A prospective investigation of bovine leukemia virus infection in young dairy cattle, using survival methods

The authors studied bovine leukemia virus infection in cattle born between July 1, 1979, and June 30, 1981, and followed up to 27 months in the University of Florida Dairy Research Unit herd, a 200-cow milking herd. Cattle were tested monthly for antibodies to the virus by agar-gel immunodiffusion with the glycoprotein-51 antigen. Of 473 live calves entering the study at birth, 54 became infected during the course of the study. Rates of detection of bovine leukemia virus infection were examined using survival methods. Plotted cumulative hazard rates revealed a nonlogarithmic survivorship function and three age-specific detection phases. These phases were 0-6 months, 7-16 months, and 17-26 months of age with estimated incidence rates of 5.25, 2.14, and 14.37 new detections per 10,000 cattle-days at risk, respectively. Significant differences were found between incidence rates of the first and second phase (p = 0.0168) and between those of the second and third phase (p less than 0.0001); corresponding relative risks were 0.41 and 6.71, respectively. The reduced incidence rate of the second phase coincided with movement of calves from a crowded calf barn to pastures. The increased incidence rate of the third phase coincided with mixing of bred heifers with bovine leukemia virus-infected dry cows. Increased incidence rates did not coincide with common-needle vaccination or artificial insemination. Evidence was not found for milk-borne infection. Age-specific detection rates were not associated with dam parity (p = 0.93), dam age (p = 0.79), breed (p = 0.60), or consumption of colostrum from cows infected with the virus (p = 0.23). However, detection rates were lower in cattle born to bovine leukemia virus-infected cows than in those born to noninfected cows (p = 0.03).