The purpose of this study was to describe the epidemiology of subclinical ketosis (SCK) in dairy cows in early lactation and determine the association of (1) days in milk (DIM) at onset of SCK, and (2) blood β-hydroxybutyrate (BHBA) concentration at onset of SCK with development of displaced abomasum (DA) and removal from herd in the first 30 DIM, conception to first service, days to conception within 150 DIM, and early lactation milk yield. Cows from 4 freestall dairy herds (2 in New York and 2 in Wisconsin) were each tested 6 times for SCK from 3 to 16 DIM using the Precision Xtra meter (Abbott Laboratories, Abbott Park, IL). Subclinical ketosis was defined as a BHBA concentration of 1.2 to 2.9 mmol/L. Mixed-effects multivariable Poisson regression was used to assess DA, removal from herd, and conception to first service. Semiparametric proportional hazards models were used to evaluate days to conception, and repeated-measures ANOVA was used to evaluate milk yield in the first 30 DIM. A total of 741 of 1,717 (43.2%) eligible cows had a least one BHBA test of 1.2 to 2.9 mmol/L. Peak incidence of SCK occurred at 5 DIM, when 22.3% of cows had their first SCK-positive test. Peak prevalence of SCK occurred at 5 DIM, when 28.9% of cows had a SCK-positive test. Median time from first positive SCK test until BHBA test <1.2 mmol/L was 5d. Cows first testing SCK positive from 3 to 5 DIM were 6.1 times more likely [95% confidence interval (CI) = 2.3 to 16.0] to develop a DA than cows first testing SCK positive at 6 DIM or later. Cows first testing SCK positive from 3 to 7 DIM were 4.5 times more likely (95% CI = 1.7 to 11.7) to be removed from the herd, were 0.7 times as likely (95% CI = 0.6 to 0.8) to conceive to first service, and produced 2.2 kg less milk per day for the first 30 DIM than cows first testing positive at 8 DIM or later. Each 0.1 mmol/L increase in BHBA at first SCK-positive test increased the risk of developing a DA by a factor of 1.1 (95% CI = 1.0 to 1.2), increased the risk of removal from herd by a factor of 1.4 (95% CI = 1.1 to 1.8), and was associated with a decrease in milk production by 0.5 kg/d for the first 30 DIM. These results show that time of onset and BHBA concentration of first SCK-positive test are important indicators of individual cow performance.

Two experiments were conducted 1) to validate a field protocol for the determination of ruminal pH and 2) to develop a strategy to interpret ruminal pH data from groups of cows. In the first experiment, ruminal fluid was collected from 30 lactating dairy cows. Ruminal fluid pH was 0.28 pH units lower for fluid collected by rumenocentesis than for fluid collected through a ruminal cannula. Concentrations of volatile fatty acids were correspondingly higher in samples collected by rumenocentesis. A portable pH meter capable of measuring pH of a very small volume of ruminal fluid yielded very similar pH readings as did a standard meter with a pH probe. Filtration or aspiration of ruminal fluid had no effect on pH. In the second experiment, a strategy was developed to use ruminal pH values from a subsample of cows to distinguish between groups fed either a low or higher forage diet. Groups could be distinguished using a cut point of 5.5 ruminal pH, a sample size of 12 cows, and a critical value of 3 or more cows below the cut point. This strategy had the lowest theoretical error rate for herds with either a high or low prevalence of cows with a low ruminal pH.

A meta-analysis of previous studies was performed to clarify the response of prepartum dairy cows to lowering dietary cation-anion difference (DCAD) and to compare different equations that have been proposed to calculate DCAD. Twenty-two published studies containing 75 treatment groups met criteria for inclusion in the meta-analysis. Five different equations used to calculate DCAD were compared for their association with clinical milk fever and urinary pH. The DCAD equation (Na + K) - (Cl + 0.6 S) was the most highly associated with clinical milk fever (R(2) = 0.44) and urinary pH (R(2) = 0.85). Lowering DCAD reduced clinical milk fever but also reduced DM intake. Lowered DCAD was associated with reduced urinary pH, blood bicarbonate, and blood CO(2), suggesting a metabolic acidosis with respiratory compensation. Blood pH was very slightly lowered by lowered DCAD. Lowering DCAD increased ionized Ca in blood before and at calving. The model predicted that lowering DCAD from +300 to 0 mEq/kg reduced risk for clinical milk fever from 16.4 to 3.2%, reduced urinary pH from about 8.1 to 7.0, and reduced DM intake by 11.3%.