A large multinational study shows that pet owners don’t think feeding raw food increases the risk of infection to household members. Raw food includes all uncooked meat and other parts of a dead animal. In fact, an infection in humans could be traced to raw pet food in just 0.24% of all surveyed households.
The percentage of pet owners who feed their pets with raw food varies from about 16% in the US and Australia, to over 60% in the Netherlands. Many consider raw food to be healthier for the pet, but others say it is less digestible, could predispose to kidney stones, and could cause leaky gut. Others, of course, fear it could lead to the transmission of food poisoning to humans.
Many people have discussed this issue, however without solid proof in the form of outbreaks of human infection spread from the raw pet food. To answer this question, a team of vets asked pet owners who gave their cats and dogs raw animal food what they thought about the matter.
Raw food denotes any meat, internal organs, bones and cartilage fed to pets uncooked. Image Credit: Johanna Anturaniemi
How was the study done?
The researchers used an Internet survey to collect information from almost 16 500 households with pets in 81 countries, about the handling of pet food, how much of the pet’s diet was made up by raw foods, and which animal it originated from, as well as the type of food borne infection it was associated with in their household, if any. The survey was available online in five languages for almost 230 days.
The results were classified into households where it was confirmed that a pathogen had spread from the pet food to a human; those where such spread was suspected; and those where no transmission occurred.
What did the study show?
Almost 74% of the respondent households named a dog as the pet fed with raw food, while in almost 20% both cat and dog were fed this way.
Only 0.24%, or 39 households in all, said that someone in the household had been infected by pet food, by a specific named pathogen which was found on laboratory analysis of the human (but not always the pet food) sample. In many of these households the pets often had <20% of their diet made up of raw meat.
The most frequently reported known pathogen causing the human infection was Campylobacteria, and Salmonella, then Escherichia coli, Clostridium, Toxoplasma, and in one case, Yersinia. However, the same pathogen had been found in the pet food in only three cases, since in none of the other households was the suspected pet food subjected to analysis.
An additional 24 households also said they had experienced human infection which they suspected to be spread from the pet food but were not able to say which infective organism was involved. These households more often allowed pets to eat meat from dead animals other than the given foods, and also fed spoiled human food to their pets, which could cause some of these infections rather than the raw pet food.
In toto, 99.9% of these households had no history of human infection spreading from the raw pet food, which they had been feeding their pets from a few weeks up to 65 years. The average period was 5.5 years. The average age of the person who became ill, reportedly from the raw pet food, was about 40 years. In the first group of known bacterial infections, four included children of 2-6 years of age, which is 10% of the infected number though 25% of households had children of this age. This might be because children are shielded from potential infection.
In two cases, of which one had cancer and the other Crohn’s disease, weakened immunity was suspected to underlie the infection, but 15% of the households had such individuals.
Earlier studies show that dogs fed on raw meat shed Salmonella and E. coli in feces and in a small percentage of meat diet samples, but another study showed that the presence of Salmonella in the raw meat was poorly correlated with fecal shedding of this organism. Poultry and beef were fed to 80% and 90% of the 39 confirmed infection households but there was no association with increased infection rates. Raw fruits and vegetables, as well as dry pet foods, are also potential sources of these infections.
Yersinia is found in raw pork, which was fed to around 50% of the household pets, but was associated with only one infection. Similar is the case with Toxoplasma, found in pork and small ruminants, or in cat feces and contaminated water, but found in only two cases, one of which did not use pork while the other beef to feed the pets.
What do we learn?
The age group of 2-6 years was the only positive risk factor for infection (not necessarily from pets) found in this study. This encourages the question of whether these infections truly came from pet food or from outside sources like daycare, or other public areas. Immunocompromised situations did not appear to predispose to such infections.
The survey also did not support any increased risk from preparing raw animal food using the same dishes and workspace as food for the family. Nor was it more dangerous to feed the pet with more than 50% raw food, or to use salmon or turkey. Good hygiene is important in handling raw pet food as well as any other raw meat product, to prevent foodborne infections.
Researcher Johanna Anturaniemi says, “It was surprising to find that statistical analyses identified fewer infections in the households with more than 50% of the pet diet consisting of raw food. Furthermore, feeding pets raw salmon or turkey was associated with a smaller number of infections.”
The findings are in sharp contrast to the repeated outbreaks of infection related to dry pet food and treats. The Dogrisk group of researchers is now planning to follow up with a study comparing the infections that are spread from pet food in situations where pets are fed with both raw and dry food.
The research is published in the BMJ journal Veterinary Record.
Owners’ perception of acquiring infections through raw pet food: a comprehensive internet-based survey. Johanna Anturaniemi, Stella Maria Barrouin-Melo, Sara Zaldivar-López, Hanna Sinkko and Anna Hielm-Björkman. http://dx.doi.org/10.1136/vr.105122.