Why Our Nutrition Facts Need an Overhaul

One bag of chips has 300 calories, the other has 360. Right now, it’s impossible to tell which one <i>actually</i> has the lower amount.

National Journal
Brian Resnick
Feb. 26, 2014, 7:32 a.m.

“To say they [the data] are im­per­fect is the equi­val­ent of say­ing the Ti­tan­ic had a float­a­tion prob­lem or a buoy­ancy prob­lem.”

Earli­er this week, news broke that the Food and Drug Ad­min­is­tra­tion is fi­nally set to re­lease new guidelines on nu­tri­tion la­bels, 20 years after it im­ple­men­ted the cur­rent ones. This move is huge. Not only does it af­fect every single food pro­du­cer in the United States (in hav­ing to re­design product pack­aging), it could also stand as a paradigm shift in the way Amer­ic­ans think about nu­tri­tion.

For in­stance, if “ad­ded sug­ar” is high­lighted more prom­in­ently on the la­bel, as some are re­port­ing it might be, that would co­di­fy sug­ar as be­ing pub­lic-health en­emy No. 1. If it provides con­sumers with more real­ist­ic serving sizes, it might in­duce second thoughts about pur­chas­ing that pint of ice cream that says it is five servings (but we all know what will really hap­pen there).

The nu­tri­tion la­bels as we know them have been around since 1994, when the 1990 Nu­tri­tion La­beling and Edu­ca­tion Act went in­to ef­fect. But nu­tri­tion is a slip­pery sci­ence, de­pend­ent on self-re­por­ted data­sets for dec­ades. What we “knew” just a few years ago — for ex­ample, that eggs are bad for heart dis­ease — doesn’t ne­ces­sar­ily hold true today. And pub­lic un­der­stand­ing of nu­tri­tion is fur­ther di­luted by the prom­ises of fad di­ets.

Which is to say that the pro­spect of cre­at­ing a nu­tri­tion la­bels that re­flects sol­id sci­ence and can re­main val­id for dec­ades is fraught. Add to that chal­lenge, the task to make that in­form­a­tion sa­li­ent to people without over­whelm­ing them, and the task seems down­right im­possible.

But change is needed. Here’s why our cur­rent la­bels are so flawed:

1. There’s no one stand­ard way to meas­ure a cal­or­ie.

Say you are look­ing at two bags of chips. Both weigh the same, and pre­sum­ably con­tain the same amount of chips. One bag has 300 cal­or­ies, the oth­er has 360. Which one ac­tu­ally has the lower cal­or­ic amount? Right now, it’s im­possible to tell.

That’s be­cause the FDA al­lows for five dif­fer­ent ways to meas­ure total cal­or­ies and al­lows for a mar­gin of er­ror of up to 20 per­cent. Ba­sic­ally, a man­u­fac­turer can either use pre­de­ter­mined con­ver­sions (one gram of car­bo­hydrate is four cal­or­ies) or burn the product to see how much en­ergy that comes off. These will provide dif­fer­ent val­ues.

And the FDA doesn’t have the greatest track re­cord of over­sight on the la­bel val­ues. In 2008, a Gov­ern­ment Ac­count­ab­il­ity Of­fice re­port found “FDA’s test­ing of nu­tri­tion in­form­a­tion has been lim­ited and has found vary­ing de­grees of com­pli­ance.” Al­though 87 per­cent to 91 per­cent of la­bels were in an al­low­able range in an audit, “com­pli­ance rates var­ied sig­ni­fic­antly for a few nu­tri­ents.” Only 47 per­cent of la­bels were in the cor­rect range for the amount of Vit­am­in A they con­tained.

2. Per­cent daily val­ues are based on very shoddy data

Much about what we “know” about nu­tri­tion comes from a four-dec­ades-deep data­set called the Na­tion­al Health and Nu­tri­tion Ex­am­in­a­tion Sur­vey, or NHANES, which is com­prised of self-re­por­ted data. And from this, ba­sic cor­rel­a­tions emerge, like how high-fat di­ets are linked to high­er weights, and so on. The sur­vey has also giv­en us a pic­ture of what the av­er­age hu­man eats. Those val­ues, in turn, are re­flec­ted in nu­tri­tion policy.

But it turns out that people are hor­rible about ac­cur­ately re­port­ing their food in­take.

“To say they [the data] are im­per­fect is the equi­val­ent of say­ing the Ti­tan­ic had a float­a­tion prob­lem or a buoy­ancy prob­lem. These data should not be used,” says Ed­ward Arch­er, an obesity re­search­er, who au­thored a sci­entif­ic take­down of the NHANES data in the Journ­al PLOS One.

That’s be­cause people would com­monly re­port levels of food that are in­cap­able of main­tain­ing hu­man life. “In no sur­vey did at least 50 per­cent of the re­spond­ents re­port plaus­ible [en­ergy in­take] val­ues,” Arch­er and his coau­thors con­cluded in their study.

3. Serving sizes do not re­flect real­ity.

Ac­cord­ing to the Cen­ter for Sci­ence in the Pub­lic In­terest, a watch­dog group, data on serving size comes from a 20-year old sur­vey (again, us­ing self-re­por­ted data) from the Ag­ri­cul­ture De­part­ment. “Four­teen new­er food con­sump­tion data show that con­sumers are eat­ing lar­ger por­tion sizes than they did in the 1970s and 1980s,” the cen­ter wrote in its 2010 case for nu­tri­tion facts re­form.

For in­stance, un­der cur­rent reg­u­la­tions, a serving size of ice cream is half a cup. Try scoop­ing only that amount. 

4. In­form­a­tion can be over­whelm­ing.

You can give people num­bers, but those num­bers may not be sa­li­ent in terms of chan­ging be­ha­vi­or. 

In the 2008 re­port, GAO found that con­sumers “gen­er­ally found nu­tri­tion la­beling con­fus­ing.” And while 30 per­cent of con­sumers say they will look at a nu­tri­tion la­bel when pur­chas­ing an item, an eye-track­ing study found that only 9 per­cent will look at the cal­or­ie counts when read­ing a la­bel.

In 2003, the FDA star­ted re­search­ing the prob­lem to try to make the nu­tri­tion in­form­a­tion more sa­li­ent. “One con­sist­ent find­ing … was that par­ti­cipants don’t like to do math, and they of­ten make mis­takes when de­term­in­ing the cal­or­ic and nu­tri­tion­al con­tent of pack­aged food,” Amy Lando, an FDA re­search­er said in a pod­cast. The agency found that con­sumers were likely to make bet­ter de­cisions when a serving size re­flec­ted the con­tain­er size or when the la­bel had an ad­di­tion­al column with the en­tire cal­or­ic count of the con­tain­er (which may be a hint as to what the new la­bels will look like).

“It ap­pears that in­form­a­tion is not one of the main obstacles to people eat­ing well,” Ju­lie Downs, a re­search­er at Carne­gie Mel­lon Uni­versity, told me. She led a study that found no changes in be­ha­vi­or after the im­ple­ment­a­tion of cal­or­ie counts in a fast-food res­taur­ant. While that’s not the same thing as buy­ing gro­cer­ies, she thinks the concept car­ries over.

We don’t think of food in terms of num­bers, and it’s hard to put 300 cal­or­ies in the con­text of an en­tire day. It’s even harder, con­sid­er­ing that in­di­vidu­al factors play a huge role. As Arch­er ex­plains, “Put­ting a num­ber on a box isn’t go­ing to tell that per­son what their body is go­ing to do with the food.” For some people, so­di­um is a prob­lem, ex­acer­bat­ing high blood pres­sure. For oth­ers, it’s not. Those who are phys­ic­ally act­ive meta­bol­ize food dif­fer­ently than those who are more sedent­ary. And we haven’t even waded in­to the de­bate over wheth­er all cal­or­ies are cre­ated equal.

The new re­com­mend­a­tions will be re­leased to­mor­row. Good luck, FDA.

1. There's no one standard way to measure a calorie.

Say you are look­ing at two bags of chips. Both weigh the same, and pre­sum­ably con­tain the same amount of chips. One bag has 300 cal­or­ies, the oth­er has 360. Which one ac­tu­ally has the lower cal­or­ic amount? Right now, it’s im­possible to tell.

That’s be­cause the FDA al­lows for five dif­fer­ent ways to meas­ure total cal­or­ies and al­lows for a mar­gin of er­ror of up to 20 per­cent. Ba­sic­ally, a man­u­fac­turer can either use pre­de­ter­mined con­ver­sions (one gram of car­bo­hydrate is four cal­or­ies) or burn the product to see how much en­ergy that comes off. These will provide dif­fer­ent val­ues.

And the FDA doesn’t have the greatest track re­cord of over­sight on the la­bel val­ues. In 2008, a Gov­ern­ment Ac­count­ab­il­ity Of­fice re­port found “FDA’s test­ing of nu­tri­tion in­form­a­tion has been lim­ited and has found vary­ing de­grees of com­pli­ance.” Al­though 87 per­cent to 91 per­cent of la­bels were in an al­low­able range in an audit, “com­pli­ance rates var­ied sig­ni­fic­antly for a few nu­tri­ents.” Only 47 per­cent of la­bels were in the cor­rect range for the amount of Vit­am­in A they con­tained.

2. Percent daily values are based on very shoddy data

Much about what we “know” about nu­tri­tion comes from a four-dec­ades-deep data­set called the Na­tion­al Health and Nu­tri­tion Ex­am­in­a­tion Sur­vey, or NHANES, which is com­prised of self-re­por­ted data. And from this, ba­sic cor­rel­a­tions emerge, like how high-fat di­ets are linked to high­er weights, and so on. The sur­vey has also giv­en us a pic­ture of what the av­er­age hu­man eats. Those val­ues, in turn, are re­flec­ted in nu­tri­tion policy.

But it turns out that people are hor­rible about ac­cur­ately re­port­ing their food in­take.

“To say they [the data] are im­per­fect is the equi­val­ent of say­ing the Ti­tan­ic had a float­a­tion prob­lem or a buoy­ancy prob­lem. These data should not be used,” says Ed­ward Arch­er, an obesity re­search­er, who au­thored a sci­entif­ic take­down of the NHANES data in the Journ­al PLOS One.

That’s be­cause people would com­monly re­port levels of food that are in­cap­able of main­tain­ing hu­man life. “In no sur­vey did at least 50 per­cent of the re­spond­ents re­port plaus­ible [en­ergy in­take] val­ues,” Arch­er and his coau­thors con­cluded in their study.

3. Serving sizes do not reflect reality.

Ac­cord­ing to the Cen­ter for Sci­ence in the Pub­lic In­terest, a watch­dog group, data on serving size comes from a 20-year old sur­vey (again, us­ing self-re­por­ted data) from the Ag­ri­cul­ture De­part­ment. “Four­teen new­er food con­sump­tion data show that con­sumers are eat­ing lar­ger por­tion sizes than they did in the 1970s and 1980s,” the cen­ter wrote in its 2010 case for nu­tri­tion facts re­form.

For in­stance, un­der cur­rent reg­u­la­tions, a serving size of ice cream is half a cup. Try scoop­ing only that amount. 

4. Information can be overwhelming.

You can give people num­bers, but those num­bers may not be sa­li­ent in terms of chan­ging be­ha­vi­or. 

In the 2008 re­port, GAO found that con­sumers “gen­er­ally found nu­tri­tion la­beling con­fus­ing.” And while 30 per­cent of con­sumers say they will look at a nu­tri­tion la­bel when pur­chas­ing an item, an eye-track­ing study found that only 9 per­cent will look at the cal­or­ie counts when read­ing a la­bel.

In 2003, the FDA star­ted re­search­ing the prob­lem to try to make the nu­tri­tion in­form­a­tion more sa­li­ent. “One con­sist­ent find­ing … was that par­ti­cipants don’t like to do math, and they of­ten make mis­takes when de­term­in­ing the cal­or­ic and nu­tri­tion­al con­tent of pack­aged food,” Amy Lando, an FDA re­search­er said in a pod­cast. The agency found that con­sumers were likely to make bet­ter de­cisions when a serving size re­flec­ted the con­tain­er size or when the la­bel had an ad­di­tion­al column with the en­tire cal­or­ic count of the con­tain­er (which may be a hint as to what the new la­bels will look like).

“It ap­pears that in­form­a­tion is not one of the main obstacles to people eat­ing well,” Ju­lie Downs, a re­search­er at Carne­gie Mel­lon Uni­versity, told me. She led a study that found no changes in be­ha­vi­or after the im­ple­ment­a­tion of cal­or­ie counts in a fast-food res­taur­ant. While that’s not the same thing as buy­ing gro­cer­ies, she thinks the concept car­ries over.

We don’t think of food in terms of num­bers, and it’s hard to put 300 cal­or­ies in the con­text of an en­tire day. It’s even harder, con­sid­er­ing that in­di­vidu­al factors play a huge role. As Arch­er ex­plains, “Put­ting a num­ber on a box isn’t go­ing to tell that per­son what their body is go­ing to do with the food.” For some people, so­di­um is a prob­lem, ex­acer­bat­ing high blood pres­sure. For oth­ers, it’s not. Those who are phys­ic­ally act­ive meta­bol­ize food dif­fer­ently than those who are more sedent­ary. And we haven’t even waded in­to the de­bate over wheth­er all cal­or­ies are cre­ated equal.

The new re­com­mend­a­tions will be re­leased to­mor­row. Good luck, FDA.

What We're Following See More »
SANDERS UP TEN POINTS
Trump Leads Tightly Packed Group Vying for Second
2 hours ago
THE LATEST

In one of the last surveys before New Hampshirites actually vote, a Monmouth poll has Donald Trump with a big edge on the Republican field. His 30% leads a cluster of rivals in the low-to-mid teens, including John Kasich (14%), Jeb Bush and Marco Rubio (13% each) and Ted Cruz (12%). On the Democratic side, Bernie Sanders leads Hillary Clinton 52%-42%.

Source:
×