National Academies Press: OpenBook

Nutrition Labeling: Issues and Directions for the 1990s (1990)

Chapter: 4. Contextual Factors Affecting Food Labeling Reform

« Previous: 3. Current Food Labeling
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 74
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 75
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 76
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 77
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 78
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 79
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 80
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 81
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 82
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 83
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 84
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 85
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 86
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 87
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 88
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 89
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 90
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 91
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 92
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 93
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 94
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 95
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 96
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 97
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 98
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 99
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 100
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 101
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 102
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 103
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 104
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 105
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 106
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 107
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 108
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 109
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 110
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 111
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 112
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 113
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 114
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 115
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 116
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 117
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 118
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 119
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 120
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 121
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 122
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 123
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 124
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 125
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 126
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 127
Suggested Citation:"4. Contextual Factors Affecting Food Labeling Reform." Institute of Medicine. 1990. Nutrition Labeling: Issues and Directions for the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1576.
×
Page 128

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

4 Contextual Factors Affecting Food Labeling Reform A number of external factors have an influence on food labeling reform. Although an exhaustive discussion of all these factors is not possible in this report, a brief review is necessary in order to convey the milieu in which proposals for changes in food labeling will be evaluated. The factors that are considered in this chapter include current dietary patterns, food marketing in the United States, consumer understanding of nutrition and use of food labels, and analytical considerations that affect food labeling information. CURRENT DIETARY PATTERNS OF AMERICANS Since me turn of the century, Americans have made extensive changes in their eating habits. More food is purchased for consumption away from home. Snacking is more common. A wider variety of foods is available year-round. Changes in the composition of foods have occurred due to improved methods of cultivation and animal husbandry, the introduction of new varieties of plants and animals bred for either nutritional or other features, and advances in food processing that permit the formulation of foods with desirable characteristics. The data on which a discussion of these changes is based come primarily from records maintained and surveys conducted by the U.S. Department of Agriculture (USDA) and the U.S. Deponent of Health and Human Services (DHHS). Most of the trend data discussed here are based on the U.S. Food Supply Series and the Nationwide Food Consumption Survey (NFCS) conducted by USDA and the National Health and Nutrition Examination Survey (NHANES) conducted by DHHS. 74

CONTEX7VAL FA=ORS AFFECTING FOOD LABELING REFORM Sources of Data and Issues of Interpretation 75 Major changes have occurred in what and where Americans eat; however, care needs to be taken in the interpretation of the available data The data that are collected represent four different levels: national food supply, use of food in households, individual food intakes, and nutritional and health outcomes that are influenced by diet. A variety of methods is used to collect the data, and the collective evaluation of such data needs to be sensitive to these different methods. Systematic collection of data on the U.S. food supply began in 1909, when USDA initiated tracking the availability of foods in the U.S. marketplace. USDA has been able to calculate annually the approximate amount of food available per individual by dividing the total amount of foodstuffs available by the civilian population of the United States at a given time. The total amount of foodstuffs is calculated as [(food produced + beginning inventories + food import - (food exported + food purchased by the military + year-end inventories + food having nonfood uses)~. Through the use of food composition tables, a rough estimate can then be made of the nutritive value of foods available for consumption by Americans. Refinements of these types of calculations have led to the development of estimates of the per capita availability of 25 nutrients in approximately 350 foods as they "disappear" into the U.S. food distribution system. Data on food availability represent quantities that are larger than those actually eaten, because the amounts do not account for the losses that occur during processing, marketing, and home use. Calculations of the nutritive value of the food supply overlook some sources of nutrients such as alcoholic beverages (which provide calories, but few nutrients) and vitamin and mineral supplements (which provide micronutrients, but essentially no calories). Because the per capita availability of food and nutrients is based on the total U.S. population in any given year, comparisons over time do not take into account the changing demographic structure of the U.S. population. With these caveats in mind, and recognizing that food supply data do not provide information on the actual foods that are eaten, they are still useful for reflecting changes in the overall patterns of the foods and nutrients available over time. At He second and third levels, data are collected by using household- based surveys (i.e., use of food in households and individual food intakes). At approximately l~year intervals since 1936, USDA has conducted its NFCS. The first four surveys collected data only on the use of foods in households. In 1965, data on the consumption of food by individuals were added to the information collected. Surveys were conducted in 1965-1966, 1977-1978, and 1987-1988. Beginning in 1985, USDA began the Continuing Survey of Food Intakes of Individuals (CSFII), which is performed annually except in years when the more comprehensive NFGS is conducted.

76 Nlm~lTION LABELfiNG DHHS conducts a variety of surveys and surveillance activities that provide information on food intakes and the nutritional and health status of the U.S. population and subgroups of the population. The Total Diet Study, which has been conducted annually since 1961 by the Food and Drug Administration (ADA), estimates intakes of certain essential minerals as well as the extent of contamination of foodstuffs by industrial chemicals and pesticides. The Ten-State Nutrition Survey was conducted from 1968 to 1970 to examine the diets and nutritional status of the poor. The first National Health and Nutrition Examination Survey, NHANES I, which was conducted between 1971 and 1974, collects an ambitious set of data, including not only He dietary intakes of individuals but also health and medical histories, physical examinations, and Laboratory data Particular attention is paid to nutrition-related diseases. NHANES II was conducted from 1976 to 1980, the Hispanic HANES (HHANES) was carried out from 1982 to 1984, and NHANES III began in 1988. In 1977, the U.S. Congress directed that a comprehensive, coordinated nutrition monitoring system be created. The National Nutrition Monitoring System (NNMS) was begun as a collaborative program between USDA and DHHS. NNMS is meant to coordinate the survey activities of these two federal departments and to issue joint survey reports through the Interagency Committee on Nutrition Monitoring. Reports issued in 1986 and 1989; these reports compared data from the 1977-1978 NFCS, 198~1986 CS~I, NHANES II, HHANES, and USDA's historical data series (DHHS/USDA, 1986; LSRO, EASED, 1989). Data from the surveys mentioned above and historical data series can be useful in assessing major shifts in eating habits and the resulting nutritional status of Americans; however, the nature of the information requires that these conclusions must be reached with consideration of the limitations of the sources. All of the information, whether derived from food supply data or dietary surveys, is dependent on the quality of estimates and assumptions that permit the calculation of nutrient intakes. Dietary intake surveys of individuals often use a technique called the 24-hour dietary recall, relying on the respondent to accurately recall and describe the foods consumed during the pawing day or in the past 24 hours. Much of the data concerning the per capita disappearance of food consists of gross estimates of the amount of food produced; no adjustments are made for discar - , wasted, and spoiled foods. In addition, the data represent the average amount of food and nutrients available to the population, regardless of differences in age, sex, race, and economic status. However, major shifts in eating habits and the resulting nutritional status of Americans can be discerned from the available data. More detail on changes in food consumption over the years than is offered here are provided in the National Research Council ARCS report, Diet and Health: Implications for Reducing Chronic Disease Risk MARC, 1989) and the

CONTEXTUAL Fx~rORS AFFECTING FOOD LA~EuNG REFORM 77 Expert Panel on National Nutrition Monitoring report, Nutrition Mon~tonng in the United States (LSRO, FASEB, 1989~. Food Energy The daily food energy content (or calories, as it is expressed on food labels) of the food supply is substantially higher (3,500 calories in 1985) than intakes recorded in surveys. The primary changes in food sources of energy in the past 70 years have been from an increase in the percentage of energy contributed by fats, oils, sugars, and sweeteners, along with a decrease in the percentage contributed by grain products. Of the percentage of calories obtained from macronutrients from 1909 to 1985, protein contributed about 11 percent, fats increased from 32 to 43 percent, and carbohy~es fell from 57 to 46 percent (NRC, 1989~. In 1985, the major sources of food energy in the food supply were fats and oils (20 percent); grains (19.9 percent); meat, poultry, and fish (19 percent); and sugars and sweeteners (17.8 percent) (LSRO, FASEB, 1989~. Data from the 1985-1986 CSPII indicate that children aged 1 to 5 have mean energy intakes that fall within the range for their age groups, with little difference observed between blacks and whites or among those from families with different income levels. The reported mean intakes by women, however, fall below the recommended range. In addition, intakes are lower in older (40 to 49 years) than in younger (20 to 29 years) women, lower in blacks than In whites, and lower in women below the poverty level than in those above the poverty level (LSRO, FASEB, 1989~. Despite the low caloric intakes report, more than a quarter of American adults are overweight. Data from NHANES I, NHANES II, and HHANES reveal that women are more likely than men to be overweight (LSRO, FASEB, 1989~. A comparison of NHANES I and health survey data from 1960 to 1962 reveals that the average weight gain was a 3-pound increase for women and a 6-pound increase for men (NRC, 1989~. Data from national surveys reveal that reported caloric intakes have de- creased over time, while the prevalence of overweight individuals has remained the same or increased slightly. Whether an individual's weight changes or is maintained depends on He balance between caloric intake and physical activity, body size, body composition, and metabolic efficiency. The paradox of He re- ported low caloric intakes in conjunction with the high prevalence of overweight individuals in the United States has yet to be fully explained. A number of reasons have been suggested, including underreporting of caloric intake, hered- ity, decreased physical activity, and metabolic mechanisms. However, JNMEC considered it probable that low levels of physical activity have a significant relationship to the high prevalence of overweight individuals observed in He United States (DHHS/US DA, 1986~. Concern exists that recommendations to

78 NUTRITION LABELING reduce caloric intake further without a concomitant increase in nutrient density may compromise nutritional status. Food energy is considered a current public health issue due to the rela- tionship of total caloric intake to both its association with body weight and the implications for overall nutrient intakes and chronic diseases (LSRO, FASEB, 1989~. Fats and Cholesterol The per capita amount of fat available in the food supply, the food sources, and the types of fat have changed considerably during the twentieth century. The amount of fat available per capita has increased to the current level of 169 g/day. The types of fat have also changed. Since 1909, the per capita amount of samraled fatty acids has remained constant at approximately 60 g/day, but the amount of monounsaturated fatty acids has gradually increased to Me current level of 68 g/day, and since the mid-1960s the amount of polyunsaturated fatty acids has doubled to 33 g/day (LSRO, FASEB, 1989~. Cholesterol in the food supply increased from 500 mg/day in 1909 to a peak of 570 mg/day in 1947. It subsequently fell to 480 mg/day in 1977, where it has remained. The decline in the availability of cholesterol is primarily due to the reduced use of eggs, from a peak of 49 pounds per person per year in 1951 to 32 pounds per person per year in 1985 ARC, 1989). Sources of dietary fats in the food supply have also changed. The proportion of total fats from meat, poultry, and fish has changed little, amounting to about 31.4 percent in 1985. Fat from whole mink has declined steadily, from 10.4 percent in the late 1940s to 3 percent in 1985, while a significant increase has occurred in the amount from fats and oils, increasing from 38 to 47 percent during the same period. The proportion of saturated fatty acids from meat, poultry, and fish has changed little since 1909, although there has been an increase in poultry consumption (LSRO, FASEB, 1989~. Since the mid-1960s, a shift in daily product consumption has occurred, with about a 50 percent decrease in whole mink consumption, a doubling of low-fat mink consumption, and a 173 percent increase in the consumption of cheeses (NRC, 1989~. Within the fats and oils group, the proportion of saturated fatty acids obtained from animal sources has declined, while the amount obtained from vegetable sources has increased, due to the use of salad and cooking oils, which has increased from 2 to 25 pounds per capita since 1909 (NRC, 1989~. In 1985, meat, poultry, and fish, fats and oils, and dairy products contributed almost all of the saturated fatty acids to the food supply. The proportion of polyunsaturated fatty acids from meat, poultry, and fish has decreased, while the amount from fats and oils has more than doubled in the past 70 years. ~ 1985, the food groups contributing

CONTEXTUAL FA=ORS AFFECI7NG FOOD lung REFORM 79 most of the cholesterol were meat, poultry, and fish (43 percent), caky products (13 percent), and eggs (39 percent) (LSRO, FASEB, 1989). The most recent data on dietary fat intakes are derived from the Away dietary intakes from the 1985-1986 CSFII. Women aged 20 to 49 and children aged 1 to 5 consumed 37 percent and 35 percent of total calories from fat, respectively. Only about 10 percent of women surveyed had fat intakes below 30 percent of total calories. Fat intakes by women in this survey were higher among whites than among blacks and were higher for those in higher socioeconomic groups. However, race and economic status have been shown to have little to do with the percentage of calories from fat. Samurai fatty acids comprised an estimated 13 percent and 14 percent of calories in the diets of women and children, respectively. Monounsa~ed fatty acids accounted for 13 percent of calories in both groups. Polyunsaturated fatty acids provided 6 percent of calories for children and 7 percent for women. Mean cholesterol intakes were 277 mg/day and 228 mg/day by women and children, respectively. More Ran 25 percent of women had mean cholesterol intakes in excess of 300 mg/day. Estimated intakes by men have remained high, at 423 to 466 mg/day (LSRO, FASEB, 1989~. Comparison of 1-day data from the 1985-1986 CSFII and the 1977-1978 NFGS reveals similar intakes of total fat by children and adult males, whereas for females aged 20 to 49 there appeared to be a decrease of about 10 percent between the surveys. The data can be considered to suggest that only women's fat intakes have actually changed. However, from 1977 to 1985 the percentage of tote calories from fat has declined for children (37 to 35 percent) and adult males and females (42 to 37 percent). This change may be the result of an increase in carbohydrate consumption (LSRO, FASEB, 1989~. Data on mean serum cholesterol levels from national surveys have been compared (NRC, 1989~. For adult men and women aged 20 to 74, the mean serum cholesterol levels have decreased 3 to 4 percent since the early 1960s, and the declines are statistically significant for both men and women, for all whites, but not for blacks (LSRO, FASEB, 1989~. According to the definition of the National Cholesterol Education Program, 36 percent of all adults aged 20 to 74 are candidates for medical advice and intervention for high blood cholesterol levels (Sempos et al., 1989~. In view of the continuing indications of the high per capita availability and higher than recommended intakes of dietary fats and cholesterol, dietary fats are considered to be a current public health issue due to their association with heart disease, certain cancers, and obesity. As a result, the high consumption of total fat, saturated fate acids, and cholesterol has a high priority in public health monitoring NERO, FASEB, 1989~.

80 Carbokydr~es Nl~lNTION LABELING The per capita amount of carbohydrates in the food supply declined from the turn of the century until Me mid-1960s. In the past 70 years there has been a significant decrease in the proportion of carbohy~es obtained from grain products and an increase in We proportion obtained from sugars and sweeteners; in particular, high-fructose corn syrup has replaced sucrose in many products since the 1960s. In 1909, Me proportion of carbohy~s in Me food supply was about ~vo-thirds from complex carbohydrates and one-third from sugar. In 1985, sugars and sweeteners contributed 39.6 percent of the carbohydrates in the food supply, while gIain products, fruits, and vegetables provided most of the remainder (35.8 percent, 6.6 percent, and 9.2 percent, respectively) (LSRO, FASEB, 1989~. Four-day data from the 1985-1986 CSFII showed Mat Me carbohydrate intake for women aged 20 to 49 was 175 g/day, providing 46 percent of calories; and the intake for children aged 1 to 5 was 184 g/day, providing 52 percent of calories (LSRO, FASEB, 1989). The 1977-1978 NFCS found that carbohydrate intake averaged 47 percent of calories for children aged 1 to 8 and 46 percent of calories for females and 45 percent of calories for males aged 9 to 18. Based on 1-day estimates, the general trend during the past two decades seems to indicate an increase in the mean percentage of calories obtained from carbohydrates in most age groups (NRC, 1989~. Dietary Fiber Data on the amount of fiber in the food supply are not available. Dietary fiber sources include whole gins, fruits, and vegetables. Oat bran, beans, and dried fruits provide soluble fiber, while wheat bran is a source of insoluble fiber. Four-day estimated mean intakes of dietary fiber from the 1985-1986 CSFII were 11 g/day for women aged 20 to 49 and 10 g/day for children aged 1 to 5. Only 5 percent of the women surveyed had dietary fiber intakes of 20 g or more per day, as currently recommended by the National Cancer Institute (NCI). One analysis indicated that vegetables, grains, and fruits supplied 50 percent, 30 percent, and 12 percent, respectively, of the dietary fiber consumed by women. finely intake data from the 1985-1986 CSFII indicate Am, on average, the dietary fiber intake by men aged 19 to 50 was 17 g/day, a level higher Man Be dietary fiber intake by women (LSRO, FASEB, 1989~. DietaIy fiber is considered a potential public health issue worthy of further study due to its possible role in reducing the risk of certain chronic diseases (LSRO, FASEB, 1989~.

CONTEXTUAL FACTORS AFFECTING FOOD LABEl]NG RE;FORAl Protein i] 81 Since early in the twentieth century, the food supply has provided approx- mately 11 percent of calories as protein, or the equivalent of about 100 g of protein per person per day ARC, 1989~. Over the years the source of protein has changed from plant sources to increased levels from animal sources. Dunng the period from 1909 to 1913, approximately 52 percent of protein came from animal sources; by 1982, the amount had increased to 68 percent as a result of the increased use of meat, poultry, ash, and daily products, with a concomitant decrease in the use of eggs, flour, cereal products, and potatoes (NRC, 1989~. The amount of protein available per capita is considerably higher Man the 1980 Recommended Dietary Allowances (RDA) for protein, which is 56 g/day for men over age 15 who weigh 70 kg and 44 g/day for women in the same age group who weigh 55 kg (LSRO, FASEB, 1989~. According to We 1977- 1978 NFCS, the average protein intake was 74.3 g/day for all respondents' with race, economic status, region, urbanization, and season having little influence on dietary protein levels (NRC, 1989~. Protein contributed an average of 17 percent of total calories in the diets of males and females. 1h 1985, meat, poultry, and fish supplied 43.4 percent, daily products supplied 20.6 percent, and grain products supplied 19 percent from the food supply (LSRO, FASEB, 1989~. Sodium aru! Potassium Data on the amount of sodium in the food supply are not available. The daily per capita amount of potassium has been declining since 1909, to a level of 3,460 mg in 1985. Dietary sources of sodium include meat, dairy products, some vegetables, and sodium-containing compounds added to foods during processing, preparation, or at the table. The major sources of potassium in the food supply are vegetables; daily products; meat, poultry, and fish; and fruits. Four-day data from the 1985-1986 CSFII provide estimates of individual sodium intakes. The mean intake by women aged 20 to 49 was 2,372 mg/day (excluding salt added at the tablet, with many exceeding the upper limit of estimated safe and adequate intakes. Intakes were slightly higher in whites, those above the poverty level, and those with higher education levels. The mean sodium intake by children aged 1 to 5 was 2,036 mg/day. Estimates from NHANES II and 1985-1986 CSFII data (l-day) reported mean intakes in excess of 3,300 mg/day by males aged 12 to 49. Sodium intake is considered a public health issue due to its relationship to hypertension. Thursday data from the 1985-1986 CSFII indicated that the mean potassium intake by women aged 20 to 49 was 2,073 mg/day, with at least 25 percent of women having intakes below the lower limit of the safe and adequate range. Intakes were higher by white women, those above the poverty level, and those with higher education levels. For children aged 1 to 5, mean intakes were almost

82 NU=TION LABELING all above the lower limit of safe and adequate intakes, and some intakes exceeded the upper limit. In 1980, only 4.5 percent of the adult U.S. population obtained potassium in the form of supplements, and only a few exceeded the upper limit of the safe and adequate level. Vitamins Vitamin A and Carotenes From 1909 to 1985 there was an increase in the daily per capita amount of vitamin A available in the U.S. food supply, and since 1965 there has been an increase in carotenes. Both nutrients reached a peak in 1985, at 1,610 retinal equivalents (RE) for vitamin A and 660 RE for carotene. The increases in availability were primarily the result of the development of new varieties of deep yellow vegetaWes with higher carotene contents and the fortification of margarine and other daily products. Vegetables have accounted for three-fourths of carotenes in the U.S. food supply, particularly dark green and yellow varieties. Vitamin A is also supplied by meat, poultry, and fish, as well as daily products (LSRO, FASEB, 1989~. Individual dietary intake data from the 1985-1986 CSFII indicate that He mean intake of vitamin A by women is 832 RE, although considerable individual variation was observed. Data from HHANES suggest that poor young children may be at risk for low serum vitamin A levels (LSRO, FASEB, 1989~. A survey of dietary supplement use found that 25 percent of the U.S. adult population obtained vitamin A from supplementary sources (LSRO, FASEB, 1989). Although the availability and intakes of vitamin A are generally adequate, it is considered a potential public health issue due to the low serum levels found in certain groups (LSRO, FASEB, 1989~. Thiamin Key sources of thiamin in the food supply include grain products (42.3 percent); meat, poultry, and fish (25.7 percent); vegetables (10.9 percent); and daily products (8 percent). About 2.2 mg of thiamin per capita per day is available in the U.S food supply, which is 40 percent higher than that in the pre-World War II era, when the level was 1.6 mg per capita per day. The introduction of enrichment of flour with thiamin is primarily responsible for the increase (LSRO, FASEB, 1989). Women aged 20 to 49 in the 1985-1986 CSI;II (delay) had a mean intake level slightly above the 1980 RDA. Only 5 percent of women had intakes that were below 50 percent of the RDA. Mean thiamin intakes by children aged 1 to 5 were above the RDA among all races and were highest among black children. Data from the 1977-1978 NFCS and the 1985-1986 CSFII (l~ay) indicate that mean intake levels increased 9.3 percent for children aged 1 to 5, 18 percent for

CONTEXTUAL FACTORS AFFECTING FOOD LABELING REFORM 83 men aged 20 to 49, and 10.8 percent for women aged 20 to 49 (LSRO, FASEB, 1989). In 1980, supplements containing thiamin were ingested by 30 percent of the population, with the median intake being about five times the 1980 RDA (LSRO, FASEB, 1989~. Riboflc*vir' Riboflavin is currently available in the food supply at about 2.3 to 2.4 mg per capita per day, which is about 30 percent higher Ran that prior to World War II, when the level was 1.8 mg per capita per day. The amount available has remained unchanged since World War II, when enrichment of flour with riboflavin was implemented (LSRO, FASEB, 1989~. Primary sources of riboflavin in the food supply include dairy products (34.7 percent); grain products (24 percent); and meat, poultry, and fish (24.3 percent). The mean intake of riboflavin by women aged 20 to 49 in Me 1985-1986 CSFII (4-day) was 12.5 percent above the 1980 RDA, with only 5 percent having intakes below 50 percent of the 1980 RDA. Mean riboflavin intakes by children aged 1 to 5 were at least 60 percent above the 1980 RDA, with 95 percent of children having intakes of at least 0.9 mg/day. Companson of intakes from the 1977-1978 NFCS and 1985-1986 CSFII shows that the mean intake levels of riboflavin increased by 4.3 percent for children aged 1 to 5, 8.1 percent for men aged 20 to 49, and 8.3 percent for women aged 20 to 49 (LSRO, PASEB, 1989~. In 1980, supplements with riboflavin were taken by 30 percent of the adult population, with the median intake being about four times the 1980 RDA (LSRO, FASEB, 1989~. Niacin The daily per capita amount of preformed niacin available in the food supply was 26 mg in 1985, which has increased since the 1940s, when enrichment of flour with niacin was instituted. Major sources of niacin in the food supply include meat, poultry, and fish (46 percent) and grain products (30 percent) (LSRO, FASEB, 1989). Four-day data from the 1985-1986 CSFII reveal Rat mean intakes of preformed niacin in all age groups of women and in children aged 1 to 5 are well above the 1980 RDA. Tryptophan conversion to niacin is not included in these values but would undoubtedly contribute to even higher niacin intake levels. There does not seem to be any real difference in niacin intake by age, race, or degree of urbanization. Comparison of data on niacin intakes from 1971 to 1986 showed a slight increase over time (LSRO, FASEB, 1989~. Supplements containing niacin are used by 30 percent of the adult popula- tion, and the median intake of niacin from these products is 190 times Be 1980 RDA (LSRO, FASEB, 1989~. Vitamin B6 The daily per capita amount of vit~nin B6 available in the food supply has changed little since early in the twentieth century, but the

84 NtJTHTION LABELING food sources have changed. The contributions from meat, poultry, and fish have increased dramatically, whereas the amounts from potatoes and grains have decreased. In 1985, the daily per capita amount in the food supply was 2.1 ma; the mayor sources were meat, poultry, and fish (41.1 percent); vegetables (21.9 percent); dairy products (10.7 percent); and fruits (10.6 percent). Stakes by children aged 1 to 5 exceeded Me 1980 RDA, as desk in We 198541986 CSFII. The mean intake by women was well below the 1980 RDA (approximately half) and varied by age. Thirty percent of the adult U.S. population consumed supplements contain- ing vitamin B6, at a median level of 1.4 times the 1980 RDA (LSRO, EASES, 1989~. Due to Me low intakes by a substantial numba of individuals, vitamin B6 is considered a potential public health issue (LSRO, FASEB, 1989~. Vitamin C The daily per capita amount of vitamin C in the food supply has fluctuated since the turn of the century, but it has not changed consistency. In 1985, 115 mg of vitamin C per capita per day was available, an amount well in excess of the 1980 RDA. Major food sources of vitamin C have changed; contributions from citrus fruits have increased whereas those from potatoes and vegetables other than dark green and deep yellow types have decreased. In 198S, the food groups that contributed the major shares of vitamin C to the food supply were vegetables (47.9 percent) and fruits (42.7 percent), especially citrus fruits (27.7 percent). Four-day data from the 198~1986 CS~I revealed that mean intakes of dietary vitamin C in women aged 20 to 49 and children aged 1 to 5 were well above the 1980 RDA. A comparison of intakes of vitamin C ding the period from 1971 to 1986 showed an increase, although results are not consistent over all surveys. Greater changes may be observed in the future with the introduction of higher levels of vitamin C fortification in a variety of foods and beverages (LSRO, FASEB, 1989~. In 1980, 35 percent of the adult U.S. population consumed vitamin C in supplements, and the median amount consumed was three times the 1980 RDA. Vitamin C is considered to be a potential public health issue due to low intakes in groups with low socioeconomic status; however, recent vitamin C fortification may have affected intakes by this group (LSRO, FASEB, 1989~. Folate The daily per capita amount of folate has not changed substantially since the early l900s which was nearly 300 fig in 1985. The contributions from meat, poultry, and fish, and fruit have increased, whereas the contribution from grain products has decreased. In 1985, the major sources of folate in Be food supply were Tom vegetables (24.8 percent), legumes, nuts, and soybeans (19.5 percent), grain products (12.7 percent), meat, poultry, and fish (12.6 percent), and fruits (12.4 percent) (LSRO, FASEB, 1989~.

CONTEXTUAL FACTORS AFFECTING FOOD LABEl]NG REFORM 85 Mean dietary folate consumption from the 198~1986 CSPII (delay) was estimate to be below the 1980 RDA by over 95 percent of women aged 20 to 49 End over 50 percent of children aged 3 to 5. In contrast' 90 percent of children aged 1 to 2 had folate intakes that were above the 1980 RDA (LSRO, FASEB, 1989). The serum and red blood cell folate levels measured in NHANES ~ are difficult to reconcile win the dietary data The interpretive criteria for the blood levels are not certain, and the prevalence of low levels of serum and red blood cell folate are low (LSRO, FASEB, 1984). However, women aged 20 to 44 appear to be at greater risk of folate deficiency than are other population subgroups DISCO, FASEB, 1989). Supplemental folate was consumed by 20 percent of the adult U.S. popula- tion at a median level of two times the 1980 RDA (LSRO, FASEB, 1989~. Folate is considered to be a potential public health issue due to He lower than recommended intakes by some groups (LSRO, FASEB, 1989~. Minerals Calcium The daily per capita amount of calcium in the food supply was greater than 900 mg in 1985, indicating that, overall, the food supply contains an adequate amount for most of the population. Daily products are the primary source, providing 76.8 percent in 1985 (LSRO, FASEB, 1989~. Four-day data from the 1985-1986 CSFII indicate that the mean intakes by women aged 20 to 49 continue to be below the 1980 RDA, causing concern about osteoporosis. For children aged 1 to 5, the mean intake of calcium was 804 mg/day; the median intake was 769 mg/day, indicating that over half of the group had intakes below the 1980 RDA (LSRO, FASEB, 1989~. In a comparison of data from NHANES I and NHANES II with data from CSFII, some tentative conclusions can be drawn about trends in calcium intakes since 1971. For men aged 20 to 49, mean calcium intakes ranged from 750 to about 1,100 mg/day, with little change over time having been observed, suggesting that their calcium intakes are adequate. For women aged 20 to 49, mean calcium intakes have ranged from 530 to 690 mg/day, which is well below the 1980 RDA. For children aged 1 to 5, mean calcium intakes ranged from 750 to 920 mg/day. In general, calcium intakes seem to have remained fairly constant over the 15-year period (LSRO, FASEB, 1989~. Data on dietary supplement use indicate that 13.5 percent of Be population and 34.9 percent of supplement users consumed calcium supplements. The median level of supplemental calcium was 16 percent of the 1980 RDA, indicating that calcium supplements are consumed at relatively low doses. Major promotion and use of calcium supplements and calcium-fortified foods as a

86 NATION L4BEllNG means of preventing osteoporosis have occurred since the 1980 survey (LSRO, FASEB, 1989~. Calcium is considered to be a current public health issue due to He low calcium intakes by vulnerable groups, especially women. The influence of the recent calcium promotion may have an impact on the overall calcium status of vulnerable groups (LSRO, FASEB, 1989~. Iron The amount of iron in the food supply increased during He 1940s due to the introduction of iron enrichment of flour. Recent increases have resulted in a level of 17 mg/day in 1985. The iron supplied by various food groups in 1985 included 41 percent from grain products, 23.8 percent from meat, poultry, and fish, and 12.6 percent from vegetables MISDO, FASEB, 1989~. Estimates of mean iron intake from NHANES I and II, the 1977-1978 NPCS, and leeway 1985-1986 CSFII data reveal remarkably close values, with a range of 9.2 to 10.8 mg/day by women of childbearing age. The mean intakes are less than 60 percent of the 1980 RDA for this group. Four-day 1985-1986 CSFII data show that over 95 percent of women aged 20 to 49 and over 90 percent of infants aged 1 to 2 have iron intakes below their age-specific 1980 RDAs. Iron intakes are below the 1980 RDA for 50 percent of children aged 3 to 5. Despite the low levels of iron intake with respect to the 1980 RDA, the prevalence of iron deficiency in the population is relatively low. Data from NHANES II and HHANES indicate that the prevalence of iron deficiency ranges from 2.4 to 14 percent in women of childbearing age. There are several explanations for the discrepancies with dietary intake data. The 1980 RDA for iron may be overly generous and, therefore, not attainable by most women who eat otherwise nutritionally adequate diets. Also, iron intake is not directly related to iron status, since absorption of iron increases substantially when body iron stores are low (LSRO, FASEB, 1989~. Supplemental iron (taken alone or in a multinutrient form) was consumed by about 22 percent of the U.S. population in 1980. Supplemental iron was consumed by 56 percent of all supplement users, and the median level consumed was 1.2 times the 1980 RDA (LSRO, FASEB, 1989~. Iron is considered to be a current public heals issue, due to the extent of low intakes by vulnerable groups such as women of childbearing age (LSRO, FASEB, 1989~. zinc The per capita amount of zinc in the food supply has remained essentially unchanged at about 12 mg/day since 1909, despite fluctuations over the years (NRC, 1989~. In 1985, primary sources included meats, poultry, and fish (48.7 percent); daily products (19 percent); and grain products (12.6 percent). A large decline has occurred in the percentage of zinc obtained from grain products (LSRO, FASEB, 1989).

CONTEXTUAL FACTORS AFFECTING FOOD LABEl~NG REFORM 87 Four~ay ~ from the 198~1986 CS1;II indicate that We mean dietary intake of zinc by women aged 20 to 49 is about half the 1980 RDA, win a large percentage of intakes falling well below that level. Zinc intakes were lower for blacks, those below the poverty level, and those with lower education levels. Mean intakes by men were higher and closer to Me 1980 RDA. For children aged 1 to 5, mean intakes were close to the 1980 RDA (LSRO, FASEB, 1989~. An estimated 13.5 percent of the adult population used supplements con- taining zinc in 1980, and Be median intake by users was 50 percent of Be 1980 RDA (LSRO, FASEB, 1989~. Zinc is considered a potential public health issue due to Be low dietary intakes by some groups, particularly women (LSRO, FASEB, 1989~. Conclusions from Survey Data The foregoing discussion of the per capita availability of food and nutrients, levels of dietary intake, and prevalence of impaired nutritional status leads to several conclusions. The supply of food is abundant and the nutrient levels in the food supply are generally adequate. The principal nutrition-related health problems experienced by Americans are related to overconsumption of food en- ergy, fat, saturated fatty acids, cholesterol, and sodium. Despite the abundant food supply, some subgroups in the population may not have sufficient food for a variety of reasons. There is evidence of inadequate individual dietary intake of specific nutrients or impaired nutritional status in some subgroups of the popula- tion. Iron deficiency continues to be the most common single nutrient deficiency observed among women of childbearing age and young children. Low calcium intakes by females from childhood through early adulthood may contribute to less than optimal bone mass that can predispose them to osteoporosis later in life. The survey evidence is less conclusive for vitamin A, vitamin C, folate, zinc, and vitamin B6 than it is for iron or calcium. Dietary surveys indicate that intakes are low with respect to recommended nutrient levels in specific subgroups of the population. However, there is limited information from these surveys or other studies to suggest that health problems exist in Be general population that would justify nutrition labeling of all these micronutrients. Factors Influencing Future Dietary Changes Several major changes in Americans' life-styles have had major impacts on eating habits. Perhaps the most significant change is in the number of meals eaten away from home. Approximately one-third of all food is now eaten away from home as packed lunches (9 percent), meals at restaurants (20 percent) and limited-menu restaurants (13 percent), and meals at schools (16 percent) and worl~laces (20 percent). Another 16 percent of meals are eaten at someone

88 N=NTION LABELING else's home. In general, the nutrient densities of meals eaten away from home are somewhat less than those of typical meals eaten at home (NRC, 1989~. Americans snack more; it has been estimated that as many as 20 percent of calories now come from snacking. The number of Americans who snack in- creased from 60 percent in 1977 to 80 percent in 1985. Fewer young people eat breakfast more people are dieting, and more vitamin and mineral supplements are being consumed. In 1977, only 35 percent of Americans took food supple- ments; Hat figure grew to between 45 and 60 percent among all age groups by 1985 (LSRO, EASES, 1989; NRC, 1989~. Other factors affecting dietary habits include many of those mentioned above: income (and related factors such as employment and household size), availability of food assistance programs, education, and number of meals each at home. In addition, race, ethnicity, geographic origin, and health status also help to determine eating habits. It can be assumed that even more changes in food consumption pawns will occur in the near future, reflecting changes in the U.S. population as well as the food supply. As the demographics of U.S. society shift to an older, but more ethnically heterogeneous composition, shifts in demand for different types of foods are expected. A higher proportion of the population with higher levels of education may portend more concern with health and nutrition. As households become smaller and as an even greater proportion of families are headed by single parents or two working parents, there may be an even greater demand for easy-to-prepare meals. Biotechnology has enormous potential to improve the composition of the U.S. food supply. Considering the aggregate effects of these and other unforeseen changes, it is anticipated that Americans' diets will continue to experience dynamic changes. FOOD MARKETING IN THE UNWED STATES The introductory language of the U.S. Department of Commerce's 1990 U.S. Industrial Outlook report illustrates the changing nature of today's food industry. The section on food, beverages, and tobacco, 'Mod Chain in Transition," opens with the following paragraph: Growing numbers of consumers are demanding foods with convenience, qualify, variety and healthful attributes. Today, 70 percent of U.S. households own mi- crowave ovens; some industry researchers estimate 15 minutes is the maximum most Americans are willing to spend preparing an ordinary meal. Demographics are changing. Only 28 percent of American families now have one spouse at home and one at work. The remainder have either two wage earners or a single parent. Dual-income families often purchase more convenience type products. TraveL eth- nic restaurants, and television have encouraged many consumers to experiment with new varieties of food. Hence, many retailers now offer numerous types of

CONTEX7VAL FACTORS AFFECTING FOOD LABELING REFORM 89 exotic foods. Concerned with health, many consigners want high fiber prodlucm genuinely tree of fat and cholesterol (DOC, 1990, p. 34-1~. It is clear that the U.S. food industry is faced with a rapidly changing marketplace. As the industry has searched for profitable marketing niches in this changing environment, the number of new products introduced annually has exploded. Until 1981, an average of 2,500 grocery products were introduced each year. By 1990, the number has grown steadily to an annual average of 12,000 items Friedman, 1990~. Many of these products are targeted directly at a more health-conscious public and, as a result, feature health and nutrition claims on the label. In addition, the decade of the 1980s produced a variety of structural changes that complicate labeling issues. This section explores the relevant structural issues pertaining to major sectors of the food processing and distribution systems, including the growing integration of the food system into the world community of trade and issues that have an impact on marketing, promotion, and labeling decisions. World Trade No consideration of food labeling issues should ignore the growing integra- lion of the U.S. food distribution system with trade throughout the world. The United States is the world's largest exporter of agricultural products, generating an agncultusl trade surplus every year since 1959. However, U.S. competitive- ness declined during the 1980s as the European Community shifted from being the world's largest importer of grains to one of the largest exporters. In 1988, the United States exported $35 billion of agricultural products, while agricultural product imports hit a record $21 billion. Imports can be classified into two general categories: (1) noncompetitive imports of products that cannot be produced at all, or at least not profitably, in the United States; and (2) competitive imports that compete directly with U.S. products. During the 1980s, competitive imports doubled, with meats becoming the largest single U.S. food product import, surpassing coffee in 1988 for the first time in 16 years (USDA, 1989b). While most agricultural trade is still in commodities rather than processed food products, the share of processed products is growing in the import market. It is also clear that more and more imports compete directly with products that are also produced in the United States. The volume of international trade is an important issue not only for U.S. consumers but for the badance of U.S. trade as well. Labeling changes must be sensitive to their international trade implications. Food Processing The food processing sector is huge and incredibly diverse. The industry is divided into 48 separate SIC (standard industry classification) codes. During

9o Nl~RITION LABEllNG 1989, food product shipments totaled over $345 billion (DOC, 1990~. A full survey of this sector is beyond the scope of this report. However, a few general observations that bear on label revisions can be made. First, the growing internationalization of the food industry is evident in Me ownership of food processing companies. From a balance of U.S. investment abroad of $8.2 billion versus foreign investment in the United States of $8.3 billion in 1984, there has been a shift over the past 5 years toward more foreign ownership in the United States. In 1988, U.S. investment in food product industries overseas totaled $13 billion versus $16A billion of direct foreign investment in the United States. The Netherlands accounted for the largest share of direct foreign investment, followed by the United Kingdom. The largest non- European Community investor was Canada, followed by Japan (DOC, 19903. Second, the pace of leveraged buyouts and takeover activities during the 1980s has had a direct impact on the food processing industry. For a variety of reasons, We food processing industry was a major target of hostile takeovers during the decade. Increased corporate debt to fund takeovers or to defend against a hostile raid diverted cash flows to cover interest payments. As a result, food processors have much less capital investment flexibility for product research and development in the l990s than they did in the 1980s. Food processing or labeling changes will take longer to implement during the l990s. Third, as consumers have become more interested in health and nutrition, more funding has been allocated toward the promotion of health-oriented products. It is estimated, for example, that one-third of the $3.6 billion spent on food advertising in 1988 contained some type of health or nutrition message (DOC, 1990~. This trend is a vivid demonstration of the fact that, in the l990s, nutrition "sells . " Food Retailing U.S. shoppers spent $410 billion on food in 1988, $155 billion of it on food consumed away from home and 5255 billion of it on food consumed at home (USDA, 1989b). This amount represents 11.8 percent of disposable personal income after taxes in 1988 versus 15.3 percent in 1965. The growth of sales in limited-menu (fast-food) restaurants represents the most significant shift in recent times in expenditures for food eaten away from home. ~ 1987, one-third of all food eaten away from home was purchased in limited-menu restaurants versus only 10 percent in 1963. Restaurants, lunchrooms, and cafeterias represented almost half of this market in 1963 but declined to 40 percent in 1987 (USDA, 1989b). Since foods sold in restaurants and noncommercial settings are an increasing part of American diets, this sector is very important in any program aimed at dietary improvement. There was a significant shift in sales at supermarkets for food consumed at home in the 1980s. Supermarket sales grew from about 45 percent of

CONTEXTUAL FA=ORS AFFECTING FOOD LABEllNG REFORM 91 the market for food consumed at home in 1963 to just over 73 percent in 1989. Sales at limited-line grocery stores declined from 30 percent in 1963 to just under 19 percent in 1989. Over this same period convenience store sales grew from almost zero to just under 8 percent (Sansolo, 1990~. Not all large supermarkets are operated by large businesses. There are approximately 147,000 grocery stores in the United States. Chains operate 17,300 units, representing 50.1 percent of total food sales; independent businesspeopleoperale 72,700 units, representing 42.2 percent of total food sales; and convenience stores number 57,000, representing 7.7 percent of total food sales. In-store information programs would, therefore, have an impact on a large number of small businesses. In addition, the major costs associated with food products incurred after they leave the farm differ between food consumed at home and that consumed away from home. The differences are greatest for processing (31 percent compared with 16 percent) and retailing (23 percent compared with 60 percent) (USDA, 1989a). Therefore, the issue of cost is a consideration in food labeling reform. Since processing and retailing costs already make up the majority of the retail cost of food, increases in costs related to food label changes for either or both of these sectors should be expected to have an impact on the prices consumers pay for foods. Reforms should be evaluated to ensure that consumers perceive the benefits of increased label information as a trade-off for any price increases that they may experience for individual foods. A more complete discussion of the costs of food labeling reform appears in Chapter 7. General Marketing Considerations The limited scope of this section permits only a cursory examination of food marketing and labeling considerations. As was made clear during the Committee's marketing workshop, because nutrition concerns help to sell products to today's consumers, the health and nutrition groups within food manufacturer organizations are becoming an increasingly integral part of product development and marketing teams. Prior to this shift, almost all decisions in this area were subordinated to the brand management team. As a result, health, nutrition, and labeling considerations are becoming an important part of the marketing strategy for a growing number of food products. This is an encouraging development in one very important respect. The marketing motivation provides a powerful incentive to develop new nutritious and healthful products. Therefore, even though this motivation may lead to aggressive, and at times even overzealous, claims on the label and in advertising, the incentive to satisfy consumer demand for more healthful products is an important asset if it is channeled in the proper direction. It was also clear during the Committee's marketing workshop that consumer feedback is a powerful motivating tool for including nutrition labeling and cre-

92 NUTRITION LABEl~NG ating more healthful products. Feedback takes many forms, including consumer purchase data, focus group responses, taste panels, complaints, and requests for additional product information through the mail or the use of toll-free telephone numbers. In short' consumer choice forms the basis for a powerful feedback mechanism that helps food product manufacturers to know whether they are in tune with today's consumers and, through trial and error, shapes the nature of product development. The cornerstone of consumer feedback is informed con- sumer choice. Improved label information will not only help consumers make more appropriate immediate decisions, but it also improves the direction and quality of future product development decisions. The views within the industry as to how much importance to place on health and nutrition issues differ, depending on the nature of the product. There are many products that the industry views as indulgence or reward foods for which they feel the consumer has little or no expectation that they will make a nutritional contribution to their diets. There are other products that the industry feels make such a minimal nutrition contribution that the consumer has little interest in detailed information (spices or condiments, for examples. The views of different elements of the industry also differ with regard to where the information is placed on the label. The principal display panel is viewed as a commercial vehicle, with its primary purposes being sales promotion and competition with rivals. Manufacturers guard this product "real estate" with a fervor not applied to other areas on the package. Design, color, visual impact, balance of presentation, and overall image of the product are the driving forces for the principal display panel. Detail and elaboration are generally left to the back or side panels. This possessive attitude helps to explain why label regulations Hat have an impact on the principal display panel are much more likely to meet with strenuous opposition than are any other kinds of labeling proposals. The product design, development, and marketing process is complicated, and thus, it is useful to place these aspects in perspective. The dominant view within the industry is that all positive attributes of a product, no matter how carefully they are developed, are lost if the product will not sell. For the product to sell, it must taste good and be attractively packaged, priced right, and convenient for the consumer to use in the way it was intended. The industry will generally support label changes that create a more informed consumer choice process. It will, however, strongly resist changes viewed as being so costly as to disturb the consumer perception of product value, so directive in tone that they erode the ability of the consumer to think in terms of a total dietary context (i.e., creating a good food/bad food image for single products), or run the risk of disclosing proprietary formulations.

CONTEXTUAL FACTORS AFFECTING FOOD LABELING REFO~ CONSUMER UNDERSTANDING OF NUTRITION AND USE OF FOOD LABELS 93 The U.S. food supply is diverse and abundant. Consequently, consumers can choose from over 20,000 items in most supermarkets. As Timmer and Nesheim point out, '-To the consumer who has the motivation, knowledge, and Uncial means, the American food system offers a diet as healthy, safe, and appetizing as any in history" (1979, p. 155~. Factors Influencing Food Consumption Behavior Despite comprehensive study, no single factor or set of factors has yet been found to adequately describe how dietary patterns are developed, maintained, and changed QIochbaum, 1981~. Factors that influence consumers' food choices are many and varied, including internal as well as external factors (Sims, 1981~. External influences include social, cultural, and economic factors. Dlamatic examples of the influence of culture, geography, and food availability on eating habits and chronic disease prevalence can be cited. Internal factors include individual physiological and psychological factors, as well as acquired preferences and knowledge about foods. In addition, interpersonal or social factors are also important; family and group situations dominate many food acquisition, preparation, and consumption situations (Glanz and Mullis, 1988~. Assuming that foods are available and can be purchased at a reasonable cost, familiarity is one factor that perhaps exercises influence over all others. Sociocultural forces dominate the associations made with foods that have been familiar and consumed since childhood. Psychological factors also influence food choices; many individuals go out of their way to find comfort foods (such as chocolate or ice cream) when they are bored or lonely. Motivations, particularly those related to health, play a strong role in influencing food choice decisions. Only the most foolhardy do not try to comply with advice from a physician or other health professional that a change in diet must be undertaken in order to ameliorate symptoms or cure a medical condition. Others have been persuaded that if they make certain dietary changes now, these changes will help to promote health and prevent disease in the future. Education and income are moderating factors that influence how amenable certain individuals are to making dietary changes. Those with more education are more knowledgeable about nutrition, and many individuals make food choices consistent with this knowledge. In addition, those with higher incomes can be expected to have more flexibility when it comes to choosing from among the foods available for purchase. The latest survey of consumer attitudes and shopping behavior conducted by the Food Marketing Institute (FMI, 1990) asked respondents to rate the importance of various factors in food selection. Over 70 percent of the consumers

94 NUTRITION LABEl~NG interviewed identified taste, nutrition, and product safety (in that order) as '~very important" factors when they purchase food. Those on medically restricted diets were significantly more likely to rate nutrition considerations as highly important, as were women and those over age 50. Nutrition is, therefore, perceived as being extremely important by consumers, but only as one part of a complex decisionmaking process, when purchasing food (FMI, 1990~. Price was deemed very important by two-thirds of the consumers in the 1990 FMI survey. FMI believes that economics has remained important to consumers over the past 10 years but that the dominant themes have now shifted to nutrition, product safety, and convenience. Consumers' Knowledge of Nutrition - Knowledge of food and nutrition is undoubtedly one factor that influences food choices and dietary behavior. Fanelli and Abernathy (1986) reported Hat for older adults, the reading of food labels was significantly related to their level of nutrition knowledge. It had long been thought that if nutrition knowledge were increased, improvements in food choices and eating habits would follow. Research in this area has shown this to be a misleading, if not false, assumption (Hochbaum, 1981~. Surveys, however, continue to assess the nutrition knowledge of target groups on the assumption that at least minimal understanding of basic nutrition must be present for an individual to make decisions Hat result in nutritionally sound food choices. Numerous surveys have confirmed that Americans are increasingly aware of and display interest in nutrition (Sloan, 1987~. Five of the six concerns listed by more than 10 percent of the sample in the FMI survey were dietary components that both the Surgeon General's (1988) and NBC (1989) reports recommended should be consumed at low levels or reduced in the diet. The sixth concern was vitamins and minerals, which moved from number 2 in rank in 1983 to number 6 in 1990 (FMI, 1990~. Despite such promising claims about increased consumer nutrition knowl- edge, it cannot be assumed that increased awareness leads to enhanced nutrition knowledge, which in turn leads to improved dietary behavior. The results of the 1988 FDA Health and Diet Survey are an example of this point (Levy and Stephenson, 1990~. Despite the public's increased awareness about dietary fats and cholesterol as risk factors for heart disease in recent years, the results of the 1988 Survey showed no increase in food and nutrition knowledge related to dietary fats and cholesterol between 1983 and 1988. Only 3 of 11 questions asked in the 1988 Survey were answered correctly by more than 50 percent of the respondents. However, consumer concerns about food components do seem to translate directly into their reported food choices. A number of FDA surveys have shown

CONTEXTUAL FACTORS AFFECTING FOOD LABEl~NG REFORM 95 that Americans are increasingly aware of health risks from sodium, fats, and cholesterol and that they report eating less salt, red meat, butter, whole milk, and eggs as a result (Heimbach, 1985, 1986). Putnam and Weimer (1981) reported that roughly two-thirds of surveyed households claimed to have made at least one change in food consumption, almost always to avoid a negative nutrient. Of the 10 most frequently mentioned reasons for changes in eating habits, nine were related to reducing the intake of negative food components. Although awareness of He relationship between diet and risk of disease seems to be at an all-time high, actual dietary behavior has not shifted dra- matically. Despite consumer reports of awareness, nutrition is only one factor, and usually not the most important one, that influences food choice. Lee 1988 Prevention Index Survey, which is conducted annually by Harris ~ Associates, concluded: '`Almost no change has taken place in the last five years in He structure of the American diet as it relates to preventing illness" (CNI, 1988, p. 1~. Likewise, after examining dietary changes of women between 1977 and 1985, the CSFII conducted by USDA reported Hat although consumption of red meat had declined substantially, the overall fat intake by these groups had not declined because women had correspondingly increased their intake of salad dressings, table spreads, and rich desserts (Harris and Welsh, 1989~. Sources of Nutrition Information As consumers have developed a more intense interest in diet and disease, they clamor for more information in an easier-to-understand form. Nutrition advice seems to dominate the airwaves and the printed page. Not only has the amount of information about diet and disease proliferated, but the sources and settings in which such information is offered have increased as well. Nutrition education programs are found in schools, community groups, fitness centers, and the workplace, as well as in various community sites such as supermarkets, restaurants, shopping malls, museums, and libraries (ADA, 19903. Surveys have confirmed that doctors are still regarded by consumers as the best source of nutrition information, although media sources are assuming increasing importance (Rahn, 1980; Woolcott, 1983~. Consumers who responded to a recent Gallup survey overwhelmingly cited He media as the source Hey used to get information about food and nutrition: television, magazines, and newspapers were named as the chief sources of nutrition information by 68 percent of the respondents ~C/ADA, 1990~. Consumers who ~cipated in focus-group discussions conducted by the Institute of Food Technologists ~-' also cited mass media, including print (newspaper and magazine articles, books), television, and radio, as the major sources of information about foods and nutrition, although a number of the p~cipants expressed concern about the reliability of the information they received (Snider et al., 1990~. A recent survey on the reading habits of Americans (Robinson, 1990) has

96 Nl~RITlON L4BEl~NG shown that, since 1965, the amount of time people spend reading has fallen by more than 30 percent, pnmanly as a result of a decline in newspaper reading. Today, television is the dominant mode of receiving infonnation for most consumers. The survey of reading habits confirmed that people who are older and well educated spend the most time reading. These findings have particular relevance for a study of nutrition labels. Venkatesan and fellow researchers (1977, 1986) concluded that the medium used for the presentation of nutrition information (television, advertisements, or labels) influenced the formats that consumers perceived to be the most useful. Television viewers appeared to find a variety of presentation formats equally useful, whereas readers of print media preferred detailed information as opposed to graphic or summary presentations. Tl~ o ;~ A_ ·~ ~ .: ~ ~ ~ · ~ . . _ ¢,¢ ~ Cup "~ usual w yloVlut; ~w~nuan 1mormatlon to a communication-satu- ated, time-starved public, the message must be simple, short, and practical (Shepherd, 1990~. The issue then becomes how to best present the nutrition information on food labels to measure up against these criteria Impact of Nutrition Information on Food Purchase Decisions Format of Information Given the level of interest in nutrition expressed by consumers but the high level of confusion over the interpretation of some of the currently available information, it is useful to examine whether point-of-purchase information makes a difference in food choices when presented in a format relevant to current dietary recommendations. Although the number of research reports is limited, several studies have demonstrated that consumers can be motivated to buy foods more consistent with current dietary recommendations by providing point-of- purchase information (Glascoff et al., 1986; Hixson et al., 1988; Light et al., 1989; Mullis et al., 1987~. In a survey conducted for the National Food Processors Association ~:PA) (ORC, 1990), consumers were asked whether ingredient and nutrition informa- tion influenced their actual purchase decisions. Pifty-one percent stated that it influenced their purchases "a great deal," while another 32 percent said their purchases were influenced"somewhat." By category, nutrition information had the most influence on the purchase of cereal (35 percent), canned products (26 percent), and prepared foods (13 percent). Nutrition information was reported to have the least influence on the purchase of snack foods (15 percent), produce (12 percent), and daily products (9 percent) (ORC, 1990~. One study that supported the contention that supermarket operators are willing to undertake informational programs at their own expense because they believe that consumers find them useful was a cooperative effort between Giant Fbod, Inc., of Washington, D.C., and EDA (Levy et al., 1985). This program,

CONTEX~IAL FACTORS AFFECTING FOOD LABEl]NG REFORM 97 called Special Diet Alert (SDA), consisted of shelf tags next to the product price information that identified products that were low in or had reduced levels of sodium, calories, fat, or cholesterol, all of which were accompanied by special media campaigns and other printed material. Consumer purchases of SDA and non-SDA products were tracked through computer-assisted checkout (scanners data in two market areas. Results over a 2-year period showed that sales of SDA-identified products increased, on the average, 4 to 8 percent more in the Washington test market than they did in the Baltimore (control) market. Undoubtedly, some segments of the population (e.g., those with medical conditions) are very interested in this lope of information, whereas others are not. Glanz and Mullis (1988) reviewed more than 20 reports of nutrition infor- mation programs in supermarkets, restaurants, and cafeterias and on vending machines, and concluded that point-of-purchase information programs in super- markets were more successful in improving nutrition knowledge and attitudes than they were in changing consumer purchasing behaviors. In studies that doc- umented significant changes toward the purchase of more nutritious foods, an emphasis on brand name choices appears to have been influential. Glanz et al. (1989) believe that the conclusions of the review neither support nor discredit the value of programs for improving diets; rather, they believe that the findings reflect weaknesses in the design of some interventions as well as identify short- comings in research designs and measures. Thus, while the idea of providing nutrition information at the point of purchase in grocery stores and restaurants is appealing and well liked by consumers, research to date has not conclusively demonstrated that such programs are directly responsible for changing consumer behavior in making more healthful food choices. Information Processing anal Behavioral Change Consumers learn based on the way in which they process the information they receive. McGuire (1969) pointed out that whether acquisition of a certain piece of information leads to some desired behavior depends, in part, on whether the information has been appropriately processed by the receiver. Thus, an educational program may deliver a nutrition message to individuals or groups, but whether it produces the desired effect depends on how the recipient of that information processes or internalizes it to make decisions or guide behavior. A change in food choices and purchases may be the ultimate effect of information programs such as food labeling, but there must be prior cognitive changes His and Staelin, 1982~. Thus, in order to judge the effectiveness of nutrition labels on food products, more reliance must be placed on an analysis of how such information is processed by individuals rather than on surveys based on consumer reports of how frequently they read the food label or how

98 NUTRlT70N LABELING well they understand the label content (Olson and Sims, 19803 Figure 4-1~. An information processing perspective allows an examination of the conditions under which information and persuasive messages may lead to behavioral changes. Social scientists believe that individuals pass through a series of steps in acquiring, processing, and using information that eventually leads to the adoption of new behavior (McGuire, 1976), a conceptualization of sequential impact that has been referred to as a "hierarchy of effects" (Russo et al., 1986~. For the purposes of this discussion, the framework suggested by Mazis and Staelin (1982, p. 3) is used: Exposure: data come into contact with one or more of the consumer's five senses. Attention: the consumer selects certain stimuli out of the environment for further processing. Comprehension: the consumer understands and assigns meaning to the message conveyed. Retention and Retrieval: information is stored in memory for later use when a decision is maple. Decisionmaking: the consumer sorts out and synthesizes information stored in memory or available at the point of sale. This framework is useful in discussing how nutrition information on food labels can ultimately affect consumer choices. Each factor in the framework and the effects it may exert in terms of consumer use of nutrition labels are discussed below. Exposure Consumers must come into contact with the information on the label before it can have any effect on their food choice decisions. Mazis and Staelin (1982) identified several roadblocks that can prevent consumers from ever being exposed to nutrition information: destruction or removal of the information, information unavailability, inappropriate timing of the message, or targeting problems. Therefore, nutrition information on the food label must be appropriate to consumer needs. For those who have been advised to follow a medically prescribed diet or who have hypersensitivities to certain food components, specific information identifying particular food components and ingredients must be placed on the label. Likewise, for those who wish to compare the nutrients in foods among various product categories, the information must be presented in a readily identifiable format. The issue of coverage of food products with nutrition labeling also relates to the notion of exposure. If only about half of packaged foods and far fewer fresh products carry nutrition information in the grocery store, and few restaurants or other sites offer nutrition information, nutrition labeling on only a limited number of products cannot be expected to have much of an impact on food choices.

99 .8 d.o l~ ~ ~W on ·: ~ 'e c) ~ A] a; I, .o ~ - ·c., ~ t 1 1 U) ·s o ~ ~ Ct - o = C) A - · . ~3 Cat 4 .= ~ .s oo ~ 3 2 8 . o E ._ by' ~ so 4) ~ o Cod 5 o ·° ~ a so oo ~ .S on ~ .5 ~ at In of o o on on A: . - ~ .m o ~ U' .5 , ~3 £ at: - ~ ..

100 NI~RlTION LABELING Frequency and Prevalence of label Reading When queried, con- sumers indicate that nutrition knowledge and the availability of nutrition in- formation on the label are important Several studies have confirmed that most consumers are aware of and want nutrition information (Heimbach and Stokes, 1979), and some studies suggest Mat consumers would be willing to pay extra for it (Daly, 1976; Lenahan et al., 19721. The survey recently conducted for NFPA, Food Labeling and Nutrition: What Arnen cans Wan`, revealed that, Most shoppers read food labels. About eight in ten consumers report that they usually read product labels for general information, nutrition information or He list of ingredients the first time they purchase a specific product. Four in ten shoppers always read this information on their first purchase. On subsequent purchases, only one in five always reads labels, while about six in ten consumers say dolt they read the labels at least sometimes for this information. Women, older shoppers, and those with more formal education are among the most frequent label readers, as are consumers who are on restricted diets (ORC, 1990, p. iii). Approximately one-third of the consumers in the FMI survey responded that they read labels for nutrition information "pretty much every time" when they buy packaged foods, with another 45 percent reading them "sometimes." The proportion of consumers who read labels for ingredient information was almost identical to the proportion who read them for nutrition information. Those on medically restricted diets were twice as likely to read labels for ingredients and nutrition information (FMI, 1990~. Consumers are significantly more likely to read labels the first time they purchase a product than they are on subsequent purchases. Ingredient information is read by 53 percent of first-time consumers, and nutrition information is read by 49 percent of consumers the first time they purchase a product, in contrast to 36 percent for ingredients and 36 percent for nutrition information by repeat purchasers. Only about 5 percent of consumers say they never read labels for nutrition or ingredient information, which is true for first-time as well as repeat purchasers of a product (FMI, 1990~. One-half of those consumers who responded to the EMI survey said they felt that the current labels provided "no more than some" of the desired information. Those on medically restricted diets were the most likely to judge nutrition labels as being inadequate. The desired changes in the label by those who need additional nutrition information were probed in the 1989 FMI survey. These changes included the need to make the label easier to understand (25 percent) and for information about calories (24 percent), saltisodium (21 percent), fat/saturated fat (18 percent), and chemicals, colorings, and dyes (10 percent) (FMI, 1989~. Attention Information is of little use unless consumers consciously decide to pay attention to it. Because consumers are exposed to so much information

CONTEX~IAL FA=ORS AFFECTING FOOD LABEl]NG REFORA! 101 simultaneously, they have neither the time nor the inclination to pay attention to all of it. It is at the attention stage that consumers sort out and actively examine stimuli (e.g., a warning or a nutrient claim) for the purpose of furler processing the information, while they ignore or filter out other messages. Attention is affected by internal as well as external factors. Consumers pay attention to the information that they feel is consistent and pertinent to their own personal needs, values, and goals (internal factors) and tend to be attracted to messages that are distinctive, have sufficient intensity, and are not subject to distraction from competing messages (external factors). Shepherd et al. (1989) examined the content and graphic design features of nutrition labels (i.e., external factors) that consumers considered to be the most appealing and useful; the features that consistently drew the most favorable reactions were bright, '~food- like" colors and organizational cues (such as indexing and clustering of related information). Motivation, an important internal factor, also plays an important role in de- termining how receptive the target audience is to new information. For example, consumers who are attracted by a new claim on the principal display panel of the food label may be curious enough to check out the actual information on the back of the label. This curiosity may not actually lead to the purchase of that food item, a behavioral indicator of label reading. However, consumers who have been advised by a physician to limit the amount of a certain food compo- nent (e.g., calories, sodium, fat, sugar, or cholesterol) may closely inspect the nutrition information that is provided on a label in order to choose the most desirable product. For example, at least one high-fiber breakfast cereal benefited from unprecedented shifts in market share because consumers were influenced by an advertising campaign that indicated the possible benefits of consuming a high-fiber, low-fat diet for preventing some types of cancer (Levy and Stokes, 1987~. Consumers pay attention to information that is presented in a style that they prefer. Although some studies have suggested that consumers prefer a graphic presentation of numeric information, other research points to consumer preferences for absolute numbers and percentages listed on the label (Geiger et al., 199~, Sims and Shepherd, 1985~. The study by Geiger et al. attributes these results to a manifestation of the respondents' preference for more information. This conclusion confirmed the results of a study by Scammon (1978) in which consumers felt more satisfied and less confused with their purchase decisions when more information was presented. Consumer preference for more detailed information on food Labels was sum stant~ated by the findings from the recent NFPA survey which asked consumers to rank six pairs of label formats according to their usefulness in making food purchase decisions (ORC, 1990~. Consumers reported that they wanted more information to be included on product labels, particularly information on key nutrients and daily dietary recommendations. They also expressed an interest

102 NUTRITION LABELING in acquiring more detailed product information, such as the breakdown of fat and carbohydrate contents compared with current dietary recommendations, and they wanted complete nutrition information, including data on micronutrients. Consumers preferred a label that provided nutrition information on key ingredi- ents and nutrients (even if they were not present in that product) to a label that provided nutrition information only for those nutrients contained in He prod- uct. Consumers preferred having quantities presented in grams and milligrams over grams alone, and they also wanted quantities expressed numerically, to the point of specificity that included decimals. Consumers expressed some interest in having sugars grouped on the ingredient label for better identification. Although presentation in a numeric format may require more information processing than graphic formats would, Bettman (1979) concluded that as long as consumers have control over the information input rate and can process the data at a comfortable pace (as when they are reading a nutrition label), presentation of larger amounts of data is unlikely to result in confusion during the decisionmaking process. After testing printed label format alternatives, Hammonds (1978) found that consumers preferred numeric data, which allowed them to make choices for themselves, over graphic representations, which consumers perceived as "leading the label reader toward a particular choice." Likewise, Muller (1985) concluded that consumers preferred numeric ratings for individual nutrients in order to determine for themselves which brands were more nutritious. The work of McCullough and Best (1980) confirmed that the appropriate information load may be determined individually. They used the marketing research technique of conjoint analysis to measure consumer perceptions of the usefulness of various nutrition label configurations, which were portrayed as five different information Me levels and three different information load levels. The results identified three groups of label readers: (1) predominantly blue- collar workers, who preferred less information and retaining the current nutrition labeling format; (2) white-collar workers, who strongly preferred increased amounts of information; and (3) another group of white-collar workers, who preferred more information in more complex forms. The investigators concluded that it is impossible to develop a single suitable label format, because changes made in label information for one group appeared to be disadvantageous for another group. Comprehension Studies that have examined consumer reactions to and use of nutrition information on food labels reveal a starding dichotomy. A1- though consumer interest in nutrition information appears to be high at present, a less positive conclusion can be drawn about the actual level of consumer comprehension of nutrition information. Some surveys report that consumers say they understand food labels, but other studies suggest that consumers often do not actually comprehend the data on current nutrition labels and, therefore, do

CONTEXTUAL FACTORS AFFECTING FOOD LABEUNG REFORM 103 not make optimal use of this information in their food purchase decisionmaking (DHEW/USDA/FIC, 1979; Heimbach, 1982; Jacoby et al., 1977~. Conflicting results appear from studies that have been conducted to assess consumer comprehension of food label information. The survey conducted for NFPA presented a rather optimistic view: The vast majority of shoppers understand food label Information. Reported lack of attention given to labels is more a result of low interest than confusion or lack of understanding. Consistent with previous research, only about one in ten shoppers claims to have a serious lack of understanding about ingredient and nutrition information. Use of chemical and technical terns causes He most confusion (ORC, 1990, p. iv). A less positive conclusion about He level of consumer comprehension of food label information emerges from results of other studies Hat have exam- ined this issue. In the 1979 Food ~ libeling Background Papers, the following statement on consumer acceptability of food label information appears: the Consumer Food Labeling Survey, 23 percent of the 64 percent of those consumers who use the nutrition label responded that at least some aspects of He label are confusing. Terminology was stated to be confusing by 79 percent of the people, the use of the metric system by 27 percent, "big words" by 16 percent, and U.S. RDAs and the use of percentages by 15 percent; the fact that they do not know what to do with the information was stated by 8 percent. In the food labeling hearings and written comments, 189 commenters stated that nutrition labels are confusing, with the single most often cited source of confusion being U.S. RDAs (90~. Sixty comments did not specify the confusing aspects of He label (DHEW/USDA/FI~C, 1979, p. 41~. Daly (1976) found Hat even though consumers expressed positive attitudes about the importance of nutrition labeling, comprehension of He terms used on the label was low. After summarizing the data from six studies, Jacoby et al. (1977) suggested that although a high percentage of consumers indicated that they were aware of and reported that they used the nutrition information provided on food labels, only a small percentage were able to define nutrition terms and accurately assess their total dietary intakes. Those investigators concluded that many consumers simply do not comprehend nutrition information as currently provided on package labels qacoby et al., 1981~. The following caveat must apply in interpreting He findings from surveys that report that a large number of respondents say they understand food label information. Undoubtedly, in some cases consumers respond with the answer they believe He questioner wants to hear. When asked a direct question about their level of understanding, a consumer who does not wish to appear to be misinformed or unintelligent will usually respond affirmatively. Or, Hey may say that they "think" that they understand the information, until they are probed on specifics or asked to apply the information to make food purchase decisions.

104 NUTRITION libeling During this comprehension stage, consumers begin to actively transform and assign meaning to the information they read. In order for information to pass successfully through this stage, it must be understandable, and it must be presented in a context that allows it to be encoded into memory. Consumers do not understand many of the terms now used on food labels, for example, scientific terms for nutrients or food components or the metric units used to indicate nutrient compositions (Achterberg, 19903. In addition, the concept of serving size has no consistent meaning, either for food manufacturers or consumers. These problems do not allow for adequate consumer comprehension of food label information. Levy and Stephenson (19903, after studying consumers' nutrition knowledge about dietary fats and cholesterol, confirmed that respondents had a number of common misconceptions, particularly about sources of cholesterol and the caloric value of saturated versus other fats. More Can half of the respondents in that study did not know the meaning of hydrogenated, and most had not heard of monounsaturated fatty acids or oils. The only subgroup of respondents who had improved their scores over the 5-year evaluation period were those who had been told they had high blood cholesterol or others who were consciously Dying to reduce their blood cholesterol through dietary means. Examination of consumer comprehension of the key nutrition concepts, terms, and numeric expressions used on food labels may help to explain the major causes for poor understanding of the information presented. One survey found that only 36 percent of the sample knew that ingredients are listed in decreasing order by amount Cocos, 1988~. Only 6 percent knew the relationship of grams to ounces. Only 44 percent could correctly calculate the reduction in milligrams when the amount of salt was reduced by one-third. A sample of Pennsylvania adults expressed the most confusion about fats and cholesterol in two areas: the difference between saturated and unsaturated fatty acids, and the difference between fat and cholesterol (Achterberg et al., 1990~. Related to the misconception about saturated versus unsaturated fatty acids was the fact that respondents did not know what the fat was samurai with. Those consumers also had difficulty differentiating between the concepts of fat and cholesterol (probably because the terms are so often used together) and between the notion of dietary cholesterol and serum cholesterol. The process that transforms data into usable information is referred to as encoding. The consumer may transform information as it is presented into what is actually intended. The use of inconsistent terminology may result in confusion. Consumers need to be able to reorganize the information in their own minds so that it becomes meaningful. When ideas must be communicated to consumers and no frame of reference exists (e.g., the term high-fat), We label could show relevant range data or reorganize information into groupings that help to facilitate memory processes. A continuing issue concerns the importance of the format in which nutrition

CONTEXTUAL FACTORS AFFECTING FOOD LABELING REFORM 105 information is presented. Russo et al. have conceptualized the "costs" faced by consumers who want to base their brand choice on nutrition information. They concluded that "currently [consumers] must collect the nutrition information from many different product labels, comprehend it, and then determine how to aggregate the different nutrients to identify the most nutritious brand" (Russo et al., 1986, p. 49~. At least three types of "effort costs" are involved: collection, comprehension, and computation of information. It appears Rat consumers conduct an informal cost-benefit analysis and use information only when they perceive that benefits exceed the costs. For those consumers who actually do such an analysis, it is likely to be done subconsciously. Even if the information per se is free (as it is for nutrition information on food labels), consumer efforts to process the information is still a cost to them. Therefore, an important tactic for making information provision programs more successful is to decrease the effort required by consumers to process the information. Russo et al. (1986) view effort reduction as changing the information environment to adapt to people and suggest that "it may be more effective to change the shopping environment [i.e., the food label] to adapt to people than to change people to adapt to an effortful environment" (p. 68~. The initial research conducted in the early 1970s to determine nutrition label format examined only numeric, percentage, pie chard and verbal presentations (Asam and Bucklin, 1973; Babcock and Murphy, 1973; Lenahan et al., 1972~. The study by Lenahan et al. (1972) on consumer reactions to nutrition compared nutrient content in terms of (1) units as a percentage of the RDA, in which a unit represented 10 percent of the RDA, (2) adjectives (e.g., excellent, very good, and fair), and (3) percentage of the RDA. Consumers reported preferring the percentage of RDA format over the other two options. Babcock and Murphy (1973) tested a food equivalent labeling system against the proposed RDA format. This system used a pie chart that graphically related a food's nutritional value to that of a reference food based on a nutrient- t~calorie ratio. Testing of the RDA label in comparison with the food equivalent pie chart label showed that the RDA label increased sales by 55 percent, while sales increased 63 percent with the food equivalent labels. They concluded that the use of a reference food utilizing a pie chart was more effective than the FDA-proposed label format in conveying nutrition information to consumers. In 1979, FDA reported on the written comments it had received on the type of nutrition label that consumers felt would be most useful: 84 [of the 494 received] supported the present system. The most popular was a pictorial or graphic display, with 151 comments. Other proposed alternatives included using words to describe the '~quality" of nutrition (64), nutrition scores (44), nutrient density (43), and a food group system (24) (DHEW/USDA/FTC, 1979, p. 41~. Because formats with such alternatives as graphic presentations were not

106 NOTATION LABELbNG included in the original consumer testing, it is important to examine those studies that have modified the label format in an effort to aid consumers in processing the nutrition information the label presents. While the concept of a nutrient density label has been recommended (Hall, 1977), this format has not been extensively evaluated. Hansen et al. (1985) strongly recommended a graphic format for the nutrition label because it rapidly conveyed important information and encouraged comparisons between brands. Mohr et al. (1980) suggested that a graphic nutrient density label was more effective in aiding consumer nutrition decisions than the traditional label fonnat was. A replication of this study by Rudd (1986), who added a graphic label, reaffirmed the findings of Mohr et al., and also demonstrated that a simple graphic label format produced the same effect as that of the graphic nutrient density format. Rudd (1989) recently found that the addition of a calorie-based identification statement to the graphic nutrient density label had an impact on consumer perceptions of nutrient quality. Geiger et al. (1990) reported the results of a study of consumer perceptions of label usefulness in purchase decisions by using the marketing research technique of adaptive conjoint analysis. Results indicated that for food purchase decisions, consumers preferred the graphic format over the others; consumers in that study also expressed a preference for the most nutrition information load, presentation of numbers and percentages, and a marranged order of nutrition information (placing those nutrients to be consumed in smaller amounts at the bottom of the label). One technique used to assess the level of information processing is cog- nitive response analysis. Leung-Chung et al. (1985) studied homemakers' cog- nitive responses to nutrition information presented in a conventional versus a graphic format. Each verbalized thought was categorized according to whether it contained nutrition or nonnutrition content references and whether the thought content was semantic (i.e., the respondent made inferences about the nutritional quality of the product, a product attribute, or label characteristics) or sensory (i.e., the respondent repeated verbatim a statement or comparisons exactly as they were presented on the label). Results showed that the graphic label elicited a greater number of semantic responses containing nutrition content Han did He conventional label. Consequently, it was concluded that a food label that pre- sented information in a graphic format, as opposed to the conventional numeric format, enhanced the processing of nutrition information in a meaningful way. Retention and Retrieval In order for information to be useful to the consumer, it must be retained in memory, activated, and retrieved for use in decisionmaking. Food labels must be designed that will be acted upon by both short- and long-term memory. Because short-term memory has very limited storage capacity, the aim must be to keep the label information as simple and relevant as possible. Information processing through long-term memory can be

CONTEXTUAL FACTORS AFFECTING FOOD LABEllNG REFORM 107 facilitated by providing appropriate cues for retrieval. Therefore, informational cues presented on the label must be congruent with the manner in which He same information is likely to be stored in memory. Venkatesan and fellow researchers (1977, 1986) concluded that in a presentation format experiment, the majority of consumers preferred a standardized listing of nutrient contents over a summary score or an index of nutritional quality. Deciszonm~ng When faced with making a decision, most consumers do not attempt to consider all the relevant factors. Ins~, they construct sim- plifying rules that allow them to reach a satisfactory decision. Mazis and Stselin (1982) described a number of general principles that should be followed in helm ing consumers make appropriate decisions. The first principle is concreteness. A decisionmaker tends to use only the information that is explicitly provided and uses it only in the form in which it is displayed. Information is more likely to be used if it is in a form that is directly compatible with the question the consumer is trying to answer, for example, "Is this product high in fat?" or "What is the sodium content of this product?" Also, if the information is poorly organized, consumers will not engage in an extended acquisition effort; thus, attention to the design of an appropriate label format is essential for facilitating consumer comprehension of label information. Another principle is anchoring and adjustment. People ease the strain of integrating information by first using a natural starting point (an anchor) and then adjusting to accommodate additional information. Those informational approaches designed to encourage more com- parison shopping must contend with this process, which helps to explain why it is difficult for consumers to learn or try something new or unfamiliar. Most people seem to think of nutrition in terms of positive attributes, that is, the components of food necessary for good health. However, consumers have recently increased their attention to certain food components that can be referred to as negative nutrients (e.g., calories, cholesterol, sodium, sugar, and various chemical additives) in an effort to avoid these items altogether or choose foods that contain lower amounts of these items (Russo et al., 1986~. Consumers think that knowledge of both positive and negative nutrients is important. When it comes to making food choices, however, they are more concerned with avoiding the negative nutrients (Heimbach, 1987; Heimbach and Stokes, 1979) than they are with choosing foods for their beneficial effects. Heimbach (1982) reported on a survey of consumers in which they were asked to rate the importance of 38 food components, 29 positive and 9 negative. All 9 negative food components were ranked in the top 12 most important components. Indeed, it has been shown that consumers read food labels more to avoid particular food constituents (such as fat, cholesterol, sugar, and sodium) than to seek out positive attributes (Heimbach, 1987~. Asam and Bucklin (1973) varied brand, price, nutrition information, and store location to determine the effect of nutrition labeling on consumer purchase

108 NUTRITION LABELING preferences. The investigators concluded that nutrition labels that used vague descriptors to indicate nutrient content did not affect consumer choice patterns. In addition, detailed nutrition labels that showed average values were used by some consumers and appeared to affect the perception of product quality, while promotional campaigns lessened the effect of nutrition labels. Some studies have suggested that nutrition information on food labels helps consumers decide which brand of a product they should buy. Id one study to determine Me impact of full disclosure on labeling sales of the leading store brands and pnvate-label brands, consumers purchased foods in a Sims supermarket environment (Yankelovich, Inc., 1971~. The results indicated that the dominant brand in each product category held its market share when full disclosure nutrition labeling appeared on a secondary brand, thus suggesting that full disclosure had its strongest effect on the purchase of secondary brands, as long as the brand was not a private label. While consumers appear to use nutrition information on food labels to make comparisons of the same food sold under different brand names, it is disappointing to note that consumers rarely used such information to choose from among general groups of foods to achieve more balanced diets (Rudd and Glanz, 1990~. Fanelli and Abernathy (1986) found Mat 40 percent of the older adults in Weir survey reported that they never read food labels to compare products or to examine ingredients. Summary of Consumer Understanding The field of nutrition is clearly more complex than it was 15 years ago, when the use of nutrition labels became effective; likewise, the amount of nutrition information to which consumers are exposed has expanded exponentially. It is the responsibility of regulatory agencies to require that food manufacturers present nutrition information on food labels in a format that consumers can readily understand and use. Several factors currently contribute to consumer confusion. About 40 per- cent of all food labels do not carry the nutrition information panel and were not designed to be directly relevant to today's dietary recommendations, particularly with respect to consumption of macronutrients, such as fat and cholesterol. In addition, current regulations allow confusion to continue in terminology (e.g., light, low, reduced, and diet) and definitions (e.g., serving sizes). These problems can largely be resolved by a new, well-designed, mandatory nutrition labeling system.

COIVTEXTUAL FACTORS AFFECTING FOOD LABELING REFORM ANALYTICAL CONSIDERATIONS AFFECTING FOOD LABELING REFORM Overview of Analytical Issues 109 The compositions of all food products vary as a result of factors that are not readily controlled or predicted. As with all biological materials, some degree of variability occurs with respect to the composition of foods. Substantial variation in the composition of plant-derived foods occurs as a result of various factors, including soil and climatic variables; maturity of plants at the time of harvesting; conditions of postharvest storage and handling; and conditions of processing, subsequent storage, and handling by consumers (Nagy and Wardowski, 1988; Salunkhe and Desalt 1988~. In addition, genetics and diet affect the composition of animal-derived foods (Froning, 1988; Ockerman, 1988; Renner, 1988~. The composition of seafood is also subject to wide variability ~zynowek, 1988~. Of the various food components, vitamins and minerals are generally subject to the greatest natural variability. For example, as much as 100 percent variation in the folate concentration has been reported among samples of certain vegetable products (Mullin et al., 1982), and extensive variation in ascorbic acid and vitamin A activities also have been reported in vegetables (Klein and Perry, 1982~. Even greater variability in the levels of trace minerals (e.g., selenium) can occur as a function of geographic variation and soil type. For the purpose of nutrition labeling, the precision of analytical methods must be considered in light of the variability of food composition. For example, the natural nutrient composition of some plant-derived foods varies so greatly among samples that determination is difficult or impossible even when the least amount of precision is required. In this case, current nutrition labeling regulations that specify that the actual nutrient content must be within 20 percent of the label claim appear to be overly stringent. The net effect of such regulations is that the stated nutrient values must be set sufficiently low that, within analytical variability, nutrient content of all samples of a product fall within the 20 percent tolerance range. In contrast, the composition of many formulated foods is much more constant, and application of the 20 percent limit is consistent with the need for accuracy and precision but may not be within the accuracy and precision of certain analytical methods. The nutritional significance of minor inaccuracies of label data is probably minimal. A major purpose of analyzing the composition of foods is to provide information concerning the quantities of the various nutrients and other food components as they relate to nutritional status and health. Consumers and health professionals use such information, along with estimates of food intake, for the assessment of dietary adequacy. In the food industry, analytical data are used to monitor processing and manufacturing procedures and as a basis for the data provided in nutrition labeling. Regulatory agencies use analytical data to

110 Nl~1791TION LABEl~NG monitor label claims and to enforce the law against products for which false or exaggerated claims are made. In each situation, the adequacy of methods of food analysis and the accuracy of the results are critical considerations. Continuous improvements are needed in the methodology of food analysis in all applications (i.e., for manufacturing quality control' and regulatory and research purposes). It should be recognized that frequently there are differences in the bioavail- abilities of nutrients; that is, there is often variation in the extent of intestinal absorption and/or metabolic utilization of various nutrients. Thus, analytical data describing food composition ideally should be interpreted with respect to Be bioavailability of each nutrient. This is a particularly significant consideration for such nutrients as iron, calcium, zinc, folate, niacin, and vitamin B6. However, current knowledge of vitamin bioavailability, particularly win respect to folate and vitamin B6, precludes a priori estimation. Similar uncertainty exists in Be case of many minerals, especially iron. This discussion covers issues concerning the analytical basis of food la- beling data and verification of data provided on labels, the adequacy of food composition data bases, and me status of analytical methods for the generation of food composition data. Current Analytical Basis of Food Labeling Information The authority to regulate nutrition labeling is divided between USDA (meat and poultry) and I;DA (all other foods). These agencies differ substantially with respect to their procedures and verification requirements, although He nutrition information provided is similar (Eiattali et al., 1988; Kessler, 1989; Kushner et al., 1990~. Summary of USDA Requirements USDA permits food processors and manufacturers to provide full nutrition labeling in the format and style specified by PDA (Houston, 1985; USDA, 1989c). In addition, USDA also permits an abbreviated format that provides only the content of calories, protein, carbohydrate, and fat per specified serving. Important features of the USDA labeling regulations are that the agency must approve a label prior to its use and that data must be provided to verify the continuing accuracy of nutrition information on the label. The information provided on the label must be accurate, whether determined by calculation from accepted references (e.g., the USDA data base) or from analyses by validated laboratory procedures, such as the Official Methods of Analysis of the Association of Official Analytical Chemists (AOAC, 1989) or the USDA Chemistry Laboratory Guidebook VISTA, 1979~. In He past, the agency had established a partial quality control (PQC) program to ensure label accuracy through formulation control with periodic laboratory analyses,

CONTEXTUAL FACTORS AFFECTING FOOD LABELING REFORM 111 laboratory analyses only, or a combination of formulation control and laboratory analyses (Houston, 1985~. The PQC program was allowed to be waived if there was an adequate data base for a manufactured product (Houston, 1985~. In the interest of accuracy and efficiency, the PQC program was phased out and in its place, USDA initiated Nutrition Label Verification ~V) procedures to verify that labeling is "reasonably accurate." The quantity of supporting data and He degree of precision required on nutrition labels under He NLV program of the Food Safety and Inspection Senice (FSIS) may impose analytical problems for many food processors (USDA, 1988~. The NLV program requires either quarterly (level I) or annual (level II) submission of analytical data on a randomly selected composite sample of the finished product. The label claim must be no less than 80 percent of the analytical values for protein, vitamins, and minerals and no more than 120 percent of the analytical data for calories, fat, cholesterol, fatty acids, sodium, and carbohydrate. It is stated in Policy Memorandum 085B (USDA, 1988) that "because some variability in analytical values can be expected, even though compositing tends to minimize this to a large extent, some over-declaration of calories, carbohydrates, etc., and some under-declaration of protein, vitamins, and minerals is acceptable . . . . However, the over or under-declamtion should be selected so as not to be excessive." USDA generally requires a label change, even when He difference in value is within the 80/120 rule, if a year's worth of data show another number is more accurate. The degree of leeway allowed is not specified. For products that are out of compliance, the manufacturer or processor must resolve the problem (e.g., reformulate) and perform additional analyses for each nutrient in question. Analytical data from at least 10 consecutive production lots of the revised, corrected formula must be submitted to indicate that the product is within the permitted tolerances (USDA, 1988~. Sunday of FDA Requirements In 21 CF-R §101.9, FDA specifies the procedures for nutrition labeling of foods, whether the information is provided voluntarily or is required when the product is formulated with the addition of any nutrient or any nutrition clailTI is made. Labels provide information on the composition of the products in the form in which they are packaged There may be a declaration of nutrient content of the product as consumed after typical (specified) preparation (e.g., breakfast cereal with milk). The nutrition information required on food labels is reviewed in Chapter 3. Although FDA differs from USDA in that it does not require prior approval of nutrition labeling information, the agency may challenge a label to ensure compliance (21 CFR § 101.9(e)~. If, on analysis, values of nutrient contents

112 NUTRITION LABELING declared on the label are not within the defined deviations, the food may be declared misbranded. For foods containing added vitamins, minerals, and proteins, the content of a nutrient in a composite sample must be at least equal to the value for that nutrient indicated on the label. In contrast, for na~aRy occurring nutrients, the nutrient content must be at least equal to 80 percent of the value declared for that nutrient on the label. A provision is made for analytical imprecision, in that no regulatory action will be made for nutrient values less than 80 percent of the label claim by a factor less than He variability generally recognized for the analytical method employed. Furthermore, the regulations specify that reasonable excesses in vitamins, minerals, and protein are acceptable and consistent with good manufacturing practices; reasonable deficiencies of fat, calories, and sodium are similarly allowable. The main specification concerning the accuracy of label claims for calories, fat, carbohydrate, and sodium is that the product will be declared misbranded if the content of each of these components is more than 20 percent in excess of the value indicated on the label. Also included in section 101.9(e) are specifications concerning the methods to be used for ensuring compliance with FDA nutrition labeling regulations. A sampling protocol is specified for taking a composite sample. The composite is to be analyzed by official methods of AOAC or, if no AOAC method is available, by reliable and appropriate analytical procedures. The regulations also indicate that alternative (i.e., non-AOAC) methods of analysis may be submitted to FDA for determination of their acceptability. The criteria used to evaluate Be acceptability of an alternative method are not indicated, however. Alternative methods may be used by the industry provided mat they have been validated (e.g., shown to be equivalent to AOAC official methods) such that they will be acceptable in the event of a regulatory challenge (Victor Hattali Center for Food Safety and Applied Nutrition, FDA, personal communication, 1989~. FDA goes through several steps to enforce compliance with regulations. If Labeling information is not comparable to the information obtained by FDA laboratory analysis, within established limits, FDA first issues a warning letter to the manufacturer. If the violation continues, the agency can seize a product, although seizure of a product for nutrition label misbranding is very rare. Requirements for AOAC or Other Off cial Methods USDA and ADA rely primarily on official methods of AOAC in the implementation and regulation of nutrition labeling. Furthermore, the analytical basis of all areas of food and drug regulation is the validation process developed by AOAC (Huts, 1985~. In this context, an official method is one that has been subjected to an interlaboratory collaborative study to demonstrate its accuracy and precision under the conditions of its intended use. Methods that exhibit suitable performance in such an evaluation are given official final action by

CONTEXTUAL FACTORS AFFECTING FOOD LABELING REFORM 113 AOAC for inclusion in the Off~czal Methods of Analysis (AOAC, 19893. These methods are published at approximately S-year intervals. It is important to recognize that the AOAC official methods are often not the methods of choice in terms of speed, simplicity, and suitability for automation; and with the time required to set up, conduct, and evaluate an interlaboratory study, they often do not represent the most current technologies. Thus, the need exists for flexibility in the selection of analytical methods for verification of food label ~ by the food industry and the regulatory agencies. In practice, the regulatory agencies apparently allow considerable latitude in the selection of the methods of analysis required for nutrition labeling. The criteria that they use for the selection of alternative (i.e., non-AOAC) methods have not been clearly state Designation of a method as official by AOAC requires that the method per- form satisfactorily in interlaboratory collaborative studies under the conditions of intended use. In an important change of policy, the Official Methods Board of AOAC recently took a highly restrictive view of applications of official methods. In a memorandum to James Tanner of EDA (May 19, 1990), AOAC indicated that "The Official Methods Board recommends that collaborative study results apply only to those commodities for which the method was approved. Extension of the method to other commodities should be subjected to a mini-collaborative study." The implications of this policy are far-reaching and detrimental to the reg- ulatory agencies that rely on official methods in regulating labeling compliance. For example, if an official method for measurement of a certain food compo- nent was subjected to collaborative evaluation with respect to an analysis of the nutrient composition of broccoli, it would not be considered applicable to other foods unless the results of further collaborative study were approved by AOAC. Although the scientific rationale is justified, from a regulatory viewpoint this policy effectively removes AOAC official method status from most analytical methods used in the regulation of compliance in nutrition labeling. This problem is particularly acute in specialized aspects of mandatory nutrition labeling, such as specified for infant formula (21 CPR §107.70~. The legal status of the AOAC methods used to monitor compliance of infant formulas, which were not subjected to an AOAC collaborative study for that application, is seriously in question. Selection and Validation of Methods of Food Analysis Among the problems associated with the acquisition of reliable data for food composition are (1) difficulties in the proper implementation of often complex methods of analysis; (2) the complexity of food composition, which makes selection of an appropriate analytical method of primary importance; (3) differences in analytical capabilities of different laboratories, such that different

114 NUTRITION L4BEllNG methods might, by necessity, be used to achieve the same measurement; and (4) the need for improvement of analytical methods as well as better Gaining of technicians (Stewart, 1988, 1989~. These factors apply equally to the generation of data for food labeling, the expansion of food composition data bases, and all aspects of food regulation. The current methods of food analysis leave much to be desired. Methods vary with respect to the suitability for various types of food composition, the convenience and expense, and the degree of analytical expertise required to perfonn tests (Beecher and Vanderslice, 1984~. As proposed by Stewart (1981), the result of an analytical test should be within 10 percent of the actual value when the nutrient of interest is present at nutritionally significant concentrations. Many methods of food analysis are not fully adequate when their accuracy and precision are critically evaluated (Beecher and Vanderslice, 1984~. Precision within a 10 to 15 percent relative standard deviation has been considered to be adequate for all nutrients, except those that are present at nutritionally low concentrations (Beecher and Vanderslice, 1984; Stewart, 1981~. Many methods of nutrient analysis fail to meet these criteria of accuracy and precision, particularly when they are evaluated on an interlaboratory basis. Methods of quality control for nutrient analysis should be strictly specified, but are lacking for nearly all current methods, including the official methods of AOAC and similar organizations. Considerable improvement is needed in the validation and standardization of methods of food analysis for use in nutrition labeling and in the generation of food composition data (Stewart, 1985~. In addition to the evaluation of the accuracy and precision of a method at the time of development, validation of each set of analytical data should also be provided (Stewart, 1989~. Particular emphasis should be placed on the validation of each analysis through proper use of standard laboratories, standard instruments, certified analysts, certified algoritluns, internal standards, pooled samples, standard reference materials, and audit trails (Stewart 1989~. The National Institute of Standards and Technology has developed several biological materials for use as standard reference materials in the determination of various inorganic and a few organic elements in foods (Alvarez, 1984~. Unfortunately, similar reference materials do not exist for many organic food components. It should be noted, however, that the American Association of Cereal Chemists does provide certified bran products suitable for use in fiber analysis and the American Oil Chemists' Society has developed standards for fatty acids and cholesterol. However, furler development and implementation of appropriate reference materials is needed.

CONTEXlVAL FACTORS AFFECTING FOOD LABELING REFORM Analytical Issues Related to Expansion of Nutrition Labeling Problems with Nutrients Proposedfor Inclusion on Nutrition Abets 115 Dietary Fiber Of all the methods for analyzing food components subject to nutrition labeling, those for measurement of dietary fiber have undergone the greatest transition in recent years but remain the most controversial. Because of its heterogeneous nature, dietary fiber is difficult to measure by either chemical or gravimetric methods. Selective hydrolysis and chemical analysis of dietary fiber components (e.g., the Southgate method) provide important qualitative and quantitative information (pizza and Butrum, 1986~. However, the lengthy sample preparation time and the multiple chromatographic analyses required for each sample render such methods unsuitable for routine use in food labeling. A gravime~ic approach for the measurement of total dietary fiber has received official final action by AOAC (Prosky et al., 1985~. This method is not well suited for rapid, repetitive analysis due to its labor-intensive nature, which minimizes automation, although there are currently no suitable alternatives. Previous methods of gravimetric analysis (e.g., crude fiber and neutral detergent fiber methods) underestimated total dietary fiber as a result of losses of soluble fiber components. Although the AOAC method provides generally acceptable results, evidence of some inaccuracy has been reported. For example, Marlett and Navis (1988) observed that the AOAC method overestimated total dietary fiber by 14 percent and 18 percent in chemical analyses of fiber components in samples of apples and a total diet composite, respectively. In view of the physiological properties of certain water-soluble forms of dietary fiber for the reduction of serum cholesterol, there is interest in extending food Labeling to include total, soluble, and insoluble categories of dietary fiber. In this regard, the AOAC method was modified to permit measurement of the soluble and insoluble fiber components (Prosky et al., 1985, 1988~. In an interlaboratory study of the modified method, generally acceptable results were obtained for most foods that were examined, although the need for further refinement of the fractionation procedure was evident. Currently, Were is no simple gravimetric method that is acceptable for measurement of soluble dietary fiber in foods. Because soluble fiber ordinarily makes up a small proportion of total dietary fiber, calculation of soluble fiber by difference (i.e., total dietary fiber minus insoluble fiber) probably would be subject to excessive imprecision. Carbohydrates For the purpose of nutrition labeling, the measurement of total carbohydrates in foods is ordinarily performed indirectly. Total nutritionally available carbohydrates are calculated as total sample weight minus analytical values for protein, fat, moisture, ash, and total dietary fiber. These values are also used in the calculation of caloric content of protein, carbohydrate, and fat

116 NUTRITION MUG on me basis of 4, 4, and 9 kcayg, respectively. At present there are no convenient methods for the direct measurement of total available carbohydrates. It should be noted that this indirect method for determining total available carbohycha~s is subject to potential inaccuracy and imprecision. Large variations have been repaved between measured values for available carbohydrates (sugars and starch) and indirectly calculated total carbohydrates based on handbook values for proximate composition (Li et al., 1988~. Recent improvements in liquid chromatographic and gas chromatographic methods of sugar analysis have provided means for convenient qualitative and quantitative analyses of sugars in foods. Similar methods are available for the measurement of starch and derived oligosaccharides by measuring the amount of glucose released during specific enzymatic hydrolysis. Thus, the potential exists for more accurate, precise, and direct measurements of total carbohy~es, specific sugars, and starches. Problems in the enzymatic hydrolysis of starches in foods that have been subjected to extensive browning (Maillard or caramelization) have been noted, although this would not preclude accurate measurement in most cases. Fats and Cholesterol The Surgeon General's report (DHHS, 1988) recommended that total fat, saturated and unsaturated fatty acids, and cholesterol contents be listed in nutrition labeling. Although this is clearly justified on a nutritional basis, it will inevitably cause substantial increases in the analytical loads of all laboratories that perform these analyses. The measurement of toml fat is a technically simple but cumbersome procedure, although it is routinely performed for all foods with nutrition labels. Extension of nutrition labeling policy to require data on the fatty acid content in foods represents a major analytical difficulty. In this regard, a nonchromatographic spectrophotometric method is available for measurement of total polyunsaturated (cis,cis-methylene-interrupted-polyunsabura~) fatty acids (AOAC, 1989, method 28.082~. The method is limited in Hat it does not identify the fatty acids present. In addition, it was designed for analysis of vegetable oils and requires extensive sample preparation if it is applied to complex samples (e.g., plant and animal tissues and multicomponent foods). Gas chromatographic determination of the individual fatty acids may be performed by using a traditional packed-column method (AOAC, 1989, method 28.060) or by a capillary-column procedure (Slover and Ianza, 1979~. These methods, especially capil~ary-column gas chromatography, provide specific information about the patterns of fatty acids, although sample preparation time is considerable and chromatographic analysis is slow, requiring about 2 hours per sample by capillary-column gas chromatography. Considerable interest exists win respect to the content of bans fatty acid isomers and the omega-3 family of fatty acids in foods. The formation of the bans fatty acid isomer occurs spontaneously in the hydrogenation of fats and

CO}VTEXlVAL FACTORS AFFECTING FOOD LABELING REFORM 117 oils, and naturally occurring bans fatty acids are present in small quantities in animal tissues and products. Trans fauy acid isomers can be Dined in margarine and shortening samples by direct infrared spectrophotomeiry (AOAC, 1989, method 28.086) or in any other food product by capillary~olumn gas chromatography (Slover and Stanza, 1979~. Similarly, in view of the uncertain nutritional benefit of omega-3 fatter acids versus those of other unsatmated fad acids, the additional expense of providing omega-3 fatter acid dam on nutrition labels does not appear to be warranted at this time. Several methods are available for the measurement of cholesterol in foods, with selection based on the type of sample and the equipment that are available. A gas chromatographic method, if properly calibrated, provides the most specific measurement of cholesterol in the presence of other sterols (AOAC, 1989, method 43.283~. Less specific enzymatic methods require careful validation if accurate results are to be obtained (Newman, 1989~. Protein Methods for the measurement of protein in foods are highly reliable. In addition, most foods exhibit relatively little variation with respect to their protein content. Thus, nutrition labeling of the protein component of foods can be accomplished with less difficulty than is encountered for other food components. Vitamins and Minerals Suitable methods appear to exist for the indi- vidual or simultaneous measurement of most minerals of interest (Beecher and Vanderslice, 1984~. However, further efforts in the standardization and valida- tion of methods, along with the further implementation of analytical quality control procedures (e.g., the use of standard reference materials), are need As summarized by Beecher and Vanderslice (1984), considerable development of reliable analytical methods for many of the vitamins is required, including the need for improved methods both for the actual measurement of various nutrients and for the extraction of vitamins from the food prior to analysis. Of the methods applied to vitamins designated as having current or potential public health significance, including folate and vitamins A, B6, and C OSLO, FASEB, 1989), only the method used to analyze vitamin C has been judged to be reliable. Considerable improvement has been made recently in the mea- surement of vitamin A in foods, although substantial uncertainty exists in the selection of appropriate methods (Parrish et al., 1985~. Several suitable liquid chromatographic methods have been developed for the measurement of vitamin B6, but how well those methods compare with traditional microbiological assay procedures (e.g., AOAC) is uncertain (Gregory, 1988~. Certainly, the AOAC methods for analyzing vitamins A and B6 do not reflect recent advances in an- alytical methodologies. Of all of the methods for analyzing vitamins of current interest in nutrition labeling and public health, the poorest are those available for folate analysis.

118 Nu~rRITIoN LABELING Recent improvements in methods for folate analysis have been made, although questions regarding the factors that affect the responses of the various types of assays (i.e., chromatographic, microbiological, or ligand binding) have not been fully resolved (Gregory, 1989~. The Surgeon General's report (DHHS, 1988) indicated that research should be directed toward the development of improved methods for the measurement of folate in foods. The AOAC methods for analyzing folate (AOAC, 1989, methods 43.183~3.190) are designed such that they cannot be applied to the measurement of total folate in foods. The organism used in AOAC assays for folate analysis, Streptococcusfaecalis, does not respond to methyl folates, the major form of the vitamin found in foods. The method also does not provide for the deconjugation of food folates, and thus would grossly underestimate the amount of the vitamin in foods. In Nutrition Monitoring in the United States (LSRO, EASER, 1989), caro- tenes were proposed as a class of nutrients having potential public health signif- icance. Currently, there appears to be little justification for the inclusion of data regarding total carotenes in nutrition labeling on analytical grounds. The AOAC method for carotene analysis is based on cumbersome open-column chromatog- raphy, and its validity is questionable for samples containing complex mixtures of carotenes, as present in most foods (Simpson et al., 1985~. Recent advances have been made in high-performance liquid chromatographic methodology, but individual quantitation of the many naturally occurring carotenes requires con- siderable effort to adapt the procedures to each different type of food sample analyzed (Khachik et al., 1989~. The existence of cis and bans isomers of carotenes further complicates the analysis. Ability of Analytical Laboratories To Accommodate Expanded Nutrition Labeling There is some question among industry representatives concerning the ability of analytical laboratories to manage the increased analytical demand imposed by expanded nutrition labeling requirements. This burden may be as serious for small companies with limited laboratory capabilities as it is for larger manufacturers with diverse product lines. The services of private analytical laboratories are commonly used by all segments of the food industry, particularly with respect to analyses involving nutrition labeling which are frequently not performed in-house on a routine basis, as is the case for quality control analyses used to monitor the formulation and processing of food products. Private analytical laboratories appear to be able to expand to the analytical demands of the marketplace (as influenced, in this case, by nutrition labeling requirements). For laboratories Hat already provide the analyses required for nutrition labeling, accommodating a greater volume of samples would be feasible but would require the addition of equipment and technical personnel. Regulatory agencies would

CONTEmIAL FACTORS AFFECTING FOOD LABEUNG REFORM 119 also require additional resources (personnel, equipment, and laboratory space) to monitor compliance. Nutrition Labeling of Non packaged Foods: Application of Food Composition Data Bases The application of current labeling procedures based principally on direct analytical data appears to be impractical for fresh foods, including meat, poultry, seafood, and produce and foods sold in restaurants. In view of the need for expansion of public access to food composition information; however, the use of information from appropriate data bases may be a useful alternative. However, the validity, analytical basis, and completeness of existing data must be examined. The most complete food composition data base for unprocessed foods is the National Nutrient Data Bank, which is maintained by the Hwnan Nutrition Information Service, USDA. This system contains reasonably complete data concerning the proximate composition of foods, although the data are less complete and potentially less reliable for nutrients for which sound analytical methods are lacking, especially dietary fiber and folate (Hepburn, 1987~. This data base needs to be completed and up some values are woefully out of date. Table 4-1 illustrates the comparatively incomplete state of the USDA data base as of 1987. Since then, progress has been made with respect to the inclusion of additional foods and modest increases in the percentage of foods for which actual analytical data are available for each nutrient (Ruth Matthews, Human Nutrition Information Service, USDA, personal communication, 1990~. The accuracy of information in the USDA data base has not been examined systematically. This is very important, especially for the micronutrients that are subject to considerable natural variation, inadequate sampling, analytical uncertainty, or to which obsolete methods are applied (e.g., cholesterol and iron). With respect to the USDA data base concerning folates in foods, Subar et al. (1989) recently reported on a preliminary comparative study of the content of folate in foods. It was concluded that, despite potential underestimation and overestimation of the actual folate content in foods, overall assessments of dietary folate may be reasonably accurate when the USDA data base is used. As discussed previously, the composition of food is subject to a high degree of variability. This is particularly true of vitamins and minerals. In addition, the amounts of certain major constituents, including dietary fiber, total fat, and the pattern of individual fatty acids, vary across different samples of the same food. To generate information for use in data bases, appropriate sampling protocols would involve selection of samples from venous sources to compensate for geographic and seasonal variation. Data bases currently in use may be biased if values are based on inappropriate sampling procedures or have been imputed without consideration of these variables.

120 Nl~RlTION LABELING TABLE 4-! Percentage of Analytical Data for a Given Nutrient in USDA Primary Data Set Nutrient Percentage All Best Foods Sources Calcium 97 Protein 97 Pat 96 Thiamin 91 Riboflavin 91 Niacin 91 Sodium 90 Potassium 90 Phosphorus 90 Iron 90 Vitamin C 83 92 Vitamin A (IU) 80 89 Cholesterol 80 Magnesium 75 72 Zinc 73 79 Copper 67 71 Vitamin B6 64 72 Vitamin B12 64 79 Vitamin A (RE) 61 73 Folate 56 69 Carotene 54 88 Dietary fiber 29 40 a-Tocopherol 28 39 NOTE: The USDA Primary Data Set contains data on basic foods, including ingredients of foods such as flour. SOURCE: Adapted from Hepburn, F. 1987. Food con- sumptiontfood composition inte~relabonships. Pp. 68 74 in Re- search on Survey Methodology, HNIS Report No. Adm-382. Hu- man Nutrition Information Service, U.S. Department of Agricul- ture, Hyattsville, Md. The information in a data base may indicate representative food composi- tion, but there may be large differences between data base values and the actual concentrations of certain nutrients in a single specimen of that food. It has been suggested that information from appropriate data bases (e.g., USDA) could be used as the basis for nutrition labeling of certain foods that cannot be read- ily analyzed during distribution (e.g., fresh meat, poultry, seafood, fmits, and vegetables). If this data base is to be used to provide nutrient composition data

CONTEXTUAL FACTORS AFFECTING FOOD LABEl]NG REFORM 121 for nonpackaged foods, this potential for substantial variability with respect to micronutrient content must be recognized. In addition to the USDA data base, other sources of food composition data are available for use in nutrition labeling of foods. Although not actually formal data bases, analytical data provided by various trade associations clearly complement other sources of information for nutrition labeling. Some of these data bases are reviewed by FDA, although no formal certification procedure currently exists. Data bases may serve as a mechanism for providing reasonably represen- tative nutrient content data for foods sold in restaurants, particularly those pro- viding packaged products in a highly standardized format (i.e., for foods sold in limited-menu restaurants). At present, there appears to be no means of analytical verification of the nutrient content of foods sold in other resultants and non- commercial food service settings, although the nutrient composition of recipe menus could be calculated by using food composition data from nutrient data bases. Committee Recommendations All nutrition labeling is predicated on acceptable accuracy and precision of the information provided, whether obtained by direct analysis or indirectly from a food composition data base. Thus, the validity of nutrition labeling ultimately depends on the adequacy of analytical methods used in food analysis and their appropriate application. Because of the key role of food analysis in nutrition labeling and in view of the analytical limitations described in this section, He Committee recommends that: · Label verification by analysis of composite samples should be made at least twice each year to ensure reasonable accuracy of nutrition labels without imposing the burden of a complete quarterly analysis. Although Mere are clear meets of the USDA system of label verification in terms of ensuring accuracy, management of the FDA system seems much less costly. · FDA and USDA should certify data from the National Nutrient Data Bank or over appropriate sources regarding the nutrient content of fresh foods and foods sold in restaurants. · FDA and USDA should allow considerable flexibility in the selection of analytical methods for label verification. The limitations of certain official methods hinder the analytical process, given the volume of analyses performed. In proposing an alternative (nonofficial) analytical method, suitable verifi- cation must be required (e.g., recovery of samples and analysis of refer- ence materials), and appropriate quality control procedures should be used in each analysis. A mechanism should be developed to verify the quality

122 NUTRITION LABELING control measures that are used whenever analytical data are submitted to a regulatory agency. · Development of additional standard reference materials for use in food analysis should be encouraged. · Funding should be provided for the development of improved analytical methods, establishment of programs for the testing of methods Trough interl~ratory studies, and development of additional standard reference materials. Completion and expansion of the USDA National Nutrient Data Bank should be continued. It is recommended that the relative merits of the various alternative food composition data bases be examined and that efforts toward He consolidation of data bases be support REFERENCES Achterberg, C. 1990. Information presented at the Workshop on Consumer Understanding and Use of Food Labels, Committee on the Nutrition Components of Food Labeling, Food and Nutrition Board, Institute of Medicine, Washington D.C. March 13. Achterberg, C., G. Auld, V. Getty, and J. Durrwachter. 1990. Misconceptions about fat and cholesterol in a sample of Pennsylvania men and women. Unpublished draft paper. Pennsylvania State University, University Park. ADA (American Dietetic Association). 1990. Position of the American Dietetic Associ- ation: Nutrition education for the public. J. Am. Diet. Assoc. 90:107-110. Alvarez, R. 1984. NBS standard reference materials for food analysis. Pp. 81-99 in Modern Methods of Food Analysis, K.K. Stewart and J.R. Whitaker, eds. AVI Publishing Co., Westport, Conn. AOAC (Association of Official Analytical Chemists). 1989. Official Methods of Analysis, 15th ed. AOAC, Washington, D.C. 1298 pp. Asam, E.H., and L.P. Bucklin. 1973. Nutrition labeling for canned goods: A study of consumer response. J. Marketing 37:32-37. Babcock MJ., and M.M. Murphy. 1973. Two nutrition labeling systems. J. Am. Diet. Assoc. 62:155-161. Beecher, G.R., and J.T. Vanderslice. 1984. Determination of nutrients in foods: Factors that must be considered. Pp. 29-56 in Modern Methods of Food Analysis, K.K. Stewart and J.R. Whitaker, eds. AVI Publishing Co., Westport, Conn. Bettman, J.R. 1979. An Information Processing Theory of Consumer Choice. Addison- Wesley, Reading, Mass. 402 pp. CNI (Community Nutrition Institute). 1988. FDA, private surveys disagree on diet trends. Nutr. Week. June 23. 18~23~:1 Daly, P. 1976. The response of consumers to nutrition labeling. J. Consumer Affairs 10:170-178. DHEW/USDA/FIC (U.S. Department of Health, Education, and Welfare; U.S. Depart- ment of Agriculture; and Federal Trade Commission). 1979. Food Labeling Back- ground Papers. Govemment Printing Office, Washington, D.C. 124 pp. DHHS (U.S. Department of Health and Human Services). 1988. The Surgeon General's

CONTEXTUAL FACTORS AFFECTING FOOD LABEllNG REFORM 123 Report on Nutrition and Health. Government Printing Office, Washington, D.C. 727 PP DHHS/USDA (U.S. Department of Health and Human Services and U.S. Department of Agriculture). 1986. Nutrition Monitoring in the United States: A Report from the Joint Nutrition Monitoring Evaluation Committee. Goverr~nent Printing Office, Washington, D.C. 356 pp. DOC (U.S. Department of Commerce). 1990. 1990 U.S. Industrial Outlook January). Government Printing Office, Washington D.C. 534 pp. Fanelli, M., and M. Abernathy. 1986. A nutritional questionnaire for older adults. Gerontologist 26:192-197. FMI (Food Marketing Institute). 1989. Trends. Consumer Attitudes and the Supermarket. FMI, Washington D.C. 65 pp. FMI (Food Marketing Institute). 1990. Trends. Consumer Attitudes and the Supermarket. FMI, Washington D.C. 70 pp. Frattali, V.P., J.E. Vandeneen, and A.L. Forbes. 1988. The role of the United States government in regulating the nutritional value of the food supply. Pp. 687-705 in Nutritional Evaluation of Food Processing, 3rd ea., E. Karmas and R.S. Hams, eds. Van Nostrand Reinhold Co., New York. Friedman, M., ed. 1990. Gorman's New Product News. Gorman Publishing Co., Chicago. January 5, vol. 25, no. 12. 46 pp. Froning, G.W. 1988. Effects of agricultural practices on poultry and eggs. Pp. 225-244 in Nutritional Evaluation of Food Processing, 3rd ea., E. Karmas and R.S. Harris, ads. Van Nostrand Reinhold Co., New York. Geiger, CJ., B.W. Wyse, C.R.M. Parent, and R.G. Hansen. 1990. The use of adaptive conjoint analysis (ACA) to determine the most useful nutrition label for purchase decisions. Abstract 4587. Annual Meeting of the Federation of American Societies for Experimental Biology, Washington, D.C. Glanz, K., and R. Mullis. 1988. Environmental interventions to promote healthy eating: A review of models, programs, and evidence. Health Ed. Q. 15~4~:395-415. Glanz, K., J. Rudy, R.M. Mullis, and P. Snyder. 1989. Point of choice nutrition information, federal regulations and consumer health education: A critical view. J. Nutr. Ed. 21~2~:95-100. Glascoff, M A., S. Taylor, and D.W. Glascoff. 1986. A social marketing approach to reducing salt intake. Health Ed. 17~2~:11-14. Gregory, J.F. 1988. Methods for determination of vitamin B6 in foods and other biological materials: A critical review. J. Food Comp. Anal. 1:10~123. Gregory, J.F. 1989. Chemical and nutritional aspects of folate research: Analytical procedures, methods of folate synthesis, stability, and bioavail~ibility of dietary folates. Pp. 1-101 in Advances in Food and Nutrition Research, vol. 33. J. Kinsella, ed. Academic Press, Orlando, Fly HalL R.L. 1977. Food additives: An industry view. FDA Consumer 11~10~:~11. Hammonds, T. 1978. Testimony before the Subcommittee on Nutrition and Investigations, Committee on Agriculture, Nutrition, and Forestry, U.S. Senate, Washington D.C. August 9-10. Hansen, R.G., C.T. Windham, and B.W. Wyse. 1985. Nutnent density and food labeling. Clin. Nutr. 4:164 170.

124 NUTRITION LABELING Hams, S., and S. Welsh. 1989. How well are our food choices meeting our nutrition needs? Nutr. Today 24:2~28. Heimbach, J.T. 1982. Public Understanding of Food Label Information. Food and long Administration, Washington, D.C. 24 pp. Heimbach, J.T.1985. Cardiovascular disease and diet The public view. Public Health Rep. 100:5. Heimbach, J.T. 1986. The growing impact of sodium labeling of foods. Food Technol. 40(12):102. Heimbach, J.T.1987. Risk avoidance in consumer approaches to diet and health. Clin. Nutr. 6:159. Heimbach, J.T., and R.C. Stokes.1979. Food and Drug Administration, 1978 Consumer Food Labeling Survey. Food and Drug Administration, U.S. Department of Health and Human Services, Washington, D.C. 133 pp. Hepbum, F. 1987. Food consumption/food composition interrelationships. Pp. 68-74 in Research on Survey Methodology, HNIS Report No. Adm-382. Human Nutrition Information Service, U.S. Department of Agriculture, Hyattsville, Md. Hixson, M.L., R.C. Lefebre, and S. Banspach. 1988. Evaluation of a grocery store point of purchase nutrition intervention program. Paper presented at the 1988 Annual Meeting of the Society for Behavioral Medicine, Boston. Hochbaum, G. 1981. Strategies and their rationale for changing people's eating habits. J. Nutr. Ed. 13(Suppl):59~5. Houston, D.L.1985. USDA's regulation of food claims. Food, Drug, Cosmetic Law J. 40:238-243. Hutt, PB. 1985. The importance of analytical chemistry to food and drug regulation. Vanderbilt Law Rev. 38:479~93. IFIC/ADA International Food Information Council and The American Dietetic Asso- ciation). 1990. How Are Americans Making Food Choices? Results of a Gallup Survey. IFIC, Washington D.C. 27 pp. Jacoby, J., R.W. Chestnut, and W. Silberman. 1977. Consumer use and comprehension of nutrition information. J. Consumer Res. 4:119-128. Jacoby, J., J.C. Olson, GJ. Sybillo, and E.W. Hart. 1979. Behavioral science perspec- tives on conveying nutrition information to consumers. P. 2a in Criteria of Food Ac- ceptance: How Man Chooses What He Eats, J. Solms and R.L. Hall, eds. Forster, Zurich. 461 pp. Kessler, D.A. 1989. The federal regulation of food labeling. Promoting foods to prevent disease. N. Engl. J. Med. 321:717-725. Khachik, F., G.R. Beecher, and W.R. Lusby. 1989. Separation, identification, and quantification of the major carotenoids in extracts of apricots, peaches, cantaloupe, and pink grapefruit by liquid chromatography. J. Agric. Food Chem. 37:1465-1473. Klein, B.P., and A.K. Perry. 1982. Ascorbic acid and vitamin A activity in selected vegetables from different geographical areas of the United States. J. Food Sci. 47:941-945. K~wok, J. 1988. Effects of handling, processing, and storage on fish and shellfish. Pp. 245-265 in Nutritional Evaluation of Food Processing, 3rd ea., E. Karmas and R.S. Harris, eds. Van Nostrand Reinhold Co., New York. Kushner, GJ., R.S. Silvemlan, S.B. Steinbom, and R.A. Johnson. 1990. A Guide

CONTEXTUAL FACT ORS AFFECTING FOOD LABELING REFORM 125 to Federal Food Labeling Requirements. Prepared for the U.S. Department of Agriculture and the U.S. Department of Health and Human Services, Washington, D.C. 42 pp. Lanza, E., and R.R. Butrum. 1986. A critical review of food fiber analysis and dab J. Am. Diet. Assoc. 86:732-743. Lecos, C. 1988. Food labels test your food label knowledge. FDA Consumer 22~2~:16- 21. Lenahan, RJ., J.A. Thomas, D.A. Taylor, D.L. Call, and P.I. Padberg. 1972. Consumer react8On to nutrition information on food product labels. Search Agric. 2~15~:1-26. Leung-Chung, E., L.S. Sims, and J. Olson. 1985. Assessing consumers' comprehension of nutrition information on food labels. Unpublished working paper. Pennsylvania State University, University Park. Levy, A.S., and M. Stephenson. 1990. Nutrition knowledge levels about dietary fats and cholesterol: 1983-1988. Unpublished paper. Food and Drug Administration, Washington, D.C. 20 pp. Levy, A.S., and R.C. Stokes. 1987. Effects of a health promotion advertising campaign on sales of ready-to-eat cereals. Public Health Rep. 102:39~403. Levy, A.S., O. Mathews, M. Stephenson, J.E. Tenney, and R.E. Schucker. 1985. The impact of a nutrition information program on food purchases. J. Marketing Public Policy 4:1-3. Lit B.W., M.W. MarshalL K.W. Andrews, and T.T. Adams. 1988. Analysis of individual foods for the validation of sugars and starch contents of composited diets. J. Food Comp. Anal. 1:152-158. Light, L, B. Portnoy, J.E. Blair, J.M. Smith, A B. Rodgers, E. nlckermanty, J. Tenney, and O. Mathews. 1989. Nutrition education in supermarkets. Family and Community Health: J. Health Promotion Maintenance 12~1~:43-52. LSRO, FASEB (Life Sciences Research Office, Federation of American Societies for Experimental Biology). 1984. Assessment of the Folate Nutritional Status of the U.S. Population Based on Data Collected in the Second National Health and Nutrition Examination Survey, 197~1980, F.R. Senti and S.M. Pitch, eds. FASEB, Bethesda, Md. 96 pp. LSRO, FASEB (Life Sciences Research Office, Federation of American Societies for Experimental Biology). 1989. Nutrition Monitonng in the United States: An Update Report on Nutrition Monitoring. Prepared for the U.S. Department of Agriculture and the U.S. Department of Health and Human Services. Govemment Printing Office, Washington, D.C. 408 pp. Marlett, J. A., and D. Navis. 1988. Companson of gravimetric and chemical analyses of total dietary fiber in foods. J. Agric. Food Chem. 36:311-315. Marlett, J.A., J.G. Chesters, MJ. Longacre, and JJ. Bogdanske. 1989. Recovery of soluble dietary fiber is dependent on the method of analysis. Am. J. Clin. Nutr. 50:479~85. Mazis, M B., and R. Staelin. 1982. Using information-processing principles in public policymaking. J. Marketing Public Policy 1:314. McCullough, J., and R. Best. 1980. Consumer preference for food label information: A basis for segmentation. J. Consumer Affairs 14~1~:180-192. McGuire, WJ. 1969. The nature of attitudes and attitude change. Pp. 136-314 in

126 NUTRITION L4BELlNG Handbook of Social Psychology, vol. 3, 2nd ea., G. Lindzey and E. Aronson, eds. Addison-Wesley, Reading, Mass. McGuire, WJ. 1976. Some internal psychological factors influencing consumer choices. J. Consumer Res. 2:302. Mohr, K.G., B.W. Wyse, and R.G. Hansen. 1980. Aiding consumer nutrition decisions: Companson of a graphical nutrient density labeling format with the current food label system. Home Econ. Res. J. 8~3~:162-172. Muller, T.E. 1985. Structural information factors which stimulate the use of nutrition information: A field experiment. J. Marketing Res. 22~2~: 143-157. Mullion, WJ., D.F. Wood, and S.G. Howsam. 1982. Some factors affecting folacin content of spinach, Swiss chard, broccoli and Brussels sprouts. Nutr. Rep. Ink 26:7-16. Mullis, R.M., M.K. Hunt, M. Foster, L. Hachfeld, D. Lansing, P. Snyder, and P. Pirie. 1987. The Shop Smart for Your Heart grocery program. J. Nutr. Ed. 19~5~:225-228. Nagy, S., and W.P. Wardowski. 1988. Effect of agricultural practices, handling, process- ing, and storage on fruits. Pp. 73-100 in Nutritional Evaluation of Food Processing, 3rd ea., E. Karmas and R.S. Hams, eds. Van Nostrand Reinhold Co., New York. Newman, A.R. 1989. Measuring the fat of the land. Anal. Chem. 61: 663A-664A. NRC (National Research Council). 1989. Diet and Health: Implications for Reducing Chronic Disease Risk. Report of the Committee on Diet and Health, Food and Nu- trition Board, Commission on Life Sciences. National Academy Press, Washington, D.C. 749 pp. Ockerman, H.W. 1988. Effects of agricultural practices, handling, processing, and storage on meat. Pp. 153-202 in Nutritional Evaluation of Food Processing, 3rd ea., E. Karmas and R.S. Harris, eds. Van Nostrand Reinhold Co., New York. Olson, J., and L.S. Sims. 1980. Assessing nutrition knowledge from an information processing perspective. J. Nutr. Ed. 12~3~:157-161. ORC (Opinion Research Corporation). 1990. Food Labeling and Nutrition: What Amer- icans Want. Survey conducted for the National Food Processors Association. ORC, Washington D.C. 178 pp. Parrish, D B., R.A. Moffitt, RJ. Noel, and J.N. Thompson. 1985. Vitamin A. Pp. 153- 184 in Methods of Vitamin Assay, 4th ea., J. Augustin, B.P. Klein, D. Becker, and P.B. Venugopal, eds. John Wiley & Sons, New York. 590 pp. Prosky, L., N.-G. Asp, I. Furda, J.W. DeVries, T.F. Schweizer, and B.F. Harland. 1985. Determination of total dietary fiber in foods and food products: Collaborative study. J. Assoc. Off. Anal. Chem. 68:677~79. Prosky, L., Nob. Asp, T.F. Schweizer, J.W. DeVries, and I. Furda. 1988. Determination of insoluble, soluble, and total dietary fiber in foods and food products: Interlabo- ratory study. J. Assoc. Off. Anal. Chem. 71:1017-1023. Putnam, JJ., and J. Weimer. 1981. Household diet changes linked to nutrition concerns. Staff paper of Economic and Statistics Service, U.S. Department of Agriculture, Washington, D.C. 11 pp. Quelch, J^. 1977. The role of nutrition information in national nutrition policy. Nub. Rev. 35~11~:289-293. Rahn, MJ. 1980. Nutrition knowledge of a sample of urban women. Unpublished M.S. thesis. Faculty of Graduate Studies, University of Guelph, Guelph, Ontario, Canada. Renner, E. 1988. Effects of agricultural practices on mink and dairy products. Pp. 203-

CONTEXTUAL FACTORS AFFECTING FOOD LABELING REFORA! 127 224 in Nutntional Evaluation of Food Processing, 3rd ea., E. Kannas and R.S. Hams, eds. Van Nostrand Reinhold Co., New York. Robinson, J.P. 1990. Thanks for reading this. Am. Demogr. 12~2~:~7. Rudd, J. 1986. Aiding consumer nutrition decisions with the simple graphic label format. Home Econ. Res. J. 14~3~:342-346. Rudd, J. 1989. Consumer response to caloric base variations on the graphical nutrient density food label. J. Nutr. Ed. 21:259-264. Rudd, J., and K. Glanz. 1990. How individuals use information for health action: Consumer information processing. In Health Behavior and Health Education, K. Glanz, F.M. Lewis, and B.K. Rimer, eds. Jossey-Bass, San Francisco. Russo, J.E., R. Stselin, CODA. Nolan, GJ. RusselL and B.L Metcalf. 1986. Nutrition information in the supermarket. J. Consumer Res. 13:48-70. Salunkhe, D.K., and B.B. Desai. 1988. Effects of agricultural practices, handling, processing, and storage on vegetables. Pp. 21-71 in Nutritional Evaluation of Food Processing, 3rd ea., E. Karmas and R.S. Harris, eds. Van Nostrand Reinhold Co., New York. Sansolo, M., ed. 1990. Progressive Grocer. April. MacLean Hunter Media, Stamford, Conn. 66 pp. Scammon, D. 1978. Information load and consumers. J. Consumer Res. 4~3~:148-155. Sempos, C., R. Fulwood, C. Haines, M. CarrolL R. Anda, D.F. Williamson, P. Remington, and J. Cleeman. 1989. The prevalence of high blood cholesterol levels among adults in the United States. J. Am. Med. Assoc. 262~1~:45-52. Shepherd, S.K. 1990. Nutrition and the consumer: Meeting the challenge of nutrition education in the l990s. Food Nutr. News 62~1~:1-3. Shepherd, S.K., US. Sims, FJ. Cronin, A. Shaw, and C.A. Davis. 1989. Use of focus groups to explore consumers' preferences for content and graphic design of nutrition publications. J. Am. Diet. Assoc. 89~11~:1612-1614. Simpson, K.L., S.C.S. Tsou, and C.O. Chichester. 1985. Carotenes. Pp. 185-220 in Methods of Vitamin Assay, 4th ed. J. Augustin, B.P. Klein, D. Becker, and PB. Venugopal, eds. John Wiley ~ Sons, New York. Sims, L.S. 1981. Further thoughts on research perspectives in nutrition education. J. Nutr. Ed. 13(1):S7~S75. Sims, L.S., and Sheperd, S.K. 1985. Further Exploration of Formatting, Structuring, and Sequencing of Nutrition Information for Household Food Managers. Final report submitted in partial fulfillment of FNS contact no. 53-3198~66. U.S. Deparunent of Agriculture, Washington, D.C. 97 pp. Sloan, A.E. 1987. Educating a nutrition-wise public. J. Nutr. Ed. 19:30~305. Slaver, H.T., and E. Stanza. 1979. Quantitative analysis of food fatter acids by capillary gas chromatography. J. Am. Oil Chem. Sac. 56:933-943. Snider, S., P. Kendal, W. Hurst, C. Bueso, and E. Burns. 1990. Regional Communicators' Focus Group Summary. Institute of Food Technologists, Chicago. 24 pp. Stewart, K.K. 1981. Nutrient analyses of food: A review and a strategy for the future. In Beltsville Symposia in Agricultural Research [4] Human Nutrition Research, G.R. Beecher, ed. Allanheld, Osmun Publishers, Totowa, NJ. Stewart, K.K. 1985. Method choice and development. Pp. 1-15 in Methods of Vitamin

128 NUTRITION LABELING Assay, 4th ea., J. Augustin, B.P. Klein, D. Becker, and PB. Venugopal, eds. John Wiley ~ Sons, New York. Stewart, K.K. 1988. Improvement of food composition data. Needs for analytical training. Editorial. J. Food Comp. Anal. 1:291-292. Stewart, K.K. 1989. Data validation. Editorial. J. Food Comp. Anal. 2:91-92. Subar, A.F., G. Block, and L.D. James. 1989. Folate intake and food sources in the US population. Am. J. Clin. Nutr. 50:50~516. Timmer, C.P., and M.C. Nesheim. 1979. Nutrition, product quality, and safety. Pp. 155- 203 in Consensus and Conflict in U.S. Agriculture: Perspectives from the National Farm Summit, B.L. Gardner and J.W. Richardson, ads. Texas A&M University Press, College Station. 280 pp. USDA (U.S. Department of Agriculture). 1979. Chemistry Laboratory Guidebook. Food Safety and Quality Service, Science. Government Printing Office, Washington, D.C. USDA (U.S. Department of Agriculture). 1988. FSIS Policy Memorandum 085B. Food Safety and Inspection Service, Washington, D.C. USDA (U.S. Department of Agriculture). 1989a. Food Cost Review. Agricultural Eco- nomic Report No. 615 Duly). Economic Research Service, Washington, D.C. 54 PP. USDA (U.S. Department of Agnculture). 1989b. National Food Review 1989 Yearbook: Food Beyond the Farm Gate (April-June). Economic Research Service, Washington, D.C. 52 pp. Venkatesan, M. 1977. Providing nutritional information to consumers. Paper presented at Special National Science Foundation~Iassachusetts Institute of Technology Con- ference on Consumer Research for Consumer Policy, Cambridge, Mass., July. Venkatesan, M., W. Lancaster, and K.W. Kendall. 1986. An empirical study of alter- nate formats for nutritional information disclosure in advertising. J. Public Policy Marketing 5:2903. Woolcott, D.M. 1983. Nutrition concerns and information seeking behavior of rural and urban women. Paper presented at the Society for Nutrition Education, Denver. Yankelovich, Inc. 1971. Nutrition labeling: A consumer experiment to determine the effects of nutrition labeling on food purchases. Chain Store Age panuary):57-77.

Next: Part II: Reforming Food Labels »
Nutrition Labeling: Issues and Directions for the 1990s Get This Book
×
Buy Paperback | $70.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Nutrition Labeling offers a thorough examination of current nutrition labeling practices and recommends ways to make food labeling information consistent with recent dietary recommendations from the U.S. Surgeon General and the National Research Council.

The volume proposes implementing a food labeling reform program, addressing such key issues as requiring mandatory nutrition labeling on most packaged foods, expanding nutrition labeling to foods that do not currently provide this information, making federal requirements uniform between agencies, and updating the nutrient content and format of food labels.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!