News came in the past week that the front-of-pack nutrition guidance program offered by Canada's Heart and Stroke Foundation, presented as a seal of approval in the form of a check mark, was being decommissioned. With all due respect to my friends at the Foundation, and the good intentions that brought the system into existence -- good riddance to it.
Before saying anything more about the Health Check Program, or any other nutrient profiling system, I hasten to disclose that as the principal inventor of the NuVal system, I have skin in this game. But that proviso warrants several immediate provisos of its own.
First, I am not opining because I was involved in developing a better system; if anything, it's vice versa. I had strong opinions about the need for such a system- expressed back in 2003 directly to the U.S. Secretary of Health and the Commissioner of the FDA, and in a subsequent op-ed in the Hartford Courant and elsewhere. Both the inclination to opine, and to develop a better system, derive from the same motivation to empower consumers. My skin in this game is much like Al Gore's investments in green economy enterprises; passion and conviction came first, and then we sought opportunity to put our time, effort and money where our mouths were, and do more than just talk.
Second, the algorithm that powers NuVal, the ONQI, was not built to be a business -- nor is it owned by one. Its development was funded by the Yale-affiliated, not-for-profit, community hospital that owns it to this day, Griffin Hospital in Derby, CT, where Yale University's Prevention Research Center is housed. When completed, in 2007, I made a trip back to Washington, D.C., and offered it to the federal authorities for public use. But in a private meeting, government scientists indicated to me that such a provocative system (i.e., it actually tells the truth about our food supply) might be held up in bureaucratic tangles for years. It was at the initial encouragement of a colleague at the FDA that NuVal as a business came into existence. Really.
I'll get back to NuVal and what makes for a good front-of-pack nutrition guidance system shortly. For now, why the good riddance to the Health Check program? The criteria for offering a check were reasonable, but not truly robust. More importantly, the evaluation on which check/no check decisions were based had to be purchased by food companies. As a result, many small food companies making exceptionally good products did not participate. The result, of course, was at times worse than no guidance at all: an item with a check might be substantially less nutritious than an adjacent item with no check simply because the latter product had never been evaluated. This would be like a GPS system leaving out streets haphazardly, and thus a perfect formula for getting quite lost. The Foundation's system has been controversial in Canada for years primarily for this reason.
Finally, there is the problem of expressing the range of nutritional quality in a supermarket with some 40,000 products with any binary system. Products get a check, or they don't. But obviously if, for instance, Cheerios gets a check, and so does broccoli, it scarcely means that Cheerios are as nutritious as broccoli. In just the same way that a pass/fail grading system does not reveal the range of academic aptitude, a dichotomous nutrition guidance system is blind to a vast expanse of nutritional variation. It's a very blunt instrument, like a GPS system that only tells you only whether your next turn is to the left or right, but not where, or when, or how far.
The Heart-Check seal of approval program offered by the American Heart Association has very similar liabilities.
These, of course, are by no means the worst systems cluttering up this space. The distinction of inventing those resides, naturally, with the very companies selling the foods in question. The one that stands out as a case study in what happens when foxes guard the hen house is the "Smart Choices" program, which famously stationed its halo over Froot Loops cereal. That one was under the threat of federal litigation when it was consigned to the dustbin of bad ideas.
But thus far, the Facts Up Front system has not been. Nor need it be, I suppose; it's not really a bad idea, since it doesn't quite rise to the level of an idea at all. It's a change of venue. Facts Up Front takes nutrition information already offered on the nutrition facts panel and puts some of it, you guessed it, up front -- on the front of the package. As a colleague once insightfully quipped, this might really help combat trends in obesity and chronic disease if the principal problem were the average consumer's inability to turn boxes around. Otherwise, not so much.
One of the world's better systems, the traffic light labeling system in use in the U.K., has considerable global traction. There was news in the past week that France was considering its use. But while good compared to bad systems, traffic light labeling has serious limitations if the objective is informed, at-a-glance decisions about the most nutritious options. The traffic light assigns one of the three associated colors to each of several nutrient properties: calories, total fat, saturated fat, salt, and sugar.
Some will no doubt immediately object to penalizing 'total fat,' and they are right to do so. Walnuts, almonds and avocado would all fare poorly in a system that penalizes total fat regardless of its quality. They would also take a hit for their high calorie content, despite the fact that these calories are "worth it," both because they are full of nutrients, and because these foods are very satiating -- leading to a lasting feeling of fullness. High-calorie foods with a high propensity to satiate us can actually help control, and reduce, total calorie intake over the course of a day. (All of these considerations are incorporated into NuVal, by the way, which awards very high scores to walnuts, almonds, and avocado.)
Perhaps more importantly, the traffic light system can be very hard to decipher with regard to the relevant question: Which product do I choose? If, for instance, one product has three greens and two reds, is that better or worse than one green, three yellows and one red? What about three greens and two reds compared to five yellows? Such questions are all but limitless, and honestly -- I don't know the answers to any of them.
There is evidence from independent research that good systems are good because they actually work better in the hands of consumers. Since NuVal is in over 2,000 supermarkets throughout the U.S., reaching some 40 million consumers, the anecdotes related to its effects abound, and including numerous reports of such benefits as more than 100 pounds of weight loss. One of the many virtues of more nutritious foods is that they fill us up on fewer calories -- the very opposite of "betcha' can't east just one" processing -- and such cases attest to that effect.
Both NuVal and the Traffic Light system fared better than the rest of the contenders, including a construct supported by the Institute of Medicine, in a test of consumers reported at last year's meeting of the American Public Health Association.
But even among good systems, NuVal is a stand-out. A study conducted at McGill University, and now in press, shows how much more efficiently NuVal, using a single number, reliably informs more nutritious choices as compared to the confusing profile offered by a traffic light system. The higher the number, the more nutritious the food is hard to beat for simplicity.
The data backing up the NuVal system are unique as well. It is the first and to date only nutrient profiling system shown to correlate directly with both the rate of total chronic disease, and all-cause mortality. In a Harvard study of over 100,000 people, higher average NuVal scores meant a lower likelihood of heart disease, diabetes, obesity, or dying prematurely from any cause over a 20-year period of observation.
Given the current plethora of passions directed at competing nutrition theories, a few words about how NuVal addresses these are warranted. The system penalizes added sugar rather severely, based on the relevant epidemiologic evidence. It differentiates added from intrinsic sugar, and applies a harsher penalty for the former. There is an additional penalty for glycemic load, which can be pushed up both by sugar, and refined starches.
Sugar substitutes are addressed as well. While products that replace sugars with artificial sweeteners benefit from the reduced calorie load, sugar load, and glycemic load alike, the algorithm scores these products in a way that addresses the uncertainties about the health effects of artificial sweeteners. This is especially true for the updated algorithm, ONQI 2.0, recently completed. Those scores are currently rolling out in waves nationwide.
Unlike the traffic light system, NuVal does not penalize total fat. Rather, it scores particular fats on their particular merits. Monounsaturated fats and omega-3 polyunsaturates are rewarded. Trans fat is penalized quite harshly, as it deserves. And saturated fat is handled based not on the currently prevailing hype, but on the basis of the weight of scientific evidence. In ONQI 2.0, the penalty for saturated fat remains, but is attenuated somewhat. More importantly, the algorithm does just what the 2010 Dietary Guidelines Advisory Committee recommended: exclude stearic acid from the saturated fat penalty. The Dietary Guidelines for Americans did not follow this advice, probably because advising Americans to restrict intake of some but not other saturated fatty acids does not translate easily into intelligible, actionable messaging. But a system that measures nutritional quality holistically and expresses it as a single number can readily do this -- and NuVal does. If, as it matures, the science exonerates other saturated fatty acids, as seems likely, for instance, for the lauric acid that predominates in coconut oil -- the algorithm can readily be adjusted to carve that out as well.
Finally, there was news this week out of Boston that Harvard Pilgrim is using a nutrition guidance system, apparently modeled after NuVal, as a basis for financial incentives. I have been advocating for such a model for some time. In the case of NuVal, well over 100,000 foods have been scored, and the evidence is in hand that the system correlates with the health outcomes that truly matter. That begins to argue for a policy that links nutrition guidance and financial incentives for the benefit of the neediest populations, namely food stamp recipients. That model, called Finger Tip, could help poor people choose good food to get to better health; save taxpayers a lot of money; and help keep the government solvent. Everybody wins.
While its current reach is fairly impressive, given its pedigree and performance, NuVal should have been adopted far more universally by now. The question then, is why hasn't it been?
One potential answer is a business model that has emphasized the real-world real estate of retail supermarkets, and done relatively little thus far in cyberspace. Interactive websites and apps are options the business of NuVal controls. Another is that because this is a public-private partnership, and a business, the algorithm is proprietary. But so are the blue prints for our cars and computers, and that doesn't prevent us from being able to judge when they work well. This is an issue, but often, I think, more an excuse.
An excuse for what? The obvious thing: concealing the at-a-glance truth about the overall nutritional quality of foods in a culture that routinely engages in lipstick-on-a-pig formulations of junk foods. Concealing the fact that throwing multivitamins into a vat of insalubrious gloop does not absolve the gloop. NuVal, which differentiates between intrinsic and added nutrients, bursts this protective bubble.
The principal liability of the NuVal system is that it truly, actually works just as intended. And it may be that for the powers that be, the single most unpalatable thing in the food supply is the unadulterated truth, on at-a-glance display.
-fin
David L. Katz, MD, MPH, FACPM, FACP is the founding director of Yale University's Prevention Research Center, President of the American College of Lifestyle Medicine, and author of 'Disease Proof.' He has authored roughly 200 scientific papers and 15 books, including three editions of a nutrition textbook for health care professionals.