Introduction

The term carbohydrate malabsorption is used to describe conditions in which carbohydrates escape digestion and/or absorption in the small intestine and reach the colon. Although this is usually considered to be a consequence of a malabsorptive disease of the pancreas or small intestine, carbohydrates also reach the colon in physiologically incomplete carbohydrate absorption. This is due to ingestion of carbohydrates for which the healthy gastrointestinal (GI) tract of the human has a limited digestive or absorptive capacity. Carbohydrate malabsorption may result in symptoms due to complete or incomplete bacterial metabolism of carbohydrates in the colon. These symptoms are bloating, abdominal cramping, passing of gas, flatulence, and diarrhea (carbohydrate intolerance). It is important for physicians and for patients to remember that carbohydrate intolerance does not necessarily imply that a malabsorptive disease is present, but may be the result of individual dietary habits or intolerances, probably related to increased intestinal sensitivities.

Symptoms of carbohydrate intolerance may result from

• 1.
  

  Inborn or acquired defects of luminal or membrane bound pancreatic or intestinal enzymes that are needed for digestion of polysaccharides, oligosaccharides, or disaccharides

  
  
• 2.
  

  Decreased absorption of monosaccharides due to primary defects in mucosal absorptive mechanisms or secondary to extensive reduction of small intestinal absorptive surface area due to small bowel diseases or extensive resections

  
  
• 3.
  

  Ingesting carbohydrates for which there is a physiologically limited or totally absent digestive and/or absorptive capacity, like fructose, mannitol, sorbitol, or dietary fibers

  
  
• 4.
  

  Adverse effects of therapy or dietary intervention with nonabsorbable carbohydrates (lactulose, lactitol), drugs interfering with carbohydrate absorption (acarbose, metformin), or sugar exchange products (sorbitol, fructose)

  
  
• 5.
  

  Antibiotics interfering with colonic salvage of malabsorbed or physiologically incompletely absorbed carbohydrates

This article will focus on the role of the colon in the pathogenesis of diarrhea in carbohydrate malabsorption or physiologically incomplete absorption of carbohydrates, and on the most common manifestation of carbohydrate malabsorption, lactose malabsorption. In addition, incomplete fructose absorption, the role of carbohydrate malabsorption in other malabsorptive diseases, and congenital defects that lead to malabsorption will be covered. The article concludes with a section on diagnostic tools to evaluate carbohydrate malabsorption.

Methods for establishing the role of carbohydrate malabsorption for the pathogenesis of diarrhea

Carbohydrates that are not absorbed in the small intestine are metabolized by colonic bacteria to organic acids. These organic acids are lactic acid and the short chain fatty acids respectively their salts acetate, propionate and butyrate. Most of the organic acids are absorbed across the colonic mucosa. Carbohydrates that are not metabolized by colonic bacteria to organic acids, and organic acids that escape absorption by the colon, remain in the colonic lumen and lead to osmotic diarrhea and fecal calorie loss.

Whether carbohydrates exert osmotic activity in the colon, thereby resulting in colonic water and electrolyte accumulation and diarrhea, depends on their molar concentration (expressed in mmol/L). For assessment of their osmotic activity, products of bacterial carbohydrate metabolism, that is organic acids, have to be taken into account. Lastly, for calculation of the osmotic activity of malabsorbed or incompletely absorbed carbohydrates, also the cations, which are bound by anionic salts of organic acids, have to be considered. The authors have previously established that 1 mmol of fecal organic acids obligates an average of 0.6 mmol cations (0.30 mmol Na, 0.21 mmol Ca, 0.07 mmol K, and 0.02 mmol Mg).

A comprehensive picture of the magnitude of carbohydrate malabsorption and its significance for the pathogenesis of diarrhea in malabsorptive diseases was gained in studies in which fecal outputs of carbohydrates and organic acids were measured in normal subjects with diarrhea induced by ingestion of lactulose, and in patients with severe malabsorption. Carbohydrates were assayed in 24-hour stool collections using anthrone, a method that measures all hexose carbohydrates, whether excreted as monosaccharides, disaccharides, or oligosaccharides. The unit of expression is grams excreted per day. This method provided a measure of total fecal carbohydrate excretion, regardless of molecular size. A reducing sugar assay was used to detect the reducing ends of carbohydrate molecules; 1 mol of starch (>50,000 g of carbohydrate) and 1 mol of glucose (180 g of carbohydrate) give the same result. Therefore, the reducing sugar assay provided information on the number of moles of excreted carbohydrate and was used to determine the osmotic effect of fecal carbohydrates. The reducing sugar assay has to be performed in lyophilized stool samples that have been reconstituted to their original weight by adding water to remove noncarbohydrate volatile reducing substances. In a later study, individual carbohydrates were measured by high-performance liquid chromatography (HPLC), and their molar or gram amounts were calculated by summing individual carbohydrates.

Fecal output of total organic acids quantifies the fraction of carbohydrates that are excreted in stool as a bacterial product of unabsorbed carbohydrates. In initial studies analyzing their osmotic effects and outputs organic acids were analyzed by titration assay, which measures total organic acid concentration ; in a later study by the authors’ group, individual acids were measured by HPLC.

In summary, to assess osmotic activity and the role of carbohydrate malabsorption for the pathogenesis of diarrhea, the sum of carbohydrates plus organic acids plus cations (that is measured organic acids in mmol/L × 0.6) has to be calculated and expressed as millimoles per liter. In diarrhea with only slightly increased stool weight, organic acids and electrolytes are the main driving force of carbohydrate-induced osmotic diarrhea. Only in severe carbohydrate malabsorption, with stool weight in excess of 500 g/d fecal, does carbohydrate excretion become a driving force.

Methods for establishing the role of carbohydrate malabsorption for the pathogenesis of diarrhea

Carbohydrates that are not absorbed in the small intestine are metabolized by colonic bacteria to organic acids. These organic acids are lactic acid and the short chain fatty acids respectively their salts acetate, propionate and butyrate. Most of the organic acids are absorbed across the colonic mucosa. Carbohydrates that are not metabolized by colonic bacteria to organic acids, and organic acids that escape absorption by the colon, remain in the colonic lumen and lead to osmotic diarrhea and fecal calorie loss.

Whether carbohydrates exert osmotic activity in the colon, thereby resulting in colonic water and electrolyte accumulation and diarrhea, depends on their molar concentration (expressed in mmol/L). For assessment of their osmotic activity, products of bacterial carbohydrate metabolism, that is organic acids, have to be taken into account. Lastly, for calculation of the osmotic activity of malabsorbed or incompletely absorbed carbohydrates, also the cations, which are bound by anionic salts of organic acids, have to be considered. The authors have previously established that 1 mmol of fecal organic acids obligates an average of 0.6 mmol cations (0.30 mmol Na, 0.21 mmol Ca, 0.07 mmol K, and 0.02 mmol Mg).

A comprehensive picture of the magnitude of carbohydrate malabsorption and its significance for the pathogenesis of diarrhea in malabsorptive diseases was gained in studies in which fecal outputs of carbohydrates and organic acids were measured in normal subjects with diarrhea induced by ingestion of lactulose, and in patients with severe malabsorption. Carbohydrates were assayed in 24-hour stool collections using anthrone, a method that measures all hexose carbohydrates, whether excreted as monosaccharides, disaccharides, or oligosaccharides. The unit of expression is grams excreted per day. This method provided a measure of total fecal carbohydrate excretion, regardless of molecular size. A reducing sugar assay was used to detect the reducing ends of carbohydrate molecules; 1 mol of starch (>50,000 g of carbohydrate) and 1 mol of glucose (180 g of carbohydrate) give the same result. Therefore, the reducing sugar assay provided information on the number of moles of excreted carbohydrate and was used to determine the osmotic effect of fecal carbohydrates. The reducing sugar assay has to be performed in lyophilized stool samples that have been reconstituted to their original weight by adding water to remove noncarbohydrate volatile reducing substances. In a later study, individual carbohydrates were measured by high-performance liquid chromatography (HPLC), and their molar or gram amounts were calculated by summing individual carbohydrates.

Fecal output of total organic acids quantifies the fraction of carbohydrates that are excreted in stool as a bacterial product of unabsorbed carbohydrates. In initial studies analyzing their osmotic effects and outputs organic acids were analyzed by titration assay, which measures total organic acid concentration ; in a later study by the authors’ group, individual acids were measured by HPLC.

In summary, to assess osmotic activity and the role of carbohydrate malabsorption for the pathogenesis of diarrhea, the sum of carbohydrates plus organic acids plus cations (that is measured organic acids in mmol/L × 0.6) has to be calculated and expressed as millimoles per liter. In diarrhea with only slightly increased stool weight, organic acids and electrolytes are the main driving force of carbohydrate-induced osmotic diarrhea. Only in severe carbohydrate malabsorption, with stool weight in excess of 500 g/d fecal, does carbohydrate excretion become a driving force.

The role of the colon for colonic salvage of malabsorbed or incompletely absorbed carbohydrates and gas production

The colon has the capacity to absorb a limited variety of nutrients and minerals. Although colonic nutrient absorption does not play a major role in health, the nutritive role of the colon in patients with severe malabsorption has been demonstrated.

In healthy people, between 2% and 20% of ingested starch escapes absorption in the small intestine ; pancreatic insufficiency and severe intestinal disorders further increase this amount. Carbohydrates that reach the colon cannot be absorbed by the colonic mucosa but can be metabolized by the bacterial flora. Anaerobic bacterial metabolism results in the breakdown of oligosaccharides and polysaccharides to mono- and disaccharides, which are metabolized further to lactic acid and short-chain (C2 to C4) fatty acids, such as acetate, propionate, and butyrate, and to odorless gases, including hydrogen, methane, and carbon dioxide. Although considerable proportions of these organic acids and gases are absorbed by the colonic mucosa, some remain in the colon, resulting in diarrhea and symptoms due to gas-like bloating, flatulence, and cramping. The same principle also applies to dietary fibers, which have been calculated to contain 2.5 to 3.1 kcal of metabolizable energy per gram fiber, which is liberated and made available for colonic absorption.

Studies in normal subjects have shown that the bacterial metabolism of starch to small carbohydrate moieties is a rapid process in the normal colon. The rate-limiting step in the overall conversion of polysaccharides to short-chain fatty acids is the conversion of monosaccharides to short-chain fatty acids. Colonic absorption of short-chain fatty acids results in a reduction of the osmotic load and in mitigation of diarrhea. In normal subjects, more than 45 g of carbohydrates must reach the colon to cause diarrhea, and up to 80 g of carbohydrates per day can be metabolized by bacteria to short-chain fatty acids; approximately 90% of these short-chain fatty acids are absorbed by colonic mucosa. Chronic carbohydrate malabsorption causes adaptive changes in bacterial metabolic activity that result in an even higher efficiency of the bacterial flora to digest carbohydrates although at the expense of increased flatus production.

Because short-chain fatty acids have caloric values between 3.4 and 5.95 kcal/g, colonic absorption of these acids decreases caloric loss in carbohydrate malabsorption. In patients with short bowel syndrome, colonic salvage of malabsorbed carbohydrates can contribute up to 700– to 950 kcal/d, provided that a substantial part of the colon remains in continuity with the small bowel.

The beneficial effects of colonic bacterial carbohydrate metabolism are accompanied by side effects due to gas production. Up to 8-fold interindividual differences in the volume of gas produced in the colon have been observed in normal persons after a dose of 12.5 g lactulose The colon can absorb gas; if intracolonic gas volumes are low, up to 90% of the volume of intracolonic gas can be absorbed. However, if gas volumes are high, this proportion decreases to as low as 20%. Therefore, persons who have the disadvantage of producing more gas in their colon have an additional disadvantage in that they absorb a smaller fraction of the gas. Gas produced from bacterial carbohydrate metabolism is odorless. The odor of flatus is due to volatile sulfur-containing substrates that result from bacterial metabolism of protein.

Impaired colonic salvage of carbohydrates has been suggested to contribute to the diarrhea in Crohn disease and ulcerative colitis. Bacterial carbohydrate metabolism may be decreased by antibiotic treatment. In some patients, antibiotic-associated diarrhea may be the result of impaired colonic salvage of physiologically incompletely absorbed carbohydrates or dietary fibers, which may accumulate in stool because of decreased bacterial fermentation.

Influence of antibiotics on colonic carbohydrate salvage

It has been suggested that in normal subjects, as much as 70 g of carbohydrates reaches the colon per day. The colon cannot absorb carbohydrates, but colonic bacteria metabolize carbohydrates as an energy source, thereby reducing osmotic load and salvaging calories for the human host.

Inhibition of bacterial carbohydrate metabolism by antibiotics may result in an increase of osmotically active solutes in the colon, resulting in osmotic diarrhea. In addition, reduced production of organic acids may result in functional disturbances of colonic mucosa. In the distal colon, the short chain fatty acid (SCFA) n-butyrate is an important source of energy for the mucosa through cellular oxidation. Reduction in SCFA production may deprive the colonic mucosa of an energy source, as demonstrated by the clinical model of exclusion colitis (diversion colitis) in patients with exclusion of distal parts of the colon from the fecal stream.

Various antibiotics have been shown to reduce colonic bacterial carbohydrate metabolism. Reduction of carbohydrate metabolism can result in osmotic diarrhea. However, the development of diarrhea is dependent on the quantity of poorly absorbable dietary carbohydrates like vegetables and dietary fibers.

An in vitro study simulating luminal contents of the proximal colon has demonstrated that clindamycin reduces anaerobes (clostridium and bacteroides species), decreases fecal carbohydrate metabolism, and decreases concentrations of SCFA. Ampicillin reduces colonic bacterial carbohydrate fermentation, as shown by a decreased breath hydrogen response and by the presence of carbohydrates in stool.

Inhibition of carbohydrate salvage contributes to the development of diarrhea, as indicated by an increase in frequency and stool weight in subjects who ingested a subdiarrheal dose of lactulose. Reduced colonic carbohydrate fermentation has also been shown to occur in healthy subjects following oral administration of ampicillin and metronidazole. Patients with diarrhea associated with pivampicillin, dicloxacillin, erythromycin, or ampicillin plus netilmicin had reduced fecal concentrations of SCFA. In the same study, another group of patients treated with erythromycin, dicloxacillin, or a combination of ampicillin, netilmicin, and metronidazole also had reduced fecal concentrations of SCFA but did not develop diarrhea. However, monotherapy with penicillin or pivampicillin did not reduce fecal SCFA concentrations or result in diarrhea.

In another study, SCFA concentrations were measured in the stool of 15 liver-transplanted patients who received bowel-suppressing antibiotics consisting of cefuroxime, tobramycin, and nystatin. Thirteen of them developed Clostridium difficile -negative diarrhea, and the levels of SCFA in the stools were very low, possibly because of nearly complete suppression of colonic bacterial fermentation. The diarrhea resolved before cessation of antibiotic therapy and normalization of the fecal SCFA levels.

Although fecal carbohydrates were measured in none of these studies, decreased SCFA concentrations most likely indicate decreased bacterial metabolism of incompletely absorbed carbohydrates.

Discrepancies between suppression of carbohydrate metabolism and manifestation of diarrhea suggest that disturbed carbohydrate metabolism was not the only mechanism responsible for diarrhea. A possible reason for the observed discrepancies may be an adaptive increase in colonic transit time. It has been shown that colonic transit time in osmotic diarrhea increases in the descending colon. This may provide more time for metabolism of carbohydrates and absorption of SCFA, water, and electrolytes.

The role of colonic transit in carbohydrate-induced diarrhea

Diarrhea that develops due to malabsorbed carbohydrates may be seen as the failure of colonic salvage. This may be due to excessive amounts of carbohydrate entering the colon, or decreased time available for metabolism of malabsorbed carbohydrates and absorption of organic acids and electrolytes. Accelerated colonic transit (that is decreased transit time) may result from an increase in the volume load to the colon, a change in the composition of the substrate entering the colon, or an alteration of colonic motility. Accelerated transit decreases time available for storage, bacterial metabolism, and absorption.

Malabsorption of carbohydrates in the small intestine can increase the volume that flows from the ileum into the colon to as much as 10 L/d. While an increase of the colonic fluid load per se accelerates colonic transit, especially transit through the proximal colon, the colon also possesses an ability to accommodate and store excess fluid volumes. When the storage capacity of the proximal colon is surpassed, the distal colon may provide reserve capacity that may provide additional time to absorb and metabolize colonic contents. The total time the colonic contents reside in the lumen is important for providing the time for bacterial metabolism of carbohydrates and for mucosal absorption of organic acids, electrolytes, and fluids. Under optimal conditions, the colon can absorb as much as 6 L of electrolyte solutions. However, the osmotic forces in carbohydrate malabsorption counteract the absorptive capacity of the colonic mucosa. Carbohydrate malabsorption results in a dose-dependent acceleration of colonic transit as a result of a fluid overload of the colon.

A study that compared the effect of malabsorbed degradable carbohydrates (lactulose) on colonic transit with an osmotically active, nondegradable solution (polyethylene glycol, PEG) demonstrated that carbohydrate malabsorption has a differential effect on colonic transit. An acceleration of colonic transit during carbohydrate ingestion takes place secondary to an increased intracolonic volume. This is partly counteracted by slowing of transit due to short chain fatty acids. The high concentration of short chain fatty acids in carbohydrate malabsorption reduces colonic motor activity. In the study comparing colonic transit in lactulose and PEG-induced diarrhea, increased short chain fatty acid concentrations resulted in an inhibition of colonic transit as compared with similar osmotic loads of PEG.

The size of the differences in colonic transit during ingestion of lactulose and PEG at doses that result in similar severity of diarrhea, as assessed by 24-hour stool weight, is quite remarkable; during ingestion of lactulose, there is a delay in mean residence time of colonic contents in the distal colon by 250% as compared with PEG ingestion. Mean residence time in the total colon after lactulose is also increased, but only by 50% as compared with PEG. Therefore due to the composition of colonic contents during carbohydrate malabsorption, the colon can accommodate up to 6 times higher osmotic loads of malabsorbed carbohydrates as compared with osmotic loads of PEG.

The complex relation between colonic contents, colonic volume, and transit during carbohydrate malabsorption provides further explanation for the nonlinear dose–effect curve of lactulose, in that bacterial metabolism of lactulose and absorption of its metabolic products (that is organic acids) do not only result in a reduction of the osmotic load, but also have slowing effects on colonic transit, which in turn may result in increased time for bacterial metabolism of carbohydrates and mucosal absorption of short chain fatty acids.

The inhibitory effects of short chain fatty acids on colonic transit deserve further evaluation for their possible clinical role in the pathogenesis and treatment of slow transit constipation. Short chain fatty acids are products of bacterial metabolism of dietary fibers, which, because of their bulking effect and the presumed acceleration of colonic transit, are a first-line treatment of constipation. Treatment failure of dietary fibers could be related to inhibitory effects of short chain fatty acids on transit in some susceptible patients.

Diarrhea in carbohydrate malabsorption—multiple influencing factors

Several effects influence the clinical manifestation of diarrhea in carbohydrate malabsorption. First, osmotic fluid retention due to carbohydrates, organic acids, and electrolytes results in a dose-dependent acceleration of colonic transit, primarily in the proximal but also in the distal colon. Second, stool weight in carbohydrate malabsorption is correlated to colonic transit, but it is primarily influenced by the unabsorbed osmotic load of carbohydrates, short chain fatty acids, and electrolytes. Third, stool consistency is influenced both by stool composition (carbohydrates, organic acids, electrolytes) and transit in the distal colon. Fourth, at similar stool weight, colonic transit is significantly slower during carbohydrate malabsorption as compared with colonic volume load of other origins.

Carbohydrate malabsorption in malabsorptive diseases: innocent bystander or relevant for calorie loss and diarrhea?

In an analysis of 19 patients with severe malabsorption syndrome, 12 patients had excessive fecal excretion of carbohydrates and organic acids. In 6 of these 19 patients, carbohydrate malabsorption was a major cause of diarrhea. Furthermore, carbohydrate malabsorption contributed significantly to fecal calorie loss in some of these patients.

Patients with combined small bowel and colon resection had abnormally high fecal excretion of carbohydrates and organic acids (between 11 and 78 g/d, normal being up to 5 g/d). Two factors probably contributed to high fecal excretion of carbohydrates and organic acids after small bowel plus colon resection. First, the shortened small intestine of these subjects might absorb an abnormally small fraction of ingested dietary carbohydrates, and second, their colonic conversion of carbohydrates to organic acids and their colonic absorption of these acids might be reduced.

Three of 5 patients with pancreatic disease also had excessive fecal excretion of carbohydrates and organic acids (9–21 g/d), but in these cases excess fecal carbohydrate excretion was relatively mild compared with fecal fat excretion, which was between 30 and 113 g/d in these patients.

Two of 6 patients with villous atrophy had excessive fecal excretion of carbohydrates and organic acids (14 and 103 g/d, respectively), and 1 patient had the highest fecal carbohydrate excretion of the entire group.

In 2 patients, fecal calories due to carbohydrates and organic acids approached fecal calories due to fat.

In 3 of the 5 patients with pancreatic diseases values for osmotically active reducing sugars plus organic acids plus obligated cations suggested, diarrhea was mainly due to the osmotic activity of malabsorbed carbohydrates. In 2 of those 3 patients, organic acids were excreted in excessive amounts more often than osmotically active carbohydrates of small molecular size. Also in 2 of the patients with intestinal resection and in 1 of the patients with villous atrophy, carbohydrate malabsorption was the dominant mediator of increased fecal water output.

In the other 13 patients with malabsorption syndrome, the data do not suggest a major role for carbohydrate malabsorption as a mediator of increased fecal water output even though several of them had increased fecal output of carbohydrates, however not in an osmotically relevant form. This means that carbohydrates were in a large molecular form (eg, starch).