Introduction

In 1997 Hudson Foods in Columbus, NE, recalled 25 million pounds of ground beef because it was positive for E. coli O157:H7. A few months later Beef America, a regional packer in Norfolk, NE, had a ground beef recall. Both recalls bankrupted the companies. That got our attention in Nebraska especially, but across the national beef industry as well. Meat and Poultry Magazine calculated that the ten-year cost of E. coli O157:H7 from ’93 to ’03 was about $2.7 billion. The public health cost was estimated at $989 million per year. We have learned a considerable amount about E. coli O157:H7 in the past 15 years.

Background on E. coli O157:H7

Prior to 1998, it was believed that the prevalence of E. coli O157:H7 in feedlot cattle was low - perhaps 2-3 per cent of all cattle being positive. New analytical techniques essentially increased this prevalence by a factor of about 10. This procedure uses an IMS technique which effectively ”finds” the E. coli O157:H7 in cattle faeces. Faeces have been the most common material measured to determine if cattle are positive for E. coli O157:H7. Cattle faeces contain hundreds of types of bacteria measured in millions or more of bacteria per gram. Some of these are coliform bacteria, including generic E. coli, which are not typically pathogens. So finding the E. coli O157:H7 is like “finding the needle in the haystack”.

Using the IMS and other more sensitive analytical techniques, we followed 100 steers through a summer feeding period, sampling them weekly. Prevalence was low (Figure 1) until mid-July and then increased from less than 10 per cent to 80 per cent over a two-week period. Prevalence then declined to about 30 per cent. We did not identify reasons for the dramatic change (Khaitsa et al., 2003). We sampled cattle in 74 pens in five commercial feedlots (Figure 2). Prevalence averaged 30 per cent in the summer and each of the 44 pens sampled in the summer had at least one positive animal (Smith et al., 2001). The range was from 1 to 80 per cent positive animals - the variation largely unexplained (Smith et al., 2001). Prevalence was only 6.1 per cent in winter months and 14 of 30 pens had no positive animals. Still the range in prevalence was 0 to 56 per cent. We concluded that probably all feedlots have E. coli O157:H7- positive cattle and prevalence is higher in the summer than winter and also that prevalence is highly variable and the variation is largely unexplained.

Elder et al. (2000) sampled cattle entering the packing plant and then the carcasses after slaughter. They showed a positive relationship between E. coli O157:H7-positive cattle entering the plant and the number of positive carcasses. It is believed the carcass gets contaminated from the hide as it is being removed. Because the contamination is on the surface of the carcass, the trim has the E. coli O157:H7 and is mixed into the ground beef as it is processed. The packing industry has done an excellent job in intervention (Koohmaraie et al., 2005). This includes carcass washes, steam pasteurization, organic acid rinses and “test and hold”. Test and hold is a procedure where trimmings are sampled and tested for E. coli and “held” before processing. Only trimmings testing negative are ground. The small number of positives are heat-processed to destroy the organism. With these excellent interventions, the percentage of positive samples declined from about 0.8 per cent in 2000 to about 0.2 per cent over the past four years.

Unfortunately, 2007 was not a good year for ground beef safety and recalls. There were eight recalls in 2006 and all of them were initiated because of company sampling. In 2007, there were 20recalls and nine of these resulted from illness investigation. This was(is) a problem for the cattle industry and health officials looked for reasons why E. coli 0157:H7 seemed to be a greater problem in 2007 than in the previous four years. Because the ethanol industry grew in 2007 and had attracted attention because of the effect on corn prices, some theorized that feeding ethanol byproducts was the cause of the E. coli O157:H7 recalls. Late in 2007, research from Kansas State showing a connection between distillers grains (DG) feeding and E. coli O157:H7 shedding was reported. The media, looking for a good story, made the Kansas State research findings look like a food safety crisis. Let’s look at the data from Kansas State and Nebraska and see if it confirms the media hype.

Distillers Grains and E. coli O157:H7 Shedding

The Kansas State researchers have reported four studies. Jacob et al. (2008a) reported a study using 370 feedlot cattle sampled at 122 and 136 days on feed. Prevalence overall was fairly low (under 10 per cent). On day 122, cattle were statistically more likely to shed E. coli O157:H7 when fed 25 per cent DG in the diet. On day 136, there was no effect of feeding DG. Jacob et al. (2008b) sampled cattle for 12 weeks during the feeding period. Faecal samples collected from the pen floor. Feeding DG significantly increased E. coli O157:H7 shedding although there was no difference on five of the 12 sampling periods.

The third Kansas State study was a challenge experiment where calves were inoculated with nalidixic acid resistant E. coli O157:H7. This allowed the researchers to estimate the number of E. coli O157:H7 shed. Faecal samples were collected for 42 days. E. coli O157:H7 shedding was not different for calves fed DG during the first five weeks but was statistically greater during the last week of sampling. Based on these three studies, the Kansas State researchers concluded that DG feeding increased E. coli O157:H7 shedding. In each of the three experiments, there were sampling times when DG statistically increased shedding, however, as with most results in E. coli O157:H7 research, the results were somewhat inconsistent, making interpretation of the results somewhat difficult.

Recently, the Kansas State researchers have reported to the Kansas Livestock Association results of an experiment supported by the Kansas Beef Council and NCBA (Nagaraja et al., 2008). Seven hundred cattle were fed for 150 days - half were fed DG. Pen floor samples were collected weekly or every two weeks and a total of 3,560 samples were collected and analyzed. This is a large-scale study with good statistical power. Overall prevalence was fairly low (5.1%). “Although prevalence of E. coli O157:H7 in pen floor faecal samples was numerically higher on some sampling weeks in cattle fed DDGS, there was no significant effect of DDGS (P = 0.2)”.

All of the Kansas State studies were conducted with steam-flaked corn (SFC) diets, with or without 25% DG(on a DM basis). This may be important as we compare the Nebraska research to the Kansas State research. Corrigan (2007) has shown that DG do not respond the same in SFC diets compared to dry-rolled corn (DRC) or high moisture corn diets (HMC) (Figure 3). If cattle gains and efficiencies respond differently to DG levels in SFC and DRC or HMC diets, it is possible that any effect on E. coli O157:H7 might respond differently as well. The Nebraska E. coli O157:H7 research is with DRC or HMC only.

It is logical that the diet fed to cattle could influence the growth of E. coli O157:H7 in the hindgut. Research has shown that the primary reservoir of E. coli O157:H7 is the hindgut and that the E. coli O157:H7 attach to the intestinal wall of the hindgut. Interestingly, the E. coli O157:H7 have no effect on cattle performance. There are two opposing theories on how the diet affects E. coli O157:H7 in the hindgut. The first theory is that starch escaping digestion in the rumen and small intestine is fermented in the hindgut producing volatile fatty acids and lowering pH. This is theorized to inhibit growth of the E. coli O157:H7. Fox et al. (2007) showed support for this theory. Steam-flaking reduced starch in the hindgut and increased E. coli O157:H7 shedding. However, Depenbusch et al. (2008) said “E. coli O157:H7 was not related to faecal pH or starch”. We re-analyzed the data of Peterson et al. (2007a) where diets with decreasing amounts of corn were fed - decreasing the amount of starch in the diet. Amount of starch in the diet was not related to E. coli O157:H7 shedding (P = .22).

The opposing theory is that starch in the hindgut is the substrate for E. coli O157:H7 so by reducing the amount of starch getting to the hindgut, E. coli O157:H7 would be reduced. Peterson et al. (2007a) and Folmer et al. (2003) showed this did not work. While it is logical that diet affects E. coli O157:H7 growth in the hindgut, clearly neither of the two opposing “starch theories” have been “proven” so we look for other possible theories.

The Nebraska research on the effect of DG on E. coli O157:H7 shedding was reported by Peterson et al. (2007b). The research was primarily focused on vaccination as an E. coli O157:H7 intervention. Because the study was superimposed on a nutrition study, we re-analyzed the data (Figure 4). Wet DG were fed as 0, 10, 20, 30, 40 and 50% of diet dry matter replacing DRC and HMC. In this experiment, samples of the hindgut mucosa were analyzed as well as faecal samples. Results were similar but more consistent for the mucosal samples (Figure 4). There was a significant effect of level of DG on E. coli O157:H7 shedding, however, it was not a linear relationship. None of the levels of DG feeding was statistically different from the control (0DG). The 10, 20, and 30 per cent DG levels numerically decreased the shedding of E. coli O157:H7. Interesting, this is within the range of feeding (25 per cent) the Kansas State researchers used. Our research is with DRC and HMC while their research is with SFC, which may well make a difference.



At the 40 and 50 per cent DG feeding levels, E. coli O157:H7 shedding numerically increased compared to the control. Note that the statistical difference is between the 10, 20, and 30 per cent DG levels and the 40 and 50 per cent levels. So does DG decrease or increase E. coli O157:H7 shedding? The media (Des Moines Register, January 27, 2008) chose the negative approach - “studies at the University of Nebraska suggest feeding DG increases levels of a DEADLY form of E. coli bacteria”.

Peterson et al. (2007b) were studying vaccination. The pattern of E. coli O157:H7 in hindgut mucosa for unvaccinated cattle is similar to that discussed previously (Figure 5). However, there was only one positive steer among the vaccinated cattle and that was at the 50 per cent DG feeding level. So which is more important, finger pointing at feed ingredients such as DG or looking for interventions? Over four studies involving 1,784 cattle, vaccination has reduced E. coli O157:H7 shedding by 65 per cent. This is equivalent to the effect of winter versus summer on shedding. Feeding a direct-fed microbial (Peterson et al. (2007a) reduced shedding over two years by 35 per cent. These two interventions plus others being researched have considerable merit.

Conclusions

1. It is reasonable to think that what we feed cattle might affect the bacterial population of the hindgut. This has already been demonstrated in different studies.
2. Research at Kansas State and University of Nebraska-Lincoln both suggest that under some feeding levels and some other, as yet unknown, conditions, DG may increase E. coli O157:H7 shedding.
3. Results of E. coli O157:H7 research in general and specifically with DG feeding are inconsistent. To date, there has been no demonstrably consistent effect of DG feeding on E. coli O157:H7 shedding.
4. Response in E. coli shedding to DG feeding may be affected by DG level and other dietary ingredients such as the corn type. The issue is highly complex and more research is needed to address the many factors, including those feed components given concurrently with DG, that could be involved.
5. Interventions and research on interventions is much more important than “finger pointing” at different feedstuffs, especially when data are inconsistent and more research is needed.
6. At this point, there is no scientific evidence that feeding DG, at least at levels being used commercially, is the cause of a food safety crisis! Additionally, there is no scientific evidence to suggest that the feeding of DG is the cause of the 2007 recalls.

References

Corrigan, Mark C., Galen E. Erickson, Terry J. Klopfenstein, Kyle J. Vander Pol, Matthew A. Greenquist, and Matthew K. Luebbe. 2007. Effect of corn processing and wet distillers grains inclusion level in finishing diets. Nebraska Beef Cattle Report. MP-90:33-35.
Depenbusch, B. E., T. G. Nagaraja, J. M. Sargeant. J. S. Drouillard, E. R. Loe, and M. E. Corrigan. 2008. Influence of processed grains on fecal pH, starch concentration and shedding of E. coli O157:H7 in feedlot cattle. J. Anim. Sci. 86:632-639.
Elder, R. O., Keen, J. E., Siragusa, G. R., Barkocy-Gallagher, G. A. Koohmaraie, M., and Laegreid, W. W. (2000). Correlation of enterohemorrhagic Escherichia coli O157 prevalence in feces, hides, and carcasses of beef cattle during processing. Proceedings of the National Academy of Sciences of the United States of America, 97-2999-3003.
Folmer, Jeffrey, Casey Macken, Rod Moxley, David Smith, Mindy Brashears, Suzanne Hinkley, Galen Erickson, and Terry Klopfenstein. 2003. Intervention strategies for reduction of E. coli O157:H7 in feedlot steers. Nebraska Beef Report. MP 80-A:22-23.
Fox, J. T., B. E. Depenbusch, J. S. Drouillard, and T. G. Nagaraja. 2007. Dry-rolled or steam-flaked grain-based diets and fecal shedding of E. coli O157:H7 in feedlot cattle. J. Anim. Sci. 85:1207-1212.
Jacob, M. E., J. T. Fox, S. K. Narayanan, J. S. Drouillard, D. G. Renter, and T. G. Nagaraja. 2008a. Effects of feeding wet corn distiller’s grains with solubles with or without monensin and tylosin on the prevalence and antimicrobial susceptibilities of fecal food-borne pathogenic and commensal bacteria in feedlot cattle. J. Anim. Sci. 2008. In press Eprint available. doi:10.2527.
Jacob, M. E., J. T. Fox, J. S. Drouillard, D. G. Renter, and T. G. Nagaraja. 2008b. Effects of dried distillers’ grains on fecal prevalence and growth of Escherichia coli O157 in batch culture fermentations from cattle. Appl. Environ. Microbial. 74:38-43.
Jacob, M. E., G. L. Parsons, M. K. Shelor, J. T. Fox, J. S. Drouillard, D. U. Thomson, D. G. Renter, and T. G. Nagaraja. 2008c. Feeding supplemental dried distiller’s grains increases fecal shedding Escherichia coli O157 in experimentally inoculated calves. Zoonoses Publ. Hlth (Accepted for Publication).
Khaitsa, M. L., D. R. Smith, J. A. Stoner, A. M. Parkhurst, S. Hinkley, T. J. Klopfenstein, and R. A. Moxley. 2003. Incidence, duration and prevalence of Escherichia coli O157:H7 fecal shedding by feedlot cattle during the finishing period. J. of Food Protection, Vol. 66, No. 11, 1972-1977.
Koohmaraie, M., T. M. Arthur, J. M. Bosilevac, M. Guerini, S. D. Shackelford, and T. L. Wheeler. 2005. Post-harvest interventions to reduce/eliminate pathogens in beef. Meat Science 71:79-91.
Nagaraja, T. G., J. Drouillard, D. Renter, S. Narayanan. 2008. Distillers grains and food-born pathogens in cattle: Interaction and intervention. KLA News and market Report Vol. 33, No. 35.
Peterson, R. E., T. J. Klopfenstein, G. E. Erickson, J. Folmer, S. Hinkley, R. A. Moxley, and D. R. Smith. 2007a. Effect of Lactobacillus Strain NP51 on Escherichia coli O157:H7 fecal shedding and finishing performance in beef feedlot cattle. J. Food Prot. 70, No. 2:287-291.
Peterson, R. E., T. J. Klopfenstein, R. A. Moxley, G. E. Erickson. S. Hinkley, G. Bretschneider, E. M. Berberov, D. Rogan, and D. R. Smith. 2007b. Effect of a vaccine product containing type III secreted proteins on the probability of E. coli O157:H7 fecal shedding and mucosal colonization in feedlot cattle. J. Food Protection. Vol. 70, No. 11:2568-2577.
Smith, D. R., M. P. Blackford, S. M. Younts, R. A. Moxley, J. T. Gray, L. L. Hungerford, C. T. Milton, and T. J. Klopfenstein. 2001. Ecological relationships between the prevalence of cattle shedding Escherichia coli O157:H7 and characteristics of the cattle or conditions of the feedlot pen. J. Food Prot. 64 (12) 1899-1903.

May 2008