Michael Pollan writes of the lethal E. coli O157:H7,
Most of the microbes that reside in the gut of a cow and find their way into our food get killed off by the strong acids in our stomachs, since they evolved to live in the neutral pH, environment of the rumen. But the rumen of a corn-fed feedlot steer is nearly as acidic as our own, and in this new, man-made environment new acid-resistant strains of E. coli, of which 0157:H7 is one, have evolved – yet another creature recruited by nature to absorb the excess biomass coming off the Farm Belt. The problem with these bugs is that they shake off the acid bath in our stomachs – and then go on to kill us. By acidifying the rumen with corn we’ve broken down one of our food chain’s most important barriers to infection. Yet another solution turned into a problem.
We’ve recently discovered that this process of acidification can be reversed, and that doing so can greatly diminish the threat from E. coli 0157:H7. Jim Russell, a USDA microbiologist on the faculty at Cornell, has found that switching a cow’s diet from corn to grass or hay for a few days prior to slaughter reduces the population of E. coli 0157:H7 in the animal’s gut by as much as 80 percent. But such a solution (Grass?!) is considered wildly impractical by the cattle industry and (therefore) by the USDA. (82)
Even before looking at the literature, I wasn’t convinced that this would be much of a solution. Impressive as an 80 percent reduction may sound, if, as Pollan has told us (in the paragraph before the ones I’ve quoted), ten of these bacteria can kill us, how much good would killing 80 percent of the bacteria do? Cutting a bacterial population of millions by 80 percent would leave many more than the ten required to cause a lethal infection. One wouldn’t expect all of the bacteria in a cow to be ingested by an unsuspecting beef-eater, but substantial risk would remain.
Then, there’s the question of what the literature actually says on the matter. As James McWilliams and Bill Marler have pointed out, a 1998 study by Russell and others has been widely misinterpreted to support claims like the one Pollan makes here. This study found that feeding cattle grass for a few days did indeed decrease the total E. coli populations in the cattles’ rumen and colon and also decreased the acid resistance of the E. coli. However, they did not detect any of the E. coli O157:H7 in the samples isolated from their cattle, even before feeding them grass for a few days. Instead, they did a lab experiment which showed that high concentrations of glucose, like those found in a grain-fed cow’s digestive system, induced acid-resistance in commercially available E. coli O157:H7. From this, the researchers expected that grass-feeding cows for a few days prior to slaughter would decrease the risk of O157:H7 infection.
However, Pollan doesn’t cite that study. Instead, his list of sources cites a book that Russell self-published in 2002, titled Rumen Microbiology and Its Role in Ruminant Nutrition. That book has since gone out-of-print, I don’t have access to a library copy, and I wasn’t able to find any peer-reviewed literature in Russell’s name that directly supported Pollan’s claim.
Even if Russell’s elusive book says exactly what Pollan claims here, a reader might do well to know a bit more. If somebody is trying to decide what to have for dinner and is concerned about E. coli, it isn’t so important what Jim Russell in particular has found. Such a reader is likely to want a more balanced view of the literature, rather than the opinion of one scientist. Russell is far from the only scientist who has conducted studies on the effect of diet on O157:H7 populations in cattle, and numerous studies have cast doubt on the conclusion that grass-feeding cattle reduces the risk of E. coli O157:H7 infection. Dale Hancock and Tom Besser, of the College of Veterinary Medicine at Washington State University compiled a review of available evidence and found that the conclusions of the 1998 paper (which were very similar to Pollan’s claim) “have not been corroborated by numerous scientific papers from research groups around the world.”
The review of Hancock and Besser (as well as the popular works by McWilliams and Marler) will give you plenty of evidence that grass-fed beef isn’t immune to the deadly E. coli, and I don’t think it would be useful for me to restate all of that information. Instead, I’ll quote what I think is some of the more compelling evidence from Hancock and Besser:
First, consider that a substantial number of papers by researchers around the world have documented that cattle on pasture or rangeland (i.e., eating grass) have E. coli O157:H7 in their feces at prevalences roughly similar to those of confined, grain-fed cattle of a similar age (Sargeant et al, 2000; Fegan et al, 2004b; Renter et al, 2004; Laegreid et al, 1999) .
They also point out,
Several groups have demonstrated that E. coli O157:H7 has a unique predilection for colonization of the recto-anal junction, different from most other E. coli that colonize cows (Naylor et al, 2003; Sheng et al, 2004; Low et al, 2005; Greenquist et al, 2005).
Thus, even if a few days of grass-feeding reduced E. coli populations in the rumen and colon, that wouldn’t necessarily be reason to expect it to decrease O157:H7 populations. For the most part, O157:H7 lives elsewhere in the cow.
While it’s possible that Russell’s self-published, out-of-print book says something in support of feeding cows grass for a few days to reduce the risk of infection, it’s far from clear that this is the solution to the E. coli problem that Pollan might lead us to believe.
Incidentally, the 1998 study of Russell and others happens to contradict an important point that Pollan makes in the excerpt that I’ve quoted above. Pollan claims that “the rumen of a corn-fed feedlot steer is nearly as acidic as our own.” I found this more than a bit perplexing, because we humans are not ruminants, and we do not have rumens. In the context, my best guess is that Pollan means to say that the rumens of corn-fed cattle are almost as acidic as human stomachs. However, the 1998 study reported that as the proportion of grain in the bovine diet increased, “ruminal pH remained essentially constant.” Even among cattle on a diet of ninety-percent grain, ruminal pH was above 6, considerably less acidic than the human stomach, which has a pH of less than 4. (The study did report that grain-feeding made bovine colons significantly more acidic.)