Are nutrition labels really accurate?

Options
2

Replies

  • Sued0nim
    Sued0nim Posts: 17,456 Member
    Options
    So in other words, the inaccuracy of the labels is going to drive people who weigh their food batty, but the rest of us don't have to worry about it because we're allowing for broad inaccuracy anyway.

    No it's not, why would it?

    Look we get it ...you set your calories at an arbitrarily low level to account for the fact that you don't weigh because weighing is inaccurate and it works for you

    You know what that's the point, find what works for you

    When people come on complaining they're not losing even though they're sticking to their calories it is easy to see that they in fact are fooling themselves in some way ...because if they were they would lose ...the retraining comes in removing as many variables as possible in this inaccurate world of estimation ...giving them a learning curve and then they can also choose to weigh / not weigh but they have a tool to get them on track

    Just because you don't weigh doesn't make you better at this game than those who do

    Just because it works for you at your arbitrary calorie number is no reason to tell others you need to cut calories ...if they are struggling they need to understand why and how to fix it before making their own decision on how to keep going.

  • TimothyFish
    TimothyFish Posts: 4,925 Member
    Options
    segacs wrote: »
    Hornsby wrote: »
    Food labels can be overestimated by 20% but can't be underestimated.

    That's not how statistical variance works. They can sometimes be underestimated, too.

    Yeah, I agree. Just picking fruit a little greener than normal can result in it having fewer calories rather than more calories. The companies that use the fruit in their products have no way of controlling that, which is why the wide variance is allowed.
  • Hornsby
    Hornsby Posts: 10,322 Member
    Options
    segacs wrote: »
    Hornsby wrote: »
    Food labels can be overestimated by 20% but can't be underestimated.

    That's not how statistical variance works. They can sometimes be underestimated, too.

    Awwe, okay. I just meant by law.

  • stealthq
    stealthq Posts: 4,298 Member
    edited January 2015
    Options
    Hornsby wrote: »
    segacs wrote: »
    Hornsby wrote: »
    Food labels can be overestimated by 20% but can't be underestimated.

    That's not how statistical variance works. They can sometimes be underestimated, too.

    Awwe, okay. I just meant by law.

    It's only certain nutrients that are not allowed to be over-represented on the label (by law). Not everything. And definitely not calories.

    ETA: I think I remember the nutrients that could not be over-represented on the label (but could be under-represented) were supplements like added vitamin D in milk, but I can't find the right FDA CFR document to confirm. Oh well.
  • lemurcat12
    lemurcat12 Posts: 30,886 Member
    edited January 2015
    Options
    Francl27 wrote: »
    Or you can eat mostly fresh food, then you don't have to worry about it too much. But it explains why a lot of people don't lose when they eat too much processed food (plus, you know, the products themselves are often not the same weight than what the label tells you anyway).

    Unfortunately, not eating exercise calories and leaving 100 extra calories as a buffer is just not possible as you get closer to your goal, as having a big deficit can be detrimental to your health.

    +1

    Also, it doesn't make a lot of sense if you are doing a LOT of intense exercise, like longer runs. There could be a huge difference between your real TDEE and your TDEE if sedentary.

    I weigh my food, but since I use the USDA numbers mostly I don't have a false idea that my numbers are perfectly accurate and don't really care. They'll average out.
  • Hornsby
    Hornsby Posts: 10,322 Member
    Options
    stealthq wrote: »
    Hornsby wrote: »
    segacs wrote: »
    Hornsby wrote: »
    Food labels can be overestimated by 20% but can't be underestimated.

    That's not how statistical variance works. They can sometimes be underestimated, too.

    Awwe, okay. I just meant by law.

    It's only certain nutrients that are not allowed to be over-represented on the label (by law). Not everything. And definitely not calories.

    ETA: I think I remember the nutrients that could not be over-represented on the label (but could be under-represented) were supplements like added vitamin D in milk, but I can't find the right FDA CFR document to confirm. Oh well.

    "The Third Group nutrients include calories, sugars, total fat, saturated fat, cholesterol, and sodium. However, for products (e.g., fruit drinks, juices, and confectioneries) with a sugars content of 90 percent or more of total carbohydrate, to prevent labeling anomalies due in part to rounding, FDA treats total carbohydrate as a Third Group nutrient instead of a Class II nutrient. For foods with label declarations of Third Group nutrients, the ratio between the amount obtained by laboratory analysis and the amount declared on the product label in the Nutrition Facts panel must be 120% or less, i.e., the label is considered to be out of compliance if the nutrient content of a composite of the product is greater than 20% above the value declared on the label. For example, if a laboratory analysis found 8 g of total fat/serving in a product that stated that it contained 6 g of total fat/serving, the ratio between the laboratory value and the label value would be (8 / 6) x 100 = 133%, and the product label would be considered to be out of compliance."

    http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformation/LabelingNutrition/ucm063113.htm


  • mustgetmuscles1
    mustgetmuscles1 Posts: 3,346 Member
    Options
    Close enough. Averages out over the long term as long as you are being honest and doing your best with your logging.
  • segacs
    segacs Posts: 4,599 Member
    Options
    Also beware of serving size tricks. Some products (e.g. cooking sprays) use a tiny serving size so that they can claim to have "zero" calories. Actually, anything that has between zero and five calories can claim to have zero calories. But since most people don't use a half-second spray but more like a 5-second spray, that "zero"-calorie spray could actually have 40-50 calories.

    Most misleading of all, IMHO, is when the package directions use a different serving size than the nutritional information. For instance, a box of dry pasta will sometimes give nutritional information for 40g, and then directions for "1 person" that tell you to cook up 80g or 100g. They really should be required to list nutritional information for a serving that a reasonable person would actually eat.
  • stevencloser
    stevencloser Posts: 8,911 Member
    edited January 2015
    Options
    Francl27 wrote: »
    Or you can eat mostly fresh food, then you don't have to worry about it too much. But it explains why a lot of people don't lose when they eat too much processed food (plus, you know, the products themselves are often not the same weight than what the label tells you anyway).

    Unfortunately, not eating exercise calories and leaving 100 extra calories as a buffer is just not possible as you get closer to your goal, as having a big deficit can be detrimental to your health.
    Fresh foods have exactly the same problem. The calorie counts you get for them are averages too, they can be higher or lower in the actual thing.
  • lemurcat12
    lemurcat12 Posts: 30,886 Member
    Options
    Hornsby wrote: »
    stealthq wrote: »
    Hornsby wrote: »
    segacs wrote: »
    Hornsby wrote: »
    Food labels can be overestimated by 20% but can't be underestimated.

    That's not how statistical variance works. They can sometimes be underestimated, too.

    Awwe, okay. I just meant by law.

    It's only certain nutrients that are not allowed to be over-represented on the label (by law). Not everything. And definitely not calories.

    ETA: I think I remember the nutrients that could not be over-represented on the label (but could be under-represented) were supplements like added vitamin D in milk, but I can't find the right FDA CFR document to confirm. Oh well.

    "The Third Group nutrients include calories, sugars, total fat, saturated fat, cholesterol, and sodium. However, for products (e.g., fruit drinks, juices, and confectioneries) with a sugars content of 90 percent or more of total carbohydrate, to prevent labeling anomalies due in part to rounding, FDA treats total carbohydrate as a Third Group nutrient instead of a Class II nutrient. For foods with label declarations of Third Group nutrients, the ratio between the amount obtained by laboratory analysis and the amount declared on the product label in the Nutrition Facts panel must be 120% or less, i.e., the label is considered to be out of compliance if the nutrient content of a composite of the product is greater than 20% above the value declared on the label. For example, if a laboratory analysis found 8 g of total fat/serving in a product that stated that it contained 6 g of total fat/serving, the ratio between the laboratory value and the label value would be (8 / 6) x 100 = 133%, and the product label would be considered to be out of compliance."

    http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformation/LabelingNutrition/ucm063113.htm


    I think you guys are saying the same thing.

    Certain nutrients (where we are particularly concerned that they not mislead you into thinking there are more whatever than there is) are supposed to be within a reasonable amount of the number when tested* but, specifically, must be at least 80% of the number given to be in compliance with the law.

    Others (and this includes calories), where we are particularly concerned that they not mislead you into thinking there is less whatever than there is, are also supposed to be within a reasonable amount of the number when tested* but, specifically, must not be more than 120% of it to be in compliance with the law.

    *Reasonable excesses of class I and II nutrients above labeled amounts and reasonable deficiencies of the Third Group nutrients are usually considered acceptable by the agency within good manufacturing practices.
  • Hornsby
    Hornsby Posts: 10,322 Member
    Options
    lemurcat12 wrote: »
    Hornsby wrote: »
    stealthq wrote: »
    Hornsby wrote: »
    segacs wrote: »
    Hornsby wrote: »
    Food labels can be overestimated by 20% but can't be underestimated.

    That's not how statistical variance works. They can sometimes be underestimated, too.

    Awwe, okay. I just meant by law.

    It's only certain nutrients that are not allowed to be over-represented on the label (by law). Not everything. And definitely not calories.

    ETA: I think I remember the nutrients that could not be over-represented on the label (but could be under-represented) were supplements like added vitamin D in milk, but I can't find the right FDA CFR document to confirm. Oh well.

    "The Third Group nutrients include calories, sugars, total fat, saturated fat, cholesterol, and sodium. However, for products (e.g., fruit drinks, juices, and confectioneries) with a sugars content of 90 percent or more of total carbohydrate, to prevent labeling anomalies due in part to rounding, FDA treats total carbohydrate as a Third Group nutrient instead of a Class II nutrient. For foods with label declarations of Third Group nutrients, the ratio between the amount obtained by laboratory analysis and the amount declared on the product label in the Nutrition Facts panel must be 120% or less, i.e., the label is considered to be out of compliance if the nutrient content of a composite of the product is greater than 20% above the value declared on the label. For example, if a laboratory analysis found 8 g of total fat/serving in a product that stated that it contained 6 g of total fat/serving, the ratio between the laboratory value and the label value would be (8 / 6) x 100 = 133%, and the product label would be considered to be out of compliance."

    http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformation/LabelingNutrition/ucm063113.htm


    I think you guys are saying the same thing.

    Certain nutrients (where we are particularly concerned that they not mislead you into thinking there are more whatever than there is) are supposed to be within a reasonable amount of the number when tested* but, specifically, must be at least 80% of the number given to be in compliance with the law.

    Others (and this includes calories), where we are particularly concerned that they not mislead you into thinking there is less whatever than there is, are also supposed to be within a reasonable amount of the number when tested* but, specifically, must not be more than 120% of it to be in compliance with the law.

    *Reasonable excesses of class I and II nutrients above labeled amounts and reasonable deficiencies of the Third Group nutrients are usually considered acceptable by the agency within good manufacturing practices.

    Agreed, I think I worded it wrong in my original post. I knew what I was saying, but just didn't say it right. lol.

  • merlewalton
    merlewalton Posts: 43 Member
    Options
    I work in food manufacturing and there are a lot of variables that can work their way into this situation. Are labels 100% correct? Nope. Can that yogurt be packed at 110%? Yes. Would that add another 10% calories on top of the label claim? Yes. Why could it be 110%?

    I'm glad you asked. There are laws against under-packing food there are no laws against giving you more than we tell you we will give you. So some products have to be over packed to a known percentage to make sure we don't under-pack them.

    Will that matter in a 10 lb tub of yogurt? Nope. Will it matter if you eat 10 single serve small containers of yogurt? Yup.

    That is just one instance where your food or ingredients can be out of alignment with the label.

    Like one of the women said before. Find what works for you and do that. It won't work for everyone but it will work for you.

    Find what DOESN'T work for you and don't do that.
  • jeanmariemcp
    jeanmariemcp Posts: 17 Member
    Options
    segacs wrote: »
    Don't let the perfect be the enemy of the good.

    LOVE THIS!!!

  • stealthq
    stealthq Posts: 4,298 Member
    Options
    lemurcat12 wrote: »
    Hornsby wrote: »
    stealthq wrote: »
    Hornsby wrote: »
    segacs wrote: »
    Hornsby wrote: »
    Food labels can be overestimated by 20% but can't be underestimated.

    That's not how statistical variance works. They can sometimes be underestimated, too.

    Awwe, okay. I just meant by law.

    It's only certain nutrients that are not allowed to be over-represented on the label (by law). Not everything. And definitely not calories.

    ETA: I think I remember the nutrients that could not be over-represented on the label (but could be under-represented) were supplements like added vitamin D in milk, but I can't find the right FDA CFR document to confirm. Oh well.

    "The Third Group nutrients include calories, sugars, total fat, saturated fat, cholesterol, and sodium. However, for products (e.g., fruit drinks, juices, and confectioneries) with a sugars content of 90 percent or more of total carbohydrate, to prevent labeling anomalies due in part to rounding, FDA treats total carbohydrate as a Third Group nutrient instead of a Class II nutrient. For foods with label declarations of Third Group nutrients, the ratio between the amount obtained by laboratory analysis and the amount declared on the product label in the Nutrition Facts panel must be 120% or less, i.e., the label is considered to be out of compliance if the nutrient content of a composite of the product is greater than 20% above the value declared on the label. For example, if a laboratory analysis found 8 g of total fat/serving in a product that stated that it contained 6 g of total fat/serving, the ratio between the laboratory value and the label value would be (8 / 6) x 100 = 133%, and the product label would be considered to be out of compliance."

    http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformation/LabelingNutrition/ucm063113.htm


    I think you guys are saying the same thing.

    Certain nutrients (where we are particularly concerned that they not mislead you into thinking there are more whatever than there is) are supposed to be within a reasonable amount of the number when tested* but, specifically, must be at least 80% of the number given to be in compliance with the law.

    Others (and this includes calories), where we are particularly concerned that they not mislead you into thinking there is less whatever than there is, are also supposed to be within a reasonable amount of the number when tested* but, specifically, must not be more than 120% of it to be in compliance with the law.

    *Reasonable excesses of class I and II nutrients above labeled amounts and reasonable deficiencies of the Third Group nutrients are usually considered acceptable by the agency within good manufacturing practices.

    Nope, not the same thing, but in the same document (BTW, this is an old version, there's a newer one). This is what I was trying to remember:

    In order to evaluate the accuracy of nutrition label information against a standard for compliance purposes, FDA regulations define two nutrient classes (Class I and Class II) (21 CFR 101.9(g)(3)) and list a third group (Third Group) of nutrients (21 CFR 101.9(g)(5)). Class I nutrients are those added in fortified or fabricated foods. These nutrients are vitamins, minerals, protein, dietary fiber, or potassium. Class I nutrients must be present at 100% or more of the value declared on the label ; in other words, the nutrient content identified by the laboratory analysis must be at least equal to the label value. For example, if vitamin C is added in a fortified product and the label states that vitamin C is present at 10% Daily Value (DV), the laboratory value must equal at least 6 mg of vitamin C/serving (i.e., 10% of the 60 mg Reference Daily Intake (RDI) for vitamin C that is specified in 21 CFR 101.9(c)(8)(iv)). The ratio between a laboratory finding of 4.8 mg vitamin C/serving (i.e., 8% DV) and the label value of 10% DV would be calculated as follows: ...
  • segacs
    segacs Posts: 4,599 Member
    Options
    segacs wrote: »
    Don't let the perfect be the enemy of the good.

    LOVE THIS!!!

    Thanks, but it's not my quote. It's usually attributed to Voltaire.
  • lemurcat12
    lemurcat12 Posts: 30,886 Member
    Options
    stealthq wrote: »
    lemurcat12 wrote: »
    Hornsby wrote: »
    stealthq wrote: »
    Hornsby wrote: »
    segacs wrote: »
    Hornsby wrote: »
    Food labels can be overestimated by 20% but can't be underestimated.

    That's not how statistical variance works. They can sometimes be underestimated, too.

    Awwe, okay. I just meant by law.

    It's only certain nutrients that are not allowed to be over-represented on the label (by law). Not everything. And definitely not calories.

    ETA: I think I remember the nutrients that could not be over-represented on the label (but could be under-represented) were supplements like added vitamin D in milk, but I can't find the right FDA CFR document to confirm. Oh well.

    "The Third Group nutrients include calories, sugars, total fat, saturated fat, cholesterol, and sodium. However, for products (e.g., fruit drinks, juices, and confectioneries) with a sugars content of 90 percent or more of total carbohydrate, to prevent labeling anomalies due in part to rounding, FDA treats total carbohydrate as a Third Group nutrient instead of a Class II nutrient. For foods with label declarations of Third Group nutrients, the ratio between the amount obtained by laboratory analysis and the amount declared on the product label in the Nutrition Facts panel must be 120% or less, i.e., the label is considered to be out of compliance if the nutrient content of a composite of the product is greater than 20% above the value declared on the label. For example, if a laboratory analysis found 8 g of total fat/serving in a product that stated that it contained 6 g of total fat/serving, the ratio between the laboratory value and the label value would be (8 / 6) x 100 = 133%, and the product label would be considered to be out of compliance."

    http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformation/LabelingNutrition/ucm063113.htm


    I think you guys are saying the same thing.

    Certain nutrients (where we are particularly concerned that they not mislead you into thinking there are more whatever than there is) are supposed to be within a reasonable amount of the number when tested* but, specifically, must be at least 80% of the number given to be in compliance with the law.

    Others (and this includes calories), where we are particularly concerned that they not mislead you into thinking there is less whatever than there is, are also supposed to be within a reasonable amount of the number when tested* but, specifically, must not be more than 120% of it to be in compliance with the law.

    *Reasonable excesses of class I and II nutrients above labeled amounts and reasonable deficiencies of the Third Group nutrients are usually considered acceptable by the agency within good manufacturing practices.

    Nope, not the same thing, but in the same document (BTW, this is an old version, there's a newer one). This is what I was trying to remember:

    In order to evaluate the accuracy of nutrition label information against a standard for compliance purposes, FDA regulations define two nutrient classes (Class I and Class II) (21 CFR 101.9(g)(3)) and list a third group (Third Group) of nutrients (21 CFR 101.9(g)(5)). Class I nutrients are those added in fortified or fabricated foods. These nutrients are vitamins, minerals, protein, dietary fiber, or potassium. Class I nutrients must be present at 100% or more of the value declared on the label ; in other words, the nutrient content identified by the laboratory analysis must be at least equal to the label value. For example, if vitamin C is added in a fortified product and the label states that vitamin C is present at 10% Daily Value (DV), the laboratory value must equal at least 6 mg of vitamin C/serving (i.e., 10% of the 60 mg Reference Daily Intake (RDI) for vitamin C that is specified in 21 CFR 101.9(c)(8)(iv)). The ratio between a laboratory finding of 4.8 mg vitamin C/serving (i.e., 8% DV) and the label value of 10% DV would be calculated as follows: ...

    My point was that what you were saying wasn't in disagreement with the point Hornsby was making. You appeared to be disagreeing but there was not actually a conflict.
  • stealthq
    stealthq Posts: 4,298 Member
    Options
    lemurcat12 wrote: »
    stealthq wrote: »
    lemurcat12 wrote: »
    Hornsby wrote: »
    stealthq wrote: »
    Hornsby wrote: »
    segacs wrote: »
    Hornsby wrote: »
    Food labels can be overestimated by 20% but can't be underestimated.

    That's not how statistical variance works. They can sometimes be underestimated, too.

    Awwe, okay. I just meant by law.

    It's only certain nutrients that are not allowed to be over-represented on the label (by law). Not everything. And definitely not calories.

    ETA: I think I remember the nutrients that could not be over-represented on the label (but could be under-represented) were supplements like added vitamin D in milk, but I can't find the right FDA CFR document to confirm. Oh well.

    "The Third Group nutrients include calories, sugars, total fat, saturated fat, cholesterol, and sodium. However, for products (e.g., fruit drinks, juices, and confectioneries) with a sugars content of 90 percent or more of total carbohydrate, to prevent labeling anomalies due in part to rounding, FDA treats total carbohydrate as a Third Group nutrient instead of a Class II nutrient. For foods with label declarations of Third Group nutrients, the ratio between the amount obtained by laboratory analysis and the amount declared on the product label in the Nutrition Facts panel must be 120% or less, i.e., the label is considered to be out of compliance if the nutrient content of a composite of the product is greater than 20% above the value declared on the label. For example, if a laboratory analysis found 8 g of total fat/serving in a product that stated that it contained 6 g of total fat/serving, the ratio between the laboratory value and the label value would be (8 / 6) x 100 = 133%, and the product label would be considered to be out of compliance."

    http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformation/LabelingNutrition/ucm063113.htm


    I think you guys are saying the same thing.

    Certain nutrients (where we are particularly concerned that they not mislead you into thinking there are more whatever than there is) are supposed to be within a reasonable amount of the number when tested* but, specifically, must be at least 80% of the number given to be in compliance with the law.

    Others (and this includes calories), where we are particularly concerned that they not mislead you into thinking there is less whatever than there is, are also supposed to be within a reasonable amount of the number when tested* but, specifically, must not be more than 120% of it to be in compliance with the law.

    *Reasonable excesses of class I and II nutrients above labeled amounts and reasonable deficiencies of the Third Group nutrients are usually considered acceptable by the agency within good manufacturing practices.

    Nope, not the same thing, but in the same document (BTW, this is an old version, there's a newer one). This is what I was trying to remember:

    In order to evaluate the accuracy of nutrition label information against a standard for compliance purposes, FDA regulations define two nutrient classes (Class I and Class II) (21 CFR 101.9(g)(3)) and list a third group (Third Group) of nutrients (21 CFR 101.9(g)(5)). Class I nutrients are those added in fortified or fabricated foods. These nutrients are vitamins, minerals, protein, dietary fiber, or potassium. Class I nutrients must be present at 100% or more of the value declared on the label ; in other words, the nutrient content identified by the laboratory analysis must be at least equal to the label value. For example, if vitamin C is added in a fortified product and the label states that vitamin C is present at 10% Daily Value (DV), the laboratory value must equal at least 6 mg of vitamin C/serving (i.e., 10% of the 60 mg Reference Daily Intake (RDI) for vitamin C that is specified in 21 CFR 101.9(c)(8)(iv)). The ratio between a laboratory finding of 4.8 mg vitamin C/serving (i.e., 8% DV) and the label value of 10% DV would be calculated as follows: ...

    My point was that what you were saying wasn't in disagreement with the point Hornsby was making. You appeared to be disagreeing but there was not actually a conflict.

    It is a disagreement, because he was originally implying all nutrients can't be underestimated by law when in actuality that only applies to some nutrients. The rest can be higher or lower than the stated amount, and that includes total calories.

    It isn't any big deal to me, except that people start defaulting to 'unpackaged food is more accurately estimated than packaged', which is not true since it's ALL variable and ALL the counts we rely on are based on a mean determined by someone. Which you already know, of course.
  • mbcieslak87
    mbcieslak87 Posts: 206 Member
    Options
    All I know is I lose weight referencing them. If you're eating at a 1lb a week deficit for example, MFP is taking 500 calories off your daily maintenance total to create a deficit of 3500 a week... or approximately a pound. So even if I am 200 calories off by label error every day that's still 2100 calorie deficit over the week - that's still 0.6 lbs and that's not going to happen day in and day out. Plus your calorie burn, at least for the vast majority of us, isn't that static... you burn more on some days than others just by living. Some days you walk 2,000 extra steps cause you went to the grocery store, some days you scrub floors, some days you're just plain lazy and watch 10 hours of netflix - if you are paying close attention to labels and following them accurately, you should be close enough to lose weight. 5 extra calories worth of ketchup won't ruin your life... don't let calories become an overwhelming obsession, but just focus on the trends you see and the better choices you are making.
  • Wiseandcurious
    Wiseandcurious Posts: 730 Member
    edited January 2015
    Options
    So in other words, the inaccuracy of the labels is going to drive people who weigh their food batty, but the rest of us don't have to worry about it because we're allowing for broad inaccuracy anyway.

    That's making quite big assumptions...

    I weigh my food *and* allow for broad inacuracy too. (Meaning, I accept that no estimate of calories in food is 100% correct and am fine with it, not that I am allowing a calorie "buffer or sth).

    The biggest advantage of weighing for me is increased awareness of how much and what I am eating and learning crucial portion control skills. For fun, I always try to estimate before I weigh and it's gratifying how better my estimates are becoming. Also, weighing is an excellent way to avoid increasing the inacuracy that's already there!

    I am not saying that weighing is the way to go for everyone, but it seems to me the quoted post paints it as some kind of OCD, freak-out-over-every-gram thing, which could drive away from weighing people for whom it would be extremely beneficial. I am sure that wasn't what the poster intended, but felt it was important to clarify.

    ETA the text in brackets.
  • lemurcat12
    lemurcat12 Posts: 30,886 Member
    edited January 2015
    Options
    stealthq wrote: »
    lemurcat12 wrote: »
    stealthq wrote: »
    lemurcat12 wrote: »
    Hornsby wrote: »
    stealthq wrote: »
    Hornsby wrote: »
    segacs wrote: »
    Hornsby wrote: »
    Food labels can be overestimated by 20% but can't be underestimated.

    That's not how statistical variance works. They can sometimes be underestimated, too.

    Awwe, okay. I just meant by law.

    It's only certain nutrients that are not allowed to be over-represented on the label (by law). Not everything. And definitely not calories.

    ETA: I think I remember the nutrients that could not be over-represented on the label (but could be under-represented) were supplements like added vitamin D in milk, but I can't find the right FDA CFR document to confirm. Oh well.

    "The Third Group nutrients include calories, sugars, total fat, saturated fat, cholesterol, and sodium. However, for products (e.g., fruit drinks, juices, and confectioneries) with a sugars content of 90 percent or more of total carbohydrate, to prevent labeling anomalies due in part to rounding, FDA treats total carbohydrate as a Third Group nutrient instead of a Class II nutrient. For foods with label declarations of Third Group nutrients, the ratio between the amount obtained by laboratory analysis and the amount declared on the product label in the Nutrition Facts panel must be 120% or less, i.e., the label is considered to be out of compliance if the nutrient content of a composite of the product is greater than 20% above the value declared on the label. For example, if a laboratory analysis found 8 g of total fat/serving in a product that stated that it contained 6 g of total fat/serving, the ratio between the laboratory value and the label value would be (8 / 6) x 100 = 133%, and the product label would be considered to be out of compliance."

    http://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformation/LabelingNutrition/ucm063113.htm


    I think you guys are saying the same thing.

    Certain nutrients (where we are particularly concerned that they not mislead you into thinking there are more whatever than there is) are supposed to be within a reasonable amount of the number when tested* but, specifically, must be at least 80% of the number given to be in compliance with the law.

    Others (and this includes calories), where we are particularly concerned that they not mislead you into thinking there is less whatever than there is, are also supposed to be within a reasonable amount of the number when tested* but, specifically, must not be more than 120% of it to be in compliance with the law.

    *Reasonable excesses of class I and II nutrients above labeled amounts and reasonable deficiencies of the Third Group nutrients are usually considered acceptable by the agency within good manufacturing practices.

    Nope, not the same thing, but in the same document (BTW, this is an old version, there's a newer one). This is what I was trying to remember:

    In order to evaluate the accuracy of nutrition label information against a standard for compliance purposes, FDA regulations define two nutrient classes (Class I and Class II) (21 CFR 101.9(g)(3)) and list a third group (Third Group) of nutrients (21 CFR 101.9(g)(5)). Class I nutrients are those added in fortified or fabricated foods. These nutrients are vitamins, minerals, protein, dietary fiber, or potassium. Class I nutrients must be present at 100% or more of the value declared on the label ; in other words, the nutrient content identified by the laboratory analysis must be at least equal to the label value. For example, if vitamin C is added in a fortified product and the label states that vitamin C is present at 10% Daily Value (DV), the laboratory value must equal at least 6 mg of vitamin C/serving (i.e., 10% of the 60 mg Reference Daily Intake (RDI) for vitamin C that is specified in 21 CFR 101.9(c)(8)(iv)). The ratio between a laboratory finding of 4.8 mg vitamin C/serving (i.e., 8% DV) and the label value of 10% DV would be calculated as follows: ...

    My point was that what you were saying wasn't in disagreement with the point Hornsby was making. You appeared to be disagreeing but there was not actually a conflict.

    It is a disagreement, because he was originally implying all nutrients can't be underestimated by law when in actuality that only applies to some nutrients. The rest can be higher or lower than the stated amount, and that includes total calories.

    He said it unclearly or confusingly (or backwards), as he clarified, but based on what he quoted I think you were trying to make the same point. But maybe you and I are disagreeing: calories are class III, so as I read it (not checking the actual current CFR, since I don't care that much) they by law must be 120% or less--in other words, are not supposed to be unreasonably different either way, but specifically are out of compliance if they are measured at 125 calories with a label of 100 calories. With something like Vitamin A, it's the opposite (as you said), they can't be 79% if they say 100% (or however they are measured).

    In other words, calories, of course, can be over- or under-estimated, but the law is stricter with respect to underestimation. With Vitamin A it's the opposite.
    It isn't any big deal to me, except that people start defaulting to 'unpackaged food is more accurately estimated than packaged', which is not true since it's ALL variable and ALL the counts we rely on are based on a mean determined by someone. Which you already know, of course.

    Yes, we agree on this. You just have to accept that everything is an estimate.