Factors associated with seroprevalence of bovine anaplasmosis in Texas

Bovine anaplasmosis (BA), caused by Anaplasma marginale, is an economically important tick-borne disease of cattle in the United States (U.S.) and worldwide. Anecdotally, Veterinary Feed Directive prescriptions in the southeastern U.S. are written mostly for treatment/prevention of BA. However, there are no recent temporal seroprevalence estimates of BA in Texas (TX). Thus, this study was aimed at determining the seroprevalence of and factors associated with BA in TX. Data were obtained from an active slaughter survey (n = 215) performed between August and December 2014 as well as from reviewing Texas A&M Veterinary Medical Diagnostic Laboratories (TVMDLs) records of specimens submitted for BA testing from January 2002 to June 2012 (n = 15,460). Irrespective of the assay used, the overall apparent seroprevalence of BA in TX between 2002 and 2012 was 15.91% (95% CI: 15.34 - 16.50%) and the yearly increase in seroprevalence followed a significant trend (P < .0001). With cELISA, the apparent seroprevalence of BA was 13.49% (95% CI: 9.56 - 18.7%) and 13.02% (95% CI: 9.74 - 17.18%) for the slaughter survey and TVMDLs records between October and December 2011, respectively. Whereas the estimated true seroprevalence for the same period was 12.35% (95% CI: 8.04 - 18.05%) and 12.78% (95% CI: 9.19 - 17.30%), respectively. Factors associated with positive BA results were age, breed, diagnostic assay used, year and quarter of the year the specimens were submitted. The odds of the outcome were 1.5 times as high when cattle were adults (vs juvenile). In comparison to other breeds, the odds of the outcome were 11.57, 7.16, and 2.5 times higher in Hereford, Angus, and mixed breeds, respectively. When compared to 2003, the odds of the diagnosis of BA was approximately 2 times as high in 2010 but 3 times as high in 2002, 2005, and 2011 and approximately 4 times as high in 2006 and 2007. In comparison to the duration from October to December, the odds of the outcome were approximately 1.5 as high from January to March and from July to September durations. Counties with specimen submissions for BA testing had a significantly greater cattle population (p = .0061) and number of cattle farms (p < .001) than counties without specimen submissions. Subsequent prevention and control measures for BA should target these factors and should prioritize on counties with higher cattle population in the eastern part of TX. Furthermore, TVMDLs records appear sufficient for the surveillance of BA in TX.