Seroprevalence and risk factors for infection with West Nile virus in Saskatchewan horses, 2003

Abstract

The primary objectives of this study were to determine the seroprevalence of West Nile virus (WNV) infection of horses in Saskatchewan in 2003 and to identify risk factors for the infection. Blood samples were collected in August and October from 212 horses in 20 herds in 5 geographic zones. After accounting for within-herd clustering, the proportion of horses that had been infected with WNV, as determined by IgG and IgM antibody response, was 55.7% (95% confidence interval, 44.9% to 65.8%). The proportion of antibody-positive horses differed among herds (0% to 100%) and across ecoregions (20% to 76%). Horses in southern ecoregions were more likely to have either IgM antibodies or IgG concentrations suggesting infection than were horses in northern ecoregions. The use of mosquito-control measures was associated with decreased risk. After accounting for ecoregion, there was no difference between recipients of an inactivated WNV vaccine and nonrecipients in the occurrence of antibodies reflecting natural infection.

Figure 1

Locations of the 16 vaccinated herds (black circles) and the 4 nonvaccinated herds (white squares). The proportion of infected horses among those sampled is indicated along with the 95% confidence interval.

Figure 2

Sample-to-positive (S/P) ratios of the optical-density results in the IgG enzyme-linked immunosorbent assay for previously exposed (A) and unexposed horses (B) given vaccine or saline injections (*) and having blood samples taken on all 4 dates. The solid lines (—) denote the median for each group. In the saline group, 1 horse was classified as having a “suspicious” result with the 1st 2 samples (S/P ratios 0.32 and 0.26) but was considered unexposed for the analysis. The broken line (- - -) shows the calculated cut-off point.