Nutrition Scoring: NuVal™ vs. ANDI – But What Info Is Really Needed?

Editor’s note:  This article precedes a series that Dr. Morss will be posting here on the economics of health care.


Over the years, we have been fed a continuous diet of ever changing and often incorrect information on what to eat. I quote from a recommendation of the British government made in 1938:

“The British people should drink 80% more milk, eat 35% more eggs, 40% more butter, and 30% more meat.”

Do you remember the cholesterol fear, and then the discovery of “good” cholesterol? And the food pyramids, initially heavy on meat and dairy products? This ever-changing advice reflects new knowledge, changing life styles, and the efforts of food industry lobbyists. In what follows, these changes will serve as the backdrop for an assessment of the latest players in the food info game: ANDI and NuVal™.


Both ANDI and NuVal™ provide nutritional food scores to supermarkets. ANDI scores about 500 foods; they appear in Whole Food stores. NuVal rates about 90,000 foods; its scores appear in the stores of 16 chains (Amigos, Big Y, Brookshires, Coborn’s, Festival, Food City, Hy-Vee, King Kullen, Mariano’s, Market Street, Meijer, Price Chopper, Scolari’s, Super 1 Foods, Tops, and United.

NuVal™ scores are generated by “the Overall Nutritional Quality Index (ONQI™), a patent-pending algorithm.”  NuVal™ LLC is a joint venture formed in 2008 by Topco Associates, LLC and Griffin Hospital, a non-profit hospital and home of the Yale-Griffin Prevention Reasearch Center.

ANDI stands for “Aggregate Nutrient Density Index”; the scores were generated by Eat Right America in cooperation with Dr. Joel Fuhrman.

Both sytems have used “nutrition experts” to develop their scoring systems.

ANDI’s Dr. Fuhrman represents their scoring formula as Health = Nutrients / Calories (H = N / C). NuVal uses a similar equation with good elements in the numerator and bad ones in the denominator. A more detailed comparison of what they use for scoring is provided in Table 1.

Table 1. – Elements Used in ANDI and NuVal Score Numerators

Source: ANDI and NuVal documentation

According to Fuhrman’s formula, ANDI only includes calories in the denominator. NuVal’s denominator, in contrast, includes fat (trans and saturated), sodium, sugar, and cholesterol.

I am an economist with no professional knowledge of nutritional science. But even so, a couple of quotes from Fuhrman caught my eye:

1. “…the last twenty years has demonstrated that colorful plant foods contain a huge assortment of protective compounds, mostly unnamed at this point. Only by eating an assortment of natural foods that are nutrient-rich, can we access these compounds and protect ourselves from the common diseases that afflict Americans.”

Excuse me? If protective compound/phytochemicals are unnamed and unmeasured, how can we be so sure they will be “healthful”?

2. “Because phytochemicals are largely unnamed and unmeasured, these rankings underestimate the healthful properties of colorful natural plant foods…. One thing we do know is that the foods that contain the highest amount of known nutrients are the same foods that contain the most unknown nutrients too.”

• How do we know the foods containing the most good nutrients contain the most unknown nutrients?
• How do we know that the unknown nutrients don’t have undesirable side effects?

Overall, both ANDI and NuVal scores represent today’s common wisdom on nutrition, i.e., leafy green vegetables are best, followed by other vegetables, fruits, nuts, beans, and fish. Stay away from dairy products, meats, and sugar.

Scoring Details

Inasmuch as the proverbial proof of pudding is in the eating, let’s take a look at the scores. More specifically, let’s examine differences between the ANDI and NuVal vegetable scores. NuVal uses a 1-100 scale, while ANDI uses a 1-1000 scale. To make comparisons easier, I have converted the ANDI scale to 1-100. Neither group was willing to provide me with all their scores – they urged me to visit their stores. OK. Fine. I visited stores and copied down as many scores as I could. That means the following comparisons are not complete. But I believe my conclusions would be the same if the data were complete.

Table 2. ANDI and NuVal Vegetable Scores

The first notable difference between the ANDI and NuVal scores is the scoring range. ANDI’s scores range from 100 to 4, with an average of 39. Most of NuVal’s scores are in the 90’s, with an average 94. How should these score differentials to be interpreted? Is it not reasonable to assume that if vegetable X has a nutritional score of 100 and vegetable Y has a nutritional score of 10, the nutritional value of Y is one-tenth that of X?

According to the ANDI scores:

• the nutrition value of green beans, garlic, potatoes and corn is less than one-tenth that of collards and
• asparagus is 23% the nutritional value of collards.

I wonder. I repeat, I have no professional knowledge of nutrition, but the NuVal scores look far more reasonable to me.

The second notable difference is the lack of correlation between the scores. In fact, the R2, a statistical measure of correlation, is only .049 (an R2 of 1.00 indicates a perfect correlation and 0.00 indicates no correlation). This is troubling. These scoring systems are constructed in a similar manner, and one would expect a very high correlation on vegetable scores.

My conclusion: I don’t trust the ANDI numbers. I do not believe green beans have only 7% of the nutritive value of collards. I do not believe asparagus has only 23% the nutritive value of collards. Maybe the NuVal green bean score of 100 is a tad generous relative to collards and kale, but probably far closer to reality.

Disclosure: I do not like collards or kale.

Consider another food group: beans. Table 3 provides the ratings. All of the ANDI scores are low, while NuVal scores are more forgiving. And once again, the correlation of the scores between the two systems is very low. ANDI and NuVal give lentils pretty high ratings. NuVal gives split peas the same high rating while ANDI gives it about half the value of lentils. Why?

Table 3. ANDI and NuVal Bean Scores

Other Reasons to Prefer NuVal

There are several other reasons to prefer NuVal to ANDI.

• Scoring Branded Products – Both ANDI and NuVal should be applauded for taking on the meat and dairy industries by giving their products low scores. But NuVal takes it one step further by rating branded products. For example, NuVal gives Snyder Sourdough Nibblers and Wise Ridgies Sour Cream Onion chips scores of 3, while Sun Chips get a 26 score. Rest assured the markets displaying NuVal score got calls from the high-ups at both Snyder and Wise.

• A Clear Statement on Organic Products – From NuVal:

“As yet, there is no widely validated evidence that organic foods have a higher nutritional value or greater nutrient density than food not grown organically. Obviously, consumers who choose organic foods may be doing so for reasons beyond nutrition – the fact that it is grown without using certain chemical controls, for example. If the scientific community puts forward well-vetted evidence that organic foods do, in fact, offer greater nutritional value, the NuVal™ System is designed to be able to incorporate that information into the Scores.”

And kudos to the markets that carry NuVal ratings. To post 6,000 scores is a formidable undertaking, costing a lot of staff time.

BUT – What Food Info Is Really Needed?

Are the ANDI and NuVal scores the most important information to have when shopping for food? Vitamin/mineral/nutrient deficiencies in the developed world are not that prevalent? What is prevalent? What should the primary nutrition concern be?

I quote from an previously written article that will be posted here in the next few days:

“The world is experiencing an obesity epidemic. The UN has conceded there are more obese than hungry people, even in developing countries! In the US, two easy statistics to remember:

• two-thirds of the population are overweight; and
• one third is obese.

And with jobs increasingly sedentary, things will probably get worse.

My recommendation: the most important information for consumers is calories. While food providers are required to include calories on their packaging, the print is often too small to read. If markets really want to help customers, they should put “calories per average serving” in large print right next to the price on all products.

I spoke to 5 doctors about this article. None wanted to be quoted on ANDI or NuVal. In essence, they all said: “keep your weight down and exercise”.  And, as to the question “what info is really needed,” The New York Times has an article today that discusses FDA efforts to ensure that proper representations are made regarding the health benefits of various foods.

Share this Econintersect Article:
  • Print
  • Digg
  • Facebook
  • Yahoo! Buzz
  • Twitter
  • Google Bookmarks
  • LinkedIn
  • Wikio
  • email
  • RSS
This entry was posted in Health Care, Studies and tagged , , , , , . Bookmark the permalink.

Make a Comment

Econintersect wants your comments, data and opinion on the articles posted. You can also comment using Facebook directly using he comment block below.