A study of leptospirosis in South African horses and associated risk factors

Abstract

Most leptospiral infections in horses are asymptomatic; however, acute disease manifestations as well as reproductive failure and recurrent uveitis have been reported. In South Africa, the epidemiology of the disease in horses is not well documented. A serosurvey to determine what serovars were present in horses from Gauteng, KwaZulu-Natal and Western Cape Provinces and to get an estimate of the seroprevalence of leptospirosis was carried out from January 2013 until April 2014 with the assistance of four large equine hospitals: Baker & McVeigh Equine Hospital (2 branches), the Equine Clinic of the Onderstepoort Veterinary Academic Hospital, Faculty of Veterinary Science, University of Pretoria and the Drakenstein Veterinary Clinic. Furthermore, associations between potential risk factors and both seropositive horses to the predominant serovar Bratislava and to Leptospira spp were statistically evaluated using univariable analysis and multivariable logistic regression models. Adult horses admitted to each equine hospital and horses seen on ambulatory service for various purposes were included in the study. Severely-ill, chronically infected, and horses with underlying immunosuppressive conditions were excluded from sampling, to control for potential bias. A total of 663 horse sera were collected and tested against a panel of 24 leptospiral serovars using the microscopic agglutination test. The most predominant serovars in Gauteng were Bratislava [32%, 95% CI: 29-35%]; Djasiman [10.4%, 95% CI: 8-12%] and Arborea [8.9%, 95% CI: 7-11%], in the Western Cape Province, Bratislava [27.35%, 95% CI: 23-32]; Djasiman [15.4%, 95% CI: 12-19%] and Arborea [14.5%, 95% CI: 11-18%] and in KwaZulu-Natal, Bratislava [39.4%, 95% CI: 34-44%]; Arborea [9.6%, 95% CI: 7-13%]; and Tarassovi [7.7%, 95% CI: 5-10%] respectively. Twenty one serovars representing 17 serogroups were detected in South African horses with serovar Bratislava being the most serodominant. The apparent prevalence to one or more serovars of Leptospira spp at a serum dilution of 1:100 in Gauteng, KwaZulu-Natal and Western Cape Provinces was 49%; 37% and 32% respectively. The true prevalence was calculated for each province taking into account the clustering effect during the sampling and was found to be between 24-74% in Gauteng; 26-39% in the Western Cape and 20-54% in KwaZulu-Natal. Nooitgedacht (South African horse breed) horses were found to be at greater risk of being seropositive to both serovar Bratislava (OR=5.08) and Leptospira spp (OR=6.3). Similarly, horses residing on properties with forestry in the vicinity were found to be at greater risk of being seropositive to both serovar Bratislava (OR=9.3) and Leptospira spp (OR=5.2). This study has shown that a high proportion of horses in South Africa are exposed to a wide range of serovars, inferring a complex epidemiology. It also describes for the first time new serovars of Leptospira in South African horses that have not previously been reported.