by Marion Nestle
Jan 2 2008

More on nutrition scoring: David Katz responds

My previous post elicited a long and thoughtful comment from Dr. David Katz. His response is well worth reading. Take a look. And thank you David for writing in. Your scoring system is an interesting experiment. Will people routinely choose foods with higher scores? Will food companies change their product formulas in order to qualify for higher scores? Will people be healthier as a result? Will they be less confused about nutrition?  I will be interested to follow the progress of your system and see how it works out. Thanks again for writing and happy new year!

  • Does this system factor for gmo’s in processed foods?

  • I’m glad to have read Dr. Katz’s argument for use of the ONQI nutrient scoring system. I am personally not in favor of scoring individual foods and, although I found the argument persuasive, after reading through a list of foods with scores on the ONQI website, my worries are reinforced.

    I can’t argue with the method to develop the scoring system. But as a nutritionist, I am very scared that now the sentence – oatmeal and apples are less healthy than strawberries – will be supported by their score. I hope a very clear message about comparing food scores is included in the roll out.

  • Responses to Gravel, and Katie:

    GMOs: no. Although the ONQI Scientific Panel was philosophically supportive of many such issues, especially giving credit to organic food, we could only include in the algorithm those factors where a health effect was at least nominally quantifiable. Issues for which the science was not yet sufficiently robust to allow for that were tabled, to be reviewed for subsequent iterations of the ONQI, which will be generated at regular intervals to keep pace with evolving science.

    As for strawberries, and oatmeal: The ONQI program will include a whole campaign of consumer education to address this concern, although it is probably acceptable for the public to learn that even among fruits and vegetables, some are better for you than others. But perhaps most importantly: across the 100-point range, scores will be color-coded in three tiers. We will likely use a modified traffic light system, so for now, let’s assume: green, yellow, orange.

    Foods of very high nutritional quality- to include apples, oatmeal, and strawberries- will be shaded green, all in the “eat often for better health” category. Middle tier foods will span a large numerical range, but will be colored yellow to indicate that they should be eaten routinely, but in moderation. Foods shaded orange will also span a considerable range of numbers, but will all be foods that should be consumed in judicious quantities.

    One more issue. On the page, you can see ONQI scores for diverse foods all together. But in the supermarket, you will see foods as they are distributed by aisle. So oatmeal won’t appear next to strawberries, but next to other cereals, where it will have one of the highest scores, appropriately indicating it is a great choice within that category.

    Bottom line: we would be no happier talking people out of oatmeal than you would be! There are and will be robust defenses in place against the system being misused in that manner.


  • Thank you for that, Dr. Katz. I look forward to learning more about the scoring algorithm. Is the griffenhealth/onqi website the best place for detailed info?

  • Kati- yes, for now- (which links to the Griffin site) is the best place for information about the system. We’ll be updating and expanding the material there soon.

    Just fyi, we will be launching the ONQI on the Internet, as well as in supermarkets, so soon there will be a fully interactive site where you can enter any food and determine its ONQI score; compare foods to other foods; generate a shopping list; and more. We expect to be announcing a partnerhip with one of the premier health Internet sites within the next several weeks. Ultimately, the site will become a link to the interactive site.

  • Pingback: » More on nutrition scoring: David Katz responds()

  • Thank you for the response, Dr. Katz. I am glad to see that there are groups making genuine efforts to provide simple health and nutrition information to the general public. If anyone ever asks me (and I’m continually surprised that this happens), I’ll recommend your system above others.

    Having said that, I am also against using scoring systems for food. My concern about them comes from the very simple position that I’m concerned about any food system which has become so complex that we need experts to come in and analyze our food for us. Nevermind concerns about which nutrients/nutrient clusters any given system gives primacy to, or if the system differentiates between nutrients found naturally in the food or artificially added, or any industry biases, or GMOs, or chemical processing content, or being susceptible to the whims & vagrancies of popular nutritional beliefs, or any other of a host of problems. I’m far more concerned that grading systems of all stripes only ultimately serve to legitimize a system of food production and distribution that is hazardous to almost every aspect of our health and environment. If people ate diets based on largely whole foods, minimally pre-processed, there would be no need for anything like a grading system, because people would be receiving the nutrients they need. The food wouldn’t be so complex that we could easily be fooled into believing we’re eating something that’s healthy when we’re not.

    Now, of course, I understand that the “whole foods” world is not the one we find ourselves in now. I’m very sympathetic to the argument Dr. Katz ran, that given our current food situation, we have to figure out some way of getting good information and easy-to-use nutrition criteria to people, and this is why I will recommend the ONQI system to those who ask me. But it still comes back to my overall concern that, rather than making a concerted effort to change our overall processed food system, these grading systems merely make it more OK to eat that stuff, since it has, for example, a “yellow light” on it. I guess I’m not naive enough to think that our agricultural industry will embrace a whole-foods diet approach anytime soon, so maybe food grading systems are the best we can do, but to me, that’s terribly unfortunate.

  • Keith Ayoob, EdD, RD

    Responses to Marion and previous responders:

    First of all, I want to thank Marion for generating this dialogue. It’s also nice to see some thoughtful responses and know that people are paying attention.

    I’m not the inventor of the ONQI, but I did serve on the scientific panel that helped develop it, so I feel some need to weigh in here.

    The concerns expressed by the responders are valid and legitimate. They are also my concerns and those of my fellow panelists, and we discussed them at length (and often vigorously) on numerous occasions. We discovered that, for the ONQI to work, we needed to establish some ground rules:

    First, we had to acknowledge that the ONQI was to be a tool: for teaching consumers, for doing research into how people eat, how they choose what to eat, and for measuring changes to how they eat.

    Next, we all had to accept that we’d have to check our personal stuff at the door. People eat food, not philosophy, and while food decisions are increasingly becoming ethical ones as well, for the ONQI to be true to its purpose we would have to base it on the best available research to date.

    We would also have to base the ONQI on nutritional content. As important as it may be to eat sustainably and locally whenever possible, these factors would be difficult if not impossible to quantify. Besides, who is to say that organic carrots shipped 3000 miles across the country arebetter or worse than conventional carrots grown 10 miles away? Issues like packaging also needed to be left out of the equation.

    In terms of why a simple, single number might be needed, I say this: it’s quantifiable, not just for us, but for the consumer as well. People like simple numbers. Such an index would also help consumers choose better foods within a category and help them move the needle a little at a time. In my 24 years of working daily with consumers (and lots of research by others on the subject), I feel confident in saying that people don’t tend to make huge changes overnight. They like small gradual changes and, if made consistently, such changes are more likely to be permanent.

    The ONQI can help people make small, gradual changes and gently move the needle to a better place by telling them how they’re doing at any given point.

    Consumers hate being told what NOT to eat. They love being told what’s good to eat (thank you for that, Marion!). What the ONQI hopes to do is tell people what’s a little BETTER to eat. Then they can make the decision for themselves.

    David often says that “perfect” is the enemy of “good.” We in the nutrition world need to remember that. When I’m working with families, I can’t expect perfection, I can only hope for improvement. Little improvmenets in diet and eating habits however, tend to be motivating, and this is where the ONQI may be especially useful.

    Of course, how the ONQI is implemented and explained will be ciritical to its usefulness, to be sure. The points made by Katie and Marion about scoring individual foods, etc. are valid concerns, and we aim to put things into the proper context when explaining the ONQI, as David indicated above. Remember also that an objective scoring system may be valuable for counteracting inappropriate marketing claims,.

    Finally, nutrition science is dynamic, and the ONQI will need regular review and updating, but that’s just what makes it all the more interesting.

    Best regards,
    Keith Ayoob, EdD, RD

  • Robyn- I posted a rather comprehensive response to your very thoughtful comments yesterday morning, only to have it swallowed up by cyber-gremlins. I am trying again, with Dr. Nestle’s encouragement that the gremlins have been apprehended. My thanks to Dr. Ayoob for his post in the interim; the gremlins obviously haven’t tracked him down yet.

    I agree entirely that we want to eat closer to nature. In fact, in a nutrition education program my wife and I developed for elementary schools, called ‘Nutrition Detectives’ (see: one of the 5 key clues we offer about identifying more wholesome, better-for-you food is a short ingredient list. This concept is, of course, epitomized by foods direct from nature, such as- a carrot, or an apple. The ingredient list for a banana is: banana. Such brevity is the soul of nutritiousness! And eating more foods devoid of chemical adulterations ’tis a consummation devoutly to be wished, we certainly agree.

    But in the real world, the simple fact is that a lot of what we all eat comes out of bags, boxes, bottles, jars and cans. And in each food category spanning the roughly 50,000 food inventory of the typical supermarket, there is a vast range in nutritional quality.

    The benefits of guiding people toward the preferable end of that range include the incremental improvements in nutrition to which Dr. Ayoob refers. Choose better bread, and cereal, and salad dressing, and pasta sauce…and…and…- and the net effect on the quality of your diet can be enormous. Dramatic net reductions in your intake of salt, sugar, saturated fat, trans fat; comparably dramatic increases in fiber, vitamins, minerals, antioxidants. Yes, the means are incremental, but the ends come pretty fast nonetheless, through the accumulation of incremental steps.

    But the benefits go far beyond this. I have long maintained that the most effective way to improve the food supply- ie, shift it closer to nature- is to improve the food DEMAND! After all, companies are in business for business- and want to keep the customer satisfied. The ONQI will, in fact, reliably guide consumers to the more wholesome, closer-to-nature products, because these consistently get the best scores (the system does distinguish between natural, intrinsic nutrients and nutrients added through fortification, by the way).

    We know that demand trumps supply, if only thanks to the late Dr. Atkins. Personally, I think his advice was silly at best, dangerous at worst. But he showed us how changing demand can change the food supply. Every supermarket in the country filled up with low-carb fare due not to legislation, or litigation, or policy reform- but due simply to the fact the people wanted it!

    Now imagine if we can get millions and millions of people to want not the silly abandonment of a macronurient class- but better overall nutrition! And imagine everyone is actually empowered to recognize better nutrition at a glance, using a universal, robust, science-based and objective system. If people care- and I both hope and believe they do- the system will shift their choices to better nutrition. Doing so will shift their choices closer to nature.

    Then the really good part: food companies will be motivated to shift their product offerings to compete in a new world where the average consumer really can choose on the basis of nutritiousness. Entire prodct categories should start to move closer to nature, trimming away the additives and adulterations that compromise nutritional quality, and lower the ONQI score.

    I am a pragmatist, and believe that a difference, to be a difference, must make a difference- in the real world. But I am also a dreamer, and dare to believe that the ONQI might prove to be the Archimedes’ lever of nutrition reform. One powerful force that can shift, little by little, the nature of the food supply.

    Our vision, I think, is a shared one. We both can perceive and aspire to the same illumination at the end of the tunnel. I have, along with my colleagues, built the ONQI to be a vehicle that runs along those tracks, and carries us there. My hope is that the ONQI proves to be means to the ends we both endorse. How well it serves that purpose, as Marion says- we will see. But I dare to hope! All aboard…


  • Thank you to both of the scientists who worked on the ONQI system–I am very glad to see the willingness to engage with us about our concerns. I am greatly heartened by Dr. Katz’s reply in particular–you do seem to understand my concerns very well and take them seriously. And I think, when I’m not falling victim myself to the “perfection vs. good” fallacy, that you’ve got some very good points. Not for nothing, what you say rings true in my own life’s experience, and my attempts to shift my family’s diet. Our own efforts began with implementing a once-a-week vegetarian day on the Catholicism model that it’s good to be reminded regularly that there are those who cannot afford to eat meat (as you see, this change was made on purely ethical grounds, rather than health grounds). Almost unwillingly, this comparatively small change began to pervade our diet. As I discovered more and more delicious vegetarian foods, and began to see the benefits to both our health and our bank account, our diet shifted further and further. It was gradual, but easy. We are by no means vegetarians now, but we eat an overall vastly healthier diet now.

    All of this is to say that I sincerely hope the ONQI system, if done properly, will offer folks the same sort of mechanism for diet change that our simple vegetarian-day change gave to us. I am looking forward to seeing how the educational materials are used, and to its overall effects. As someone working on beginning a cooperative “natural foods” market in my hometown, I will be particularly interested in how this system could be implemented in a consumer setting. Should our store wish to employ such a device (a real possibility), we will take yours very seriously.

    Thank you again.

    (And I’m also looking into a new, non-toxic, sustainable-and-environmentally-friendly anti-internet-gremlin trap which is safe for both kids and pets!)

  • Mark

    I guess my problem with nutrient density scores is that they imply that the goal is to mostly eat foods that have a broad balance of nutrients. This is not necessary for health. The goal is for your weekly intake of food as a whole to have a high nutrient density.

  • Mark- the ONQI goes well beyond measures of nutrient density, also measuring the biological quality of protein; the glycemic load, the ratio of healthful to unhealthful fats; and so on. It truly is a measure of overall nutritional quality.

    The ONQI takes a diet eye view, not a food eye view, of nutritional quality. It asks and answers the question: how does the concentration of various nutrients in a food compare and contribute to the recommended concentration of those nutrients in the diet overall? I could embellish this topic at great length, but I’ll let it go a that.

    Let’s use the common analogy: the forest, and the trees. Yes, the forest matters more- and should not be lost for the trees. But, of course: the forest IS the trees, in the aggregate. There is a limit to how many unhealthy trees there can be, before the forest is unhealthy! Conversely, if all the trees are vital and flourishing, de facto, so is the forest.

    Diet and foods are the same. The overall diet is what matters. But the aggregation of the foods we choose IS our diet. Lots of healthful foods in the right combinations add up to a healthful diet. Conversely, the quality of the diet can be, and all too often is, severely compromised one ill-informed food choice at a time.