Symposium review: The impact of absorbed nutrients on energy partitioning throughout lactation*

Most nutrition models and some nutritionists view ration formulation as accounting transactions to match nutrient supplies with nutrient requirements. However, diet and stage of lactation interact to alter the partitioning of nutrients toward milk and body reserves, which, in turn, alters requirements. Fermentation and digestion of diet components determine feeding behavior and the temporal pattern and profile of absorbed nutri-ents. The pattern and profile, in turn, alter hormonal signals, tissue responsiveness to hormones, and mammary metabolism to affect milk synthesis and energy partitioning differently depending on the physiological state of the cow. In the fresh period (first 2 to 3 wk postpartum), plasma insulin concentration and insulin sensitivity of tissues are low, so absorbed nutrients and body reserves are partitioned toward milk synthesis. As lactation progresses, insulin secretion and sensitivity increase, favoring deposition instead of mobilization of body reserves. High-starch diets increase ruminal propionate production, the flow of gluconeogenic precursors to the liver, and blood insulin concentrations. During early lactation, the glucose produced will preferentially be used by the mammary gland for milk production. As lactation progresses and milk yield decreases, glucose will increasingly stimulate repletion of body reserves. Diets with less starch and more digestible fiber increase ruminal production of acetate relative to propionate and, because acetate is less insulinogenic than propionate, these diets can minimize body weight gain. High dietary starch concentration and fermentability can also induce milk fat depression by increasing the production of biohydrogenation intermediates that inhibit milk fat synthesis and thus favor energy partitioning away from the mammary gland. Supplemental fatty acids also impact energy partitioning by affecting insulin concentration and insulin sensitivity of tissues. Depending on profile, physiological state, and interactions with other nutrients, supplemental fatty acids might increase milk yield at the expense of body reserves or partition energy to body reserves at the expense of milk yield. Supple-mental protein or AA also can increase milk production but there is little evidence that dietary protein directly alters whole-body partitioning. Understanding the biology of these interactions can help nutritionists better formulate diets for cows at various stages of lactation.


INTRODUCTION
Energy partitioning is the result of complex mechanisms that involve a variety of hormones and tissues and is affected by absorbed nutrients and the physiological state of the cow.In this review, we define the physiological state of a cow as her overall bodily condition characterized by a dominant biological process (e.g., growth, lactation, gestation, disease, maintenance) and associated with specific metabolic and endocrine profiles.In broad terms, the energy provided by absorbed nutrients or mobilized metabolites is used first for maintenance, and then for milk production, growth, gestation, and body reserves depending on the stage of lactation and level of intake relative to requirements.One goal when feeding and managing cows is to supply the nutrients needed by the cow to support her various needs, especially milk production.Another goal, and the focus of this review, is to stimulate the flow of nutrients toward the mammary gland and to control partitioning toward body reserves, sometimes increasing deposition (or decreasing mobilization) and sometimes decreasing deposition and maintaining reserves.Nutrients from the diet can affect partitioning to milk and body reserves through endocrine regulators, tissue receptors, and cellular signaling pathways.Endocrine Symposium review: The impact of absorbed nutrients on energy partitioning throughout lactation* P. Piantoni1 † and M. J. VandeHaar 2regulators, available nutrients, environment, management, and genetics can affect the synthetic capacity of the mammary gland and increase the pull of nutrients to produce milk (Figure 1).Therefore, formulating diets for dairy cows requires knowledge on how physiological states of cows vary throughout a lactation, and how absorbed nutrients can interact with physiological state and affect mammary gland synthetic capacity and energy partitioning.This is the third of 4 review articles from a symposium conducted to recognize the contributions of Professor Michael S. Allen in the field of dairy nutrition.The first 2 articles (Albornoz et al., 2023;Oba and Kammes-Main, 2023) and this one provide the background for the last article, which integrates control of intake and energy partitioning concepts into ration formulation (Allen, 2023).The objective of this review is to cover mechanisms involved in the regulation of energy partitioning, how they vary throughout lactation, and how absorbed nutrients can affect them.We will discuss the effects of carbohydrates, fatty acids (FA), and protein, and briefly discuss methods to assess partitioning.

REGULATION OF ENERGY PARTITIONING
Homeorhesis is the coordination of metabolism to support a dominant physiological process, such as growth, pregnancy, or lactation (Bauman and Currie, 1980).Nutrient supply is important for partitioning, as will be discussed later, but the endocrine system plays a major role in determining whether energy is parti-tioned toward a growing fetus, milk, or body reserves.Endocrine changes involving growth hormone (somatotropin, GH) and insulin, as well as tissue responsiveness to these hormones, are central to the regulation of nutrient partitioning (Tucker, 2000;Baumgard et al., 2017).Stricker and Grüter (1928) showed that pituitary extracts injected into ovariectomized rabbits induced mammary gland development and milk production, demonstrating the central role of the pituitary gland on lactation.Growth hormone is one of the protein hormones secreted by the pituitary gland, and it stimulates cell proliferation and anabolic pathways primarily through increased synthesis of IGF-I, which has endocrine, paracrine, and autocrine mitogenic actions.Although IGF-I cannot fully explain the galactopoietic effect of GH, IGF-I helps maintain the secretory capacity of the mammary gland (Tucker, 2000).The IGF-I produced by the liver is secreted into the blood and is involved in a negative feedback loop, inhibiting GH release when its concentration is high.In addition to stimulating IGF-I synthesis, GH also has direct effects on tissues to decrease sensitivity to insulin.Most importantly, GH increases lipolysis when cows are in negative energy balance and decreases lipogenesis when cows are in positive energy balance.It also decreases insulindependent glucose uptake in adipose tissue through decreased glucose transporter 4 (GLUT4) translocation and decreases insulin stimulation of lipogenesis in adipose tissue and inhibition of gluconeogenesis in liver (Etherton and Bauman, 1998).
Piantoni and VandeHaar: RUMINANT NUTRITION SYMPOSIUM Figure 1.Feed availability and diet nutrient profile determine nutrient intake, which is affected by endocrine regulators such as gut peptides and by other mechanisms of control of intake such as hepatic energy charge or gut fill.Nutrients will be used for maintenance, pregnancy, body reserves, or milk depending on stage of lactation or gestation.Endocrine regulators, such as growth hormone and insulin, and tissue responsiveness to them will determine the partitioning of nutrients between milk and body reserves.Several factors can affect the synthetic capacity of the mammary gland, including endocrine regulators (e.g., growth hormone), available nutrients (e.g., certain amino acids and fatty acids), environment (e.g., heat stress), management (e.g., milking frequency), and genetics, and therefore, the pull of nutrients to produce milk.
In early 1920s, the successful isolation of insulin from pancreatic extracts and treatment of patients with diabetes opened the door to understanding the effects of insulin on metabolism (Banting et al., 1922).Insulin is a protein hormone secreted by the pancreas under a variety of endocrine and nutrient signals, such as glucose (Hove, 1978), butyrate and propionate (Manns and Boda, 1967;Manns et al., 1967), and certain AA (Hertelendy et al., 1970).Therefore, increasing intake usually increases insulin concentration.However, it is not only intake that affects insulin concentration, but intake relative to requirements.Bradford and Allen (2007) showed that phlorizin (a compound that decreases renal reabsorption of glucose) decreased insulin despite no change in intake; increasing glucose loss in urine would have increased nutrient requirements, and because intake did not change, the level of intake relative to requirements decreased, decreasing insulin concentration.Consistent with this, high-producing cows typically have lower insulin concentrations than lower-producing cows (Lucy et al., 2009;Zinicola and Bicalho, 2019).Many of the metabolic effects of insulin are opposite to those of GH.Insulin increases glucose uptake by muscle and adipose tissue by stimulating translocation of GLUT4 to the cell membrane, and it inhibits protein degradation in muscle, lipolysis in adipose, and gluconeogenesis in liver (De Koster and Opsomer, 2013).The capacity of insulin to elicit a biological response depends on the preservation of intracellular signaling pathways within target cells (Taniguchi et al., 2006).Factors that can decrease insulin activity, in addition to GH concentration, are the concentration of proinflammatory cytokines such as tumor necrosis factor-α (Kushibiki et al., 2001) and leptin (Müller et al., 1997).When GH concentration is high and insulin is low, energy partitioning would be shifted toward milk production and away from body reserves; the opposite will occur when GH concentration is low and insulin is high.

CHANGES IN PHYSIOLOGICAL STATE THROUGHOUT THE LACTATION CYCLE
Metabolic and endocrine adaptations as the dairy cow goes through lactation have been studied for decades and have been the topic of multiple reviews (Hart, 1983;Bell., 1995;Bauman, 2000;Bell and Ehrhardt, 2000;De Koster and Opsomer, 2013;Baumgard et al., 2017).These adaptations favor mobilization of body reserves to support fetal growth later in gestation and mammary functions around the time of parturition (Bell, 1995).As lactation progresses, feed intake increases, and mobilization is no longer needed.Later in lactation, especially in pregnant cows, body reserves are replenished as milk production declines, and cows are usually dried-off approximately 2 mo before expected calving to allow for the mammary gland to involute and remodel before the cycle starts again.
In late gestation, glucose sparing becomes important when fetal growth rate is high and the mammary gland prepares for the next lactation, with an increase in mammogenesis during the last month and lactogenesis during the last week before parturition (Bell, 1995;Bell and Ehrhardt, 2000).Insulin and IGF-I concentrations begin to decrease approximately 3 wk prepartum and concentrations of nonesterified fatty acids (NEFA) increase in the last week prepartum with a sharp increase close to parturition (VandeHaar et al., 1999;Doepel et al., 2002); these changes are most pronounced in cows fed low-energy diets and are more pronounced in cows than heifers (VandeHaar et al., 1999).In addition, in the last weeks prepartum, insulin resistance (i.e., the state in which a higher concentration of insulin is needed to elicit a certain biological effect) of extrahepatic tissues increases (Bell, 1995;Zachut et al., 2013).In late gestation, GH concentration is low and starts to increase 1 wk before parturition (Simmons et al., 1994).Glucose transporter 4 expression in muscle and adipose tissue is high during the dry period (De Koster and Opsomer, 2013).Nutrient mobilization (i.e., NEFA from adipose, AA from muscle) and ketone production in the liver increase because of tissue insulin resistance, low insulin concentration, and elevated GH concentration.Mobilized nutrients will be used as sources of energy by the cow to allow for glucose uptake by the fetus and by the mammary gland for mammogenesis and lactogenesis before parturition.
Around parturition, expression of GH receptor 1A in the liver is downregulated and IGF-I synthesis and release decrease, despite normal concentrations of GH in blood (Lucy et al., 2001).A lower circulating IGF-I concentration decreases the negative feedback on the pituitary gland and GH secretion, allowing for an increase in circulating GH.This is referred to as the uncoupling of the somatotropic axis, composed of GH, GH receptor, and IGF-I (Lucy et al., 2001), which seems to be most pronounced in cows with the most negative energy balance (Lucy et al., 2009).During the fresh period (first 2 to 3 wk postpartum), mammary epithelial cell number and activity and liver size and gluconeogenic capacity increase (Bauman, 2000), increasing both glucose utilization and production (Figure 2).Plasma insulin concentration remains low (Doepel et al., 2002), and insulin resistance of extrahepatic tissues (Bell, 1995) and GH concentration are high (Doepel et al., 2002), favoring partitioning of nutrients toward the mammary gland for milk production and away from body reserves.Important to note is that even though extrahepatic tissues are insulin resistant around parturition, hepatic tissue is not, and the increase in gluconeogenesis postpartum (crucial for lactose synthesis and, therefore, milk production) is likely supported by the lower insulin concentration observed in this period (Zachut et al., 2013).Whereas GLUT4 expression decreases in adipose and muscle tissue, resulting in decreased glucose uptake, GLUT1 expression in mammary tissue increases (Zhao, 2014), increasing glucose uptake for milk production.The result of these hormone concentrations and tissue sensitivities in the fresh period is a lipolytic state with increasing circulating NEFA that allows glucose to be spared for milk production.
In the liver, NEFA is either completely oxidized for energy, partially oxidized into ketone bodies, or esterified and exported or stored as triglycerides.Ketone bodies are exported and serve as a source of energy for tissues or as a precursor for milk fat synthesis, whereas triglycerides contribute to milk fat production when exported and hepatic steatosis when stored.Hepatic oxidation of NEFA increases the pool size of acetyl CoA, and hepatic acetyl CoA content has been negatively related to intake in the postpartum cow (Stocks and Allen, 2012).Lower intakes relative to requirements keep insulin concentrations low, and insulin is required to decrease mobilization of body reserves.The hepatic oxidation theory suggests that, ultimately, what controls intake is the energy charge of the liver, ([ATP] + 0.

[ADP])/([ATP] + [ADP] +
[AMP]), and not the oxidation of fuels per se (Allen et al., 2009;Allen and Piantoni, 2013;Allen, 2020).Therefore, when hepatic energy charge is low enough, the cow will eat, regardless of the level of hepatic acetyl CoA and the oxidation of NEFA or other nutrients.The flow of gluconeogenic precursors (particularly propionate, but also gluconeogenic amino acids, lactate, and glycerol) to the liver increases during a meal and "activates" the tricarboxylic acid cycle, acetyl CoA is oxidized, and hepatic energy charge increases until a threshold is met, and the meal ends (Allen and Piantoni, 2013;Allen, 2020).Insulin increases during meals, inhibiting lipolysis and decreasing NEFA concentration in plasma.The magnitude of the decline in NEFA concentration in plasma after feeding might be related to insulin sensitivity (i.e., state in which a lower concentration of insulin is needed to elicit a certain biological effect).Piantoni et al. (2015c) showed that the decline in NEFA concentrations after the first 4 h of feeding was linearly and positively related to intake in postpartum dairy cows.We speculate that this was related to higher insulin sensitivity of extrahepatic tissues, and that higher insulin sensitivity allowed cows to reach higher intakes immediately postpartum.
In early lactation (~21 to 100 DIM), mammary synthetic capacity and liver gluconeogenic capacity are high, insulin concentration increases, and insulin sensitivity increases.As the cow reaches positive energy balance, the uncoupling of the somatotropic axis is reversed (Sharma et al., 1994;Kobayashi et al., 1999), and this reversal appears to be positively affected by insulin concentrations (Butler et al., 2003).Consequently, GH concentration decreases (Koprowski and Tucker, 1973) with increasing IGF-I, and expression of GLUT4 in muscle and adipose tissue increases.Because the synthetic capacity of the mammary gland is high, nutrients continue to be partitioned to milk production, but the mobilization of body reserves decreases as insulin sensitivity and DMI increase.
As lactation progresses and cows move into mid (~100 to 200 DIM) and late (beyond ~200 DIM) lactation, the concentration of circulating insulin increases further, the concentration of GH remains low, and insulin sensitivity and GLUT4 expression in muscle and adipose tissue are high.These changes in hormone concentrations and tissue sensitivity progressively shift nutrient partitioning from milk to body reserves.
Other physiological states such as heat stress and inflammation can also affect energy partitioning.Wheelock and collaborators (2010) showed that during heat stress, milk synthesis decreases, decreasing the pull of nutrients by the mammary gland.The decreased mammary pull might explain the findings that heatstressed cows also had elevated plasma insulin and they were not mobilizing body reserves despite the decrease in DMI.In pair-fed cows that were not heat-stressed, lower intake decreased insulin concentration and increased mobilization of reserves as the cows continued to try to maintain high milk synthesis (Wheelock et al., 2010).The effect of heat stress on metabolism has been reviewed by Baumgard and Rhoads (2013) and is a good example of how environment and management can affect physiological state and energy partitioning in the dairy cow.Recently, Baumgard et al. (2017) reviewed metabolic responses to inflammation.During this physiological state, intake usually decreases but insulin concentration increases, and a state of insulin resistance develops, increasing glucose availability for the immune system.In addition, milk synthesis is reduced, further increasing glucose availability.The increase in insulin resistance is likely related to the increase in proinflammatory cytokines such as tumor Piantoni and VandeHaar: RUMINANT NUTRITION SYMPOSIUM necrosis factor-α, which has been shown to decrease glucose clearance from blood after an intravenous injection of insulin (Kushibiki et al., 2001).Inflammation does not occur only due to infection, and all cows may experience systemic inflammation in the immediate postpartum period (Bradford et al., 2015), which might be partly responsible for the increase in insulin resistance observed.

Carbohydrates
Diet composition, as well as fermentation and digestion of diet components, determine the temporal pattern and profile of absorbed nutrients.For example, a diet high in fermentable starch increases propionate absorption in the rumen, whereas a diet with less fermentable but still highly digestible starch increases glucose absorption in the small intestine.The pattern and profile of nutrients absorbed, in turn, alter hormonal signals to affect energy partitioning (Allen and Piantoni, 2014).In the rumen, dietary carbohydrates are degraded by microbial enzymes into glucose, and glucose is fermented by microbes to produce energy for growth and VFA and gases (i.e., CO 2 , CH 4 ) as waste.The rumen is the major site of degradation for fiber but not necessarily for starch.The amount of starch fermented in the rumen for a given feed will depend on factors such as source, processing, conservation method, and length of storage.High-starch diets increase ruminal propionate production, the flow of gluconeogenic precursors to the liver, and blood insulin concentrations (Oba and Allen, 2003a).In high-producing dairy cows, the glucose produced is preferentially used by the mammary gland for milk production.Considering that as much as 85% of rumen absorbed propionate would be converted to glucose in the liver (Steinhour and Bauman, 1988), a proportion of this VFA should be available in arterial blood to reach the pancreas and stimulate insulin secretion.In contrast, high-fiber diets increase the production of acetate, which is not taken up by liver and instead is used as an energy source for mammary and extrahepatic tissues or building block for milk fat synthesis.Because acetate is less insulinogenic than propionate (Manns and Boda, 1967), diets with less starch and more digestible fiber that increase ruminal production of acetate relative to propionate can minimize BW gain.
Several experiments have evaluated the effect of replacing dietary starch with various NDF sources on production performance of dairy cows and showed that this strategy can maintain milk energy output while decreasing body reserves gain.Oba and Allen (2000) demonstrated that when cows fed a corn silage-based ration were switched to a lower-starch/higher-forage NDF diet with less than 6% added corn grain compared with a higher-starch/lower-forage diet with more than 25% added corn grain, FCM was maintained with 2 kg/d lower DMI and gain in BW decreased from 1 to 0 kg/d.Others substituted increasing levels of nonforage fiber sources (NFFS) for cereal grains and observed similar results, and the negative effects on intake and lactation performance were only observed at the highest levels of replacement.Ipharraguerre et al. (2002) showed that increasing levels of soyhulls by replacing ground corn in diets linearly decreased BW gain and increased milk fat content and yield, with no effect on yield of FCM.Voelker and Allen (2003) replaced high-moisture corn with beet pulp and showed that decreasing starch linearly decreased DMI and tended to increase FCM yield at intermediate levels of beet pulp supplementation.Increasing beet pulp linearly decreased plasma insulin and tended to decrease BCS.Potts et al. (2015) consolidated results from 4 separate experiments and reported that overall, decreasing dietary starch concentration from 30 to 14% tended to decrease DMI, and decreased milk yield and BW gain, but maintained milk energy output.In addition, lowstarch diets decreased plasma concentrations of insulin by 20% and increased plasma concentration of NEFA by 42% compared with high-starch diets, likely due to increased lipolysis from lower insulin (Burczynski, 2014).Lowering dietary starch decreased milk protein content and yield (Oba and Allen, 2000;Potts et al., 2015), and this effect seemed to be dose dependent (Ipharraguerre et al., 2002;Voelker and Allen, 2003).Overall, results from these experiments suggest that FCM yield and BW can be maintained when replacing dietary starch with a source of fermentable fiber, and that potential losses in milk protein output should be considered and monitored.
Studies reporting effects of ruminal infusions of acetate or propionate can help explain the effects of replacing starch with fermentable fiber.Rook and Balch (1961) showed 60 yr ago that continuous intraruminal infusions of propionic acid increased milk protein yield but decreased milk fat yield, whereas acetic acid infusions increased milk fat yield.Consistent with this, intraruminal infusions of acetate increased milk fat yield (Urrutia and Harvatine, 2017).
Importantly, dietary starch level interacts with level of production of cows to affect performance.Boerman et al. (2015a) showed that, compared with low-starch diets, high-starch diets did not affect milk or fat yield in low-producing cows (~30 kg/d milk yield) but increased milk yield by ~9 kg/d with no change in milk fat content in high-producing cows (~60 kg/d milk yield).In other words, feeding high starch, which likely increased rumen propionate production more than acetate, was most beneficial for the highest-producing cows-those with the greatest need for glucose.However, no interaction between level of production and dietary starch concentration was detected for BW change in that experiment (Boerman et al., 2015a).Grouping cows by level of production is likely more relevant than DIM Piantoni and VandeHaar: RUMINANT NUTRITION SYMPOSIUM if cows are in positive energy balance (Allen, 2023).Experiments where dietary treatments are compared in cows with a wide range of milk production are critical to understanding interactions between diets and level of production on lactation performance to improve feed allocation.
We hypothesize that the capacity of hepatic gluconeogenesis is at times overwhelmed when feeding highly fermentable starch diets, so that unusually high concentrations of propionate reach the pancreas.This could occur earlier in lactation when gluconeogenic capacity is still limited or later in lactation as the need for glucose by the mammary gland decreases (Oba and Allen, 2003b;Bradford et al., 2006).Consistent with this, propylene glycol increased insulin concentration when drenched but not when mixed in the diet in late-lactation dairy cows at similar doses (Chung et al., 2009).Moreover, Sutton et al. (1985Sutton et al. ( , 1986) ) showed that feeding a higher-concentrate diet increased mean daily plasma propionate and insulin concentration and decreased milk fat yield, and that increasing the feeding frequency of this diet decreased plasma insulin and NEFA concentration and increased milk fat yield without affecting mean plasma propionate concentration.We suggest that increased feeding frequency would allow highly fermentable starch diets to be fed in a way that decreases average insulin concentration within a day and thus decreases BW gain.
Increasing propionate production in the rumen with starch sources, or acetate with forage or NFFS can affect energy partitioning partly through an effect on circulating insulin.In addition, effects on energy partitioning could be related to an effect on rumen environment and microbiome (Pitta et al., 2018), and therefore, the flow of trans-FA [biohydrogenation (BH) intermediates] to the duodenum (Bauman and Griinari, 2003).Trans-FA decrease milk fat synthesis in the mammary gland (Baumgard et al., 2002) and increase lipogenesis in adipose tissue (Harvatine et al., 2009).The risk for a shift in BH pathways is greater in diets with higher concentrations of rapidly fermented starch or UFA, and the risk is even greater in diets that contain large amounts of both.In addition, the stage of lactation can interact with diet fermentability and UFA levels to affect rumen BH.A highly fermentable starch diet (i.e., 32%) with no supplemental FA decreased milk fat content in lower-producing but not higher-producing cows, and the decrease in milk fat was related to an increase in milk trans-FA content; milk fat yield was not reported by production level (Bradford and Allen, 2004).In contrast, Rico et al. (2014) showed that a calcium soap of palm FA supplement increased milk trans-FA yield and decreased milk fat content in high-producing cows but not in low-producing cows; researchers speculated that this was likely related to their higher passage rates.Even though higher UFA decreased milk fat content only in high-producing cows, they decreased milk fat and de novo FA yield in both high-and low-producing cows.Diets fed by Bradford and Allen (2004) did not contain any supplemental FA source, and FA fed were mostly in triglyceride form.Therefore, longer retention in the rumen of cows with lower passage rates could have benefited their release, hydrolysis, and later BH, which in this experiment was altered due to the high fermentable starch content of the diet.High-producing cows likely had faster passage rates, which could have decreased FA exposure and BH, preventing the drop in milk fat content and the potential effect on energy partitioning.In the experiment conducted by Rico et al. (2014), lower-producing cows had less time with rumen pH<5.6 compared with high-producing cows, which could have prevented the shift in rumen BH and decreased the dissociation of calcium soaps, increasing the flow of intact FA to the duodenum.Alternatively, and because of a slower passage rate, dissociated UFA could have been completely biohydrogenated before leaving the rumen, explaining the maintenance of milk fat content compared with control.The increased time under rumen pH 5.6 and the higher passage rates could have favored dissociation of calcium soaps and a shift in rumen BH, explaining the increase in trans-FA in milk and the decline in milk fat content in high-producing cows.
Starch sources that are less fermentable in the rumen but still digestible in the small intestine increase supply of absorbed glucose.Glucose is not taken up by the liver but used by the mammary gland for milk production or body reserves depending on stage of lactation.In addition, glucose could be used by the portal-drained viscera as an energy source, sparing glucogenic AA, which in turn, will be available to extrahepatic tissues or the liver for gluconeogenesis (Nocek and Tamminga, 1991).Brown and Allen (2013) infused 1 kg/d of glucose intravenously for 12 d after calving and reported that glucose increased insulin concentration and decreased NEFA but had no overall effect on BW change or ECM yield.However, glucose infusion interacted with the base diet fed, so that it increased milk yield by 5 kg/d without affecting intake in a high-starch/ low-protein diet and numerically depressed intake by 3 kg/d while maintaining milk yield in a low-starch/highprotein diet.In early-lactating, higher-producing cows, prolonged (14-d periods) abomasal infusion of partially hydrolyzed starch also increased circulating insulin and decreased NEFA concentrations and tended to increase milk production (Knowlton et al., 1998).In contrast, Amaral et al. (1990) showed that intravenous infusion of glucose (11-d periods) only numerically increased milk yield in early lactation cows while also increasing insulin concentration.Bartley and Black (1966) reported that exogenous supply of glucose decreased hepatic gluconeogenesis, which could partly explain the lack of a significant effect of glucose infusion on milk yield observed by Amaral et al. (1990).However, Amaral et al. (1990) reported no effect of glucose infusion on estimated endogenous glucose production but an increase in glucose utilization (milk, body reserves, or oxidation).Inconsistent responses in milk yield between Knowlton et al. (1998) and Amaral et al. (1990) could be from doses of glucose used, interactions with dietary nutrients, and characteristics of the cows used (e.g., level of production or insulin sensitivity) that determined the fate of the glucose infused.More recently, Al-Trad et al. (2009) observed that prolonged (28-d periods) intravenous infusion of increasing levels of glucose linearly increased insulin concentration, BW, and backfat thickness and did not affect milk or ECM yield in mid-lactation low-producing cows.The increase in insulin concentration was accompanied by a decrease in NEFA, which was expected considering the antilipolytic effects of insulin.Performance responses in these lower-producing dairy cows were likely due to a less demanding mammary gland and higher insulin sensitivity.In general, intravenous or abomasal infusions of glucose increase insulin concentration, decrease lipolysis and circulating NEFA, and increase milk yield or body reserves depending on stage of lactation; only longer infusion studies (>10-d periods) were discussed because they might more closely reflect feeding situations.

Fatty Acids
Depending on the profile of the FA being fed and interactions with dietary components and physiological state of cows, FA supplementation can also shift energy partitioning through direct effects on the mammary gland and by affecting insulin concentration and insulin sensitivity of tissues.A clear example of how FA can affect partitioning by affecting the pull of the mammary gland would be milk fat depression (MFD) syndrome, which is characterized by a decrease in milk fat yield, with no change in yield of other milk components or milk.During MFD, a shift in the normal ruminal BH process increases production of trans-FA such as trans-10,cis-12 C18:2, cis-10,trans-12 C18:2, and trans-9,cis-11 C18:2 (Bauman et al., 2011).Small amounts of these trans-FA reaching the small intestine for absorption can have a great effect on milk fat production.In an abomasal infusion study, Baumgard et al. (2001) showed that 0.016% of diet DM of trans-10,cis-12 C18:2 decreased milk fat synthesis by 25%.Later, Baum-gard et al. (2002) showed that trans-10,cis-12 C18:2 decreased FA in milk through the downregulation of key mammary lipogenic enzymes.Harvatine et al. (2009) evaluated adipose tissue gene expression from cows abomasally infused with trans-10,cis-12 C18:2 and observed an upregulation in key lipogenic enzymes in adipose tissue, which could indicate a direct effect of trans-FA on adipose tissue or an indirect effect through increased fuel availability from decreased milk fat synthesis on energy partitioning.
Nutritional factors that can cause MFD are increased diet fermentability (sometimes associated with lower ruminal pH), increased load of dietary UFA, and inclusion of ionophores in the diet (Bauman et al., 2011;Jenkins and Harvatine, 2014).Ruminal pH has been positively related to milk fat percent in dairy cows (Allen, 1997), and even though low ruminal pH favors synthesis of trans-FA and the risk of MFD, a low ruminal pH is not a prerequisite for the shift in BH to occur (Jenkins and Harvatine, 2014).Harvatine and Allen (2006a,b) showed that increasing the supplementation of mono-and polyunsaturated 18-carbon FA linearly decreased DMI and FCM yield compared with an SFA supplement.The decline in FCM yield was due to a decrease in milk yield and milk fat content, partly related to the drop in intake.Despite the drop in intake, UFA increased empty BW gain, and this was not related to an increase in insulin concentration but to an increase in the content of trans-FA in milk, indicating an altered rumen BH.Thus, the direct inhibition of milk fat synthesis by trans-FA likely explains the shift in energy partitioning from milk to body reserves.Liu et al. (2020) compared their findings feeding soybean oil with those of Boerman et al. (2015b) feeding a higherstarch diet.Starch increased insulin and trans-FA in milk but had almost no effect on milk energy output, whereas UFA caused a smaller increase in insulin, increased trans-FA, and decreased milk energy output.Both starch and UFA increased the proportion of energy partitioned to body stores.
Fatty acids could also affect energy partitioning through changes in plasma insulin concentration.In perfused pancreases of rats, both SFA and UFA increased insulin secretion, but the insulinotropic action of SFA was higher than that of UFA (Stein et al., 1997).In addition, FA stimulated insulin more when glucose concentration was greater.Consistent with these findings, insulin responses to FA supplementation in dairy cows are varied, with some studies showing an increase (Piantoni et al., 2013(Piantoni et al., , 2015b;;Prom et al., 2021) and some a decrease (Harvatine and Allen, 2006b;de Souza et al., 2019); these varied insulin responses may be partly due to differences in FA profiles.Chilliard (1993) suggested that the difference in insulin response to FA Piantoni and VandeHaar: RUMINANT NUTRITION SYMPOSIUM supplementation is related to the response in DMI, the ingredients that were removed from the diet to add FA, or the glucose-sparing effect of a decrease in milk fat synthesis.In addition, parity, stage of lactation, and interactions with other dietary components can influence the effect of fat supplements on insulin secretion.For example, palmitic acid increased ECM yield regardless of parity, but particularly in multiparous cows, and did not affect BW gain in multiparous cows but increased it in primiparous cows, and this increase in BW was related to an increase in circulating insulin (de Souza and Lock, 2018).
Research suggests FA can also affect insulin resistance.In fresh cows, de Souza and Lock (2019) showed that palmitic acid increased ECM yield by 4.7 kg/d and BW loss by 0.76 kg/d without affecting DMI, and that the decline in BW loss was at least partially related to a 13% decrease in circulating insulin (de Souza et al., 2019).The reported decrease in BW loss could also be related to an increase in insulin resistance.Rico et al. (2016) reported that palmitic acid increased circulating ceramides in cows, and because ceramides can decrease insulin sensitivity in adipocytes (Rico al., 2018), palmitic acid might be able to shift energy partitioning toward milk production and away from body reserves (McFadden, 2020).Oleic acid might have opposite effects to those of palmitic acid on insulin resistance.de Souza et al. (2021) showed that decreasing the ratio of palmitic to oleic acid from 80:10 to 60:30 linearly increased DMI and decreased BW loss in fresh cows.Regardless of palmitic to oleic ratio, FA supplements increased ECM by 4 kg/d compared with control.The reduction in BW loss observed in cows fed higher levels of oleic acid could have been related to an increase in insulin sensitivity, previously reported in adipose tissue of fresh cows abomasally infused with oleic acid (Laguna et al., 2019).Results from de Souza et al. ( 2021) and Piantoni et al. (2015c) suggest that developing strategies to increase insulin sensitivity of extrahepatic tissues and decrease mobilization of fat more quickly during meals in the fresh period might be beneficial; a decrease in hepatic oxidation of NEFA would decrease satiety signals and allow for longer meals, potentially increasing intake over a day and improving performance.On the other hand, developing strategies to increase insulin resistance later in lactation could minimize excessive body reserves gain.
Supplementation of high levels of UFA has been related to lower DMI (Allen, 2000), which has sometimes prevented nutritionists from feeding highly unsaturated FA supplements to fresh cows that are already struggling with lower intakes.Consistent with this understanding, dos Santos Neto et al. (2021) showed that a higher oleic acid supplement (35% oleic acid) fed at 1.5% of diet DM decreased DMI without affecting FCM yield or BW loss compared with a control diet with no supplemental FA in fresh cows.In addition, researchers showed that a palmitic acid supplement increased FCM yield but did not affect BW loss compared with a control diet with no supplemental FA, which is in contrast with previous reports of palmitic acid increasing BW loss in the fresh period (de Souza and Lock, 2019;de Souza et al., 2021).The level of UFA (not only oleic acid, but also linoleic and linolenic acids) that could be fed to fresh cows without depressing intake while still increasing insulin sensitivity of extrahepatic tissues should be determined.
Limited research indicates that the effect of FA on energy partitioning might depend on the base diet being fed.In an experiment with early lactation cows, Weiss and Pinos-Rodriguez (2009) showed that a long-chain SFA supplement (a mixture of stearic and palmitic acids) increased BCS without affecting milk production when fed in a high-forage diet, but increased milk yield without affecting BCS when fed in a low-forage diet.In partial agreement, Piantoni et al. (2015a) reported that the same FA supplement decreased BCS loss regardless of forage level, but particularly when fed in a low-forage/high-starch diet in fresh cows.However, when the FA was fed in the low-forage/high-starch diet during the fresh period, it decreased ECM yield by ~7 kg/d in early lactation, when all cows were fed a common diet.Overall, SFA supplementation improved DMI and increased baseline insulin during glucose and insulin tolerance tests (Piantoni et al., 2015b).The glucose tolerance test indicated that FA supplementation increased insulin secretion capacity, and the insulin tolerance test indicated that it decreased glucose clearance capacity, possibly indicating an increase in insulin resistance.We would expect higher insulin resistance to be related to higher milk production.However, the increase in plasma insulin concentration might have counteracted it, increasing nutrient partitioning toward body reserves.Results presented by Weiss and Pinos-Rodriguez (2009) and Piantoni et al. (2015a,b) clearly indicate that FA supplementation can interact with the diet being fed to affect energy partitioning and that more research in this area is warranted.

Protein
Unlike dietary carbohydrates and FA, protein has a clear nutritional requirement.Protein deficiency impairs milk production at any stage of lactation, and increasing the supply of MP and AA supplements will increase the production of milk, milk protein, or both.Although accurately determining the optimal MP and AA supply for a cow to reach her genetic potential might be difficult, the fact that cows respond to MP or supplementation of AA is without question (Moraes et al., 2018;NASEM, 2021).Dietary protein supply alters milk production for various reasons, the most obvious being that AA are the building blocks of not only milk proteins but also of enzymes and other proteins needed for milk synthesis; insufficient building blocks limits synthesis of the protein.In addition, some AA play an active role in regulating the activity of intracellular signaling proteins, such as the mechanistic target of rapamycin (mTOR; Arriola Apelo et al., 2014).Although dietary protein alters milk synthesis and the flow of nutrients to the mammary gland, there is little evidence that dietary protein alters whole-body coordinated changes in partitioning.Rather, increased protein supply may act at the level of the gland to increase milk protein and lactose synthesis (NASEM, 2021); increasing the synthetic capacity of lactose, in turn, would increase the pull of glucose and other non-amino nutrients from blood.As these nutrients are pulled to make more milk, tissue mobilization or decreased tissue gain could help meet the mammary demand, especially in the short term, but Moraes et al. (2018) found no evidence for this in cows after the fresh period.Instead, changes in milk energy output and changes in body tissue gain seem positively correlated.Liu et al. (2021) conducted several studies feeding high-or low-MP diets to cows in mid and late lactation and found that feeding low MP decreased both milk yield and body storage.Compared with high-MP diets, low-MP diets decreased ECM yield by 4.5 kg/d and BW gain by 0.33 kg/d in mid-lactation cows and ECM yield by 2.4 kg/d and BW gain by 0.46 kg/d in late-lactation cows.Thus, the proportion of the total response to increased MP in the diet that was BW was greater in late lactation, as BW gain is a higher priority as lactation or pregnancy progresses.In both stages of lactation, low dietary MP content decreased DMI, and the responses in milk yield and DMI could be observed within 2 d of diet changes, so the mechanism for the response was not clear.One complication in any study manipulating dietary protein is that added protein replaces another nutrient, often starch.In Liu et al. (2021), protein replaced both starch and fiber.
Although many assume that fresh cows fed more protein will produce more milk at the expense of body reserves, there is little published literature to support this idea.Brown and Allen (2013) fed cows high-or low-MP diets starting at calving; the high-MP diet was 18% CP and 22% starch and the low-MP diet was 14% CP and 30% starch.High MP increased ECM yield, concentrations of NEFA and BHB, liver triglyceride content, and loss of BCS, with no change in DMI.We suggest that the higher MP increased milk yield due to a direct effect on mammary gland intracellular regulators such as mTOR, but that responses related to the increase in partitioning of body reserves might have been caused by the decrease in dietary starch.Tebbe and Weiss (2021a,b) fed diets varying in MP and AA profile to fresh cows and showed that the high-MP diets increased milk yield, and when AA profile was balanced, they also increased DMI.Diets did not consistently affect BW, BCS, or empty body composition.However, multiparous cows fed more MP had less empty body protein loss as assessed by urea dilution.Whether the cows ate more because they produced more milk or vice versa was not clear as the DMI and milk responses occurred concurrently.In any case, there was no evidence for the level of MP or AA supplementation having an effect on partitioning per se.Several studies have examined the effects of protein by infusions of AA.Gallagher et al. (2022) infused branched-chain AA to fresh cows and observed a 6-kg increase in ECM within 1 wk of calving with an increase in blood ketones.However, they detected no difference in circulating NEFA concentrations, BW change, or DMI.Thus, increasing dietary protein may increase milk production in fresh cows but any changes in partitioning of nutrients is likely due to the pull of the mammary gland.In summary, feeding optimal concentrations of protein and AA, compared with deficient concentrations, increases milk production and feed intake but there is little evidence to suggest that dietary protein alters partitioning toward or away from body tissues.

MEASURES OF ENERGY PARTITIONING
No method to determine energy partitioning is perfect, and the use of several methods concurrently is recommended.Methods that could be used include energy balance using gas exchange, nitrogen balance, and changes in BCS, BW, rump fat thickness, and body composition as assessed by urea or deuterium oxide dilution.Measurement of circulating insulin and NEFA concentrations can also be helpful.Consistency in the measurement (e.g., time of day, evaluator, methods, equipment) is essential.Changes in the ratio of N-methyl histidine to creatinine in urine are considered to indicate loss of body protein but can be difficult to interpret (VandeHaar et al., 1999).
Two of the most common measures are changes in BW and BCS.For BCS, we suggest the use of 3 trained evaluators blinded to treatment with measurements to the nearest 0.25 unit on a scale of 1 to 5. For BW, consistency in time of day relative to feeding and milking and frequent calibration of scales are essential; measurement of BW for 3 d at the beginning and end of a 4-wk observation period is adequate to detect meaning-Piantoni and VandeHaar: RUMINANT NUTRITION SYMPOSIUM ful differences in BW for ~24 cows over a 4-wk period (Liu et al., 2021).For transition cow trials, changes over 2 wk prepartum in BW, BCS, and backfat thickness caused by diets could all be detected and were of similar direction (VandeHaar et al., 1999).We suggest that BW and BCS on the day of or after calving are essential to understand BCS or BW changes over time postpartum.New automated systems for measuring BW after each milking are promising.Body weight change should be corrected for change in intake using the formula: Δ EBW = Δ BW -5.2 × Δ DMI, which indicates that if cows ate 1 kg less, we would expect a decrease in BW of 5.2 kg (NASEM, 2021).
Rump fat thickness with ultrasound is considered more objective but requires consistency in probe pressure and placement.Nitrogen balance requires accurate measures of intake, fecal, urine, and milk N; these are not as easy to assess as we might hope.In a recent meta-analysis, Spanghero and Kowalski (2021) found that N retention in lactating cows is generally overestimated, likely because N losses in feces and urine are generally underestimated.When assessing relative changes in N balance, MUN might be a reasonable approach to estimating urinary N (Spek et al., 2013).Changes in body protein have little effect on energy but could have a big impact on BW gain because fat-free tissue is 22% protein, 6% ash, and 73% water.Changes in energy balance require measurement of gas exchange so are impractical in most situations.

CONCLUSIONS
Intricate mechanisms are in place to help the cow sustain a developing calf and, once the calf is born, produce milk and get pregnant again to start the cycle once more.The physiological state of cows determines whether energy is partitioned toward milk or body reserves, and changes over a lactation.Absorbed nutrients are not simply building blocks and fuels; they can alter hormonal signals, tissue responsiveness to hormones, and mammary metabolism to affect partitioning depending on physiological state.Shifting the rumen fermentation profile toward higher propionate production increases insulin and milk yield in highproducing, early-lactation cows, and increases body reserves as cows become more sensitive to insulin.On the other hand, increasing acetate production with the use of highly degradable forages or NFFS has the potential to maintain both milk energy output and body reserves.Fatty acid supplementation affects energy partitioning depending on FA profile of the supplement used and stage of lactation and parity of the cows fed, but dietary protein does not seem to affect it.Understanding the biology of these interactions can help nutritionists better formulate diets for cows at various stages of lactation.In addition, developing strategies to increase insulin sensitivity of extrahepatic tissues in the fresh period and insulin resistance after peak and later in lactation could increase production performance and health of cows.

Figure 2 .
Figure 2. Physiological state changes that cows undergo throughout lactation (i.e., fresh, early, and mid to late lactation) that determine whether energy is partitioned preferentially toward milk or body reserves.