Nutritional Management

Intestinal Failure

Intestinal failure (IF) describes a state of insufficient intestinal capacity to fulfill nutritional demands, resulting in dependency on the use of PN, although the definition of IF has been revised by multiple sources. The European Society for Clinical Nutrition and Metabolism (ESPEN) is the first scientific society to issue a formal definition for intestinal failure. In the ESPEN definition, 2 criteria must be present: a “decreased absorption of macronutrients and/or water and electrolytes due to a loss of gut function” and the “need for parenteral support.” IF is further classified into Types I, II, and III. Type I IF is acute, short-term, and due to a self-limiting condition such as ileus following abdominal surgery, which may require a brief period of nutritional support. Type II IF results from a prolonged acute condition, often in metabolically unstable patients who require IV supplementation over periods of weeks or months, and may be reversible or irreversible. Type II IF patients may recover fully or progress to type III intestinal failure, which is a chronic state of IF requiring long-term nutritional support, typically in the form of home PN. Etiologies include Crohn disease (CD) (see Chapter 115 ), radiation enteritis (see Chapter 41 ), intestinal obstruction (see Chapter 123 ), dysmotility (see Chapter 99, Chapter 100 and 124 ), intestinal trauma, congenital disorders (see Chapter 98 ), intestinal fistulae (see Chapters 29 and 115 ), and vascular complications (see Chapter 118 ).

Short bowel syndrome (SBS) due to intestinal malabsorption associated with a functional small intestine length of less than 200 cm is a common cause of IF (see Chapter 106 ). After extensive intestinal resection, 3 clinical stages have been described. The first stage occurs during the first few weeks after resection and is characterized by significant fluid and electrolyte shifts that require copious amounts of IV fluids to prevent dehydration. During the second stage, which may last for up to 2 years, there is both structural adaptation (increase in size and absorptive surface as a result of cellular hyperplasia) and functional adaptation (slowing of bowel transit to allow increased time for absorption). The third stage is a stable phase during which no further improvement or adaptive changes occur. Nutritional management of SBS depends on the amount and location of small intestine removed, because the intestine has the ability to adapt and increase its absorptive function over time. Initially, PPIs are used to reduce gastric hypersecretion, and anticholinergic agents and opioids are used to slow intestinal transit. Patients may require larger doses of anticholinergics than are usually recommended, because absorption of the oral medication may be limited. PN is prescribed during stage 1 to meet nutritional needs.

During stage 2, oral feedings are gradually started, and the volume of PN is reduced as oral feedings are increasingly tolerated. Patients should eat small, frequent meals—avoiding simple sugars, fiber, and nutrient-poor foods—and separate the times of fluid and solid food ingestion. Lactose is usually well tolerated, unless the proximal jejunum is resected. Dietary intake should be increased by at least 50% because most stable adult SBS patients absorb only half to two thirds as many calories as normal. Such a hyperphagic diet is best tolerated when consumed as 5 to 6 meals throughout the day. Diarrhea may result from oral feedings and may limit weaning from PN. While percutaneous endoscopic gastrostomy (PEG) placement in the management of SBS is controversial, enteral tube feeding administered continuously over 12 to 24 hours is usually better tolerated than intermittent bolus feeding because of greater nutrient absorption and less osmotic diarrhea. EN is usually slowly advanced while PN is isocalorically decreased over several months, with frequent monitoring of tolerance as determined by the development of symptoms, amounts of food and fluid intake, stool and urine output, body weight, hydration status, and micronutrient levels.

Following extensive resection of the small intestine, 3 distinct clinical types of SBS can be identified. In type I SBS, patients have only jejunum remaining with an end jejunostomy and no colon. These patients experience massive fluid shifts, show little signs of adaptation over time, and are more likely to be PN dependent. In type II SBS, patients have variable length of jejunum connected in series with some portion of colon. Clinically they show greater signs of adaptation but demonstrate slow deterioration of nutritional status over time without parenteral support. Finally, in type III SBS, intestinal rehabilitation of the remaining small intestine is most likely to be successful (meaning the patient can resume intake of adequate oral nutrition) because the colon has been preserved and is in continuity with the small intestine and the ileocecal valve is maintained. Production of glucagon-like peptide (GLP)-1 by the remnant of terminal ileum has a trophic effect and stimulates SB adaptation, as a result of which these patients rarely need EN or PN. Clinical factors useful in predicting the success of intestinal rehabilitation include the presence of residual disease in the remnant bowel, bowel length, the degree to which adaptation has occurred, and the duration of time on PN. Intestinal autonomy is defined by the ability of a SBS patient to live without PN and may be expected if a patient has 70 to 90 cm of small bowel and an intact colon, or 130 to 150 cm of small bowel with no colon. Through the period of adaptation, intestinal autonomy may be achieved more readily for calories than for fluid and electrolytes. Citrulline, a non-protein amino acid produced by intestinal mucosa, has been proposed as a predictor of permanent versus transient IF. In one study, a plasma citrulline level of below 20 μmol/L identified patients destined to have permanent IF, with positive and negative predictive values of 95% and 86%, respectively.

Patients with severe SBS (<200 cm small bowel remaining) usually require a glucose-electrolyte oral rehydration solution (ORS). Ingestion of an ORS containing glucose with a sodium concentration of at least 90 mmol/L aids in water absorption by making use of sodium-glucose co-transporters in the jejunum (see Chapter 101 ). Between 2 and 3 L of an ORS solution should be sipped throughout the day. Hypo-osmolar fluids should be avoided, as their absorption is dependent mostly on passive diffusion. Hyperosmolar fluids also should be avoided in patients with SBS, because they lead to fluid shifts into the bowel lumen, worsening diarrhea. If a patient has had a partial ileal resection (resection of <100 cm) and has an intact colon, the bile-binding resin cholestyramine can be used to reduce bile salt-induced diarrhea. In patients with a limited amount of ileum remaining (>100 cm of ileum resected) and an intact colon, however, cholestyramine can increase diarrhea by depleting the bile salt pool. In general, fat restriction should not be used for SBS type I patients who do not have a colon, but it may be beneficial in reducing diarrhea for SBS types II and III, in which some length of colon remains. Even in these latter types, however, fat restriction may not change the volume of diarrhea and polyphagia is made more difficult without the ingestion of calorically-dense fatty foods. Vitamin B ₁₂ injections should be administered monthly if more than 50 to 60 cm of terminal ileum has been resected. The somatostatin analog octreotide has been shown to prolong small intestinal transit time and decrease GI secretions, but its use remains controversial because it is also associated with gallstone formation and decreased splanchnic protein synthesis, and has not been shown to eliminate the need for PN.

The use of growth hormone and glutamine to promote small intestinal mucosal hypertrophy and improve absorption is controversial. A single randomized controlled trial (RCT) demonstrated decreased PN volume, calories, and number of infusions with this approach, although the effects were short-lived, and measured parameters returned to baseline after cessation of therapy. Use of a GLP-2 analog has been shown to be a trophic stimulator of small intestinal mucosa, resulting in improved absorption. Teduglutide, a GLP-2 analog, was shown in a RCT to significantly reduce the volume and number of days of PN for patients with intestinal failure, and resulted in weaning from PN dependency in a small percentage of patients (see Chapter 106 ).

Pancreatitis

Nutritional therapy in the management of acute pancreatitis (AP) has undergone a significant paradigm shift from previous practices. Historically, patients with AP were kept NPO to avoid the potential risk of further stimulating exocrine pancreatic secretion and worsening inflammation. Over the past 2 decades there has been a shift toward early EN in AP patients, but also in all critically ill patients. Data on early feeding (initiated within 24 to 36 hours of admission) in AP have demonstrated lower risk of multi-organ failure (MOF), operative interventions, systemic infections, septic complications, and even mortality compared with what had been standard therapy (no EN/PN) or delayed EN. The PYTHON (Pancreatitis, Very Early Compared with Selective Delayed Start of Enteral Feeding) study, an RCT in which patients were allocated to receive EN within 24 hours through a nasojejunal (NJ) catheter or be given an on-demand oral diet over the first 4 days (and then only start EN if the oral diet were not tolerated), seemed to refute these results. Outcomes showed that rates of major infection or death were no different between groups, although only 20% of patients in this study were admitted to the ICU, and less than 8% had severe AP (defined by persistent MOF > 48 hours). The key issue in nutritional therapy is severity of the systemic inflammatory response syndrome (SIRS). If the SIRS response is severe enough to require admission to the ICU (especially if the patient is placed on mechanical ventilation), a NG/NJ tube should be placed and EN initiated within 24 to 36 hours of admission. If the SIRS response is minimal and the patient can be managed on the hospital wards, then oral diet should be offered as tolerated and EN considered only when there is failure to advance the diet after 4 days. Early initiation and advancement of EN should be performed with caution pending adequate resuscitation, as hemodynamically unstable patients requiring inotropic support may be at an increased risk of non-occlusive mesenteric ischemia (see Chapter 118 ).

Two meta-analyses comparing EN with PN in severe AP have shown a significant 2-fold reduction in the risk of total and pancreatic infectious complications, and a 2.5-fold reduction in the risk of death in patients receiving EN. In an American Gastroenterological Association (AGA) technical review of 12 RCTs, EN reduced the risk of infected peripancreatic necrosis (OR 0.28, 95% CI 0.15 to 0.51), single organ failure (OR 0.25, 95% CI 0.10 to 0.62), and MOF (OR 0.41, 95% CI 0.27 to 0.63) ²³ compared with PN. Stratifying patients with severe AP from those with mild-to-moderate disease helps triage patients needing admission to the ICU, adequate hydration, treatment for early organ failure, and provision of early EN. Moreover, 3 RCTs comparing gastric with jejunal feeding in severe AP showed no significant difference in the levels of infusion between groups with regard to tolerance or clinical outcome.

Patients with chronic pancreatitis (CP) often have weight loss associated with hypermetabolism, and their nutrient intake can be further compromised by abdominal pain, malabsorption, and diabetes. Jejunal feeding in such patients has been used to improve weight as well as to reduce abdominal pain, GI side effects, and narcotic use. Oxidative stress has been implicated in the pathophysiology of CP and antioxidant supplementation with selenium, ascorbic acid, β-carotene, α-tocopherol, and methionine has been shown to provide pain relief in patients with CP.

In a randomized, placebo-controlled trial of antioxidants used to treat the pain of CP, although therapy significantly raised blood levels of antioxidants, it did not have any beneficial effect on pain or quality of life of the 92 patients studied. These findings refute another RCT of 127 patients with CP who were treated with antioxidant supplementation that found 32% of patients became pain free in the antioxidant group compared with 13% in the placebo group. There was also a reduction in the number of painful days per month as well as the number of analgesic tablets used per month. It should be noted that in this study, the mean patient age was 30 years and only slightly over a quarter had alcohol as the etiology for CP. Taken altogether, it would seem that elderly patients with alcohol as the etiology for their CP are less likely to benefit from antioxidant therapy than younger patients with a non-alcoholic etiology. Patients with CP should consume small, frequent meals and avoid foods that are difficult to digest (e.g., legumes). Fat restriction is no longer recommended. In patients with weight loss, medium-chain triglycerides (MCTs) may be useful to provide extra calories without causing steatorrhea. MCTs, however, may be poorly tolerated because they are foul-tasting and can cause cramps, nausea, and diarrhea. Fat-soluble vitamins, vitamin B ₁₂ , and calcium should be replaced as clinically indicated.

Crohn Disease (CD)

CD (see Chapter 115 ) may be associated with malnutrition secondary to anorexia, malabsorption, increased intestinal losses, and the catabolic effects of systemic inflammation. Deficiencies of magnesium, selenium, potassium, zinc, iron, and vitamin B ₁₂ are common. Vitamin D deficiency is seen in about 50% of patients with CD. Besides a role in osteoporosis, vitamin D has also been implicated in the pathogenesis of CD, as it may down-regulate TNF-α–related genes. Although dietary therapy in IBD (with such commercial diets as the Colitis 5-Step, Atkins, South Beach, Specific Carbohydrate, Elimination, or the Maker’s Diet) has been proposed to help reduce symptoms, few well designed placebo-controlled studies exist. The use of EN, however, is an important component of IBD therapy for patients who cannot eat. RCTs in Asia comparing whole-day or half-day use of oral enteral formulas versus ad lib oral diets have shown significant improvement in IBD symptoms with EN. Although glucocorticoid therapy has been shown to be more effective than EN for inducing clinical remission of CD in adults, EN in children is just as effective as glucocorticoids in inducing clinical remission. An enteral formula with transforming growth factor-β is marketed specifically for IBD, but there is no robust evidence to recommend its use at this time.

Among hospitalized patients with active CD, no significant differences in improvement have been shown in patients randomized to receive EN or PN. Use of PN to achieve remission is problematic, as symptoms invariably recur once PN is stopped and an oral diet is resumed. The use of PN in IBD should be restricted to patients who have not responded to medications or in whom EN cannot be delivered. Bowel rest is not necessary to achieve remission in CD. Malnutrition is an independent risk factor for postoperative complications. The ESPEN 2017 guidelines suggest that CD patients who require emergency surgery should receive additional nutrition in the form of EN or PN to optimize postoperative outcomes. There are a paucity of high-quality studies, however, that specifically address the issue of nutritional support to prevent postoperative complications in CD patients. One meta-analysis showed that CD patients who received preoperative EN or PN were 74% less likely to have postoperative complications compared with standard of care without EN or PN (20.0% vs. 61.3%, respectively). A separate meta-analysis showed that in selected patients, preoperative PN resulted in improved nutritional status, fewer postoperative complications, and reduced disease severity.

Liver Disease

Malnutrition is common in advanced liver disease patients, with a prevalence of 50% to 90% in those with cirrhosis, depending on the methods used for nutritional assessment. Malnutrition leads to more complications (e.g., ascites, hepatorenal syndrome) and has been shown to be an independent predictor of survival. Among patients with cirrhosis and portal hypertension who are malnourished, the crude in-hospital mortality rate is 14.1%, compared with 7.5% in those who are not malnourished. Nutritional deficiencies among patients with liver disease result from many different factors acting in combination: malabsorption, altered metabolism, decreased nutrient storage, increased nutrient requirements, and decreased dietary intake resulting from anorexia, altered taste, and dietary restrictions. The etiology of anorexia is multifactorial and includes mechanical compression of the stomach by ascites as well as alterations in inflammatory and appetite mediators (e.g., increases in TNF-α and leptin). Patients with cirrhosis have been found to have dysgeusia, which can result from magnesium deficiency. Restriction of dietary sodium and protein to manage ascites and hepatic encephalopathy, respectively, can further lead to reduced food variety and poor oral intake. In addition, patients with alcoholic cirrhosis often substitute alcohol for nutrient-rich foods. Decreased bile salt production results in an intolerance to high-fat foods and the development of fat-soluble vitamin malabsorption, especially in patients with cholestatic liver disease. Nutrient absorption may be further compromised by hypoalbuminemia, which results in edema of the small intestine. Also, some patients with alcoholic cirrhosis have CP, with resultant maldigestion of protein and fats.

Portosystemic shunting can cause nutrients to bypass the liver, preventing them from being metabolized. Gluconeogenesis and protein catabolism are up-regulated, glycogenolysis is down-regulated, and insulin resistance occurs, leading to a depletion of muscle mass and fat mass because of their use as energy sources. Patients with cirrhosis have increased protein needs, and limiting their protein intake to prevent hepatic encephalopathy will only further accelerate protein-calorie malnutrition. It has been shown that diets with a normal protein intake are well tolerated and do not worsen hepatic encephalopathy (see Chapter 94 ). Patients should be fed according to their protein needs, and portosystemic encephalopathy treated with medications if it develops. Despite the high prevalence of insulin resistance among cirrhotic patients, carbohydrate restriction is not recommended as a method to prevent the hypoglycemia associated with impaired glycogen synthesis and hepatic stores. The branched-chain amino acids (BCAAs) valine, leucine, and isoleucine are used preferentially as a protein source by patients in liver failure because they are metabolized by the muscle, kidney, adipose, and brain tissue. In contrast, the aromatic amino acids (AAAs) tyrosine, phenylalanine, and methionine are metabolized and deaminated solely by the liver. Normal serum AA concentrations are altered in cirrhosis, with a rise in AAAs and a fall in BCAAs. We are not certain at this time whether AA concentrations are deranged in cirrhotics with or without portal hypertension, or in both. It is postulated that the rise in concentration of AAAs precipitates hepatic encephalopathy because they act as false neurotransmitters. In the past, BCAA supplementation has been shown to reduce hyperammonemia, because the metabolism of BCAAs by skeletal muscle supplies carbon skeletons for the formation of α-ketoglutarate, which combines with 2 ammonia molecules to become glutamine. Newer guidelines suggest that use of a formulation enriched in BCAAs should not be expected to improve patient outcomes compared with standard whole-protein formulations in critically ill patients with liver disease. Outpatient RCTs suggest that long-term nutrition supplementation with oral BCAA granules may be useful in slowing progression of hepatic disease and prolonging event-free survival. Patients with hepatic encephalopathy already receiving antibiotics and lactulose derive no further improvement in outcome (mental status or coma grade) by adding BCAAs to their therapy.

Micronutrient deficiencies can occur in patients with cirrhosis. Water-soluble vitamin (vitamin B complex and C) deficiencies can occur in both alcohol-associated and non-alcohol-associated liver disease. Thiamine deficiency can lead to Wernicke encephalopathy and Korsakoff dementia, not only in alcoholics, but also in patients with HCV-related cirrhosis. As a result, thiamine supplementation is recommended in all patients with cirrhosis. Decreased levels of folate and vitamin B ₆ have been reported in HCV infection, and it is noted that pegylated interferon and ribavirin therapy further decrease levels of thiamine, riboflavin, and vitamin B ₆ . Fat-soluble vitamin deficiencies occur more frequently with cholestatic than parenchymal liver disease. Vitamin A deficiency has been reported in cirrhosis and is considered a risk factor for cancer, including HCC. Vitamin D levels are low in patients with liver disease and fall as liver disease progresses, as a result of which there is a high prevalence of osteoporosis in both cholestatic and non-cholestatic liver diseases. Use of immunosuppressives, including glucocorticoids, as part of the treatment regimen for autoimmune hepatitis and following liver transplantation increases the risk for metabolic bone disease. Other risk factors for osteoporosis in patients with chronic liver disease include advanced age, low BMI, hypogonadism, estrogen deficiency, low calcium intake, excessive alcohol use, tobacco use, and physical inactivity. Vitamin E deficiency has been reported in both cholestatic and alcohol-associated liver disease, and low levels may facilitate progression of fatty liver to steatohepatitis. Zinc deficiency is also associated with liver disease, especially alcohol-associated liver disease, and may lead to anorexia, altered taste and smell, immune dysfunction, altered protein metabolism, hepatic encephalopathy, and impaired glucose tolerance. Zinc supplementation improves glucose metabolism, but data for improvement in hepatic encephalopathy are conflicting. Because copper and manganese are excreted into bile, these trace elements should be decreased or omitted from PN formulas in patients with cirrhosis or cholestatic liver disease.

EN is preferred over PN in patients with cirrhosis who require nutritional support, because liver function can worsen, and patients with ascites may not be able to tolerate the large fluid volumes associated with PN. Excess dextrose and glucose can lead to steatosis, and patients receiving long-term PN can develop cholestasis, fibrosis, and cirrhosis. There is an increased risk for catheter sepsis secondary to immune dysfunction and increased intestinal permeability. Compared with cirrhotic patients, those with chronic liver disease have better outcome with greater caloric intake. Lack of liver glycogen stores and reduced capacity for gluconeogenesis can lead to hypoglycemia and reductions in lean body mass during prolonged NPO periods. For these reasons, cirrhotic patients should not go more than 3 hours without eating, and a bedtime snack should be ingested. Patients with liver disease complicated by severe malnutrition have been shown to have more infections, longer ICU stays, and longer hospitalizations after liver transplantation, but there currently are no RCTs to show that preoperative nutritional support improves clinical outcomes of liver transplantation. Early postoperative EN has been shown to reduce the incidence of sepsis, and postoperative PN has been shown to reduce the length of ICU stay. Two randomized studies using a symbiotic (prebiotic and probiotic) regimen given enterally after liver transplantation have shown a reduction in bacterial infections.

Diverticular Disease

Patients with diverticular disease (see Chapter 121 ) are often provided with incorrect nutritional information. Patients are told to avoid nuts or foods that contain seeds because of fear that the hard, small particles may lodge in a diverticulum and precipitate diverticulitis, despite evidence showing no harm. A number of previous retrospective and epidemiologic studies have suggested benefit from fiber in preventing symptomatic diverticular disease, but no well-designed RCTs support this practice. Fiber intake should be at least 25 g/day and provided as insoluble fiber, such as that contained in wheat bran, bran muffins, and fiber-based cereals. The use of probiotics has had some success in treating and preventing diverticulitis. The most frequently investigated probiotics studied to date have been different strains of Lactobacilli. A systematic review of 11 studies on probiotics in the treatment of diverticular disease showed regression or reduction of symptoms in a majority of the 764 patients evaluated. However, high-quality data on prevention of complications and recurrence are scant and currently the AGA guidelines recommend against the use of probiotics after uncomplicated diverticulitis. Obesity and physical inactivity have been shown to increase risk of symptomatic diverticular disease in both men and women.

Dumping Syndrome

Dumping syndrome, which occurs when food passes too rapidly into the small intestine, can occur after gastrectomy, vagotomy, or esophageal surgery, and is increasingly being seen with the growing popularity of bariatric surgery (see Chapter 8 ). Two types of dumping syndrome are recognized. Early dumping syndrome occurs within 30 minutes of a meal and is characterized by abdominal pain, diarrhea, borborygmi, bloating, nausea, and vasomotor symptoms including flushing, sweating, tachycardia, hypotension, and syncope. This syndrome results from shifting of fluids out of the intravascular space and into the hyperosmolar environment of the duodenal lumen, as well as from enhanced release of GI hormones caused by a high carbohydrate load entering the small intestine. These hormones, including enteroglucagon, pancreatic polypeptide, peptide YY, vasoactive intestinal polypeptide, and neurotensin, have been implicated in the etiology of the vasomotor symptoms by causing systemic and splanchnic vasodilation. Late dumping syndrome occurs 1 to 3 hours after a meal and is characterized by hypoglycemia, sweating, hunger, fatigue, and syncope and is thought to be related to hypoglycemia from the rapid (earlier) increase in insulin via GLP-1 in response to the excessive carbohydrate load in the jejunum. Dumping syndrome is diagnosed by clinical assessment and a modified oral glucose tolerance test. Nutritional therapy for dumping syndrome is described later in this chapter.

Cancer

Protein-calorie malnutrition is a common problem for cancer patients. At presentation, the frequency of weight loss associated with cancer varies based on tumor type: 31% to 40% in patients with sarcomas, breast, and hematologic cancers; 54% to 64% in patients with colon, prostate, and lung cancers; and over 80% in patients with pancreatic and gastric cancers. Malnutrition is not only affected by the type of cancer, but also by the specific antitumor therapy regimen and patient characteristics (age, gender, and comorbidities such as diabetes and GI disorders). Malnutrition has been associated with reduced effectiveness of anti-cancer therapies, leading to longer duration of treatment and hospital stays, increased cost, and increased morbidity and mortality. The etiology of malnutrition in cancer patients is multifactorial and includes cancer cachexia (which is caused by tumor-induced metabolic abnormalities), impaired caloric intake, maldigestion, malabsorption, and GI toxicity from the cancer therapies themselves. Important mediators of cancer cachexia are thought to include proteolysis-inducing factor and lipid-mobilizing factor, which are produced by tumors, as well as alterations in the balance of neurohormones like neuropeptide Y (appetite stimulating) and pro-opiomelanocortin (anorexigenic stimulating). Proteolysis-inducing factor leads to decreased protein synthesis, increased protein degradation, and an increase in pro-inflammatory cytokines (interleukin-6 and interleukin-8). Lipid-mobilizing factor has been shown to increase lipolysis, leading to decreased body fat and weight, independent of caloric intake.

Most practitioners start a dietary regimen for cancer patients by modifying the amount of food and pattern of eating (i.e., small, frequent meals), using additional foods or supplements, or changing the formulation of the food (liquids, pureed foods, etc.). Appetite stimulation with glucocorticoids and megestrol acetate has been used successfully in cancer patients with mild malnutrition. Although both agents lead to improved appetite and weight gain, use of megestrol acetate is associated with a higher risk of DVT. Routine use of nutritional support in non-malnourished cancer patients who are undergoing chemotherapy, radiation, or surgery is not recommended unless they are unable to ingest or to absorb adequate nutrients for a prolonged period of time, in which case studies have shown an improvement in weight and nitrogen balance, but not survival. EN has been used successfully in patients with head and neck cancer to prevent weight loss, reduce hospitalizations, and reduce interruptions in chemotherapy and radiotherapy, although dependence on PEG feedings may delay return of swallowing function and PO intake. Routine use of PN during chemotherapy has not been shown to decrease toxicity, improve tumor response, or decrease mortality. PN has been found beneficial in patients who have developed severe GI mucositis after bone marrow or hematopoietic stem cell transplantation. In hematopoietic stem cell transplant patients, EN compared with PN is associated with increased morbidity, diarrhea, and hyperglycemia and delayed time to engraftment but less weight and body fat loss. Use of PN support in the cancer patient should be restricted to those patients with a reasonable life expectancy and a sufficient quality of life (Karnofsky Score >50), who are not expected to maintain their nutritional needs for a prolonged period (see Chapter 132 ).

Obesity

Obesity is the second leading cause of preventable death, due to its association with type 2 diabetes, hypertension, coronary artery disease, cerebrovascular disease, obstructive sleep apnea, cancer, osteoarthritis, and depression. Despite abundant adipose tissue, obese critically ill patients should receive early and timely nutrition therapy. Nutritional assessment of critically ill obese patients can be difficult. Current equations to estimate energy expenditure are invalid in this population, so indirect calorimetry remains the gold standard (see Chapter 5 ). In an obese patient with glucose intolerance or diabetes, the concentrated glucose solution in PN can lead to hyperglycemia, which in critically ill patients has been shown to increase the risk for nosocomial infection, weaken the immune response, delay wound healing, and increase overall mortality. During metabolic stress, protein breakdown leads to gluconeogenesis. Several studies have suggested that protein-rich hypocaloric feeding (2 g protein/kg ideal body weight [IBW]/day and 65% to 70% of caloric requirements), also known as permissive underfeeding, is advantageous over standard nutritional regimens because oxidation of endogenous lipid stores supplies the energy source while protein supplementation is used to promote protein anabolism; this approach, however, is still debated. With permissive underfeeding, hyperglycemia is seen less often, and weight loss may occur with maintenance of lean body mass. In critically ill obese patients, the American Society for Parenteral and Enteral Nutrition recommends 11 to 14 kcal/kg actual body weight per day for BMI 30 to 50 and 22 to 25 kcal/kg IBW/day for BMI greater than 50. The recommended dose of protein is 2 g/kg IBW for BMI of 30 to 40, and 2.5 g/kg IBW for BMI ≥40.

The United States Preventive Service Task Force recommends that all adults be screened for obesity, and those with a BMI of 30 kg/m ² or higher be offered a referral to intensive multicomponent behavioral interventions. Bariatric surgery is recommended for individuals with a BMI greater than 40 kg/m ² and for those with a BMI greater than 35 kg/m ² who have obesity-related comorbidities (see Chapter 8 ). Micronutrient deficiencies, particularly iron and vitamin D, are commonly present in obese patients and should be corrected preoperatively. After bariatric surgery, patients are sequentially advanced from a clear liquid diet to a solid diet, and postoperative nutritional guidance by a dedicated bariatric dietician is highly encouraged. In patients with Roux-en-Y gastric bypass, limitation in oral intake is necessary because of the small size of the gastric pouch. The shorter the length of the common channel in the Roux-en-Y gastric bypass, the more likely there will be micronutrient and macronutrient deficiencies. The laparoscopic adjustable gastric band, which is least likely to cause nutritional problems, is gradually being phased out because of its high rate of complications requiring removal in greater than 40% of cases. The vertical sleeve gastrectomy is being increasingly utilized because of its less disruptive effect on GI physiology and its effective impact on weight reduction and decreased risk of diabetes. Not only is understanding a patient’s postsurgical anatomy important, but also a basic knowledge of the various sites of nutrient absorption is essential to prevent and diagnose nutrient deficiencies ( Table 6.1 ). Post-bariatric surgery nutritional deficiencies can be divided into 3 types: protein-calorie malnutrition, vitamin and mineral deficiencies, and dehydration. Lifelong vitamin supplementation is started shortly after hospital discharge to prevent development of nutritional deficiencies, which can develop gradually and may take years to manifest.

Deterioration of nutritional status can occur post-operatively for a variety of reasons, including gastrojejunal anastomotic strictures and an excessively long segment of bypassed biliopancreatic limb. Micronutrient deficiencies can lead to anemia (iron, copper, zinc, folate, vitamins B12, A, E), metabolic bone disease (calcium, vitamin D), encephalopathy (thiamine), polyneuropathy and myopathy (thiamine, copper, vitamins B12 and E), visual disturbance (thiamine, vitamins A and E), and rash (zinc, essential fatty acids, vitamin A) (see Chapter 103 ). Iron, folate, calcium, and vitamin B12 deficiencies can occur after Roux-en-Y gastric bypass. After biliopancreatic diversion, zinc, sodium, chloride, magnesium, and fat-soluble vitamin deficiencies can occur. Dehydration is common after bariatric surgery, especially in warm weather and after vigorous exercise. The patient’s ability to drink large amounts of fluid is restricted because of the reduced size of the stomach. Approximately 2 L of fluid intake is usually recommended per day, though the amount can vary depending on the specific patient and his or her daily activity.

For those not meeting criteria to undergo bariatric surgery, pharmacotherapy is an option. According to the 2013 joint guidelines from the American College of Cardiology, the American Heart Association, and the Obesity Society for the management of overweight and obesity in adults, and the Endocrine Society’s clinical practice guidelines on the pharmacologic management of obesity, pharmacotherapy for obesity should be considered if patients have a BMI of ≥30 kg/m ² or a BMI of ≥27 kg/m ² with weight-related comorbidities, such as hypertension, dyslipidemia, type 2 diabetes, or obstructive sleep apnea. There are 6 main anti-obesity medications: phentermine, phentermine/topiramate, orlistat, lorcaserin, bupropion/naltrexone, and liraglutide (see Chapter 7 ).

Phentermine is the most commonly prescribed medication for obesity in the USA. It is a schedule IV controlled substance and classified as an adrenergic agonist that functions to increase resting energy expenditure and suppress appetite. As monotherapy, it is indicated for short-term use (3 months), as there are no long-term safety trials. When combined with topiramate ER, phentermine may be used long term. Orlistat promotes weight loss via inhibition of pancreatic and gastric lipases, thus preventing the absorption of fat. Orlistat also has the added benefit of lowering serum glucose and improving insulin sensitivity. Lorcaserin is a selective serotonin-2C receptor agonist for long-term treatment of obesity. The 5-hydroxytryptamine 2C receptor has a role in the regulation of the dopamine system, and is postulated to affect food-related behaviors. Two agents—bupropion (a dopamine and norepinephrine re-uptake inhibitor) and naltrexone (an opioid antagonist)—were FDA approved separately for the treatment of opioid dependence and alcohol use disorder. Together, naltrexone/bupropion affects the arcuate nucleus of the hypothalamus and the mesolimbic dopamine reward circuit, acting to regulate appetite and food cravings. Liraglutide was approved by the FDA for chronic weight management and treatment of type 2 diabetes. It mimics GLP-1, which is released in response to food intake, and acts to reduce hunger, decrease food intake, and delay gastric emptying.

Critical Illness

Determination of nutritional risk in critically ill patients is important, as it emphasizes the fact that risk from a nutritional standpoint is twofold, driven both by deteriorating nutritional status and disease severity. While patients with higher nutritional risk tend to have greater degrees of GI intolerance resulting in more difficulty in complying with the prescribed EN regimen, they are more likely to show improved outcomes from nutritional interventions than those with lower nutritional risk. American Society for Parenteral and Enteral Nutrition guidelines for nutritional therapy in the critically ill adult patient recommend that all patients admitted to the ICU undergo an initial nutritional risk screening. While the concept of nutritional risk is very important, use of the tools to determine such risk (the Nutritional Risk Screening 2002 or NRS-2002 and the Nutrition Risk in the Critically Ill or NUTRIC Score) is problematic. Patients determined to be at high nutritional risk (NRS-2002 >5 or NUTRIC score ≥5) should have EN started early within 24 to 36 hours of admission to the ICU. Advancement to goal should take 3 to 4 days in order to minimize the chance for overfeeding, as exogenous feeds are additive to endogenous gluconeogenesis by the liver; GI intolerance is monitored; electrolytes are scrutinized for evidence of refeeding hypophosphatemia; hypotensive patients on vasopressive agents are stabilized; and the process of autophagy is supported. Getting to the protein goal sooner (up to 2.0 g/kg/day) is more important than getting to the caloric goal (20 to 25 kcal/kg/day). Efforts to provide a goal of approximately 80% of estimated or calculated goal energy requirements should be made in order to achieve the clinical benefit of EN over the first week of hospitalization while avoiding risk of overfeeding. Although trophic feeds provide the non-nutritional benefit of feeding by preventing mucosal atrophy and maintaining intestinal integrity in low-to-moderate risk patients, they are insufficient to achieve the usual endpoints sought for EN therapy in high-risk patients. More than 50% to 60% of goal energy is required to prevent increases in intestinal permeability and systemic infection in burn and bone-marrow transplant patients, to promote faster return of cognitive function in head injury patients, and to reduce mortality in high-risk hospitalized patients. Likewise, a prospective study of high-risk surgery patients (NRS-2002 ≥5) who received sufficient preoperative nutrition therapy (>10 kcal/kg/day for 7 days) had significant reductions in nosocomial infections and overall complications compared with patients who received insufficient therapy.

Once EN has commenced, patients should be monitored daily for tolerance. Inappropriate interruption of EN should be avoided, with NPO orders surrounding the time of tests and procedures kept to a minimum to avoid propagation of ileus and prevention of inadequate nutrient delivery. Gastric residual volumes (GRVs) should not be used as part of routine care to monitor ICU patients on EN, as GRVs do not correlate with the incidence of pneumonia, regurgitation, or aspiration, and are a poor surrogate marker for gastric emptying. EN protocols should be designed and implemented to increase the overall percentage of goal calories provided, with consideration for a volume-based feeding protocol to clearly define the daily goal volume of EN to be infused. Appropriate adjustments to protocol should be made in the setting of hemodynamic compromise or instability. In this scenario, EN should be withheld if vasopressive therapy is being initiated. Initiation or re-initiation of EN may be considered with caution once the patient is fully resuscitated and stable for 24 to 36 hours and pressor support has begun to be withdrawn. Critically ill patients are at increased risk for subclinical ischemia/reperfusion injury to the intestine, but data show that ischemic bowel is a very rare complication of EN, and that patients on stable low doses of vasopressors or even multiple vasopressors had lower ICU and hospital mortalities when EN was used than when it was withheld.

In patients at low nutritional risk, PN should not be used until after the first 7 days following ICU admission if the patient cannot maintain volitional intake and early EN is not feasible. If EN is not feasible in patients at high nutritional risk, it is appropriate to initiate PN as soon as possible after hospital admission, once resuscitation has been performed. In any critically ill patient, regardless of risk, who is already on EN tube feeding but receiving less than 60% of the prescribed goal regimen, addition of supplemental PN should be withheld until after 7 to 10 days from admission. As tolerance to EN improves, the amount of PN energy should be reduced and finally discontinued when the patient is receiving more than 60% of target energy requirements from EN.

In all ICU patients who require PN, high protein hypocaloric PN should be considered initially over the first week, with provision of 80% of energy requirements (20 kcal/kg actual body weight/day). Compared with eucaloric PN, permissive underfeeding has been shown to reduce the incidence of hyperglycemia, infections, ICU and hospital lengths of stay, and duration of mechanical ventilation.

A combination of antioxidant vitamins (including vitamins E and C) and trace minerals (including selenium, zinc, and copper) should be given enterally or parenterally to all critically ill patients receiving nutrition support, as these may reduce mortality for patients with burns, trauma, and critical illness requiring mechanical ventilation.

Enteral Nutrition

EN supports both the structural and functional integrity of the GI tract. EN sustains structural integrity by maintaining mucosal mass and villus height, stimulating epithelial cell proliferation, promoting the production of brush border enzymes, and maintaining the secretory immunoglobulin (Ig)A-producing immunocytes, which make up the gut-associated lymphoid tissue (GALT). EN also maintains the functional integrity of the GI tract by maintaining tight junctions between the intraepithelial cells, sustaining a thick mucus layer, stimulating blood flow, and inducing the production and release of various trophic endogenous agents, including gastrin, CCK, bombesin, and bile salts. The provision of EN supports the role of the commensal microbiome and helps prevent the emergence of a virulent pathobiome in response to critical illness.

In those patients who will not or cannot eat because of some dysfunction of the GI tract, a feeding tube is necessary to provide feedings. The radiologist, gastroenterologist, or surgeon usually places these enteral access devices. This can be done at the bedside, fluoroscopically, endoscopically, or in the operating room, depending on the specific device and the expertise available.