You are on page 1of 7

Changes in Intermediary Metabolism in Severe Surgical Illness

Abstract. Under normal circumstances there is a reciprocal relation between the availability of free fatty acids (FFAs) and
glucose in plasma. In the fasted state, FFAs predominate in both availability and the relative contribution to energy
production, whereas the same is true for glucose in the fed state. The extent of glucose oxidation is directly determined
by its availability, whereas FFAs are normally available well in excess of their rate of oxidation. The rate of FFA oxidation
is determined by the rate of transfer into the mitochondria via the carnitine palmitoyltransferase (CPT) enzyme system,
which in turn is regulated by the metabolism of glucose. With critical illness the stress response involves mobilization of
both plasma glucose and FFAs simultaneously in both the fed and fasted states. In the situation of excess availability of
substrates, the metabolism of glucose limits the oxidation of FFAs, thereby channeling those fatty acids into triglyceride
(TG) stores in the muscle and the liver. The high FFA concentrations and increased tissue TG stores can limit glucose
clearance from the blood, thereby contributing to the development of hyperglycemia. Also, the excessive metabolism of
glucose can result in lacticacidemia and can contribute to the depletion of muscle glutamine. The nutritional treatment of
such patients must account for these underlying metabolic responses to avoid amplifying potentially detrimental
responses to the excess availability of substrates already present in the fasting state.

Hyperglycemia, hypertriglyceridemia, high lactate levels, and high free fatty acid (FFA) concentrations are indicators of
major changes in intermediary metabolism in critically ill patients. Although alterations in the normal concentrations of
these plasma indicators of intracellular metabolism in critical illness have been recognized for years, there is still only
limited information available regarding the exact nature of the metabolic response at the cellular level in human patients.
We review the normal regulation of substrate metabolism and, within the context of those control mechanisms, discuss
the response to critical illness insofar as possible.

Interorgan Substrate Kinetics

The availability of energy substrates plays a potentially important role in intermediary metabolism at the physiologic level.
Regardless of intracellular signals and enzyme activities, the rate of metabolism of a substrate is affected to at least some
extent by its availability. In this light, the generalized response to stress involves mobilization of energy substrates. Energy
substrate kinetics in the normal fasted and fed states are shown in Figure 1, and the alterations that occur in these
pathways during critical illness are shown in Figure 2. In the fasted state the plasma glucose concentration is maintained
relatively constant by regulated output of glucose from the liver. The liver glucose production arises from glycogen
breakdown, synthesis from recycled carbons (i.e., lactate and glycerol), and (to a much lesser extent) de novo synthesis
from amino acids such as alanine. Under normal conditions the peripheral metabolism of glucose almost exactly matches
the rate of appearance of glucose in the blood. Most glucose uptake is completely metabolized to CO2 and water, although
in some tissues (represented by muscle in Figure 1), glucose that is taken up is not only oxidized but also partially
metabolized via pyruvate to lactate via glycolysis. Lactate is in turn released into the blood and largely cleared by the liver,
when the carbons can be recycled into newly produced glucose.

Plasma FFA originates primarily from the hydrolysis of triglycerides in adipose tissue. In contrast to the situation with
glucose, FFAs are generally released into the plasma at a rate well in excess of the requirement for oxidation. Thus,
whereas some tissues (e.g., muscle) (Fig. 1) clear plasma FFA entirely for the purpose of oxidation, FFAs can also be cleared
by the liver. Once they have entered the hepatic fatty acid pool, some fatty acids are oxidized, and others are reesterified
into triglyceride (TG). Under normal conditions, newly esterified hepatic TG is packaged into very low density lipoprotein
(VLDL) and secreted into the blood, where the TG is transported to peripheral tissues, including muscle and adipose tissue.
The enzyme lipoprotein lipase hydrolyzes the circulating TG, enabling the cells to take up the fatty acids. In muscle, the
fatty acids can be subsequently oxidized or stored as TG. In adipose tissue, fatty acids are reesterified back into TG. In
addition, some of the circulating FFAs are also taken back up by the adipose tissue and reesterified. The release of fatty
acids from adipose tissue and the ultimate reesterification of those fatty acids is called the triglyceride–fatty acid substrate
cycle [1]. The glycerol- 3-P that forms the backbone of the newly formed triglyceride in adipose tissue arises from the
metabolism of plasma glucose, as it cannot be produced directly from glycerol in those cells. In the fasted state, FFA
predominates as the energy substrate, as the liver does not provide enough glucose to the blood to supply more than a
fraction of the energy substrate requirements.

In the fed state the balance of substrate metabolism changes. The ingested carbohydrate, which ultimately enters the
blood as glucose, now provides sufficient substrate to satisfy most energy requirements. An increase in glucose
concentration signals stimulation of insulin release, which in turn suppresses lipolysis and stimulates glucose uptake in
certain tissues, including muscle. The balance of substrate metabolism switches to a predominance of carbohydrate. The
increased glucose metabolism in muscle not only involves accelerated glucose oxidation but also increases lactate
production and release into blood. The liver and adipose tissue turn into organs of storage, with glycogen being deposited
in the liver and TG in adipose tissue. Although hepatic triglyceride stores may increase, there is an approximate balance
between the rate of TG formation in the liver and secretion of VLDL-TG [2].

In the fasted state FFA availability normally predominates, whereas in the fed state glucose availability predominates.
With critical illness the normal inverse relation between fatty acid and glucose availability is disrupted. The generalized
stress response involves mobilization of energy substrate such that both glucose and fatty acid kinetics mobilization are
enhanced simultaneously. In the fasted state, FFA release is increased to a greater extent than is normally the case, even
though overall energy requirements are not much different from free-living controls. This later point is due to the fact that
an elevation in resting energy expenditure above normal in critical illness is balanced by the inactivity due to bed rest [3].
Concurrent with the stimulation of lipolysis, hepatic glucose output is also elevated owing to increased rates of both
glycogenolysis and gluconeogenesis. The high rate of glucose uptake is associated with a rapid rate of release of lactate,
usually leading to an increase in the blood lactate concentration. The high rate of lipolysis causes plasma FFA
concentrations to increase. Because muscle FFA oxidation is not changed in proportion to the increase in plasma FFA
availability, hepatic clearance of fatty acids increases, leading to increased production of hepatic TG and secretion of VLDL-
TG. The rate of recycling of FFAs back to adipose tissue for storage is accelerated severalfold.

In the fed state the normal suppressive effects of ingested carbohydrate on hepatic glucose output are diminished, leading
to hyperglycemia [4]. The hyperglycemia during carbohydrate ingestión is further amplified by a relative inefficiency of
insulin in stimulating muscle glucose uptake [5]. The surfeit of energy substrates in the fed state is compounded by
persistent lipolysis, which not only causes plasma FFA concentrations to be high, but also results in increased flux and
concentration of VLDL-TG [6, 7]. An accelerated rate of de novo synthesis of fatty acids also contributes to the increased
hepatic TG production [8].

Alterations in intermediary metabolism do not arise because of an insufficient availability of energy substrates. Rather,
the observed changes at the cellular level are, at least in part, probably the consequence of the need to deal with the
excess supply of substrates. The consequences of the responses of intermediary metabolism needed to deal with the
excess substrates likely lead to many of the complications associated with critical illness.

Metabolic Basis for Lactic Acidosis

An elevation of lactate concentration is common in critically ill patients and may reach levels as high as those achieved
during strenuous exercise by a normal individual. Whereas it is posible that this reflects impaired tissue oxygenation,
lacticacidemia can persist even in patients with normal arterial PO2. The high rate of lactate production in this
circumstance does not reflect tissue hypoxia but, rather, excess production of pyruvate as a consequence of accelerated
glycolysis stemming from the increased glucose uptake and glycogen breakdown. This was demonstrated by the
observation that the infusion of dichloroacetate (DCA) lowers lactate concentration and production in critically ill patients
[9]. DCA stimulates pyruvate dehydrogenase (PDH), which is the enzyme responsible for pyruvate entering the TCA cycle
for complete oxidation. In severely burned patients, the complete oxidation of glucose is increased by DCA, indicating that
PDH activity is rate limiting for the oxidation of glucose. The high rate of lactate production thus does not reflect a specific
deficiency in the ability to oxidize glucose. In fact, stable isotopic studies of glucose and pyruvate oxidation indicate that
once in the cell, the proportion of pyruvate that is converted to acetyl-CoA via PDH and enters the tricarboxylic acid (TCA)
cycle for oxidation is not different from normal [9]. Glucose-6-P is produced within cells at an accelerated rate owing to
increased glucose uptake and increased glycogen breakdown. Glycolysis is correspondingly accelerated, and the high rate
of lactate production that results is the consequence of the high rate of pyruvate production, rather tan a deficiency in
the ability of pyruvate to enter the TCA cycle.

Relation of Altered Glucose Metabolism to Protein Catabolism

Accelerated catabolism of muscle protein is a universal problema in critically ill patients; it may impair recovery if severe
enough and certainly limits the return of patients to normal function after recovery. It has been proposed that this
catabolic response in part follows from accelerated oxidation of branched-chain amino acids (BCAAs), which in turn occurs
because of the inability of muscle to oxidize glucose completely. However, as pointed out above, there is not a marked
deficiency in the ability to oxidize glucose during critical illness. It is therefore not surprising that when DCA was used to
stimulate glucose oxidation in muscle there was no effect on the oxidation of leucine, a representative BCAA [10]. Rather,
a further increase in glucose oxidation was paralleled by a corresponding decrease in fatty acid oxidation. It is more likely
that increased BCAA oxidation is the consequence of activation of BCAA dehydrogenase resulting from the abundant
intracelular supply of BCAA due to the accelerated protein breakdown. Whereas the accelerated glycolysis and pyruvate
production likely have little direct effect on muscle protein kinetics, there may nonetheless be an effect mediated via
changes in intramuscular glutamine. Glutamine is the most abundant intramuscular free amino acid; and changes in
intramuscular glutamine concentration have been proposed to play a role in the regulation of muscle protein synthesis
and breakdown [11]. In critically ill patients the intramuscular concentration of glutamine may fall by as much as 80% to
90% [12]. Part of this fall is due to accelerated outward transport, as the intramuscular store of glutamine serves as a
reservoir to supply the immune system and gut with extra glutamine needed for stressful situations. Another aspect of
the decrease in glutamine concentration is that the de novo synthesis of glutamine is either decreased or increases
insufficiently to balance the accelerated efflux [12]. The limitation of glutamine production is linked to the accelerated
rate of pyruvate by the mechanisms shown in Figure 3.

Glutamate serves as the precursor for both glutamine and alanine. In the case of the formation of alanine, pyruvate is the
immediate precursor, such that glutamate is converted to a-ketoglutarate and pyruvate is transaminated to form alanine.
Alternatively, glutamate can be converted to glutamine by adding ammonia via the glutamine synthetase reaction. Under
a variety of circumstances the formation of alanine from glutamate is the preferred pathway. In any circumstance where
pyruvate production is accelerated, conversion of glutamate to alanine is also increased, often in the absence of any
change, or even a decrease, in the rate of glutamine production. For example, pyruvate formation is increased severalfold
during exercise. As a consequence, alanine formation and release by muscle is also increased severalfold, although there
is little change in the rate of glutamine production [13]. In the presence of critical illness the same basic response occurs,
except that the prolonged nature of the response, compared to that seen with exercise, results in depletion of the
availability of glutamate as a precursor for glutamine synthesis. The importance of these relations in critically ill patients
was demonstrated by the fact that decreasing the availability of pyruvate availability by infusion of DCA decreased alanine
production [9] and increased intramuscular glutamine concentration, presumably by increasing synthesis [14].

Intracellular Relations between Glucose and Fatty Acids

We focus this discussion primarily on the liver because few data are available for other tissues with regard to the response
to critical illness. Under normal conditions in the fasted state there is rapid uptake of FFAs by the liver and oxidation to
CO2 (Fig. 4A). Glucose is released from the liver into the blood. When plasma glucose concentration increases as a
consequence of carbohydrate intake, the liver changes from producing glucose to taking up glucose (Fig. 4B). As a
consequence, the production of pyruvate via glycolysis in the cytoplasm is stimulated. Cytosolic pyruvate can then enter
the mitochondria, where it has two major fates: (1) It can be decarboxylated to form acetyl coenzyme A (CoA) for entry
into the TCA cycle for oxidation; or (2) the pyruvate can enter the TCA cycle anapleurotically (via pyruvate carboxylase to
form oxaloacetate), and the resulting citrate can be shuttled back to the cytoplasm. In the cytoplasm, citrate can directly
activate acetyl CoA carboxylase (ACC), which is the enzyme responsible for the addition of CO2 to acetyl CoA to form
malonyl CoA, and be converted back to acetyl CoA to serve as substrate for ACC. Increased malonyl CoA, in turn, can bind
to the enzyme carnitine palmitoyl transferase I (CPT-I), thereby allosterically inhibiting CPT-I activity. CPT-I is responsible
for transporting long-chain fatty acids into the mitochondria for oxidation, so inhibition of CPT-I reduces FFA oxidation.
Increased malonyl CoA also leads to increased de novo fatty acid synthesis, but this pathway is minimal after a single meal.
The inhibition of fatty acid oxidation does not limit hepatic fatty acid uptake but, rather, channels nonoxidized fatty acids
to TG synthesis. Hypertriglyceridemia does not normally result, however, because the availability of FFAs from the plasma
is decreased by inhibition of peripheral lipolysis.

The central point of this perspective is counter to the traditional notion of glucose–fatty acid interactions, as first
articulated by Randle and associates in 1963 [15] when they described the so-called glucose–fatty acid cycle. Their
perspective was that fatty acid oxidation inhibited glucose oxidation, which is in contrast to our perspective that if glucose
is available it inhibits fatty acid oxidation (glucose–fatty acid cycle reversed [16]). This difference of perspective is not of
great practical significance under normal conditions because the reciprocal changes in the availability of glucose and fatty
acids dominate the nature of fuel selection at the cellular level. However, in insulin-resistant states, such as critical illness,
the nature of the intracellular relations between fatty acids and glucose become of great importance because both glucose
and FFAs are available in abundance in the plasma in both the fasted and the fed states (Fig. 5).

According to our perspective, in the fed state hyperglycemia due to peripheral insulin resistance leads to increased hepatic
glucose uptake and metabolism because hepatic glucose uptake is responsive to the circulating glucose, but not the
insulin, concentration. The glucose in turn leads to inhibition of CPT-I and fatty acid oxidation by the mechanism described
above. Because hepatic FFA uptake is elevated owing to the higher plasma concentrations, inhibition of oxidation leads to
more than normal synthesis of TG. Also, chronic hyperglycemia/hyperinsulinemia further activates ACC and fatty acid
synthetase. De novo fatty acid synthesis is thus stimulated, thereby also contributing to the increased cytoplasmic fatty
acyl CoA pool and thus TG synthesis. Hypertriglyceridemia results; and if the rate of hepatic synthesis of TG exceeds its
rate of secretion, accumulation of hepatic TG results.

In the fasted state during critical illness, the liver produces much more glucose than normal. In contrast to the normal
postabsorptive situation, the rate of pyruvate metabolism remains high because of an accelerated rate of lactate uptake
from the blood and conversion to pyruvate. The high lactate level in blood results from rapid peripheral uptake of glucose
and partial metabolism. Thus even in the fasted state ACC remains activated, and CPT-I is inhibited (Fig. 5A).

Certain aspects of the above scenario in insulin resistance are established. With critical illness, FFA flux is elevated to a
greater extent than fat oxidation, even during hyperglycemia/hyperinsulinemia, due to b-adrenergic stimulation [6];
peripheral insulin resistance leads to hyperglycemia, particularly during carbohydrate intake [5]; hepatic uptake of lactate
is accelerated in the fasting state [4]; and the plasma (and hepatic) TG concentration is elevated [7]. Furthermore, acyl
CoA synthetase activity was decreased in mitochondria and increased in microsomes in the livers of rats treated with
endotoxin [17]. Acyl CoA synthetase is needed to convert fatty acids to acyl CoA before being oxidized (in mitochondria)
or esterified (in cytosol). Thus the differential response of hepatic acyl CoA synthetase is consistent with the notion of
decreased fatty acid oxidation and increased fatty acid esterification. Finally, the rate of de novo synthesis of fatty acids,
representing activation of ACC and production of malonyl CoA as an intermediate, is accelerated in burn patients [8].
Other aspects of the scheme are predicted on the basis of studies in normal volunteers or animals, or on the results of in
vitro studies.

Effects of Accelerated Hepatic Production of Pyruvate: Increased Malonyl CoA and Inhibition of CPT-I Activity, Limiting
Fatty Acid Oxidation

The fact that accelerated hepatic production of pyruvate increases malonyl CoA and inhibits CPT-I activity, thereby limiting
fatty acid oxidation, is central to the proposed mechanism by which hepatic fatty acid oxidation is inhibited in critical
illness. It is based on two lines of reasoning.

First, it is based on the notion that hepatic pyruvate oxidation is accelerated with a burn injury. We have previously found
in patients with severe burns that during glucose intake and in the postabsorptive state the whole-body rate of pyruvate
production and oxidation (via acetyl CoA) are accelerated severalfold above normal [18]. Because the liver is capable of
oxidizing glucose extensively [19], it is reasonable to propose that the whole-body data reflect accelerated conversion of
pyruvate to acetyl CoA in the liver. In the fed state, pyruvate in the liver comes from glucose uptake. In the fasted state,
the rate of lactate uptake by the liver is increased [18], and the lactate is rapidly converted to pyruvate [20].

Second, this hypothesis is based on the notion that increased hepatic acetyl CoA production from pyruvate oxidation leads
to increased malonyl CoA and therefore inhibition of CPT-I activity. This concept is supported by the results of our own
study in normal volunteers in whom we demonstrated that the infusion of glucose and insulin inhibited the functional
activity of CPT-I across the splanchnic bed [21]. The evidence that the inhibition of CPT-I by glucose is via the malonyl CoA
system derives from animal experiments. For example, liver malonyl CoA concentration is high and ACC activity is
increased after refeeding in rats [22]. Also, glucose infusions during exercise prevented the normal decrease in liver
malonyl CoA [23]. Thus liver ACC and malonyl CoA are influenced by carbohydrate availability and the prevailing insulin
concentration.

There are metabolic pathways that can explain the relation between pyruvate metabolism and the intracellular malonyl
CoA concentration. Pyruvate that enters the mitochondria can be shuttled back into the cytoplasm as citrate and as
acetylcarnitine. In the cytosol, citrate lyase splits citrate into oxaloacetate and acetyl CoA. Thus an increase in pyruvate
due to an increase in glycolysis or lactate uptake (or both) increases cytosolic acetyl CoA levels [24]. This increase in
cytosolic acetyl CoA serves as a substrate for the enzyme ACC, which is activated by the concomitant increase in cytosolic
citrate [24]. In addition, during periods of high glucose availability the insulin level is elevated. Insulin is a potent activator
of liver ACC [25]. Thus when cytosolic levels of citrate and acetyl CoA are increased, ACC is maximally activated. This action
leads to the formation of malonyl CoA, which is the first step in lipid biosynthesis. Malonyl CoA inhibits CPT-I and thus
decreases the rate of long-chain fatty acids entering the mitochondria for oxidation [26, 27]. Given the metabolic profile
in critically ill patients, it is reasonable to extrapolate from the findings cited above and to propose that hepatic malonyl
CoA is increased and CPT-I activity inhibited as a consequence of accelerated production of pyruvate.

Hepatic FFA Uptake: A Function of Delivery, Irrespective of the Metabolic Fate of Fatty Acids in the Liver

We recently performed a study in normal volunteers in which the plasma FFA concentration was maintained constant,
and the response of the splanchnic bed to hyperglycemia/hyperinsulinemia was tested [28]. We found that although the
splanchnic oxidation of FFAs was inhibited by hyperglycemia/hyperinsulinemia, splanchnic FFA uptake remained constant.
These results are consistent with a large body of data dating back 30 years or more [29] showing a direct relation between
fatty acid concentration and uptake at the whole-body level.

Effect of Changes in FFA Availability on Fatty Acid Oxidation If CPT-I Activity Is Suppressed

In isolated hepatocytes, for any given CPT-I activity, an increase in fatty acid concentration stimulates fatty acid oxidation
to some extent [26]. However, when CPT-I is in the inhibited state, changes in FFA availability have a reduced effect on
fatty acid oxidation [26]. This point is pertinent to the schemes shown in Figures 4 and 5, because fatty acids that are taken
up by the liver but not oxidized are channeled into TG.

Fatty Acid–Glucose Interactions in Muscle

As in the liver, when plasma FFAs enter the myocyte, the first important step in their oxidation is the formation of fatty
acyl CoA. The enzymatic regulation of CPT-I then determines if the fatty acyl CoA is transferred into the mitochondria for
subsequent oxidation. Once inside the mitochondria, the fatty acyl CoA is then acted on by the enzymes of the b-oxidation
pathway to generate adenosine triphosphate (ATP). This sequence of events has three potential sites of regulation: (1)
transport of fatty acids into the cell; (2) enzymatic regulation, primarily of CPT-I; and (3) enzymatic regulation of b-
oxidation within the mitochondria.

Fatty Acid Transporters

Researchers have identified a family of plasma membrane proteins that bind FFAs in several tissues, including skeletal
muscle [30 –32]. Although there is a positive correlation between total plasma fatty acid concentration and total fatty acid
turnover, evidence in perfused rat skeletal muscle [32, 33] and exercising human muscle [34] reveals that unbound fatty
acid uptake is a saturable process, suggesting that FFA uptake by the myocyte may involve a carrier-mediated process.
Once inside the muscle cell, traditional thought has maintained that passive diffusion of FFAs occurs, but a family of
cytoplasmic fatty acid-binding proteins (FABPc) has been identified [35]. It is thus likely that fatty acid transporters and
binding proteins exist in human skeletal muscle, but their physiologic relevance has yet to be determined. Correlational
evidence that oxidative fibers contain large amounts of these proteins and that alterations occur in the amount of these
proteins during conditions of elevated FFA oxidation implies a physiologic function. On the other hand, there is
considerable circumstantial evidence that under resting conditions transporters play no regulatory role. For example, the
muscle uptake and oxidation of FFAs increases severalfold at the immediate onset of exercise, a response that would be
impossible if transport had been rate-limiting in the resting state. Also, an acutely increasing fatty acid concentration can
result in a corresponding increase in fatty acid uptake in a variety of physiologic settings [36, 37]. Thus it is likely that
factors inside the muscle cell regulate fatty acid oxidation rather than limitations in transport.

Enzymatic Regulation

It now seems likely that intermediary metabolism is regulated by similar mechanisms in muscle and liver. This was initially
thought not to be the case, as malonyl CoA plays a central role in the liver, and it is the first committed product in fatty
acid biosynthesis. It was therefore not expected to play a regulatory role in nonlipogenic tissues such as muscle. However,
malonyl CoA has been found in muscle tissue [38], and CPT-I is more sensitive to malonyl CoA in skeletal muscle than in
liver [38]. Consequently, it has been proposed that the same general mechanism for regulating intermediary metabolism
occurs in both muscle and liver [39]. Glucose availability and metabolism control the oxidation of fatty acyl CoA by
regulating CPT-I activity via changes in the malonyl CoA concentration.

Enzymatic Regulation within Mitochondria

The enzymatic regulation within mitochondria is an aspect of fatty acid metabolism control that has not been directly
assessed in critically ill patients; indications from other studies, however, are that it is not a likely point of control. For
example, in critically ill patients TCA cycle function appears to be unimpaired in the metabolism of acetyl CoA derived from
pyruvate [9], so there is little reason to believe there would be a problem regarding TCA cycle activity and the metabolism
of fatty acids. Furthermore, in other circumstances in which the oxidation of long-chain fatty acids are inhibited despite
constant FFA uptake, including lipid/ heparin infusion plus hyperglycemia [16] or plus strenuous exercise [40], the
oxidation of octanoate was not found to be altered. Octanoate is a short-chain fatty acid that enters the mitocondria by
diffusion, but within the mitochondria it is oxidized by the same b-oxidative pathway as long-chain fatty acids.
Consequently, as in liver, fatty acid oxidation in muscle is likely limited by transport into the mitochondria. Because the
rate of fatty acid uptake is largely determined by the rate of delivery, the limitation of fatty acid oxidation in muscle leads
to channeling of nonoxidized fatty acids into the intramuscular TG pool. It would be anticipated that the increase in the
intramuscular TG pool may contribute to the insulin-resistant state [41].

Consequences of Alterations in Intermediary Metabolism

The most prominent clinical manifestations of the alterations in substrate metabolism in critical illness are hyperglycemia
and fat accumulation in the liver. Direct evidence regarding the importance of these responses is lacking, but there is
reason to believe that complications can ensue from both responses. It has long been recognized that hyperglycemia per
se impairs wound healing by promoting the growth of infections [e.g., 42]. This is consistent with unfavorable clinical
results when patients are allowed to be markedly hyperglycemic [e.g., 43]. It seems likely that controlling hyperglycemia
by administering insulin can effectively overcome some of these potentially detrimental effects [44], but this approach
could potential increase hepatic TG stores. The clinical significance of fat accumulation in the liver (hepatic steatosis) is
less clear, as hepatic steatosis occurs naturally in about 3% of the population without symptomatology. On the other hand,
it is well known that livers infiltrated with fat provide a poor outcome in liver transportation. Furthermore, hepatic
steatosis in rats increases mortality and the risk of liver injury after administration of endotoxin [45, 46], perhaps due to
inability of the liver to process bacterial toxins and other mediators [45]. Thus hepatic steatosis is a risk factor for mortality
and progressive liver disease, but disease progression requires another insult that promotes inflammation [46]. A
deficiency in the ability of the liver to clear and detoxify bacterial toxins and other mediators would not necessarily be
reflected in traditional liver function tests and may not become apparent unless another stress (i.e., sepsis) is present.
Consequently, it is reasonable to assume that the accumulation of TG in the liver is detrimental to the recovery process.

Clinical Implications of Changes in Intermediary Metabolism

From the above discussion it should be clear that whereas the schemes in Figures 4 and 5 are based on experimental
evidence all aspects have not been directly confirmed in critically ill patients. Consequently, this presentation should be
considered a plausible hypothesis, rather than a well documented scheme. Nonetheless, two important points can be
made from the detailed information provided above. (1) The physiologic stress response involves excessive mobilization
of energy substrates; and (2) the primary complications of clinical significance are a consequence of this excessive
availability. It then follows that the protein catabolic response is not likely due to a deficiency in energy substrates, and in
fact limited available data support this notion. Consequently, administration of nonprotein calories to critically patients
must be undertaken with caution. It is in fact possible that the only beneficial aspect of nonprotein calories in this
circumstance is the role of exogenous glucose in stimulating insulin release, which itself has a pronounced anabolic effect,
particularly in conjunction with an exogenous supply of ample amino acids [44]. Alternative substrates, such as medium-
chain fatty acids [47], dicarboxylic acids [48, 49], and triacetin [50], might prove useful because they circumvent the need
for CPT-I, thereby being efficiently oxidized. Because there is little evidence that critically ill patients have diminished
ability to oxidize the substrates already available, it is unclear if these “alternative” substrates provide a unique advantage.
Thus the primary clinical message of the comments herein is that the pathways by which the detrimental responses of
hyperglycemia and hypertriglyceridemia occur should be kept in mind when selecting the nature and quantity of
nutritional support. In that light, it is clear that excessive intake of fat or carbohydrate would likely have detrimental
effects.

Acknowledgments

This study was supported by NIH grants DK34817 and DK38010 and Shriners Hospitals for Children 8490.