Nutritional support to ARDS patients either shares metabolic concepts common to critically ill patients or generates particular concern due to the limitation of oxygen supply and the difficulties of carbon dioxide disposal. Metabolic response to injury and sepsis is characterized by the increased secretion of pro-inflammatory cytokines (IL-1, IL-6, and TNF) [1], catabolic hormones (cortisol, glucagon, and catecholamines) [2], as well as increased insulin resistance [3]. All these lead to glycogenolysis, increased gluconeogenesis and lipolysis, and enhanced muscle protein breakdown ensuring both energy and amino acids for wound repair and immune function.

Particularly, glucose is an indispensable substrate for host response in damaged tissues and immune cells. Endogenous glucose production from amino acids, lactate, and glycerol is necessary because glycogen stores are limited. Lipolysis is also necessary because of the energy required by gluconeogenesis. In fact this cycling of substrates costs in terms of ATP and is responsible in part of the increased energy expenditure. Amino acid from muscle protein is in part converted to glutamine and alanine and then to glucose and urea. This is mainly the fate of branched chain amino acids. Despite the increased protein synthesis of acute phase protein, the dissimilarity of different amino acids’ catabolism explains the large amount of wasted muscle tissue (500–1000 g per day) [4].

The stress response can only be reversed by resolution of the triggering event as infection control, wound healing, and so on. However, the rate of muscle tissue loss can be reduced by an adequate amount of energy and protein intake. This is important because body wasting impacts on tissue functions and hence on the recovery that can be delayed or even prevented by the loss of functional tissue.

For this reasons, metabolic support would comprise energy and protein/amino acids. Substrates can be administered either parenterally or by the enteral route. Parenteral nutrition can be given via the central or peripheral vein according to the osmolality; enteral tract can be accessed at the gastric or jejunal level either by nasal tubes or ostomies. Jejunal approach is preferred when the stomach fails to empty adequately. The relationship of gastric residual volume and diet tolerance is not completely understood; however, gastric residuals up to 500 mL can be tolerated, providing that in the end, the diet progresses to the jejunum [5]. Ostomies are indicated for long-term support.

Enteral nutrition is preferred because it maintains the absorptive function and the integrity of the intestinal barrier and leads to fewer complications, and also because it costs much less than parenteral nutrition. Enteral nutrition can be contraindicated (complete intestinal obstruction) or problematic (severe inflammation, high-output fistula) and sometimes is difficult to cover all the nutritional needs with the enteral administration. In these cases, it is necessary to substitute or integrate the enteral diet with parenteral nutrition in order to fulfill the requirements. The timing and the amount of this substitution/integration is a matter of debate [6, 7]; what is certain is that sooner or later, protein and energy deficits have negative effects on prognosis and that the negative effects are in some way proportional to the severity of the patient [8–10].

If some is good, more is not always better, and even if lack of energy and protein has negative effect, an excess of diet can be dangerous as well, both because of diet-induced thermogenesis (DIT) that furtherly increases energy expenditure and because excess substrates have to be stored or disposed giving way to waste product that can challenge organ functions of critically ill patients. For ARDS patients, energy excess can be particularly difficult to manage as will be explained ahead.

In the absence of the measurement of energy expenditure by indirect calorimetry, current guidelines suggest 20–30 kcal/kg/day [11–13]. The recommended protein/amino acid intake has increased in the last years up to 2 g/kg/day [14]. When nutritional support is administered enterally, the diet composition is predetermined. A standard diet is characterized by a ratio of 50:30:20 that represents the proportion of carbohydrates (CHO), fats, and protein in calories. Diets may differ for concentration (hypo- or hypercaloric) or composition. Differences in composition can depend upon relative quantity (high protein or fat) with respect to standard diet or upon quality of nutrients. Proteins’ quality regards the source (casein, whey protein, soy, egg) and the degree of hydrolysis (from whole protein to single amino acid passing through oligopeptides). Lipid quality is represented by the length of chain and by the content of polyunsaturated fatty acid with double bond on carbon w3 or w6. CHO generally consist of maltodextrins, and their content varies to allow the variation of protein or lipid content.

The differences for parenteral nutrition are more or less the same; however, protein content consists of synthetic amino acids and, allowing chemical stability, can be varied in many ways. Lipid content can be really different comprising w9 polyunsaturated fatty acid. CHO are represented by glucose that as we have seen before is an indispensable substrate in critical illness.

Apart from the concentration that allows different water loads, detailed description of the indication of different diets is beyond the purpose of this chapter; however, any difference in composition of interest for ARDS patients will be discussed later.

Whichever the source of CHO, they are mainly metabolized to glucose that is the primary source of easily available energy for all tissues. However, stress hyperglycemia may have deleterious effect on outcome. For this reason, glycemic control, after many years of conflicting studies, stands yet as a significant target in critically ill patients. The key questions are the appropriate target ranges in different patients’ population and the way to obtain it; both these goals should satisfy safety and feasibility. Glucose control is obtained by standardized i.v. insulin infusion. Protocols can either consist of intuitive rules or they can be paper-based or computerized algorithms; however, they generally do not take into account factors such as the degree of patient stress, i.e., insulin sensitivity, and the amount of nutritional support. Recently, a pathophysiology-based glucose control protocol which takes into account patient CHO load intake and patient insulin resistance has been proposed [15]. Computerization allows increasing complexity of the decisional tree, while diminishing the work effort; moreover, it performs better in terms of glucose control either as time in target or preventing hypoglycemia [16]. The more complex computer programs are able to react to or to anticipate glucose changes; future development in this field will probably be a closed-loop insulin infusion sustained by continuous glucose monitoring.

Critically ill patients receiving nutritional support may develop depletion of vitamins or trace elements; in some cases they can be already depleted. Usually, an adequate intake of calorie by the enteral route ensues the needs of microelements and vitamins; in case of inadequate intake or of parenteral support, trace elements and vitamins have to be supplemented.

Pharmakonutrition (i.e., nutritional support enriched with anti-inflammatory or immunomodulating substrates) is supported by many biochemical studies. It includes w3 polyunsaturated fatty acids, glutamine, arginine, and antioxidants. Larger clinical studies have demonstrated deleterious effect particularly of the amino acids glutamine and arginine as well as of antioxidants [17, 18]. These results have to be interpreted cautiously because of excessive dosage or due to different actions in different patient population (e.g., trauma vs septic shock patients). Particular effect of pharmakonutrients on ARDS patients will be discussed ahead.

ARDS is a heterogeneous syndrome of acute onset of respiratory insufficiency, characterized by increased pulmonary permeability and subsequent pulmonary edema. Patients with ARDS typically have altered lung and respiratory system elastance and an increased shunt fraction and dead space; as a result, gas exchange is severely impaired, with refractory hypoxemia and hypercapnia [19]. The cause underlying the development of ARDS is generally infectious (pneumonia, sepsis with secondary development of pulmonary lesions), or infections can arise as a complication of the first diagnosis (trauma, pancreatitis, visceral sepsis).

From a metabolic point of view, ARDS is characterized by a pro-inflammatory response associated with hypercatabolism that could lead to significant nutritional deficits [20], which can modify respiratory muscle function, ventilatory drive, and lung defense mechanisms [21–23].

In these critically ill patients, elevated resting energy expenditure (REE), reaching up to 126% of that predicted by their body size, means increased oxygen consumption (VO₂) and carbon dioxide production (VCO₂), i.e., increased cardiac work and increased respiratory exchange demand, in spite of impaired lung and sometimes cardiac function.

Similar to other critically ill patients, some days of starving or hypoalimentation lead to caloric deficit; this is even more relevant in ARDS. Nutrition support is then necessary to prevent cumulative caloric deficits, malnutrition, loss of lean body mass, and deterioration of respiratory muscle strength [13, 24–26].

Several aspects of energy consumption physiology and of substrate supply, disposal and transformation, oxidation, or storage, generally not considered in the absence of respiratory failure, may have a relevant impact and have to be carefully taken into account when feeding critically ill patients with severe respiratory failure.

The human body extracts the energy that is needed to comply to processes requiring mechanical (muscular contraction), chemical (biosynthesis), and osmotic (active transport) work by oxidizing a mixture of CHO, lipids, and protein to produce carbon dioxide (CO₂), water, and nitrogen. Part of the energy packed in substrates is in fact utilized to generate adenosine triphosphate (ATP), while the remainder is converted to heat used to maintain body temperature. The energy ingested in excess to the expenditure is inevitably stored [27, 28]. As only small amounts of energy can be derived by phosphorylation at substrate level (glycolysis), energy is mainly released from substrate oxidation, so that VO₂ and REE are closely correlated.

The goal of nutritional support remains to provide sufficient amount of calories to satisfy both the basal metabolic rate and the demands for the synthesis of protein to maintain lean body mass. National and international guidelines suggest administration of nutritional support to all patients suffering from ARDS and undergoing mechanical ventilation [11, 13]. An observational trial [37] and three RCTs [63–65] in critically ill mechanically ventilated patients tested the effect of energy prescription based on indirect calorimetry, and all reported clinical benefits.

However, all the abovementioned limitation needs to be considered when dealing with nutritional support in an ARDS scenario. As a result, adequate metabolic support of this category of patients is challenging. A first strategy is the reduction of REE that can be achieved in different ways.

At first, the reduction in the amount of physical activity (including the work of breathing), the control of body temperature, and the use of sedation are essential strategies.

Secondly, a daily amount of energy intake (EI) greater than needed need to be carefully avoided, as the impact of substrate disposal on REE (DIT increase) is certainly dangerous [27, 28, 66–68]. When circulatory function is suboptimal, this results in difficult heat dispersion, with hyperthermia [69], and in an increased ventilator and cardiocirculatory demand [22, 70]. Of interest, the increase in caloric provision per se, rather than excessive carbohydrate substrate, correlated more closely with VCO₂ [71]. Then, the recommended daily supply is 25 kcal/kg; however, when gas exchange is severely impaired, calorie load can be reduced to 20 kcal/kg, at least for the first phase of illness. In this scenario, be particularly aware of the likely development of an energy-protein deficit.

Third, the use of continuous feeding is protective. In fact, DIT is minimal during enteral feeding, both in health and disease [72, 73], provided that energy supply is less than 1–1.3 times the REE, while it increases to up to 20% of total EE when energy supply approaches two times the REE. This was interpreted as a lower energy cost of digestion and absorption of nutrients in comparison with the cost of nutrient storage. Indeed, oral administration of nutrients (i.e., as a bolus) induces a larger thermogenic response than continuous enteral administration [72, 73], due to the necessity of storing the ingested substrate until their utilization. During critical illness, moreover, VO₂ and VCO₂, and consequently DIT, are effectively blunted even with continuous TPN, provided that EI = REE [74].

Fourth, the composition of macronutrients can be altered. This can help minimize the requirements for mechanical ventilation, taking the pharmacological effect of nutrients into account and the possibility to manipulate DIT and VO₂/VCO₂. Indeed, macronutrient composition is much less likely to affect CO₂ production when the design of the nutrition support regimen approximates energy requirements [13, 67, 68, 71]. Then, the traditional suggestion that up to 50% of the nonprotein portion of caloric intake consists of lipid, to be able to limit VCO₂ and minute ventilation, should be abandoned. Instead, a more physiological composition (60% CHO/40% lipids) should be delivered, to allow the anabolic effects of insulin to take place. The use of EN formula can generally solve this problem, while careful attention should be paid to the nonprotein composition of the standardized commercially available PN formulas.

This said, a relevant issue is that of the type of lipids provided with artificial nutrition. Intravenous administration of high amounts of the w6 linoleic acid, even at slow rates of infusion, appears to be undesirable in patients with severe pulmonary failure [13, 75, 76], due to the pro-inflammatory effects and the interference with hypoxic pulmonary vasoconstriction.

To reduce this problem, several studies in patients with ARDS have shown that the use of w3 fatty acids, specifically EPA and GLA, possibly coupled with antioxidants, may confer biochemical and clinical advantage in ARDS by preventing oxidative cellular injury, modifying the metabolic stress response, and modulating immunity and inflammation [76–84]. Other authors [85, 86] and a recent meta-analysis [87] found that such an immunomodulatory diet did not show any significant reduction in the risk of all-cause mortality in ARDS patients, nor it increased the ventilator- and ICU-free days.

Actually, many of these studies have significant heterogeneity, and the majority share biases on the study design: the comparator diet provided in the control group often supplies a relatively high amount of w6 fatty acids, rather than the usual amount provided with commonly used balanced formulas, or they provide different amount of proteins among groups. As a consequence, the more-recently published guidelines on nutrition support [13] do not recommend the use of any enteral formula with an anti-inflammatory lipid profile and antioxidants in ARDS patients. Instead, the recommended strategy is the provision of an enteral/parenteral formula in which the w6 content is balanced by the presence of w3 and w9 long-chain fatty acids.

However, this strategy can often be vanished by the often underlooked issue of the energy contained in (w6) lipids used for sedation (i.e., propofol), which may represent up to 15% of the total energy administered and may lead to a profound unbalance in the composition of the diet [88]. The use of high-concentrated emulsion coupled with sedation strategies that use other compounds, such as enteral drugs [89], may give the advantage of reducing/eliminating this issue.

However, the sometimes necessary EI reduction during EN may result in a “caloric shortfall” and, even more dangerously, in a protein shortfall [37–40]. Bearing in mind that protein retention can improve by increasing either protein (main determinant) or even energy intake [90], if kidney function is normal, the protein shortfall can be limited by the use of a protein-strengthened enteral formula, yielding a supply of at least 0.8–1 g/kg. Still, under these circumstances, the early provision of adequate nutritional support is deeply relevant, so as to limit the unavoidable energy deficit. The role of enteral/parenteral glutamine supplementation is highly controversial [13], and in case of caloric shortfall, it should not be used at the expense of an optimal amount of proteins.