Nutrition

I. CAUSES OF PROTEIN ENERGY WASTING (PEW) IN CKD PATIENTS. Metabolic and nutritional derangements are common in patients with chronic kidney disease (CKD), especially in patients on maintenance dialysis therapy (Ikizler, 2013). These derangements are called protein energy wasting (PEW) of CKD. Patients with this syndrome have increased rates of hospitalization and mortality (Kalantar-Zadeh, 2004). There are multiple etiologies of PEW (Table 31.1), including decreased nutrient intake; metabolic derangements such as metabolic acidosis, dialysis-associated catabolism, uremic toxins; and comorbid medical conditions such as diabetes mellitus and cardiovascular disease (Carrero, 2013). PEW affects approximately one-third of hemodialysis and peritoneal dialysis patients (Pupim, 2006). The sequelae of PEW in kidney disease are numerous and include malaise, fatigue, poor rehabilitation, impaired wound healing, increased susceptibility to infection, increased cardiovascular disease risk, and increased rates of hospitalization and mortality. In most instances, serum levels of inflammatory markers are increased, and numerous causes of chronic inflammation may be present (Kaysen, 2001). Proinflammatory cytokines can cause anorexia with suppression of nutrient intake (Kaizu, 2003). Chronic inflammation also is associated with cytokine-mediated hypermetabolism and resistance to anabolic actions of insulin leading to increased net protein catabolism (Siew, 2010). Disruption of the growth hormone (GH) and insulin-like growth factor-1 (IGF-1) axis leads to decreased protein synthesis. Increased leptin concentrations may worsen anorexia due to central effects.

A. Obesity. Concern has always focused on wasting in CKD patients, because the death rate increases steeply with evidence of skeletal muscle loss, weight that is below peer weight, or body mass index. There is, however, an increasing incidence of obesity among patients initiating maintenance dialysis therapy (Kramer, 2006). Although obesity has traditionally been defined in terms of body mass index, some dialysis patients who are normal or overweight by BMI have been found to be obese as defined by percent body fat (Gracia-Iguacel, 2013). Studies of the effect of obesity on survival in dialysis patients are difficult to interpret because of the observational nature of the studies, differences in analysis techniques and definitions of obesity, and confounding (Stenvinkel, 2013).

TABLE	Causes of Kidney Disease Wasting
31.1

Decreased nutritional intake

Overzealous dietary restrictions

Delayed gastric emptying and diarrhea

Intercurrent illnesses and hospitalizations

Decrease in food intake on hemodialysis days

Medications causing dyspepsia (phosphate binders, iron preparations)

Suppression of oral intake by peritoneal dialysate glucose load

Inadequate dialysis

Monetary restrictions

Inability to prepare or acquire food due to physical limitations

Poor dentition or severe gum disease

Neurologic disorders that impair eating/swallowing

Depression

Altered sense of taste

Increased losses

Gastrointestinal blood loss (100 mL blood = 14–17 g protein)

Intradialytic nitrogen losses (hemodialysis, 6–8 g amino acid per procedure; peritoneal dialysis, 8–10 g protein per day)

Severe proteinuria (>8–10 g/d)

Increase in protein catabolism

Intercurrent illnesses and hospitalizations

Other medical comorbidities, including diabetes mellitus, cardiovascular disease, infection

Metabolic acidosis (promotes protein catabolism)

Catabolism associated with hemodialysis (due to activation of proinflammatory cytokines)

Dysfunction of the growth hormone–insulin growth factor endocrine axis

Insulin resistance

Catabolic effects of other hormones (parathyroid hormone, cortisol, glucagon)

II. NUTRITIONAL ASSESSMENT

A. Patient interview and physical examination. Symptoms of nausea, vomiting, and anorexia, as well as recent changes in body weight, should be carefully evaluated to ascertain cause. Nonuremic causes of changes in weight and/or food intake must be kept in mind, including severe congestive heart failure, diabetes, various gastrointestinal diseases, and depression. Phosphate binders or oral iron preparations can cause dyspepsia and other gastrointestinal symptoms.

B. Assessment of food intake. Patient recall of food intake should be determined on both dialysis and nondialysis days, performed biannually (Kopple, 2001); intake on dialysis days typically is about 20% lower (Burrowes, 2003). Food frequency questionnaires may also provide useful information (Kalantar-Zadeh, 2002).

C. Nutritional screening tools: A variety of screening tools are available, such as the Malnutrition Universal Screening Tool (MUST), Mini Nutritional Assessment (MNA), and others. All these tools require a brief patient interview. Questions that are common to all screening tools include information about body weight changes within a given time frame, amount of oral intake, or lack of appetite. Because of its simplicity and reliability, the Malnutrition Screening Tool (MST), which is simpler to apply than the more comprehensive MUST, is suggested as the first option. The MST includes two questions about weight loss and one question about appetite. When scores for answers to these questions are added together and the result is greater than 2, the patient is at risk for malnutrition/PEW, and a nutritional assessment is recommended.

C. Nutritional assessment tools

1. Body composition

a. Body weight and body mass index. One should compare ideal or median standard weight (see Appendix B) with actual body weight. Comparison with prior values is important as both body weight and lean body mass decrease over time in hemodialysis patients (Di Filippo, 2006; Rocco, 2004). Although BMI is easy to calculate and is used in many nutritional guidelines, it should be emphasized that this metric is a poor estimate of fat mass and its distribution within the body, especially in patients with CKD.

b. Anthropometry. The waist–hip ratio (WHR) and skinfold thickness are superior to BMI for the correct classification of obesity in CKD in cross-sectional studies. Skinfold thickness measured at the biceps or triceps provides an estimate of body fat, whereas midarm circumference can be used to estimate muscle mass. These measures can be compared with reference ranges established in well-nourished dialysis patients (Chumlea, 2003). Patients with values below the 25th percentile for either middle upper arm circumference or triceps skinfold thickness are likely to be malnourished.

c. Bioimpedance. Bioimpedance analysis is based on the measurement of resistance and reactance when a constant alternating electrical current is applied to a patient. Empirical equations are used to predict total-body water from resistance and total-body mass from the ratio of resistance to reactance or from its geometrical derivative, the phase angle. Phase angle correlates strongly with anthropometric measures of nutritional status and with serum albumin levels. For reproducibility, bioimpedance measurements should be performed within 120 minutes of the end of a dialysis treatment (Di Iorio, 2004). Low phase angle measurements are associated with an increased risk of mortality (Mushnick, 2003). An international study using bioimpedance spectroscopy has noted an impairment in the lean tissue index in all dialysis patients, with better preservation in peritoneal dialysis patients versus hemodialysis patients (van Biesen, 2013).

d. Dual energy x-ray absorptiometry (DEXA). This test was developed to measure bone density, but was later adapted to quantify soft tissue composition, including fat and fat-free mass. A DEXA scan takes only 6–15 minutes, involves minimal radiation exposure, and hence can be used serially to follow changes over time. At present, DEXA is used mostly for research purposes; it is more costly, and there are no data relating DEXA results to outcome in patients with advanced kidney disease. DEXA findings must be evaluated with hydration status in mind as well.

2. Composite indices. Subjective global assessment (SGA) is a clinical method for evaluating nutritional status that includes history, symptoms, and physical parameters. The history component focuses on five areas: (a) percentage of body weight lost in the previous 6 months; (b) dietary nutrient intake; (c) the presence of anorexia, nausea, vomiting, diarrhea, or abdominal pain; (d) functional capacity; and (e) metabolic demands in view of underlying disease state. Physical parameters focus on assessment of subcutaneous fat; muscle wasting in the temporal area, deltoids, and quadriceps; the presence of ankle or sacral edema; and the presence of ascites. The SGA has a good reproducibility and correlates strongly with outcomes in end stage kidney disease (ESKD) patients (Duerksen, 2000). Other scoring systems that have been proposed include the modified SGA (Churchill, 1996), the Dialysis Malnutrition Score, and the Malnutrition Inflammation Score (Kalantar-Zadeh, 2001), all of which use a combination of objective and subjective factors. The Geriatric Nutritional Risk Index (GNRI) consists of only three objective parameters—body weight, height, and serum albumin level; the score is predictive of mortality (Kobayashi, 2010).

E. Laboratory tests

1. Serum albumin. Low levels are a strong predictor of mortality, and hospitalization risk rises dramatically and logarithmically as levels decline below 4.0 g/dL (40 g/L). The assay method used can change results by as much as 20%. Serum albumin levels correlate modestly with other nutritional measures, and hypoalbuminemia may be due to a low nutrient intake, protein losses, increased catabolism or some combination of these mechanisms. Additional assessment including, but not limited to, physical exam, dietary recalls and measurement of acute phase reactants (e.g., plasma C-reactive protein level) is necessary for appropriate management of the patient.

2. Predialysis serum urea nitrogen (SUN). The predialysis SUN level reflects the balance between urea generation and removal. Thus, a low SUN level could occur in a very well-dialyzed patient with good protein intake, or in an inadequately dialyzed patient with poor protein intake. Also, a low SUN may reflect substantial residual kidney function or a markedly anabolic state (such as during rapid recovery from an intercurrent illness). Therefore, it is difficult to infer the level of protein intake from the SUN directly.

3. Urea nitrogen appearance (g). This measurement can be used to estimate protein intake. This is because, in the absence of marked catabolism or anabolism, the urea nitrogen appearance rate reflects protein intake. In catabolic or anabolic patients, protein intake will be over- or underestimated, respectively. As discussed in Chapter 3, in hemodialysis patients, the g can be computed using a pre- and postdialysis SUN. In patients with AKI, g can be estimated by measuring SUN at two time points, usually 24 hours apart, after making an estimate of total body water. Another method used in computing g for both hemodialysis and peritoneal dialysis patients is to collect aliquots of the spent dialysate as well as urine, and to measure the amount of urea nitrogen in each.

4. Protein equivalent of total nitrogen appearance (PNA). Several formulas are available for the calculation of PNA from g, given that, on average, the percentage of nitrogen from protein that winds up as urea is known. Dialysis modeling programs usually normalize the PNA to the “kinetic” body weight; the latter is estimated as urea distribution volume divided by 0.58. The kinetic weight (which usually is an internal number and is not reported) is usually, but not always, close to the actual body weight. Dividing PNA by the kinetic weight gives a “normalized” PNA, or nPNA in units of g/kg per day.

5. Clinical utility of the nPNA. The utility of the PNA in terms of predicting outcomes has been questioned. In the HEMO trial, as well as in observational data sets, once serum albumin and creatinine were controlled for, the PNA had little, if any, additional predictive power in terms of outcome. In the HEMO trial, PNA was a very poor predictor of dietary protein intake. It was assumed that the dietary recall method used was not sufficiently sensitive to show a relationship, but alternative explanations are possible as well.

6. Other laboratory measures. Serum transferrin is low in almost all dialysis patients and is influenced by changes in iron stores, presence of inflammation, and changes in volume status; it is not a good indicator of nutritional status. Serum prealbumin levels may be elevated because of interaction of prealbumin with retinol-binding protein and decreased renal clearance. C-reactive protein (CRP) is an acute phase reactant that correlates negatively with albumin and other visceral protein concentrations. When serum levels of albumin or prealbumin are low, it is appropriate to check CRP levels to help uncover potential covert inflammation. CRP levels are highly variable in ESKD patients, reducing their practical utility, but serial CRP measurements may provide valuable information.

III. DIETARY REQUIREMENTS. The recommended average levels of nutritional intake are listed in Table 31.2, and include recommendations that are generally consistent with the National Kidney Foundation’s (NKF) Kidney Disease Outcome Quality Initiative (KDOQI) 2001 guidelines on nutrition (NKF, 2001) and the European best practice guidelines for nutrition (Dombros, 2005).

TABLE	Daily Dietary Recommendations For Dialysis Patients^a
31.2

Nutrient or Substance	Hemodialysis	Peritoneal Dialysis
Protein (g/kg)	>1.2	>1.2; >1.5 with peritonitis
Calories (sedentary, kcal/kg)	30–35^b	30–35^{b, c}
Protein (%)	15–25
Carbohydrate (%)	50–60^d	50–60^{c, d}
Fat (%)	25–35
Cholesterol	<200 mg (0.52 mmol)
Saturated fat (%)	<7
Crude fiber (g)	20–30
Sodium	80–100 mmol^e
Potassium	< 1 mmol/kg if elevated	Usually not an issue
Calcium	2.0 g (50 mmol)^f
Phosphorus	0.8–1.0 g (26–32 mmol)^g
Magnesium	0.2–0.3 g (8–12 mmol)
Iron	See Chapter 34
Vitamin A	None
β-carotene	None
Retinol	None
Thiamine (mg)	1.5
Riboflavin (mg)	1.7
Vitamin B6 (mg)	10
Vitamin B12 (mg)	0.006
Niacin (mg)	20
Folic acid (mg)	>1.0
Pantothenic acid (mg)	10
Biotin (mg)	0.3
Vitamin C (mg)	60–100
Vitamin E	None
Vitamin D	See Chapter 36
Vitamin K	See text

^aAll intakes calculated on the basis of normalized body weight (i.e., the average body weight of normal persons of the same age, height, and sex as the patient).

^b35 kcal/kg body weight per day if <60 years of age; 30–35 kcal/body weight per day if >60 years of age.

^cIncludes glucose absorbed from dialysis solutions.

^dCarbohydrate intake should be decreased in patients with hypertriglyceridemia.

^eLower sodium intakes, in the range of 1.0–1.5 g (43–65 mmol), may result in better control of blood pressure in peritoneal dialysis patients and a lower dialysis solution glucose load, and are recommended if this can be done while maintaining energy intake.

^fThe total dose of elemental calcium provided by the calcium-based phosphate binders should not exceed 1,500 mg (37 mmol) per day, and the total intake of elemental calcium (including dietary calcium) should not exceed 2,000 mg (50 mmol) per day.

^gFor patients with serum phosphorus level >5.5 mg/dL (1.8 mmol/L); use phosphate binders if elevated.

A. Need for individualization. A “renal” diet has numerous restrictions, and so adherence to such a diet can be difficult and stressful. Prescribed diets should be individualized to help accommodate each patient’s unique circumstances in terms of palatability, cost, comorbid medical conditions, and cultural eating habits. Specific nutritional issues in the diabetic dialysis patient are discussed in Chapter 32. Too many restrictions should be avoided as they may lead to poor intake. The nutritional recommendations need to be reinforced by all members of the health care team. Compliance should be assessed on a regular basis, even monthly at the initiation of dialysis or for those with a previous history of noncompliance.

B. Peer rather than actual body weight. One problem with dietary intake recommendations for dialysis patients, who often have PEW, is the choice of weight to use in the denominator. For example, if a patient has lost body mass such that his or her weight is now 50 kg versus a premorbid weight of 90 kg, ingestion of an “adequate” amount of protein or calories based on actual weight may maintain the patient at that lower body weight, but may not be optimal for regaining lost weight, assuming that this is desired. Protein and caloric recommendations should be based on the median standard (or “peer”) body weight (see Tables B.1 and B.2 in Appendix B) for healthy subjects of the same sex, height, age, and body frame size as the patient. On the other hand, in obese patients, adjusted body weight should be used (Adjusted BW = Peer Body Weight + 0.25 × (actual Body Weight – Peer Body Weight).