TABLE 142-4.
Tests for the Evaluation of Malabsorption*
| TEST | COMMENTS |
|---|---|
| GENERAL TESTS OF ABSORPTION | |
| Quantitative stool fat test | Gold standard test of fat malabsorption, with which all other tests are compared. Requires ingestion of a high-fat diet (100 g) for 2 days before and during the collection. Stool is collected for 3 days. Normally, <7 g/24 hr is excreted on a high-fat diet. Borderline abnormalities of 8-14 g/24 hr may be seen in secretory or osmotic diarrheas that are not caused by malabsorption. There are false-negative findings if fat intake is inadequate. False-positive results can occur if mineral oil laxatives or rectal suppositories (e.g., cocoa butter) are given to the patient before stool collection. |
| Qualitative stool fat test | Sudan stain of a stool sample for fat. Many fat droplets per medium-power field (×40) constitute a positive test result. The nuclear magnetic resonance method determines the percentage of fat in the stool (normal, <20%). The test depends on an adequate fat intake (100 g/day). There is high sensitivity (90%) and specificity (90%) with fat malabsorption of >10 g/24 hr. Sensitivity drops with stool fat in the range of 6-10 g/24 hr. |
| d-Xylose test | A test of small intestinal mucosal absorption, used to distinguish mucosal malabsorption from malabsorption due to pancreatic insufficiency. An oral dose of d-xylose (25 g/500 mL water) is administered, and d-xylose excretion is measured in a 5-hr urine collection. Normally, >4 g of d-xylose is excreted in the urine over 5 hr. The test also may be positive in bacterial overgrowth owing to metabolism of d-xylose by bacteria in the intestinal lumen. False-positive test results occur with renal failure, ascites, and an incomplete urine collection. Blood levels at 1 and 3 hr improve sensitivity. May be normal with mild or limited mucosal disease. |
| Hydrogen breath test | Most useful in the diagnosis of lactase deficiency. An oral dose of lactose (1 g/kg body weight) is administered after measurement of basal breath H2 levels. The sole source of H2 in the mammal is bacterial fermentation; unabsorbed lactose makes its way to colonic bacteria, resulting in excess breath H2. A late peak (within 3-6 hr) of >20 ppm of exhaled H2 after lactose ingestion suggests lactose malabsorption. Absorption of other carbohydrates (e.g., sucrose, glucose, fructose) also can be tested. |
| SPECIFIC TESTS FOR MALABSORPTION | |
| Tests for Pancreatic Function | |
| Secretin stimulation test | The gold standard test of pancreatic function. Requires duodenal test intubation with a double-lumen tube and collection of pancreatic juice in response to IV secretin. Allows measurement of bicarbonate (HCO3−) and pancreatic enzymes. A sensitive test of pancreatic function, but labor intensive and invasive. |
| Fecal elastase-1 test | Stool test for pancreatic function. Equal sensitivity to the secretin stimulation test for the diagnosis of moderate-to-severe pancreatic insufficiency. More specific than the fecal chymotrypsin test. Unreliable with mild insufficiency. False-positive results occur with increased stool volume and intestinal mucosal diseases. |
| Tests for Bacterial Overgrowth | |
| Quantitative culture of small intestinal aspirate | Gold standard test for bacterial overgrowth. Greater than 105 colony-forming units (CFU)/mL in the jejunum suggests bacterial overgrowth. Requires special anaerobic sample collection, rapid anaerobic and aerobic plating, and care to avoid oropharyngeal contamination. False-negative results occur with focal jejunal diverticula and when overgrowth is distal to the site aspirated. |
| Hydrogen breath test | The 50-g glucose breath test has a sensitivity of 90% for growth of 105 colonic-type bacteria in the small intestine. If bacterial overgrowth is present, increased H2 is excreted in the breath. A hydrogen level (within 2 hr) of >20 ppm suggests bacterial overgrowth. False-negative results occur with non-hydrogen-producing organisms. |
| 14C-d-xylose breath test | This test uses 1 g of carbon 14–labeled d-xylose. It has a sensitivity and specificity >90% for growth of 105 test colonic-type bacteria in the small intestine. Bacteria metabolize d-xylose with release of 14CO2, which is absorbed and exhaled. Non-degraded d-xylose is absorbed in the small bowel and does not reach the colon, yielding a greater specificity than the lactulose H2 breath test. A nonradioactive 13C-d-xylose breath test is suitable for children and pregnant women. |
| Tests for Mucosal Disease | |
| Small bowel biopsy | Obtained for a specific diagnosis when there is a high index of suspicion for small intestinal disease. Several biopsy specimens (4-5) must be obtained to maximize the diagnostic yield. Distal duodenal biopsy specimens are usually adequate for diagnosis, but occasionally enteroscopy with jejunal biopsy specimens is necessary. Small intestinal biopsy provides a specific diagnosis in some diseases (e.g., intestinal infection, Whipple's disease, abetalipoproteinemia, agammaglobulinemia, lymphangiectasia, lymphoma, amyloidosis). In other conditions, such as celiac disease and tropical sprue, the biopsy specimens show characteristic findings, but the diagnosis is made on improvement after treatment. |
| Tests of Ileal Function | |
| Schilling test | A test of vitamin B12 absorption (see Table 167-1 in Chapter 167). |
| 75SeHCAT test | This is a test of bile acid absorption. Seven days after ingestion of radiolabeled synthetic selenium-homocholic acid conjugated with taurine (75SeHCAT), whole body retention is measured by a gamma-counting device. The result is expressed as a fraction of baseline ingestion. Retention values of less than 10% are abnormal and indicate bile acid malabsorption with a sensitivity of 80-90% and specificity of 70-100%. The radiation dose is equivalent to a plain chest x-ray. Liver disease and bacterial overgrowth may give false results. Not approved for use in the United States. |
*
Not all these tests are readily available. A strong suspicion for any disease may warrant foregoing an extensive work-up and obtaining the test with highest diagnostic yield. In some cases, empirical treatment, such as removing lactose from the diet of an otherwise healthy individual with lactose intolerance, is warranted without any testing.