Risk factors associated with clinical signs of vesicular stomatitis and seroconversion without clinical disease in Colorado horses during the 2014 outbreak.

Vesicular stomatitis (VS) is caused by a contagious rhabdovirus that affects horses, cattle, and swine. Clinical signs of vesicular stomatitis virus (VSV) infection in pigs and cattle are indistinguishable from foot-and-mouth disease (FMD), a foreign animal disease and reportable disease in the United States (Rodriguez et al., 2000). A VS epidemic occurred in the Rocky Mountain region in 2014-15. A study was conducted in Colorado to evaluate horse- and management-level factors associated with VS. For a horse to be considered a clinical VS horse, there were two requirements. First, clinical VS horses had to have clinical signs consistent with VS, including one or more of the following: vesicles, ulcers, erosions or crusting on the muzzle, nares, lips, oral or nasal mucosa, ears, ventrum, udder or penile sheath, or coronary band lesions. Second, clinical VS horses had to have laboratory confirmation of VSV exposure via virus isolation from lesions or a positive complement fixation test performed on sera. All non-clinical horses residing on VSV-affected premises enrolled in the study were evaluated for exposure (i.e., seroconversion) to VSV. Overall, management and housing data were collected from 334 horses on 48 premises in Colorado. Approximately one-third (31.4%) of enrolled horses were clinical cases and two-thirds (68.6%) were controls. Three premises-matched logistic regression models were constructed in SAS using backward elimination (P-value < 0.05) after univariate screening of a priori-selected variables (P-value < 0.20). Model outcomes included differences in characteristics and management of 1) clinical and nonclinical horses, 2) exposed and unexposed horses, and 3) exposed nonclinical and unexposed nonclinical horses. Overall, factors most strongly associated with risk of being a VS clinical horse were access to pasture (P-value = 0.002), and pregnancy status (P-value = 0.001). Factors most strongly associated with VSV exposure among horses were access to pasture (P-value = 0.003) and lack of any insect control (P-value = 0.001). The only factor associated with VSV-exposed nonclinical horses compared with unexposed VSV horses was contact with clinical horses (P-value = 0.013). There were no associations identified regarding clinical horses compared with exposed nonclinical horses. With regard to severity of lesions (severe vs. moderate or mild), no variables met the criteria for inclusion in the multivariable model. Results of this study provide evidence that pasture access and fly control are important factors associated with VSV exposure.