP design to include two columns: one that contains nutrition information for a single serving and one that contains nutrition information for the entire container, particularly if it is a packaged food or beverage typically consumed in one sitting.
A study by Antonuk and Block52 randomized undergraduate students to either a single- or dual-labeled NFP appearing on a package of 50 M&Ms they could eat while watching a short video. The study found that non-dieters exposed to the dual-column NFP reduced their consumption of M&Ms; dieters M&M consumption did not differ between groups. Although dieters ate significantly fewer M&Ms compared with non-dieters when exposed to the single-column label, the dieters and non-dieters in the dual-column group did not significantly differ in the amount of M&Ms consumed. These results suggest that the dual-label column approach has the potential to encourage healthier eating behavior.
In addition, Lando and Lo53 conducted an online study examining the dual-column NFP format. Approximately, 9500 participants recruited from an Internet panel were randomized to one of the 40 study arms. The study used a 10 (label format) × 2 (product category: frozen meal or a bag of chips) × 2 (healthy versus less healthy food) design. The tested labeling formats presented nutrition information as either: (i) two servings per container with nutrient information listed per serving in a single column (five different versions), (ii) two servings per container with a dual column: one column listing nutrient information per serving and the other listing information per package (three different versions), or (iii) one serving per container, with nutrient information listed per serving in a single column (two different versions). The different label versions also involved the removal of calories from fat and/or enlarged font for calories. The current NFP, with two servings per container, served as the control.
Results revealed that participants rated products as less healthful when they were labeled with one serving per container. Relative to the current NFP, participants could more accurately determine the nutrient content of a product when it was labeled with a single column containing one serving or when information was presented in dual columns (per serving information in the first column and information per package in the second column). When products had the same NFP format, there were no significant differences in participants’ ability to select the more healthful of the two products. However, when comparing products with different NFP formats, the greatest proportion of participants could accurately identify the more healthful product (75%) and calories per container (68%) when a dual-column label was compared with a two servings, single-column label (the current NFP format). Enlarging the font size for calories and removing ‘calories from fat’ did not independently affect label usability.
The findings from these two studies suggest that the addition of a second column presenting nutrient and calorie information for an entire package, rather than per serving, might be more helpful for the consumer. However, such a format would mean adding more information to an already complex and busy label. Therefore, the option of a single column for products typically consumed in one sitting, with the serving size based on the entire package, might be preferred. However, before adopting this new labeling scheme for the NFP, additional research should compare the dual- and single- column labels to even simpler presentation formats that provide less information and use creative methods to interpret the information for the consumer, including Traffic Light labels or other graphical displays. Such labeling schemes must also be tested during real-world shopping trips.
Reducing the amount of complicated information on the NFP
Taylor and Wilkening26 explain that great care was taken when designing the original NFP to consider research ‘about comprehension, legibility, and literacy, taking into account the needs of the elderly and others with sight limitations.’54 For example, specific design elements were added to improve usability, such as the inclusion of lines between nutrients, the removal of punctuation marks, the use of larger type and upper and lower case letters, instead of only uppercase, and the bolding of important nutrient information. The NFP is also displayed in a box with a white background to make it stand out from the food packaging.
Although designed to be easy-to-use, infrequent use of the NFP, particularly by certain demographic groups, might be partially explained by the large amount of complicated information presented on the label. In Graham and Jeffery32 eye-tracking study, most consumers typically only viewed the top five lines of the label, suggesting that much of the additional information may rarely get read, except perhaps by highly nutrition-conscious consumers. The bottom half of the label also presents additional information about grams/milligrams of nutrients based on a 2000 versus 2500 kilocalorie diet. As the FDA discusses altering the NFP to improve usability, it would be worth considering whether all of this information should remain or if a better approach is to include less but more meaningful and salient information. The small font of the NFP has also been cited as a deterrent to its use.34
Another concern with the NFP is the use of percentages, which were originally included to put the nutrition information in the context of an overall daily diet and enable easy comparison across nutrients.26 However, research has found that consumers have trouble understanding and using percentages on food labels.47, 48, 49, 50 One solution proposed by the Center for Science in the Public Interest is the inclusion of high/med/low text next to nutrient amounts to aid understanding of the percentage of daily values.55 The inclusion of such text has been found to improve FOP label understanding, especially among groups of lower socioeconomic status and education levels.49 In FDA online educational materials (see Figure 1), consumers are informed that <5% of a nutrient is ‘low’ and >20% of a nutrient is ‘high.’ These criteria could also be used as the basis for text indicators and/or text could replace percentages entirely. FDA online materials also use different colors and text to educate consumers about the nutrients that should be limited (for example, total fat, cholesterol, sodium) and those consumers must ‘get enough of’ (for example, dietary fiber, vitamins).56 These kind of text labels might further aid NFP comprehension and should be studied. NFP clarity might also be improved by sacrificing technical accuracy to communicate more effectively with the consumer. For example, ‘dietary fiber’ could be listed as ‘fiber’55 and ‘sodium’ as ‘salt.’ Overall, more research is needed to identify strategies to communicate complicated nutrition information to consumers in meaningful ways, rather than relying exclusively on numeric data (for example, kcal, grams, milligrams, percentages). This is especially important given that those with low literacy and/or numeracy skills have particular difficulty comprehending the NFP.