Editor’s note: We take a break from our regularly scheduled program of food nutrients and depression, analysis of beef liver, phytic acid, surviving rattle snakes and forest fires, and enjoying fantastic sunflower pictures, to bring you a lively discussion of raw milk. Below is my own response to an analysis of my “raw milk white paper” by the Weston A. Price Foundation’s Ted Beals.
Raw milk is a great drink but there are a good many people who will ascribe to it many diverse properties. One of these properties is that the beneficial bacteria and enzymes in the milk protect it from contamination. There actually is not debate over whether raw milk has some pathogen-killing properties. It does. The key question is whether there is enough research-based evidence that raw milk kills pathogens quickly and thoroughly enough to provide safety guarantees to consumers. Raw milk consumers should be keenly interested in this topic.
As background, I write about raw milk politics on the Ethicurean and recently published a “white paper” on raw milk’s ability to kill pathogens. A good part of my rationale in writing the white paper is posted on the Ethicurean (Memo to raw-milk advocates: Improve information or get sued, an article that is still receiving lively comments).
Basically, raw milk advocates have a poor track record in providing raw milk safety information to consumers. If you have heard that raw milk can never be contaminated because it has built-in safety mechanisms that kill pathogens, then you do not have the full story. Raw milk, like all other food on the planet, offers no safety guarantees. It might kill pathogens; it might not kill pathogens. That conclusion is at least what I found in reviewing the research on raw milk’s ability to kill pathogens in my “raw milk white paper.” The issue is even more complex because each pathogen (e.g, Campylobacter, E. coli O157:H7, Listeria, Salmonella) behaves differently in raw milk: some populations die-off over time, some grow for awhile then die off, some show no growth or death when put in raw milk. Temperature is also important; most pathogens will grow in raw milk left out on the counter at room temperature; Listeria may die.
In general, Beals does not address the key concern in my paper: Does the research on competitive exclusion provide consumers a guarantee of raw milk safety? Does raw milk kill pathogens quickly and completely so that consumers can be sure of safe milk even when the milk is subject to low levels of contamination? I do hope that if there is an on-going discussion of my white paper, the discussion centers on this key question. However, in this particular instance, it did not.
Appeal to authority
Beals opens his piece by incompletely describing my background and questioning my motives. I have a Ph.D. in political science and I had an expert in infectious disease review my paper to ensure that I described the research properly. Recognizing that I am not a scientist, I have no issue revising the white paper (or anything else I have written) so that it provides the most useful information to consumers. This material is not being presented in order to begin a debate or to “win” an argument, but rather to provide an unbiased viewpoint on a body of research that I read about raw milk killing pathogens.
Beals positions himself as an expert and should provide us with expert quality analysis. He has many years of experience in the medical field and has made a study of dairy microbiology in his retirement. He provided expert testimony for California’s raw milk producers last year in an on-going political battle. He lends his expertise to the WAPF and the FTCLDF. You, the reader, may have paid for his analysis of my white paper or at least for its production and distribution as a member of the WAPF. Beals is one of the foundation’s biggest guns and I feel honored that they used their best for this particular task. His analysis should be of the highest quality even if it does not address the core concern in my white paper. Let’s take a look.
Pathogens are killed
Beals’ key argument is that raw milk does kill pathogens, a point I do not actually take issue with. However, in his review of the same scientific literature I examine, he comes to a different conclusion than I. On the second page of his review he discusses my analysis: “She fails to inform readers that in each of these papers, the results and conclusions of these authors is that the inoculated pathogens are killed.”
More specifically, Beals reviews a study by Massa and colleagues on E. coli O157:H7 and states: “Even with these huge numbers of pathogens, the E. coli O157:H7 strains failed to grow and died off gradually.”
I direct readers to the abstract of the Massa paper because that is as far as we need to go to examine Beals’ claim. In the abstract, Massa and colleagues state:
“There was essentially no change in the viable population of three EHEC strains for up to 14 days. The remaining four strains showed an increase in population from <2 log to 3 log cfu ml^−1 in a time period of between 9 and 17 days.”
If you own the white paper, you can see the data visualization that makes the finding clear. However, let’s review this point for further clarity:
Beals: “..the E. coli O157:H7 strains failed to grow and died off gradually.”
Massa et al.: “The remaining four strains showed an increase in population from < 2 to 3 logs cfu / ml in a time period of between 9 and 17 days.”
Rose: There is no evidence that raw milk kills pathogens enough to provide consumer guarantees. In fact, this paper shows that some deadly strains of E. coli O157:H7 actually grow in raw milk.
Let’s take a look at another example of Beals’ analysis.
On the second page of his review, Beals describes the Pitt paper on Listeria monocytogenes. He says: “After fifty-six hours no viable cells of L. mono were detected.” As I describe in my paper, that experiment was at room temperature and it does provide evidence that raw milk kills Listeria monocytogenes. Beals and I agree. But the story does not end there. In reference to Beals’ more general criticism of my white paper – that the research shows “the inoculated pathogens are killed” – this particular paper on Listeria monocytogenes finds something different when the milk is kept at refrigeration temperatures:
“After incubation at 4°C and 7°C for 288 h (12 days) the true psychrotrophic nature of L. monocytogenes was apparent. The concentration of L. monocytogenes increased from an initial inoculum of approximately 10^4 cfu/mL to a final density of 10^6-10^7 cfu/mL at 4°C and 10^7-10^8 cfu/mL at 7°C.”
“Psychotropic” means that the bacteria can survive or thrive in a cold environment. In contrast, bacteria such as Salmonella generally do not grow in the refrigerator, but can grow rapidly at room temperature (again illustrating the complexity of the topic since each pathogen, and sometimes even different strains of a species, behave differently in raw milk).
Again, for clarification on the Pitt paper:
Beals: “She fails to inform readers that in each of these papers, the results and conclusions of these authors is that the inoculated pathogens are killed.”
Pitt et al.: “The concentration of L. monocytogenes increased from an initial inoculum of approximately 10^4 cfu/mL to a final density of 10^6-10^7 cfu/mL at 4°C and 10^7-10^8 cfu/mL at 7°C.”
Rose: Raw milk does kill pathogens in some circumstances, as I describe in my paper. In other cases the pathogen counts stay the same or grow. As stated in the paper, Listeria monocytogenes numbers increased (were not killed) in a cold environment.
In the literature on competitive exclusion, researchers inoculate milk with large numbers of pathogens – seven or more logs (tens of millions of pathogenic cells). These large numbers are not a realistic snap shot of actual outbreak situations. That is a point I make in my paper. Beals spends a great deal of ink space describing why these “huge number” studies are flawed and why they should not serve as a basis for judging the pathogen-killing ability of raw milk. Beals’ review reads as if the weakness is in my analysis – as if I had never considered this issue.
Reading Beals, I recognized a circular argument – WAPF appears to be using these same “large number” studies to tell consumers that raw milk kills pathogens and yet Beals is critical of their application to this research question. To give him the benefit of the doubt, I asked him by email if he has other studies (using smaller numbers of inoculum) that support his conclusions that raw milk kills pathogens. He has not responded as of the time of writing this response.
As it turns out, we need to look no further than my own paper for studies that used smaller numbers of pathogens to determine if they die, stay the same, or grow in raw milk.
A 1997 study of E. coli O157:H7 examined the growth and decline of the bacteria in pasteurized and raw milk (Wang et al. 1997) using inocula of two sizes: 3 logs and 5 logs. If we are concerned that the “huge number” studies represent situations in which the raw milk pathogen-killing properties are overwhelmed and, at the same time, we argue that raw milk does quickly and completely kill pathogens in a small contamination event, we would expect smaller inocula studies to show increasing effectiveness in pathogen-killing.
Put simply, the less E.coli you spike the milk with, the better job the milk should do killing it under Beals’ argument.
In fact, we do not see this. The Wang study shows that the 3 logs of pathogenic E. coli grow at temperatures of 8 degrees Celsius and above for the first few days. The bacteria do then begin to decline. At 8 degrees, 15 degrees, and 22 degrees Celsius, the pathogen grows first before it begins to decline. In only one experiment was it killed completely — at 22 degrees Celsius the researchers do not recover the bacteria on Day 14, but did so on Day 7. Indeed, consumers could chug the milk on Day 14 without worry if they kept it under those warm conditions. Those who chugged it on Day 7 or earlier (and perhaps a few days later) could well end up in a hospital. In the best case, the E. coli in milk kept at 5 degrees Celsius did not grow at all and did decline over the four-week period. Consumers would have been zapped by this milk as well (because the pathogen was never killed completely and the infectious dose is low), but it declines in raw and pasteurized milk nevertheless.
The researchers do not present the data from their 5 log experiments. They do state that the results from the two sets of experiments are similar, contrary to what Beals would predict.
As a second test of Beals’ argument, there is a study underway at UC Davis by Linda Harris in which researchers inoculate commercial raw milk from Organic Pastures and Claravale with Salmonella. They inoculate the milk with only 2 logs of Salmonella and hold the milk at room temperature and at refrigeration temperature. As an aside, Harris’ research is noteworthy in that it uses raw milk from legal, licensed and inspected raw dairies in its experiments.
Since Harris is using smaller counts of Salmonella – hundreds of cells instead of tens of millions – we would expect to observe the milk’s good bacteria and enzymes working hard to protect the consumer. Minimally, it should be less “overwhelmed” by the lower levels of Salmonella put into the milk by the researchers. Harris’ work is unpublished but you can see preliminary evidence in slides 15 and 18 of this PDF.
From page 15 of version 1.0 of my raw milk white paper:
Harris also studied fairly low levels of Salmonella in milk – two logs of Salmonella or hundreds of cells. This low level of Salmonella may be closer to real-life contamination events, particularly compared to studies that added seven logs of the pathogen (10,000,000+ cells).
Preliminary findings from Harris’ work show that the Salmonella numbers do not decrease over a seven-day period when the milk is held at refrigeration temperatures. Held at room temperatures, the Salmonella numbers grew from hundreds of cells to hundreds of thousands of cells within two days.
Let’s read Beals again: “She fails to inform readers that in each of these papers, the results and conclusions of these authors is that the inoculated pathogens are killed.”
Now read Harris as cited by Jay-Russell: “Conclusion: 3+ log increase of Salmonella at room temperature” (slide 18 here: PDF).
Rose: Raw milk, like life, provides no guarantees. In fact, raw milk from California’s licensed raw dairies can grow Salmonella if left to clabber outside of the refrigerator.
In fairness to Beals, the Harris research is not yet published but I do know that the researcher is easily found and, in fact, I describe the study in my white paper. I asked Beals for clarification on this omission but he has not responded to my email.
Allow me to add as well that I learned about the “huge numbers” literature in the first place from raw milk advocacy sources. If the “huge numbers” problem is such a critical flaw in the research, why then do advocacy sources cite it as evidence that raw milk is safe? It is not logical that the Doyle study would be cited for years as evidence that raw milk is “uniquely safe” only now to find out that the Doyle experiment was flawed in the first place according to WAPF because Doyle used “huge numbers” to conduct the experiment. Where are the non-flawed experiments that WAPF bases its claims on? They should form the core part of a WAPF rebuttal should there be one.
The most intriguing part of Beals’ analysis is that I fail to discuss the origin of pathogen strains. Let’s discuss that now.
As background, Beals offers a graph of the decline in Campylobacter jejuni in raw milk. He questions why I do not use it. I did not include the graph because I needed permission from the publisher and I did not have the support of a million dollar foundation to get permission for me. Basically, the graph I use and the graph he uses tells the same story: consumers have no guarantee that raw milk will kill Campylobacter jejuni quickly enough and thoroughly enough to guarantee consumer safety. I have no intent to deceive people using a different graph. Again, look closely at the graph, especially the x axis that shows the number of days it takes before the Campylobacter population is completely dead. Beals’ blanket (and oversimplified) statement that raw milk kills pathogens does no service in educating consumers. Consumers would be sick if they drank this milk fresh.
It is from this graph that Beals’ comment emerges about the nonhuman source of that Campylobacter jejuni strain.
Referring to the graph, Beals implies that only three of the strains actually matter: “Only three of these strains are from human sources (a fact that Rose omits from her report). All strains do show reduction in pathogens with time. The only line without a steep decline (the steeper the line the faster the pathogens were dying) tracks a nonhuman strain.”
Indeed, the strain that does not die off very quickly and is still present in large amounts on Day 21, is not of human origin. What that means is that the strain was isolated in a nonhuman specimen. The strain may or may not make humans sick. We do not know. Even Campylobacter jejuni isolated in humans may or may not have made those same humans sick. In this particular Campylobacter jejuni study, these researchers had access to these various strains in their lab and they used them.
Our key question: Does the origin of the strain matter? Not really.
What matters are these questions:
(1) Can the strain make us sick?
(2) Can the strain make it into raw milk?
I do not have information on the first point. Some Campylobacter jejuni strains make us sick, some do not. I asked an expert in Campylobacter jejuni to confirm whether this particular strain has been implicated in an outbreak. The strain number could not be found in the DNA typing databases where outbreak strains are catalogued. What this means is that we do not have enough information to know if the strain is pathogenic. It is like all Campylobacter jejuni strains: it may or may not be pathogenic.
Let’s explore the second question: Can the strain (the one that won’t die off completely) make it into raw milk?
Let’s consider some examples. If we are concerned about pathogens in raw milk, what sources would concern us most?
Would chicken sources of Campylobacter jejuni concern us?
Perhaps. Chickens hang out near cows and goats and could possibly contaminate raw milk.
Would human sources of Campylobacter jejuni concern us?
Perhaps. People milk cows and goats and could contaminate the milk.
Would bovine sources of Campylobacter jejuni concern us?
Absolutely. Bovine are cattle and cattle give milk. Perhaps not all bovine strains would make people sick, but some do.
Back to the campy graph. What is the source of that one strain that does not die off?
Take a guess.
Remember Beals’ words:
“Only three of these strains are from human sources (a fact that Rose omits from her report). All strains do show reduction in pathogens with time. The only line without a steep decline (the steeper the line the faster the pathogens were dying) tracks a nonhuman strain.”
What is the source of the strain?
The answer: Bovine.
We should only be concerned about this strain if our topic has something to do with cows. The strain that does not die off completely is a bovine strain. To answer our first question, “Can the strain make it into raw milk?,” the answer is “Yes, if a cow is involved.”
I asked the Campylobacter jejuni expert if bovine sources of Campylobacter jejuni have made people sick. She responded, “There are many Campylobacter jejuni strains of bovine origin in these databases that have caused human illness.” She added an educational tidbit: “Humans have not been shown to be reservoirs (carriers) of Campylobacter jejuni. It is a transient infection, so the ultimately humans are picking up these strains from animals or a contaminated environment (food, water, mud, etc.).
And it may all be irrelevant
This week, food injury attorney Bill Marler posted on his blog part of his own response to a WAPF review of his work (Marler response to WAPF). Part II includes the issue I raise in my white paper: “Does raw milk kill pathogens quickly and thoroughly enough to ensure consumer safety?”
What is fascinating about the review is not that Bill Marler and I agree on something related to raw milk, but the comments left by a couple of scientists. Read the Marler review and read the comments carefully because these experts in microbiology are saying that even though the pathogen levels appear to decline in some of these lab experiments, the milk may still be infective. Campylobacter jejuni, in particular, may appear to decline in lab tests but the apparently dead cells may still be able to infect you. Apparently the cells are actually dormant and come back alive when they enter a better environment like a human gut. One commenter also mentions that the infective dose of Campylobacter jejuni in milk is much lower than it is in other foods. These issues are far beyond this review, but perhaps are issues the WAPF can tackle in its rebuttal.
The core of my argument in the white paper is that raw milk kills pathogens except when it doesn’t. The rest of the details in the Beals’ review are somewhat irrelevant but I’ll address them briefly. Read Beals’ review for background on these points (PDF).
• 5 log reduction
In my paper, readers get a realistic tool for judging the evidence and that tool is this: Does raw milk kill pathogens thoroughly enough and quickly enough to ensure consumer safety? I describe how to examine the data to determine the answer to that question. The 5 log reduction is a measurement used in food science. Beals criticizes me for mentioning this measurement but what is surprising is that Beals does not define what he means by “kills pathogens.”
If my raw milk has 500 cells of E. coli O157:H7 and I take a big swig, would the death of one cell in 500 be enough for Beals to declare that the “raw milk kills pathogens?” If so, would that definition be useful to me as a consumer? As a consumer myself, I do not want to split the proverbial scientific hair over the definition of “pathogen-killing,” I want to know if the milk would make me sick. Indeed, that is the framework I provide in my paper that Beals criticizes.
• “Competitive exclusion” versus “competitive inhibition”
Beals is concerned that I use the term “competitive exclusion” instead of his favored term “competitive inhibition.” WAPF uses the term “competitive exclusion” in a 2007 press release. Both terms are appropriate. What I cannot imagine is that consumers would care about this terminology issue when print space could be used educating them about whether raw milk kills pathogens enough to guarantee its own safety.
I do make reference to Listeria monocytogenes and miscarriage in the context of the question: Does raw milk kill Listeria monocytogenes enough that pregnant women need not be concerned? Personally, I think it is a good question and I certainly do not have the answer. The minimum infective dose of Listeria monocytogenes is about 1,000 cells and we end up with fluish symptoms. The minimum number of cells that need to cross the placenta is apparently one cell. The odds of a cell making it into the placenta are awfully long, but if it makes it, baby is very likely to be born early or die. That too is a fact. I do not know what the odds are of that happening and neither does Beals. I do not write this to “scare” people away from raw milk. I know women who drink raw milk, eat soft cheeses, and sushi in their pregnancies knowing that they are consuming high-risk foods. It is their decision. I feel great when I drink raw milk and I understand the appeal of choosing to drink it in pregnancy.
I discuss none of my own choices in the white paper because they really are irrelevant but I will say here that I did not consume raw milk in my recent pregnancy. With my history of depression in pregnancy, the last thing I needed was to worry about what was in my food. I grew my own food and avoided all high-risk foods whenever possible. I do not expect to be judged harshly for my decision by raw milk consumers because I know that raw milk consumers place a high value on the right to choose. I exercised my own right to choose what was best for my family under my own unique circumstances.
• Raw milk kills pathogens better than pasteurized
An implied criticism of me is that I leave out a study on Listeria monocytogenes that shows that raw milk outperforms pasteurized milk in killing pathogens. Beals mentions the study at the bottom of the second page of his review. Someone sent me the study in August and I considered adding it to the paper at that time but it offers no new evidence for the paper unless I want to extend the review to other pathogens. I will include it in version 2.0. To clarify, I do already discuss the issue of raw versus pasteurized milk in the paper and the evidence is favorable toward raw milk.
Criticizing me for leaving out that particular study does make me wonder where all the citations are that support Beals’ argument. From his second page: “A lot more on this subject can be found in the scientific literature, but I will focus on the reports that Rose cites in her paper.” Indeed, he does add a paper I missed and criticizes me for excluding it.
The body of literature on this subject appears thin, and many papers are old (the Doyle paper was published in the early 1980′s, and the raw milk was probably nothing like today’s WAPF style raw dairies). I implore Beals to share additional research papers on this subject, especially anything that is more current.
I asked Beals by email for his omitted citations but he has not responded.
• “Position Paper”
Beals refers to my paper as a “position paper.” I actually take no position in the paper and that may be part of the confusion. Consumers need to make their own decision and I do not actually care one way or another what they choose. If folks are like me and know there is a real long chance of getting a parasite from sushi or a bad bug from raw milk and consume it anyway, more power to them.
Beals argues that my white paper is emotional — I do mention miscarriage and preterm birth as a possible consequence of a Listeria monocytogenes infection. I never considered that speaking about issues of concern to us women of child-bearing age might appear emotional to a 75-year-old man. So be it.
If this current writing appears emotional, it is because I am right now feeling profound and utter disappointment that the analysis of my white paper by a respected physician was so superficial and full of inaccuracies about the scientific literature.
Black and white
Some people need their world to be white and see the rest of the world as black. You will find interesting black-and-white information at the RealMilk.com website. I hold the apparently controversial notion that the world is far more complicated. If you enjoy reading more diverse opinions, you may be interested in the Raw Milk White Papers website. The tag line at the site is:
“I don’t tell you about raw milk research, I show you.”
I provide data visualizations that help you understand the research on raw milk. I do not expect raw milk to be more than it is – a really great food. I do not expect it never to be contaminated. It can fall victim to this deteriorating and pathogenic world just like everything else can. It is neither “uniquely safe” nor “uniquely dangerous.”
As I get a wild hair, I will put some useful information there for consumers. I sell the white papers to support the site, but I include less thorough articles for free.
I should mention that the white paper in question is available there for seven bucks. That seven bucks is the only way I can figure out how to fund my no-nonsense approach. (Come to think of it, however, your money may be better spent buying my book on nutrients and depression since we all may have a mild case after reading these 4,000 words — I leave you with a nice image below.)
If you are a blogger or food writer and would like a review copy, feel free to contact me.