FDA Warning Letters this week: A dirty glove is a dirty hand

The week goes by and new letters come out. Below I’ve selected a few of the food-oriented warning letters and discuss/explain some of the more interesting findings.

Don’t know what warning letters are? Check out this post for a brief overview of what they are and why FDA sends them out.

A dirty glove is a dirty hand

This week I’m going to do something a little different, rather than review the entire contents of one or several letters, I wanted to address some common findings that showed up in some truly awful facilities that received warning letters recently. Among a huge number of general findings of insanitary conditions like dirty equipment, rotting food, and poor general upkeep, the FDA inspectors made some pointed comments about hand contact when gloves were used.

WL# 18-ATL-01

Subject: CGMP/Manufacturing, Packing or Holding Human Food/Adulterated/Insanitary Conditions
   You failed to ensure all personnel working in direct contact with food, food contact surfaces, and food packaging materials conform to hygienic practices while on duty to the extent necessary to protect against contamination of food as required by 21 CFR Part 110.10(b). For example:
 a.  Employees failed to maintain gloves used in food handling in an intact, clean, and sanitary condition as required by 21 CFR Part 110.10(b)(5). For example, an employee in the front kitchen was observed using gloved hands to assemble sanitized production equipment. This employee stopped assembly to manipulate the hose that had been on the floor and then resumed assembling the sanitized equipment without first washing and sanitizing or changing the gloves.
b.  Employees did not wash hands thoroughly in an adequate hand-washing facility at any time their hands may have become soiled or contaminated as required by 21 CFR Part 110.10(b)(3). For example:
i.    Multiple employees were observed returning from breaks without washing their hands or washed their hands and then put on hairnets and earphones, without first re-washing their hands, before resuming production activities in the front and back kitchens and the spice room.
ii.    Throughout the inspection the front and back kitchen mixer employees were observed touching non-food contact surfaces such as shovels, buttons, equipment, buckets, cardboard boxes, waste bins, drums, kettles, and tables with gloved hands, then using the same gloved hands, without first sanitizing or changing, to handle raw ingredients and subsequently handle finished product and food contact surfaces.
iii.    A production employee was observed after a break in production to come directly from the break area into the spice room, handle non-food contact surfaces including a door handle, tables, a screw driver and then handle food production equipment such as a mixer, paddle, and a hose without first washing or sanitizing their hands.
iv.    Employees were observed entering the front kitchen production room from the spice room, handling non-food contact equipment such as a radio, boxes, and a trash can and then not washing or sanitizing their hands before performing work that included their bare hands touching the interior of “(b)(4)” finished product containers.

WL# CIN-17-510879-04

Subject: CGMP Food/Prepared, Packed or Held Under Insanitary conditions/adulterated
Employees were not observed washing their hands thoroughly in an adequate hand-washing facility before starting work, after each absence from the work station, and at any other time when the hands may have become soiled or contaminated, as required by 21 CFR 110.10(b)(3). Specifically…Employees engaged in food preparation were observed leaving the food preparation area, retrieving additional ingredients or packaging supplies, and taking phone calls. The employees were observed changing their gloves between these events, but they were not observed washing their hands prior to changing their gloves.
So what risk do the findings pose?

So….here’s the deal. Hands are obvious and effective vectors for disease. In general we tend to focus on the fecal-oral route (fingernails and cheap toilet paper kill people, wash up friends), but in a food plant, it’s important to watch where employees frequently interact with areas where they might touch “unclean” spaces and bring things back to the product, like bacteria on a dirty pallet or allergens from a snack/lunch while on break.

What’s missing here to evaluate the risk is some context. Generally employees can move from sanitary “food contact” surfaces without having to “intervene” on their hands, and some of the surfaces identified such as equipment buttons or screwdrivers could very well have been sanitized at the beginning of the shift so that clean-handed employees could use them without issue. That would be an efficient approach to startup sanitation. However, the other findings like coming back from break or handling items on the floor or obviously dirty equipment is normally out of the realm of best practice, and based on the awful sanitary state of the plants mentioned in the warning letters, it’s very likely that those surfaces/practices resulted in frequent cross-contamination. Good auditor.

What about the gloves?

This is more interesting, FDA specifically went after both manufacturers for using dirty gloved hands, because gloves aren’t magical. However, in the second warning letter above, FDA was pretty strict in that they didn’t consider the glove change adequate as an intervention. The general thought process here is that if you take off a pair of dirty gloves, your hands get dirty in the process. Therefore putting on a new pair of gloves with “dirty” hands means that your new gloves get contaminated regardless.

The “gloves or no gloves” debate typically focuses on the fact that it has been shown that glove use actually reduces hand washing compliance in healthcare and food settings, because people think the gloves make their hands cleaner. While dirty hands get washed (eew, there’s dirt on my hand), dirty gloves get worn.

The other half of the risk assessment though is based on bacterial transfer rates. And on this half, it has been shown that bacteria transfers from food to bare hands at about a 10% rate while through a glove it only transfers at a 0.01% rate. So even if you contaminated your hands while taking off an old pair of dirty gloves, only a tiny amount of those bacteria would actually make it back through a new pair of gloves. Though that doesn’t matter if you then contaminate the contact side of a new pair of gloves with your dirty hands, and it has been shown that handwashing between glove changes reduces that incidence! BUT it turns out that surgeons do this to protect themselves from getting contaminated during surgery without handwashing when changing gloves aseptically, so it’s common practice in other aseptic fields!

*pant, pant*

So…it’s complicated. But there is a potential third option, even if it isn’t generally considered in the guidance (by the way, mandatory washing between glove changes isn’t actually codified anywhere, but is enforced guidance and part of GFSI schemes).

So we’ve concluded that putting on gloves with dirty hands results in dirty gloves. Fine. But once you’ve put on a smooth, impervious, surface that can be cleaned and sanitized, you no longer have a glove, but a “food contact surface” just like a spoon! And spoons (if visibly clean) can be sanitized before returning to use.

We know that, in general, current handwashing recommendations can result in a 2-3.5 log reduction of bacteria on hands. This is the best we can realistically hope to do with current practice, and we rely on handwashing to prevent either low levels of contamination, or to simply remove visible filth that carries higher numbers of bacteria. Further transfer to food depends on transfer rates that are highly variable, and actual food safety risk depends on what’s transferred and whether it has the ability to keep proliferating.

We also know that gloves can be removed with minimal transfer per medical literature, and that only 0.01% of bacteria will make it from a “dirty” hand through the glove. So we just need to take care of those pesky bacteria on the outside of the glove. Too bad there isn’t some kind of way we could quickly treat a visibly clean food contact surface…say a 5-log reduction in like, 30 seconds to a minute…

Enter glove/hand dips. A method where in critical areas you maintain a “bath” of sanitizer prepared to manufacturers specifications that will provide a 5 log reduction of pathogens on a clean surface. By changing your gloves and sanitizing them by dipping them in your solution, you can empiracally achieve a much higher reduction in organisms on the contact surface than you would achieve by de-gloving>handwashing>re-gloving. Even better, you can do it just as quickly has a hand wash and can put these hand dips in convenient locations. Basically, you treat your hands as if they were a utensil like a spoon instead of hands.

FDA references these “hand dip” or “glove dip” stations in the Food code (2-301.16), refers to them in guidance related to nut processing, warning letters, and specifically states in a risk assessment on sprouts processors (emphasis mine):

Hand dip stations are important to sanitize clean hands and gloves.

Important language distinction here is that they said “clean gloves”, they still need to meet a visibly clean standard for the sanitizer to work, but if there is no visible contamination, you’re ready to sanitize, and a fresh pair of gloves out of the box will normally meet this standard.

Hand dip stations are also referenced in some older FSIS publications.

So, from a scientifically validated approach, glove dip stations should be as or more efficient as handwashing between glove changes. Unfortunately because of the abundance of state and FDA guidance mandating a “wash” between glove changes, different auditors may be more or less open to this procedure even if it has been properly validated/verified. However, the entire point of HACCP or HARPC is that you can establish whatever procedures work best for your plant/process that can be substantiated scientifically. Try it out, collect enough data to demonstrate that it works, and slip this blog post in your citations if it helps ;).

Literature:

Fuller, Christopher, et al. ““The dirty hand in the latex glove”: a study of hand hygiene compliance when gloves are worn.” Infection Control & Hospital Epidemiology 32.12 (2011): 1194-1199.

Montville, Rebecca, Yuhuan Chen, and Donald W. Schaffner. “Glove barriers to bacterial cross-contamination between hands to food.” Journal of Food Protection 64.6 (2001): 845-849.

Robinson, Andrew L., et al. “Adequate Hand Washing and Glove Use Are Necessary To Reduce Cross-Contamination from Hands with High Bacterial Loads.” Journal of food protection 79.2 (2016): 304-308.

Jensen, Dane A., et al. “Quantifying the Effects of Water Temperature, Soap Volume, Lather Time, and Antimicrobial Soap as Variables in the Removal of Escherichia coli ATCC 11229 from Hands.” Journal of Food Protection 80.6 (2017): 1022-1031.

 

Presence of fecal coliforms does not mean “there’s poop here”, but for aspiring poop-hunters is there an alternative?

Image created by Austin Bouck. Dear god let there be no egregious mistakes…

Recently a podcast I often listen to has discussed several articles with a similar note in their conclusions.

Nearly half of the 90 beverages from soda fountain machines in one area in Virginia tested positive for coliform bacteria — which could indicate possible fecal contamination

The study, conducted over six months at six licensed day-care centers in California, found high levels of coliform contamination, particularly in kitchen areas and on the hands of day-care workers. Coliform bacteria are transmitted through feces.

But recent reports reveal that the coffee or tea you’re sipping  – and even the cash you used to pay for it –  also carry bacteria found in feces.

Seven out of 10 samples of Costa ice were found to be contaminated with bacteria found in faeces.

40 percent of office coffee mugs contain coliform bacteria, which can be found in feces.

I don’t need to rehash arguments that have already been made here. Fecal coliform is a group of organisms historically used to identify whether drinking water could be contaminated with sewage or fecal material. Because it’s still a valid test to determine if water has been contaminated (they don’t survive forever in clean water, so if they’re present it’s because there’s a likely a failure in the system), the CDC and tons of state regulations still falsely state that these organisms are only found in fecal material.

Part of the issue is that coliforms as an indicator in general were extrapolated to environmental samples and food. NACMCF summarized the most recent scientific conclusion:

…whether coliforms, fecal coliforms, Enterobacteriaceae or E. coli. Kornacki and others (Kornacki et al., 2013) provide an historical evaluation of these criteria for foods and their utility based on current knowledge. None of these criteria accurately and consistently reflect fecal contamination of raw and processed foods nor are they useful or reliable as index organisms predicting the presence of pathogens.

The Kornacki reference is from the Compendium of methods for the Microbiological Examination of foods, which while expensive is an excellent reference for any food micro lab:

The debunking of the coliform=feces absolute conclusion was well covered by Doyle and Erickson in 2006, and more recently some thoughts were posted by barfblog.

But I’m writing this post today because I have a follow up question,

How exactly would someone actually test for fecal material?

Apparently, this is actually a problem in forensics, as an article published as recently as 2013 stated”no sensitive and simple fecal identification method using molecular biological techniques has been reported.” (Nakanishi et al., 2013). Here were the methods I was able to find with a cursory review of the literature.

1. Direct (macro) observation

A.k.a, the stain on the tighty-whities. Feces can be identified from it’s characteristic green-brown color from bile digestion and characteristic odor…this detection method isn’t really worth going into and I apologize for the mental image/odor.

2. Microscopy

Now we’re getting sciencey. Fecal material is a hodgepodge of various microscopic indicators. An experienced eye would be able to find bacteria (the bulk of the dry weight), but also undigested food particles and cellulose cell walls from plant material. Some epithelial and mucousal cells from the GI tract would also be visible. Interestingly, one way to determine if the fecal material was from a person or animal would be to look for excess hair in the sample, either from digesting prey or from grooming. The hair has to go somewhere, and I’m really curious if you would see a difference in hair mass between bearded and clean shaven men…

3. Chemical Indicators

Urobilinogen is a byproduct of bilirubin metabolism, and will be found in animals which consume meat and/or are otherwise digesting blood.  This appears to be a fairly classic test to discriminate fecal material from other bodily fluids like sweat or saliva and can be performed on site using a fluorescence indicator solution. Unfortunately, urobilinogen can also be found in urine since it is also returned to the kidneys for excretion.

4. Microbiological Profile

Well, crap. Here we come full circle. The most recent research on this subject seems to be using newer sequencing techniques to identify the unique organisms in the microbiome of feces. The goal is to find specific organisms or genes that would allow forensic scientists to discriminate between fecal material and other bodily fluids.

Here’s the thing though, none of the research identifies coliforms as a group of interest. It’s too broad and unhelpful! Rather than traditional “fecal bacteria”, the Bacteriodes genus has been identified as the predominant organism group in feces. Specific organisms identified were B. uniformis, B. vulgatus, and B. thetaiotaomicron. However, the state of California has specifically identified these organisms thriving in marshlands, yet still attribute them directly to feces. It would seem that again we can infer that fecal material is likely to contain these organisms, but it seems improper to assume that the presence of the organisms means that fecal material was the source.

I’d love to hear from some forensic scientists on what I got right and wrong here. From what I can tell from the literature it seems like there isn’t as much forensic interest in fecal material. The authors referenced how it is hard to isolate DNA due to interference from bile enzymes and microorganisms, which would reduce its value as evidence.

It seems like at this time a definitive test for “there is poop on this plate/ice/food/hand” doesn’t exist.

Much to the chagrin of PCRM, who would like us to declare the invisible feces on our meat.

Maybe we can just stop making the correlation between feces and food hygiene and instead focus on pathogen detection/prevention/pervasiveness as a means to evaluate foods on the market. While we talk about the fecal-oral route a lot, we’ve known for a long time that pathogens can be found almost anywhere if you start looking, so let’s look for them instead of fecal coliform clickbait (which this post totally is).

 

 

Resources for fecal identification forensics:

Drexler, Judith Z., et al. “Marsh Soils as Potential Sinks for Bacteroides Fecal Indicator Bacteria, Waccamaw National Wildlife Refuge, Georgetown, SC, USA.” Water, Air, & Soil Pollution 225.2 (2014): 1861.

Forensic Resources.Serology – Blood and other Bodily Fluids. http://www.ncids.com/forensic/serology/serology.shtml

James, Stuart H., Jon J. Nordby, and Suzanne Bell, eds. Forensic science: an introduction to scientific and investigative techniques. CRC press, 2002.

Li, Richard. Forensic biology. CRC Press, 2015.

Nakanishi, Hiroaki, et al. “Identification of feces by detection of Bacteroides genes.” Forensic Science International: Genetics7.1 (2013): 176-179.

Virkler, Kelly, and Igor K. Lednev. “Analysis of body fluids for forensic purposes: from laboratory testing to non-destructive rapid confirmatory identification at a crime scene.” Forensic Science International 188.1 (2009): 1-17.

Zou, Kai-Nan, et al. “Identification of vaginal fluid, saliva, and feces using microbial signatures in a Han Chinese population.” Journal of forensic and legal medicine 43 (2016): 126-131.

FDA Warning Letters this week 11/2/17: A gift from FDA to QA staffers reviewing product labels

The week goes by and new letters come out. Below I’ve selected a few of the food-oriented warning letters and discuss/explain some of the more interesting findings.

Don’t know what warning letters are? Check out this post for a brief overview of what they are and why FDA sends them out.

WL# 17-ATL-12

Subject: Food Labeling/Misbranded/False and Misleading

The labeling requirements for food are so complicated, that the guidance documents for them are in the hundreds of pages. It’s the responsibility of those folks involved in compliance to review labels and claims to ensure that their company doesn’t end up in trouble, but that can mean being the bearer of bad news. While QA/regulatory folks get buried in the details, other stakeholders only understand labels from what they see in the market from products they buy or competitor products.

Oftentimes, QA folks have the fun job of saying “that’s not technically what the law says”. This is the risk assessment version of label review. The next step then, is risk management, in which other folks may ask the question of whether FDA/FTC are actually enforcing these rules, some of which are nestled in very grey areas and/or not commonly followed!

Often when evaluating the enforcement side of the law, good QA personnel turn to warning letters to see which portions FDA is actually enforcing. Unfortunately when labeling is enforced on the food side via warning letter, it’s never for the soft grey area stuff. FDA mostly enforces misrepresentation of product, health/function/unapproved drug claims, having components straight up missing (e.g. “you didn’t have an ingredient statement“), or allergen mislabeling (which is a food safety issue). We get very few examples where FDA has actually verified nutrition values (for which there wasn’t a claim), corrected serving size errors, or identified other interpretation issues.

So, in a blog that offers plenty of criticism, let’s take a moment to say:

Thank you FDA!

Visible enforcement actions on labeling deviations help everyone keep their marketing teams in check. Regulatory professionals have so much more ability to push for accurate and fair labeling  when they can actually demonstrate that it’s being enforced elsewhere.

Okay, so exposition over, let’s get into the actual findings of this warning letter:

In June, 2017, FDA actually collected a sample of this product (crackers) from a retail store and tested it for nutrition against the values in the nutrition facts panel.

This is super cool. And while FDA is clear in the regulations that it has this power (and explains to manufacturers how these samples are collected) you never really hear about it happening.

 The product label states one serving is one package/ 6 crackers and contains 10% of the Daily Value for iron. However, the sample analysis performed by FDA found the product to contain 51.7% (original) and 57.2% (check) of the declared amount….Based on these findings, your product is misbranded under Section 403(a)(1) of the Act [21 U.S.C. § 343(a)(1)] in that the label is false or misleading. Specifically, the iron content is less than 80% of the amount declared on the label [21 CFR 101.9(g)(4)(ii)].

Okay, so real quick and dirty discussion about nutrition labeling and how far your numbers are allowed to be off.

How far off from label is too far?

Basically, while still operating within “good manufacturing practice”, FDA created three groups of ingredients.

Class I ingredients  are vitamins and minerals that are added to “fortified” foods. These are ingredients that consumers see as beneficial, that you’ve now made marketing claims about, and that you would have a economic incentive to over declare. So  when your food is tested, these nutrients need to be present at 100% of the labeled level or higher. If you say it has 40mg of vitamin C, you better have at least 40mg. It can contain 50 or 60 with no problem, but it shouldn’t ever be below what you’re claiming, which would be interpreted as misleading your customer to sell more products without adding the ingredient.

Class II ingredients are the “good” stuff in normal foods that aren’t claiming to be “fortified”. Things like vitamins, minerals, protein, total carbohydrates, dietary fiber, unsaturated fat, and potassium that people like to see in their food (e.g. “I’m looking for a high protein snack“). These again are “beneficial” ingredients to consumers, so you would have an economic incentive to say they were higher than they actually are (or over declare them). When your food is tested, these nutrients need to be present at 80% of the labeled level or higher. If you say it has 8g of protein per serving, FDA gives you a little wiggle room since values will vary between crop years etc. (potatoes will change a little year to year based on rain etc.), but would rather you under declare to make sure no one is misled by the label.

Class III ingredients are the “bad” stuff that people are trying to avoid. Things like calories, total fat, saturated fat, cholesterol, and sodium (and sometimes carbohydrates/sugars). These you would have an incentive to under declare or say there was less of the bad stuff than there really was (40 fewer calories per serving than our competitors!). When your food is tested, these nutrients need to be present at 120% of the labeled level or lower.

Back to the letter…

So this company got in trouble for violating the rule for class II ingredients, in that the actual tested value for iron was more than 20% less than the label said. Assuming that this was an honest mistake, this company would have done well to truncate their labeled values for iron downward, as FDA says in their guidance:

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.

So don’t just label your protein as zero to make sure you don’t get in trouble, but if you tested at 2 grams and decided to label just 1 gram to make sure you were good to go, FDA is okay with that.

The information panel labeling does not meet the requirements in 21 CFR 101.2(e) because the bar code is intervening material within the ingredient list.

The “information panel” is a term that refers to the section of the label you decided to place the nutrition facts, ingredient statement, and place of manufacture. Technically it appears directly to the right of the “principal display panel” which is what you see facing you on the shelf, but that rule is very rarely enforced provided you don’t deliberately hide it.

The nutrition facts, ingredient statement, and place of manufacture (signature line) have a requirement that they all need to be touching each other with “no intervening material”. This is a common one for companies trying to make their labels look pretty to mess up. Things like logos, preparation instructions, or romantic copy (things like company story or description of the product) can’t be stuck in there between your ingredients, nutrition, and contact info, they need to be in one single unit so that people don’t need to go looking for them all over your package.

Intervening material...is a crisis
See how you can’t separate out the regulatory portion with “stuff”?

 

The product label does not declare the street address of the firm. Under 21 CFR 101.5(d), the street address is required unless it is shown in a current city directory or telephone directory.

We don’t have the label to review here, so we can’t say what information was on the label. But I have a haunch as to what happened with this product. It’s for sale on Amazon in large multi-packs, but fda stated that their package was 1.75 oz. I’m guessing complete contact info was available on the larger, bulk package, but the retailer was selling individual 1.75 oz packages. They may not have included all info since they didn’t wholesale the individual packets, just the cases. But since the individual packets were consumer facing packaging, they are required to contain all the pertinent information just like every candy bar and stick of gum.

On a separate note, in the internet age, the 1994 standard of “shown in a current city directory or telephone directory” is absolutely an obsolete one. I can readily find this business and contact information online, and it’s silly to assume that the same access isn’t available through the internet as a phone book. The minimum standard of name/city/state/zip is still helpful to search for businesses with generic names that may be harder to find.

As you may already know, FDA finalized new requirements for nutrition labeling on May 27, 2016. The new Nutrition Facts label includes updates to the required nutrient declarations and formatting requirements. Although FDA has proposed  to extend the initial compliance dates of July 26, 2018 or July 26, 2019 (depending on the size of the manufacturer), manufacturers may begin revising labels now to meet the new requirements.

Bold, FDA….bold.

This isn’t a finding, just a comment that FDA would like to see the updates to this label include switching to the new nutrition facts panel format.  However, I take issue with them trying to rush this particular manufacturer given that FDA is supporting extension of the deadlines despite many manufacturers already having completed in the 16 months since the rule became final, and the original deadline still 9 months out for large businesses.

Summary

This is a great warning letter to keep around. It’s a clear example of FDA both performing market surveillance on single serve products for compliance, as well as actually testing products to compare to the nutrition facts panel in the absence of a nutrition claim, which is very uncommon.

For businesses evaluating their labels for risk of enforcement, I offer the following business risk approach if you find yourself making decisions in the “grey area”.

  1. Food safety issues are non negotiable. Label content must NEVER put a consumer at risk.
  2. Has this issue been enforced based on the evidence available in warning letters or class action lawsuits?
  3. Is your product at “high risk” for label scrutiny by either FDA or consumer advocacy groups? Labels that are easy targets are those that: are dietary supplements, are fortified, contain nutrition claims on the label, contain health claims on the label, is intended for children, or are connected to a brand that makes nutrition/health connections in their marketing materials or website.
  4. Does the discovery of your use of the “grey area” show that you had an economic motivation to stretch the rule? I.e. did you deliberately mislead potential customers in order to gain a business advantage over your competitors or current industry trends?
  5. Is your decision defensible within the “must/shall” of the CFR?
  6. Always ask your marketing team, why do you want this? If the answer is that not doing it would affect purchase intent, that’s an obvious sign that consumers do value the issue, and you’re intentionally misleading customers.

Use these questions to decide how hard to push for the “spririt” of the labeling regulations, which is always to protect consumers from misleading business practices, as well as encourage fair competition among food producers.

Sell your product, not the label.

 

FDA’s Bad Bug Book refers consumers to Wikipedia for information on Listeria

I was perusing the Bad Bug Book while doing some research on the recent Blue Bell outbreak and came across a hyperlink. After hearing “do you want to know more?” in my head, I clicked through on some non-L. mono species of Listeria and was…confused. I quickly doubled back, thinking that maybe I had been redirected, but there it was.

FDA Bad Bug Book linking directly to wikipedia
FDA Bad Bug Book linking directly to Wikipedia

FDA describes the reference as “current information about the major known agents that cause foodborne illness.” Descriptions also include a statement that it should not be used as a comprehensive or clinical reference. However, this isn’t an excuse for making a consumer and industry reference link to a completely uncontrolled document source. The Bad Bug Book (2nd ed.) is a wonderfully written resource, both for a lay and industry audience; but the fact that the authors of the Listeria page referred to Wikipedia as an ongoing resource, without knowing or being able to control the content presented to consumers, is irresponsible. A nefarious Wikipedia troll could at any moment have an article claiming that L. grayi is a GMO herbicide borne bacteria found in bananas that causes uncontrolled crying and hair growth, and have the full support of the FDA behind their article.

Please don’t write that article.

A  currently live example of why this was such a poor decision is that if you click through to some of the pages, they don’t exist (as of 7/27/15). I don’t know if the author intended to write them him/herself and never got around to it, or if they simply assumed the pages existed, and then didn’t bother to review the content. I’m not satisfied with either of those answers, and if alternatively the reference articles were removed at some point, that also highlights what a poor decision those links were.

Given the sheer number of PhD’s involved in the book’s creation, I think taxpayers should expect a resource with material actually reviewed and sanctioned by FDA. The poor editing here is unacceptable and a change should be made to the current edition of the book.

Many of the other pages in the book name multiple related species, but either included links to NIH or CDC or included no link at all, both of which are acceptable alternatives. I won’t name the authors and editors of the book here, anyone who wants to know can find them at the front of the document. If you’re interested in bringing this to FDA’s attention in your own way, they’re on twitter as @US_FDA and additional points of contact are available at www.fda.gov.
ResearchBlogging.org

Food and Drug Administration (2012). Listeria Monocytogenes Bad Bug Book, Foodborne pathogenic microorganisms and natural toxins. Second Edition, 99-100

Anatomy of a “Serving”

NFPIs it how much you’re expected to eat? How much you should eat? Pardon me, but if I want to eat an entire bag of potato chips, that’s exactly what I’m going to do.

I say this as the guy who helps create nutrition facts panels for food products for a living. I have read, reread, proof read, and colored red hundreds of these little tables in my time, and believe it or not, people aren’t lining up at parties to hear my thrilling stories.

I know!

Right now you’re thinking, “But Austin, I remember back in 2008 the FDA called 2,584 adults from the US to ask them questions about their diet. And 24% of the respondents said they had no idea if serving sizes were determined by government rules or by manufacturers.”

Your oddly specific observation would be correct, and I should have at least ¼ of the room hovering around me in rapt attention, waiting for me to clarify this confusing point. Time has shown however that everyone is clearly too intimidated to approach and ask the simple questions, even when I’m subtly firing off labeling trivia from the empty cracker box carelessly left by the cheese platter, or establishing my mastery of the dance floor.

Fact: Every party has a dance floor, here’s the label. dance floor Well let’s clear this up right now while I’ve got you at home/work/somewhere, on your computer/phone/tablet (scary that I know where you are, isn’t it?). Who determines serving sizes, manufacturers or the government?

Answer: The government!…ish.

Well that was unsatisfying, but it’s the most accurate answer I can give. Essentially what happens is that our government, via the FDA and FSIS (Food Safety Inspection Service), determine how serving sizes are to be calculated and presented, but also leave manufacturers options in specific situations.

So how are serving sizes determined?

Step 1, what are you eating?

he first step companies need to take when determining serving size is to determine what type of product they are selling. Back in 1993 when they had to decide all of this stuff, FDA determined that they could use data collected in the NHANES dietary surveys conducted in 1978 and 1988 to set these standards. These were nationwide surveys that collected all sorts of data, including nutritional intake and food frequency data. With this information, FDA created “Reference Amounts Customarily Consumed,” or RACC, for different categories of food.

The first step is easy, find the category that a food falls into, and look up the RACC used to determine the serving size in 21 CFR 101.12 (for non-meat items). For example, if I was making mashed potatoes, my category would be “Potatoes and Sweet Potatoes/Yams: Mashed, candied, stuffed, or with sauce” and my RACC would be 140 grams.

Step 2, how can someone measure it out?

Here’s where the variation begins! You might just want to use that 140 grams we saw above; however, not everyone has a scale in their kitchen, and let’s see you try to guess how much mashed potato makes up 140 grams. Can you think of the last time you weighed your food, much less with metric weights (provided you live in the U.S.)?

So at this juncture, the government instructs manufacturers to determine what the closest “common household measurement” to one RACC of your product is. So if we take 140g of our potatoes and see that that’s about 2/3 of a cup, our serving size becomes 2/3 of a cup!

>THIS IS IMPORTANT<

Imagine we made a new, super fluffy mashed potato with more butter, and 140 grams of these potatoes actually wind up closer to ¾ of a cup. This means that even though both potato products were based on a RACC of 140g, they might have two completely different serving sizes, and the manufacturer arrived at each using the same government reference amount!

So there you have it, two serving sizes created based on a government standard, but completely different once observed on the store shelf. How could there be even more variation?

Single Serving

Ah, right. For many products, if the entire container contains less or near 200% of the RACC amount, then there are different rules to play by. In most cases, the product will be considered a single serving, but in others, manufacturers have the choice to label them as one or two servings. This is why you see different types of labeling in small containers such as ice cream, muffins, soda, and other “single serving” containers that appear significantly larger than the usual RACC amount.

As eaten, not as sold

Ah, and this is critical. When you ask someone how much cake they eat, they typically don’t respond with, “about 1/3 of a box of cake mix”. RACC values are based on products as they are consumed. However, serving sizes are based on products as they are sold. The reasoning? Because it would be bizarre to buy a bag of flour and see “two slices of bread” for a serving size. This makes more sense for some products than others, but ultimately serving sizes for products that require further preparation are the amount of packaged product it takes to make about 1 RACC of product as eaten. And remember, this must be rounded off at a common household measure!

Final thoughts

As we realize that our beloved nutrition facts panel is now old enough to drink (enforcement began in ’94), we look back and start to wonder if that data from the 70’s and 80’s used to determine RACC values still holds water. I can’t think of anyone who eats ½ cup of ice cream in a sitting, nor leaves the potato chip bag untouched after their first 10-20 chips.

But how about we think about RACC values in a different way. These values were never intended to be an expectation, but simply a way to bring nutrition information into context using consumer data. The thousands of calories in a 20 lb. bag of rice don’t have a lot of context when I eat it one bowl at a time, but that’s also not to say that I’ll never eat an obscene amount of rice in one sitting just because I’m starving.

Instead think of it this way, if these values are simply references to what we customarily consume at a time, we’ve got a great tool on our hands. I wouldn’t expect you to eat only ½ cup of ice cream, but have you noticed that many ice cream scoops happen to portion about ½ cup of ice cream at a time? And while I’ve been known to turn a bag of kettle fried chips into a meal, I still eat them one handful at a time, which just so happens to contain approximately 10-15 chips.

If only some sort of reference was available so that I could tell about how many calories I ate with each handful…

To learn more about how serving sizes are determined for all food products, check out the labeling and nutrition documents on the FDA website, this PowerPoint provided by the FSIS, or the Guide to Federal Food Labeling Requirements for Meat, Poultry, and Egg Products. Check out what consumer opinions of labels are looking like since 2008 in the FDA’s consumer research.

Choinière C. & Lando A. (2008). 2008 Health and Diet Survey, DOI:

ResearchBlogging.org

Conrad J. Choinière, & Amy Lando (2008). 2008 Health and Diet Survey FDA Consumer Behavior Research Foos Safety Surveys (FSS)