Policy Research for Front of Package Nutrition Labeling: Developing and Testing a Summary System Algorithm. 4.4.3 Final Algorithm Development and Testing Using the MAXR Approach

05/01/2011

The beta coefficients from the eight-nutrient/component models per 100 kcal (Table 4-3) and per RACC (Table 4-5) were used to create algorithms to score foods. The algorithms are depicted in Figure 4-6.


Figure 4-6. Final Algorithms Using the MAXR Approach

NDSKCAL [(1.40 * Protein g per 100 kcal/50 g) + (3.13 * Fiber g per 100 kcal/25 g) + (1.00 * Calcium mg per 100 kcal/1,000 mg + (2.51 * Unsaturated fat g per 100 kcal/44 g) + (0.37 * Vitamin C mg per 100 kcal/60 mg)
− (2.95 * Saturated fat g per 100 kcal/20 g) - (0.52 * Added sugars g per 100 kcal/50 g) - (1.34 * Sodium mg per 100 kcal/2,400 mg) ] * 100
NDSRACC [(1.19 * Protein g per RACC/50 g) + (3.45 * Fiber g per RACC/25 g) + (0.95 * Calcium mg per RACC/1,000 mg + (1.98 * Unsaturated fat g per RACC/44 g) + (0.39 * Vitamin C mg per RACC/60 mg)
− (2.99 * Saturated fat g per RACC/20 g) - (0.64 * Added sugars g per RACC/50 g) − (1.45 * Sodium mg per RACC/2,400 mg) ] * 100
Positive nutrients were capped at 100% of recommended intake.

Methods for Subgroup Analyses for Final Algorithms

These algorithms are inherently different from the previous algorithms tested in Section 4.3 because different weights are applied to each nutrient and the values for each nutrient are summed rather than taking an average of the positive and negative nutrients. The scores were calculated for all foods in the NHANES database, and scores for the 570 foods were examined by food groupings as described previously. The food scores were regressed on HEI scores for all individuals and for the following subpopulations:

  1. age groups (children 2 to 18 years, adults 19+ years, and adults 50+ years)
  2. ethnic groups (non-Hispanic white, non-Hispanic black, and Mexican American)
  3. income status (poverty income ratio ≤1.85, and >1.85)
  4. weight status (for adults, obese: body mass index [BMI] ≥ 30, overweight: BMI ≥ 25 and < 30, normal: BMI < 25; for children, obese: BMI-for-age z-score ≥ 95th percentile, overweight: BMI-for-age z-score ≥ 85th and < 95th percentile, normal: BMI-for-age z-score < 85th percentile)
  5. lipid status (normal LDL: <130 mg/dl, elevated LDL: ≥130 mg/dl)

Regression models were conducted as previously described in Section 4.1, using SUDAAN Version 10 to account for the complex sampling design of NHANES. Covariates included age (as a continuous variable), gender, and ethnicity. Models for subpopulations by age included covariates for gender and ethnicity, and models for ethnic groups included covariates for age and gender.

Methods for Categorical Ratings for Final Algorithms

The continuous nutrient density-based scores were converted into 3- and 5-point scales or ratings using tertiles of scores for the 3-point rating and quintiles for the 5-point rating. To create a normal distribution of scores, the scores were standardized within each major USDA food grouping using the following equation:

standardized score within food group = (NDS food score - median NDS score
within food group)/interquartile range of NDS scores within food group

and across all food groups using this equation:

standardized score across food groups = (NDS food score - median NDS score
of all foods)/interquartile range of NDS scores of all foods

Next, separate tertile cutoffs were identified for the within-food group and across all foods' standardized scores. Each standardized score was assigned a rating of 1 to 3 based on its corresponding tertile. Next, separate quintile cutoffs were identified for the within-food group and across all foods standardized scores, and each standardized score was assigned a rating of 1 to 5 based on its corresponding quintile.

The categorical scores are described and compared for each of the two final algorithms shown in Figure 4-6.

Performance of Final Algorithms in Ranking the Nutritional Quality of Foods and Food Groups

The mean and distribution of algorithm scores for all foods and major food groupings are presented in Figures 4-7 and 4-8 for NDSKCAL and NDSRACC, respectively. The absolute algorithm scores cannot be compared with previous algorithms tested in Section 4.3 because in these final algorithms different weights were applied to each nutrient (Figure 4-6), and the values for each nutrient were summed rather than taking an average of the positive and negative nutrients.


Figure 4-7. Box Plot of Nutrient Density Scores per 100 Kcal for Final Algorithm (NDSKCAL)

Box plot shows the mean and distribution of nutrient density scores for all 570 foods and for major food groups. The mean is shown by the diamond; the left side of the boxes represents the 25th percentile of scores, and the right side of the boxes represents the 75th percentile. The minimum and maximum scores are represented at the ends of the horizontal lines.



Figure 4-8. Box Plot of Nutrient Density Scores per RACC for Final Algorithm (NDSRACC)

Box plot shows the mean and distribution of nutrient density scores for all 570 foods and for major food groups. The mean is shown by the diamond; the left side of the boxes represents the 25th percentile of scores, and the right side of the boxes represents the 75th percentile. The minimum and maximum scores are represented at the ends of the horizontal lines.

The overall mean score of the 570 foods was slightly higher, and the range of scores was wider, on a per 100 kcal unit basis than on a per RACC basis. The highest ranking mean score for a major food group was fruit on a per 100 kcal basis and legumes on a per RACC basis. The dairy mean score ranked eighth on both unit bases. Fats and oils had the lowest mean score ranking on a per 100 kcal basis, while the sweets and beverages mean score ranked lowest on a per RACC basis.

Mean scores of subgroupings of foods and scores of selected foods are presented in Appendix H. The top 10 highest scoring foods were all vegetables on a per 100 kcal basis, mostly raw and leafy green vegetables. Among the top 10 highest ranking foods on a per RACC basis were avocado, Fiber One bar, calcium-fortified orange juice, almonds, pinto beans, raw oranges, and strawberries.

Among the dairy products, nonfat yogurts and milk scored highest, and ice cream and cheeses scored lowest. Among legumes and nuts, the scores per RACC were higher for nuts because of the larger portion size of a RACC (1 ounce) compared with a 100 kcal portion. Among the highest scoring foods in the grain group were fortified bars and cereals, and popcorn. The range of scores for grains was much larger on a per RACC basis than on a per 100 kcal basis. Cakes and pastries had much lower scores on a per RACC basis than on a per 100 kcal basis because of the larger portion sizes and, therefore, higher fat and sugar content. Avocados ranked much higher than other fruits on a per RACC basis, with a score of 175 compared with a score of 98 for the next highest scoring fruit (calcium-fortified orange juice). The mean score of vegetables was much higher on a per 100 kcal basis than on a per RACC basis, because of the exaggerated scores of low-calorie leafy green vegetables.

Performance of Final Algorithms in Predicting HEI Scores for All Individuals and Subpopulations

The performance of the final algorithms in predicting overall dietary quality was assessed by calculating composite algorithm scores for 1-day food intakes in the NHANES sample and regressing composite algorithm scores on HEI scores for all individuals and for subpopulations. Results of regression models are presented in Table 4-7. Overall, the models explained approximately two-thirds of the variance in HEI scores, compared with the best of the previous models (Section 4.3.1) that explained approximately one-half of the variance in the HEI. The model with the final algorithm scored on a per 100 kcal basis had a higher R2 (64.76%) than the per RACC algorithm (60.42%). We would expect similar R2s from these regressions of HEI scores on food scores using the final algorithms as the R2 from the eight-nutrient model in Tables 4-3 and -5, because they include the same nutrients and beta-coefficients. In addition, we can examine how well the algorithm works for various subpopulations.


Table 4-7. Weighted Mean Scores of Foods Consumed by Individuals in NHANES 2005-2008 and Linear Regression Models on Healthy Eating Index Using Final Algorithmsa
Per 100 kcal (NDSKCAL) n Weighted
Mean
Scores (SE)
Beta
Coefficient
(SE)
p-value R2
- Means no data.
a Regression models were adjusted for complex sampling of NHANES and included covariates for age, gender, and ethnicity.
b Regression models included covariates for gender and ethnicity.
c Regression models included covariates for age and gender.
Overall 16,587 7.40 (0.29) 1.00 (0.02) <0.0001 64.76%
Age groups b - - - - -
Children (2-18 y) 6,706 5.08 (0.23) 1.12 (0.02) <0.0001 64.48%
Adults (19+ y) 9,881 8.15 (0.35) 0.96 (0.02) <0.0001 65.54%
Older adults (50+ y) 4,792 10.58 (0.33) 0.96 (0.02) <0.0001 66.81%
Ethnicity c - - - - -
Non-Hispanic white 6,700 6.99 (0.37) 1.00 (0.02) <0.0001 66.07%
Non-Hispanic black 4,110 6.31 (0.33) 0.99 (0.02) <0.0001 60.05%
Mexican Americans 3,781 9.89 (0.30) 0.99 (0.03) <0.0001 60.11%
Poverty income ratio - - - - -
≤1.85 7,353 6.83 (0.49) 1.00 (0.01) <0.0001 64.63%
>1.85 8,175 7.53 (0.29) 1.00 (0.02) <0.0001 64.87%
Child BMI-for-age status - - - - -
Obese (≥95th pct) 1,262 4.87 (0.50) 1.02 (0.06) <0.0001 65.00%
Overweight (≥85th and <95th pct) 996 4.76 (0.63) 1.11 (0.04) <0.0001 69.17%
Normal weight (<85th pct) 4,448 5.20 (0.26) 1.15 (0.02) <0.0001 63.44%
(continued)



Table 4-7. Weighted Mean Scores of Foods Consumed by Individuals in NHANES 2005-2008 and Linear Regression Models on Healthy Eating Index Using Final Algorithmsa (continued)
Per RACC (NDSRACC) n Weighted
Mean
Scores (SE)
Beta
Coefficient
(SE)
p-value R2
- Means no data.
a Regression models were adjusted for complex sampling of NHANES and included covariates for age, gender, and ethnicity.
b Regression models included covariates for gender and ethnicity.
c Regression models included covariates for age and gender.
Overall 16,587 1.16 (0.29) 1.00 (0.01) <0.0001 60.42%
Adult BMI status - - - - -
Obese (≥30) 3,457 7.58 (0.35) 0.97 (0.03) <0.0001 64.19%
Overweight (≥25 and <30) 3,316 8.72 (0.41) 0.95 (0.03) <0.0001 65.05%
Normal weight (<25) 2,965 8.03 (0.46) 0.98 (0.03) <0.0001 67.70%
LDL status - - - - -
Normal (<130 mg/dl) 4,132 7.13 (0.34) 0.95 (0.04) <0.0001 64.10%
Elevated (≥130 mg/dl) 1,450 7.38 (0.60) 0.95 (0.03) <0.0001 65.10%
Age groups b - - - - -
Children (2-18 y) 6,706 −0.67 (0.28) 0.91 (0.02) <0.0001 62.73%
Adults (19+ y) 9,881 1.76 (0.32) 1.02 (0.02) <0.0001 60.22%
Older adults (50+ y) 4,792 4.01 (0.27) 1.11 (0.02) <0.0001 63.03%
Ethnicity c - - - - -
Non-Hispanic white 6,700 0.66 (0.35) 1.04 (0.02) <0.0001 61.43%
Non-Hispanic black 4,110 −0.04 (0.36) 0.92 (0.03) <0.0001 57.82%
Mexican Americans 3,781 4.26 (0.30) 0.91 (0.02) <0.0001 57.08%
Poverty income ratio - - - - -
≤1.85 7,353 0.77 (0.52) 0.95 (0.03) <0.0001 59.78%
>1.85 8,175 1.20 (0.25) 1.03 (0.02) <0.0001 60.77%
Child BMI-for-age status - - - - -
Obese (≥95th pct) 1,262 −0.99 (0.60) 0.87 (0.04) <0.0001 62.62%
Overweight (≥85th and <95th pct) 996 −1.37 (0.68) 0.92 (0.06) <0.0001 66.11%
Normal weight (<85th pct) 4,448 −0.44 (0.33) 0.91 (0.02) <0.0001 62.02%
Adult BMI status - - - - -
Obese (≥30) 3,457 1.02 (0.29) 1.08 (0.03) <0.0001 61.01%
Overweight (≥25 and <30) 3,316 2.27 (0.37) 1.04 (0.03) <0.0001 61.10%
Normal weight (<25) 2,965 1.77 (0.40) 1.00 (0.03) <0.0001 60.58%
LDL status - - - - -
Normal (<130 mg/dl) 4,132 0.78 (0.33) 1.01 (0.03) <0.0001 60.76%
Elevated (≥130 mg/dl) 1,450 0.90 (0.54) 1.00 (0.03) <0.0001 60.71%

As shown in Table 4-7, the algorithms explained more of the variance in HEI scores for some subpopulations than others. The R2 of the model with older adults was higher than models with all adults or children only. Older adults also had more nutrient-dense diets; the weighted mean score was higher than the other age groups. Mexican Americans had a higher weighted mean score than the other ethnic groups but that model had a lower R2 than the model with non-Hispanic whites. A more nutrient-dense diet as measured by weighted mean score does not equate to a higher R2 because the R2 is driven by better agreement between algorithm scores and HEI scores across all levels of the HEI. This point is illustrated in Appendix I, which contains the mean scores by deciles of HEI. Although Mexican Americans had higher HEI scores and higher mean algorithm scores than whites, Mexican Americans had lower algorithm scores per 100 kcal than whites in both the lowest and highest deciles of the HEI. Models with overweight children had higher R2s than models with normal or obese children. There was little difference in R2s in models by poverty status or LDL status.

Results of Categorical Ratings Based on Final Algorithms

Category ratings within or across food groups for the 4,059 foods in NHANES 2007-2008 were assigned to foods according to tertiles or quintiles of standardized scores of final algorithms, NDSKCAL and NDSRACC. The ranges of the standardized scores corresponding to each categorical rating are provided in Tables 4-8 (for the three-category rating scheme) and 4-9 (for the five-category rating scheme). By definition, the median value of the standardized scores is zero, and each category (tertile or quintile) has approximately the same number of food items.


Table 4-8. Range of Standardized Nutrient Density Scores for Three-Category Rating Scheme for 4,097 Foods in NHANES 2007-2008
Standardized Scores Tertile 1 Tertile 2 Tertile 3
Category ratings within and across food groups were assigned to foods according to tertiles of standardized scores on final algorithms, NDSKCAL or NDSRACC.
Category Rating 1 2 3
Per 100 kcal (NDSKCAL)
Within food groups −37.55 to −0.34 −0.35 to 0.27 0.28 to 15.52
Across food groups −30.02 to −0.21 −0.22 to 0.35 0.35 to 7.94
Per RACC (NDSRACC)
Within food groups −65.63 to −0.33 −0.34 to 0.10 0.11 to 11.32
Across food groups −17.92 to −0.17 −0.18 to 0.35 0.36 to 7.84



Table 4-9. Range of Standardized Nutrient Density Scores for Five-Category Rating Scheme for 4,097 Foods in NHANES 2007-2008
Standardized Scores Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5
Category ratings within and across food groups were assigned to foods according to tertiles of standardized scores on final algorithms, NDSKCAL or NDSRACC.
Category Rating 1 2 3 4 5
Per 100 kcal (NDSKCAL)
Within food groups −37.55 to −0.68 −0.69 to −0.16 −0.17 to 0.08 0.09 to 0.62 0.63 to 15.52
Across food groups −30.02 to −0.62 −0.63 to −0.05 −0.06 to 0.19 0.20 to 0.78 0.79 to 7.94
Per RACC (NDSRACC)
Within food groups −65.63 to −0.81 −0.82 to −0.14 −0.15 to 0.05 0.06 to 0.34 0.35 to 11.32
Across food groups −17.92 to −0.54 −0.55 to −0.02 −0.03 to 0.12 0.13 to 0.78 0.79 to 7.84

The category ratings of a selection of 100 foods based on per 100 kcal standardized NDSs are presented in Table 4-10 for three- and five-category ratings within and across food groups. These 100 foods were selected to illustrate a variety of foods, with some examples of foods that rated differently across or within food groups. Foods in Table 4-10 are grouped by major food groupings and sorted by descending NDSs, and the three- and five-category ratings appear from highest (3 or 5) to lowest (1). Categories shown in Table 4-10 could be considered, for example, for an FOP system with an overall "low, medium, high" rating scheme (three-category rating) or for an FOP system with a 5-point rating scheme (five-category rating). The results for selected foods in Table 4-10 illustrate some characteristics of category rating systems based on the final NDS algorithm, NDSKCAL.


Table 4-10. Categorical Ratings of 100 Foods Based on the per 100 Kcal Standardized Nutrient Density-Based Scores, for Three- and Five-Category Ratings Within and Across Food Groups
Food Code Food Description Nutrient
Density
Score
Three-
Category
Ratings
Within
Food
Groups
Three-
Category
Ratings
Across
Foods
Five-
Category
Ratings
Within
Food
Groups
Five-
Category
Ratings
Across
Foods
Foods are sorted by major food groups and descending nutrient density-based scores. The three- and five-category ratings appear from highest (3 or 5) to lowest (1).
DAIRY FOODS
11433500 Yogurt, fruit variety, nonfat milk, sweetened with low-calorie sweetener 67.71 3 3 5 5
11411300 Yogurt, plain, nonfat milk 54.99 3 3 5 5
11113000 Milk, cow's, fluid, skim or nonfat, 0.5% or less butterfat 54.73 3 3 5 5
11112210 Milk, cow's, fluid, 1% fat 28.39 3 3 5 4
11411200 Yogurt, plain, low-fat milk 27.80 3 3 5 4
11511400 Milk, chocolate, low-fat milk-based 19.07 3 2 4 4
14204010 Cheese, cottage, low-fat (1-2% fat) 14.51 3 2 4 3
11432500 Yogurt, fruit variety, low-fat milk, sweetened with low-calorie sweetener 14.31 3 2 4 3
11433000 Yogurt, fruit variety, nonfat milk 11.67 3 2 4 3
11112110 Milk, cow's, fluid, 2% fat 7.62 3 2 4 3
11511200 Milk, chocolate, reduced fat milk-based, 2% (formerly "low-fat") 7.00 3 2 4 3
12210200 Cream substitute, liquid 1.55 2 2 4 2
11432000 Yogurt, fruit variety, low-fat milk 1.44 2 2 4 2
11511100 Milk, chocolate, whole milk-based −3.10 2 1 4 2
14107030 Cheese, Mozzarella, part skim −4.49 2 1 3 2
11111000 Milk, cow's, fluid, whole −6.63 2 1 3 2
11411100 Yogurt, plain, whole milk −9.41 2 1 3 2
13160410 Fat-free ice cream, chocolate −10.31 2 1 3 2
11431000 Yogurt, fruit variety, whole milk −11.93 2 1 3 2
13130310 Light ice cream, chocolate (formerly ice milk) −18.71 2 1 2 1
13110110 Ice cream, regular, chocolate −32.38 1 1 2 1
14104010 Cheese, natural, Cheddar or American type −35.98 1 1 1 1
13110130 Ice cream, rich, chocolate −39.47 1 1 1 1
MEATS, POULTRY, AND FISH
24122120 Chicken, breast, roasted, broiled, or baked, skin not eaten 37.87 3 3 5 4
26137120 Salmon, baked or broiled 34.10 3 3 5 4
21401120 Beef, roast, roasted, lean only eaten 21.31 3 2 4 4
27111410 Chili con carne with beans 20.02 3 2 4 4
28141250 Chicken with rice-vegetable mixture (diet frozen meal) 19.59 3 2 4 4
22101120 Pork chop, broiled or baked, lean only eaten 18.07 3 2 4 4
21501000 Ground beef, less than 80% lean, cooked (formerly regular) −0.94 2 1 2 2
25210110 Frankfurter, wiener, or hot dog, NFS −26.73 1 1 1 1
EGGS
31103000 Egg, whole, boiled 5.04 1 2 2 2
LEGUMES AND NUTS
42100100 Almonds, NFS 74.53 3 3 5 5
41104020 Pinto, calico, or red Mexican beans, dry, cooked, fat not added in cooking 66.28 3 3 5 5
41208030 Pork and beans 44.07 2 3 3 4
42110000 Mixed nuts, NFS 43.05 2 3 2 4
42202150 Peanut butter, reduced fat 31.91 1 3 2 4
42202000 Peanut butter 31.32 1 3 2 4
GRAIN PRODUCTS
53540300 Fiber One Chewy Bar 80.61 3 3 5 5
53544450 PowerBar (fortified high energy bar) 50.30 3 3 5 5
57123000 Cheerios 49.47 3 3 5 5
51601020 Bread, multigrain 41.34 3 3 5 4
56203010 Oatmeal, cooked, regular, fat not added in cooking 35.46 3 3 5 4
54403000 Popcorn, popped in oil, unbuttered 25.90 3 3 5 4
51301010 Bread, wheat or cracked wheat 21.94 3 3 4 4
53542210 Granola bar, nonfat 21.27 3 2 4 4
53540400 Kellogg's Nutri-Grain Cereal Bar 20.27 3 2 4 4
58301110 Vegetable lasagna (frozen meal) 16.89 3 2 4 4
55201000 Waffle, plain 16.44 3 2 4 4
51101000 Bread, white 9.49 2 2 3 3
53234000 Cookie, peanut butter 7.15 2 2 2 3
57135000 Corn flakes, Kellogg's 3.87 1 2 2 2
58130011 Lasagna with meat 2.90 1 2 2 2
58304200 Ravioli, cheese-filled, with tomato sauce (diet frozen meal) 1.79 1 2 2 2
54101010 Cracker, animal 1.15 1 2 2 2
53204010 Cookie, brownie, without icing 0.78 1 2 2 2
58106225 Pizza, cheese, regular crust 0.07 1 2 2 2
54408010 Pretzels, hard −0.41 1 2 2 2
58145110 Macaroni or noodles with cheese −1.17 1 1 2 2
53121260 Cake, yellow, with icing, made from home recipe or purchased ready-to-eat −6.86 1 1 1 2
FRUIT
61119010 Orange, raw 115.15 3 3 5 5
61210250 Orange juice, with calcium added, canned, bottled or in a carton 91.76 3 3 5 5
63105010 Avocado, raw 82.10 3 3 5 5
63101000 Apple, raw 65.53 3 3 4 5
63143010 Plum, raw 58.07 2 3 4 5
61210620 Orange juice, frozen (reconstituted with water) 48.81 2 3 3 5
63107010 Banana, raw 45.35 2 3 2 5
63149010 Watermelon, raw 42.05 1 3 2 4
63101120 Applesauce, stewed apples, unsweetened 36.01 1 3 2 4
63123000 Grapes, raw, NS as to type 29.49 1 3 1 4
63135150 Peach, cooked or canned, drained solids 22.76 1 3 1 4
64104010 Apple juice 20.97 1 2 1 4
62125100 Raisins 20.19 1 2 1 4
VEGETABLES
72125100 Spinach, raw 215.74 3 3 5 5
72201100 Broccoli, raw 163.50 3 3 5 5
71201010 White potato, chips 33.46 2 3 3 4
74303000 Tomato and vegetable juice, mostly tomato 26.00 2 3 3 4
71401030 White potato, French fries, from frozen, deep fried 18.85 2 2 3 4
71101000 White potato, baked, peel not eaten 17.85 2 2 3 4
75510000 Olives, NFS 17.23 2 2 3 4
FATS, OILS, AND DRESSINGS
82104000 Olive oil 30.83 3 3 5 4
83107000 Mayonnaise, regular 20.85 3 2 4 4
81102000 Margarine, NFS 12.83 2 2 4 3
83104000 French dressing 6.50 2 2 3 3
83202020 French dressing, reduced calorie −22.12 1 1 1 1
81101000 Butter, stick, salted −90.35 1 1 1 1
83205500 Italian dressing, reduced calorie, fat-free −115.13 1 1 1 1
SWEETS AND BEVERAGES
92101000 Coffee, made from ground, regular 48.61 3 3 5 5
92550030 Fruit juice drink, low calorie, with high vitamin C 47.98 3 3 5 5
92306000 Tea, herbal 14.89 2 2 4 3
92410320 Soft drink, cola-type, sugar-free 8.07 2 2 3 3
91201010 Sugar substitute, aspartame-based, dry powder 1.66 2 2 3 2
92541010 Fruit flavored drink, made from powdered mix −3.01 2 1 2 2
91401000 Jelly, all flavors −13.25 1 1 2 2
91405500 Jelly, reduced sugar, all flavors −14.83 1 1 2 2
91745020 Hard candy −17.08 1 1 2 1
92410310 Soft drink, cola-type −24.77 1 1 2 1
91101010 Sugar, white, granulated or lump −26.86 1 1 1 1
92560100 Gatorade Thirst Quencher sports drink −28.56 1 1 1 1
91705010 Milk chocolate candy, plain −35.79 1 1 1 1
  • Among dairy foods, nonfat milk and yogurts were rated highest both across all foods and within the group of dairy foods. Low-fat milk and yogurt were also rated highest across all foods and within dairy foods with the three-category scheme, but using the five-category scheme they were rated highest (5) within the group and next highest (4) across foods. Whole milk was rated intermediate (2) within the group and lowest (1) across foods with the three-category scheme, and intermediate (3) within dairy foods and next lowest (2) across all foods using five categories.
  • Among the meats, poultry and fish group, baked chicken and salmon ranked highest within and across foods using three categories and highest (5) within the group and next highest (4) across foods using five categories.
  • Among legumes and nuts, almonds scored in the highest category both within the group and across foods using either three or five categories. Peanut butter had a score of lowest (1) within the legume and nut group despite scoring in the highest category (3) across all foods using three categories.
  • Grain products contain many varied foods. Foods that scored high both across foods and within the food group include fortified energy bars, Cheerios, oatmeal, and whole wheat bread. Some of the lowest scoring grains within the grain group with the three-category scheme scored intermediate (2) across foods (e.g., animal crackers, cheese pizza, and pretzels); with the five-category scheme, these foods scored next lowest (2) both within the group and across foods.
  • Fruits that scored highest across foods using three categories but lowest within the fruit group included applesauce, grapes, canned peaches, and watermelon. Grapes and peaches also scored lowest (1) within the fruit group using five categories and scored next highest (4) across foods.

Categorical ratings of the same 100 selected foods based on the final algorithm per RACC, NDSRACC, are presented in Table 4-11. There were some differences in categorical scores compared with the per 100 kcal rankings (Table 4-10), attributable to the differences in algorithm scores with the different unit bases; however, the differences were small, usually one category difference. Standardization of scores normalizes the distribution of scores. The normalization probably contributes to more similar categorical ratings per 100 kcal and per RACC than if the absolute scores were categorized without standardization and compared between the two unit bases.


Table 4-11. Categorical Ratings of 100 Foods Based on the per RACC Standardized Nutrient Density-Based Scores for Three- and Five-Category Ratings Within and Across Food Groups
Food Code Food Description Nutrient
Density
Score
Three-
Category
Ratings
Within
Food
Groups
Three-
Category
Ratings
Across
Foods
Five-
Category
Ratings
Within
Food
Groups
Five-
Category
Ratings
Across
Foods
Foods are sorted by major food groups and descending nutrient density-based scores. The three- and five-category ratings appear from highest (3 or 5) to lowest (1).
DAIRY FOODS
11433500 Yogurt, fruit variety, nonfat milk, sweetened with low-calorie sweetener 64.61 3 3 5 5
11411300 Yogurt, plain, nonfat milk 60.71 3 3 5 5
11113000 Milk, cow's, fluid, skim or nonfat, 0.5% or less butterfat 40.05 3 3 5 5
11411200 Yogurt, plain, low-fat milk 30.01 3 3 5 5
11432500 Yogurt, fruit variety, low-fat milk, sweetened with low-calorie sweetener 25.40 3 3 5 4
11112210 Milk, cow's, fluid, 1% fat 22.39 3 3 5 4
11511400 Milk, chocolate, low-fat milk-based 22.02 3 3 5 4
11433000 Yogurt, fruit variety, nonfat milk 11.94 3 2 5 4
11511200 Milk, chocolate, reduced fat milk-based, 2% (formerly "low-fat") 5.31 3 2 5 4
14204010 Cheese, cottage, low-fat (1-2% fat) 2.82 3 2 4 3
11112110 Milk, cow's, fluid, 2% fat 1.47 3 2 4 3
12210200 Cream substitute, liquid −1.65 3 2 4 2
14107030 Cheese, Mozzarella, part skim −11.75 2 1 3 2
11432000 Yogurt, fruit variety, low-fat milk −11.93 2 1 3 2
11511100 Milk, chocolate, whole milk-based −16.62 2 1 3 2
11111000 Milk, cow's, fluid, whole −18.83 2 1 2 2
13160410 Fat-free ice cream, chocolate −19.43 2 1 2 1
11411100 Yogurt, plain, whole milk −21.61 2 1 2 1
13130310 Light ice cream, chocolate (formerly ice milk) −29.77 1 1 2 1
11431000 Yogurt, fruit variety, whole milk −49.36 1 1 1 1
14104010 Cheese, natural, Cheddar or American type −53.58 1 1 1 1
13110110 Ice cream, regular, chocolate −54.48 1 1 1 1
13110130 Ice cream, rich, chocolate −84.67 1 1 1 1
MEATS, POULTRY, AND FISH
24122120 Chicken, breast, roasted, broiled, or baked, skin not eaten 38.00 3 3 5 5
27111410 Chili con carne with beans 37.80 3 3 5 5
26137120 Salmon, baked or broiled 33.16 3 3 5 5
28141250 Chicken with rice-vegetable mixture (diet frozen meal) 28.67 3 3 5 4
21401120 Beef, roast, roasted, lean only eaten 20.17 3 3 5 4
22101120 Pork chop, broiled or baked, lean only eaten 13.19 3 2 4 4
21501000 Ground beef, less than 80% lean, cooked (formerly regular) −23.43 1 1 2 1
25210110 Frankfurter, wiener, or hot dog, NFS −62.82 1 1 1 1
EGGS
31103000 Egg, whole, boiled −3.13 3 2 5 2
LEGUMES AND NUTS
42100100 Almonds, NFS 114.43 3 3 5 5
41104020 Pinto, calico, or red Mexican beans, dry, cooked, fat not added in cooking 70.54 3 3 4 5
42110000 Mixed nuts, NFS 63.32 2 3 3 5
41208030 Pork and beans 52.82 2 3 2 5
42202150 Peanut butter, reduced fat 45.48 1 3 2 5
42202000 Peanut butter 41.54 1 3 2 5
GRAIN PRODUCTS
53540300 Fiber One Chewy Bar 118.62 3 3 5 5
53544450 PowerBar (fortified high energy bar) 71.03 3 3 5 5
57123000 Cheerios 53.93 3 3 5 5
51601020 Bread, multigrain 52.70 3 3 5 5
56203010 Oatmeal, cooked, regular, fat not added in cooking 49.86 3 3 5 5
58301110 Vegetable lasagna (frozen meal) 33.95 3 3 5 5
54403000 Popcorn, popped in oil, unbuttered 33.70 3 3 5 5
55201000 Waffle, plain 31.56 3 3 5 5
53542210 Granola bar, nonfat 28.14 3 3 5 4
51301010 Bread, wheat or cracked wheat 25.78 3 3 5 4
53540400 Kellogg's Nutri-Grain Cereal Bar 24.33 3 3 5 4
51101000 Bread, white 8.82 2 2 3 4
57135000 Corn flakes, Kellogg's 2.76 2 2 2 3
53234000 Cookie, peanut butter 0.17 2 2 2 3
54408010 Pretzels, hard −3.38 1 2 2 2
54101010 Cracker, animal −4.12 1 2 2 2
58130011 Lasagna with meat −4.89 1 2 2 2
53204010 Cookie, brownie, without icing −7.18 1 1 2 2
58304200 Ravioli, cheese-filled, with tomato sauce (diet frozen meal) −8.99 1 1 2 2
58106225 Pizza, cheese, regular crust −17.70 1 1 2 2
58145110 Macaroni or noodles with cheese −31.79 1 1 1 1
53121260 Cake, yellow, with icing, made from home recipe or purchased ready-to-eat −38.51 1 1 1 1
FRUIT
63105010 Avocado, raw 174.89 3 3 5 5
61210250 Orange juice, with calcium added, canned, bottled or in a carton 98.42 3 3 5 5
61119010 Orange, raw 93.81 3 3 5 5
63107010 Banana, raw 60.68 3 3 4 5
63101000 Apple, raw 51.91 2 3 3 5
61210620 Orange juice, frozen (reconstituted with water) 51.43 2 3 3 5
63143010 Plum, raw 39.59 1 3 2 5
63149010 Watermelon, raw 36.38 1 3 2 5
63123000 Grapes, raw, NS as to type 29.99 1 3 2 5
62125100 Raisins 25.39 1 3 1 4
64104010 Apple juice 24.20 1 3 1 4
63135150 Peach, cooked or canned, drained solids 23.85 1 3 1 4
63101120 Applesauce, stewed apples, unsweetened 23.02 1 3 1 4
VEGETABLES
72201100 Broccoli, raw 77.00 3 3 5 5
72125100 Spinach, raw 50.92 3 3 5 5
71201010 White potato, chips 42.89 3 3 5 5
74303000 Tomato and vegetable juice, mostly tomato 32.55 3 3 4 5
71401030 White potato, French fries, from frozen, deep fried 32.16 3 3 4 5
71101000 White potato, baked, peel not eaten 18.20 2 3 2 4
75510000 Olives, NFS 1.04 1 2 2 3
FATS, OILS, AND DRESSINGS
82104000 Olive oil 22.85 3 3 5 4
83107000 Mayonnaise, regular 10.45 3 2 5 4
81102000 Margarine, NFS 1.05 3 2 4 3
83104000 French dressing −4.38 2 2 3 2
83205500 Italian dressing, reduced calorie, fat-free −18.07 1 1 2 2
83202020 French dressing, reduced calorie −20.25 1 1 2 1
81101000 Butter, stick, salted −98.00 1 1 1 1
SWEETS AND BEVERAGES
92550030 Fruit juice drink, low calorie, with high vitamin C 43.95 3 3 5 5
92101000 Coffee, made from ground, regular 0.95 2 2 4 3
92306000 Tea, herbal 0.30 2 2 3 3
92410320 Soft drink, cola-type, sugar-free 0.15 2 2 3 3
91201010 Sugar substitute, aspartame-based, dry powder 0.05 2 2 3 3
91101010 Sugar, white, granulated or lump −5.12 2 2 2 2
91405500 Jelly, reduced sugar, all flavors −6.46 1 1 2 2
92541010 Fruit flavored drink, made from powdered mix −6.95 1 1 2 2
91401000 Jelly, all flavors −8.58 1 1 2 2
91745020 Hard candy −12.39 1 1 2 2
92560100 Gatorade Thirst Quencher sports drink −21.33 1 1 1 1
92410310 Soft drink, cola-type −27.31 1 1 1 1
91705010 Milk chocolate candy, plain −86.49 1 1 1 1