Quantification of the Effects of Ketosis on Reproduction in Dairy Cows by Meta-Analysis

Effects of ketosis on reproduction were reviewed and summarised according to meta-analysis methods. Thirteen peer-reviewed papers in English based on data issued after 1960 from dairy intensive regions were selected. Summary estimates and 95% confidence intervals were calculated according to general variance-based methods for meta-analysis. Possible reasons for heterogeneity of the results were investigated among origin of the data and study design. Clinical ketosis and ketosis diagnosed by a test for ketone bodies in milk or blood were considered separately.

Effects varied widely between studies. Summary estimates [95% CI] of effects of clinical ketosis were 2.5 [1.0, 4.1] more days to first service, 3.8 less % conception at first service, 5.9 [3.0, 8.8] more days to conception, 0.13 [0.06, 0.19] more services per conception, and a hazard ratio of 0.87 [0.82, 0.91] to conceive between 56 and 120 days postpartum.

Different thresholds were used in studies detecting ketotic cows by a test. Results were discrepant and studies with the lowest thresholds concluded in the highest effects. Clinical cases are more likely to be detected and treatment probably contributes to limit their effect.

In Holstein cows from US or Canada herds, no significant effect was evidenced. In Swedish Red and White and Swedish Friesian cows, effects were small or non significant. The highest effects were evidenced in Finncattle, Finnish Friesian or Finnish Ayrshire cows. Results suggest that under current farming conditions, effects of ketosis on reproduction are on average limited, but show large variations in different production systems.