Nutrition Labels: What You Don’t Know Could Hurt You


Nutrition LabelsI’ve recently done a bunch of research into nutrition labels — you know, the “Nutrition Facts” panel that appears on the back of most packaged foods. I’m doing it for the book I’m writing about vitamins, but my obsessive dive into the details of FDA labeling regulations revealed certain facts about nutrition labels that I think all people with diabetes should be aware of. Prepare to be freaked out.

First, the basic back story: the Nutrition Facts panels, as you know them today, are the result of the 1990 Nutrition Labeling and Education Act, which required nutritional information to be printed on nearly all packaged foods. (Before that, such labeling was voluntary unless the food made a health claim.) Every bit of that label has been discussed and debated, from the font and size for “Nutrition Facts” (Franklin Gothic Heavy or Helvetica Black, flush left and flush right, no smaller than 13 point) to the thickness of the bar that separates calories from the rest of the DVs (3 point) to the fact that the ingredients are printed in all caps. The result is a pleasing consistency in appearance between labels, which presumably makes the information easier to process and understand. 

But here’s the question: where does that information actually come from? And how accurate is it? For anyone who counts carbohydrates, that second question is obviously particularly important.

So you may be surprised (and, if you’re like me, frustrated and a little scared) to find out that when it comes to carbohydrates, the FDA’s rules state that they must be present in the food at “at 80 percent or more” of the amount listed on the label.  They’re not supposed to exist in amounts beyond a “reasonable excess,” but that term is not defined.

If you do the math, this means that a food that says it has 20 grams of carbohydrate could be anywhere from 16 to, well, whatever the manufacturer considers a “reasonable excess” to be. That may not seem like a lot, but the possible inaccuracy is greater the more grams of carbohydrate the food contains. Say you wanted to eat a bagel — which, if it’s 4.5 inches across, supposedly contains about 70 grams of carb. Well, with a 20 percent margin of error at the lower end, that means that you bagel could contain 14 fewer grams of carb and the label would still be considered accurate. So that 70-gram carb bomb may actually contain anywhere from 56 grams of carb to, again, who knows? If you use a 1:15 insulin to carb ratio and eat that bagel when your blood sugar is 100 mg/dl, you can see how you might run into some serious problems. (Side note: I am not advocating the consumption of bagels. Eating 70 grams of carb at one time would screw me up for the rest of the day.)

Worried yet? Here’s another thing: the accuracy of the labels is the responsibility of the food manufacturer, not the FDA. The FDA doesn’t have the time or resources necessary to measure the nutritional profile of every food available in the grocery store — I get that. But the result is that even if you take the carb counts on the label with the FDA-mandated grain of salt, there’s still no assurance that you can trust that the food falls within range. That only works if the company is actually playing by the rules (or, to be a little more gracious about it, if their analytical methods and/or food nutrient databases, which are often proprietary, are actually accurate). 

I think what really gets me about this isn’t the labels themselves — though I am dismayed to know that in this day and age of prediabetes and Type 2 diabetes, there’s no upper limit on carbs. What gets me about it is that none of the diabetes educators or nutritionists or endocrinologists I have seen since I was diagnosed nearly twelve years ago has ever mentioned this to me. Instead, I repeatedly have heard the message that if I just count my carbs right and figure out my insulin-to-carb ratio, I will have perfect control of my blood glucose values. The translation of this message: if carb counting does not seem to work for me — which it doesn’t — it is not a fault of the method itself; it is because I am doing something wrong. 

This is something I think that diabetes educators need to be aware of and to educate their patients about, because it might alleviate a bit of guilt people feel when they try to do everything “right,” and yet their blood glucose values don’t seem to behave. As we all know, there are myriad variables in our bodies themselves that affect blood glucose values — stress, sleep, time of day, hormones, etc. But when you throw in the fact that the one variable that’s supposed to be set — the number of carbs in a serving of a packaged food — is actually not set at all, well, it’s enough to make you go a little crazy. By raising awareness of this, I hope that more diabetes educators, nutritionists and doctors will adopt the attitude held by my current, great endocrinologist: that yes, managing diabetes is your responsibility, but it is an art, not a science. 

Lastly, now that I’ve gotten that out of my system, here are some other things about the  Nutrition Facts that you might be interested to know: 

– calories are allowed to be present at up to 120 percent of the value listed on the label. So that aforementioned bagel, which says it has 360 calories, could have 432 and still be considered legit. (And if you expand this further, it means that your whole diet could have 20 percent more calories than you think.)

– the Daily Values for micronutrients like vitamins and minerals are based on the Recommended Dietary Allowances (RDAs) from 1968. Yes, nearly a half a century ago. Suffice it to say that nutritional understanding has changed a wee bit since then, and that the Food and Nutrition Board at the Institute of Medicine, which is the organization that actually sets the RDAs (though they’re not even called RDAs anymore — now the FNB uses a more nuanced set of values called Dietary Reference Intakes) has come out with numerous updates since then. (The most recent were the ones for calcium and Vitamin D, published in 2010.) 

I could go on, but the bottom line, in terms of diabetes, is this: you can’t trust the accuracy of the carbohydrate count on the nutrition labels. They’re better than nothing, for sure. But if you find that your best efforts at diabetic math aren’t working the way you want them to, I hope that this will give you at least a bit of comfort: as is often the case when you try to turn your body into a word problem, it’s not necessarily your fault if the answer comes out wrong. 

Leave a Reply

8 Comments on "Nutrition Labels: What You Don’t Know Could Hurt You"

Notify of
Sort by:   newest | oldest | most voted
Jonathan Lindberg
I work in the food industry, and specifically I develop the Nutrition Facts panels that are put onto our packaging. We are way MORE accurate than 20% – but you are right about the way the “Food Labeling Act” is written, there is wiggle room.  For our products, we know down to the 0.000X gram, exactly how much of each ingredient is in the final product. And that is how we calculate our nutrition facts. (The website link I provided talks about how “sneaky” trans fat labels can be.) At least for our company, there is never any intention to… Read more »
A belated response to Dina: most of the information in those databases and apps is derived from the USDA’s food database: .  I need to do more research on this but my understanding is that those values are based on an average of several different samples of the food at hand. I would think that in the case of, say, a particular cracker, this would mean that they are relatively accurate. But when you have foods with more variability — different varieties of apples or bananas, for example — then they’re not going to be as spot-on. Not only… Read more »

Do you think the inaccuracy in carb values is also present in books and apps that give you the nutritional values of foods? For example, I use Calorie King.


Great post.  I have read about the 20% rule before and it did freak me out.  And what I immediately thought of was the 20% rule that exists for glucometers. I’m sorry if I made you more frustrated.

Scott K. Johnson

Brilliant post, Catherine.  Thank you!


Scott — that’s a great way to phrase it. Carb counting really is the “best of the worst” of blood glucose management methods. I’m definitely aware of how many grams of carb I am (presumably) consuming, but I also know that the very same number can affect me differently day by day. Thanks for the comment!

The prevailing attitude among diabetes educators is that if a patient follows the rules, as you state, “if carb counting does not seem to work for me — which it doesn’t — it is not a fault of the method itself; it is because I am doing something wrong” … part of this is because diabetes educators have been told that carb counting is, in fact, scientifically proven.  However, the evidence suggests that while glycemic management is improved with carb counting (especially relative to alternative management options), there is actually slim evidence that glucose is truly “controlled” with carb counting.… Read more »
Ed Terry

My favorite example of misleading nutrition labeling is “Wheaties Fuel”. In a 55 g serving, the label states that it only has 14 g sugars. But if you examine the list of ingredients, you’ll see, in addition to Wheat, Rice, Oats, Corn Bran, Wheat bran, the following ingredients: Honey, Maltodextrin, Corn Starch, Brown Sugar Syrup, Corn Syrup Solids. These last 5 ingredients break down to sugar instantly as soon as they hit the gut.

Copyright © 2009-2016 Diabetes Media Foundation, All Rights Reserved.
ASweetLife™ is a trademark of the Diabetes Media Foundation, All Rights Reserved.