Nutrition Labeling Will Never Work

Michael Jacobson, Executive Director for The Center for Science in the Public Interest, has been at the forefront in consumer nutrition education for over 40 years; and he has written a piece in the Washington Post (1.10.13) about new trends and ideas in consumer nutrition education, especially product labeling.

When I first moved to Washington, the Center was in a funky building near Dupont Circle and had just issued one of their first publications – a chart which displayed all the various additives to food and their potential danger or benefits.I had just returned from five years in India where I had been working on one of the first social marketing campaigns – efforts to use modern techniques of market research and advertising to “sell” nutrition; and I was encouraged to see that some innovative educational efforts were underway in this country. 

Prior to my India project, nutrition education was a dry, formal matter, and the most celebrated visual aid was a flip-chart of the famous Four Basic Food Groups.  Female extension agents would organize village women to attend discussion groups where the idea of a balanced diet were presented.

These proscriptive, academic lectures took no note of the various socio-economic, cultural, and market factors which determine dietary preference, purchase, preparation, and physiological assimilation.  It was the duty of the agricultural extension agents to present the Four Basic Food Groups and the duty of village women to listen, but little else.  Our project (CARE/USAID) sought to change that.  If we could carefully research out target audience to assess the determinants of diet and gauge the likelihood of modifying it; and if we could transform that information into an attractive and persuasive media campaign, people would listen, and their diet would improve.

The idea was sound, the execution was faulty.  Green leafy vegetables, a source rich in iron, was to be the magic bullet that raised iron levels, defeated anemia, and improved the overall health of millions of women.  What we neglected to consider, however, was the expense of growing vegetables of any kind and the subsequent lack of a market for them.  We underestimated the economic constraints for families living on the margins – any change in an already straitened diet was an unacceptable risk.  We never even considered the psychological satisfaction of a full stomach – i.e. one filled with rice and dal and not leaves, however nutrient-rich.  We learned too late about phytate buildup – the process during carbohydrate loading which prevents the body from assimilating iron. 

Although commercial advertising in the  US  - our model – had always been successful in changing consumer behavior, the success was most often for low-price retail items where the risk of changing brands was minimal.  It didn’t matter if you didn’t like Colgate toothpaste, you could switch to Crest.  The amount of money you risked and lost was insignificant.  Modifying diet on the margins had a huge risk and potential cost associated with it. 

Finally we overestimated the power of the media and the amount of investment in time and resources needed to change even a fraction of one percent of behavior. Hundreds of millions of dollars are spent by US advertising to effect a miniscule shift in brand purchase.  We had peanuts. 

Many organizations later thought that they could tweak the advertising-marketing model, do comprehensive consumer research, concentrate resources in a small, manageable area, and change behavior significantly and quickly.  Most failed.  Behavior change in nutrition is a complex, intransigently difficult affair.

Armed with all the knowledge that failure provides, I returned to the United States to see if there were to make a dent in the increasing problem of dietary fats, salt, and sugar and the rising rates of obesity and poor diet-related health.  Here it would be easy to mount a campaign similar to he one I implemented in India.  After all, this was the mother lode, the home of commercial success. 

Within a few months I had run into the buzz saw of commercial lobby groups.  The milk, egg, cheese, oil, fast-food, soft drink, and snack lobbies had a stranglehold on Congress and by extension the Department of Agriculture and their embryonic consumer protection agency.  Whereas in the centrally-planned Indian economy, government could program by fiat (Tell the people what they need to know), the US was a different, more diverse, fragmented, and complex mess of competing interests.

The Center for Science in the Public Interest was one of the first to take on the behemoth and benefitted from the Sixties’ promotion of ‘natural’ foods.  Many of us in the Dupont Circle area had their poster on our refrigerators and followed it religiously. The government, on the other hand, was hamstrung and could do nothing, however innocuous, to promote what scientists knew to be better nutrition.  Someone would complain and worse, threaten a lawsuit.

Finally in 1990 Government acted and food labels began to appear on most packaged food products.  These labels, now familiar, indicated the Recommended Daily Allowances (RDA) for each nutrient and the percentage of those allowances provided by a particular product.  A consumer could look at a can of vegetables and measure sodium, fat, sugar (as well as positive nutrients such as niacin, magnesium, folic acid, etc.) against maximum limits. A serving of peas might contain 20 percent of the RDA for sodium, a cookie 25 percent for fat.  Calories are indicated in reference to a standard average adult dietary requirement – e.g. based on a 2500 calorie diet – so a consumer could tell what percentage of his total calorie limit would be available in, say, a handful of Doritos.

It all seems simple and straightforward, but it is not.  First most consumers have no idea what a serving size is, and usually make assumptions that it is far larger than what is written on the package (in ounces).  One serving of peas is nowhere near the generous helping that hungry diners heap over the mashed potatoes.  Second, few people have any idea of what their caloric intake is.  Whether they consume 2500 or 3500 calories per day is beyond them, so any calculations made on the basis of the ‘standard’ baseline will be flawed and off by many factors.  Third, the percentage figures on such nutrients as salt mean little.  It is OK, say many consumers, if I eat a handful of salted peanuts, because it is only 15 percent of my daily requirements.  What they do not calculate is the processed ham for lunch, the tortilla chips for snacks, and the salt-infused pizza for dinner.  Many Americans go way beyond their RDA for sodium without batting an eye.  The same is true for calories.  Unless a consumer can visualize what a 2500 calorie day is like, there is no way that he can quickly calculate the relative importance of, say, a cookie or a slice.

In other words, the only way anyone can make sense out of the nutrition labels is to take a short course in nutrition. Start with body weight, actual calorie intake, calorie expenditure, dietary intake of nutrients; then do practice runs on quick mental calculations of relative value – Can I really afford to eat that piece of pie?  Unless this entire calculus is understood, the labels are meaningless. “Too much salt” is a relative term, and that relativity makes it difficult to define.

To give some idea about the complexity of the issue, here is an excerpt from an FDA site on serving sizes in the question-and-answer format:

L20. We produce a cookie assortment containing various percentages of 6 different cookies. What nutrition format should be used?
Answer: The manufacturer may choose to use: (1) a separate Nutrition Facts label for each variety of cookie in the package, (2) an aggregate label (i.e., a single Nutrition Facts label including nutrient content information and % DVs in separate columns for each variety), or, (3) if it is likely that one person would eat an assortment of the cookies at the same time, a composite label that provides one set of nutrition information based on a weighted average of all of the cookies in the assortment. 21 CFR 101.9(h), 21 CFR 101.9(d)(13)

L63. What terms must be used for the serving size?
Answer: The serving size declaration is made up of two parts: a “household measure term” followed by its metric equivalent in grams (g). For beverages, the household measures may be declared as either fluid ounces, cups, or fractions of a cup with the metric equivalent in milliliters (mL).

So 1 cookie can be expressed as 28g on the label.  Since we are not on the metric system, to whom will this be readily understandable?  And even if it were in ounces, who has a kitchen scale handy?  Europeans cook by weight measure; Americans by fluid ounces.

Acknowledging the difficulty of this mathematical puzzle, Government decided to present this concept of relative importance graphically:

As anyone can quickly see this chart doesn’t help at all.  What does “Small amounts of fats, oils, and sugar” mean?  As discussed above, relative value is difficult to calculate in nutrition; and what may be a small amount to one person may be a lot to another.  The issue of serving size will not go away. “Two servings each” of fruits and vegetables mean of just those shown, or others; and how many others?  “Fruits and vegetables” show only a selected few, and tomatoes don’t happen to be on the list; so does that mean no pizza sauce?  Or if so, how do you calculate a serving of tomato sauce on pasta? According to this chart the meat, fish, and dairy products are lumped together.  While the reason for this is clear – all are protein sources – the consumer could eat only cheese omelets and pork chops washed down with milk and consume more fat that would be recommended.

An article in the NY Times in 2010 reported on the FDA deliberations to ‘bring RDA into line with reality’ – bureaucratic-speak for upping the size given the huge amounts people already eat.  So instead of having mini-bite cookies, let’s go for real, true, big ones.  Some improvement!

Now comes another idea – rate foods with stars to give people an idea of the overall value of a particular food:

Perhaps the smartest approaches rate foods by giving them credit for the nutrients we should be eating more of and subtracting credit for the bad stuff. The Hannaford Brothers and Food Lion supermarket chains give all foods zero to three gold stars, depending on the balance of nutrients. Everyone can understand these simple ratings, which are printed on shelf tags rather than on labels.

The NuVal system, which is used on shelf tags by Raley’s, Hy-Vee and two dozen other supermarket chains, rates all foods on a scale of 1 to 100. That approach allows people to scan dozens of salad dressings, breakfast cereals or frozen dinners to see which have the highest, or best, numbers.

The problem – again – is with the concept of relative value. Whole milk should be rated high for protein and Vitamins A&D but low for fat. For healthy, active young adults who need calories and protein, milk might be considered a power drink; but for overweight, sedentary older people, milk should not be an option…except in small quantities.  So how do you rate milk?  Science aside, what would the milk lobby say if FDA ranked its product with no stars?

I have been following this labeling/nutrition education debate for years and have been unable to see a clear way forward.  I think that government attempts to educate adults have failed, and there has been no concerted effort to teach comprehensive nutrition in schools.  If anything, the focus is on obesity.  A revision in the nutritional norms for school lunches has resulted in even more plate waste than before because children simply will never prefer carrot sticks over chicken nuggets.  Experiments in California with ‘interesting’ food such as Thai dishes appealing to foodies, were routinely tossed in the garbage by school children.

The answer, perhaps, lies in the market and in the eventual change in cultural norms.  Where I live, most people shop at Whole Foods and come back with a market basket filled with fish, fresh fruits and vegetables, non-fat Greek yoghurt, salt-free canned goods, and the finest assortment of cage-free, open range, organic poultry and meats. If any of my neighbors were to catch a glimpse of my shopping bags filled with Fruit Loops, Oreos, frozen pizzas, and lunchmeats, I would be dunned out of the neighborhood.

The good food fad has already moved somewhat out of wealthy enclaves into the mainstream, but it has often been coopted by manufacturers who have tried to label their products ‘natural’ and ‘healthy’ without a shred of proof.  This cat-and-mouse game with the FDA takes time and money, so in the meantime caveat emptor.

It is likely that if incomes rise across the board, those people eating cornpone and fatback might opt for a few veggies in the mix, and eventually diets will improve.  As I have written previously, the principal factor in obesity is poverty – if you don’t have money, then you will be forced into a high-calorie high-fat, high processed food diet.

Culture does matter, and whatever comes out of California seems to be adopted by the nation; so let’s hope that the foodie culture, now a religion out there, will make its way east.