Herd-level risk factors for infection with bovine leukemia virus in Canadian dairy herds

- Seventy-eight percent (213/272) of the herds were identified as BLV-positive.
- Herds with a history of clinical leukosis had larger proportion of infected animals.
- Herds that purchased animals with unknown infection status had higher seroprevalence.
- Herds from eastern Canadian provinces had lower odds of infection with BLV.

Enzootic bovine leukosis (EBL) is an economically important infection of dairy cattle worldwide, which is caused by bovine leukemia virus (BLV). The prevalence of infection in Canadian dairy herds is high and continues to increase; however, there has not been a national program to control BLV. This cross-sectional study was conducted to identify potentially important risk factors for BLV infection on Canadian dairy herds, which is a prerequisite to developing an effective control program.During 1998-2003, based on a stratified two-stage random sampling process, 315 dairy farms from seven provinces of Canada were selected. Within each farm, 9-45 cows were bled and tested with a commercial serum ELISA kit for BLV antibodies. A comprehensive questionnaire, targeting potentially important herd-level management indicators, was successfully administered in 272 herds. A zero-inflated negative binomial (ZINB) regression model was fit to the resulting data to assess the potential associations between BLV seropositivity and a variety of herd-level factors.Seventy-eight percent of the herds were identified as BLV-positive (had one or more test positive animals). In the negative-binomial part of the final ZINB model, herds with clinical cases of leukosis during the 12 months prior to sampling, as well as herds which purchased animals with unknown BLV infection status in the last five years, had a significantly larger proportion of BLV positive animals. Based on a significant interaction between two of the risk factors, changing gloves between cows during pregnancy examination was not statistically associated with lower proportion of infected cows compared with not changing gloves, in the western Canadian provinces. In the logistic part of the model, herds from eastern Canadian provinces and those not purchasing cows in the last five years had increased odds of being free from BLV. The high prevalence of infection across Canada should be addressed through the development and implementation of a nationwide control program which will address the regional and herd-level risk factors for BLV infection identified in this study.