(I have spent much of the last week reading a 189-page report issued jointly by the U.S. Food and Drug Administration and Health Canada, with the dry title, “Joint FDA/Health Canada Quantitative Assessment of the Risk of Listeriosis from Soft-Ripened Cheese Consumption in the United States and Canada”. The reading is as dry as the title suggests. It’s full of technical statistical and research terminology, terms like “risk characterization,” “sensitivity analysis,” “mitigation,” “rank correlation,” and on and on. What that meant was that I had to read everything two, three, and four times before I could begin to make sense of it. But the more I read, the more upset I became, because I realized this is a very important document, one that could have a huge effect on food availability. If the authors of this report are successful in accomplishing what they want to accomplish with brie and camembert cheeses, you can be sure they will continue on to other kinds of cheese, and then other entire categories of food products, in their endless search for supposedly serious pathogen dangers. Equally troubling, the FDA considers this report ”science based and transparent,” when it is anything but. I wrote the following analysis to try to get my thoughts down in an orderly way. I encourage you, after you’ve read my assessment, to try your own hand at reading the report, or at least the summary, and then to take the opportunity the FDA is offering to provide comments, and let the agency know in no uncertain terms what you think about this particular piece of literature. )
Nearly 15 years ago, a business book came out with the strange title, Who Moved My Cheese?
It was the story of mice in search of cheese that had disappeared, a parable about how people need to prepare for change, in their business and personal lives, and it became a huge best seller.
The parable may be playing out literally in real life before too long for raw milk soft cheese, if regulators in the U.S. and Canada have their way. A newly released 189-page report from the U.S. Food and Drug Administration and Health Canada concludes that there is “a 50- to 160-fold increase in the risk of listeriosis from a serving of soft-ripened raw-milk cheese, compared with cheese made from pasteurized milk.” As a result, the regulators suggest they want to see raw milk cheeses like camembert and brie either subject to unprecedented testing, processing similar to pasteurization, or else banned completely. I should note they also offer the option of doing away with the 60-day aging requirement for cheese, as a possible way to reduce the time pathogens have to multiply in the cheese, but it’s offered as kind of a straw man, since it “does not consider the effect of removing the regulation on the risk of illness from other pathogens…” (That risk assessment should be worth another few lengthy reports.)
The risk certainly sounds serious…until you read closely the full 189-page report and learn that the FDA-Health Canada conclusion about “a 50- to 160-fold increase in the risk” is based entirely on estimates and mathematical predictions, rather than real-life data on illnesses from the soft raw milk cheeses.
Even more remarkable, the actual real-life data presented in the report of illnesses worldwide from listeriosis in soft cheese over a 23-year period between 1986 and 2008 show not a single documented illness in the U.S. from listeriosis due to tainted brie or camembert. That data, in a table on page 17, documents four outbreaks of listeriosis in cheese in the U.S. over the 23-year period, but all four are from raw milk queso fresco cheese, a soft cheese served fresh, without being aged the required 60 days, and thus illegal in the U.S. (It gets made illegally, often in the Hispanic community, sometimes even mixed in bathtubs, which has earned it the nickname “bathtub cheese.”)
In Canada, two cheese-related outbreaks, which sickened 58 people, are attributable to “multiple types” of cheese.
Indeed, the researchers were only able to document 20 outbreaks of illness from listeria in all cheesesfrom all kinds of milk worldwide over the same 23-year period—less than one per year…and, according to the report, “half involved cheese made from unpasteurized milk.”
So in actuality, we’re talking about ten listeriosis outbreaks worldwide from raw milk cheeses over a 23-year period—a tiny number by any stretch of the imagination.
By way of comparison, the U.S. Centers for Disease Control documents more than 1,000 outbreaks, resulting in between 15,000 and 30,000 illnesses, from all foods each year in the U.S.
So, given the absence of a single documented illness from 60-day-aged soft raw milk cheeses over 23 years in the U.S., and just the possibility of a few dozen illnesses in Canada, how do the FDA and Health Canada come to their conclusion that such cheeses are up to 160 times more risky than pasteurized soft cheeses? As I recall my grade school math, zero times any number equals zero. The regulators do it via “mathematical / probabilistic modeling… to estimate the risk per serving of Camembert-like cheese in both countries,” according to the report.
The source of their data? “Literature data, previous risk assessments …and expert sources” in Canada and the U.S. What about information from real-life farms and cheese producers? There is none: “Risk was expressed on a ‘per serving’ basis because (of) the lack of data on overall levels of cheese production (particularly for small cheese makers)…”
In the absence of real-life farm or producer data, the report relied heavily on a single study published in 2003 “that include values for both the prevalence and the level of contamination of soft-ripened cheeses in the United States and Canada …from a random sample of cheeses obtained at retail in Maryland and California … as part of a larger survey of ready-to-eat foods.”
Moreover, while the 2003 survey measured the presence of listeria monocytogenes (L. monocytogenes) in various foods, including deli meats and salads, along with assorted cheeses (the overall presence of L. monocytogenes. was 1.82%) there is no mention in the study of any illnesses resulting from the pathogens the researchers found, strongly suggesting that no illnesses resulted from the L. monocytogenes (the sample foods were collected over 14 to 23 months, plenty of time to have recorded illnesses).
So essentially what happened is that the FDA-Health Canada report writers worked backwards, and extrapolated what might happen on farms and at cheese-making facilities, based on the presence of L. monocytogenes shown in the 2003 study of ready-to-eat foods in retail markets. And the summary report notes near the end: “The prevalence and level of contamination in Canada and in the U.S. rely on a single study to infer in-plant environmental L. monocytogenes contamination. Additional information about the prevalence of contaminated lots and contaminated cheeses within lots is needed.”
Despite such obvious and serious limitations, the FDA-Health Canada report spends most of its 189 pages describing complex mathematical formulas and computations and providing diagrams of the cheese making process before, presto, spitting out its conclusion that anyone who consumes a serving of raw milk soft cheese has a significant chance of becoming ill, up to a 160 times greater chance than if he or she ate pasteurized cheese.
All the fancy mathematical gyrations and fear mongering come in the face of a simple and seemingly successful public health initiative implemented after World War II, whereby both the U.S. and Canada require all soft and hard cheeses to be aged a minimum 60 days, to allow for any pathogens that might be present to die off. (Quebec has no aging requirement, meaning that soft cheeses can be sold immediately after production.)
Serious makers of raw milk soft cheeses have long chafed under the 60-day aging requirement, since such cheeses don’t age well in terms of taste, and are best served shortly after production. But American and Canadian makers have learned to adjust—some even using pasteurized sheep, goat, and cow’s milk to avoid the 60-day aging requirement–and as the FDA-Health Canada’s own 23-year analysis showed, there have been remarkably few illnesses.
But the problem gnawing at the FDA, in particular, has been the growing popularity of raw milk, including raw milk cheeses. Actual sales of soft raw milk cheeses in the U.S. aren’t known, but the American Cheese Society reported in a December 2012 survey that 59 per cent of more than 200 respondents to a survey it conducted of members last year said they make raw milk cheese. The ACS has grown to 1,500 members since its founding in 1983.
The current assault on soft raw milk cheeses actually began nine years ago, when the head of the FDA’s dairy division, John Sheehan, concluded as the lead author of a 2004 article in Food Safety magazine, that there were fundamental problems with the FDA’s 60-day cheese-aging rule. New research, he wrote, suggested “that 60-day aging is largely ineffectual as a means of reducing levels of certain pathogens in cheeses. With this information in hand, FDA is now developing a risk profile for raw milk cheeses, which will aid in the Agency’s assessment of the requirements for processing these cheeses.”
There was particular concern, he and his co-authors stated in the article, over listeria in cheese. They wrote about an “action plan” for cheeses that “have resulted in new efforts to improve the safety of cheeses.”
Shift to the report just issued, of which Sheehan is an author. Instead of acknowledging the paucity of recorded illnesses from raw milk soft cheeses during the 23-year period covered by the report, it suggests there is some kind of lurking problem: “The United States and Canada continue to experience sporadic illnesses and outbreaks of listeriosis associated with the consumption of cheese, particularly soft and soft-ripened cheese.” Presumably the “sporadic illnesses and outbreaks of listeriosis” refer to the four outbreaks from queso fresco and the two outbreaks in Canada over the 23-year period. By the way, the report also notes that the overall incidence in the U.S. of listeriosis is “approximately 3 cases per 1 million individuals…” That is very low, especially when compared to something like autism rates, which have soared to one in each 88 children.
As further justification, the report notes that over the same 23-year period “there were a total of 137 recalls of various types of cheeses, of which 108 (79%) were Listeria-related. In Canada from 2004 through mid-2009 there were 15 cheese recalls, of which 11 (73%) were Listeria-related. A wide variety of cheeses were involved in these recalls.”
But as evidenced by the number of outbreaks and illnesses in the U.S. and Canada, those recalls were carried out after cells of L. monocytogenes were discovered in the cheeses, but not as a result of actual illnesses. Recalls nearly always occur well after whatever foods in question have been sold, and thus consumed. And it’s been well established that the presence of small numbers of L. monocytogenes cells rarely cause illnesses; for that reason, the European Union doesn’t institute recalls for food containing small numbers of cells.
Now, I could accept a nearly completely conceptual analysis of this sort (aside from the terrible expense and time sink associated with it) if it was issued with the intent of sparking additional reality-based research. But it’s not. It’s not only a solution in search of a problem, but it has been issued with one intent and one intent only: to fear monger about the supposedly huge risk associated with eating raw milk soft cheeses like brie and camembert…as a lead-in to stricter rules against raw milk soft cheeses.
And the report has succeeded in that respect. A number of publications that cover the food safety arena have picked it up and run with the fear mongering about raw milk cheeses as if it is fact, notably Food Safety News and Barfblog, among others.
What’s most troubling is that the FDA has shown itself so completely opposed to raw milk and raw milk products under any circumstances that it would move to limit or ban such products even without evidence of a single illness. Talk about an ideological commitment.
Consumers have until April 29 to submit comments on this study. I suggest you let the FDA know about any concerns you might have. If there aren’t many comments, you know the FDA will simply assume everyone supports their direction, and their direction is crystal clear after nine years of effort.© Food Safety News