Articles on Medical Diseases and Conditions

Blog

  • Gluten-Induced Enteropathy

    Gluten-induced enteropathy includes sprue and celiac disease (childhood nontropical sprue), both diseases that affect the small intestine (predominantly the duodenum and jejunum). Both conditions are caused by immune reaction mediated by T-lymphocytes against gluten, a mixture of proteins found in wheat, rye, barley, and possibly oats. Celiac disease is found predominantly in Europeans, uncommonly in African Americans, and rarely in Asians. It is about 10 to 15 times more common in IgA-deficient persons and to a lesser extent (1%-3%) in patients with type I (insulin-dependent) diabetes mellitus. There is also an association with GI tract T-cell lymphoma and juvenile rheumatoid arthritis. There is increased incidence of the class II histocompatibility complex antigens (HLAs) HLA-DRw3 (about 80%-90% of Northern European-descent patients) and HLA-B8 to a lesser extent. HLA-DR7 is often seen in Southern Europeans. Some feel that HLA-DRw2 is even more important in all affected Europeans. Adults tend to have overt diarrhea more often than children with celiac disease; children are more likely to have anemia, nonspecific chronic illness, or short stature. A substantial number of patients have subclinical or mild disease. In those who are symptomatic there is often some degree of carbohydrate and fat malabsorption that may be accompanied by vitamin B12 and folic acid malabsorption if the ileum is affected.

    Small intestinal mucosa typically first shows blunting (mild widening and shortening) of the mucosal villi and then flattening and loss of the villi with infiltration by lymphocytes. Mucosal biopsy (particularly the proximal jejunum) is still considered the gold standard for diagnosis. About 30%-50% of adult patients (not children) with nontropical sprue are reported to have some evidence of splenic atrophy, which may be seen as RBC Howell-Jolly bodies and thrombocytosis. Gliadin is the toxic protein component of gluten. Antigliadin antibodies (AGA) are found in 95% or more of patients with nontropical sprue or celiac disease. AGA-IgG are present in about 96%-97% (range, 91%-100%) of untreated patients, but only about 80% (range, 58%-95%) have celiac disease. In contrast, AGA-IgA is found in about 75% of patients (range, 42%-100%) but is about 95% specific for celiac disease (range, 80%-100%). One or both AGA are elevated in about 45%-85% of patients with dermatitis herpetiformis, which is a bullous skin disease also triggered by gluten sensitivity. False positive results in either AGA-IgA or IgA are most commonly due to ulcerative colitis or Crohn’s disease.

    Antireticulum antibodies are reported to be elevated in about 50%-60% (range, 28%-97%) of patients, but specificity is said to be about 98% (range, 97%-100%). Endomysial antibodies (EMyA, reacting against the endomysial component of smooth muscle) for unknown reasons is elevated in 90%-95% of celiac disease patients (range, 85%-100%), with specificity of nearly 100%. However, both antireticulum and antiendomysial antibodies have been introduced relatively recently and have been evaluated less frequently than antigliaden antibodies. All of these antibody evaluations have been performed using homemade antibodies, which always differ somewhat and therefore make interpretation of reference laboratory results more difficult. In general, investigators seem to be currently screening patients using AGA-IgG, with ARA or EMyA either concurrently or as confirmation.

    D-xylose absorption testing is said to be positive in about 80% (range, 43%-92%) of cases, but specificity for celiac disease is only about 50%. Another nonspecific absorption test recently advocated for screening purposes is the “differential sugar test,” based on poor absorption of smaller nonmetabolized carbohydrate molecules relative to larger molecules in patients with malabsorption due to small intestine mucosal dysfunction. Although several nonmetabolized sugars have been used, the most current ones are mannitol (small molecule) and lactulose (large molecule). A lactulose/mannitol urine excretion ratio greater than 0.10 was considered abnormal. Sensitivity in celiac disease was said to be 89% with specificity for celiac disease of 54%-100% (depending on the control group).

    Workup for possible malabsorption

    This section has considered certain tests for intestinal malabsorption. In the opinion of many gastroenterologists, the most useful tests in initial screening for malabsorption syndromes are the D-xylose, plasma carotene, and qualitative fecal fat. If the results of all these tests are normal, the chances of demonstrating significant degrees of steatorrhea by other means are low. If one or more results are abnormal and confirmation is desirable, it may be necessary to perform a 72-hour fecal fat study.

    Once steatorrhea is strongly suspected or established, many investigators proceed to a D-xylose test. If the test is abnormal, many will attempt a small intestine mucosal biopsy, both to confirm a diagnosis of sprue and to rule out certain other conditions such as Whipple’s disease. Some prefer a trial period of oral antibiotics to eliminate the possibility of bacterial overgrowth syndrome and then repeat the D-xylose test before proceeding to biopsy. In some centers a biopsy is done without a D-xylose test. If gluten-induced enteropathy is suspected, antigliadin (or possibly antiendomysial) antibody assay could be useful as a screening procedure. If results of the small intestine biopsy are normal, possibilities other than sprue are investigated, such as pancreatic enzyme deficiency and bile salt abnormalities. A trial of pancreatic extract or an endoscopic retrograde cholangiopancreatography examination to investigate pancreatic or common bile duct status might be considered in these patients. Small intestine x-ray series may demonstrate some of the secondary causes of steatorrhea, such as lymphoma, diverticula, blind loops, and regional enteritis. However, barium may interfere with certain tests, such as stool collection for Giardia lamblia.

    If small intestine biopsy is not available, a therapeutic trial of a gluten-free diet could be attempted. However, this requires considerable time and patient cooperation, and adults with nontropical sprue may respond slowly.

    If pernicious anemia is suspected, a Schilling test is the best diagnostic procedure. Bone marrow aspiration could be performed prior to the Schilling test to demonstrate megaloblastic changes, which, in turn, would be useful mainly to suggest folate deficiency if the Schilling test result is normal. It may also reveal coexistent iron deficiency. More widely used than bone marrow aspiration are serum B12 and folic acid (folate) assay.

  • Malabsorption

    The function of the gastrointestinal (GI) tract is to perform certain mechanical and enzymatic procedures on food to prepare food for absorption, to absorb necessary dietary constituents into the bloodstream, and to excrete whatever is not absorbed. When the usual dietary constituents are not absorbed normally, symptoms may develop that form part of the syndrome known as malabsorption. There are three basic types of malabsorption. The first type involves the interruption of one of the stages in fat absorption; this concerns primarily fat absorption and also those substances dependent on the presence of lipid. The second type is related to intrinsic defect of the small bowel mucosa (category IV, A–D). In this kind of malabsorption there is interference not only with fat and fat-soluble substances but also with absorption of carbohydrates and many other materials. The third type of malabsorption is associated with altered bacterial flora (category IV, E) and also with the deficiency disease called “pernicious anemia.” In malabsorption of this kind, lack of a single specific substance normally produced by the GI tract leads to malabsorption of other substances dependent on that substance for absorption. Some of these conditions and their laboratory diagnosis are discussed in detail elsewhere.

    Clinical findings. Steatorrhea, or the appearance of excess quantities of fat in the stool, is a frequent manifestation of most malabsorption syndromes. Many patients with steatorrhea also have diarrhea, but the two are not synonymous; a patient can have steatorrhea without diarrhea. On the other hand, some reports indicate that moderate or severe diarrhea can induce some degree of steatorrhea in about 20%-50% of patients. In children, the principal diseases associated with steatorrhea and malabsorption are celiac disease and cystic fibrosis of the pancreas. In adults, the most common causes are tropical sprue, nontropical sprue (the adult form of celiac disease), and pancreatic insufficiency. The clinical picture of all these diseases is roughly similar but varies according to etiology, severity, and duration. The most common chief complaints in severe malabsorption are diarrhea and weakness, weight loss, and mild functional GI complaints (anorexia, nausea, mild abdominal pain). Physical findings and laboratory test results tend to differ with the various etiologies. In severe cases of sprue, tetany, bone pain, tongue surface atrophy, and even bleeding may be found. Physical examination may show abdominal distention and also peripheral edema in nearly half the patients. In pancreatic insufficiency, physical examination may be normal or show malnutrition. Neurologic symptoms are found with moderate frequency in pernicious anemia but may be present in malabsorption of other causes as well.

    Laboratory findings. Laboratory test findings vary according to severity and etiology of the malabsorption, but in sprue they most often include one or more of the following: anemia, steatorrhea, hypoproteinemia, hypocalcemia, and hypoprothrombinemia. In pancreatic insufficiency, the main laboratory test abnormalities are steatorrhea and decreased carbohydrate tolerance (sometimes overt diabetes). In pernicious anemia the patient has only anemia without diarrhea, steatorrhea, or the other test abnormalities previously listed. Most stomach operations do not cause diarrhea or abnormalities of fat absorption.

    Steatorrhea is caused by excess excretion of fat in the stools due to inability to absorb lipids. Anemia associated with steatorrhea is most often macrocytic but sometimes is caused by iron deficiency or is a mixed type due to various degrees of deficiency of folic acid, vitamin B12, and iron. Calcium may be decreased both from GI loss due to diarrhea and from artifact due to hypoalbuminemia. Prothrombin formation by the liver is often impaired to some degree because of lack of vitamin K. Vitamin K is a fat-soluble vitamin that is obtained from food and also is produced by bacteria in the small bowel. Long-term oral antibiotic use may reduce the bacterial flora by killing these bacteria and thus may interfere with vitamin K formation. Inability to absorb fat secondarily prevents vitamin K and vitamin A, which are dependent on fat solubility for intestinal absorption, from entering the bloodstream. Malnutrition resulting from lack of fat and carbohydrate absorption leads to hypoalbuminemia because of decreased production of albumin by the liver. This also contributes to the peripheral edema that many patients develop.

    Classification of Malabsorptive Disorders (With Comments on Occurrence and Associated Abnormalities)

    I. Inadequate mixing of food with bile salts and lipase. Mild chemical steatorrhea common, but clinical steatorrhea uncommon. Actual diarrhea uncommon. Anemia in approximately 15%-35%; most often iron deficiency, rarely megaloblastic.
    A. Pyloroplasty
    B. Subtotal and total gastrectomy (occasional megaloblastic anemias reported)
    C. Gastrojejunostomy
    II. Inadequate lipolysis — lack of lipase or normal stimulation of pancreatic secretion. Steatorrhea only in far-advanced pancreatic destruction, and diarrhea even less often.
    A. Cystic fibrosis of the pancreas
    B. Chronic pancreatitis
    C. Cancer of the pancreas or ampulla of Vater
    D. Pancreatic fistula
    E. Severe protein deficiency
    F. Vagus nerve section
    III. Inadequate emulsification of fat — lack of bile salts. Clinical steatorrhea uncommon, sometimes occurs in very severe cases. Usually no diarrhea.
    A. Obstructive jaundice
    B. Severe liver disease
    IV. Primary absorptive defect—small bowel.
    A. Inadequate length of normal absorptive surface; unusual complication of surgery
    1. Surgical resection
    2. Internal fistula
    3. Gastroileostomy
    B. Obstruction of mesenteric lymphatics (rare)
    1. Lymphoma
    2. Hodgkin’s disease
    3. Carcinoma
    4. Whipple’s disease
    5. Intestinal tuberculosis
    C. Inadequate absorptive surface due to extensive mucosal disease; except for Giardia infection and regional enteritis, most of these diseases are uncommon; steatorrhea only if there is extensive bowel involvement
    1. Inflammatory
    a. Tuberculosis
    b. Regional enteritis or enterocolitis (diarrhea very common)
    c. Giardia lamblia infection (diarrhea common; malabsorption rare)
    2. Neoplastic
    3. Amyloid disease
    4. Scleroderma
    5. Pseudomembranous enterocolitis (diarrhea frequent)
    6. Radiation injury
    7. Pneumatosis cystoides intestinalis.
    D. Biochemical dysfunction of mucosal cells
    1. “Gluten-induced” (steatorrhea and diarrhea very common)
    a. Celiac disease (childhood)
    b. Nontropical sprue (adult)
    2. Enzymatic defect
    a. Disaccharide malabsorption (diarrhea frequent symptom)
    b. Pernicious anemia (deficiency of gastric “intrinsic factor”)
    3. Cause unknown; uncommon except for tropical sprue (which is common only in the tropics)
    a. Tropical sprue (diarrhea and steatorrhea common)
    b. Severe starvation
    c. Diabetic visceral neuropathy
    d. Endocrine and metabolic disorder (e.g., hypothyroidism)
    e. Zollinger-Ellison syndrome (diarrhea common; steatorrhea may be present)
    f. Miscellaneous
    V. Malabsorption associated with altered bacterial flora (diarrhea fairly common)
    1. Small intestinal blind loops, diverticula, anastomoses (rare)
    2. Drug (oral antibiotic) administration (infrequent but not rare)

    The majority of patients with malabsorption usually present in one of two ways. In the first group the major finding on admission is anemia, and once malabsorption is suspected, either by the finding of megaloblastic bone marrow changes or by other symptoms or signs suggestive of malabsorption, the problem becomes one of differentiating pernicious anemia from other types of malabsorption. In the second group, the patients present with one or more clinical symptoms of malabsorption, either mild or marked in severity. The diagnosis must be firmly established and the etiology investigated. There are several basic tests for malabsorption which, if used appropriately, usually can lead to the diagnosis and in some cases reveal the cause.

    Useful individual laboratory tests

    Qualitative fecal fat. Fat in the feces can be stained with Sudan III dye. Neutral fat can be seen as bright orange droplets, but the fatty acids normally do not stain. Both these fatty acids and the original neutral fat can be converted to stainable fatty acids by heat and acid hydrolysis. The preparation is then stained and examined a second time to determine if the number of droplets has increased from the first examination. The reliability of this procedure is debated in the literature, but it is reported to be reasonably accurate if the technician is experienced. However, it is sometimes difficult to be certain whether Sudan-positive droplets are fat or some other substance. Naturally, there will be difficulty in distinguishing normal results from low-grade steatorrhea. It is possible to get some idea of etiology by estimating the amount of neutral fat versus fatty acid: lack of fatty acid suggests pancreatic disease.

    Quantitative fecal fat. The basic diagnostic test for steatorrhea is quantitative fecal fat. Stool collections are taken over a minimum of 3 full days. The patient should be on a diet containing approximately 50-150 gm/day of fat (average 100 gm/day) beginning 2 days before the test collection. It is necessary to make sure that the patient is actually eating enough of this diet to take in at least 50 gm of fat/day; it is also obviously important to make sure that all the stools are collected and the patient is not incontinent of feces. Patient noncompliance with complete stool collection is probably the most common cause of false negative results. If the patient is constipated (and some are), it may be necessary to use a bedtime laxative. Normal diet results in an average excretion of less than 7 gm of fat/24 hours. Excretion of 5-7 gm/24 hours is equivocal, since many patients with minimal steatorrhea and a small but significant percentage of normal persons have excretion in this range. There are some reports that 24-hour fecal fat excretion less than 9.5 gm/100 gm of stool favors nontropical sprue and celiac disease, whereas excretion greater than 9.5 gm/100 gm of stool favors pancreatic insufficiency, bacterial overgrowth, or biliary tract disease. Finally, some patients with partial or complete malabsorption syndromes may have normal fecal fat excretion. This is most common in tropical sprue.

    Plasma carotene. Carotene is the fat-soluble precursor of vitamin A and is adequately present in most normal diets that contain green or yellow vegetables. Normal values are considered 70-300 µg/100 ml (1.3-5.6 µmol/L). Values of 30-70 µg/100 ml are usually considered moderately decreased, and levels less than 30 µg/100 ml (0.56 µmol/L) indicate severe depletion. Other causes of low plasma carotene, besides malabsorption, are poor diet, severe liver disease, and high fever. There is considerable overlap between the carotene values of malabsorption and the carotene values in normal control patients, but such overlap usually is over the 30-µg level. However, this test is valuable mostly as a screening procedure. The patient must be eating a sufficient quantity of carotene-rich food to draw valid conclusions from a low test result.

    X-ray examination. A small bowel series is done by letting barium pass into the small intestine. There are several changes in the normal radiologic appearance of the small bowel that are suggestive of malabsorption. These changes appear in 70%-90% of patients, depending on the etiology and severity of the disease and the interpretative skill of the investigator. The radiologic literature agrees that many chronic diseases, especially when associated with fever and cachexia, may interfere with digestion so severely as to produce a pattern that may be confused with sprue. Secondary malabsorption cannot be distinguished from primary malabsorption except in certain rare cases such as tumor. The so-called diagnostic patterns of sprue are thus characteristic of, but not specific for, primary small intestine absorption and are not present in about 20% of patients.

    Schilling test. This is the classic test for vitamin B12 malabsorption, since it can differentiate pernicious anemia from other malabsorption etiologies affecting the terminal ileum where B12 is absorbed. This is discussed in Chapter 3.

    D-xylose test. Besides quantitative (fecal) fat studies, the most important test for malabsorption is the D-xylose test. Rather than a screening test for malabsorption per se, it is a test that identifies the sprue-type diseases and differentiates them from other malabsorption etiologies. Originally, an oral glucose tolerance test (OGTT) was used in malabsorption, since it was found that most patients with sprue showed a flat curve. A “flat” OGTT is usually defined as an OGTT peak with a value no greater than 25 mg/100 ml (1.4 mmol/L) above the baseline fasting level, although there is some disagreement about this criterion. However, some patients with obvious malabsorption have a normal curve, and it was also found in several large series that up to 20% of apparently normal persons had a flat curve, so this test was abandoned.

    Test protocol. D-xylose is a pentose isomer that is absorbed in much the same manner as glucose from the jejunum. The standard test dose is 25 gm of D-xylose in 250 ml of water, followed by another 250 ml of water. The patient is fasted overnight, since xylose absorption is delayed by other food. After the test dose, the patient is kept in bed for 5 hours without food. The normal person’s peak D-xylose blood levels are reached in approximately 2 hours and fall to fasting levels in approximately 5 hours. D-xylose is excreted mostly in the urine with approximately 80%-95% of the excretion in the first 5 hours and the remainder in 24 hours. Side effects of oral D -xylose administration are mild diarrhea and abdominal discomfort in a small number of patients.

    Interpretation. Normal values for (2-hour) blood D-xylose levels are more than 25 mg/100 ml (1.66 mmol/L); values of 20-25 mg/100 ml are equivocal, and values less than 20 mg/100 ml (1.33 mmol/L) are strongly suggestive of malabsorption. The 5-hour urinary D-xylose normal values are more than 5 gm/5 hours. It is obviously very important to make sure that urine collection is complete and that there is no fecal contamination of the urine. A catheter may have to be used if the patient is incontinent of urine or if there is a question of fecal contamination, but catheterization should be avoided if at all possible. Two main physiologic circumstances may affect the 5-hour urinary excretion: renal insufficiency and advanced age. There may be abnormally low 5-hour urinary excretion of D-xylose in persons over age 65. However, one study claims that the 24-hour urine collection is normal (also >5 gm) unless actual malabsorption is present. If the serum creatinine level is borderline or elevated, the 5-hour urinary D -xylose excretion is also likely to be abnormally low, and again the 24-hour excretion may be useful. In these cases, however, the 2-hour blood levels may help, because they should be normal and are not affected by age or renal insufficiency. Otherwise, the 5-hour urinary excretion is more reliable than the blood levels, which tend to fluctuate.

    Clinical correlation. The D-xylose test may be helpful in determining if the patient has malabsorption and also provides some clues as to etiology. Most patients with cystic fibrosis and pancreatic insufficiency are said to have normal urinary D-xylose values. This is also true of most patients with liver disease. Patients with classic pernicious anemia have normal D-xylose test results, although it must be remembered that many of these patients are aged and for that reason may have low 5-hour urine results. Some patients with megaloblastic anemia of pregnancy have abnormal D-xylose test results, although probably the majority have normal values. A small percentage of patients with partial gastrectomy reportedly have abnormal urine values. Patients with functional diarrhea and duodenal ulcer have normal results.

    In malabsorption diseases, there is excellent correlation of D-xylose excretion with proved sprue and celiac disease. The urine results are more often clear-cut than the blood levels. There is no correlation with the degree of steatorrhea. Patients with regional enteritis involving extensive areas of the jejunum may have abnormal results, whereas normal results are associated with this disease when it is localized to the ileum. Patients with Whipple’s disease, “blind loop” syndrome (isolated small intestine area of bacterial overgrowth), postgastrectomy, small intestine lymphoma, multiple jejunal diverticula, and some infants with cow’s milk allergy may also have abnormal results. Certain patients with diseases other than classic malabsorption may have an abnormal D-xylose test result. These diseases include myxedema, diabetic neuropathic diarrhea, rheumatoid arthritis, acute or chronic alcoholism, and occasionally severe congestive heart failure. Ascites is reported to produce abnormal urine excretion with normal plasma levels. Although D-xylose excretion is frequently depressed in myxedema, abnormal test results occur in only a small number of patients with the other conditions listed. Some of these conditions can produce decreased absorption of substances other than D-xylose, although such problems are usually mild or moderate in degree and may be due to multiple factors.

    D-xylose tests thus may be abnormal in diseases other than sprue. In nontropical sprue, about 10% of untreated patients have normal D-xylose 5-hour urine test results (literature range, 0%-40%). There is also sufficient overlap in the urine excretion range of 4-5 gm/5 hours from persons without primary small intestine malabsorption to warrant routine collection of a 19-hour urine specimen immediately following the 5-hour specimen (to have a total of 24 hours, if necessary). Several studies found better results in children less than 12 years old using a 5-g oral dose of D-xylose and obtaining a serum specimen (no urine specimen) 1 hour after D-xylose administration. The lower limit of serum reference range was 20 mg/100 ml.

    Hydrogen breath test. An oral test dose of a specific carbohydrate is administered. If the carbohydrate is not absorbed normally in the small intestine, it reaches the colon, where bacteria metabolize the carbohydrate and release various metabolic products, among which is hydrogen gas. About 15%-20% of the hydrogen is absorbed and then released from the lungs in expired air. Expired air is collected in a single-breath collection bag or other apparatus and analyzed for hydrogen content by some variant of gas chromatography. This technique has been used to test for various types of malabsorption. It has proved most useful in diagnosis of deficiency involving the enzyme lactase, small intestine “blind loop” syndrome, and some cases of rapid intestine transit (“intestinal hurry syndrome”). In the case of the blind loop syndrome or rapid transit, the test dose consists of a carbohydrate that normally is not absorbed. The hydrogen breath test has not proved reliable in diagnosis of sprue or glutin-associated malabsorption.

    There are various conditions that interfere with the test. Recent use of antibiotics can affect the bacterial flora, sometimes when discontinued as long as 2 weeks before the test. Use of colon enemas can partially wash out some of the flora. Delayed or unusually swift gastric emptying can change the quantity of the carbohydrate or the time that it reaches the colon. Breath collected during sleep contains 2-3 times the amount of hydrogen obtained when the patient is awake. Cigarette smoking produces large amounts of hydrogen. Finally, the collection apparatus and the analysis equipment are relatively expensive, and the test is usually available only in larger institutions or medical centers.

    Small intestine biopsy. In classic sprue, both tropical and nontropical, the mucosa of the small intestine shows characteristic histologic abnormalities. Instead of the normal monotonous fingerlike villous pattern, the villi are thickened and blunted, with flattening of the cuboidal epithelium, and the villi may eventually fuse or disappear altogether. Depending on the degree of change in the villi, biopsy results may show moderate or severe changes. These same changes may be found to a much lesser degree in many of the other conditions causing malabsorption, including even subtotal gastrectomy. However, these usually are not of the severity seen in sprue and generally can be differentiated by the clinical history or other findings. Other causes of malabsorption, such as the rare Whipple’s disease (which characteristically shows many periodic acid-Schiff-positive macrophages in the mucosa) may be detected on biopsy. In infants less than 1 year of age, transient small intestinal mucosal abnormalities similar to those of sprue have been reported in some patients with acute gastroenteritis and in some with cow’s milk allergy. Eosinophilic gastroenteritis is another cause in older children.

  • Evaluation of Protein-Calorie Nutritional Status

    Various studies have shown that a significant degree of malnutrition is frequent in hospitalized persons, ranging from 25%-50% of patients (depending on whether the population screened was a general or specialty group). In one report, 97% of surgical patients had at least one abnormal result on tests for nutritional status.

    Classification of protein-calorie malnutrition

    Although protein or caloric deficiency exists in grades of severity, a classification of patients according to pathophysiology and laboratory abnormalities requires analysis of late-stage deprivation. At a late stage, three basic patient types have been described: kwashiorkor, marasmus, and a mixed picture.

    Kwashiorkor results from protein deficiency without total calorie deficiency. This condition is produced by a diet with adequate calories that are obtained almost exclusively from carbohydrate. This may result from a nonhospital diet that is low in protein or may be seen in hospitalized patients who are maintained primarily on IV dextrose. Severe stress, major illness, or surgery results in greatly increased utilization of body protein and may rapidly lead to protein depletion if there is not adequate replacement. These patients externally seem to be of normal weight or even overweight and may have edema. Kwashiorkor involves depletion of primarily visceral (nonmuscle) protein.

    Marasmus is produced by prolonged deficiency of both protein and carbohydrates. Examples are starvation due to inability to eat and anorexia nervosa. These patients lose both fat and muscle mass and appear emaciated. Marasmus involves loss of primarily somatic (fat and muscle) protein rather than visceral protein.

    The mixed category combines various degrees of protein deprivation with various degrees of carbohydrate and total calorie deficiency. It may also result from end-state marasmus (when both somatic and visceral protein are consumed) or may occur in a patient with moderate marasmus who undergoes severe stress, thus accelerating visceral protein loss.

    Besides classic end-state cases there are far greater numbers of patients with malnutrition of lesser severity. In general, the greater the degree of deficiency, the greater the chance of unwanted consequences. These include increased postoperative morbidity and mortality and certain defined complications such as increased tendency toward infection, poor wound healing, and extended hospitalization.

    Tests useful in patients with malnutrition

    Functional categories of tests (i.e., information that tests can supply) in protein-calorie malnutrition include the following:

    1. Tests that screen patients for protein-calorie deficiency
    2. Tests that assess degree of deficiency
    3. Tests that differentiate the various types of deficiency
    4. Tests used to guide therapy

    Procedures or tests available

    1. Anthropometric measurements: triceps skinfold thickness; midarm circumference
    2. Calculation of undernourishment based on body height and weight (percent of ideal weight or percent of preillness weight)
    3. Biochemical tests reflecting visceral (nonmuscle) protein status: serum albumin and serum transferrin (also, serum iron-binding capacity, retinol-binding protein, serum prealbumin)
    4. Metabolic indices: creatinine-height index (somatic protein or muscle mass) and urinary nitrogen excretion (protein catabolism)
    5. Tests of immune status: skin tests with various antigens, total lymphocyte count

    Anthropometric measurements

    These procedures are designed to estimate fat and muscle wasting, which reflects somatic protein depletion. Triceps skin fold measurement is performed with special calipers and measures fat reserves. Midarm circumference is used to estimate lean body mass. Patient measurements are compared with values in standard tables.

    Weight deficiency (weight loss)

    Weight deficiency may be calculated either as a percentage of preillness weight (current weight/preillness weight) or as a percentage of ideal weight (current weight/ideal weight). Ideal weight requires measurement of height and use of ideal weight tables. Percent weight loss after hospitalization is also useful. In all the various measurements, a 10% weight loss is considered suspicious for protein-calorie deficiency. If this occurred before admission, it probably developed over a relatively extended period of time (there is no standard time period, but at least 4 weeks has been suggested and seems reasonable) rather than over a short period of time, which is more likely to be a fluid problem. After hospitalization, there is no time limit if the weight loss is not due to diuretics, fluid removal, or some other obvious cause. Edema and obesity may produce error in nutritional assessment by these methods.

    Tests for visceral protein status

    Serum albumin. Albumin is the major force maintaining plasma oncotic pressure. It is synthesized by the liver from amino acids. Decreased serum albumin levels result from decreased production, either from defective synthesis because of liver cell damage, deficient intake of amino acids (absolute protein intake deficit); or from disease- or stress-induced catabolism of body protein, which increases the need for dietary protein without a corresponding increase in dietary protein intake (relative protein intake deficit). The serum albumin level is thus considered an indicator of visceral (nonmuscle) protein status. Other serum proteins that have been used for the same purpose include transferrin, prealbumin, and retinol-binding protein.

    Serum albumin has a serum half-life of about 20 days and begins to decrease about 2 weeks after onset of protein depletion. Mild protein deficiency is said to correlate with albumin levels of 3.0-3.5 gm/100 ml (30-35 g/L); moderate deficiency, 2.1-3.0 gm/100 ml (21-30 g/L); and severe deficiency, less than 2.1 gm/100 ml (21 g/L). Other etiologies for albumin decrease besides deficient protein intake include disorders of liver synthesis (cirrhosis, severe acute liver disease), extravascular protein loss (nephrotic syndrome, acute or chronic protein-losing enteropathy, extensive burns), and hemodilution (congestive heart failure). Albumin decrease, whether within or below reference limits, is seen in many severe acute and chronic illnesses. The exact mechanism is frequently uncertain or is sometimes due to more than one cause. Overhydration and dehydration also may change apparent albumin levels.

    Serum transferrin. Transferrin has a serum half-life of about 9 days and begins to decrease about 1 week after onset of protein depletion. However, chronic iron deficiency, therapy with estrogen or estrogen-containing contraceptives, and the same severe acute and chronic illnesses that decrease albumin levels tend to elevate serum transferrin levels and could mask early change due to nutritional depletion. Transferrin can be measured directly in a variety of ways, usually by immunologic (antitransferrin antibody) techniques, or can be estimated using serum total iron-binding capacity (TIBC). TIBC is easier and less expensive for most laboratories. The formula most commonly used is: transferrin = (0.8 Ч TIBC) – 43. Mild protein deficiency correlates with transferrin levels of 150-175 mg/100 ml (1.5-1.7 g/L); moderate deficiency, 100-150 mg/100 ml (1.0-1.5 g/L); and severe deficiency, less than 100 mg/100 ml (1.0 g/L).

    Serum prealbumin. Prealbumin is a carrier protein for retinol-binding protein and for a small part of serum thyroxine. Its serum half-life is about 2 days. Its serum concentration decreases in many severe illnesses. It is measured by immunologic techniques, most commonly by radial immunodiffusion or immunonephelometry. Prealbumin levels begin to decrease within 48-72 hours in response to protein malnutrition. However, like albumin, it is decreased by severe liver disease or as a temporary short- or long-term result of many severe acute or chronic illnesses.

    Retinol-binding protein. Retinol-binding protein is the specific binding protein for vitamin A. Its serum half-life is only about 10 hours. It begins to decrease within 48-72 hours after onset of protein malnutrition and otherwise behaves like prealbumin. However, in addition, retinol-binding protein may decrease when renal function decreases (as frequently occurs in severely ill persons).

    Most investigators have accepted serum albumin as the most practical marker for visceral protein depletion. When serial measurements of visceral protein indices are necessary, transferrin may be substituted because it responds faster than albumin to change in nutrition status. Transferrin can also be used if albumin measurement is invalidated by therapeutic administration of albumin. Comparisons of serum albumin with other parameters of malnutrition have generally shown that serum albumin levels have the best single-test correlation with patient outcome.

    Metabolic indices

    Creatinine height index (CHI). This calculated value is thought to provide an estimate of lean body mass, based on the theory that urinary creatinine (UC) output is related to body muscle mass. Height is used to relate patient data to data of normal (“ideal”) persons. The creatinine height index (CHI) has the advantage that it relates to skeletal muscle mass (somatic protein) rather than to liver production of serum protein (visceral protein). In classic marasmus, the CHI is markedly decreased, whereas the serum albumin concentration may be either within reference range or close to it. Another advantage is that edema does not greatly affect the CHI, whereas it might affect arm circumference measurement. The major disadvantage is need for accurate 24-hour urine collection. Also, urine ketone bodies can interfere with creatinine assay. Data for ideal urine creatinine excretion are presented in Table 37-11.

    Nitrogen balance estimates. Nitrogen balance (NB), as measured by urine urea nitrogen (UUN), may be helpful in assessing the quantitative and compositional adequacy of nutritional therapy. The formula most often used is:

    NB = (Protein intake / 6.25) – (UUN + 4)

    when nitrogen balance is in terms of net gain (+) or loss (–) in grams of nitrogen per day and both protein intake and urine urea output are in grams per day. Reference limits are +4 to –20 gm of nitrogen/day. Protein intake (in grams/day) is usually estimated but can be measured if feeding is entirely through nasogastric tube or hyperalimentation. If the patient has oral food intake, only the food actually eaten rather than food provided should be used in the calculation. The 4-gm correction factor is supposed to compensate for urine non-urea nitrogen loss additional to the urea nitrogen. If additional loss occurs from the GI tract, fistulas, and so forth, such loss must be estimated and incorporated into the correction factor. The 24-hour urine collection must be complete, because incomplete collection results in a falsely higher value.

    Tests of immune status

    Moderate or severe protein-calorie malnutrition of any type often results in depression of the body immune system. This is reflected in decreased immune response (especially, delayed hypersensitivity response) to various stimuli.

    Total lymphocyte count. There is a rough correlation of total lymphocyte count with degree of malnutrition. The correlation is closest in kwashiorkor (visceral protein depletion). Total lymphocyte counts of 1,200-2,000/mm3 (1.2-2.0 x 109/L) are said to be associated with mild protein depletion; counts of 800-1,200/mm3 with moderate depletion, and counts of less than 800/mm3 with severe depletion. However, there is considerable overlap between immunologic impairment and nonimpairment with counts higher than 1,000/mm3. Values less than 1,000/mm3 (1.0 x 109/L) are generally considered evidence of definite significant immunologic impairment. Total lymphocyte count is easy to obtain (WBC count x percent lymphocytes in the peripheral smear WBC differential result). Most investigators have found less correlation with eventual patient outcome than with serum albumin levels. One reason is that various other conditions may cause a decrease in total lymphocytes. Best correlation of total lymphocyte count to patient well-being seems to be associated with infection and with cancer therapy. When both albumin and total lymphocyte counts are significantly decreased, there is some reinforcement of significance compared with abnormality in either test alone.

    Skin tests. Skin test response to various intradermally injected delayed hypersensitivity antigens such as Candida, mumps, and streptokinase-streptodornase provides an in vivo method to evaluate immune response. Reactions are measured at 24 and 48 hours. Various studies have shown a substantial correlation between lack of skin test reactivity to multiple antigens and an increased incidence of sepsis or postoperative complications and mortality. There is some disagreement in the literature on the predictive value of skin tests versus serum albumin levels, with the majority opinion favoring albumin. The major drawbacks to skin testing are the time interval required and the necessity for good injection technique.

    Current status of tests for protein-calorie malnutrition

    Patient screening for protein-calorie malnutrition. Although various institutions have different protocols, the most common practice that is emerging is to determine the percent of weight loss and the serum albumin level, either one alone or in combination, and to take into consideration the type of illness or therapy involved. Total lymphocyte count also seems to be widely used as an adjunctive test. The CHI is helpful if marasmus is suspected; serum albumin levels could be misleading if albumin were the sole criterion for possible malnutrition.

    Tests to assess degree of deficiency. Serum albumin level is the most widely used single test. The CHI is also widely employed if marasmus is present.

    Nutritional Deficiency Syndromes and Screening Tests
    Overall nutritional status
    1. Percent weight loss (at least 10% nondiuretic loss)
    Marasmus (somatic protein and fat deficit due to total calorie deficiency; somatic protein = skeletal muscle lean body mass)
    1. Somatic protein estimation
    Creatinine-height index
    Midarm circumference
    2. Fat
    Triceps skin fold thickness
    Kwashiorkor (visceral protein deficit due to protein intake deficiency; visceral protein = liver-produced protein, including plasma proteins)
    1. Serum albumin (or transferrin, prealbumin, retinol-binding protein)
    2. Total lymphocyte count
    3. Cell-mediated immunity
    Mixed kwashiorkor and marasmus (deficit in both protein intake and total calories)
    1. Tests abnormal in both deficiency groups.

    Anthropometric measurements, serum transferrin determination, and tests of immune function are available but seem to be ordered more in university or research centers.

    Tests to differentiate categories of malnutrition. In classic kwashiorkor, anthropometric measurements and CHI values are relatively normal, whereas serum albumin levels and other tests of visceral protein status are decreased. In classic marasmus, anthropometric measurements and the CHI are decreased, whereas results of visceral protein adequacy tests may be normal. Although immune status test results are depressed in severe marasmus, over the broad spectrum of marasmus severity they are not as severely affected as in kwashiorkor. It must be emphasized that patients may have some combination of overall protein-calorie deficiency and of severe protein loss and therefore may not have clear-cut differential test patterns.

    Tests to guide nutritional therapy. Serum albumin levels and total lymphocyte count are still the most commonly used parameters of therapeutic response. Because serum albumin has a relatively long half-life and changes relatively slowly in response to renourishment, and also because fluid shifts influence albumin levels, an increasing number of investigators use serum transferrin or prealbumin levels rather than albumin levels to monitor therapy. Some find that urinary nitrogen excretion data are very helpful in determining when a positive nitrogen balance has been achieved, but others do not believe that it is necessary.

  • Trace Elements

    Zinc

    Zinc is a component of certain important enzymes, such as carbonic anhydrase, lactic dehydrogenase, alkaline phosphatase, DNA and RNA polymerases, and d-aminolevulinic acid dehydratase. Zinc is obtained primarily through food. About 30% of that ingested is absorbed in the small intestine. About 80% of zinc in blood is found in RBCs, mostly as part of the enzyme carbonic anhydrase. Of that portion not in RBCs, about 50% is bound to albumin, about 30% is bound to alpha-2 macroglobulin or transferrin, and about 5% is bound to histidine and certain other amino acids, leaving about 15% free in plasma. Excretion occurs predominately in the stool, with a much smaller amount excreted in urine and sweat.

    Zinc deficiency is usually not clinically evident until it becomes severe. Severe deficiency may produce growth retardation, delayed sexual development, acrodermatitis enterohepatica (dermatitis, diarrhea, and alopecia), decreased taste acuity, and poor wound healing. Acrodermatitis enterohepatica can either be congenital (autosomal recessive trait) or can appear in severe acquired zinc deficiency.

    Conditions producing zinc deficiency include inadequate diet intake (most often in hospitalized patients on IV feeding, including hyperalimentation patients), conditions that interfere with intestinal absorption (high-fiber or phytate diet, prolonged severe diarrhea, steatorrhea), excess zinc loss (sickle cell anemia), increased zinc requirement (pregnancy, lactation, wound healing), and certain diseases such as alcoholism and cirrhosis.

    Assay of plasma zinc is usually done by atomic absorption spectrophotometry. Contamination is a major problem. Rubber stoppers or gaskets are well known for this. Glassware must be specially prepared. Another major problem is considerable variation in reference range between laboratories for both adults and infants. Serum is reported to have slightly higher levels than plasma. Since a substantial amount of plasma zinc is bound to albumin, changes in albumin can change (total) plasma zinc levels without reflecting patient body zinc levels. Hemolysis invalidates zinc measurement due to the high levels in RBCs. There is a circadian rhythm, with values somewhat higher in the morning.

    Aluminum

    Normally, small amounts of aluminum are ingested with food. Other sources include aluminum leached from aluminum cooking utensils by acidic juices, and alum-containing baking soda, processed cheese, and beer. Aluminum in serum is predominantly bound to transferrin, with a small amount bound to citrate in extracellular fluid. The only source of excretion is the kidney. Most interest in aluminum is focused on aluminum toxicity in patients with renal failure. These patients develop microcytic anemia, osteodystrophy (osteomalacia) resistant to Vitamin D, and encephalopathy. The osteodystrophy is caused by deposition in bone of excessive aluminum, where it interferes with bone mineralization. Initially, aluminum toxicity was thought to be due to aluminum contamination of water used for renal dialysis. More recently, the source of aluminum has been traced to aluminum-containing preparations used to bind phosphates in the GI tract to prevent phosphate absorption, with subsequent phosphate accumulation and development of secondary hyperparathyroidism. The gold standard test for aluminum osteodystrophy is bone biopsy with either (or both) chemical analysis or histochemical staining for aluminum content. However, other procedures have been used to estimate likelihood of aluminum bone toxicity. The serum aluminum level has been most commonly used for this purpose. Values greater than 50 ng/ml are generally considered abnormal, and values greater than 100 ng/ml are generally considered suggestive of possible aluminum bone toxicity. However, in one series about 30% of patients with serum aluminum values greater than 100 ng/ml failed to show definite evidence of aluminum toxicity on bone biopsy (70% specificity), and about 20% with serum values less than 100 µg/ml had biopsy evidence of aluminum toxicity (80% sensitivity). Many patients with aluminum toxicity develop a microcytic anemia (although it must be emphasized that microcytic anemia is not specific for aluminum toxicity). In one series, about 20% of renal hemodialysis patients with a microcytic mean cell volume (MCV) had a serum aluminum level less than 50 ng/ml, about 40% had a level between 50 and 100 ng/ml and about 40% had a level more than 100 ng/ml. About 25% of patients with serum aluminum values greater than 100 ng/ml had a normal MCV. However, no patient with a value more than 140 ng/ml had a normal MCV. Another test involves infusion of a chelating drug desferrioxamine, which extracts some aluminum from tissues and increases serum aluminum levels by a certain amount over baseline if the tissues contain excess aluminum. Serum aluminum assay is difficult and is available only at large reference laboratories and some medical centers. The major problem is contamination by aluminum in laboratory apparatus and in the environment. Some sources of contamination include aluminum needles used for specimen collection, rubber stoppers on blood tubes, contaminated pipets or other glassware, and aluminum contamination of environmental dust.

  • Phosphorus and Phosphate Abnormalities

    Phosphorus and phosphate are often spoken of interchangeably, although phosphorus is only one component of phosphate. The semantic problem is even more confusing because an order for “phosphate” assay usually results in laboratory measurement of inorganic phosphorus. However, much of the body phosphorus is a part of phosphate compounds. About 80%-85% of body phosphorus is found in bone and about 10% in skeletal muscle. Most body phosphorus is intracellular, where it represents the most abundant intracellular anion. Phosphorus is a part of phospholipid compounds in all cell membranes, adenosine triphosphate energy-transfer compounds, nucleic acids, the compound 2,3-diphosphoglyceric acid (which regulates oxygen affinity for hemoglobin), various enzymes, and the principal urinary acid-base buffer system. Phosphorus is acquired through food and absorbed through the small intestine. About 90% is extracted from serum by the kidney with about 85%-90% being normally reabsorbed by the renal proximal tubules. Serum phosphorus values change considerably during the day (variation of 2 mg/100 ml [0.65 mmol/L] within a reference range of 2.5-4.5 mg/100 ml [0.81-1.45 mmol/L]), with lowest values at 10-11 A.M. and highest at 10 P.M. -3 A.M. Therefore, values are usually higher in the late afternoon and evening than in the morning. Some of these changes are due to dietary factors and some to shifts between intracellular and extracellular localization. Phosphate excretion is low about 9 A.M.-1 P.M., high about 3 P.M. -8 P.M., low again about midnight-1 A.M., and high again about 3-5 A.M. Administration of glucose leads to a temporary shift of phosphorus from an extracellular to an intracellular location. If a glucose load is given orally, trough serum phosphorus values are found about 2 hours postprandially, and preingestion values are regained about 5 hours postprandially.

    Hypophosphatemia

    Most clinical abnormalities involving phosphorus are associated with hypophosphatemia. Symptoms include confusion, disorientation, delirium, and sometimes seizures, thus resembling the symptoms of hyponatremia and other metabolic encephalopathies. In addition, there is skeletal muscle weakness that may progress to actual myopathy. In chronic severe hypophosphatemia there may be bone abnormalities such as osteomalacia and pseudofractures, as well as hematologic abnormalities such as decrease in oxygen delivery by RBCs and a tendency toward hemolysis. WBC function may be disturbed, with an increased incidence of fungal and bacterial infection. Mild hypophosphatemia, on the other hand, is usually asymptomatic clinically and biochemically.

    The overall incidence of hypophosphatemia ranges from 2%-22% in hospitalized patients. The great majority of patients demonstrate only a mild degree of abnormality and no clinical effects. The box lists conditions more likely to be associated with severe hypophosphatemia. For example, severe hypophosphatemia may appear in chronically malnourished persons who undergo rapid refeeding with low-phosphate nutrients (nutritional recovery syndrome). In many of the conditions listed under severe phosphate deficiency, onset of hypophosphatemia may not appear until 1 or more days after onset of illness. The other conditions on the list frequently produce hypophosphatemia but usually only of moderate degree. Even more conditions may produce mild disorder. There is an association of hypophosphatemia with hypomagnesemia, especially in alcoholics.

    Selected Disorders Associated With Serum Phosphate Abnormality

    Phosphate decrease*
    Parenteral hyperalimentation
    Diabetic acidosis
    Alcohol withdrawal
    Severe metabolic or respiratory alkalosis
    Antacids that bind phosphorus
    Malnutrition with refeeding using low-phosphorus nutrients
    Renal tubule failure to reabsorb phosphate (Fanconi’s syndrome; congenital; vitamin D deficiency)
    Glucose administration
    Nasogastric suction
    Malabsorption
    Gram-negative sepsis
    Primary hyperthyroidism
    Chlorothiazide diuretics
    Therapy of acute severe asthma
    Acute respiratory failure with mechanical ventilation
    Phosphate excess
    Renal failure
    Severe muscle injury
    Phosphate-containing antacids
    Hypoparathyroidism
    Tumor lysis syndrome

    __________________________________________________________
    *Low phosphate diet can magnify effect of phosphorus-lowering disorders.

    In one study, the most common etiology was medication known to induce hypophosphatemia without phosphate supplements. This most often occurred in association with surgery. The second most common etiology was gram-negative sepsis.

    Hyperphosphatemia

    The most common cause of hyperphosphatemia is renal failure. Other causes are listed in the box on this page. Hyperlipidemia or RBC hemolysis may produce artifactual phosphate elevation. Hyperphosphatemia may lead to hypocalcemia.

  • Serum Magnesium Abnormalities

    Magnesium is the fourth most common body cation (after sodium, potassium, and calcium) and the second most common intracellular cation (after potassium). About half is located in soft tissue and muscle cells and about half is in bone. Only 1%-5% is extracellular. Most body magnesium is derived from food intake. About one third of dietary magnesium is absorbed, with the absorption site being the small intestine. Body magnesium is excreted by the kidney, primarily through glomerular filtration. Some tubular reabsorption also takes place. About 33% of serum magnesium (literature range, 15%-45%) is bound to serum proteins, 15% is complexed, and about 50% is free in the ionized form. Of the protein-bound fraction, about 75% is attached to albumin and most of the remainder to alpha-1 and alpha-2 globulin. Albumin thus carries about 30% (range, 25%-33%) of total serum magnesium. PTH is a very important regulator of magnesium blood levels through regulation of renal tubule reabsorption.

    Magnesium is important in protein synthesis, enzyme activation, and oxidative phosphorylation. It influences renal exchange of potassium and hydrogen ions and affects calcium levels. It also has an important role in nervous system control of muscle at the neuromuscular junction, where it slows neuromuscular impulse transmission by inhibiting acetylcholine. The major clinical symptoms of magnesium disorders are neuromuscular. Magnesium deficiency enhances muscle fiber excitability due to increased activity of acetyl-choline; this is manifested by muscle tremor, which can progress to seizures and tetany. Mental abnormalities include confusion, anxiety, and hallucination. Magnesium excess conversely displays antagonism of nerve impulse transmission and results in muscle weakness. Magnesium also exerts some effect on heart muscle. Decreased magnesium levels may produce or aggravate cardiac arrhythmias, whereas toxic levels of magnesium may be associated with heart block. Hypomagnesemia also potentiates the toxic effects of digitalis.

    Magnesium deficiency. Magnesium deficiency has been reported in about 10% (range, 7%-11%) of hospitalized patients. Some of the etiologies of hypomagnesemia are listed in the box. In addition, it has been reported that hypomagnesemia frequently accompanies hyponatremia (22%-27% of hyponatremic patients); hypocalcemia (22%-32% of patients); and hypophosphatemia (25%-29% of patients). Several studies also found that hypomagnesemia is frequent in patients with hypokalemia (38%-42%), but one study reported only 7%. Postoperative patients on IV feeding are reported to have a short-term, temporary 20% decrease in serum magnesium levels. Similar findings have been reported 12-24 hours after acute myocardial infarction, returning to previous levels by 48 hours, but not all studies agree.

    Excess magnesium. Increased serum magnesium levels are most often due to oliguric renal failure, which prevents excretion of magnesium. Overuse of magnesium compounds is an occasional etiology.

    Laboratory tests. RBCs contain 2-3 times the concentration of magnesium found in serum. Artifactual hemolysis thus may produce a significant increase in assay values. Skeletal muscle contains about 10 times the serum concentration. Since about 30% of serum magnesium is bound to albumin, and assays measure total magnesium levels, hypoalbuminemia will artifactually decrease serum magnesium levels.

    Magnesium Disorders
    Magnesium deficiency
    Alcoholism
    Malabsorption
    Malnutrition
    IV fluids without magnesium
    Severe diarrhea
    Diabetic ketoacidosis
    Hemodialysis
    Hypercalcemia
    Congestive heart failure
    Artifact (hypoalbuminemia)
    Certain medications
    Loop and thiazide diuretics
    Cyclosporine
    Cisplatin
    Gentamicin
    Magnesium excess
    Oliguric renal failure
    Overuse of magnesium-containing compounds
    Artifactual (specimen hemolysis)

    Various reports emphasize that serum magnesium values may not always truly reflect total body magnesium levels, since serum values may be falsely elevated in dehydration and falsely decreased in hemodilution with or without clinical edema or hypoalbuminemia. However, this problem is not unique to magnesium.

  • Hypocalcemia

    Hypocalcemia may be subdivided into nonionized hypocalcemia (decrease in serum total calcium value) and true hypocalcemia (decrease in ionized calcium value).

    Selected Etiologies of Hypocalcemia
    Artifactual
    Hypoalbuminemia
    Hemodilution
    Primary hypoparathyroidism
    Pseudohypoparathyroidism
    Vitamin D-related
    Vitamin D deficiency
    Malabsorption
    Renal failure
    Magnesium deficiency
    Sepsis
    Chronic alcoholism
    Tumor lysis syndrome
    Rhabdomyolysis
    Alkalosis (respiratory or metabolic)
    Acute pancreatitis
    Drug-induced hypocalcemia
    Large doses of magnesium sulfate
    Anticonvulsants
    Mithramycin
    Gentamicin
    Cimetidine

    The most common cause of nonionized (“laboratory”) hypocalcemia is a decrease in the serum albumin level, which lowers the total serum calcium value by decreasing the metabolically inactive bound fraction without changing the nonbound “ionized” metabolically active fraction. Therefore, this type of hypocalcemia is artifactual as far as the patient is concerned, since the metabolically active fraction is not affected. Sometimes nonionized hypocalcemia occurs with serum albumin values within the lower part of the population reference range, presumably because the previous albumin level was in the upper portion of the reference range. Although laboratory hypocalcemia is fairly common in hospitalized patients, true hypocalcemia is considerably less common than hypercalcemia. In one study, only 18% of patients with a decreased total serum calcium level had true hypocalcemia. Symptoms of decreased ionized calcium include neuromuscular irritability (Chvostek’s or Trousseau’s sign), which may progress to tetany in severe cases; mental changes (irritability, psychotic symptoms); and sometimes convulsions. Some causes of hypocalcemia are listed in the box on this page.

    Neonatal hypocalcemia. Neonatal serum calcium levels are lower than adult levels, with adult levels being attained at about 2 weeks of life for full-term infants and at about 4 weeks for premature infants. Neonates may develop hypocalcemia early (within the first 48 hours of life) or later (between age 4-30 days). Late-onset hypocalcemia can be due to a high-phosphate diet (cow’s milk), malabsorption, dietary vitamin D deficiency, alkalosis, and congenital disorders. The etiology of early-onset hypocalcemia is poorly understood. Symptoms include muscular twitching, tremor, and sometimes convulsions. Since one or more episodes of tremor or twitching are not uncommon in neonates, hypocalcemia is a rather frequent consideration in the newborn nursery. Conditions that predispose to early-onset neonatal hypocalcemia include maternal insulin-dependent diabetes, birth hypoxia, acidosis, respiratory distress, and low birth weight (usually associated with prematurity). There is a general inverse relationship between serum calcium level and birth weight or infant gestational age. Infants who are severely premature or have very low birth weight tend to develop hypocalcemia very early; in one study of such patients, one third became hypocalcemic by 15 hours after birth. In adult hypocalcemia, the diagnosis can be easily made with a serum total calcium assay if the patient has typical symptoms and if hypoalbuminemia is excluded. Ionized calcium assay may be necessary in equivocal cases. Although several formulas exist to predict ionized calcium using total calcium and serum albumin data, there is considerable disagreement in the literature whether these formulas are sufficiently accurate to be clinically useful. In one study on seriously ill adult patients, only about 20% of those who had formula-predicted ionized calcium deficit had measured ionized calcium abnormality. In newborns, serum calcium assay is much more difficult to interpret. First, neonatal calcium reference values increase with increasing gestational age, so that the reference range for prematures is different from the range for term infants. Second, there are surprisingly few data on neonatal reference ranges for calcium in the literature, and the data available are contradictory. For example, in laboratories with adult calcium reference range values of 9-11 mg/100 ml (2.25-2.75 mmol/L), the lower limit for premature infants varies in the literature from 6.0 to 8.0 mg/100 ml (1.50-2.0 mmol/L), and for full-term infants, from 7.3 to 9.4 mg/100 ml (1.83-2.35 mmol/L). If some other laboratory’s adult reference range were lower than 9-11 mg/100 ml, presumably the neonatal reference lower limit could be even lower than those quoted. High levels of bilirubin or hemoglobin (hemolysis) can affect (falsely decrease) several methodologies for serum calcium. Thus, laboratory results in possible early-onset neonatal hypocalcemia may be difficult to interpret.

    Laboratory tests

    Laboratory tests helpful in differential diagnosis are serum albumin, BUN, calcium, phosphorus, pH, and PCO2. These help to exclude hypoalbuminemia, chronic renal disease (BUN and phosphorus levels are elevated, pH is decreased), and alkalosis (respiratory or metabolic). Medication effect can be detected by a good patient history. Serum magnesium assay can exclude magnesium deficiency. If malabsorption is possible, serum carotene is a good screening test (Chapter 26). PTH assay is needed to diagnose hypoparathyroidism (PTH deficiency with decreased PTH levels) or pseudohypoparathyroidism (renal or skeletal nonresponse to PTH with increased PTH levels). N-terminal or “intact” PTH assay is better for this purpose than midregion or C-terminal assay if the patient has renal failure, since midregion and C-terminal fragments have a long half-life and thus accumulate in renal failure more than intact PTH or N-terminal fragments. If the BUN level is normal, there should be no difference between the various PTH assays.

    Vitamin D compound assay. Vitamin D is a fat-soluble steroid-related molecule that is absorbed in the small intestine. After absorption it is carried in chylomicrons or bound to an alpha-1 globulin called “transcalciferin.” Normally, about one third is metabolized to calcidiol (25-hydroxy-vitamin D) in the liver, and the remainder is stored in adipose tissue. Calcidiol is primarily regulated by the total amount of vitamin D in plasma from exogenous or endogenous sources; therefore, calcidiol is an indicator of vitamin D body reserves. Estrogen increases calcidiol formation. Calcidiol is altered to calcitriol (1,25-dihydroxy-vitamin D, the active form of vitamin D) in kidney proximal tubules by a 1-hydroxylase enzyme. Normal values decline with age. About 10% is metabolized to 24,25-dihydroxy-vitamin D by a different enzyme. As noted previously, calcitriol has actions affecting calcium availability in bone, kidney, and intestine. PTH and blood phosphate levels can influence the hydroxylase enzyme, with the effects of PTH being produced through its action on cyclic AMP.

    The vitamin D group includes two other compounds: Vitamin D2(ergocalciferol), derived from plant sources; and vitamin D3 (cholecalciferol), synthesized in the epidermis and therefore a naturally occurring form of vitamin D in humans.

    Laboratory assays for both calcidiol and calcitriol are available in some of the larger reference laboratories. These assays are useful mainly in patients with possible vitamin D overdose (hypervitaminosis D), in children with rickets, and in some adults with osteomalacia (the adult equivalent of rickets). Both osteomalacia and rickets are characterized by defective calcification of bone osteoid, and both involve some element of vitamin D deficiency.

    Vitamin D excess can produce hypercalcemia, hyperphosphatemia, soft tissue calcification, and renal failure. Calcidiol assay is the test of choice; the calcidiol level should be considerably elevated. In some patients with PHPT, the serum calcium level may be normal or borderline, and PTH assay may be equivocal. In these few patients, calcitriol assay may be useful, since it should be elevated in PHPT.

  • Serum Parathyroid Hormone-Related Protein (PTHrP)

    Since many patients (50% or more) with cancer and hypercalcemia do not have demonstrable bone metastases or PHPT, it has long been suspected that the cancer could be producing a parathyroid hormonelike substance. The parathyroid hormone-related protein (PTHrP) molecule has a C-terminal end and an N-terminal end like PTH; in addition, a portion of the PTHrP amino acid sequence is identical to that of PTH, although the majority of the PTHrP molecule is not. Also, it has been found that certain normal tissues can produce PTHrP (including the keratinized layer of skin epidermis, lactating breast tissue, placenta, adrenal, and a few others). PTHrP has recently been isolated and cloned, and antibodies have been obtained that react against it. Several investigators have reported results using homemade test kits, and one commercial kit is now available. Results thus far with these first-generation kits show that about 50% (range, 20%-91%) of patients with solid malignancies and hypercalcemia have increased PTHrP levels. Another 20% have bone metastases that could account for hypercalcemia without hormonal basis. It is currently thought that the other 30% may be producing some type of altered PTHrP that is not being detected by current antibodies. PTHrP assay may be useful when PTH assays fail to give expected results in patients with malignancy or give results that are borderline or slightly overlapping in nomogram areas between PHPT and tumor patients. However, PTHrP assays are not all alike and it is necessary to find a laboratory or kit that gives superior results.

  • Hypercalcemia and Malignancy

    In confirmed hypercalcemia, differential diagnosis is usually among PHPT, malignancy (metastatic to bone or the ectopic PTH syndrome), and all other etiologies. In most cases the differential eventually resolves into PHPT versus hypercalcemia of malignancy (HCM). There is no single laboratory test that can distinguish between PHPT and HCM every time with certainty. As noted previously, the better midmolecule PTH assays usually can differentiate normal from either PHPT or HCM and frequently can differentiate PHPT from HCM. If PHPT and HCM are not clearly separated, intact PTH assay might be obtained since it is generally better at separating PHPT and HCM. In any case a nomogram containing a scattergram of known cases is necessary. If the different PTH assays are not available, some other tests might indirectly provide evidence one way or the other. Hand x-rays are helpful if typical changes of PHPT are found (but this occurs in only a small percentage of cases). Renal stones are common in PHPT and uncommon in tumor. The quickest and easiest screening test for myeloma is serum protein electrophoresis, although serum and urine immunoelectrophoresis is more sensitive. A serum chloride value at the upper limit of the reference range or above is evidence against metastatic tumor. A bone scan and x-ray skeletal survey are useful to detect metastatic tumor. Some investigators advocate the assay of calcitonin, which is elevated with varying frequency in tumors associated with hypercalcemia and is usually not elevated in PHPT (some investigators report mild elevation in some patients). Unfortunately, regardless of the test results, PHPT may be present concurrently with malignancy in about 5% of patients with cancer.

    Serum calcitonin assay. Calcitonin (thyrocalcitonin, TCT) is secreted by nonfollicular C cells of the thyroid. An increased serum calcium level induces thyroid C cells to produce more calcitonin as part of hypercalcemia compensatory mechanisms. A major exception is PHPT, where the TCT level is usually normal or low, for poorly understood reasons (one report indicates an elevation in 10% of cases). The TCT level may be elevated in a considerable percentage of certain tumors known to metastasize to bone, such as lung carcinoma (about 30%-50% of cases; literature range 21%-62%) and breast carcinoma (about 50%; range, 38%-75%). Medullary thyroid carcinoma (MTC) produces elevated basal TCT in about 75% of cases (range, 33%-100%). Total serum calcium in MTC is usually normal. MTC or C-cell hyperplasia is found in >95% of patients with multiple endocrine neoplasia (MEN) syndromes type 2A and 2B. Type 2A also includes pheochromocytoma (about 50% cases) and PHPT (10%-25% cases). PHPT also is part of MEN type 1, which does not include MTC. The TCT level may be increased in the Zollinger-Ellison syndrome, as well as in certain nonneoplastic conditions such as chronic renal failure or pernicious anemia, and values may overlap with MCT. In summary, an elevated TCT level in a patient with possible PHPT raises the question of medullary carcinoma of the thyroid or some other malignancy, if the patient is not in renal failure.

    Ectopic parathyroid hormone syndrome.

    Nonparathyroid tumors that secrete PTH or PTH-like hormones (ectopic PTH syndrome) can produce considerable diagnostic problems. In one study, 19% of patients with tumor-associated hypercalcemia had no evidence of bone metastases. On the average, PTH assay values in ectopic PTH syndrome are lower than PTH values in PHPT. Although there is some overlap, the degree of overlap depends on the individual antiserum. There is disagreement regarding the nature of the ectopically produced hormone; that is, whether it is true PTH or a nonidentical molecule with a similar structure and PTH-like action that cross-reacts with most current PTH antisera. To further confuse matters, it is estimated that 5%-10% of patients with malignancy and hypercalcemia also will have a coexisting parathyroid adenoma with PHPT. It has also been stated that 15% of patients with PHPT have some coexisting disorder that could produce hypercalcemia.

  • Parathyroid Hormone (PTH)

    PTH is secreted in a discontinuous (pulsatile) fashion. There is a diurnal variation, with highest values at 2 A.M. (midnight to 4 A.M.) and lowest at noon (10 A.M. -2 P.M.. The parathyroids synthesize intact PTH, consisting of 84 amino acids in a single chain. Metabolic breakdown of intact PTH occurs both inside and outside of the parathyroids; outside the parathyroid, breakdown takes place in the liver and to a much lesser extent in the kidneys. This breakdown results in several fragment molecules: a small amino-terminal (N-terminal) fragment containing the PTH amino acid sequence 1-34; a larger midregion fragment containing amino acids 44-68; and a relatively large carboxy-terminal (C-terminal) fragment containing amino acids 53-84. Intact PTH and the N-terminal fragments have metabolic activity but not the midregion or C-terminal fragments. In a normal person, intact PTH constitutes 5%-15% of circulating PTH molecules. All PTH fragments are eliminated by the kidney, primarily through glomerular filtration. The measurable serum half-life of intact PTH is only about 5 minutes; that of the N-terminal fragment is about 2-3 minutes; and that of the C-terminal fragment is about 30 minutes. Renal function impairment will decrease elimination of the C-terminal fragment and also to a lesser extent the N-terminal fragment. In renal failure the C-terminal half-life lengthens to 24-36 hours and the N-terminal half-life lengthens to 30 minutes.

    PTH is measured by immunoassay. Original methods were based on antibodies against either the N-terminal fragment or the C-terminal fragment. Current tests use antibody against synthetic portions of the PTH chain, resulting in somewhat better sensitivity and reliability. These tests primarily detect either the C-terminal, the midregion fragment, or the intact PTH molecule. Actually, the C-terminal and midregion assays detect more fragments than the principal one indicated by their name:

    Assay
    Intact PTH
    N-terminal
    C-terminal
    Midregion (sometimes called “total” PTH)

    Assay includes
    Intact PTH only
    N-terminal fragment Intact PTH
    C-terminal fragment Intact PTH
    Midregion combined with C-terminal fragment
    Midregion fragment Intact PTH
    Midregion combined with C-terminal fragment

    At present, midregion assays have generally been more sensitive in detecting PHPT and separating PHPT from normal persons than C-terminal or intact PTH assays have been. Although there is considerable variation in reported sensitivity due to different kit antibodies and other technical factors, the best midregion kits are claimed to achieve 95% or greater sensitivity in detecting primary hyperparathyroidism. However, they are generally not as good in differentiating hypercalcemia due to PHPT from hypercalcemia of malignancy (the midregion PTH levels of 20%-25% of these cancer patients are normal or sometimes slightly increased rather than suppressed below reference range by the hypercalcemia). Intact PTH, on the other hand, generally is best at separating PHPT and hypercalcemia of malignancy (the PTH values of cancer patients are usually below intact PTH reference range or are in the lower part of the range, whereas the levels of PHPT patients are elevated or in the upper part of the range). The best intact assays are reported to detect PHPT almost as well as the better midmolecule assays. Intact PTH is also more reliable in patients with poor renal function. In azotemia, serum C-terminal and midregion fragments increase much more than intact PTH because of decreased excretion by the diseased kidneys.

    In some cases, detection of abnormality can be assisted by correlating PTH values with serum calcium values. PTH values may be within the upper part of the reference range but may still be higher than expected for the degree of serum calcium elevation. A PTH/calcium nomogram should be constructed for each PTH antiserum.

    Parathyroid hormone assay interpretation.

    Among diseases associated with hypercalcemia, PTH values are elevated in PHPT, in most cases of ectopic PTH syndrome, and in most cases of tertiary hyperparathyroidism. The actual percentage of elevated results in each category varies with the particular antiserum used (e.g., 8%-73% of PHPT patient values have been reported to be within the reference range with different anti-sera). In metastatic carcinoma to bone, the PTH value is normal with the majority of antisera, but there is overlap with PHPT in a significant minority of patients with nearly all antisera (the exact percentage varying with the particular antiserum). Parathyroid hormone values are usually normal or decreased in other conditions producing hypercalcemia.

    Parathyroid hormone values are elevated in many (but not all) conditions associated with true hypocalcemia (false hypocalcemia from hypoalbuminemia must be excluded). These include osteomalacia, vitamin D deficiency of dietary origin and in some patients with malabsorption, renal failure, and pseudohypoparathyroidism (congenital nonresponse of kidney to PTH). In PHPT, serum PTH and serum calcium levels are both increased.

    There are additional factors in PTH assay interpretation. PTH has a diurnal variation, with the lowest values (trough, nadir) about noon (10 A.M.-2 P.M.) and the peak about 2 A.M. (midnight-4 A.M.). Specimens should be drawn when patients are fasting at about 7-8 A.M. without using specimen anticoagulants. Specimens should be processed at cold temperatures, frozen immediately, and transported in dry ice. Assay of PTH at present is difficult. Reliable antisera are not yet readily available from commercial sources, and, as noted previously, homemade antisera in reference laboratories differ in reactivity.

    Problems and some solutions in PTH assay. Theoretically, any PTH assay should differentiate parathyroid tumor from various other etiologies of hypercalcemia, since PHPT should have increased PTH serum levels and hypercalcemia of all other etiologies should show decreased PTH secretion. Unfortunately, when tested with presently available antisera, some patients with PHPT may have PTH values within laboratory reference range (5%-73% in reports from different laboratories). In addition, some patients with hypercalcemia not due to PHPT may have values that are within the reference range rather than decreased. In most publications, results in hypercalcemia of malignancy fall within the reference range or below, but a few antisera permit some elevated values. In most reports there is substantial overlap between patients with PHPT and patients with hypercalcemia of malignancy when their values fall within the reference range, averaging about 10%-20% (literature range, 0%-73%). Diagnosis of parathyroid adenoma can be assisted by correlating PTH assay with serum calcium levels, based on the fact that the PTH level normally decreases as the serum calcium level increases. A parathyroid adenoma may produce a PTH level that is within population normal range but is higher than expected in relation to the degree of calcium elevation. A nomogram should be constructed for each PTH antiserum, correlating PTH values with serum calcium values obtained from patients with surgically proved PHPT. This nomogram also should provide data on PTH and calcium findings in other calcium-phosphorus disorders, such as metastatic carcinoma to bone, ectopic PTH syndrome, myeloma, and renal disease. The nomogram may permit separation of these conditions when it would be impossible with the numerical values alone. For example, in one report 45% of PHPT patients had PTH values within the PTH reference range; but with the use of the nomogram, 87% of the PHPT patients could be separated from normal persons. Therefore, with the majority of antisera, such a nomogram is almost essential when PTH values are interpreted, especially since results from different reference laboratories on the same patient specimens have shown considerable difference in behavior among different antisera when tested on patients with calcium disorders. These differences in results exist not only between C-terminal, midregion, and N-terminal categories of antisera but also between individual antisera within the same category. Use of a nomogram can reduce overlap between PHPT and malignancy to 5%-10% (literature range, 0%-15%).

    If the serum albumin level is low, the calcium (total calcium) level will be falsely decreased and could alter patient position in the nomogram. Correction of the calcium-albumin relationship by formula may help but may not be accurate. Also, the diagram block areas appear to clearly separate different categories of calcium disease, but, in fact there may be overlap between patient values in certain disorders, and the amount of overlap is different with each individual antiserum. To choose the best laboratory for PTH assay, I strongly suggest that each laboratory under consideration be required to supply a nomogram for each type of PTH assay they perform showing actual values from patients with proven calcium diseases, including hypercalcemia of malignancy, plotted on the diagram in the form of individual symbols, the symbols (dots, circles, triangles) representing the different diseases. It is necessary to have a substantial number of patient results in each disease category, especially in both PHPT and malignancy, to obtain an accurate picture. This way, it is possible to obtain a more meaningful comparison of actual PTH test results in different laboratories. The best PTH assay is one that not only clearly separates different diseases from the reference range but also has the least overlap between disease categories, especially in the area between PHPT and hypercalcemia of malignancy.

    If a patient has significant hypoalbuminemia, it may be better to ask for measurement of ionized calcium (which is not affected by the albumin level as total calcium is) and a PTH-calcium nomogram using ionized calcium and PTH values. The nomogram should have a scattergram of known PHPT and hypercalcemia of malignancy cases, not an empty block diagram only.

    Lesser used or historically important tests

    Tubular reabsorption of phosphate (phosphate reabsorption index). This procedure indirectly measures PTH by estimating PTH action on renal phosphate reabsorption. The patient should be on a normal phosphate (PO4) diet; a low-PO4 diet (<500 mg/day) raises tubular reabsorption of PO4 (TRP) normal values, whereas a high-PO4 diet (3,000 mg/day) lowers TRP normal values limits.

    The patient drinks several glasses of water and then voids completely. One hour after voiding, a blood sample is obtained for phosphorus and creatinine measurement. Exactly 2 hours after beginning the test, the patient again voids completely, and the urine volume and urine concentration of creatinine and phosphate are determined. It is then possible to calculate the creatinine clearance rate and find the amount of phosphorus filtered per minute by the glomeruli. Comparing this with the actual amount of phosphate excreted per minute gives the amount reabsorbed by the tubules per minute, or the TRP value. A rough approximation is afforded by the formula:

    %TRP = [1 – (UrinePO4 x serum creatinine /  Urine creatinine x serum PO4)]

    An index value of less than 80% means diminished TRP value and suggests PHPT. This test becomes increasingly unreliable in the presence of renal insufficiency. About 5% of patients with renal stones but without parathyroid tumor have TRP values of 70%-80%, whereas about 20% of patients with parathyroid tumors have normal TRP values. Therefore, a TRP reduction is more significant than a normal result. Hypercalcemia due to malignancy is usually associated with a decreased TRP value. In addition, some patients with other conditions such as sarcoidosis and myeloma have been reported to have reduced TRP values.

    X-ray findings. Bone changes highly suggestive of hyperparathyroidism may be found radiologically in about 15% of PHPT patients (literature range, 9%-36%), although the older literature reports some type of change in up to 46% of cases with skeletal surveys. The incidence of bone change has considerably decreased because of earlier diagnosis. The most typical findings are subperiosteal cortical bone resorption in the phalanges. Patients with chronic renal disease (secondary or tertiary hyperparathyroidism) may also demonstrate these abnormalities but do not have elevated serum calcium levels (except in tertiary hyperparathyroidism, in which case there should be obvious long-term renal failure). Serum alkaline phosphatase elevation in PHPT is highly correlated with the presence of bone changes. It would be unlikely to find skeletal changes in hand x-ray films if the serum alkaline phosphatase level is not elevated. Of course, the serum alkaline phosphatase level could be elevated for a variety of reasons in any individual patient with hypercalcemia.

    Serum chloride. Primary hyperparathyroidism tends to develop a hyperchloremic acidosis. Serum chloride is often elevated in PHPT (40%-50% of cases, if one excludes patients with conditions that lower serum chloride levels such as vomiting, diarrhea, or diuretic use). Less than 10% of patients with non-PHPT etiologies of hypercalcemia have elevated serum chloride levels. In one series, these were all patients with thyrotoxicosis or the ectopic PTH syndrome. A chloride/phosphorus ratio has also been proposed. This was found to be greater than 33 in about 94% of PHPT patients (without renal failure). However, results in other hypercalcemias have been variable, with the percentage of patients reported with a ratio greater than 33 having ranged from 4%-39%.