Notes

Abstract
Nutrition information on food labels is an important source of nutrition information but is typically underutilized by consumers. This review examined whether consumer nutrition knowledge is important for communication of nutrition information through labels on packaged foods. A cognitive processing model posits that consumers with prior knowledge are more likely to use label information effectively, that is, focus on salient information, understand information, and make healthful decisions based on this information. Consistent with this model, the review found that nutrition knowledge provides support for food label use. However, nutrition knowledge measures varied widely in terms of the dimensions they included and the extensiveness of the assessment. Relatively few studies investigated knowledge effects on the use of ingredient lists and claims, compared to nutrition facts labels. We also found an overreliance on convenience samples relying on younger adults, limiting our understanding of how knowledge supports food label use in later life. Future research should 1) investigate which dimensions, or forms, of nutrition knowledge are most critical to food label use and dietary decision making and 2) determine whether increases in nutrition knowledge can promote great use of nutrition information on food labels.

Keywords
Nutrition knowledge; Food label use; Nutrition labels; Claims; Ingredient lists
Introduction
Nutrition information on food labels could be a cost-effective method of communicating nutrition information to consumers because the information appears at the point of sale for most packaged foods (Campos, Doxey, & Hammond, 2011). Although consumers value nutrition when deciding which foods to buy (Glanz, Basil, Maibach, Goldberg, & Snyder, 1998), nutrition information on food labels is complex and does not always live up to its potential to communicate effectively (Drichoutis et al, 2009; Hieke, Taylor, 2012; Lin, Yen, 2010 ; Wills et al, 2009). Prior knowledge has been shown to support performance on complex tasks in the cognitive literature; however, its role in food label use is less clear. In this review, we examine the literature surrounding the effects of nutrition knowledge on food label use to examine the state of literature on whether knowledge is important for food label use.

We draw on the cognitive science literature to illustrate how knowledge could support food label use. In particular, we assume that food label use relies on a set of interrelated processes centered on comprehension: attention, comprehension and memory, and decision making (see shaded portion of Fig. 1). Consumers pay attention to information on a food label, comprehend it, and store the information at least long enough to apply it to a food-related decision.

Cognitive processes underlying use of food labels.
Fig. 1.
Cognitive processes underlying use of food labels.
Figure options
Knowledge has been credited with providing the power to perform these key cognitive processes. The phrase “knowledge is power,” often credited to Sir Francis Bacon, has been used widely to convey the centrality of knowledge to human and artificial intelligence (Feigenbaum, 1989). The Long-Term Working Memory model (Ericsson & Kintsch, 1995) describes how knowledge supports cognition. Specifically, the model states that knowledge facilitates cognition by providing retrieval structures which link information in working memory (a limited attention buffer) with long-term memory (stored knowledge), so that newly learned information can be integrated with existing knowledge stores for later use. This results in a long-term working memory system, which represents the speed of access, associated with working memory, with the durability and capacity associated with long-term memory. Knowledge is powerful because it renders attention, comprehension, memory, and decision making processes more efficient (Chiesi et al, 1979 ; Ericsson, Kintsch, 1995).

Based on this work, as well as findings surrounding the effects of knowledge on perceptual processes and information overload (Charness et al, 2001 ; Jacoby et al, 1974), nutrition knowledge could support the use of nutrition information on food label use in at least three ways. First, prior knowledge could enable consumers to pay attention to important information on a food label, and to ignore marketing features that do not reflect salient nutritional qualities, which in turn minimizes information overload. Second, prior nutrition knowledge can facilitate comprehension of, and memory for, food label nutrition information (e.g., determining whether 700 mg represents a little or a lot of sodium). Third, prior nutrition knowledge could support the application of the comprehended and remembered information to food choice.

Nutrition knowledge could be important for dietary choice in other ways, for example, by having direct effects on food choice, without food label information, or by impacting attitudes or beliefs. In addition, food label use could be a moderator of the association between nutrition knowledge and dietary behaviors (Fitzgerald et al, 2008 ; Satia et al, 2005). There have been excellent reviews conducted in the past 5 years addressing knowledge effects on dietary intake (Spronk, Kullen, Burdon, & O'Connor, 2014) as well as a broad range of consumer attributes and behaviors such as attitudes, perceptions, and food choice (Bonsmann, Wills, 2012; Campos et al, 2011; Hieke, Taylor, 2012; Lähteenmäki, 2013 ; Wills et al, 2012). However, in this review, we limit the focus of our inquiry to the effects of knowledge on food label use in an attempt to better understand whether and how knowledge supports food label use.

Food label use constructs and information type

We review the literature on food label use related to three types of food label information that are most central to conveying nutrition and health information: nutrition labels, ingredient lists, and claims. Typically, food label use studies focus on nutrition labels; however, ingredient lists and health/nutrient claims also play important roles in conveying the products' diet and health information to consumers and, for this reason, are regulated in the US by the Food and Drug Administration (FDA). The European Commission's regulation of food labels was limited to claims until very recently, although food producers voluntarily provided nutrition labels and ingredient lists on most packaged foods (Bonsmann & Wills, 2012). Drawing on past research (Campos et al, 2011 ; Mhurchu, Gorton, 2007), we adopt two broad categories to organize the literature on food label use: whether or how often food labels are used (frequency) and the ability to understand labels (comprehension). Frequency of use and comprehension measures can be further subdivided into subjective (e.g., self-reported assessment of frequency, self-ratings of ability to locate and/or apply information) and objective measures (e.g., experimenter's observation of consumer food label consultation or experimenter's assessment of comprehension using questions scored for accuracy).

Nutrition labels

Over 98% of FDA-regulated processed, packaged foods have Nutrition Facts panels (NFPs) in the US (Legault et al., 2004) and roughly 84% of products in Europe have nutrition labels (Bonsmann, Celemin, & Grunert, 2010). Nutrition labels typically contain information on calories, serving size, and amounts and/or daily values of several macronutrients, vitamins, and minerals (e.g., fats, carbohydrate, calcium). In the US, the content of NFPs is government regulated and must include serving size, calories, nutrients, and percent of daily values of each nutrient. Close to two-thirds of respondents in a survey report using NFPs to make purchasing decisions (Ollberding, Wolf, & Contento, 2010). Most individuals are able to understand at least some basic nutrition information on food labels (Graham, Jeffery, 2011; Grunert, Wills, 2007; Levy, Fein, 1998 ; Miller et al, 1997). However, comprehension accuracy decreases for more complex tasks. For example, Levy and Fein (1998) found that most consumers (78%) accurately identified nutrient differences between two products; however, far fewer (20%) were able to calculate the contribution of a single food to a total daily intake.

Ingredient lists

In addition to non-nutrition information (e.g., additives), ingredient lists contain important nutrition information that can contribute to the consumer's assessment of a food's healthfulness. The US Dietary Guidelines 2010 states that: “The ingredients list can be used to find out whether a food or beverage contains synthetic trans fats, solid fats, added sugars, whole grains, and refined grains.” Ingredient lists provide an account of ingredients within a product in descending order of proportion by weight (i.e., ingredients at the end of the list are present in smaller quantities). The FDA recommends that lists conform to a variety of specifications to enable consumers to be informed (Food and Drug Administration, 2014). For example, basic components of foods must be listed and products containing ingredients consisting of several components must list the components in parentheses. Font size and presentation should conform to federal regulations to maximize readability, but even when they do, font size is a frequent problem for consumers' use of ingredient lists (Mackey & Metz, 2009). Consumers frequently consult the ingredient list portion of food labels. For example, self-reported frequency of ingredient list use (as well as use of nutrition labels and claims) was 52% in one study (Ollberding et al., 2010) and even higher (78%) in another (Norazmir, Norazlanshah, Naqieyah, & Anuar, 2012).

Health and nutrient claims

Health claims are intended to communicate scientifically proven health benefits associated with consuming a particular food (Ippolito, Mathios, 1991 ; Williams, 2005), for example, “low fat diets rich in fiber may reduce the risk of some types of cancer.” One goal of nutrient content claims is to communicate the value or relative amount of a specific nutrient within a food product (e.g., good source of fiber, fat free, low calorie). Claims have been shown to impact how other food label information is processed and to influence other dietary behaviors (Mathios, Ippolito, 1999 ; Williams, 2005). For example, consumers sometimes use claims in place of NFPs (Labiner-Wolfe, Jordan Lin, & Verrill, 2010). On the other hand, claims sometimes have little impact on product evaluations (Garretson & Burton, 2000) and may even be misleading and confusing (Hasler, 2008). However, claim comprehension is higher among those with greater experience and education (Dean et al, 2011 ; Verbeke et al, 2009).

Nutrition knowledge construct

Nutrition knowledge, broadly defined, refers to knowledge of concepts and processes related to nutrition and health including knowledge of diet and health, diet and disease, foods representing major sources of nutrients, and dietary guidelines and recommendations (Axelson, Brinberg, 1992; McKinnon et al, 2014; Moorman, 1996 ; Parmenter, Wardle, 1999). Although some have argued that a narrower definition of nutrition knowledge may be desirable (Axelson, Brinberg, 1992 ; Li et al, 2000), Parmenter and Wardle (1999) suggest that a broad definition of nutrition knowledge is needed to capture the complex and wide-ranging nature of the information used to inform dietary choice. We make a similar argument that the ability to use food labels draws on a wide range of situations and behaviors that could potentially draw on many areas of nutrition knowledge. For example, knowledge of the relationship between diet and cancer may enable consumers to focus on fiber information presented on the nutrition label and whole grains in the ingredient list. Knowledge of dietary recommendations may support applying these pieces of nutrition information to decide whether the food product represents a healthful choice within the context of other foods the individual consumes that day. Consistent with the cognitive literature, the various dimensions of nutrition knowledge may be connected in such a way that they support each other, as an integrated semantic network. In this review, we categorize the literature in terms of the number of dimensions included in the nutrition knowledge assessment.

Materials and methods
The review was restricted to empirical, English-language, peer-reviewed studies examining knowledge effects on food label use. Searches were conducted in electronic databases (CINAHL, Cochrane, PubMed, Proquest, Psychinfo, ScienceDirect, Web of Science) and reference lists of relevant articles and reviews, that were published between June 1999 and June 2014 (including in online first print in 2015). The Nutrition Labeling and Education Act of 1990 mandated compliance with a new set of regulations by May of 1994. We used this time frame to allow a sufficient gap in time for consumers to become familiar with the new labels and researchers to examine consumers' familiarity with labels, which is an important factor for label use (Bialkova & van Trijp, 2010). Similarly, we omit studies investigating relatively new forms of nutrition information, namely, front-of-package symbols, which appear on some products (Hawley et al, 2013; Hersey et al, 2013 ; Vyth et al, 2012).

The following key word combinations to search each database: “knowledge” AND “consumer” OR “label use” OR “use of *labels” OR “attention” OR “comprehension” AND “nutrition * panel OR nutrition* label OR food label*” OR “ingredient list” OR “health claim” OR “nutrition claim” yielded 55 abstracts. Articles were screened for quality in terms of clarity of the descriptions of measures, methods, and findings. We excluded studies that did not include sufficient details of the nutrition knowledge measure to evaluate whether it assessed nutrition knowledge rather than another type of knowledge (e.g., functional foods, diabetes), did not differentiate between nutrition knowledge and constructs such as beliefs, confidence, or attitudes, did not describe in detail or provide examples of food label use questions, or did not differentiate between nutrition knowledge and food label use (n = 13). We also excluded studies with adequate measures of nutrition knowledge and food label use when associations between the two measures were not reported (n = 8). We coded food label use measures in terms of frequency of use and comprehension, and within that, self-reported and objective measures; we coded nutrition knowledge assessments in terms of self-reported and objective measures. These criteria were coded by the authors; agreement between raters was good (over 95%), and discrepancies were resolved through discussion.

Results
The final pool of articles (n = 34) is shown in Table 1. Each was coded in terms of the location, sampling method, food label area examined, and dimensions included in the nutrition knowledge assessment as well as the source of the measure. We found wide variation in sampling methods and thus representativeness of the samples, including convenience samples from college students, online panels, random samples of food shoppers in one or more stores, as well as random selections of households representing the entire country. We also found a variety of nutrition knowledge assessments, ranging from a single-nutrient focus to a multidimensional approach, most typically employing Parmenter and Wardle's (1999) measure. Table 2 summarizes the findings in terms of which studies reported a positive association between nutrition knowledge and food label use by type of measure. In the sections that follow, we present the findings for each food label area. Although we did not exclude studies based on age, none of the studies included individuals under the age of 17.

Table 1.
Studies of nutrition knowledge and food label use for Nutrition Labels (NL), Claims (Cl), and Ingredient Lists (IL).