FDA Label Tolerances

Hi everyone,

So this is my first post, but I've been reading this forum for a while and have really learned a lot. I've not been in QA for long, but I recently became PCQI certified and have working to bring our Food Safety plan up to FSMA standards. I have a very specific question about product labeling. We produce savory snacks (chips, popcorn) and we are in the process of setting up labels for a new customer. The problem is that we have been getting variations in our macros even on products of the same batch, specifically in the protein. We mix the product with a protein (whey) powder, but the problem is that the process we use makes it almost impossible to get the exact same distribution of powder across each batch, so when we send it out for testing, different packages within a batch can differ by as much as 0.4 gram per serving in protein, which makes a difference because it would two different label results. (The average across batches are the same.) My boss has said he heard from another QA person that the FDA allows larger tolerances for things like our products because of the greater chances for variability. Does anyone know if this is true, and, if so where I can find it in the Code of Federal regulations?

The only thing I can find is from 21 CFR 101.9 (g), but it simply says the claims per serving for protein for Class II must be 80 percent of the value declared on the label. Since none of our ingredients contain added nutrients, we should classify as Class II, and, as I understand it, as long as we have 12 subsamples to be taken from 12 randomly chosen cases to be representative of a lot that fit the rounding requirements set out elsewhere in the Code, we should be good. 

Anyone see any problem with that? I would appreciate any assistance. 

kc

(ii) When a vitamin, mineral, protein, total carbohydrate, polyunsaturated or monounsaturated fat, or dietary fiber meets the definition of a Class II nutrient, the nutrient content of the composite must be at least equal to 80 percent of the value for that nutrient declared on the label. Provided, That no regulatory action will be based on a determination of a nutrient value that falls below this level by a factor less than the variability generally recognized for the analytical method used in that food at the level involved.

My read of FDA thinking on these things is that certain nutrients are "good", such as protein and fiber, and you can round up for those. Nutrients that are "bad" such as cholesterol can be rounded down. Calories can be rounded to the nearest 5; that is your calorie number can end with 0 or 5.

And as can be seen from the text, they make an allowance for quirks in the analytical methods used, which will include variance.

Thanks Hank,

What I understand from the code is that between 2 and 5 grams of protein (which is what I am concerned with now) are rounded in 0.5 gram increments. I'm just wondering about the tolerances. I think I'm reading the code right; if any one else has dealt with this specifically, I would appreciate it. 

(ii) The gram or milliliter quantity equivalent to the household measure should be rounded to the nearest whole number except for quantities that are less than 5 g (mL). The gram (mL) quantity between 2 and 5 g (mL) should be rounded to the nearest 0.5 g (mL) and the g (mL) quantity less than 2 g (mL) should be expressed in 0.1-g (mL) increments.

***

7) "Protein": A statement of the number of grams of protein in a serving, expressed to the nearest gram, except that if a serving contains less than 1 gram, the statement "Contains less than 1 gram" or "less than 1 gram" may be used as an alternative, and if the serving contains less than 0.5 gram, the content may be expressed as zero. When the protein in foods represented or purported to be for adults and children 4 or more years of age has a protein quality value that is a protein digestibility-corrected amino acid score of less than 20 expressed as a percent, or when the protein in a food represented or purported to be for children greater than 1 but less than 4 years of age has a protein quality value that is a protein digestibility-corrected amino acid score of less than 40 expressed as a percent, either of the following shall be placed adjacent to the declaration of protein content by weight: The statement "not a significant source of protein," or a listing aligned under the column headed "Percent Daily Value" of the corrected amount of protein per serving, as determined in paragraph ©(7)(ii) of this section

FDA Rounding rules -

Thanks KfromIA,

I'm aware of that document, but hopefully it will be helpful for others. Just so everyone understands, I'm trying to verify whether what my boss heard from another QA person (that different products have different tolerances) is true.

I don't think it is, as I can't find it anywhere in Code. All I can find in CFR 101.9 is the distinction between Class I and Class II nutrients. Class I defines added nutrients, and Class II defines naturally occurring nutrients. 

3) Two classes of nutrients are defined for purposes of compliance:

(i) Class I. Added nutrients in fortified or fabricated foods; and

(ii) Class II. Naturally occurring (indigenous) nutrients. When a nutrient is naturally occurring (indigenous) in a food or an ingredient that is added to a food, the total amount of such nutrient in the final food product is subject to class II requirements, except that when an exogenous source of the nutrient is also added to the final food product, the total amount of the nutrient in the final food product (indigenous and exogenous) is subject to class I requirements.

The rounding requirements for these two classes are different. The former must have all samples meet the minimum label claims for, say protein; the latter needs to have 12 random samples meed 80 percent of say, claims for protein. 

(i) When a vitamin, mineral, protein, or dietary fiber meets the definition of a Class I nutrient, the nutrient content of the composite must be formulated to be at least equal to the value for that nutrient declared on the label.

(ii) When a vitamin, mineral, protein, total carbohydrate, polyunsaturated or monounsaturated fat, or dietary fiber meets the definition of a Class II nutrient, the nutrient content of the composite must be at least equal to 80 percent of the value for that nutrient declared on the label. Provided, That no regulatory action will be based on a determination of a nutrient value that falls below this level by a factor less than the variability generally recognized for the analytical method used in that food at the level involved.

I guess I'm asking whether anyone knows of anything different in actual code. (I don't).

My read is that it means you can be off by 20% from your label's statement due to natural variation. To me it means the FDA and the public doesn't really care.  Every decade or so, a news outlet will report on how the food labels are largely incorrect, and then nothing will happen because nobody cares.

Thanks KfromIA,

 

I'm aware of that document, but hopefully it will be helpful for others. Just so everyone understands, I'm trying to verify whether what my boss heard from another QA person (that different products have different tolerances) is true.

 

I don't think it is, as I can't find it anywhere in Code. All I can find in CFR 101.9 is the distinction between Class I and Class II nutrients. Class I defines added nutrients, and Class II defines naturally occurring nutrients. 

 

3) Two classes of nutrients are defined for purposes of compliance:

(i) Class I. Added nutrients in fortified or fabricated foods; and

(ii) Class II. Naturally occurring (indigenous) nutrients. When a nutrient is naturally occurring (indigenous) in a food or an ingredient that is added to a food, the total amount of such nutrient in the final food product is subject to class II requirements, except that when an exogenous source of the nutrient is also added to the final food product, the total amount of the nutrient in the final food product (indigenous and exogenous) is subject to class I requirements.

 

The rounding requirements for these two classes are different. The former must have all samples meet the minimum label claims for, say protein; the latter needs to have 12 random samples meed 80 percent of say, claims for protein. 

 

(i) When a vitamin, mineral, protein, or dietary fiber meets the definition of a Class I nutrient, the nutrient content of the composite must be formulated to be at least equal to the value for that nutrient declared on the label.

(ii) When a vitamin, mineral, protein, total carbohydrate, polyunsaturated or monounsaturated fat, or dietary fiber meets the definition of a Class II nutrient, the nutrient content of the composite must be at least equal to 80 percent of the value for that nutrient declared on the label. Provided, That no regulatory action will be based on a determination of a nutrient value that falls below this level by a factor less than the variability generally recognized for the analytical method used in that food at the level involved.

 

I guess I'm asking whether anyone knows of anything different in actual code. (I don't).

I don't know of any either. Like Hank Major said. Do the best you can. I have a feeling if the FDA questions anything- they'll go after labels with health claims. 

No, what your boss said is NOT true.  Doesn't matter the type of product or process used to create the product.  All must follow the FDA labeling regulations.

Thanks all. And yes, KfromIA, I will do the best I can. 

I have always rounded up on the bad nutrients (potentially over-declaring) and rounded down on the good nutrients (potentially under-declaring). That way any errors would be to the consumer’s favor. Make sure you round to the proper level (.5g, 10 mg, or whatever applies to the specific nutrient), and FDA does allow different tolerances for different nutrients.