Zoonotic Diseases from Horses: A Systematic Review

Literature search

In December 2016, we searched the following databases: PubMed, Web of Science: Core Collection, Zoological Record, and BIOSIS, ProQuest: Ag Science Collection, EBSCO Host: CAB, and Wiley Online Library. We used the PRISMA method (Moher et al. 2009) (Supplementary Data). Search terms included: Equine(s), horse(s), equid(s), “Equus ferus” zoonotic, zoonoses, zoonosis, “Animal to human transmission,” and “equine exposure” (see Supplementary Table S1 for full search strings). Only journal articles were used. No time-period was specified. Only English language articles were included. The gray literature was not included. Duplicate articles were removed.

Literature analysis

Titles and abstracts were reviewed to determine if articles referred to diseases in horses or humans that were zoonotic between horses and humans. For the purpose of our study, zoonoses was defined as a disease to which humans and horses were both susceptible that could be transmitted either directly or indirectly from horses to humans. Vectored diseases were included. Prevalence, surveillance, outbreaks, and case studies/series in humans and horses were collected. The articles that were not in English, review articles, and any citation for a format that was not a journal article (books, thesis, and conference proceedings/presentations) were removed.

Full-text review on the remaining articles was performed to verify that the diseases were zoonotic and involved horses in their transmission cycle. Articles were reviewed both by A.S. and F.S.O. and included by consensus. Further information was extracted from each article by A.S. for pathogen, transmission, and methodology. Articles that only dealt with molecular characteristics or policy were also removed.

Additional analyses

Further information about the identified pathogens was gathered using online information from the Centers for Disease Control (CDC), World Health Organization (WHO), and the Food and Agriculture Organization (FAO), as well as the online version of the Merck Veterinary Manual (January 2017). Information was collected on taxonomic grouping, morphology, and transmission, as well as organ systems affected available diagnostic tests. Pathogens were categorized as in categories A, B, or C according to the National Institute of Health's National Institute of Allergy and Infectious Diseases (NIAID) Emerging Infectious Diseases/Pathogens list (National Institute of Health 2016).

Transmission routes: ingestion

According to the U.S. CDC, one in six people get sick each year from eating or handling contaminated foods. Similarly, ingestion is the most common route of infection for equine zoonotic pathogens. For contaminated foods, cooking meat thoroughly and washing hands and equipment after handling meat will prevent most foodborne pathogens. Handwashing with alcohol-based products is not sufficient for every pathogen. For instance, Cryptosporidium parvum cannot easily be destroyed with alcohol handwashing treatments, and there are records of direct transmission from foals to humans (Galuppi et al. 2016).

Cooking meat properly is important for several other diseases. Trichinella spp. are a recognized complication of undercooked horsemeat, especially in Italy and France (Pozio et al. 2001). Similarly, Toxoplasma gondii has also been found in horsemeat meant for human consumption (Osiyemi et al. 1985, Paştiu et al. 2015). Both the category A equine zoonotic pathogens (B. anthracis and C. botulinum) have been reported as transmissible by ingestion of infected horse meat. According to the CDC, dogs should be kept away from raw meat and organs from livestock, including horses, and where possible, dogs should be kept away from livestock and livestock feed. For Fasciola hepatica, an aquatic snail is necessary for its transmission, as the eggs pass from livestock and humans to develop into the infective stage in the snail. However, in a study where pigs were fed with mouse liver infected with juvenile flukes, the pigs became infected, suggesting that direct infection is possible when consuming undercooked pork liver (Taira et al. 1997). Since horses are known to be able to be infected with Fasciola spp., proper cooking of liver is suggested (Haridy et al. 2002).

Meat is not the only horse product with the potential for infection. In other livestock species, Brucella spp., Mycobacterium spp., Salmonella spp., Listeria monocytogenes, Campylobacter spp., Coxiella burnetii, and E. coli have all been transmitted in unpasteurized milk (John 2006). Adding to the risk from these pathogens, several zoonotic bacteria have been found in equines with multidrug-resistant strains, including Actinobacillus equuli, E. coli, and Campylobacter spp. (Ahmed et al. 2010, Jokisalo et al. 2010, Ewers et al. 2014, Komba et al. 2014).

Transmission routes: inhalation

While fewer horse pathogens are spread by inhalation than ingestion, these include many pathogens with high mortality rates in human. Much of the recent biosecurity work involving horses with infection by inhalation has been done around the Hendra virus, which has been associated with numerous human and horse fatalities during outbreaks in Australia (Paterson et al. 1998, Westbury 2000, Hanna et al. 2006). A review of safety procedures used by veterinarians found that 58% of clients were receptive to biosecurity procedures. However, cited among issues veterinarians had with clients adopting biosecurity measures was that the clients assumed that one had to physically touch the horses to be at risk (Mendez et al. 2014). Personnel protective equipment is recommended if coming within 5 meters of an Hendra virus infected horse (Mendez et al. 2016). Another high-risk pathogen spread by inhalation is B. mallei, which has higher than 40% mortality rate in humans even with antibiotic treatment. Recent outbreaks in nonendemic areas have occurred from recently imported horses (Kettle and Wernery 2016). Equine influenza is another potentially zoonotic repository pathogen with a history of spread from recently imported horses (Xie et al. 2016).

Rabies virus has been transmitted by inhalation in the laboratory setting, and rabies inhalation is considered a low but potential risk in a necropsy setting (Winkler et al. 1973). Risks for human infection with the Australian bat lyssavirus are similar to those for rabies (Annand and Reid 2014). While rare other neurologic diseases, including West Nile virus, are at risk for transmission during necropsy (Venter and Swanepoel 2010, Venter et al. 2010).

Risks for human infection with MRS and A. equuli are increased in human and equine hospital settings (Weese and Lefebvre 2007, Jokisalo et al. 2010, Cuny and Witte 2016). Immunocompromised individuals are at special risk for Rhodococcus equi (Topino et al. 2010), which is still a risk to immunocompetent individuals (Ulivieri and Oliveri 2006). Parturition is a time of increased risk for respiratory transmission, including from A. equuli and C. burnetii (Tozer et al. 2014, 2016).

Transmission routes: wounds, bites, or cutaneous exposures

Cutaneous exposure is the third most common route of exposure for potential equine zoonoses and can occur either with or without penetration from bites or other wounds. A comprehensive review of horse bite injuries to humans found that five genera of bacteria were spread by equine bites (Langley and Morris 2009). Rabies virus is commonly associated with bites, and horses can present with misleading symptoms (Merini et al. 2010).

Wounds from knives, scalpels, or needles while handling dead horses, during slaughter or necropsy, are other sources of infection. A case of A. equuli infection occurred in a butcher who had cut his hand while at work (Ashhurst-Smith et al. 1998). Contamination of wounds can occur when the zoonotic pathogen has access to open wounds. Also, infected wounds in horses can serve as a source of infection for humans. MRSA is a common wound contaminant (Köck et al. 2014). In one case, multiple veterinarians and students without open wounds caught MRSA from a foal with a contaminated wound (Schwaber et al. 2013). In another case, a fatal infection of S. equi subspecies zooepidemicus was diagnosed in a man caring for a horse with an open wound (Kawakami et al. 2016). In general, skin conditions in horses are visible and affected horses should not be directly handled. Dermatophytosis is found worldwide in horses, including performance and working horses (ElAshmawy and Ali 2016, Maurice et al. 2016). The CDC suggests washing hands often when touching farm animals regardless of the presence of visible lesions.

Transmission routes: vector-borne

Of the approximate one quarter of potential equine zoonotic diseases that involved vectors for transmission, mosquitoes are the most common vector, followed by ticks and flies. Bartonella spp. are spread by flies, lice, or fleas. All but one of these pathogens, C. burnetii, require a vector for transmission.

All seven of the equine zoonotic diseases transmitted by mosquitoes are viruses (Table 1). All but one, Sindbis virus, of the mosquito-borne viruses cause encephalitis. For eastern equine encephalitis virus, Ilheus virus, Japanese encephalitis virus, Sindbis virus, St. Louis encephalitis virus, Western equine encephalitis virus, and West Nile virus, horses are presumed to be a dead-end host and are not part of the pathogens' active transmission cycles. Horses and humans do not reach a high enough viremia to infect mosquitoes (Lundström et al. 2010, Díaz et al. 2012, Pauvolid-Corrêa et al. 2013, Go et al. 2014). Horses do serve as sentinels for human infections as both are often infected from common avian reservoirs (Ludu et al. 2014, Saegerman et al. 2016). Horses with Venezuelan equine encephalitis virus can present with a viremia and thus serve as a source of infection (Estrada-Franco et al. 2004). Mosquito control and the prevention of mosquito bites can reduce horse and human infections from these pathogens.

Equine zoonotic diseases from ticks have been found in humans, ticks, and horses (Otomura et al. 2010, Toledo et al. 2011, Butler et al. 2016a, 2016b). Prevention requires tick control measures, as there is no direct transmission for these diseases except for C. burnetii. All five equine zoonotic pathogens carried by ticks are gram negative bacteria (Table 1). Flies and other insects can also transmit diseases from horses to humans. Flies spread Leishmania and Onchocerca spp. (Beaver et al. 1974, 1989, Ali-Khan 1977, Burr et al. 1998, Filho et al. 2008). There are species of Onchocerca that persistently cause disease in humans, but Onchocerca cervicalis from horses is considered a rare occurrence (Beaver et al. 1974, 1989, Ali-Khan 1977, Burr et al. 1998). In both cases, there was a possibility that the cases were caused by Onchocerca gutterosa, which is zoonotic from cattle (Ali-Khan 1977, Beaver et al. 1989). The pathogen, Bartonella spp. is spread by flies, lice, and fleas. There is some discussion of ticks being involved in transmission as well, but evidence is considered inconclusive (Telford and Wormser 2010).

Study strengths and limitations

The study is unique in that we used PRISMA methodology with a goal of including all potential zoonotic pathogens associated with horses. We identified and characterized a relatively large number of such pathogens. However, more than a third of the pathogens identified in this review were only mentioned in one article. In addition, most reports supporting direct transmission of equine pathogens to humans were case reports with insufficient data to determine transmission prevalence or risk. Some might argue that the sparse zoonotic disease data are at least partially due to a lack of awareness of equine zoonoses among health care providers. Others might argue that some of these reports were describing aberrant transmission and the case reports may not represent true zoonotic risks. Despite these uncertainties, it seems clear that there exist major gaps in the medical literature regarding the epidemiology of equine zoonotic diseases. We also did not find extensive literature supporting a number of widely accepted equine zoonoses. For instance, in this review, West Nile virus was found in 14 reports, eastern equine encephalitis in four reports, Venezuelan equine encephalitis found in one report, but western equine encephalitis virus was found only in review articles. Since review articles were outside the scope of the review, western equine encephalitis was not included in the 56 zoonotic pathogens. Future reviews might include other equids, such as donkeys, mules, and zebras, to improve our understanding.