Various studies have shown that cows with clinical and subclinical Johne's disease (JD) are likely to produce between 4 and 24% less milk than test negative cows. The objective of this study was to quantify the difference between the milk production of JD milk ELISA positive cows and test negative cows, as well as to evaluate variations in the test results due to breed.

The data included the test day information from 47,418 cows from 817 CanWest DHI herds in Ontario, Manitoba, Saskatchewan, Alberta and British Columbia. The cows were tested between March 2005 and April 2007 with the AntelBio Johne's Milk ELISA. The optical density (OD) cut points for a suspect or positive test result were set at > 0.065 and > 0.1, respectively. A mixed model was fitted to investigate the association of a positive milk ELISA test with the estimated cumulative 305 day milk production. Only factors with a p-value < 0.01 were retained in the model. A second model was fitted to examine the relationship between breed and observed test result. The SAS Glimmix procedure was used, including only herds with at least one positive cow. Herd was included in both models as a random effect.

The number of tested cows per herd varied between 1 and 342 (Interquartile range: 35 -71). The breeds were Holstein (n= 44,420), Jersey (n= 2,078), Brown Swiss (n= 361), Ayreshire (n= 308), Guernsey (n=141), and Milking Shorthorn (n= 110). The majority of the cows (n= 46,515; 98.1%) tested negative, while 753 tested positive (1.59%) and only 150 cows (0.32%) had suspect test results. Therefore, positive and suspect cows were combined into one JD positive group. The percent of positive cows by province varied from 0.74% to 2.27% (p < 0.0001).

The average 305 day milk production was 9,696 kg/ cow. After adjusting for breed, lactation number, season of calving, somatic cell count, days in milk on test day as well as several interactions among these factors, cows that tested positive for JD produced 748 kg or 7.7% less milk over the estimated 305 day lactation compared to milk ELISA test negative cows (p < 0.001).

The between breed comparison was adjusted for the lactation number, province, days in milk on test day and somatic cell count. Guernsey and Jersey cows were 3.9 and 2.3 times as likely to test positive for JD as Holstein cows (p ≤ 0.0001).

These findings indicate differences in JD prevalence across the 5 western most provinces in Canada. Furthermore, cows testing suspect or positive with the milk ELISA test produced significantly less milk than their negative herd mates and cows of the Channel breeds were more likely to test positive than Holsteins. Further research is needed to examine the differences in dairy herd management practices among the provinces that might explain the variations in the JD prevalence, and the cause for the variation in the test results among the different dairy breeds.