The milk production of dairy cows has been continuously increased during the last century and will probably be raised in the future. The increase is the result of the genetic selection for more milk, proper animal management and the provision of an adequate environment. Yields of 10.000 kg milk per lactation (305 d) or even more are not unusual under intensive production conditions. This development has been carefully outlined by Baumgard et al. [1] for the USA, in which milk yield (MY) has increased from ca. 2.500 kg in 1950 to 10.000 kg in 2016. Similarly, in Germany, a rise has been reported from 2.600 kg in 1950 to 8557 kg in 2022 as the mean for all breeds [2]. This rapid expansion in production “will be recognized as the “Golden Age” of lactation biology” [3].

Nevertheless, dairy cows experience a negative energy balance (NEB) postpartum (p.p.) which is compensated by the mobilization of some energy stores and their later replenishment during the lactation period. This cyclic change is normal [4], although excessive and long-lasting mobilization p.p. can lead to health problems, such as ketosis, the impairment of fertility or so-called “production diseases” [5]. A clear-cut transition from physiology to pathophysiology is difficult to discern.

The aim of the current overview is to present the hypothesis that increasing milk production is associated with an inadequate dry matter intake (DMI) leading to a rising NEB. The discrepancy between input (DMI) and output (MY) is covered by the mobilization, primarily, of energy (fat). Hence, postpartum MY, DMI, mobilization and NEB represent traits that guarantee the nutrition of the calf, on the one hand, even under tough environmental conditions, whereas increasing MY under intensive farming conditions is associated, on the other hand, with an inadequate DMI and, consequently, with an expansion of the extent and duration of NEB. The cascade behind these events will be described here and includes (a) DMI during transition; (b) milk yield and NEB; (c) mobilization, metabolites and hormones; (d) the pathogenesis of lipidosis as a precondition of ketosis and its association with “production diseases”; and, finally, (e) the possible genetic background.

2. Dry Matter Intake during Transition

During the last 15–10 d antepartum (a.p.), DMI continuously decreases and is 20–40% lower at the day of parturition. The decline is more pronounced in overconditioned cows [6], and DMI a.p. is genetically correlated with DMI p.p. [7]. The drop at parturition is slowly compensated by increasing DMI p.p. until a maximum intake is achieved between 8 and 12 weeks. MY grows much faster, with a peak at 6–8 weeks, resulting in a gap between the requirement for milk production and DMI with a subsequent NEB. These cyclic changes with a lag between milk production and DMI p.p. have been studied in the past with no satisfactory explanation [6] and are still poorly understood [8]. Knight [9] has summarized these uncertainties as follows: “Perhaps, the real conundrum is why so much effort is exerted by agricultural nutritionists and dairy farmers in trying to persuade the early-lactation cow to eat more. She knows better!” An explanation for this assumption “She knows better” has been suggested by Vernon and Pond [10]: “The inappetance around parturition is probably a throwback to the wild state when mothers would need to remain at the nest for a period of time and would be unable to feed”. Hence, the increase in milk production at insufficient DMI with the subsequent NEB has a probable biological and possible genetic background for ensuring the nutrition and survival of the calf. The amount of milk for the nutrition of one calf is limited, as is the NEB, and is a negligible challenge for the metabolism and health of the cow. This biological situation, namely, low MY and minor NEB, has been altered by the selection for higher milk production and frequent milking for maximal MY at a remaining inadequate DMI. The changed shape of the lactation curve has noticeably contributed to this situation, as Gravert [11] has stated: “The curve of feed intake (p.p. the author) still corresponds to the non-domesticated cow in according with the requirement of the suckling calf while the lactation curve has been altered by artificial selection for higher milk yields”. The possible drawbacks of the insufficient DMI have been known for decades. Balch [12] emphasized this problem in the introduction to a meeting about feed intake in dairy cows: “Feed intake regulation: A limiting factor in animal production (the dairy cow, the author)”. This shortcoming was expressed again in the 1980s: “…that selection of milk yield would not automatically increase feed intake of dairy cows in the first part of lactation” [13]. Subsequently, Arendonk et al. [14] pointed out: “Intake capacity might be considered as an additional trait in the selection goal to avoid an increase of negative energy balance in early lactation.” Some 23 years (!) later, von Leesen et al. [15] proposed an improvement in the breeding of dairy cows: “Thus, an energy balance (indicator) should be included in future breeding programs”, which again has been underlined by Rodehutscord and Titze [16]: “A remaining goal for reducing the energy deficit is a higher ranking of DMI in the breeding index.” However, the new German breeding index for Holstein cows does not include this trait, because the measurement of DMI is difficult and the corresponding data for genetic selection are not sufficiently available. This is, to the knowledge of the author, a worldwide problem.

3. Milk Yield and Negative Energy Balance

An increasing MY requires an augmented DMI because the requirement for maintenance is exceeded by a factor of 3–5 within a few weeks p.p. [17] and is accompanied by a heavy metabolic load, as expressed by Sheldon et al. [18]: “For a typical dairy cow producing 40 L of milk/d, the metabolic energy requirements for milk production are about 200 MJ/d, whereas only about 65 MJ/d is needed for maintenance. An equivalent metabolic demand for humans is running 3 marathons per day”. Hence, the dairy cow is a “marathon runner”, but over a time-span of weeks. The rapid change in metabolism and its metabolic load require comprehensive regulation, which has been named homeorhesis, in order to provide “orchestrated changes for priorities of a physiological state” [19], but it has the important consequence that milk production (“physiological state”) takes priority [19] over other physiological functions that are then possibly restricted ^{[19][20][21][22]}[19,20,21,22]. Obviously, feed intake does not belong to the priority of milk production. On the contrary, “feed intake during early lactation was negatively correlated with milk yield” [23], a finding later confirmed by Karacaören et al. [24] and by Liinamo et al. [25]. Similarly, Krattenmacher et al. [26] observed a moderate negative genetic correlation between MY and energy balance in early lactation, making understandable the early conclusion of Buttchereit et al. [27]: “that continued selection for high milk production will lead to a further increase in the postpartum energy deficit, unless energy balance is directly or indirectly included in the breeding programs with appropriate economic weights”. The production-dependent increase in NEB was demonstrated decades ago. Brand et al. [28] measured an NEB of 1445 MJ_NEL (the mean of all cows) with a duration of 11 weeks p.p. in the first lactation of German Holstein Friesian cows and correlated the NEB with MY: MY < 25 kg/d = 580 MJ_NEL NEB, 25–30 kg/d = 1323 MJ_NEL NEB and >30 kg/d = 1956 MJ_NEL NEB. This MY-dependent increase in NEB was confirmed by Friggens et al. [29] by a comparison of MY and the nadir of NEB in Jersey, Danish Red and Danish Holstein cows. The nadir (MJ/d) was augmented at higher MY: Danish Holstein > Danish Red > Jersey. Insufficient DMI p.p. was further demonstrated in an international co-operation (Austria, Germany, Switzerland) by Gruber et al. [30]. The increase in DMI p.p. was 0.1 kg DM/kg milk, clearly indicating an increasing deficit with increasing MY. As mentioned above, MY p.p. is partly uncoupled from DMI, a characteristic that has a biological and probable genetic background. The possible benefits are that the nutrition of the calf is independent of DMI and independent of the search for food by the cow, which can then concentrate her time on supervising her newborn calf. Even a shortage of food under tough environmental conditions does not challenge the nutrition of the calf. The deficit is compensated by the mobilization of reserves. Since the requirement (and the appetite) of the calf is limited, nutrition at the cost of the mother is rarely a challenge for the metabolism or health of the cow. Data from Hart et al. [31] support this assumption: British Hereford cows with milk production only for the nutrition of the calf exhibit a gain in bodyweight (BW) p.p., which nevertheless does not mean a lack of BW loss, because the increase in DMI and thus gut fill might compensate lost BW. However, the NEB in beef cattle (natural) is of minor importance in comparison with that of dairy cows, and the magnitude can only be estimated, with a range of perhaps 100–200 MJ. In contrast, Hart et al. (1974) detected a diminished BW in British Holstein dairy cows for ca. 10 weeks p.p. in their study [31]. Hence, the current NEB is a man-made problem in the modern high-genetic-merit dairy cow with its higher MY. The situation can be considered as “natural” in Hereford or beef cattle, but not in high-producing dairy cows. The genetic trait of milk production is uncoupled from sufficient DMI and, hence, the continuous selection for more milk has exacerbated the NEB now to 2000 MJ or more with a duration of 2–3 months ^[32][33][32,33], a situation that “exhibits characteristics of chronic under-nutrition” [20]. The 2000 MJ lost represent ca. one ΔBCS (5-point), 80 kg BW or 40–50 kg fat (39.7 MJ/kg) and ca. 10 mm fat thickness on the cow’s back [34]. An amount of 1 kg mobilized BW includes 64% fat, 8% protein and 28% water [35]. Hence, the current NEB in high-producing cows is far above the original (biological) loss of BW but is still often considered as “natural”, because weight loss in mammals after parturition is well known; this biological event is thus frequently used to declare the NEB of the dairy cow as “normal”. For example, Bauman [36] and Horst et al. [37] mention the heavy weight loss in bears, seals and whales p.p. However, these animals belong to species with “adaptive fasting” [38] and do not eat at all for weeks or months p.p. They totally rely on their reserves during the time spent raising their offspring. This strategy is obviously optimal for the nutrition of their young and can hardly be compared with that of the dairy cow. Moreover, the milk of the above species does not contain glucose [39] and no incidence of ketosis has been reported in these animals. Large amounts of glucose are required for the synthesis of lactose in cow’s milk [40], and glucose is without doubt the metabolic “eye of the needle” during early lactation. In dairy cows, glucose is mainly produced by gluconeogenesis, which can scarcely be maintained without an adequate DMI.

4. Mobilization of Reserves: Hormones and Metabolites

The rapid and massive changes in the metabolism in the cow pp. are accompanied by a plethora of alterations in various hormones, e.g., leptin, prolactin, ghrelin, resistine, corticoid, thyroxine, adiponectin and FGF21. Furthermore, an increase in growth hormone (GH) and a decrease in insulin (Ins) attended by insulin resistance (InsR) and reduced IGF-1 concentrations occur, which together are highly likely to determine the observed metabolic adaptation. GH and insulin accompanied by InsR and IGF-1 heavily influence metabolism p.p. High GH, low Ins and InsR, and low IGF-1 are the “orchestrated (hormonal) changes for priorities of a physiological state” [19], i.e., milk production p.p. during a period of NEB, and lead to enhanced lipolysis and the release of NEFAs for energy production (ATP) and milk fat synthesis, reduced lipogenesis, the stimulation of gluconeogenesis in the liver and the reduced uptake of glucose in muscle and fat tissue, with partitioning of the spared glucose for the synthesis of lactose in milk. The combined effects of GH and the GH-dependent InsR are under the feedback control of IGF-1 within the so-called GH-IGF-1 axis. GH, via the growth receptor 1A in the liver, stimulates IGF-1 synthesis and its release into the blood ^[41][44]. IGF-1 reduces the extrusion of GH from the pituitary gland, and this negative feedback finally controls catabolic reactions. Unfortunately, the fine-tuning signal cascade of the GH-IGF-1 axis can be uncoupled by the reduced expression of GHR-1A in the liver of dairy cows ^[42][45], although not in Angus beef cattle ^[43][46]. The degree of uncoupling of the GH-IGF-1 axis evidently depends on MY. The interruption of the GH-IGF-1 axis signal cascade is caused by the reduced expression of GHR-1A in the liver, resulting in decreased IGF-1 and, because of the lack of feedback, increasing GH concentrations in the blood. The decrease in GHR-1A expression is probably related to the low insulin concentration p.p., because the infusion of insulin + glucose p.p. increases its expression and “recouples” the GH-IGF-1 axis ^[44][43]. Thus, high-producing and “uncoupled” dairy cows p.p. are characterized by high concentrations of GH and low concentrations of insulin and IGF-1 in the blood. These alterations in hormones and metabolites and the possible health risks arising from them have been demonstrated in two publications ^[45][46][48,49], the data of which can be used as examples for the complex and voluminous adaptation and maladaptation that occur p.p.

5. NEFAs, Lipidosis and Diseases

The rapid increase in NEFAs p.p. is the result of enhanced lipolysis overwhelming the metabolic capacity of the cow ^[47][50] for the production of energy (ATP) and milk fat. The surplus of NEFAs causes the ectopic deposition of triglycerides in the muscle and liver ^[48][51]. Arshad and Santos ^[46][49] have demonstrated an almost linear correlation between NEFA concentration in the blood and lipidosis of the liver. One important direct effect of lipidosis is the positive and linear correlation with BHB blood values ^[46][49] that explains the risk of subclinical or clinical ketosis. These associations suggest a sequence of NEB, NEFAs, lipidosis and, finally, ketosis. Hence, lipidosis precedes ketosis or vice versa: ketosis p.p. is always connected with lipidosis. Furthermore, the degree of lipidosis is associated with various clinical diseases. A morbidity of 36.2%, even in the group with a physiological TGs (triglycerides) of 2.5%, is high and can hardly be accepted as normal. Remarkably, all listed diseases other than displaced abomasum exhibit an increased incidence with the degree of lipidosis. A causal correlation probably exists between TGs and subclinical ketosis, because TGs and BHB are correlated ^[46][49]. The association between TGs and the other diseases does not indicate a clear-cut causality and supports the conclusion of Sundrum ^[49][54]: “Metabolic disorders in the transition period indicate that the dairy cows’ ability to adapt is overstressed”, which is probably also the reason for the numerical decrease in survival ^[46][49] and agrees with the correlation of hyperketonemia and an increased culling rate ^[50][55]. The association between subclinical hypocalcemia (SCH) and TGs possibly has a metabolic (energetic) background. Cows with SCH (≤2.14 mmol/L) had higher NEFA and BHB concentrations ^[51][56], and cows with hypocalcemia (<1.0 mmol/L ionized Ca) and high NEFAs on the day of calving had significantly more lipids in their liver at day 7 ^[52][57]. “Hypocalcemic cows appeared to experience a more severe energy balance on the day of calving” ^[52][57], suggesting “that SCH increases lipomobilization” ^[51][56], which predisposes the cows to lipidosis. Indeed, cows with hypocalcemia (<2.00 mmol/L) had low insulin and high glucose concentrations, which “support the observation that physiological Ca concentration are required for glucose stimulation of insulin secretion” ^[53][58]. Experimental SCH in non-pregnant and non-lactating cows reduces DMI ^[54][59], possibly worsening the NEB p.p.

6. Conclusions

The increased MY during the last century is the result of genetic selection for higher production and of proper management with improved feeding and environment. Unfortunately, this progress has not been accompanied by adequate DMI p.p. to parallel the raised milk requirement. A growing body of evidence indicates that the decline in DMI around parturition has a genetic background with a biological function: securing the nutrition of the calf independently of DMI. Thus, dairy cows p.p. experience an NEB that is short and low at an MY based on the sole nutrition of one calf and results in no serious challenge for the health of the mother. Importantly, here, the amount of milk is limited by the appetite of the calf and, hence, the small NEB is “natural”. This limitation has been neglected during the selection for more milk and for frequent milking ^[55][129] with the goal of maximal milk production. The genetic correlation between MY and insufficient DMI causes the now long-lasting and deep NEB. The homeorhetic priority of MY and the resource allocation theory predicts restrictions of other (health) functions during NEB. Hence, homeorhesis for the “physiological state” [19] of MY is a two-edged sword. At low MY, it secures the nutrition of the calf uncoupled from DMI, as is the case in beef cattle and rarely with health problems. Unfortunately, the advantages of uncoupled DMI from MY at low production are turned, with rising MY, into a pronounced NEB with hormonal and metabolic alterations that challenge the health of the cow, because the prioritization of MY causes a subordination of health functions. The missing energy is covered by lipolysis and the release of NEFAs, but, critically, above the metabolic capacity of the cow, leading to the ectopic accumulation of TGs in the liver. Lipidosis is clearly correlated with subclinical and clinical ketosis and regularly precedes ketosis. Hence, the association of ketosis with “production diseases” probably represents an association between lipidosis and diseases. BHB is the “gold standard” of ketosis and mirrors this disease with its impaired function in gluconeogenesis ^[56][130], the detoxification of ammonia ^[57][131] and urea production ^[58][132], but is probably not the cause of the clinical signs of ketosis ^[59][133]. The cascade of events (“production diseases”) is practically unknown in beef cattle ^[60][134], with each cow only producing milk for the nutrition of one calf. This underlines the disadvantage of selecting dairy cows for higher MY without adequate DMI. This drawback has been realized and many efforts have been made to breed for higher DMI in early lactation [26] improving feed efficiency ^[61][135] and targeting BCS at parturition and limiting BCS loss p.p. ^[62][136]. The Gordian knot for the dairy cow is the discrepancy between MY and DMI at high MY with the health problems of subclinical and clinical ketosis and, most probably but indirectly, with the “production diseases”, including inflammation, oxidative and endoplasmatic stress, and immunosuppression. In ancient mythology, the problem of the Gordian knot was solved by its being cut with the sword of Alexander the Great. Such a “cut” for the health problems of the dairy cow remains out of sight, although a better understanding of the biological and genetic constitution and the inclusion of this knowledge in the breeding index could pave the way for sustainable milk production. More milk at inadequate DMI with a high incidence of diseases, early culling, increasing mortality and thus short but intense lifelong production will raise public concerns about production in agriculture and is a serious matter of animal welfare. A lower incidence of diseases and traits of energy metabolism (NEB, BCS, DMI and fat-to-protein ratio) should have high priority in the breeding indices during early lactation. Such a change is long overdue, because the deficits in the animal welfare of high-producing dairy cows have clearly been raised in the literature several times in the past ^[63][64][65][137,138,139].