Frontiers in Veterinary Science (Feb 2020)
Leptospira Seroprevalence Among Ugandan Slaughter Cattle: Comparison of Sero-Status With Renal Leptospira Infection
Abstract
Despite evidence of both human and animal Leptospira exposures in Uganda, the epidemiology of the disease is still not well-investigated. Contact with animals and their environments have been pointed out as potential source of infection with Leptospira species in humans; and cattle may be an important reservoir in Uganda. In this cross-sectional study, we estimated the prevalence of anti-Leptospira antibodies by the standard microscopic agglutination test (MAT); and associated risk factors among slaughtered cattle. We also compared the performance of the MAT used in this study against a lipL32 based real time PCR (qPCR) assay previously conducted on the kidneys and urine of the same slaughter cattle as tested in this reported study. Of 500 cattle sampled, 27.8% (95% CI 23.9–32.0) tested positive (titer ≥ 100) to at least one Leptospira serovar, with the majority of seropositive cattle reacting to serovars Tarassovi (sg Tarassovi) (11.6%), Sejroe (Sg Sejroe) (7.8%), and Australis (Sg Australis) (5.2%). Older animals had 2.8 times (95% CI 1.0–8.2, p-value 0.055) greater odds of being seropositive than younger ones (<1.5 years). The sensitivity and specificity of the MAT over the qPCR were 65.9% (95% CI 50.1–79.5) and 75.9% (95% CI 71.7–79.7), respectively; with a negative predictive value of 95.8% and positive predictive value of 20.9%. In conclusion, slaughter cattle in this study were significantly exposed to pathogenic Leptospira species of mainly the Tarassovi, Sejroe, and Australis serogroups, with seroprevalence being higher among older cattle. The high specificity and negative predictive value of MAT as used in this study when compared to the qPCR assay may imply a rather strong association between seronegativity and absence of renal Leptospira infection. However, MAT predictability for renal Leptospira infection may be interpreted cautiously since predictive values of diagnostic tests are dependent on prevalence.
Keywords