During the first 30 days of lactation, cows are at high risk for both infectious and metabolic disease issues such as hypocalcemia, ketosis, retained placenta (RP), metritis and abomasum displacement (DA).

Michael W. Overton

These and other diseases that occur during the early postparturient period are detrimental because they decrease milk production, increase treatment costs, negatively impact reproductive performance and increase mortality and culling risk. The two most costly of these issues that also result in a high level of antimicrobial use on dairies are mastitis and metritis.

Clinical mastitis, more specifically mastitis occurring in the first 30 days in milk (DIM), results in a number of negative outcomes including pain and inflammation, reduced milk production (5 to 8 percent total lactation loss), decreased reproductive performance (four- to eight-day delay to first service, 23 percent reduction in first A.I. conception rate, 15 percent reduction in percent pregnant by 320 DIM and 21 to 25 more days open), approximately double the risk of premature culling and a small increase in risk of death.

Due to a combination of a high incidence and the antimicrobial components of commonly used treatment protocols, this disease likely accounts for more cows receiving antimicrobial exposure than any other disease condition.

Based upon recent data, regional incidence for mastitis in early lactation varies by geographical region and season of calving with ranges from 5.5 to 26.1 percent based on on-farm records.

Metritis is also a very common and complex disease condition in postparturient dairy cattle characterized by inflammation of the uterine tissues, and based upon previously mentioned farm-level studies, region average whole-herd incidence varies from 18.5 to 27.6 percent.

In general, first-lactation cows appear to be having a higher incidence, but mature cows tend to have a higher percent of cases that are severe in nature.

Severity can range from very mild to severe, life-threatening forms. It is most often diagnosed during the first 14 DIM, but it can negatively affect milk production, reproduction and culling risk well beyond early lactation.

Numerous studies have demonstrated both direct and indirect negative impacts of uterine disease on overall dairy herd performance and profitability. California researchers found that cows with metritis averaged 4.9 pounds per day less milk over the first 120 days of lactation compared to normal herdmates. Other work has shown that the level of milk loss varied by DIM:

• Cows with metritis that were culled during the first 30 DIM produced 15.1 pounds less milk per day and had a median days-to-exit of 10.
  
  
• Cows with metritis that were culled during 31 to 60 DIM produced an average of 9.1 pounds less milk per day and had a median days-to-exit of 42.
  
  
• Cows with metritis that survived past 60 DIM experienced an average of 6.2 pounds of milk lost per day over the first 110 DIM and then no difference from herdmates.

Metritis has been shown to have varying influence on culling, with some studies showing no effect while others show cows that experience metritis are approximately twice as likely to be culled.

Metritis also has significant negative impacts on reproduction. It has been associated with seven more days to first service, a 20 percent lower first-service conception rate, 13 to 19 more days open and a significant reduction in pregnancy rate.

The goal of this project was to evaluate commercial herd data from across the U.S. to evaluate the measurable impact of on-farm recorded mastitis and metritis on milk production, reproduction and culling.

Herds enrolled in Elanco’s Dairy Data Access System that recorded milk production had at least 1 percent recorded mastitis incidence in the first 30 DIM, and at least 3 percent recorded metritis incidence were eligible for inclusion.

Records were extracted from the DairyComp 305 program and included Holstein cows that calved anytime from Aug. 1, 2013, to Aug. 1, 2014, and were followed for 12 months or until culled. A total of 55,643 cows from 20 herds were part of the final dataset.

Descriptive statistics were performed to demonstrate that the herds were representative of commercial herds and included frequency histograms for parity (lactation = 1, 38 percent; lactation = 2, 29 percent; and lactation greater than 2, 34 percent), calving dates, season of calving, age at first calving (mean of 23.3 months) and previous days dry (mean of 54.7 days).

Herd averages for 305me milk (29,560 pounds) were also calculated. Metritis incidence ranged from 4.2 to 29.2 percent with a mean of 16 percent. Mastitis within the first 30 DIM ranged from 1.5 percent to 15.2 percent with a mean of 5 percent.

First, multivariate nominal logistic fit models were created to evaluate risk factors for metritis and mastitis. In each case, separate models were created for primiparous versus multiparous cows.

The primiparous model for metritis (log odds of yes/no) offered the following potential explanatory variables for consideration: herd, season fresh, calf outcome (M, F, or Twin), RP (yes/no) and calving ease score (1 to 5). Due to inconsistencies in recording across herds, calving ease was dropped.

Having a male calf or twin resulted in higher odds of metritis, with twin calvers much higher than those delivering males. Age at calving was not significantly associated with risk of metritis. Heifers calving in the spring had higher odds of metritis than those calving in fall or winter.

A similar model was created for multiparous cows except previous days open was used instead of age at calving and lactation group (lactation = 2 or greater than 2) was added. Mature cows were at higher odds of metritis than second-lactation cows, as were cows that had an RP. Having longer previous days open was associated with higher odds of metritis.

Again, delivering male or twin calves resulted in higher odds of metritis, with twins associated with much higher odds than either male or female calves. Summer calvers had higher odds for metritis relative to spring- or fall-calving cows.

Likewise, the primiparous model for mastitis (log odds of yes/no) offered the following potential explanatory variables for consideration: herd, season fresh, calf outcome and age at calving. Again, herd was significant, and heifers calving in the spring or summer were at higher odds for mastitis. Heifers calving at an older age were associated with higher odds of mastitis.

For the multiparous mastitis model, herd, lactation group and previous days dry were all significantly associated with risk of early lactation mastitis, but season fresh and previous lactation 305me were not. Longer days open and older cows were all associated with higher odds of mastitis. Higher dry log somatic cell count was significantly associated with odds of early mastitis.

To evaluate the relationship of mastitis and metritis on milk production, a multivariate model to fit least squares means was created using 305me milk production for each of the two parity groups: primiparous and multiparous.

Variables offered to the primiparous model included herd, season fresh, calf outcome, mastitis (yes/no), metritis (yes/no), RP (yes/no) and DA (yes/no). Adjusting for the impact of these variables, having mastitis in the first 30 DIM was associated with a reduction in 305me of 2,496 pounds.

Metritis was associated with a reduction in 305me of 610 pounds. RP and DA were associated with losses of 495 and 1,721 pounds each.

For the multiparous cows, a similar model was created except that previous lactation 305me and lactation group were added. Adjusting for the impact of these variables, a case of mastitis was estimated to result in 2,637 pounds less 305me, while metritis resulted in 974 pounds less 305me milk. RP and DA were associated with 723 and 2,449 pounds less milk as well.

To examine the impact of mastitis and metritis on culling and reproduction, separate Cox Proportional Hazards models were created to examine time to event through 300 DIM. Again, separate models were created for the two parity groups.

For both parity-based models, herd, season fresh, 305me milk, metritis (yes/no), mastitis (yes/no), RP (yes/no) and DA (yes/no) were included. In addition, age at calving and lactation group were added to the primiparous and multiparous models, respectively.

In the primiparous culling model, herd, season fresh, age at calving 305me, mastitis, metritis and DA were significantly associated with risk of culling. As expected, a higher 305me was protective against culling. An extra month of age at calving was associated with 1.03 times higher risk of culling. Risk ratios for mastitis, metritis and DA were 1.23, 1.12 and 1.89, respectively. RP was not a significant predictor of culling risk.

In the multiparous model, all variables were significant and remained in the model except for RP and DA. As in the primiparous model, a higher 305me was associated with reduced risk of culling. Adjusting for the other variables, mastitis and metritis in the first 30 DIM were associated with 1.12 and 1.07 times higher risk of culling by 300 DIM, respectively.

In the primiparous reproductive model, 305me and mastitis in the first 30 DIM were not significant predictors of time to pregnancy. A lower age at calving was associated with greater likelihood of successful pregnancy by 300 DIM. Metritis, RP and DA were associated with a 34 percent, 35 percent and 34 percent reduction in likelihood of pregnancy by 300 DIM, respectively.

In the multiparous model, a higher 305me was associated with reduced risk of pregnancy. Mastitis, metritis, RP and DA were all associated with reduced risk ratios for time to pregnancy of 0.85, 0.67. 0.81 and 0.85, respectively.

What does it all mean? Cows with mastitis were predicted to lose approximately 2,500 pounds of 305me milk, to experience a 1.12 to 1.23 times higher odds of premature culling and a 6 to 15 percent reduction in odds of pregnancy by 300 DIM. Cows with metritis were predicted to lose 600 to 975 pounds of 305me milk, experience 1.07 to 1.12 times higher odds of culling by 300 DIM and a 34 percent reduction in odds of pregnancy by 300 DIM.

Due to inconsistent disease definitions, detection and possibly recording across these 20 commercial dairy herds, the true impact is likely greater than this project reveals. Epidemiologically speaking, failing to properly record disease leads to a recording/reporting bias, which leads to an underestimation of the influence of the recorded disease on the herd’s health and production.

By failing to identify all affected animals, the statistical comparisons are now being made between a subset of affected animals and the remaining population that contains both non-affected and affected cows.

This misclassification results in an underestimation of the actual clinical impact and falsely suggests the diseases are having less of a negative impact than is actually occurring.

Conclusions

Mastitis and metritis were demonstrated to negatively affect milk production, culling and reproductive performance in commercial dairy cows from 20 herds across the U.S. The apparent impact, thought significant and costly, is likely underestimated due to a variety of issues including inconsistent disease definition or detection/recording intensity across herds.

Improved disease recording and the inclusion of other potential explanatory variables not routinely recorded on most commercial dairies are likely to affect the measured impact of these common fresh cow disease issues.  PD

References omitted but are available upon request. Click here to email an editor.

• Michael W. Overton

• Elanco Knowledge Solutions – Dairy
• Email Michael W. Overton