Duringthe transitionperiod, dairycows undergolarge metabolic adaptations in glucose, fatty acid, and mineral metabolismtosupportlactationandavoidmetabolicdys- function. The practical goal of nutritional management during this timeframe is to support these metabolic ad- aptations. The National Research Council addressed nu- tritionalmanagementoftransitioncowsforthefirsttime in 2001; however, a substantial amount of research has been reported since this publication was released. Re- sults support 2-group nutritional strategies for dry cows to minimize overfeeding of nutrients during the early dry period but increase nutrient supply to facilitate met- abolic adaptation to lactation during the late dry period. Increasing the amount of energy supplied through di- etary carbohydrate during the prepartum period results in generally positive effects on metabolism and perfor- manceoftransitioncows.Recentresearch,however,sug- gests that the form of that carbohydrate (i.e., starch vs. highlydigestibleneutraldetergentfiber)maybeoflesser importance. Attempts to increase energy supply by feed- ing dietary fat sources or decrease energy expenditure by supplying specific fatty acids such as trans-10, cis- 12 conjugated linoleic acid to decrease milk fat output during early lactation do not decrease the release of nonesterified fatty acids (NEFA) from adipose tissue. Although the view that nutritional means have limited ability to enhance hepatic export of NEFA as triglycer- ides in lipoproteins in ruminants has become dogma, recent evidence suggests that nutrients such as choline or specific fatty acids may enhance this process in transi-

Our aim in this Journal of Dairy Science centennial review is to describe the evolution of focus on metabolic indicators, from discovery and description to evaluation at the individual cow and subsequently herd levels, over the past 100 yr. Furthermore, we discuss current and future technologies that will be used in the dairy industry to utilize these indicators widely going forward. Knowledge of chemical changes in various fluids (e.g., blood, urine, and milk) accompanying numerous metabolic disease states in the dairy cow has existed since almost the beginning of the Journal of Dairy Science 100 yr ago. However, only during the last 25 yr have these metabolic indicators been developed into useful tools for cow- and herd-level monitoring for disease and management. From the 1920s through the 1940s, our understanding of the changes in blood chemistry accompanying milk fever and ketosis increased, as did our understanding of the underlying biology. In the 1950s and 1960s, workers studying ketosis and energy metabolism began to evaluate changes in lipid metabolism reflected by concentrations of circulating nonesterified fatty acids; furthermore, initial development occurred for on-farm tests of milk ketones. During the 1970s, blood metabolic profiling was applied to dairy farms but found to be of varied and limited usefulness. The turning point occurred when large epidemiologic studies of periparturient cow disease were pioneered in the United States, Canada, and Europe in the 1980s; these studies further solidified our understanding of risk factors and epidemiological interrelationships among disease, production, and reproduction. In the early 1990s, scientists first incorporated indicators of metabolic health into large observational studies and determined important epidemiological relationships between these indicators and outcomes of interest. This field of study blossomed during the 2000s as several research groups conducted multiple investigations into metabolic indicators related to energy metabolism and began to develop cow-level thresholds and herd-level alarms for use in monitoring and management. This work was accompanied by additional studies to validate point-of-care instruments that could be used to implement these strategies at the cow and herd levels. Work in the 2000s continued to identify and evaluate other physiological indicators of inflammation and oxidative stress; however, these have yet to be incorporated into large-scale cohort studies. Finally, use of technology (e.g., activity monitoring, cow-monitoring collars and tags, milk-based analysis using Fourier transform infrared spectroscopy) continues to receive significant attention going forward to eventually allow for real-time and automatic monitoring of metabolic indicators and improved health and herd management on dairy farms.

Four multiparous lactating cows (175 to 220 d in milk [DIM]) were used in a 4 x 4 Latin square design to assess the effects of four doses (0.0, 0.5, 1.0, and 1.5 microg/kg of body weight) of lipopolysaccharide (LPS; Escherichia coli 0111:B4) on performance and plasma metabolite and hormone concentrations. In addition, effects of immune activation on in vitro hepatic metabolic capacity were evaluated in 12 multiparous lactating cows (150 to 220 DIM) infused with 0 (n = 6), 1.0 (n = 4) or 2.0 (n = 2) microg of LPS/kg. Milk production and DMI decreased linearly with LPS dose for 24 h after LPS infusion. Overall mean plasma tumor necrosis factor-alpha, insulin, glucagon, and cortisol concentrations increased linearly with LPS dose, and plasma beta-hydroxybutyrate decreased linearly by dose after LPS infusion. Infusion of LPS decreased the insulin:glucagon molar ratio, but did not affect plasma concentrations of growth hormone, insulin-like growth factor-1, leptin, or L-(+)-lactate. Plasma concentrations of glucose tended to increase initially and subsequently decrease, and there was a quadratic tendency for increased plasma nonesterified fatty acid concentrations after LPS administration. In vitro hepatic capacity for conversion of [1-(14)C]L-(+)-lactate and [1-(14)C]palmitate, but not [1-(14)C]propionate or [1-(14)C]L-alanine, to CO2 increased after LPS administration. Hepatic capacity to convert [1-(14)C]propionate to glucose tended to increase, but neither esterification nor the conversion of palmitate to acid soluble products was altered by LPS. The LPS infusion resulted in significant changes of endocrine mediators responsible for regulation of energy metabolism of lactating cows and tended to alter subsequent in vitro hepatic metabolic capacity.

Forty-eight multiparous Holstein cows were fed treatments consisting of either 0, 45, 60, or 75 g/d of a rumen-protected choline (RPC) source in a completely randomized design from 21 d before expected calving to 63 d postpartum to determine whether choline supplementation to the diet would affect hepatic fatty acid and glucose metabolism, key metabolites in plasma, and cow performance. Dry matter intake (DMI), milk yield, body condition score, and body weights (BW) were similar for cows receiving the four treatments. Feeding RPC tended to increase yields of milk fat, 3.5% fat-corrected milk, and total solids. Plasma concentrations of nonesterified fatty acids and beta-hydroxybutyrate were not different among cows fed the four treatments. Concentrations of triglycerides in liver were similar, but concentrations of glycogen in liver increased as cows consumed increasing amounts of RPC. Hepatic capacity for storage of [1-(14)C]palmitate as esterified products within liver slices tended to decrease as the amount of RPC consumed by cows increased; however, effects of treatment on hepatic capacity for oxidation of [1-(14)C]palmitate to CO2 were not significant. These data imply that choline may increase the rate of very low density lipoprotein synthesis and secretion of esterified lipid products from liver. Hepatic capacities for conversion of [1-(14)C] propionate to CO2 and to glucose in liver were similar among cows fed the four treatments. Collectively, these results suggest that hepatic fatty acid metabolism and cow performance are responsive to increasing the supply of choline during the periparturient period.

Holstein cows (n = 30) entering second or greater lactation were fed fat supplements (90 g/d of fatty acids) consisting of Ca salts of either palm fatty acid distillate (control) or a mixture of palm fatty acid distillate and mixed isomers of conjugated linoleic acid (CLA, 30.4 g/ d) from 2 wk prepartum through 20 wk postpartum to determine whether CLA would inhibit milk fat synthesis during early lactation and, in turn, affect energy metabolism of dairy cows during the transition period and early lactation. Feeding CLA did not affect DMI or plasma concentrations of glucose, nonesterfied fatty acids, or beta-hydroxbutyrate during the prepartum period and did not affect postpartum DMI. Feeding CLA reduced milk fat content by 12.5% during early lactation; however, cows fed CLA tended to produce approximately 3 kg/d more milk during the first 20 wk of lactation. Feeding CLA tended to decrease the contribution of short- and medium-chain (C < or = 16) fatty acids to milk fat. Changes in milk yield, milk fat content, and milk fatty acid composition were not apparent until after the second week of lactation. Yield of 3.5% fat-corrected milk, milk protein content, milk protein composition, and calculated energy balance were not affected by treatment. Postpartum concentrations of glucose, nonesterfied fatty acids, and beta-hydroxbutyrate in plasma and hepatic content of glycogen and triglycerides were similar between treatments. These data imply that with CLA treatment in early lactation, dairy cows decreased milk fat synthesis and appeared to respond by partitioning more nutrients toward milk synthesis rather than improving net energy balance.

Twenty Holstein cows in early lactation (7 d in milk) were administered 100 microg of Escherichia coli lipopolysaccharide (LPS) dissolved in 10 mL of sterile 0.9% NaCl saline (treatment; TRT) or 10 mL of sterile saline (control) into both right mammary quarters to test the hypothesis that acute experimental mastitis would have negative impacts on aspects of energy metabolism that might lead to the development of metabolic disorders. A primed continuous intravenous infusion (14-micromol/kg of BW priming dose; 11.5-micromol/kg of BW per h continuous infusion) of 6,6-dideuterated glucose was used to determine pre- and posttreatment glucose kinetics using steady-state tracer methodologies. The LPS-treated cows displayed productive, clinical, and physiological signs of moderate to severe inflammation; control cows displayed no signs of immune activation. Pretreatment glucose rates of appearance (Ra) into plasma were similar (715 and 662 +/- 33 mmol/h for TRT and control, respectively) between treatment groups. Intramammary LPS infusion into TRT cows resulted in increased glucose Ra relative to control cows (mean glucose Ra from 150 through 270 min after intramammary infusion were 815 and 674 +/- 21 mmol/h for TRT and control cows, respectively). Furthermore, plasma concentrations of glucose increased, whereas plasma nonesterified fatty acids, glycerol, and beta-hydroxybutyrate concentrations decreased, in TRT relative to control cows. Interestingly, plasma insulin concentration increased dramatically in TRT cows and occurred prior to the small increase in plasma glucose concentration. Although these results only represent the early stages of inflammation, they are not consistent with a causal relationship between mastitis and energy-related metabolic disorders and instead suggest a coordinated protective effect by the immune system on metabolism during the early stages of mammary insult.