OSMOTIC DIARRHEA

Ingestion of poorly absorbed cations and anions or poorly absorbed sugars or sugar alcohols (e.g., mannitol, sorbitol) accounts for most osmotic diarrheas.⁹ Ions that are poorly absorbed include magnesium, sulfate, and phosphate. These ions are transported actively by mechanisms that are saturated at low intraluminal ion concentrations and passively by mechanisms that are slow. Together, these processes limit total absorption to a fraction of the amount that can be ingested. Because neither the small intestine nor colon can maintain an osmotic gradient, unabsorbed ions (and their counter ions) that remain in the intestinal lumen obligate retention of water to maintain an intraluminal osmolality equal to that of body fluids (about 290 mOsm/kg). Therefore, approximately 3.5 mL of water (1000 mL/kg divided by 290 mOsm/kg) are retained for every 1 mOsm of retained ions or molecules.^9–12

Sugars and sugar alcohols are the other category of substances that cause osmotic diarrhea.¹² Monosaccharides, but not disaccharides, can be absorbed intact across the apical membrane of the intestine. When disaccharides such as sucrose and lactose are ingested, absence of the appropriate disaccharidase will preclude absorption of the disaccharide or its component monosaccharides (see Chapter 101). The most common clinical syndrome of disaccharidase deficiency is acquired lactase deficiency, which accounts for lactose intolerance in many adults.¹³ Lactase is present in the brush border of the small intestine of most immature mammals but disappears in adult mammals, including 70% of adult humans.¹⁴ The main exceptions are persons from the northern European gene pool, who typically maintain lactase activity into adult life. Lactase activity often falls with age even in this group, however. Congenital deficiency of lactase is rare and seems to be the result of a mutation in a gene distinct from that for lactase-phlorizin hydrolase (the gene affected in adult lactase deficiency).¹⁵ Acquired deficiencies also may be associated with mucosal diseases of the upper small intestine. Congenital sucrase and trehalase deficiencies are rare and prevent the adequate digestion of sucrose (table sugar) and trehalose (a sugar found in mushrooms), respectively. Lactulose is a synthetic disaccharide that cannot be hydrolyzed by the human intestine and is not absorbed intact in more than trace amounts. It thereby causes an osmotic diarrhea when given in sufficient quantity to overwhelm the metabolic capacity of colonic bacteria (about 80 g/day).

The essential characteristic of osmotic diarrhea is that it disappears with fasting or cessation of ingestion of the offending substance. This characteristic has been used clinically to differentiate osmotic diarrhea from secretory diarrhea, which typically continues with fasting. Electrolyte absorption is not impaired in osmotic diarrhea, and electrolyte concentrations in stool water are usually low.^10–12

SECRETORY DIARRHEA

Secretory diarrhea has many causes, and the mechanism of this type of diarrhea is always net secretion of anions (chloride or bicarbonate) or net inhibition of sodium absorption.¹⁶ The stimuli for secretion arise from the intestinal lumen, subepithelial space, or systemic circulation and substantially alter the messenger systems that regulate ion transport pathways. In some cases, congenital absence of a specific transport molecule limits sodium or chloride absorption and results in diarrhea; in others, lack of sufficient absorptive surface area limits electrolyte, particularly sodium, absorption critically.

The most common cause of secretory diarrhea is infection.¹⁶ Enterotoxins from a host of infectious agents (primarily bacteria but also parasites and viruses) interact with receptors that modulate intestinal transport and lead to increased anion secretion. Enterotoxins also may block specific absorptive pathways in addition to stimulating secretion. Most enterotoxins inhibit Na⁺-H⁺ exchange in the small intestine and colon, thereby blocking one of the important driving forces for electrolyte and fluid absorption.^17,18

Peptides, such as vasoactive intestinal peptide, produced by endocrine tumors, cause secretory diarrhea by stimulating secretion by epithelial cells, as do peptides released from subepithelial neurons and inflammatory cells (see Chapter 32).¹⁹ Neurotransmitters such as acetylcholine or serotonin (5-hydroxytryptamine, 5-HT) and other modulators such as histamine and inflammatory cytokines also are potent secretory stimuli.^20,21 Most of these endogenous regulators of intestinal transport elicit diarrhea by altering intracellular messengers, such as cyclic adenosine monophosphate (cAMP), cyclic guanosine monophosphate (cGMP), and calcium, that control specific transport pathways.²² In addition, peptides and other regulators may affect the synthesis, localization, and degradation of individual transport proteins. Exogenous agents, such as drugs and some poisons, lead to secretory diarrhea, presumably by interacting with intracellular regulators or intracellular messengers of the enterocytes.

The absence or disruption of a specific absorptive pathway may cause diarrhea. For example, rare congenital syndromes, such as congenital chloridorrhea and congenital sodium diarrhea, are caused by the absence of a specific transport molecule.²³ In chloridorrhea, Cl⁻-HCO₃⁻ exchange in the ileum and colon is defective, thereby transforming chloride into a poorly absorbed ion. Diarrhea resulting from chloridorrhea can be reduced by limiting oral chloride intake or chloride secretion (i.e., by reducing gastric acid secretion with a proton pump inhibitor) or by stimulating chloride absorption in the colon by enhancing short-chain fatty acid absorption.²⁴ Several transporter defects have been proposed for congenital sodium diarrhea.²⁵

More commonly, a relative shift in the balance of absorptive and secretory pathways may contribute to diarrhea in clinical settings associated with epithelial injury or changes in cell proliferation. For example, viral gastroenteritis and celiac disease may disproportionately compromise transporters (e.g., disaccharidases, Na⁺-coupled absorption) that mediate absorption on the villous surface, whereas secretory pathways in the crypt are unchanged or increased (see Chapters 104 and 107).

For intestinal fluid and electrolyte absorption to be complete, the intestine must have an adequate surface area and adequate contact time with luminal contents. Substantial loss of surface area, as in celiac disease or inflammatory bowel disease (IBD) or after resective surgery, may compromise water absorption. Even though the reserve absorptive capacity in the small intestine and colon is large, sufficiently long surgical resections inevitably cause diarrhea. In some cases, the problem is temporary, because over time the intestine may improve its capacity for absorption by the process of adaptation.²⁶ Such compensation is impossible following resection of certain segments of the intestine with highly specific absorptive functions that simply cannot be assumed by other segments of the bowel. For example, ileocecal resection is followed by permanent inability to absorb sodium chloride against a concentration gradient²⁷ and, if sufficient ileum is resected, by failure to absorb vitamin B₁₂, intrinsic factor, and normal amounts of conjugated bile acids (see Chapter 103).

Abnormal motility may lead to diarrhea that has secretory and osmotic components.⁸ For fluid and electrolyte absorption to be complete, the contact time between luminal contents and the epithelium must be sufficient to permit absorption. In some patients, abnormal motility produces intestinal “hurry.”^28,29 Because rapid transit prevents adequate time for absorption, diarrhea results despite intact mucosal absorptive capacity, as measured by intestinal perfusion studies during which contact time is maximized by rapid infusion of fluid into the intestine.⁸ In some patients with intestinal hurry, the oral-cecal transit time may be as short as 10 minutes. Under such circumstances, the diarrhea is exacerbated by malabsorption of nutrients that produces an osmotic component to diarrhea. In disorders such as diabetes mellitus and postvagotomy diarrhea, intestinal hurry has been linked to abnormal enteric nervous system function.³⁰ In other clinical settings, such as amyloidosis, postprandial diarrhea, and irritable bowel syndrome (IBS), enteric nervous system dysfunction is suspected, but unproved.^31,32 Many endocrine diarrheas, such as those caused by peptide-secreting tumors or hyperthyroidism, may lead to diarrhea not only by affecting intestinal electrolyte transport, but also by accelerating intestinal motility.³³

Conversely, slow intestinal transit may lead to a secretory diarrhea by promoting small intestinal bacterial overgrowth.^34,35 Excess bacteria in the small intestine disrupt digestion and may alter electrolyte transport. The best documented example of diarrhea related to this mechanism is scleroderma. Although diabetes mellitus is often suspected of causing diarrhea by slow transit and stasis, as seen in scleroderma, such a pathophysiology is not always established (see Chapter 102).^36,37

Evaluation of the role of intestinal motility in the pathogenesis of diarrhea has been limited by the lack of the necessary tools to measure the interactions among motility, propulsive forces, and transit time. Except for intestinal perfusion studies, during which the effect of motility on electrolyte transport is eliminated, no methodology exists to dissociate the effects of intestinal transport and motility on net absorption.⁸ Thus, consensus has not been achieved on whether too much or too little motility causes diarrhea, nor on how luminal factors may alter intestinal smooth muscle function.

Reduced intestinal blood flow has an important but as yet poorly defined role in diarrhea. Whether mesenteric ischemia has a direct effect on absorption or whether low blood flow prompts secondary responses (e.g., via cytokines or neurotransmitters) that affect fluid transport and produce a secretory diarrhea is not clear. Radiation enteritis also produces an abnormal intestinal microcirculation associated with persistent diarrhea that may be difficult to treat (see Chapters 39 and 114).

COMPLEX DIARRHEA

Although classification of diarrhea as osmotic or secretory may be instructive in thinking about the pathophysiology of diarrhea, cases of pure secretory or pure osmotic diarrhea are uncommon. Most clinically important diarrhea is complex in pathogenesis; rather than being produced by a single pathophysiologic mechanism, several mechanisms are involved. Causes may include the effects of substances released by enteric endocrine cells, cytokines released by local and remote immunologically reactive cells, activity of the enteric nervous system, and peripherally released peptides and hormones (paracrine, immune, neural, and endocrine systems; see Chapter 1).

Further complicating the understanding of diarrhea is the recognition that certain mediators affect not only epithelial or muscle function, but also each other. For example, enteric nerves may stimulate mast cells, and products released from mast cells (particularly histamine) may alter enteric neuron functions.³⁸ A single agonist, such as prostaglandin, may have multiple simultaneous effects on epithelial function, muscle contraction, and the paracellular pathway, thereby leading to effects on ion transport, motility, and mucosal permeability.³⁹ Therefore, a number of modulators and effectors contribute to the final clinical picture. A full appreciation of the pathophysiology of diarrhea requires consideration of a regulatory system known as PINES (paracrine, immune, neural, and endocrine system modulators; Fig. 15-2).

Figure 15-2. PINES regulatory system in the intestine. The regulatory system of the intestine integrates paracrine, immune, neural, and endocrine systems and produces coordinated changes in mucosal and muscular function that permit adaptive responses to changing conditions. The regulatory system can widen or narrow the paracellular pathway that governs passive transmucosal permeability of electrolytes, accelerate or retard the transepithelial transport of nutrients and electrolytes by affecting membrane channels and pumps, alter motility by relaxing or contracting the various muscle layers in the intestine, and increasing or decreasing mucosal blood flow, thereby influencing intestinal metabolism. Diarrhea may be an appropriate response to acute infection. Maladaptive responses may be responsible for chronic diarrhea.

(From Sellin JH. Functional anatomy, fluid and electrolyte absorption. In: Feldman M, Schiller LR, editors. Gastroenterology and hepatology. The Comprehensive Visual Reference, vol 7: Small Intestine. Philadelphia, Pa: Current Medicine; 1997. p 1.11.)

An example of the complexity of the pathophysiology of a diarrheal syndrome is cholera. Cholera is often cited as the paradigm of a pure secretory diarrhea: cholera toxin targets the epithelial cell, increases the second messenger, cAMP, which opens apical chloride channels to stimulate chloride secretion, and results in diarrhea; however, the actual mechanism whereby cholera induces diarrhea is far more complex.⁴⁰ Cholera toxin stimulates endocrine cells and neural elements that reinforce its direct secretory effect on enterocytes.⁴¹ In addition, cholera toxin causes distinct changes in intestinal motility. Other toxins produced by Vibrio cholerae target tight junctions and thereby alter mucosal permeability (Fig. 15-3; see Chapter 107).⁴²

Figure 15-3. Pathophysiology of cholera. Vibrio cholerae produces several toxins that interact with adenylate cyclase in the enterocyte and several elements of the regulatory system of the intestine, including enteric neurons and enterochromaffin (EC) cells, to produce a secretory state and voluminous diarrhea. In addition to the classic enterotoxin, cholera toxin (CT), the bacterium also produces zona occludens toxin (ZOT), which increases the permeability of the tight junction between enterocytes, and accessory cholera enterotoxin (ACE), which has unclear effects on enterocytes. In addition to cyclic adenosine monophosphate (cAMP) generated by adenylate cyclase in response to CT, secretory stimuli include prostaglandin (PG), serotonin (5HT), and vasoactive intestinal polypeptide (VIP) released by macrophages, EC cells, and enteric neurons. CT has been shown to activate adenylate cyclase on the basolateral membrane of the cell in experiments in which the transcellular transport of CT to the basolateral membrane was inhibited by Brefeldin A (jagged arrow).

(From Sellin JH. Functional anatomy, fluid and electrolyte absorption. In: Feldman M, Schiller LR, editors. Gastroenterology and Hepatology. The Comprehensive Visual Reference, vol 7: Small Intestine. Philadelphia, Pa: Current Medicine; 1997. p 1.14.)

Another example of dysregulation of PINES is IBD.⁴³ Diarrhea in patients with IBD involves more than just exudation into the lumen as a result of destruction of the mucosa. Intact enterocytes are barraged by multiple secretagogues released by immune cells in the intestine and by bacterial toxins that may influence enterocyte function. Although initial models of diarrhea in IBD suggested altered fluid transport driven by chloride secretion, more recent studies have demonstrated that the diarrhea in IBD is mediated by an antiabsorptive effect associated with down-regulation of sodium channels and pumps.^44–46 The pathophysiology of diarrhea in IBD is even more complex if we consider the role of luminal bacteria. Bacterial proteins, such as flagellin, may stimulate the production of cytokines, such as interleukin (IL)-8, which further attract inflammatory cells.⁴⁷ Cytokines and immune cells also may influence tight junction barrier function and enterocyte secretory and absorptive pathways directly.^48,49 Conversely, epithelial cells may secrete cytokines, such as IL-6, that enhance polymorphonuclear leukocyte (neutrophil) function.⁵⁰

IBS is another example of a disorder with a complex pathophysiology. A constellation of factors, such as altered motility,³² bile acid malabsorption,⁵¹ and compromised rectal reservoir capacity,⁵² may aggravate symptoms in patients with IBS. At a more fundamental level, alterations in mast cell or enterochromaffin cell number, serotonin content, and serotonin reuptake and transport may contribute to the development of diarrhea (see Chapter 118).^53–55

Complex pathophysiology also may be observed in malabsorption syndromes and functional disorders, particularly those characterized by rapid transit. Failure to absorb carbohydrates may lead to osmotic diarrhea, but failure to absorb long-chain fatty acids may complicate matters by impairing electrolyte absorption by the colon.^12,56 Common postprandial functional diarrhea probably involves an interplay between motility and transport functions. Diarrhea caused by food allergy also involves activation of immunologic, paracrine, and neural mechanisms that regulate vascular permeability, electrolyte transport, and motility (see Chapters 9 and 101).⁵⁷

ACUTE VERSUS CHRONIC DIARRHEA

The time course of diarrhea can help direct management. Acute diarrhea (<four weeks) usually is caused by an infection, which is generally self-limited or easily treated.⁵⁸ Most common bacterial and viral causes of diarrhea run their course within seven days. Persistence of an acute diarrheal episode beyond seven days raises the possibility of a protozoal cause, such as giardiasis or cryptosporidiosis. Although few infectious agents (e.g., Aeromonas or Yersinia spp.) cause prolonged diarrhea in an immunocompetent person, chronic diarrhea is usually not caused by an infectious agent. Therefore, when confronted with a patient with chronic diarrhea, the clinician must consider noninfectious causes (see later).

LARGE-VOLUME VERSUS SMALL-VOLUME DIARRHEA

Differentiation of the cause of diarrhea on the basis of the volume of individual stools (rather than the total daily stool output) rests on the premise that the normal rectosigmoid colon functions as a storage reservoir. When that reservoir capacity is compromised by inflammatory or motility disorders involving the left colon, frequent small-volume bowel movements ensue. If the source of diarrhea is in the right colon or small bowel and if the rectosigmoid reservoir is intact, individual bowel movements are less frequent and larger. Therefore, frequent, small, painful stools may point to a distal colonic site of pathology, whereas painless large-volume stools suggest a right colonic or small bowel source. Although patients have difficulty quantifying stool volume accurately, the distinction between small- and large-volume stools may guide further diagnostic studies.

The daily total stool output may also provide hints about the cause. IBS often results in normal or only slightly elevated 24-hour stool weights, whereas diarrhea of other causes may produce greater stool weights. The stool weight can be estimated by the patient’s history: patients with diarrhea that produces dehydration (in the absence of vomiting or limited oral intake) typically have stool weights more than 1000 g and therefore are unlikely to have IBS (see Chapter 118).

OSMOTIC VERSUS SECRETORY DIARRHEA

Distinguishing diarrhea that results from intestinal malabsorption of ingested nonelectrolytes (osmotic diarrhea) from diarrhea that results from malabsorption or secretion of electrolytes (secretory diarrhea) helps separate the small number of cases of osmotic diarrhea from the much larger number of cases of secretory diarrhea. This distinction is based on the measurement of stool electrolyte concentrations.¹⁰ In secretory diarrhea, sodium, potassium, and accompanying anions account almost entirely for stool osmolality, whereas in osmotic diarrhea poorly absorbable solutes within the lumen of the intestine account for much of the osmotic activity of stool water (see later discussion). Because osmotic diarrhea is caused by the ingestion of some poorly absorbed substance, it abates with fasting. Secretory diarrhea typically continues during fasting, although stool output may decrease modestly because of reduced endogenous secretions.

WATERY VERSUS FATTY VERSUS INFLAMMATORY DIARRHEA

When diarrhea is chronic (>four weeks), the differential diagnosis can overwhelm even the most experienced clinician. By characterizing stools as watery, fatty, or inflammatory on the basis of simple stool tests, evaluation of the patient can be expedited by limiting the number of conditions that must be considered in the differential diagnosis.³ Watery diarrhea implies a defect primarily in water absorption as a result of increased electrolyte secretion or reduced electrolyte absorption (secretory diarrhea) or ingestion of a poorly absorbed substance (osmotic diarrhea). Fatty diarrhea implies defective absorption of fat and perhaps other nutrients in the small intestine. Inflammatory diarrhea implies the presence of one of a limited number of inflammatory or neoplastic diseases involving the gastrointestinal tract.

EPIDEMIOLOGIC FEATURES

One of the most useful clinical approaches to narrowing the differential diagnosis is to relate diarrhea to its setting. For example, a soccer mom and a backpacker from Nepal conceivably could have the same cause of the diarrhea but are more likely to have different causes. Some common clinical scenarios and the diagnoses that should be considered are shown in Table 15-2.

Table 15-2 Likely Causes of Diarrhea in Well-Defined Patient Groups or Settings

Travelers

Epidemics and Outbreaks

Diabetic Patients

Patients with Acquired Immunodeficiency Syndrome

Institutionalized and Hospitalized Patients

HISTORY

A carefully taken medical history is the key to the evaluation of a patient presenting with diarrhea. An essential feature is the duration of symptoms. Patients with acute diarrhea (<four weeks in duration) should be distinguished from those with chronic diarrhea, in whom the differential diagnosis is much broader. The severity of the diarrhea should be ascertained. Stool frequency is the easiest characteristic of diarrhea for patients to define but does not necessarily correlate with stool weight; some persons pass small amounts of stool frequently, but others have less frequent and more voluminous evacuations. Patients have a poor notion of stool volume, but symptoms such as dry mouth, increased thirst, decreased urine output, and weakness suggest dehydration resulting from a high stool output. Acute weight loss is also a good marker of severity. Chronic weight loss may suggest intestinal malabsorption or a serious constitutional process, such as malignancy, IBD, or hyperthyroidism.

Stool characteristics, such as the presence of blood, mucus, pus, oil droplets, or food particles, are also important. Blood in the stool signals the possibility of malignancy or IBD, although it is often caused by hemorrhoids in patients with frequent evacuations. In patients with acute infectious diarrhea, visible blood in the stool is highly specific for infection with an invasive organism.⁶⁵ Watery stools suggest an osmotic or secretory process, and the presence of oil or food particles is suggestive of malabsorption, maldigestion, or intestinal hurry. The phenomenon of floating stools generally represents an increase in the gas content rather than the fat content of the stools. The physician should also ask about the relationship of defecation to meals or fasting, passage of stool during the day versus the night, and presence of fecal urgency or incontinence. Urgency and incontinence are not indicative of voluminous diarrhea but suggest a problem with rectal compliance or with the muscles regulating continence. Nocturnal diarrhea that awakens the patient from sleep strongly suggests an organic rather than a functional disorder such as IBS. Other coexisting symptoms such as abdominal pain, flatulence, bloating or gaseous distention, cramps, fever, and weight loss should be noted. Excessive flatus suggests increased fermentation of carbohydrate by colonic bacteria as a result of ingestion of poorly absorbable carbohydrate or malabsorption of carbohydrate by the small intestine.

Because iatrogenic causes of diarrhea, such as drugs, previous surgery, or radiation therapy, are common, the physician should explore the history thoroughly for prior abdominal surgeries and ingestion of prescription drugs and over-the-counter remedies, including nutritional and herbal therapies. The patient’s diet should be reviewed thoroughly because diarrhea may result from ingestion of large quantities of poorly absorbable carbohydrates, such as fructose, or sugar alcohols, such as sorbitol or mannitol, which may be consumed in fruit juices and soda (which contain fructose and high-fructose corn syrup) or as dietetic, sugar-free candies and chewing gums (which contain sorbitol and mannitol).⁶⁶ Excessive coffee consumption also may be associated with diarrhea.

Epidemiologic clues also should be pursued (see Table 15-2). For example, recent foreign travel, particularly to undeveloped countries, makes the diagnosis of travelers’ diarrhea likely. The globalization of commerce has increased the frequency of once exotic infections in those without grossly obvious exposures.⁶⁷ The physician also should consider the patient’s residence in a rural or urban environment, the source of the patient’s drinking water, and the patient’s occupation, sexual orientation, and use of alcohol or illicit drugs. Potential secondary gains from illness or a history of attempted weight loss and fixation on body image should raise the possibility of laxative abuse (see later).

The patient’s history is essential in differentiating patients with IBS from those with other functional disorders or organic conditions that cause diarrhea. Current definitions of IBS emphasize the presence of abdominal pain associated with defecation.⁶⁸ Additional factors that suggest a diagnosis of IBS include a long history that usually extends back to adolescence or young adulthood, passage of mucus, and exacerbation of symptoms by stress. Factors that argue against a diagnosis of IBS include the recent onset of diarrhea, especially in older patients, diarrhea that awakens the patient from sleep, weight loss, the presence of blood in the stool, and stool weights more than 400 to 500 g daily. Abnormal blood test results, such as a low hemoglobin level, low serum albumin concentration, or high erythrocyte sedimentation rate, also argue against a diagnosis of IBS (see Chapter 118).

Painless diarrhea should no longer be considered a form of IBS. The Rome III committee has defined functional diarrhea as “at least 3 months, which need not be consecutive, in the preceding 6 months of liquid (mushy) or watery stools more than three-quarters of the time; and no abdominal pain.”⁶⁸ Obviously, many patients with chronic diarrhea will not have a readily defined cause of diarrhea identified when first seen and could be characterized as having functional diarrhea. Physicians should not rush to make this diagnosis without exploring alternative possibilities, particularly those that can produce episodic and variable diarrhea, such as small intestinal bacterial overgrowth and carbohydrate malabsorption. Functional diarrhea must also be distinguished from idiopathic secretory diarrhea (see later).

PHYSICAL EXAMINATION

Physical findings are usually more useful in determining the severity of diarrhea than in determining its cause. The patient’s volume status can be assessed by looking for orthostatic changes in blood pressure and pulse. Fever and other signs of toxicity should be noted. A careful abdominal examination is important, with particular attention to the presence or absence of bowel sounds, abdominal distention, localized or generalized tenderness, masses, and an enlarged liver.

On occasion, the physical examination may provide more direct evidence of the cause of diarrhea. Characteristic physical findings may be seen in mastocytosis (urticaria pigmentosa), amyloidosis (macroglossia, waxy papules, pinch purpura), Addison’s disease (increased pigmentation), glucagonoma (migratory necrotizing erythema), carcinoid syndrome (flushing), Köhlmeier-Degos disease (malignant atrophic papulosis), and celiac disease (dermatitis herpetiformis). Peripheral neuropathy and orthostatic hypotension may be the only clues to a diagnosis of amyloidosis. A thyroid nodule with cervical lymphadenopathy may be the only lead to the presence of medullary carcinoma of the thyroid. Tremor and other systemic signs should lead to consideration of hyperthyroidism. Right-sided heart murmurs, as well as an enlarged hard liver, may be present with carcinoid syndrome. Evidence of arthritis may be noted in IBD, Whipple’s disease, and some enteric infections. Lymphadenopathy might suggest acquired immunodeficiency syndrome (AIDS) or lymphoma. Signs of peripheral vascular disease with or without an abdominal bruit may suggest chronic mesenteric ischemia. Evidence of chronic liver disease may suggest advanced primary sclerosing cholangitis in a patient with colitis. A careful rectal examination may disclose defective sphincter or pelvic floor muscle function that could produce fecal incontinence.

FURTHER EVALUATION OF ACUTE DIARRHEA

Most cases of acute diarrhea are caused by infectious diseases that have a limited course, from a few days to a few weeks, and do not require a physician’s intervention unless the patient’s immune system is compromised or the patient develops complications of volume depletion or other evidence of severe toxicity, including inability to ingest fluid, frequent vomiting, and debilitating muscle or joint pain.⁵⁸

When these complications are present or when the diarrhea has persisted for more than a few days, a more comprehensive evaluation is warranted. In such patients, a complete blood count should be done to look for anemia, hemoconcentration, or an abnormal white blood cell count. Patients with a viral cause of diarrhea usually have normal white blood cell and differential counts or lymphocytosis, but those with bacterial infections, particularly caused by organisms that invade the intestinal mucosa, have a leukocytosis with an excess of immature white blood cells. Neutropenia, however, can occur in patients with salmonellosis. Measurements of serum electrolyte concentrations and blood urea nitrogen and serum creatinine levels can be used to assess the extent of fluid and electrolyte depletion and its effect on kidney function.

Stool samples should be examined for white blood cells to identify inflammatory diarrhea.⁶⁹ The standard method for detecting white blood cells in stool is with a Wright stain and microscopy. The accuracy of the test depends on the experience and skill of the observer, because false-positive and false-negative results are common. Tests for the neutrophil products, calprotectin and lactoferrin, are sensitive and specific for the detection of neutrophils in stool and may be a useful alternative to microscopy.⁷⁰ Studies have suggested that stool cultures are unlikely to grow pathogenic bacteria in the absence of fecal leukocytes; therefore, a Wright stain of stool or a fecal lactoferrin assay can be used to decide which stool samples should be sent for bacterial culture, thereby minimizing expense. This approach may be of more value in outpatients than in patients hospitalized with diarrhea, because the latter have toxicity or have failed to resolve spontaneously within a few days and must have stool cultures sent. Routine stool cultures are of little use for hospitalized patients in whom acute diarrhea develops while the patient is in the hospital; testing for Clostridium difficile toxin is likely to be more helpful. The diagnostic value of examination of stool for ova and parasites depends on the pretest probability of a parasitic infection and the experience of the observer. Enzyme-linked immunosorbent assays (ELISAs) for giardiasis and cryptosporidiosis and serologic testing for amebiasis are more accurate tests than stool microscopy and should be ordered, even in the absence of fecal leukocytes.⁷¹ Patients who have been treated with antibiotics in the preceding three months or those in whom diarrhea develops in an institutional setting should be tested for C. difficile toxin.⁷² With the increase in community acquired C. difficile infections, physicians should consider this treatable cause of acute diarrhea, even in the absence of a prior history of antibiotic use (see Chapters 107 to 109).

Abdominal radiographs should be obtained in toxic patients to assess for colitis and to look for evidence of ileus or megacolon. Proctoscopy or flexible sigmoidoscopy also should be considered in patients who are clearly toxic with infection, have blood in the stool, or have persistent acute diarrhea. Sigmoidoscopy is probably adequate as an early investigation in such cases of severe acute diarrhea. In patients with AIDS-related diarrhea, colonoscopy is preferable because a substantial proportion of infections and lymphomas may be present only in the right colon,⁷³ although this approach has been called into question.⁷⁴ If sigmoidoscopy or colonoscopy is done, mucosal biopsy specimens should be obtained, even if the mucosa does not appear to be grossly inflamed, because pathologic examination can identify important clues to facilitate a specific diagnosis.⁷⁵ An algorithm for the evaluation of patients with acute diarrhea is shown in Figure 15-4.

Figure 15-4. Algorithm for the evaluation of patients with acute diarrhea. WBC, white blood cells.

(From Schiller LR. Diarrhea. Med Clin North Am 2000; 84:1259-74.)

FURTHER EVALUATION OF CHRONIC DIARRHEA

Because the differential diagnosis of chronic diarrhea is more extensive than that of acute diarrhea, evaluation of patients with chronic diarrhea is more complex.^76,77 Initially, the physician should categorize the diarrhea as watery, inflammatory, or fatty (Fig. 15-5). In addition to the history, physical examination, and routine laboratory tests already mentioned, analysis of a stool sample can be used to categorize the diarrhea and thereby limit the number of conditions to be considered in the differential diagnosis. Stool analysis can be obtained on a random sample or a timed collection (i.e., 24-, 48-, or 72-hour stool sample). The value of analyzing a timed collection is that the stool weight and hence the output of stool components, such as fat, can be measured accurately. The daily stool weight is perhaps the best clue to the potential metabolic impact of the diarrhea. In the absence of a timed collection, however, assessments of other stool characteristics on a random, or spot, collection still provide many clues to the correct diagnosis.³ These assessments include stool sodium and potassium concentrations, stool pH, a fecal occult blood test, and an examination of stool for white blood cells or a test for the presence of a surrogate marker, such as fecal lactoferrin or calprotectin.⁷⁰ In appropriate circumstances, stool samples can also be analyzed for fat content and for laxatives, including magnesium, phosphate, sulfate, bisacodyl, and anthraquinones (see later discussion of factitious diarrhea).

Figure 15-5. Algorithm for the initial evaluation of patients with chronic diarrhea. AIDS, acquired immunodeficiency syndrome; Ig, immunoglobulin; OTC, over-the-counter; Rx, prescription; WBCs, white blood cells.

(From Fine KD, Schiller LR. AGA technical review on the evaluation and management of chronic diarrhea. Gastroenterology 1999; 116:1464-86.)

Although stool collections are often viewed by patients and physicians as being messy and distasteful, they usually can be done easily and successfully at home or in the hospital. Perhaps the biggest hurdle is in dealing with laboratories that are inexperienced or uninterested in stool analysis. Commercially available collection units that fit into a commode and allow separation of stool and urine facilitate the collection, as does the use of preweighed plastic or metal containers and a small refrigerator or picnic cooler to keep the specimens cold. Patients should continue their regular activities and should consume a regular diet, including a fat intake of 80 to 100 g of fat daily, during the collection. Keeping a diary of food and liquid ingested facilitates estimation of the patient’s fat and calorie intake by a dietitian. When evaluating the results of timed stool collections, it is important to recognize that patients often do not eat diets this high in fat. During the collection, diagnostic tests that might alter stool output or composition, such as barium radiograph studies, should be avoided and only essential medication should be given. Any antidiarrheal medications should be withdrawn. For most patients with diarrhea, a 48-hour collection is sufficient. If stool output is not representative during that time, the collection can be extended. Occasionally, stool output is measured during fasting; if the diarrhea is caused by some ingested substance, fasting should abolish the diarrhea. Persistence of diarrhea during fasting is one criterion for secretory diarrhea.

Measurement of stool sodium and potassium concentrations allows the physician to calculate an osmotic gap in stool water. The osmotic gap is calculated by subtracting twice the sum of the sodium and potassium concentrations from 290 mOsm/kg, the osmolality of stool in the body.¹⁰ The concentration is doubled to account for anions that accompany these cations. A small osmotic gap (<50 mOsm/kg), which signifies that the osmolality of stool water is attributable mostly to incompletely absorbed electrolytes, is characteristic of secretory diarrhea (Fig. 15-6). On the other hand, a large osmotic gap (>100 mOsm/kg) indicates that much of the stool osmolality is composed of nonelectrolytes. A large gap is characteristic of an osmotic diarrhea, usually resulting from ingestion of some poorly absorbed substance, such as magnesium salts. When the sum of sodium and potassium concentrations doubled is higher than 290 mOsm/kg, ingestion of a poorly absorbed multivalent anion, such as phosphate or sulfate, is likely.¹⁰ Such a negative osmotic gap is the result of an excess of cations obligated by multivalent anions. The actual measurement of stool osmolality is of value only in detecting samples that have been contaminated by the addition of water or hypotonic urine. Such samples have an osmolality lower than 290 mOsm/kg. Stool osmolality tends to rise once the stool has been collected because of continuing bacterial fermentation in vitro.¹² Therefore, measured osmolality should not be used to calculate the fecal osmotic gap.

Figure 15-6. Fecal electrolytes and the fecal osmotic gap. The osmolality of colonic fluid is in equilibrium with body fluids and is approximately 290 mOsm/kg. The total concentration of electrolytes, therefore, cannot exceed 290 mmol/L. In secretory diarrhea, almost all the osmotic activity of colonic contents is caused by electrolytes and, therefore, the estimate of electrolyte content by the formula 2 × ([Na⁺] + [K⁺]), is approximately 290 mmol/L. In osmotic diarrhea, electrolytes account for only a small part of the osmotic activity; unmeasured osmoles resulting from the ingestion of a poorly absorbed substance account for most of the osmotic activity, and the calculated osmotic gap will be high.

(From Schiller LR. Chronic diarrhea. In: McNally P, editor. GI/Liver Secrets. 2nd ed. Philadelphia, Pa: Hanley & Belfus; 2001. p 411.)

The pH of stool water provides useful information about the possibility of carbohydrate malabsorption.¹⁰ Carbohydrate that reaches the colon is promptly fermented by the bacterial flora, with release of CO₂ and H₂ gases and short-chain fatty acids. As a result of fermentation, the pH is acidic, usually dropping to less than 6, a finding that indirectly indicates excess carbohydrate fermentation in the colon.

Fecal occult blood testing and examination of stool for leukocytes allow one to identify inflammatory diarrhea resulting from colitis or malignancy. Other diarrheal conditions that cause occult bleeding include lymphoma of the small intestine, celiac disease (fecal occult blood in 50% of cases), and refractory sprue (fecal occult blood in 70% of cases).⁷⁸

Stool fat output can be measured quantitatively by chemical means on a timed (48- to 72-hour) collection or estimated qualitatively by use of a Sudan stain on a random specimen. Steatorrhea is defined as excessive loss of fat in the stool (more than 7 g, or 9% of intake, for 24 hours). This definition, however, may not be valid for the diagnosis of fat malabsorption or maldigestion in all patients with chronic diarrhea. In one study,⁷⁹ diarrhea induced with laxatives produced mild steatorrhea in 35% of normal subjects. In patients with diarrhea, fat excretion in the range of 7 to 14 g/24 hr has a low specificity for the diagnosis of defective fat absorption. Fat excretion more than 14 g/24 hr, however, strongly indicates a problem with fat absorption.⁷⁹ Fat intake during a quantitative collection should be estimated from diet diaries, because patients with diarrhea frequently have anorexia or early satiation that may reduce their fat intakes substantially, thereby reducing fat excretion. For a valid study, patients should consume 70 to 100 g of fat daily for a few days before and during the timed collection. Measurement of fat excretion as a measure of malabsorption also can be compromised by ingestion of the lipase inhibitor orlistat or the fat substitute olestra.⁸⁰

When only a random sample of stool is available, qualitative estimation of fat excretion by means of a Sudan stain of a fecal smear may be helpful.⁸¹ Semiquantitative methods can be applied to measure the number and size of fat globules, and these methods produce results that correlate well with quantitative collections (see Chapter 101).

In patients in whom surreptitious laxative ingestion is suspected, stool water can be analyzed for laxatives by chemical or chromatographic methods. As currently done commercially, this analysis is subject to error.⁸² If positive, the test for the laxative should be repeated on another stool sample to confirm the finding before confronting the patient with this discovery.

Stool samples can also be assayed with a chemical test for carbohydrate (anthrone reagent)¹² and for α₁-antitrypsin clearance to detect protein-losing enteropathy.⁸³ These tests have limited clinical usefulness and should not be used routinely for the initial evaluation of a patient with chronic diarrhea.

Once the stool analysis is completed, chronic diarrhea can be categorized as watery (and secretory or osmotic), inflammatory, or fatty. This classification allows more direct evaluation of the cause of diarrhea.

Chronic Watery Diarrhea

Secretory diarrhea has a broad differential diagnosis, as indicated previously (see Table 15-3), and a wide investigative net must be cast to identify a specific cause (Fig. 15-7).

Figure 15-7. Algorithm for the evaluation of chronic secretory diarrhea. ACTH, adrenocorticotropic hormone; CT, computed tomography; 5-HIAA, 5-hydroxyindoleacetic acid; TSH, thyroid-stimulating hormone; VIP, vasoactive intestinal peptide.

(From Fine KD, Schiller LR. AGA technical review on the evaluation and management of chronic diarrhea. Gastroenterology 1999; 116:1464-86.)

Infection should be excluded by stool culture for bacteria and special tests for other organisms. The patient’s human immunodeficiency virus (HIV) status should be clarified at this point, because patients with AIDS are more likely than others to have an infectious cause of chronic diarrhea (see Chapter 33).⁸⁴ Although most bacteria that cause diarrhea are cleared spontaneously within four weeks, some organisms, such as Aeromonas and Plesiomonas, may produce chronic diarrhea.^85,86 Special culture techniques may be needed to detect these organisms. Special techniques are also required to find other pathogens. For example, coccidia and microsporidia require special microbiologic techniques such as polymerase chain reaction methodology for optimal detection.⁸⁷ Giardiasis and cryptosporidiosis are sometimes difficult to diagnose by a standard examination for ova and parasites, and ELISA testing for Giardia antigen and cryptosporidium antigen should be ordered.⁷¹ Examination of mucosal biopsy specimens with special stains or electron microscopy may be needed to find pathogens.

Small intestinal bacterial overgrowth may result in secretory diarrhea, presumably caused by toxins, as well as fatty diarrhea caused by bile salt deconjugation (see later). The glucose-hydrogen breath test (see later) can be used to screen for this condition, but the gold standard for diagnosis of small intestinal bacterial overgrowth remains quantitative culture of a small bowel aspirate, if available (see Chapter 102).⁸⁸ Structural diseases, such as short bowel syndrome, gastrocolic or enteroenteric fistula, mucosal diseases, IBD, and tumors, including lymphomas, should be sought by means of radiographic and endoscopic techniques. Small bowel radiographs remain an important method for detecting structural small bowel diseases. Computed tomography (CT) or magnetic resonance imaging (MRI), particularly CT or MRI enterography, is of value for detecting not only small bowel and colonic diseases, but also diseases extrinsic to the bowel that can cause diarrhea, such as pancreatic tumors.

Visualization and biopsy of the mucosa of the small bowel by endoscopy or enteroscopy can be valuable, although whether push enteroscopy adds much to standard esophagogastroduodenoscopy for the evaluation of diffuse small bowel diseases is uncertain.⁸⁹ Diseases that may be detected by small intestinal biopsy include celiac disease, Crohn’s disease, giardiasis, intestinal lymphoma, eosinophilic gastroenteritis, tropical sprue, Whipple’s disease, lymphangiectasia, abetalipoproteinemia, amyloidosis, mastocytosis, and various infectious processes. Patients with many of these disorders usually, but not always, present with steatorrhea (see Chapters 27, 29, 35, 101, 104 to 109, and 111).

The role of wireless capsule endoscopy in the diagnosis of diarrhea caused by small bowel disease is evolving rapidly.⁹⁰ Subtle lesions not appreciated by other diagnostic modalities can be seen, and new methods of deep enteroscopy (double-balloon and spiral overtube enteroscopy) allow access to and biopsy of most of these abnormalities (see Chapter 19). Sigmoidoscopy or colonoscopy can be used to visualize the colon and permit directed biopsies. Because colonic causes of chronic secretory diarrhea tend to produce diffuse changes throughout the colon, sigmoidoscopy usually is adequate for this purpose.⁹¹ Colonoscopy is preferable if the patient is older (and thus requires screening for colon cancer), has blood in the stool, is suspected of having right colonic or ileal disease, or has AIDS.^73,92 Chronic disorders that can be diagnosed by inspection of the colonic mucosa include pseudomelanosis coli, polyps, tumors, Crohn’s disease, ulcerative colitis, amebiasis, and nonspecific ulceration. All patients with undiagnosed chronic secretory diarrhea should have mucosal biopsy specimens obtained from the colon, even when the mucosa appears grossly normal.³ Random biopsies should include multiple samples from several locations to give the pathologist the best chance of making a diagnosis. Diseases in which the colonic mucosa appears normal endoscopically, but which can be diagnosed histologically, include microscopic colitis (lymphocytic and collagenous colitis, see later), amyloidosis, granulomatous infections, and schistosomiasis (see Chapters 35, 110 to 112, and 124). The diagnostic yield of colonoscopy or sigmoidoscopy with biopsy in patients referred for chronic diarrhea is approximately 15% to 30%.

The next level of investigation is selective testing for diarrhea caused by peptide-secreting tumors, an intellectually interesting form of chronic watery diarrhea that is quite rare. The pretest probability of having a peptide-secreting tumor in a patient with chronic diarrhea is so low that screening these patients with a panel of serum peptide levels is far more likely to produce a false-positive than a true-positive result.⁹³ Testing for elevated serum peptide levels or urinary metabolites of endocrine mediators, such as 5-hydroxyindoleacetic acid (5-HIAA), metanephrine, and histamine, should be limited to those patients who have chronic diarrhea with symptoms and signs consistent with a tumor syndrome, such as flushing or a large hard liver in carcinoid syndrome, ulcer disease suggestive of Zollinger-Ellison syndrome, headache, flushing, and urticaria pigmentosa in mastocytosis, or patients who have a CT scan that shows a tumor.¹⁹ Scintigraphy using radiolabeled octreotide, especially combined with positron-emission tomography (PET) or CT, also can be used to identify peptide-secreting tumors (see Chapters 31 and 32).

More common endocrinologic diseases that cause diarrhea are diabetes mellitus, hyperthyroidism, and Addison’s disease. In many cases, other symptoms and signs, such as an enlarged thyroid or skin pigmentation characteristic of Addison’s disease, suggest the presence of these conditions. Blood glucose, thyroid-stimulating hormone, and serum cortisol levels before and after injection of an adrenocorticotropic hormone analog should be measured selectively in patients who might have one of these conditions. Other blood tests that may be relevant in evaluating secretory diarrhea include serum protein electrophoresis and immunoglobulin electrophoresis. Selective immunoglobulin A (IgA) deficiency may present with recurrent intestinal infections such as giardiasis, whereas combined variable immune deficiency can be associated with a variety of puzzling intestinal findings that sometimes mimic celiac disease.⁹⁴ Testing for HIV and HIV-2 may be appropriate (see Chapters 33 and 35).

Osmotic diarrhea has a much more limited differential diagnosis, and its evaluation is much simpler (Fig. 15-8).⁹ For practical purposes, osmotic diarrhea is caused by one of three conditions—ingestion of exogenous magnesium, consumption of poorly absorbable carbohydrates, or carbohydrate malabsorption. Ingestion of other osmotically active substances is unusual. Fortunately, these conditions can be differentiated by taking a careful history and performing simple stool tests.

Figure 15-8. Algorithm for the evaluation of chronic osmotic diarrhea.

(From Fine KD, Schiller LR. AGA technical review on the evaluation and management of chronic diarrhea. Gastroenterology 1999; 116:1464-86.)

Magnesium can be measured directly in stool water by atomic absorption spectrophotometry.¹¹ Excretion of more than 15 mmol (30 mEq) of magnesium daily or concentrations in stool water of more than 44 mmol/L (90 mEq/L) strongly suggests magnesium-induced diarrhea. The ingestion may be intentional, as in a patient with surreptitious laxative ingestion, or accidental, as in a patient who uses magnesium-containing antacids or mineral supplements.

Ingestion of poorly absorbed carbohydrates or carbohydrate malabsorption typically leads to a low fecal pH because of bacterial fermentation in the colon. A fecal pH lower than 6 is highly suggestive of carbohydrate malabsorption.^10,12 More generalized malabsorption that involves fecal loss of amino acids and fatty acids in addition to carbohydrate may produce a somewhat higher pH (e.g., pH = 6 to 7.5). Isolated carbohydrate malabsorption is usually caused by ingestion of a poorly absorbable carbohydrate, such as lactose in a person with lactase deficiency. Other common causes include ingestion of poorly absorbed sugar alcohols that are used as artificial sweeteners, such as sorbitol or mannitol, or excessive ingestion of sugars with a limited absorption capacity, such as fructose.⁶⁶ Therapeutic use of inhibitors of carbohydrate absorption, such as acarbose, may also lead to carbohydrate malabsorption.⁹⁵ Because fermentation produces not only short-chain fatty acids that acidify the stool, but also carbon dioxide and hydrogen, complaints of gas and bloating by the patient are clinical clues to the presence of carbohydrate malabsorption, although these symptoms are fairly nonspecific (see Chapters 16, 100, and 101).⁹⁶

Once the clinical picture or stool analysis suggests carbohydrate malabsorption, a careful review of the patient’s diet may indicate the likely source. In some persons, a hydrogen breath test with lactose as the sugar substrate can confirm lactose intolerance as the diagnosis.⁹⁷ In this test, a previously fasting patient ingests 25 g of lactose dissolved in water, and exhaled breath is assayed for hydrogen content at baseline and at intervals for several hours. Because hydrogen is not a normal product of human metabolism, any increase in breath hydrogen concentration is the result of bacterial fermentation and indicates that unabsorbed lactose has reached the colon. This principle also has been applied to the assessment of sucrase deficiency after administration of a sucrose load and to the assessment of fructose malabsorption after a fructose load.⁹⁸ Breath hydrogen testing has been adapted to detect small intestinal bacterial overgrowth with the use of glucose, a substrate that ordinarily should be absorbed completely before reaching the colon.⁸⁸ Lactulose, a nonabsorbable but easily fermented disaccharide, also has been used to detect small intestinal bacterial overgrowth, but because of the wide variability of intestinal transit time, use of lactulose for this purpose is problematic. Lactulose can be used as a substrate for determining the oral-cecal transit time. Breath hydrogen testing after administration of d-xylose has been advocated as a screening test for generalized intestinal malabsorption.⁹⁹ For most purposes breath hydrogen testing provides only supportive evidence when the pretest likelihood of a particular diagnosis is high (see Chapter 101).

Once a specific cause of osmotic diarrhea has been postulated, a therapeutic trial of an elimination diet can confirm the diagnosis.

Chronic Inflammatory Diarrhea

Patients with chronic diarrhea and white blood cells or blood in the stool are classified as having inflammatory diarrhea. These characteristics indicate that the mucosa is disrupted and inflamed. Diagnostic considerations include IBD, infections, pseudomembranous enterocolitis, mesenteric ischemia, radiation enteritis, and neoplasia. Because these conditions may produce a secretory diarrhea without markers of inflammation in the stool, they must be considered in the differential diagnosis of secretory diarrhea as well (see Chapters 39, 104 to 112, 114, 115, and 121 to 124).

Sigmoidoscopy or colonoscopy should be undertaken to look initially for structural changes, because colitis is a common cause of inflammatory diarrhea and a neoplasm can be life threatening (Fig. 15-9). Sigmoidoscopy can detect most causes of inflammatory diarrhea but can miss disorders localized to the right colon and ileum. Because preparation for this test is simpler than that for colonoscopy and the frequency of complications is lower, sigmoidoscopy is preferred by some physicians. Others prefer to examine the entire colon and terminal ileum in patients with inflammatory diarrhea, especially if occult blood is detected in the stool.⁹² The choice of test depends on the circumstances of the individual patient. For example, a patient older than 50 years who has not undergone previous screening for colon cancer would benefit by undergoing colonoscopy in this setting. Whichever test is selected, biopsy specimens must be obtained from the colon to aid in making the correct diagnosis.

Figure 15-9. Algorithm for the evaluation of chronic inflammatory diarrhea. CT, computed tomography.

(From Fine KD, Schiller LR. AGA technical review on the evaluation and management of chronic diarrhea. Gastroenterology 1999; 116:1464-86.)

Infection can cause chronic inflammatory diarrhea or aggravate existing inflammatory diarrhea caused by ulcerative colitis or Crohn’s disease. The pathogens most likely to cause chronic inflammatory diarrhea are C. difficile, cytomegalovirus, Entamoeba histolytica, Yersinia spp., and Mycobacterium tuberculosis. C. difficile and cytomegalovirus are notorious for causing exacerbations of IBD (see Chapters 111 and 112).¹⁰⁰ In addition to biopsies, appropriate cultures and serologic tests should be obtained to exclude these infections.

Chronic Fatty Diarrhea

Steatorrhea implies the disruption of fat solubilization, digestion, or absorption in the small intestine. The evaluation of chronic fatty diarrhea is designed to distinguish maldigestion, inadequate luminal breakdown of triglyceride, from malabsorption, inadequate mucosal transport of the products of digestion (see Chapter 101).

The major causes of maldigestion are pancreatic exocrine insufficiency (e.g., chronic pancreatitis) and lack of bile (e.g., advanced primary biliary cirrhosis). Mucosal diseases (e.g., celiac disease) are the usual causes of malabsorption. The absolute amount of steatorrhea and the fecal fat concentration (grams of fat/100 g of stool) provide clues to the cause of steatorrhea.¹⁰¹ The degree of steatorrhea tends to be higher with maldigestion (as in pancreatic insufficiency)—often more than 30 g fat/day—than with mucosal disease, because of the greater disruption of fat assimilation with maldigestion. Additionally, the fecal fat concentration tends to be higher with maldigestion than with mucosal disease because mucosal disease often is associated with poor fluid and electrolyte absorption, so that the stool fat content is diluted by unabsorbed water. Also, because fat digestion usually is intact in mucosal disease, triglycerides are broken down to fatty acids in the small intestine and pass into the colon, where they inhibit electrolyte and water absorption and thus further dilute the fat content of stool.⁵⁶ By contrast, maldigestion caused by pancreatic and biliary disorders typically reduces triglyceride hydrolysis and does not result in delivery of excess fatty acids to the colon or inhibition of fluid and electrolyte absorption. Therefore, the unabsorbed fat is dispersed in a smaller stool volume and is thus more concentrated. A fecal fat concentration more than 9.5 g/100 g in a patient with suspected maldigestion strongly suggests a pancreatic or biliary cause of steatorrhea.

The further evaluation of patients with chronic fatty diarrhea is relatively straightforward (Fig. 15-10). The first step is to look for a structural problem involving the small bowel or pancreas. The evaluation may include small bowel radiography or endoscopy with small bowel biopsy, CT, and MR cholangiopancreatography. When a small bowel biopsy is performed, luminal contents should be aspirated and a sample sent for quantitative culture to exclude small intestinal bacterial overgrowth. Because celiac disease is the most common cause of mucosal disease that leads to malabsorption, tissue transglutaminase antibodies and endomysial antibodies should be determined (see Chapters 102 and 104).¹⁰²

Figure 15-10. Algorithm for the evaluation of chronic fatty diarrhea.

(From Fine KD, Schiller LR. AGA technical review on the evaluation and management of chronic diarrhea. Gastroenterology 1999; 116:1464-86.)

If no intestinal abnormalities are discovered or radiographic evidence of chronic pancreatitis is detected, abnormal pancreatic exocrine function should be considered. Available tests of pancreatic function all have limitations. The secretin stimulation test, in which exogenous secretin is used to stimulate the pancreas and bicarbonate output is measured by aspiration of duodenal contents, is the most time-honored of these tests but is rarely performed because of its complexity.¹⁰³ For the test to be done properly, the duodenum and stomach have to be intubated and the samples obtained from the duodenum need to be analyzed promptly. Attempts to update the secretin stimulation test by combining it with endoscopic retrograde cholangiopancreatography have been attempted, but these modifications have not been adopted widely.¹⁰⁴ Determination of pancreatic enzyme concentrations in stool has been advocated as a simpler screening test for pancreatic exocrine insufficiency. Direct measurement of stool chymotrypsin activity has poor sensitivity and specificity in patients with chronic diarrhea.¹⁰⁵ Measurement of fecal elastase has only somewhat better reliability.¹⁰⁶ In reality, the best way to determine pancreatic exocrine insufficiency may be a therapeutic trial of pancreatic enzyme supplementation. If such a trial is conducted, high doses of enzymes should be prescribed and some objective measurement, such as fecal fat excretion or weight gain, should be monitored to assess response (see Chapter 59).¹⁰⁷

Inadequate bile salt solubilization of dietary fat can usually be inferred from the patient’s history or physical examination (e.g., cholestatic jaundice, ileal resection, or a known enterocolic fistula). If proof of the mechanism is required, analysis of a postprandial duodenal aspirate can demonstrate reduced conjugated bile acid concentrations. This test may not be available outside specialized centers, and a therapeutic trial of exogenous conjugated bile acids may be the best way of establishing the diagnosis. Supplementation with bile acids reduces steatorrhea and can often improve the patient’s nutritional status without aggravating diarrhea.¹⁰⁸

EMPIRICAL THERAPY OF ACUTE DIARRHEA

Because infection is a frequent cause of acute diarrhea, empirical trials of antibiotic therapy are often considered by physicians.¹¹³ If the prevalence of bacterial or protozoal infection is high in a community or a specific situation, empirical use of an antibiotic is logical, as in the treatment of travelers’ diarrhea with a fluoroquinolone or rifaximin, even without bacteriologic proof of infection.¹¹⁴ Antibiotic therapy also is often used empirically for more severely ill patients while bacterial culture results are pending. This approach has been called into question with the observation that patients in whom hemolytic-uremic syndrome develops in response to infection with Escherichia coli are more likely than those without hemolytic-uremic syndrome to have received empirical antibiotic therapy, although the occurrence of the syndrome may relate to the specific antibiotic and dose prescribed.¹¹⁵ Experts also advise against empirical antibiotic treatment of salmonellosis unless enteric fever is present.¹¹⁶ For patients with persistent diarrhea (lasting more than one week), an empirical trial of metronidazole or nitazoxanide for a protozoal infection is sometimes considered (see Chapters 107 and 109).¹¹⁷

Nonspecific antidiarrheal agents can reduce stool frequency, stool weight, and coexisting symptoms, such as abdominal cramps (Table 15-6). Opiates such as loperamide or diphenoxylate with atropine frequently are prescribed.¹¹⁸ The concern that these antiperistaltic agents slow the clearance of pathogens from the intestine has largely not been substantiated. Intraluminal agents, such as bismuth subsalicylate (Pepto-Bismol) and adsorbents (e.g., kaolin), also may help reduce the fluidity of bowel movements. Racecadotril, a drug that inhibits enkephalinase and thereby increases the effects of endogenous opiates on the mu opiate receptor, is available for the treatment of acute diarrhea in some countries.¹¹⁹

Table 15-6 Nonspecific Drug Therapy for Chronic Diarrhea

* Not yet approved in the United States.

EMPIRICAL THERAPY OF CHRONIC DIARRHEA

Empirical therapy is used in patients with chronic diarrhea in three situations: (1) as temporizing or initial treatment before diagnostic testing; (2) after diagnostic testing has failed to confirm a diagnosis; and (3) when a diagnosis has been made but no specific treatment is available or specific treatment has failed to produce a cure. Generally, empirical antibiotic therapy is less useful for chronic diarrhea than for acute diarrhea, because infection is a much less likely cause. Although some clinicians try an empirical course of metronidazole or a fluoroquinolone before committing a patient to extensive diagnostic testing, this approach is not supported by data and is not recommended.

In the appropriate clinical setting, therapeutic trials of pancreatic enzyme replacement and conjugated bile acid supplementation in patients with unexplained steatorrhea may be diagnostic and therapeutic (see earlier). By contrast, when pancreatic enzyme supplements or bile acid–binding resins are tried empirically for so-called idiopathic chronic diarrhea, they rarely yield satisfactory results (see later).

Symptomatic treatment with an opiate often is necessary in patients with chronic diarrhea because specific treatment may not be available.¹¹⁸ Potent opiates such as codeine, opium, or morphine are underused in the management of these patients, largely because of fear of abuse. In fact, these agents are rarely abused by patients with chronic diarrhea, especially if a few simple measures are taken. First, the patient needs to be informed about the abuse potential of the medication and should be warned not to increase the dose without consulting the physician. Second, the dose should be low initially and titrated up until efficacy is achieved. Third, use of the opiate should be monitored closely, and the prescription should not be refilled until an interval appropriate with the anticipated usage has passed.

Other agents that are sometimes used as nonspecific antidiarrheal agents include octreotide and clonidine. Octreotide, a somatostatin analog, has been shown to improve diarrhea in patients with the carcinoid syndrome and other endocrinopathies, dumping syndrome, chemotherapy-induced diarrhea, and AIDS.¹²⁰ The benefit in other diarrheal diseases is less clear. Clonidine, an α-adrenergic agent that has effects on intestinal motility and transport,¹²¹ may have a special role in diabetic diarrhea, but its hypotensive effect limits its usefulness in many patients with diarrhea.¹²²

Interest in and evidence to support the use of probiotics, ostensibly good bacteria (e.g., certain strains of lactobacilli), as therapy for diarrhea has been increasing. By modifying the colonic flora, these agents may stimulate local immunity and speed the resolution of travelers’ diarrhea, antibiotic-associated diarrhea, and infantile diarrhea.^123–125 Herbal remedies for diarrhea include those containing berberine (goldenseal, barberry), which appears to stimulate fluid and electrolyte absorption, and arrowroot, the mechanism of which is unknown.^126,127

Stool-modifying agents such as psyllium alter stool consistency but do not reduce stool weight.^128,129 They can be helpful in patients with coexisting fecal incontinence and in some patients with low stool weights (see earlier). The change from watery to semiformed stools may be sufficient to alleviate symptoms. In addition, pectin may delay transit through the proximal intestine and increase luminal viscosity, thus serving as an adjunctive empirical treatment.

IRRITABLE BOWEL SYNDROME AND FUNCTIONAL DIARRHEA

Undoubtedly, the most common diagnosis made in patients with chronic diarrhea is IBS, yet only a fraction of patients with chronic diarrhea actually meet the current criteria for the diagnosis of IBS, for which abdominal pain is a central feature^68,130 and diagnostic criteria have changed. In the past, painless diarrhea was considered to be a subtype of IBS, but such patients are now excluded from a diagnosis of IBS and are labeled as having functional diarrhea instead.

Diarrhea in patients with IBS and functional diarrhea tends to be variable and sometimes alternates with periods of constipation.⁶⁸ When measured, daily stool output is relatively low, typically less than 400 g/24 hr. Consistency varies from loose to soft and rarely is watery. The diarrhea does not awaken patients from sleep. Rectal urgency and fecal incontinence may be pronounced, especially during periods of psychological stress. Weight loss and evidence of chronic illness are uncommon (see Chapter 118). In some patients, IBS seems to be a late consequence of acute gastroenteritis.¹³¹

When typical features of IBS are absent, other diagnoses should be considered. An alternative diagnosis is carbohydrate malabsorption which can produce diarrhea of variable severity, depending on the amount of the malabsorbed carbohydrate consumed.¹³² In addition, cramps, excessive flatus, and bloating may be present with carbohydrate malabsorption. A carefully taken dietary history and a stool pH value lower than 6 distinguish this disorder from IBS. Caution should be exercised when using breath testing to confirm lactose intolerance in a patient with chronic diarrhea.¹³³ Another condition that can simulate IBS is small intestinal bacterial overgrowth. The diagnosis of this condition is complex and controversial (see earlier).⁸⁸ Bile acid–induced diarrhea also can vary in severity, depending on the rate of delivery of bile acids to the colon. Response to a therapeutic trial of a bile acid–binding resin may be a reasonable diagnostic test for this condition (see later). Celiac disease is another diagnosis that often is considered but is found in only 3% to 5% of patients who meet criteria for a diagnosis of IBS.¹³⁴ Most patients with other causes of chronic diarrhea have been misdiagnosed as IBS at some point before the correct diagnosis is discovered.

MICROSCOPIC COLITIS

The terminology associated with microscopic colitis has been confusing. The term collagenous colitis was used initially to describe the histologic findings of subepithelial fibrosis and inflammation in the rectal mucosa of a woman with chronic watery diarrhea who had a normal gross appearance of the mucosa by proctoscopy. Subsequently, the term microscopic colitis was introduced to describe the histologic findings of mucosal inflammation without fibrosis in patients with chronic diarrhea and a normal endoscopic appearance to the colonic mucosa. The term lymphocytic colitis was coined later to emphasize the presence of intraepithelial lymphocytosis in these patients.

The clinical and pathologic spectrum of microscopic colitis has been reviewed in the 2000s (see Chapter 124).^135,136 The interrelationship between the different histologic diagnoses is not entirely clear. Because their clinical presentations are so similar and their pathologic appearances differ only in the presence or absence of a thickened subepithelial collagen table, current usage now includes lymphocytic colitis and collagenous colitis as histologic subtypes of microscopic colitis. Therefore, microscopic colitis is defined as a syndrome of watery diarrhea characterized by a normal colonoscopic appearance and histologic changes of lymphocytic-plasmacytic inflammation in the lamina propria and intraepithelial lymphocytosis, with or without thickening of the subepithelial collagen table.

Microscopic colitis is a fairly common cause of chronic diarrhea of obscure origin in the general population, as well as at referral centers. At a tertiary referral center, microscopic colitis was discovered in 10% of patients with chronic diarrhea, with an even division between the lymphocytic and collagenous subtypes.⁹³ A population-based epidemiologic study in Sweden has shown that the annual incidence of microscopic colitis is similar to that of Crohn’s disease, and the disorder was diagnosed in 10% of patients who presented with chronic nonbloody diarrhea.¹³⁷ The keys to making the diagnosis are remembering to obtain biopsies of normal-appearing colonic mucosa in patients who present with chronic watery diarrhea, taking a sufficient number of biopsy specimens (6 to 10) to avoid sampling error, and having a skilled pathologist review the biopsy slides.

Like idiopathic IBS, the cause or causes of microscopic colitis remain unknown. The disease occurs most frequently in middle-aged women and often occurs in association with autoimmune diseases, such as arthritis and hypothyroidism. A history of nonsteroidal anti-inflammatory drug use is frequent. Most fascinating is the tight linkage of lymphocytic and collagenous colitis with human leukocyte antigen (HLA)-DQ2 and HLA-DQ1,3 (including the HLA-DQ1,3 subtypes HLA-DQ1,7, HLA-DQ1,8, and HLA-DQ1,9), as in celiac disease, suggesting the possibility that similar immune mechanisms are involved in the pathogenesis of microscopic colitis and celiac disease.¹³⁸ In fact, microscopic colitis has been associated with celiac disease and may be a cause of persistent diarrhea in patients with celiac disease who are treated with a gluten-free diet.¹³⁹ Gluten is almost certainly not the causative antigen in microscopic colitis, however, because many patients with celiac disease continue to have histologic evidence of lymphocytic colitis despite treatment with a gluten-free diet, and elimination of gluten from the diet is not effective in other patients with microscopic colitis.¹⁴⁰ Bacterial antigens in the colonic lumen, rather than dietary antigens, might play an important pathogenic role in microscopic colitis.

Whatever the cause, mucosal inflammation is largely responsible for the diarrhea of microscopic colitis. Colonic perfusion studies have shown that absorption of water and salt is impaired in lymphocytic colitis and collagenous colitis.¹⁴¹ In vitro studies using human colon specimens have demonstrated decreases in sodium chloride absorption accompanied by changes in diffusion and the function of tight junctions.¹⁴² Colonic water absorption correlates inversely with the cellularity of the lamina propria, but not with the thickness of the collagen table. Net secretion of water and salt by the colon is not noted frequently, however. Typical stool weights of 500 to 1000 g/24 hr are consistent with little or no fluid absorption by the colon. Bile acid malabsorption also may play a role in the pathogenesis of diarrhea in this condition.¹⁴³

A Cochrane analysis of interventions for treating collagenous colitis has concluded that budesonide has the best evidence for efficacy.¹⁴⁴ Less well-established agents for the treatment of collagenous colitis include bismuth subsalicylate, mesalamine, and prednisone. Studies of budesonide in the treatment of lymphocytic colitis are fewer in number, and the overall evidence in support of a benefit is weaker.¹⁴⁵

Microscopic colitis remits in most patients with time, and symptomatic therapy with an antidiarrheal drug may be an appropriate option.¹⁴⁶ Bile acid sequestrants also may improve diarrhea in patients with this condition.¹⁴³

POSTSURGICAL DIARRHEA

Gastrointestinal and biliary tract surgical procedures produce a number of changes in intestinal function that may lead to diarrhea. Although surgery for peptic ulcer is performed much less commonly now than in the past, other operations performed on the gastrointestinal and biliary tracts continue to be complicated by diarrhea.

BILE ACID–INDUCED DIARRHEA

The importance of bile acid malabsorption as a mechanism for producing chronic secretory diarrhea in the absence of ileal resection or disease is controversial.^98,153,158 Bile acid malabsorption has been well described as the mechanism of diarrhea when ileal disease or resection allows excessive amounts of conjugated bile acid to enter the colon. Concentrations of bile acid in the colon higher than 3 to 5 mmol/L can inhibit electrolyte absorption and stimulate secretion by the colonic mucosa.¹⁵⁹

How often this mechanism produces chronic watery diarrhea when there is no overt ileal disease or resection is unclear. In rare congenital cases, primary bile acid malabsorption is caused by mutations in the ileal sodium-dependent bile acid transporter gene (see Chapter 64), but most cases of adult-onset bile acid malabsorption are not associated with a discrete molecular defect.^160–162 Accelerated transit through the small bowel and colon could account for impaired ileal bile acid absorption in the absence of ileal disease or molecular defects.¹⁶³ Studies from Europe and the United States have indicated that bile acid malabsorption is common in patients with idiopathic chronic diarrhea. Reports of the therapeutic effect of bile acid–sequestering resins in this setting, however, are discrepant. In many European studies, a high proportion of patients with idiopathic secretory diarrhea responded to therapeutic doses of bile acid–sequestering resins, whereas in American studies no consistent beneficial effect has been observed.

The clinical implications of these reports are clear. First, performing a sophisticated test for bile acid malabsorption, such as the selenium-75 labeled homotaurocholic acid (SeHCAT) retention test (not available in the United States) or fecal radiolabeled bile acid study is not worthwhile, because diarrhea itself may make the test abnormal. Second, such a test may not be predictive of successful therapy with a bile acid–sequestering resin. In patients with undiagnosed idiopathic diarrhea, therefore, it makes more sense to prescribe an empirical trial of a bile acid–sequestering resin. If the medication controls the diarrhea, bile acid malabsorption may be playing a role. Because such agents may also bind toxins or other luminal agonists, the possibility of a nonspecific effect must be considered.

DIARRHEA IN HOSPITALIZED PATIENTS

Diarrhea frequently develops during hospitalization, particularly in severely ill patients hospitalized for protracted periods. Common causes of diarrhea in this setting include medications, especially antibiotics, tube feedings, intestinal ischemia, and fecal impaction.

Diarrhea is a side effect of many medications, including those frequently used in hospitalized patients (see Table 15-5).⁵⁹ Antibiotic therapy is particularly likely to cause diarrhea by at least two main mechanisms, impairing carbohydrate metabolism by the colonic bacterial flora and facilitating overgrowth of C. difficile and production of toxins by the bacteria.¹⁶⁴ In some cases, erythromycin produces diarrhea by its motilin-like effect on gastrointestinal transit.

Impaired bacterial metabolism can cause diarrhea by allowing carbohydrates and associated water to remain in the intestinal lumen.¹⁶⁵ Ordinarily, all dietary fiber and about 20% of wheat starch evade digestion and absorption and reach the colon. Colonic bacteria ferment these carbohydrates to short-chain fatty acids, hydrogen, and carbon dioxide. These fermentation products and associated water are rapidly absorbed by the colon, and therefore diarrhea does not result. By contrast, when an antibiotic kills some of the normal colonic flora, fermentation decreases; undigested fiber and carbohydrates as well as water are retained in the lumen, thereby leading to an osmotic diarrhea. In some persons, intestinal transit may be modified by illness or by other drugs given concomitantly, thus leading to greater delivery of carbohydrate to the colon and further aggravating the diarrhea. Such diarrhea should subside when the patient is fasting.

C. difficile-related diarrhea is a more serious concern.¹⁶⁶ Hospitalized patients and residents of nursing homes are likely to be colonized by this organism; approximately 20% of hospitalized patients become colonized with C. difficile. Physical proximity and poor hand hygiene are major factors in its spread within institutions. Factors precipitating diarrhea in this condition include antibiotic therapy, chemotherapy, and altered immunity, including reduced gastric acidity resulting from administration of a proton pump inhibitor.¹⁶⁷ C. difficile-related disease ranges in severity from modest diarrhea to life-threatening colitis, particularly when patients are colonized with a virulent strain of C. difficile (see Chapter 108). Patients may have severe pain, abdominal tenderness, and marked polymorphonuclear leukocytosis, with an increased percentage of immature forms. In less severely ill patients suspected of having this infection, stool should be tested for C. difficile toxin and, if the result is positive, appropriate antibiotic therapy should be initiated (see Chapter 108). In more severely ill patients, analysis of stool samples for toxin may not be sufficient,¹⁶⁸ and sigmoidoscopy is performed to identify the characteristic findings of pseudomembranous colitis. Occasionally, the colitis is evident only more proximally in the colon, and colonoscopy is required to confirm the diagnosis. Therapy with metronidazole or vancomycin often suppresses the diarrhea, but recurrence is common because the organism forms spores. Probiotics, toxin-absorbing resins, and vaccines are being studied as ways to reduce recurrence.⁷²

Diarrhea also may be a complication of enteral nutrition, although it is often the result of coexisting problems.¹⁶⁹ Tube feeding, although more physiologic than parenteral nutrition, is still different from the normal presentation of nutrients to the intestine by oral ingestion of food, and the regulatory system of the gastrointestinal tract may not adapt to tube feeding. Some tube feeding formulas are hypertonic and may induce diarrhea by a mechanism similar to that of dumping syndrome. In such cases, a change in formula to one that is isotonic may be of benefit. In other cases, slowing the rate of infusion and thereby decreasing the delivery of nutrients to the intestine may be helpful, but this approach may be of limited value if the patient’s nutritional needs are not being met at the slower rate of infusion. Addition of an antidiarrheal agent such as loperamide or tincture of opium to the tube feeding may be necessary, although this approach has limitations, especially for patients in whom ileus is a risk (see Chapter 120).

Intestinal ischemia may develop in some hospitalized patients, especially those with hypotension or shock.¹⁷⁰ These patients are at risk of developing bloody diarrhea caused by ischemic colitis or more profound diarrhea if small intestinal ischemia develops (see Chapter 114).

The risk of fecal impaction is increased in older adults, patients on prolonged bowel rest, and those taking constipating drugs. Paradoxical or overflow diarrhea with incontinence may be the first clue to an impaction. Hospitalized patients in whom diarrhea develops should undergo a digital rectal examination to exclude fecal impaction.¹⁷¹

FACTITIOUS DIARRHEA

Surreptitious laxative abuse should be considered for persons in whom diarrhea remains undiagnosed, especially in those who fit into one of the following four categories (Table 15-7): (1) patients with anorexia nervosa or bulimia who use laxatives as a way of adjusting body weight (see Chapter 8); (2) those who use laxative-induced illness for secondary gain, such as disability income, or to generate concern in other persons; (3) patients with Munchausen’s syndrome, who feign illness to confound physicians; and (4) persons who are being poisoned with laxatives by their caregivers.^172,173

Table 15-7 Groups of Patients Predisposed to Laxative Abuse

A high index of suspicion is required for detecting laxative abuse. Physicians usually assume that patients are being truthful, but up to 15% of patients who undergo an evaluation for chronic diarrhea may be abusing laxatives surreptitiously.¹⁷⁴ Clues to laxative abuse may be found during the evaluation for chronic diarrhea. For example, hypokalemia may suggest ingestion of stimulant laxatives, such as senna. Detection of pseudomelanosis coli, a brownish pigmentation of the colonic mucosa, suggests chronic ingestion of anthracene laxatives, such as senna or cascara (see Chapter 124). The presence of a large fecal osmotic gap suggests magnesium ingestion.

In patients who belong to one of the groups listed in Table 15-7 and in those with diarrhea that remains undiagnosed after evaluation, stool samples should be analyzed for laxatives with standardized methods. Most laxatives can be detected by spectrophotometry or chromatography, but the accuracy of commercial analysis has been called into question.⁸² Because some patients exaggerate stool volume by adding urine or water, stool osmolality should be measured as well; a value lower than 290 mOsm/kg suggests dilution of the stool with water or hypotonic urine. Admixture of stool with hypertonic urine often leads to an impossibly high fecal osmolality (typically, >600 mOsm/kg) and to a negative fecal osmotic gap because of high concentrations of sodium and potassium in the urine. A negative fecal osmotic gap also can be noted when phosphate or sulfate osmotic laxatives are ingested. In today’s legal environment, unauthorized room searches for laxatives should not be conducted.

When a diagnosis of laxative abuse is made, an effort should be made to confirm the diagnosis with repeated stool analyses before discussion with the patient or family. The patient should be confronted with the findings, but not before plans for the aftermath are made. Psychiatric consultation should follow the discussion with the patient because some persons who abuse laxatives become suicidal after being discovered and all patients who abuse laxatives need counseling. In cases of laxative administration by a parent or caregiver, legal proceedings should be instituted to separate the patient from the abuser.

Few outcome studies of the effect of discovery of laxative abuse are available. In one study of 11 patients seen at the Cleveland Clinic, 6 said that they were improved and 5 claimed no benefit; 4 of these 5 unimproved patients sought further medical attention elsewhere for chronic diarrhea.¹⁷⁵

IDIOPATHIC SECRETORY DIARRHEA

When an exhaustive evaluation fails to reveal a cause of chronic diarrhea and stool analysis suggests a secretory diarrhea, the diagnosis of idiopathic secretory diarrhea is made. This condition often starts suddenly in a previously healthy person and is differentiated from the many similar acute diarrheal illnesses by persisting beyond four weeks. It occurs in two forms, epidemic and sporadic.^63,64

The epidemic form of secretory diarrhea occurs in outbreaks seemingly linked to contaminated food or drink.^63,176–180 The initial description of this condition resulted from an outbreak in Brainerd, Minnesota, thus giving this condition its common appellation, Brainerd diarrhea. Several outbreaks have been described in the literature in different communities and even on a cruise ship. Although the epidemiology suggests an infectious cause, no causative agent has been identified in these outbreaks.

Sporadic idiopathic secretory diarrhea affects persons in a fashion identical to that of the epidemic form but does not seem to be acquired easily by family members or others.⁶⁴ Many affected persons give a history of travel, but to destinations not usually associated with travelers’ diarrhea. Diarrhea begins abruptly and reaches its maximum intensity soon after onset. Weight loss of up to 20 pounds is characteristic and almost always occurs within the first few months of illness and not thereafter. Empirical trials of antibiotics and bile acid–binding resins are ineffective. Nonspecific opioid antidiarrheal agents may provide symptomatic improvement.

Both forms of idiopathic secretory diarrhea have a self-limited course and usually resolve within two years of onset. The resolution of idiopathic secretory diarrhea occurs gradually over two to three months. Understanding this natural history can be a solace to patients, who may otherwise feel mired in an unending illness. Idiopathic secretory diarrhea may share several clinical characteristics with functional diarrhea but in general has a more discrete onset and is associated with higher stool volumes.

DIARRHEA OF OBSCURE ORIGIN

Physicians sometimes fail to make a specific diagnosis in patients with chronic diarrhea, despite an elaborate evaluation, and may refer these patients to centers interested in this condition. Common diagnoses resulting from reevaluation of these patients are shown in Table 15-8.

Table 15-8 Frequent Diagnoses in Patients with Diarrhea of Obscure Origin

Although unusual or obscure conditions that require special tests might be expected to predominate in this group of patients, most of the eventual diagnoses are straightforward and could have been made sooner.⁹³ Fecal incontinence and iatrogenic diarrhea could be recognized with a careful history. Surreptitious laxative ingestion and microscopic colitis could be diagnosed with an appropriate index of suspicion and testing (e.g., laxative screen and colonic biopsy, respectively). Bile acid–induced diarrhea, small intestinal bacterial overgrowth, pancreatic exocrine insufficiency, and carbohydrate malabsorption could be discovered with a detailed history and a properly conducted therapeutic trial. Peptide-secreting tumors are rare, but serum peptide assays and scanning techniques (e.g., CT scanning and octreotide scanning) are widely available. Failure to make a diagnosis typically results from failure to appreciate the evidence at hand and to think through the differential diagnosis of chronic diarrhea.