Whole foods are almost always the healthiest choices (Monteiro 2011; Moubarac 2013; Popkin 2012; Rauber 2014). Yet our busy lives mean that most of us sometimes turn to processed foods. We want to help consumers find less processed versions of their favorite foods. This turned out to be a challenging task, given that we have very limited information on how any particular ingredient is produced or any particular product is made.
Our intention was to develop a unique method of calculation, or algorithm, that could help guide consumers in their shopping decisions. EWG's processing score reflects our best estimate of the extent to which a particular food product has been processed. To our knowledge, this is the first time such an algorithm has been developed and applied on this scale.
The major factors that we considered in developing the processing concern algorithm include:
These factors are weighted based on the extent to which they are likely to alter the food's nature or nutritional content and then combined to determine a product's overall processing score. For the individual ingredient scores, both the source of the ingredient and the processes used to manufacture it are taken into consideration.
Our scoring system relies on in-depth analysis of processes used in food manufacturing, based on information in industry handbooks, chemical encyclopedias, patents, government and health agency reports and food manufacturing manuals.
The scores of individual ingredients range from 0.2 to 5. All synthetic chemicals (also described as industrial chemicals) were assigned a maximum processing score of 5. Both artificial and natural "flavors" are also given this maximum processing score since it is often unclear exactly they consist of and what kinds of preservatives and additives they contain.
For all ingredients other than synthetic industrial chemicals and "flavors," a score is calculated based on the processes used to prepare an ingredient and the source of the ingredient. The sources are assigned a weight of either 1, for ingredients derived from petroleum, or 0 for other ingredients. Each ingredient is scored according to this equation:
Ingredient score = 0.2 + source weight + sum (processing weights)
With the ingredient score formula used by EWG, each ingredient either has a minimum processing score of 0.2 or a higher score determined by the source and the processes used to obtain the ingredient. For four categories of ingredients, a specialized baseline score is assigned: 3 for non-artificial added sweeteners; 3 for chemically extracted vitamins and minerals; 0.7 for spices; and 0.5 for herbs. For ingredients in these three categories, the higher of the baseline score or the calculated score is used as the final score.
Individual ingredient scores are combined based on the listed order of ingredients on the Nutrition Facts panel of a given product. The U.S. Food and Drug Administration requires manufacturers to list ingredients so that the ingredient that weighs the most is listed first, and the ingredient that weighs the least is listed last (FDA 2013). EWG accounted for this in the equation by weighting the first two ingredients more heavily than the others:
Combined ingredient score = 5 X (1st ingredient score) + 3 X (2nd ingredient score) + sum (remaining ingredient scores)
The combined ingredient score is added to the product level score to obtain the overall processing score.
In instances where multiple ingredients in one product are processed the same way, EWG groups those ingredients and accounts for those processes only once in order to avoid inflating the combined ingredient score. For example, a multi-grain bread might contain five different grains that have likely been processed in similar fashion. In such a case, EWG's combined ingredient score for the grains would only reflect the processing of the most highly processed grain.
Products with multiple fruit or vegetable ingredients were scored similarly. EWG grouped the fruit and vegetable ingredients and accounted for their processing by taking the average score of each of the ingredients.
EWG also identified innocuous ingredients, such as water, that do not contribute to the score.
Other ingredient groupings are listed in Table 1.
Table 1. Group scoring method for various ingredient types
Group Scoring Method | Ingredient Types |
---|---|
No scoring contribution | Innocuous ingredients (water, carbonated water, etc.) |
Average of ingredient scores | Dairy other than from cows; fruit and vegetables (powders, dried, concentrates, purees and sauces, freeze-dried, whole); juice concentrates |
Highest scoring ingredient | Dairy (milk, cheese, cream) from cows; flavors (artificial and natural); grains; herbs and spices; juice; juice concentrates (used as added sugar); meat (beef, bison, chicken, goat, pork, sheep and lamb, turkey, veal, seafood, other non-poultry); vitamins and minerals |
The product scoring system emphasizes the manufacturing processes that typically occur at the product level. For example, the processing score for breads takes into account the processes that are used to manufacture the ingredients in the bread as well as the processes used to manufacture the finished product - in the case of bread, kneading and baking. To evaluate the extent of processing that could occur in each manufacturing step, EWG assigned a processing weight to each operation (Table 2). The processing operation weight was assigned based on the EWG research team's review of the research literature.
Product-level processing scores are calculated as follows:
Product-level score = sum (weighted processes)
The product-level score is combined with the ingredient score to determine the overall processing score for a product. A factor of 5 is added to the formula to highlight the importance of product-level processing.
Overall processing score = 5 X product-level score + combined ingredient-level score
Table 2. Extent of processing in different food manufacturing operations
Processing Weight Categories | Manufacturing and Processing Steps |
---|---|
0 | washed |
0.1 | sifted/sorted/separated; rolled; filtered or purified; chopped, sliced, diced, fileted or cut; sun-dried; dehydrated or evaporated; steeped in water; pH-adjusted |
0.3 | pressed or extracted; ground, milled or crushed; de-skinned or de-hulled; other mechanical processing; pasteurized; refrigerated or frozen; cooked, boiled or heated; enzymatically produced; fortified or enriched; curdled; fermented; added sweetener; emulsified |
0.5 | freeze-dried; distilled; crystallized or precipitated; cured; color added; brined or salted |
1 | extremely heated; chemically synthesized; hydrolyzed; hydrogenated; chemically extracted; bleached; added preservative; other chemical reaction |
Once the overall processing score is calculated, it is scaled from 1 (best) to 10 (worst), using the following equation:
Final processing score = (overall processing score - 1) / 10
The final processing score determines the processing concern level on the product page. Products with no identified processing concerns have a final processing score of 0 and are generally whole foods without additives. Products with low processing concerns have a final processing score between 0.1 and 1.9. They generally have few artificial ingredients, more ingredients that are either whole foods or are minimally processed and fewer ingredients overall. Products with moderate processing concerns score from 2.0 to 6.4, and those with high processing concerns score from 6.5 to 10. These products generally have more artificial ingredients, more ingredients that have been significantly modified from whole foods and more ingredients overall.
Example 1: Organic Girl Five Happiness Tasty Salad
Ingredients listed on the product label: BABY BOK CHOY LEAVES, BABY RED CHARD, TANGO, MIZUNA, BABY ARUGULA, BABY SPINACH, BABY GREEN CHARD, BABY GREEN ROMAINE, BABY GREEN OAKLEAF LETTUCE
Product level score = 0.0
Individual ingredient scores*:
- Baby bok choy leaves = 0.2 (minimum processing score)
- Baby red chard = 0.2 (minimum processing score)
- Tango = 0.2 (minimum processing score)
- Mizuna = 0.2 (minimum processing score)
- Baby arugula = 0.2 (minimum processing score)
- Baby spinach = 0.2 (minimum processing score)
- Baby green chard = 0.2 (minimum processing score)
- Baby green romaine = 0.2 (minimum processing score)
- Baby oakleaf lettuce = 0.2 (minimum processing score)
*All of these ingredients fall under the fruit & vegetable ingredient group and are therefore scored according to the average ingredient score.
Combined Ingredient Score = 5 X (0.2) = 1
Overall Processing Score = 5 X (0) + 1 = 1
Final Processing Score = (1 - 1) / 10 = 0 => No Concern Identified
Example 2: Coca-Cola
Ingredients listed on the product label: CARBONATED WATER, HIGH FRUCTOSE CORN SYRUP, CARAMEL COLOR, PHOSPHORIC ACID, NATURAL FLAVORS, CAFFEINE
Product Level Score = 0
Individual Ingredient Scores:
- Carbonated water* = N/A
- High fructose corn syrup = 3
- Caramel color = 3.2
- Phosphoric acid = 5
- Natural flavors = 5
- Caffeine = 2.5
*Carbonated water is an innocuous ingredient and does not contribute to the processing score. Therefore high fructose corn syrup is considered the first ingredient.
Combined Ingredient Score = 5 X (3) + 3 X (3.2) + 5 + 5+ 2.5 = 37.1
Overall Processing Score = 5 X (0) + 37.1 = 37.1
Final Processing Score = (37.1 - 1) / 10 = 3.6 => Moderate processing concerns
FDA. 2013. Guidance for Industry: A Food Labeling Guide (6. Ingredient Lists). http://www.fda.gov/food/guidanceregulation/guidancedocumentsregulatoryinformation/labelingnutrition/ucm064880.htm#descend. Accessed Aug. 20, 2014
Monteiro CA, Levy RB, Claro RM, de Castro IR, Cannon G. Increasing consumption of ultra-processed foods and likely impact on human health: evidence from Brazil. Public Health Nutr. 14(1): 5-13.
Moubarac JC, Marins AP, Claro RM, Levy RB, Cannon G, Monteiro CA. Consumption of ultra-processed foods and likely impact on human health. Evidence in Canada. Public Health Nutr. 16(12): 2240-8
Popkin BM, Adair LS, Ng SW. Global nutrition transition and the pandemic of obesity in developing countries. Nutr Rev. 70(1): 3-21.
Rauber F, Campagnolo PD, Hoffman DJ, Vitolo MR. Consumption of ultra-processed food products and its effects on children's lipid profiles: A longitudinal study. Nutr Metab Cardiovasc Dis. doi: 10.1016/j.numecd.2014.08.001. [Epub ahead of print]
Angell PM. 1960. Process of and apparatus for producing cracker meal and the like. Patent US2956519
Barbaroux O. 1990. Production, Properties, and Uses of Alginate, Carrageenan and Agar. Technical Resource Papers Regional Workshop on the Culture and Utilization of Seaweeds 2, 180
Bunke PR, Ekanayake A. 2012. Process for producing cassava flour. Patent WO2012003225 A1
Chahal SP, Starr JN. 2006. Lactic Acid. Ullmann's Encyclopedia of Industrial Chemistry, 219-226. Hoboken: John Wiley & Sons, Ltd.
Chao KC, McCarthy EF, McConaghy GA. 1980. Yeast Autolysis Process. Patent US4218481
Davenport AL, Staats LJ. 1998. Maple Syrup Production for the Beginner, 1-4
European Association for Specialty Yeast Products. 2011. Yeast Extract-Naturally Good. http://www.dsm.com/content/dam/dsm/foodandbeverages/en_US/documents/eurasyp-brochures.pdf. Accessed Nov. 8, 2013.
FAO. 1990. Training manual of Gracilaria culture and seaweed processing in China. FAO Corporate Document Repository, 107
FAO. 2003. A guide to the seaweed industry, 3. Agar. FAO Fisheries Technical Papers
FAO. 2011. Caramel colours. FAO JECFA Monographs 11
FAO: Animal Production and Health. 2012. Slaughtering Process. Accessed at http://www.fao.org/ag/againfo/themes/en/meat/slaugh_process.html, Dec. 6, 2013.
Fellows P. 1997. Guidelines for small-scale fruit and vegetable processors. FAO
Grace MR. 1977. Cassava processing. FAO Plant Production and Protection Series No. 3
Hui YH (Editor). 2006. Handbook of Food Science, Technology, and Engineering. Boca Raton: CRC Press, Taylor & Francis Group
Hui YH (Editor). 2007. Handbook of Food Products Manufacturing. Hoboken: John Wiley & Sons, Ltd.
Institute of Food Technologists. 2014. Making Peanut Butter. Accessed at http://www.ift.org/Knowledge-Center/Learn-About-Food-Science/Become-a-Food-Scientist/Introduction-to-the-Food-Industry/Lesson-2/Making-Peanut-Butter.aspx, Feb. 10, 2014
James WH. 1960. Process for production of buttermilk. Patent US2958601 A
Jiménez González AT. 1995. Milling process of durum wheat. Durum wheat quality in the Mediterranean region. International Centre for Advance Mediterranean Agronomic Studies, 43-51
Lusas EW, Rooney LW. 2001. Use of Dried Potatoes in Snack Foods, Snack Foods Processing. CRC Press, 257-258
Mazza G. 1993. Storage, Processing, and Quality Aspects of Buckwheat Seed. New crops, 251-254
Murray BC, Gross DH, Fox TJ. 1994. Starch Manufacturing: A Profile. http://www.epa.gov/ttnecas1/regdata/IPs/Starch%20Manufacturing_IP.pdf. Accessed Aug. 22, 2013
North American Millers' Association. 2011. Wheat Milling Process. http://www.namamillers.org/education/wheat-milling-process/. Accessed Nov. 14, 2013.
Office of Air Quality Planning and Standards, US EPA. 2010. Available and Emerging Technologies for Reducing Greenhouse Gas Emissions from the Pulp and Paper Manufacturing Industry. EPA: Office of Air and Radiation, 3
Pickford K. 2003. Breadcrumb Coating for Food Products. Patent US 6,613,370 B1
The Salt Institute. 2013. Salt 101: Production & Industry. Accessed at http://www.saltinstitute.org/salt-101/production-industry/, Oct. 17, 2013. Sprague WA. 1928. Carnauba Wax and the preparation thereof. Patent US 1658062 A
The Sugar Association. 2013. How We Get Sugar. http://www.sugar.org/how-we-get-sugar/. Accessed Oct. 21, 2013
Tan HZ, Li ZG, Tan B. 2009. Starch Noodles: History, Classification, Materials, Processing, Structure, Nutrition, Quality Evaluating, and Improving. Food Research International 42, 551-576
USDA. 1999. Safe Practices for Sausage Production, 1-20
USDA. 2011. Technical Evaluation Report on Tartaric Acid: Handling/Processing. Agricultural Marketing Service, 1-12
USDA. 2013. Bacon and Food Safety. Food Safety and Inspection Service, Food Safety Information
Verhoff F. 2011. Citric Acid. Ullmann's Encyclopedia of Industrial Chemistry 9, 197-202. Hoboken: John Wiley & Sons, Ltd.
Warne D. 1988. Manual of Fish Canning. FAO Fisheries Technical Paper 285 Xiaodang L. 2011. Pure chicken powder and production process thereof. Patent CN101999683 A
The Eat Well Guide helps consumers find locally grown and sustainably produced food. Listings include farms, restaurants, stores, farmers' markets, and CSAs throughout the United States.
Visit the Eat Well GuideSign up to receive email updates, action alerts, healthy eating tips, promotions to support our work and more from EWG! You can opt out at any time.