Chronic Diarrhea:

December 1, 2005
Angela A. Carollo, MD

,
Lawrence R. Schiller, MD

Chronic diarrhea presents difficulties for clinicians as well as for patients. Because the differential diagnosis is enormous, management can be challenging. In this article, we present a strategy for quickly narrowing the differential based on a simple analysis of stool characteristics. We then describe an appropriate workup for each of the basic types of diarrhea.

Chronic diarrhea is defined as diarrhea that lasts for more than 4 weeks and that often persists unless therapy is instituted (unlike acute diarrhea, which is usually self-limited). This common problem has a negative impact on patients' quality of life and results in considerable health care costs.1-3

Chronic diarrhea presents difficulties for clinicians as well as for patients. Because the differential diagnosis is enormous, management can be challenging. In this article, we present a strategy for quickly narrowing the differential based on a simple analysis of stool characteristics. We then describe an appropriate workup for each of the basic types of diarrhea.

PATHOPHYSIOLOGY OF DIARRHEA

The fundamental pathophysiology of all diarrheal illnesses is incomplete absorption of water from the lumen. This can result from one of two mechanisms:

  • Reduced rate of net water absorption (related to impaired electrolyte absorption or excessive electrolyte secretion).

  • Osmotic retention of water intraluminally.4

In reality, however, the pathophysiology of diarrhea is often mixed. Although incomplete absorption is now regarded as the basic mechanism of the condition, rapid transit (which has traditionally been cited as its cause) is often an important factor in its development. Motility changes occur in many diarrheal diseases, either as a result of increased intraluminal volume or as motor effects of the same mediators that alter transport. By speeding intraluminal contents past absorption sites, rapid transit may be a major contributor to reduced net absorption.4 In many patients with diarrhea, several mechanisms are at work simultaneously. For example, the diarrhea associated with inflammatory bowel disease may be caused by a combination of inflammatory mediators, microbial toxins, and alterations in both mucosal permeability and motility.

CLASSIFICATION OF CHRONIC DIARRHEA BY STOOL TYPE

It is helpful to divide diarrhea into 3 broad categories based on stool characteristics:

  • Watery.

  • Fatty.

  • Inflammatory.

Watery diarrhea can be further divided into secretory and osmotic diarrhea. In secretory diarrhea, the excess stool water results from inadequate electrolyte and fluid absorption. The excess stool water in osmotic diarrhea is the result of luminal water retention caused by the presence of poorly absorbed substances.

Watery diarrhea. This is characterized by fluid stools that typically are pourable. No blood, pus, or fat is evident. Usually, the patient is either afebrile or only minimally febrile. Watery diarrhea can be characterized as either osmotic or secretory by fecal electrolyte analysis and calculation of the fecal osmotic gap (FOG).

Figure

The FOG is a measure of the nonelectrolyte contribution to the intraluminal osmolality.5 The Figure shows the typical fecal electrolytes found in secretory and osmotic diarrhea. Intraluminal osmolality is equal to plasma osmolality (290 mOsm/kg). The electrolyte contribution = 2 ([Na] + [K]). (The doubling accounts for anions as well as cations.) FOG = 290 2 2 ([Na] + [K]). In osmotic diarrhea, the electrolyte concentrations are low and the FOG is high (greater than 50 mOsm/kg). In secretory diarrhea, the electrolyte concentrations are high and the FOG is low (less than 50 mOsm/kg).5,6

Fatty diarrhea. In diarrhea of this type, gross steatorrhea may be present. The patient may actually see oil droplets in the commode. In the absence of obvious steatorrhea, fatty diarrhea may be identified by a positive Sudan stain or quantitative fat analysis. Other clues to the presence of fatty diarrhea are weight loss and gaseousness.

Inflammatory diarrhea. This is usually associated with the presence of blood and pus. Results of a fecal occult blood test are typically positive, and a fecal leukocyte count is elevated. The patient may be febrile and may complain of tenesmus. The stool weight is moderate.

INITIAL EVALUATION

In any patient with chronic diarrhea, it is important to take a thorough history and to perform a physical examination. Begin by ascertaining the onset and duration of the condition and whether it has been continuous.

Table 1

Table 2

Table 3

Ask about the pattern of bowel movements and the characteristics of the stool. Keep in mind that what patients mean when they say they have "diarrhea" varies considerably--and may not correlate with the scientific understanding of the term. Many complain of diarrhea when their stool is loose in consistency, while others complain only when their stool frequency increases.7,8 The traditional scientific definition of diarrhea is an increase in stool weight to more than 200 g/d; however, the stool weight in some patients may be normal despite loose stools.9 Moreover, some patients complain of diarrhea when they are actually experiencing fecal incontinence; be sure to rule out the latter.10 (A lax anal sphincter on rectal examination may be an important clue to the presence of fecal incontinence.) Elicit information that might help you categorize a patient's stools as watery, fatty, or inflammatory.

Note associated symptoms, such as weight loss or abdominal pain. Inquire about epidemiologic risk factors, such as travel or consumption of unsanitary food or water, and note any aggravating or mitigating factors, such as diet or medications. The presence of systemic disease or a history of previous surgery or radiation exposure may also help elucidate the ultimate diagnosis.

On physical examination, look for clues such as flushing, abdominal tenderness, or edema. Initial laboratory tests should include a complete blood cell count and chemistry panel.

APPROACHES TO INITIAL MANAGEMENT

After the initial evaluation of a patient with chronic diarrhea, the clinician may choose from 3 different approaches to management:

  • Empiric therapy.

  • Test and treat.

  • Categorize, test, and treat.11

Empiric therapy. This strategy bypasses the step of establishing a specific diagnosis and skips to treating the diarrhea with either nonspecific antidiarrheal agents or a series of more specific treatments. Empiric therapy is only feasible if life-threatening and specifically treated conditions can be excluded by history, physical examination, and possibly, simple tests. This approach relies on the assumption that the diarrhea most likely is caused only by functional or self-limited conditions. It must be used with caution, since diarrhea that seems to be functional or self-limited may, in fact, be specifically treatable. Because it may result in the delay of appropriate, potentially curative treatment, empiric therapy is potentially perilous for the patient.11

Test and treat. This is the traditional approach to the evaluation and management of chronic diarrhea; it involves the development of a differential diagnosis, followed by testing for each possibility until a diagnosis is made. Specific treatment then can be initiated. In most patients with diarrhea, this approach is impractical because the differential diagnosis is so extensive. However, if the pretest probability of a certain diagnosis is high and the diagnostic tests involved have high sensitivity and specificity, the "test and treat" approach may be appropriate. An example in which this approach might be useful is a patient who presents with diarrhea, flushing, a heart murmur, and a large liver. The pretest probability of a diagnosis of carcinoid is high; thus, proceeding with specific testing to support this diagnosis would be a reasonable next step.11

Categorize, test, and treat. This is the preferred approach in most patients with chronic diarrhea. In this strategy, a few preliminary tests are performed to help narrow the differential diagnosis. This approach is used frequently because the results of the history taking and physical examination are often nonspecific. The first step in the "categorize, test, and treat" approach is to characterize the patient's diarrhea as watery, fatty, or inflammatory.

Stool analysis. Characterization of diarrhea usually requires formal stool analysis. Stool analysis should include:

  • Fecal weight.

  • Fecal electrolyte analysis (including calculation of FOG).

  • pH.

  • Occult blood test.

  • Fecal leukocyte count.

  • Either qualitative or quantitative fecal fat analysis.

  • Measurement of magnesium output.

Consider including a laxative screen in patients at risk for laxative abuse. Ideally, also perform a quantitative stool collection (typically lasting 48 hours); however, if this is not feasible, a spot stool collection often can provide much of the information needed. Tips on how to perform several of these tests appear in the Box.

Use the results of these tests in combination with the history and physical examination findings to characterize the diarrhea as watery, fatty, or inflammatory and to develop a differential diagnosis. Once the diarrhea has been categorized, the differential diagnosis becomes more manageable (Tables 1, 2, and 3) and a more focused series of investigations can be done.11

Workup of secretory diarrhea.

Secretory diarrhea may be caused by congenital syndromes, bacterial toxins, ileal bile acid malabsorption, inflammatory bowel disease, vasculitis, drugs and poisons, stimulant laxatives, or disordered regulation (resulting from neuropathy). It may also represent endocrine diarrhea (caused by hormones produced by tumors), or idiopathic secretory diarrhea.

4

Algorithm I

Algorithm II

Algorithm III

Algorithm IV

The first step in the workup of secretory diarrhea is to exclude infection by using stool cultures (Algorithm I). Next, exclude structural diseases with either direct endoscopic imaging, which also allows for biopsies, or with computerized tomography and/or small-bowel radiographs. If the diagnosis is still unclear, you may want to try an empiric trial of cholestyramine for bile acid diarrhea--or you can look for rare causes of secretory diarrhea, such as VIPomas or carcinoids.

Workup of osmotic diarrhea. Osmotic diarrhea may be caused by carbohydrate malabsorption or ingestion of poorly absorbed ions, such as magnesium. Distinguishing between these causative mechanisms can be done fairly easily on the basis of the fecal pH and the magnesium output (Algorithm II). A low pH is associated with carbohydrate malabsorption. A dietary review, breath hydrogen test, or mucosal lactase assay can be used to confirm lactose intolerance. Also consider ingestion of nonabsorbable carbohydrates, such as sorbitol, which is found in some sugarless gum. Osmotic diarrhea caused by magnesium ingestion is associated with a high fecal magnesium output. Excessive magnesium ingestion may be either inadvertent (eg, the result of antacid use) or deliberate (as in surreptitious laxative abuse).

Workup of fatty diarrhea. Fatty diarrhea may be caused by malabsorption or maldigestion. Malabsorption can be the result of mucosal disease, short bowel syndrome, postresection diarrhea, small-bowel bacterial overgrowth, or mesenteric ischemia. Maldigestion can result from pancreatic exocrine insufficiency or inadequate luminal bile acid concentration.4

In fatty diarrhea, it is important to exclude structural disease (Algorithm III). This can be done with small-bowel radiographs and/or CT scans. To exclude bacterial overgrowth, consider endoscopy with small-bowel biopsy and aspiration for quantitative culture. The role of capsule enteroscopy has not yet been defined in this setting.

Pancreatic exocrine insufficiency is another possible cause of fatty diarrhea. Often, an empiric trial of pancreatic enzymes can establish this diagnosis. If you choose this approach, administer a high dose of enzymes initially and monitor fat output to establish effectiveness objectively. Measurement of stool chymotrypsin activity or a secretin test can also be used to test for pancreatic exocrine insufficiency.

Workup of inflammatory diarrhea. Inflammatory diarrhea may be caused by inflammatory bowel disease, invasive infections, ischemia, radiation enteritis, or neoplasms.4 Here, too, it is important to exclude structural disease, such as inflammatory bowel disease (Algorithm IV). This can be done in a number of ways, including endoscopy, CT scans, and small-bowel radiographs. Capsule enteroscopy may also have a role in this setting. In addition, look for and exclude infections.

IF THE DIAGNOSIS REMAINS ELUSIVE

Most of the time, the "classify, test, and treat" approach results in a diagnosis. When it does not, consider "difficult-to-diagnose" diarrhea, also referred to as diarrhea of obscure origin (DOO). The differential diagnosis of DOO is fairly brief and includes both common problems that are easily overlooked and rare syndromes (Table 4).4

Table 4

Fecal incontinence. This fairly common problem may present as diarrhea. Many patients are hesitant to discuss it because of embarrassment. However, it is important to ask about fecal incontinence because the condition requires further evaluation. Effective therapy exists, and no patient should suffer with fecal incontinence needlessly.

In addition, it is important to differentiate fecal incontinence from overflow leakage around a fecal impaction. The latter condition should be excluded by rectal examination, especially in elderly, bedbound patients.

Autonomic neuropathy. Diarrhea can result from dysregulation of either absorption/secretion or motility. Diarrhea associated with neuropathy is seen in patients with diabetic enteropathy or amyloidosis. Diabetic enteropathy may be more likely in patients with other complications of poorly controlled diabetes. An investigation for amyloidosis may be warranted in patients with diarrhea and evidence of heart failure, autonomic neuropathy, or renal disease of unknown origin.

Drug-induced diarrhea. A high index of suspicion is required to make the diagnosis of drug-induced diarrhea. Take a thorough drug history, including use of over-the-counter agents and herbal preparations. Establishing causation often is as simple as stopping the suspected drug or reducing the dosage. Some of the drug classes that can cause diarrhea include:

  • Antibiotics.

  • Antineoplastic agents.

  • Antiarrhythmics.

  • Antihypertensives.

  • Antacids.

  • H2-receptor antagonists.

  • Proton pump inhibitors.

  • Prostaglandins.

Surreptitious laxative ingestion. This should be suspected in patients with eating disorders, in patients with secondary gain from the "sick" role, in those with Munchausen syndrome, and in Polle syndrome (poisoning by a caregiver, usually to create an opportunity to prove how caring he or she is). Assaying stool for laxatives may establish the diagno-sis. Historically, room searches were used to find the offending agent, but this approach is no longer considered appropriate.12 When you suspect laxative abuse, confirm the diagnosis, confront the patient, and then consult a psychiatrist to arrange for evaluation of the patient.

Small-bowel bacterial overgrowth (SBBO). Consider SBBO in patients who have areas of small-intestine stagnation and dysmotility--for example, patients who have had gastrojejunostomies or other gastroenteric surgeries. Other disorders that may be associated with bacterial overgrowth include scleroderma, radiation enteropathy, adhesions, lymphoma, and tuberculosis. Conditions that predispose to hypochlorhydria, because they interfere with the gastric acid barrier to survival of swallowed bacteria, may also predispose to SBBO.13

Microscopic colitis. This diagnosis may be difficult to make. Grossly, the colonic mucosa appears normal; however, biopsies reveal one of two histologic abnormalities: lymphocytic or collagenous colitis.14 The diarrhea results from decreased colonic water and electrolyte absorption. Microscopic colitis is a common cause of secretory diarrhea, especially in older patients. Clinical clues to the diagnosis include associated rheumatologic disease, celiac sprue, and fecal incontinence. Stool weight is typically about 500 g/24 h (always less than 1000 g/24 h). To exclude microscopic colitis in a patient with chronic diarrhea, biopsy specimens must be obtained from normal-appearing colonic mucosa.

Carbohydrate malabsorption. This is another very common cause of DOO. Clinically, carbohydrate malabsorption presents as episodic diarrhea with gas. Stools are acidic, and a fecal osmotic gap is present (greater than 50 mOsm/kg, but usually less than 100 mOsm/kg). Breath tests may be inaccurate; thus, to make the diagnosis, a dietary history is essential. Commonly ingested carbohydrates that may be malabsorbed include lactose, fructose, sorbitol, and fiber supplements.

Bile acid diarrhea. Bile acids are normally absorbed by the ileum. Thus, consider this diagnosis in patients who have ileal disease or who have had a bowel resection.15 Cathartic concentrations of bile acids may be present in the colon. This type of diarrhea abates with cholestyramine therapy or use of other bile acid binders. The role of bile acid malabsorption in idiopathic diarrhea (without ileal disease or bowel resection) is somewhat controversial. Most patients with chronic idiopathic diarrhea have demonstrable bile acid malabsorption, but they are not necessarily helped by cholestyramine. Thus, bile acid malabsorption may be less important than previously thought in this group of patients.16-18

Pancreatic exocrine insufficiency. Clinical clues to the possibility of this diagnosis include known alcoholism as well as recurrent bouts of abdominal pain of unknown origin. In addition, weight loss is usually present. Failure to diagnose pancreatic exocrine insufficiency is often the result of an inadequate empiric trial of pancreatic enzyme replacement.

Endocrine tumors. These lesions are rare but can be a cause of DOO. Such tumors can usually be classified based on the type of hormone they produce. Examples of endocrine tumors include carcinoid, gastrinoma, VIPoma, medullary carcinoma of the thyroid, somatostatinoma, pheochromocytoma, mastocytosis, and glucagonoma. It is of little diagnostic help to measure any of the peptides produced by these tumors because the pretest probability of finding a peptide-secreting tumor is initially very low. Instead, the diagnosis depends mainly on the use of imaging procedures such as high resolution CT or radio-labeled octreotide scanning, which can locate a tumor that may be secreting the offending hormone.4

Idiopathic secretory diarrhea. This is diagnosed when a patient presents with chronic secretory diarrhea, persistent loose stools, no previous GI surgery or systemic disease, and a negative diagnostic evaluation.8 This disorder tends to develop in previously healthy persons, is of abrupt onset, and is associated with frequent, watery stools (mean, 10/d; range, 5 to 25/d) and moderate weight loss soon after onset.4 Epidemiologic factors associated with idiopathic secretory diarrhea include recent travel, often to domestic recreation areas; rarely, it is seen in the household contacts of a person with primary disease. Although the epidemiology suggests an infectious cause, patients show no response to antibiotics. The clinical course is spontaneous resolution with no recurrence.19 The mean duration is 15 months (range, 7 to 31 months). An epidemic form of this type of diarrhea is known as Brainerd diarrhea.20,21