If you're new here, you may want to subscribe to my RSS feed. Thanks for visiting!
During one of our many recent conversations about nutrition and fitness, my wife wondered why food labels always had whole numbers for calorie content. Further investigation into the foods in our cupboards showed that the calorie contents are rounded even more than to the nearest whole number–in fact, they appeared to be rounded to the nearest whole number divisible by 5!
The table below shows a comparison between the calories listed on specific food labels found in my kitchen and the calculated calories. I based my calculations on the following criteria:
- Fat contains 9 calories/gram.
- Protein contains 4 calories/gram.
- Carbohydrate contains 4 calories/gram.
- I didn’t pay attention to the dietary fiber–which is sometimes subtracted from the total carbohydrate content to make food labels look better than they actually are.
Comparison between the calories listed on the FDA nutrition labels and a calculation based on grams of fat, carbohydrate, and protein listed on the labels
| Food | Protein Grams | Carb Grams | Fat Grams | Listed Calories | Computed Calories | Variance |
|---|---|---|---|---|---|---|
| Fred Meyer Yogurt | 10 | 47 | 2.5 | 250 | 250.5 | 0.20% |
| Tree Top Applesauce | 0 | 18 | 0 | 70 | 72 | 2.86% |
| Arthur's Parmesan Cheese | 2 | 0 | 1.5 | 20 | 21.5 | 7.50% |
| Whole Milk | 8 | 12 | 8 | 150 | 152 | 1.33% |
| Blue Diamond Smokehouse Almonds | 6 | 5 | 16 | 170 | 188 | 10.59% |
| Campbells Bean & Ham Soup | 11 | 30 | 2 | 180 | 182 | 1.11% |
| S&W Garbanzo Beans | 7 | 19 | 1 | 80 | 113 | 41.25% |
| Bisquick | 3 | 26 | 5 | 160 | 161 | 0.63% |
| Ritz Crackers | 1 | 10 | 4.5 | 80 | 84.5 | 5.63% |
| Nestle White Chocolate Morsels | 0 | 9 | 4 | 70 | 72 | 2.86% |
I was amazed to see how different my calculations were compared to the listed values on the FDA’s nutrition labels. They ranged from .2% all the way to 41.25%. The garbanzo beans by far had the biggest discrepancy and I’m sure that’s due to the fiber content of the beans. I’m not a nutritionist, so I don’t know how much of that fiber is digested or not, but I’m assuming some of it is.
My cupboard research left me with one question: What are the guidelines provided by the FDA to food companies regarding the rounding of their calorie labels? After a long search, I was able to find this. Here are their guidelines in a nutshell:
- If the food has less than 5 calories/serving, they can label it as 0 calories.
- If the food has between 5 and 50 calories, they can round it to the nearest 5 calories.
- If the food has greater than 50 calories, they can round it to the nearest 10 calories.
So unless my understanding of rounding is incorrect, the biggest number of calories/serving that a food can be off by is 5. I wonder how S&W gets by with a 53 calorie error? Either they’re wrong, or I need to research (and eat) more about fiber.
Addendum: The Garbanzo beans contain 5 g of fiber per serving. If you subtract the 5 grams from the carbohydrate listing, you still get 14 grams of carbs/serving. Based on my calculations, there are still 93 calories/serving versus the listing of 80 calories/serving for a % difference of 16.25 %.
J.D.
13 responses so far ↓
1 Red // Feb 22, 2008 at 12:35 pm
This combined with the occasional meal at a non-calorie listing restaurant really makes figuring out what you’re consuming a real shot in the dark.
2 Corey // Feb 22, 2008 at 1:22 pm
To help you out a little here, this is my understanding of the issue at hand:
Dietary fiber are complex carbohydrates that aren’t digested, and hence yield no caloric content. In addition to this, many legumes (beans) also have simple carbohydrates that act similarly to fiber in that they are also non-digestible.
I hope this helps clarify things for everyone. Essentially, foods high in fiber are _very good_ for you. They have numerous health benefits, and can help with proper weight maintenance.
3 Di // Feb 22, 2008 at 2:00 pm
Its my understanding that the manufacturers don’t use the standard table of cal content per fat/protein/carb gram, instead the food sample gets scientifically tested. Maybe that counts for the discrepancy?
4 Uncle Midriff // Feb 22, 2008 at 2:04 pm
Do we actually need to be counting calories this closely?
My understanding was that, due to the fact that food manufacturing and packaging isn’t all that precise, the calorie information is just there to give you a general idea of what to expect.
As an example, the brand of tortilla chips I buy says that a serving is “about 12 chips.” That is nice to know, but it’s also very imprecise. Even disregarding the whole “about” bit, if the twelve chips I pull out of the bag happen to be slightly bigger or thicker than average due to some kind of manufacturing error, I’ll be getting more than the listed number of calories. There’s just no practical way to know *exactly* how many calories you’re consuming.
I agree that food manufacturers should strive for accuracy when labeling their food, but given that the whole endeavor is pretty imprecise, there’s going to be some inconsistencies.
And even in your impromptu study, the biggest difference between your calculations and the label is 33 calories, and I’d argue that that piece of data should be thrown out because you yourself admit that you’re not sure you calculated it correctly. Regardless, the average difference between what’s listed and what you calculated is 6.47 calories. Without the beans, the average difference falls to about 3.52 calories. I fairly certain that even if everything I ate in a given day had 3.52 more calories than I was expecting, or even 6.47 more calories than I was expecting, it wouldn’t wreck my diet.
5 Zulu // Feb 22, 2008 at 2:36 pm
As a mathematical exercise and look into how the FDA guidelines are setup, it’s interesting.
But, I have to question the assumptions the whole thing is based on. Aren’t there different types of Fats, Proteins and Carbohydrates that will all have different calorie yields? I must confess ignorance to the matter at hand though.
6 Anne Keckler, Personal Trainer // Feb 22, 2008 at 3:03 pm
Most of those discrepancies are too small to worry about. For one thing, the macronutrients could be rounded up, so there might be slightly less than what the label indicates, which would account for some of the differences in calorie counts.
The beans, however, make no sense at all. I’ve tried calculating it different ways, and it just doesn’t add up. The only other time I have encountered a large discrepancy was when I looked at the nutrition page on the website of a favorite fast food place.
When it comes to fiber, keep in mind that there is soluble fiber and insoluble fiber.
Even though it’s not absolutely precise, I still recommend counting calories if someone is trying to lose weight and it’s not happening. It’s a good way to become aware of the caloric content of the foods we eat, so we know where to make changes.
7 BabyPop // Feb 22, 2008 at 3:18 pm
Holy moly! I had no idea…wonder if that could be part of the reason why I’m staying within my suggested calorie range for weight loss but still not losing?
8 Corey // Feb 22, 2008 at 3:18 pm
Anne, refer to my post concerning the “bean problem.” Also, no dietary fiber is digested. Soluble fiber just means that it is water soluble, so it breaks down in the large intestines to promote healthy activities; insoluble fiber generally just helps the “flow” of things.
Remember that nutrition labels are regulated; there aren’t any absurd lies on the labels, they are merely rounded for convenience in a similar manner to asking someone the time.
9 mrs darling // Feb 22, 2008 at 11:07 pm
All I know is that when I keep track of my calories I lose weight. Even if I am a few calories off it doesnt matter. If I eat less than I was eating before than I lose.
Ive been trcking my food intake since the 7th of January and Ive lost 18 pounds. Im happy with that regardless of the conflicting information.
10 Jill // Feb 23, 2008 at 8:19 am
I think that where this is most useful is when you’re thinking of consuming 2 or 3 servings of something– that’s when all that rounding can really add up and come back to haunt you. I remember running up against this when I tried Weight Watchers. 1 of portion of a food would be 1 “point”, then suddenly 2 portions of the same food would be 3 “points” due to the way NI was rounded up or down– which was always an unpleasant shock, but it served as a good reminder to keep portions under control.
11 monica // Feb 23, 2008 at 4:04 pm
As others have noted, high-fiber foods are typically calculated as 4-4-9 calories per gram of protein, carbohydrate
minus insoluble fiber, and fat. This is based on the Atwater system for determining energy values. However, Atwater’s original paper highlights that the actual number of grams per nutrient varies by food (e.g. eggs have 4.5 kcal per gram of protein while fruits have about 3.95 kcal per gram of protein). The 4-4-9 are based on averages. (The USDA has a link to the paper here.)
However, the S&W chickpea thing is downright wrong. The USDA food database record for 130g of “Chickpeas (garbanzo beans, bengal gram), mature seeds, canned” shows
1.4g fat
29.4g carbs
6.4g protein
154.7 kcal
Looks like the S&W folks are under-valueing their carbs! Somebody should tell their labelling manager that Dr. Atkins died and they should stop being scared of carb-haters!
On a different note, is it really “healthy” to scrutinize everything you eat to this level? This exercise seems to invite obsessive compulsive behavior in a system that is riddled with imprecision . The value in recording food is less about precise quantities and more about relative measurements (kind of like weight, right?).
12 Shanti @ Antishay Ventenne // Feb 27, 2008 at 6:24 pm
Great post! I have always known that there is some variance, but different companies will hold themselves to a higher standard.
I worked for a couple of years running a business that made cookies for Starbucks. While incredibly lucrative, the time it took to deal with Sbux almost made it not worth it. They are completely anal about everything (well, they have to be, I guess). They required that your nutritional content label have a variance of no more than 2 calories! We gathered cookies once a day for testing so that by the time they were all made and ready for labeling (a few weeks later), we would have actual results for the cookies that we could present with almost no variance.
I know that SBUX food that’s in the display window doesn’t have labels, but the stores are required to provide nutritional information upon request, and it’s all listed online.
13 Carol // Mar 1, 2008 at 1:31 pm
I’m with a company that does nutrition analysis of food products using database software. This is the most common (and cheapest) method of analysis for smaller food companies. Larger companies send the product to a lab for chemical analysis (more precise and much more expensive). Either method can and does often result in errors, usually human errors — e.g., when someone typed the wrong number in the wrong place and no one with knowledge of nutrition reviewed the data before it got printed. It is VERY common, so don’t expect too much accuracy from food labels (restaurant data is even more prone to errors and lack of accuracy). This is why a few calories up or down is not significant (not to mention that your body doesn’t notice 5-10 calorie differences… it’s so small). There is also a lot of variation in caloric content in just about every food in nature — one apple will have more sugar/calories than another of the same size depending on ripeness, variety, etc.
Leave a Comment