May 2008 Issue
Savvy Shopping: Make Way for Food Rating Systems
By Rita E. Carey, MS, RD, CDE
Today’s Dietitian
Vol. 10 No. 5 P. 46
They may evaluate foods using different criteria, but all nutrition profiling systems have the same general goal: to guide consumers toward healthier products and help them more easily assess the nutritional value of their choices.
In the introduction to the 2005 Dietary Guidelines for Americans, the Health and Human Services (HHS)/USDA Web site states that despite a growing body of evidence that suggests following the recommendations will lower mortality rates and the risk of chronic disease, Americans’ adherence to these guidelines is low. The Web site also notes that Americans are more overweight and obese than ever before.
The guidelines attempt to address undernutrition and overnutrition in the United States by recommending that consumers ingest more nutrient-dense and low-calorie foods. However, given their relative ineffectiveness to date, it is unlikely these guidelines can effect much change on their own.
Hence, before the 2005 guidelines were published, government agencies began looking for outside help in developing new ways to guide consumer diet choices.1 This search, combined with consumer confusion about dietary guidelines and food labeling, has provided the impetus for both retailers and health professionals to devise nutrition rating systems to make eating well easier for Americans. Nutrition profiling is the term used to describe this new process of rating the relative healthfulness of foods. Time will tell whether this approach will help improve consumers’ health.
Guided by the Stars
If you live on the East Coast, you may shop at a Hannaford Brothers supermarket—the first in the United States to develop and implement a nutrition profiling system. If you don’t have a Hannaford near you, a quick trip to the company’s Web site (www.hannaford.com) will give you a virtual view of its Guiding Stars program, developed to help consumers quickly identify the healthiest choices on store shelves.
Hannaford bases its food rating system on the nutrients listed on a standard Nutrition Facts label and the ingredients list for a product. Foods gain points for fiber, whole grains, vitamins A and C, iron, and calcium and lose points for excessive sodium, cholesterol, added sugar, and total saturated and trans fat content. Meats, seafood, poultry, dairy, and nuts are rated differently from other foods to avoid excessive scoring penalties for their natural lack of fiber or their cholesterol content. Nutrient content is evaluated per 100-calorie portions of foods.
The Guiding Stars rating system was developed by a panel of health and research professionals who referred to the human nutrition guidelines established by the FDA, the USDA, HHS, the National Academies of Science, and the World Health Organization to establish a proprietary formula for determining the healthfulness of foods.
Hannaford got some interesting results (published on its Web site) when it evaluated the more than 25,000 foods sold in its stores and rated them on a scale of 0 to 3 stars. Of the foods rated—water, alcohol, and other items with less than 5 calories per serving were excluded—only 28% of them received one or more stars. In fact, some foods that are currently labeled and marketed with health claims (eg, some low-sodium products) received no stars, a result that could elicit complaints from food manufacturers.
Health professionals and dietitians, however, may interpret Hannaford’s results differently. When foods are evaluated from a holistic perspective (considering the overall nutrient value instead of the lack of one or two undesirable nutrients), as the Guiding Stars system attempts to do, products that make health claims according to FDA guidelines do not always stand up to higher levels of scrutiny.
As dietitians know, “healthy” foods are typically defined by their low levels of certain undesirable elements (eg, sodium, fat, cholesterol, and simple sugars) or their relatively high levels of individual nutrients (eg, fiber, calcium, and iron). The black licorice in my desk drawer sports a “low-fat food” label, but the label does not state that it is also a significant source of calories, being composed primarily of refined wheat flour and sweeteners.
Nutrient content claims such as “low fat,” “reduced,” or “good source of” certainly have value, but they can also be misleading if they divert focus from the overall nutritional quality and caloric content of a food to the presence or absence of one or more individual elements. And confusion can cause consumers to make less-than-optimal choices—for example, I’ve counseled clients who shake their heads in discouragement when they discover that the low-fat food they have been purchasing is extraordinarily high in sodium, added sugars, or calories.
Nutrition profiling systems attempt to relieve the nutrition tunnel vision many experience by providing a snapshot evaluation of overall nutrient and caloric density. These systems align themselves with the HHS/USDA dietary guidelines, which encourage individuals to consume lower calorie foods with a higher nutrient density. Because the U.S. government does not define nutrient or caloric density, a variety of complicated algorithms have been devised to do just that.
More Systems on the Scene
The Guiding Stars system bases its estimates of nutrient density on the information available on any Nutrition Facts label. Other systems (eg, the Naturally Nutrient Rich [NNR] score and the Overall Nutrition Quality Index [ONQI]) are more complex. While Guiding Stars and the NNR and ONQI represent profiling systems developed in the United States, it is important to note that a number of systems are being developed and used in the European Union—examples being the traffic light system endorsed by the Sainsbury’s grocery chain and Unilever’s profiling formula, both designed to guide improvement of the nutritional quality of their products. The European Union also plans to use nutrition profiling as the basis for regulating nutrition and health claims on foods sold in member countries.2
The NNR score is calculated as the mean ratio of the amounts of 15 health-promoting nutrients relative to their established recommended daily intakes (RDIs) in 2,000 kilocalories of food.3,4 The nutrients included in the NNR algorithm are compared with other nutrition profiling systems in Table 1. Unlike the Guiding Stars system, the NNR does not fault a food for fat, sodium, or sugar content, though foods with higher energy content do receive lower scores because nutrients are calculated per calorie. Each nutrient considered in the NNR formula is weighted equally, and the maximum allowable nutrient levels per 2,000 kilocalories are capped, preventing large quantities of single nutrients from skewing the total score. The score can be used in making comparisons between foods in the same category and foods in different categories. For example, the nutrition density of pears vs. bananas and bananas vs. marshmallows could be fairly assessed.
The NNR score has been tested against other algorithms for identifying nutrient- and energy-dense foods and is currently being modified to include fat, sodium, and sugar content for greater accuracy. This new rating formula is called the Nutrient-Rich Foods (NRF) index.
Adam Drewnowski, PhD, of the School of Public Health and Community Medicine at the University of Washington in Seattle has led development of the NNR score and NRF index. Drewnowski states in several papers that because current concepts of healthful, nutritious foods are not based on consistent standards in the literature and because the term nutrient dense has not been defined, consumer confusion is inevitable.3,4 Creating a standardized nutrient-density score such as the NNR will help individuals identify foods with the highest levels of nutrients per calorie—essentially helping them get more nutritional bang for their buck.
“I think the main issue [with the obesity epidemic and poor health of Americans] is that the diet has really become energy rich but nutrient poor. As we have been saying, people are becoming undernourished yet overfed. Anything that focuses attention on the nutrients foods contain—always in relation to calories—is a step in the positive direction,” Drewnowski says.
The ONQI considers an even more complex array of nutritional indices than the NNR. Developed by a team of 12 nutrition experts, including Walter Willett, MD, DrPH, from Harvard, and Barbara Rolls, PhD, from Penn State, the ONQI assigns nutrient scores to foods based on a formula that includes 30 different factors (see Chart).5 Unlike the NNR, which gives equal value to all nutrients entered into its equations, the ONQI assigns greater value to some nutrients that are strongly associated with specific health outcomes, such as cancer or heart disease.
The ONQI rates foods on a scale of 1 to 100 and considers factors other than the relative concentration of different nutrients. The algorithm also gives value to glycemic load, energy density, and protein quality in addition to nutrients that currently have no official RDI, such as omega-3 fatty acids, bioflavonoids, and carotenoids. Ratings posted at www.onqi.com are so far unsurprising: Mustard greens and strawberries have a score of 100, salmon gets an 87, and regular soda receives a 1.
David L. Katz, MD, MPH, director of the Yale Griffin Prevention Research Center and chair of the team that created the ONQI, describes its validation as a multilayered process. The ONQI advisory panel first ranked various foods on the basis of individual assessment of nutrition quality. These foods were then rated using the ONQI algorithm, and a correlation analysis was performed. The panel also reviewed any anomalous scores produced by the algorithm, checking nutrient details of the foods in question to ensure accuracy. These reviews resulted in a few changes to the algorithm.
The ONQI has also been market tested with approximately 600 consumers to date, and the team is collaborating with the Harvard School of Public Health to validate ONQI scores with the food frequency questionnaires and health outcomes of individuals involved in the Nurses Health Study and Health Professionals Follow-Up Study. The ONQI is expected to debut in August in roughly 1,000 stores across the United States and online.
The Proof is in the Pudding…
Nutrition profiling scores have the potential to vastly improve consumer food choices and even influence food policy and manufacturing, but their validity in the marketplace still needs to be tested. Issues to examine include the following:
• Will nutrient-density scores actually improve diet quality, and will this improvement occur across all cultural and income groups?
• Not all cultures within U.S. borders eat the same foods, so will nutrition profiling be applicable to all diets?
• Lower calorie and nutrient-dense foods often cost more than nutritionally inferior products, so will cost be prohibitive?
Because each current nutrition profiling system utilizes different algorithms that place value on a wide variety of nutrients, it will be essential to standardize a system that can be used in all markets nationwide and, ultimately, globally. Creating a global system presents its own challenges, as nutrient availability varies across countries and cultures. For example, protein is ubiquitous in most American diets but is not in some areas of Africa or Asia.
Complaints about nutrition profiling by the food industry are an interesting consideration. Food manufacturers have expressed concern that making comparisons across food groups will essentially label foods as being good or bad. Directly comparing the nutrition quality of carrots with potato chips, for example, will obviously suggest that chips are a nutrient-poor, energy-dense food and carrots are the healthier choice.
But making such value judgments does not seem to be the purpose of nutrition profiling. Consumers likely know that carrots are healthier than chips, yet chips are often the snack of choice. The goal of the 2005 Dietary Guidelines for Americans is to encourage individuals to choose more nutrient-dense foods and watch their calorie intake, not to avoid energy-dense and nutrient-poor foods entirely. If nutrition profiling can direct consumers more frequently to the produce section and less often to the snack aisle, it certainly will have served its purpose. And food manufacturers should fear not: There will always be a market for chips.
The ultimate test of nutrition profiling’s value will be its effect on the health of Americans and others who utilize it. If ratings on product labels or store shelves help consumers assess the true nutritional value and caloric density of foods, people may be inclined to purchase healthier foods. Diets may become more nutrient dense and less calorie dense. Food manufacturing and marketing may change for the better when nutrition-savvy consumers opt for better products. Nutrition profiling may also help focus attention on the quality of food available to people with restricted incomes or living in poor or inner-city neighborhoods. It is important, however, that any system adopted for nationwide use be standardized and evaluated for validity. Consumers do not need to be bombarded with more confusing nutrition information.
— Rita E. Carey, MS, RD, CDE, is a clinical dietitian and diabetes educator at Yavapai Regional Medical Center and the Pendleton Wellness Center in Prescott, Ariz.
References
1. McClellan MB. Speech before Harvard School of Public Health. July 1, 2003. Available at: http://www.fda.gov/oc/speeches/2003/harvard0701.html
2. Drewnowski A, Maillot M, Darmon N. Testing nutrient profile models in relation to energy density and energy cost. Eur J Clin Nutr. 2008; Epub ahead of print.
3. Drewnowski A, Fulgoni V III. Nutrient profiling of foods: Creating a nutrient-rich food index. Nutr Rev. 2008;66(1):23-39.
4. Drewnowski A. Concept of a nutritious food: Toward a nutrient density score. Am J Clin Nutr. 2005;82(4):721-732.
5. The Overall Nutrition Quality Index. Yale Griffin Prevention Research Center Scientific Conference. Available at: http://www.griffinhealth.org/Research/ONQI.aspx
Table 1
| Profiling System | Factors Rated |
| Guiding Stars | Protein, vitamins A and C, iron, calcium, fiber, total fat, saturated fat, trans fat, cholesterol, sodium, added sugar, whole grains |
| NNR | Protein, fiber, monounsaturated fat, thiamin, riboflavin, folate, calcium, iron, zinc, potassium, vitamins A, B12, C, D, and E |
| NRF | Protein, fiber, monounsaturated fat, thiamin, riboflavin, folate, calcium, iron, zinc, potassium, fat, sodium, added sugar, vitamins A, B12, C, D, and E |
| ONQI | Fiber, folate, potassium, calcium, zinc, omega-3 fatty acids, bioflavonoids, carotenoids, magnesium, iron, total fat, saturated fat, trans fat, sodium, added sugar, cholesterol, fat quality, protein quality, energy density, glycemic load, vitamins A, C, D, E, B12, and B6 |
— Sources: www.hannaford.com; www.onqi.com; Drewnowski A, Fulgoni V III. Nutrient profiling of foods: Creating a nutrient-rich food index. Nutr Rev. 2008;66(1):23-39.