A longitudinal observational study of the dynamics of Mycoplasma bovis antibodies in naturally exposed and diseased dairy cows.

Mycoplasma bovis is an important pathogen causing disease and substantial economic losses in cattle. However, knowledge of the dynamics of antibody responses in individual cows in the face of an outbreak is currently extremely limited. The use of commercial antibody tests to support clinical decision-making and for surveillance purposes is therefore challenging. Our objective was to describe the dynamics of M. bovis antibody responses in 4 Danish dairy herds experiencing an acute outbreak of M. bovis-associated disease, and to compare the antibody dynamics between dairy cows with different disease manifestations. A total of 120 cows were examined using a standardized clinical protocol and categorized into 4 disease groups: "mastitis," "systemic," "nonspecific," and "none." Paired blood and milk samples were collected and tested using a commercial M. bovis antibody-detecting ELISA. Plots of raw data and generalized additive mixed models with cow and herd as random effects were used to describe serum and milk antibody dynamics relative to the estimated time of onset of clinical disease. Cows with mastitis had high optical density measurement (ODC%) of antibodies in both milk and serum at disease onset. The estimated mean ODC% in milk was below the manufacturer's cut-off for the other groups for the entire study period. The estimated mean serum ODC% in the "systemic" group was high at onset of disease and stayed above the cut-off until 65 d after disease onset. However, the lower 95% confidence interval (CI) for the mean ODC% was only above the manufacturer's cut-off between 7 and 17 d after onset of disease. The CI of the "systemic" and "none" groups did not overlap at any time between the day of disease onset and 65 d after disease onset, and the estimated mean ODC% for both the "nonspecific" and "none" groups were generally below the cut-off for the majority of the study period. In conclusion, the serum antibody responses were highly dynamic and showed a high level of variation between individual cows. This strongly suggests that serology is unlikely to be useful for individual diagnosis of M. bovis-associated disease in dairy cows. However, it might still be useful for herd- or group-level diagnosis. Antibodies in milk were only increased in cows with M. bovis mastitis, indicating that milk antibody measurements only have diagnostic utility for cows with mastitis.