Salmonella and the Raw Diet
Lesley New – BSc Nutrition and Food Science
Critical look at the literature

The biggest barrier encountered when proposing feeding a raw diet are those concerning microbial risks. Individuals opposed to this feeding method state that there is evidence lacking to prove that dogs can deal with the resultant risk of the bacteria that may be associated with eating raw meat. A critical look at the literature is therefore necessary to evaluate if this concern is valid.

One popular study that most recently tackled this issue is a paper by Joffe assessing the risk of salmonella infection in dogs fed a BARF (Bones And Raw Food) diet10. He proposes that dogs fed a raw meat diet are a “public health concern”. Looking objectively at his methods and results fail to prove this to be the case. His study used 10 client owned raw fed dogs and 10 client owned dogs fed a dry food diet as controls. One meal-sized food sample and one stool sample were collected from each dog and tested for salmonella. The owners were responsible for the collection of the samples and as is stated in the paper “were aware of the purpose of the study prior to collection”10. This alone is questionable and can create bias. Controlled scientific studies should be set up so that they are blind to the participants in order to prevent their influence on the results. Nowhere in the paper does it mention control measures surrounding feeding (including, but not specific to sanitation), storage of food, collection procedures or sample handling (temperature regulation, time between collection and testing). There is also no information given with regards to the subjects (age, breed, and/or clinical history). Were any of these dogs previously treated for gastroenteritis? Are they predisposed to developing such conditions? Were the dogs tested for the presence of salmonella in their stool prior to the onset of the study? Were any other environmental exposures considered as the source of the salmonella (especially in the cultures that did not directly reflect the bacteria seen in the food)?

The results themselves seem suspect and conclusions that are drawn from them do not necessarily fit with the break down of the facts. In fact Joffe helps to prove that dogs can deal with salmonella in their system. He states that 80% of the raw food samples tested positive for salmonella10. Of those eight dogs two of them showed salmonella in their stool, only one of which showed the same serovar that was present in the food10. How can he then dismiss the suggestion that dogs fed infected food do not necessarily shed salmonella in their stool? Interestingly none of the dogs were reported to show clinical symptoms of Salmonellosis even though they were fed food that contained the bacterium. This is further supported by Ettinger and Feldman who state that “isolation from gastrointestinal tract or its secretions does not indicate that the organisms are causing clinical disease” 7.

Curiously none of the stool samples of the dog’s fed a dry diet showed positive results for salmonella. Given that 36% of healthy dogs carry salmonella in their digestive tract and may shed the bacterium7, 9, 18 it seems suspect that the results fail to show even a small percentage of shedding in these dogs. Joffe also states that a dog’s mouth must be considered a source of infection. Curious, we had oral and anal swab tests conducted by an independent lab on some of our dogs that are fed exclusively raw diets. These dogs were purposely fed meat that had been left out at room temperature for several days before feeding. Results were negative for salmonella in all of the dogs, contraindicating his concern. Interestingly enough Joffe states that although his results, as produced, are “suggestive, they are not statistically significant owing to the small number of dogs studied”10. This remark clearly opens the door to further study and warrants caution to those taking his conclusions as the final word.

Another important paper to consider in this argument is that by Day, James and Heather6. Their purpose was to look at the response of dogs to salmonella presented in their food, either existing naturally or added at predetermined dosages. All other aspects were controlled for to limit the possibility of outside sources of contamination, including disinfecting the cages and all food and water bowls daily. The first series of experiments involved feeding dry dog food that was determined to be naturally contaminated for a duration of 16 days and then again with another set of dogs for 50 days. Anal swabs were collected daily and examined for enteric bacteria. Of the dogs fed for 16 days no positive samples for salmonella were produced. With the longer feeding duration, however, a shedder state was produced but no evidence of clinical symptoms was reported 6. These results show that although contaminated food may produce a shedder state in some dogs, no signs of infection were actualized. In the second series the dogs were fed, along with their food, a large single dose of salmonella. The strain used was isolated from human patients showing clinical signs of Salmonellosis. Again none of these dogs developed clinical signs of infection and the shedding of the pathogen was very sporadic6. It is important to note that the dogs involved in the testing were 3 months of age. This is significant because younger dogs appear to be more susceptible to infection than do mature dogs15, 16, due in part to their developing immune systems. The results from this data show that dogs openly fed salmonella did not develop any clinical signs of infection, suggesting that their systems can handle continual exposure to the bacterium without displaying symptoms of illness.

A look at earlier studies done in the area may also help to give credence to the argument that a dogs system can handle exposure to enteric bacteria. The results in Kintner’s study lean towards this hypothesis. Early work done at the Ohio State University with 71 dogs showed that 18% of the dogs tested for salmonella were positive 12. Further studies were conducted in an attempt to infect several of the dogs with Salmonella typhimurium and Salmonella enteritidis. The dogs were dosed with Salmonella either through their food or via a stomach tube and monitored for signs of infection. In all of the dogs organisms were recovered in their feces but were negative three to eight days later. None of the animals showed a clinical state of infection or any evidence of ill effect12. Two of the dogs were then given S. typhimurium intravenously and subsequently developed Salmonellosis, however both recovered without treatment in 4-5 days12. These findings are important because although Salmonellosis was evident in these two dogs it was only after the dose was given directly into the bloodstream. In order for salmonella to cause a systemic infection fed via the oral route the bacterium must adapt to the acidic environment it encounters in the stomach and then resist being killed by the white blood cells of the immune system18. In this study the only findings of infection occurred after administering a dose through a route which has limited significance once taken out of the laboratory.

Incidence and Prevalence

Many critics believe that dogs fed a raw diet are a reservoir for transmitting salmonella to humans. For this reason it is important to consider the prevalence in dogs and how this relates to humans. Information from the BC Center for Disease Control for 2001 shows that 80% of the positive results for salmonella were from humans sources3. The remaining 20% were isolated from non-human sources, which could include many potential origins (contaminated food, environment, reptiles, dogs and/or cats). This limits the significance of cases that were of the direct result of contact with dogs alone. Statistics from the CDC show that in 2001 the total number of isolations of salmonella from domestic animals and environmental sources (both clinical and non-clinical presentations) was approximately 7.9%5. It is important to note that this also included environmental sources of salmonella as well as other domestic animals such as cats. Therefore the percentage of isolations from dogs alone would be a fraction of this percentage, leaving a marginal window of cases where infections could be passed to humans.

After examining canine zoonoses and their risk to man Baxter and Leck state that direct canine-human transmission has occurred in relatively few cases2. They further state that in reports from the UK dogs account for only about 1% of animal isolations of salmonella and speculate that they may be involved in a similar number of human cases2, although no definitive data to theorize such conclusions is given. Pelzer states that although salmonella has been isolated from dogs, it is difficult to assess the risk of humans acquiring infection because the infective dose is difficult to define, and in most cases the number of organisms isolated from the animals is not given17. In cases where the same serotype is isolated from both humans and animals in the home the “question arises as to whether one or the other was the culprit or whether both were victims of a common source”11. A study by Galton et al did not find significant evidence to conclude that salmonella infections in dogs were a source of infection in man8. The paper further states that infections from their study were in part “either derived from common sources, spread from animal to man or possibly the reverse”8. Mackel et al suggested that instances of infection also relate to the numerous reservoirs of Salmonella in both human and animal environments14. While dogs could be considered a potential source of infection1, 16 evidence has yet to prove that they are a definitive source of infection or a “public health concern”10. To further elicit this point a look at a study conducted by Caraway et al is warranted4. The study involved sentry dogs that were shedding salmonella in their stool with unknown origin for the point of infection. While the source of the infection was not isolated the paper did show, after testing the handlers, that even though salmonella was found to be carried by over 78% of the dogs the bacterium had not been passed to any of the handlers4. As stated by Ettinger and Feldman “most cases of human enteric disease caused by these bacteria are not associated with pet exposure”7.

Relationship to raw feeding

As can be seen in the studies examined dogs exposed to salmonella in their food did not develop clinical Salmonellosis. According to Pelzer CDC reports in the UK from 1973- 1984 state that the number of organisms in contaminated meat is low (1/100g), a rate of contamination that, for most people, is below the infective dose17. Considering that dogs are more resistant to infection than humans, proper handling and storage of raw meat products should minimize the risk. In some of the studies the bacteria could be found in the stool of some dogs after exposure, however since 36% of healthy dogs carry salmonella in their digestive tracts7, 9, 18 these findings were not surprising. Due to this fact an important issue that must be remembered when discussing salmonella risk in dogs, regardless of diet, is hygiene7,13. The source of the pathogen is generally proved to be feces15, 17, 18, therefore the risk of transmission can be minimized or eliminated by maintaining certain levels of sanitation and hygiene. Additional measures should be taken in dogs fed a raw diet to ensure sufficient levels of safety. This includes properly cleaning all feeding bowls, contact surfaces and utensils using hot soapy water and limiting the time spent above refrigeration temperature (5°C) during storage and feeding.

Sources Cited:
  1. Adler, HE, Willers, EH and Levine, M. Incidence of Salmonella in Apparently Healthy Dogs. JAVMA 1951; 118:300-304.
  2. Baxter, DN and Leck, I. The Deleterious Effects of Dogs on Human Health: 2. Canine Zoonoses. Community Medicine 1984; 6:185-197.
  3. BC Center for Disease Control – 2001 British Columbia Annual Summary of Reportable Diseases/ Laboratory Services Annual Report. March 14, 2003.
  4. Caraway, CT, Scott, AE, Roberts, NC and Hauser, GH. Salmonellosis in Sentry Dogs. JAVMA 1959; 135:599-602.
  6. CDC – Salmonella Annual Summary 2001. April 2, 2003.
  7. Day, WH, James, E and Heather, CD. Salmonellosis in the Dog. Am J Vet Res 1963; 24:156-157.
  8. Ettinger, SJ and Feldman EC (1995) Textbook of Veterinary Internal Medicine 4th Edition. Harcourt Brace. Pp. 366-368.
  9. Galton, MM, Scatterday, JE and Hardy, AV. Salmonellosis in Dogs I: Bacteriological, Epidemiological and Clinical Considerations. J Infect Dis 1952; 91:1-5.
  10. Hand, MS, Thatcher, CD, Remillard, RL and Roudebush P (2000) Small Animal Clinical Nutrition. Mark Morris Institute. Pp. 188.
  11. Joffe, DJ and Schlesinger, DP. Preliminary Assessment of the Risk of Salmonella Infection in Dogs Fed Raw Chicken Diets. Can Vet J 2002; 43:441-442.
  12. Kaufmann, AF. Pets and Salmonella Infection. JA VMA 1966; 149:1655-1661.
  13. Kintner, L. Canine Salmonellosis. Veterinary Medicine 1949; 44:396-398.
  14. LeJeune, JT and Hancock DD. Public Health Concerns Associated with Feeding Raw Meat Diets to Dogs. JAVMA 2001; 219:1222-1225.
  15. Mackel, DC, Galton, MM, Gray, H and Hardy, AV. Salmonellosis in Dogs IV: Prevalence in Normal Dogs and Their Contacts. J Infect Dis 1952; 91:15-18.
  16. Morse, EV and Duncan MA. Canine Salmonellosis: Prevalence, Epizootiology, Signs and Public Health Significance. JAVMA 1975; 167:817-820.
  17. Morse, EV, Duncan, MA, Estep, DA, Riggs, WA and Blackburn, BO. Canine Salmonellosis: A Review and Report of Dog to Child Transmission of Salmonella enteritidis. Amer. J Pub Health 1976; 66:82-84.
  18. Pelzer, KD. Salmonellosis. JA VMA 1989; 195:456-463.
  19. Sanchez, S, Hofacre, CL, Lee, MD, Maurer, JJ and Doyle, MP. Animal Sources of Salmonellosis in Humans. JAVMA 2002; 221:492-497.