Effect of Psyllium, Calcium Polycarbophil, and Wheat Bran on Secretory Diarrhea Induced by Phenolphthalein

ANDREAS J. EHERER, CAROL A. SANTA ANA, JACK PORTER, and JOHN S. FORDTRAN

Department of Internal Medicine, Baylor University Medical Center, Dallas, Texas

Results

Effect of Psyllium, Calcium Polycarbophil, and Wheat Bran

As shown in Table 1, average stool frequency was 3.8 times per day in subjects taking placebo, and frequency was not changed by psyllium, calcium poly- carbophil, or bran. Total fecal weight, fecal water, fecal solids, and fecal cations (Na⁺ and K⁺) were significantly increased by all three agents; however, because fecal water and fecal solids were increased proportionally, percent water and percent solid content of the feces did not change. The absolute increase in fecal solids was 10 g/day when 18 g/day of psyllium was ingested, 14 g/day when 6 g/day of calcium polycar- bophil was ingested, and 20 g/day when 18 g fiber per day was ingested as wheat bran.

Average stool viscosity of subjects taking placebo was 16,470 cp. Viscosity was not changed significantly by calcium polycarbophil or bran but was increased to 35,622 cp with psyllium therapy (P < 0.001). In every subject, the viscosity was higher with psyllium than with placebo, as shown in Figure 1.

Subjective description of stools is provided in Table 2. For subjects taking placebo, most of the stools were judged to be semiliquid, and this was not altered by calcium polycarbophil or by bran. On the other hand, when subjects were taking psyllium the stools had a greater consistency, usually described as soft (P < 0.01). Table 2 also shows the results of an unblinded evaluation of urgency and/or fear of fecal incontinence. Although these symptoms were reported less during psyllium treatment, the differences did not reach statistical significance.

Comparison of Different Doses of Psyllium

As shown in Table 3, increasing doses of psyllium caused a progressive increase in fecal water out-

Figure 1. Effect of psyllium, calcium polycarbophil, and wheat bran on stool viscosity in 9 subjects with secretory diarrhea induced by phenolphthalein. Data represent the average viscosity of each subject during a 3-day stool collection.put, fecal solids, and fecal cations. Because total water and total solids increased proportionately, there was no change in percent total water or percent solid content of stools. Increasing doses of psyllium also tended to increase stool consistency by visual description and increased stool viscosity in an approximately linear fashion.

Assuming that psyllium is not degraded or absorbed as it passes through the intestine,³ the psyllium concentration in stool can be calculated by dividing the ingested psyllium dose by the total stool weight. Doses of 9, 18, and 30 g of psyllium per day resulted in calculated psyllium concentrations in stool of 1.4%, 2.8%, and 4.2%, respectively.

Discussion

Many patients with chronic diarrhea have underlying diseases that are incurable, and they therefore must rely on nonspecific therapy. Although opiate derivatives are useful for this purpose, response to such drugs is highly variable and often incomplete.⁴ Therefore, some patients continue to have chronic diarrhea in spite of treatment with opiates. If water-holding or fiber products improve stool consistency in patients with chronic diarrhea without clinically important side effects, these products may improve the well-being of such patients.

We examined the effect of water-holding and fiber products in normal subjects in whom diarrhea was induced for experimental purposes by phenolphthalein. Phenolphthalein causes secretory diarrhea by inhibition of Na/K adenosine triphosphatase, inhibition of chloride/bicarbonate exchange, and stimulation of electrogenic anion secretion⁵ in both small bowel and colonic mucosa.⁶ In our subjects treated with placebo, phenolphthalein increased stool weight from an expected normal of about 150 g/day⁷ to about 600 g/day, with an increase in stool frequency from an expected normal of 1 to 3.8 times per day. Stools were usually

semiliquid in consistency. Therefore, this model results in fecal characteristics similar to those of the majority of patients with chronic diarrhea that have been reported in the literature.^4,8,9,10

Because the most likely possible benefit of the products we tested was an improvement in the consistency of stools, and because stool consistency has no precise definition and cannot be objectively quantitated, we decided to measure stool viscosity as an index of consistency. The numerical results obtained are based on the force required to rotate a cylinder of defined dimensions in a mixed stool specimen. We found the method gave reproducible values for any specific diarrheal stool specimen. Moreover, when average values for viscosity are considered, there was good agreement between viscosity and stool consistency by subjective visual inspection.

Calcium polycarbophil is a ''hydrophilic drug" that is currently marketed as a treatment for both constipation and diarrhea.¹¹ According to one report, polycarbophil exhibits a water holding capacity 3 times that of psyllium.² In our studies, calcium polycarbophil was effective as a water holding agent because it increased fecal excretion of water； however, it did not increase stool viscosity or improve stool consistency. The mechanism by which it holds water is unknown, but it does not "sequester" stool water in such a way that stool viscosity is increased in people with diarrhea.

The ratio of fecal water to fecal solids is one of the factors that supposedly determines fecal consistency. If bran increased fecal solids to a greater extent than to which it increased fecal water, bran might improve fecal consistency in people with diarrhea. However, our results showed that bran increased water and solid output commensurately, and there was no improvement in either fecal consistency or in measured viscosity. Wheat bran also increased fecal losses of sodium and potassium. Our studies therefore suggest that bran, like calcium polycarbophil, has only adverse effects in people with diarrhea.

Table 3. Effects of Different Doses of Psyllium in Six Subjects with Secretory Diarrhea Induced by Phenolphthalein

NOTE. Data are expressed as mean 士 SEM.

Recent reviews do not mention psyllium as a useful remedy for diarrhea,¹² and we are not aware of any psyllium products that are marketed for antidiarrheal activity. Although two reports^2,13 have suggested that psyllium might be helpful in patients with diarrhea, these did not compare psyllium with a placebo, and in one instance psyllium was used in combination with calcium carbonate¹³ so that the perceived benefit might not have been due to psyllium per se. Our results in normal subjects with phenolphthalein-induced diarrhea show that 18 gof psyllium per day improved fecal consistency, doubled fecal viscosity, and increased the fecal output of water, cations, and solids. The enhanced output of water was presumably due to binding of intestinal water by psyllium, preventing such water from being absorbed, and thus obligating its excretion in stool. This bound water may have originated from water ingested with psyllium or with the diet, or from various gastrointestinal secretions. The percent increase in fecal water and cation outputs were similar, suggesting that psyllium "holds" or "binds" intestinal solution, not just intestinal water. The increase in fecal solids can be explained by the nonabsorbability of ingested psyllium.

Previous investigators have given psyllium in similar doses to other types of people and measured the effect on water output. When these previous results and our current results are normalized per unit intake of psyllium, 1 g of psyllium per day increased water output by 3.0 g/day in normal subjects,¹⁴ by 5.9 g/day in subjects with phenolphthalein-induced diarrhea, and by 12.8 g/day in patients with an ileostomy.¹⁵ At least two possible explanations for these differences can be suggested. First, if psyllium is to some extent degraded during passage through the colon, psyllium might obligate more water excretion in people with an ileostomy; moreover, shorter transit time through the colon in people with diarrhea might reduce psyllium degradation and thereby increase water output when compared with people without diarrhea. Second, the water-absorptive mechanisms of the small and large intestines may compete against psyllium for water in the intestinal lumen. The small intestine may compete least effectively, and people with normal colon (no phenolphthalein, for example) may compete most effectively.

It seems rather remarkable that an agent that increases fecal water output would simultaneously increase fecal viscosity and consistency. The mechanism of this psyllium effect is not known. Because psyllium increased fecal solids and fecal water commensurately, improved viscosity and consistency was not due to an increase in the ratio of fecal solids to fecal water. However, psyllium may have sequestered fecal water and thereby decreased the relative amount of "free" water available to interact with other stool solids. On the other hand, psyllium forms a viscous gel when it is added to pure water, so the psyllium effect on stool viscosity and consistency may be related to its gel forming property per se and may not be dependent on the relative amounts of free water and nonpsyllium solids in stool.

Although it is clear that our results in this experimental model cannot be extrapolated directly to patients with diarrhea, three clinical points nevertheless deserve consideration. The first is the extent to which patients with diarrhea would be benefited by psyllium therapy. In our study, psyllium significantly improved stool consistency as evaluated by a technician who was unaware of the treatment the subjects were receiving. The subjects themselves reported less urgency and/or fecal incontinence during psyllium therapy but this symptomatic improvement is inconclusive because the subjects were not blinded to the treatment they were receiving and because the differences were not statistically significant. Further studies designed specifically to evaluate symptoms in patients with diarrhea will be necessary to establish whether or not improved fecal viscosity by psyllium will translate into clinically significant symptomatic relief.

Second, assuming that increasing fecal consistency would benefit patients with diarrhea, it is interesting to consider the dose of psyllium that should be used for this purpose. In our experimental model of human diarrhea in which stool weight in the absence of psyllium therapy averaged about 600 g/day, stool viscosity was increased in a linear fashion as the daily dose of psyllium was increased from 9 to 18 to 30 g/day. Stool viscosity was substantially increased when the calculated psyllium concentration in stool was about 3%- 4%. Many patients with chronic diarrhea have stool weights between 200 and 600 g/day,^4,8,9,10 and in such patients psyllium in doses of 9 to 30 g/day (depending on stool weight) should achieve an apparent psyllium concentration in stool of 4% or higher, which should cause a substantial increase in stool viscosity. On the other hand, even 30 g/day of psyllium is not likely to substantially increase stool viscosity in a patient whose baseline stool weight exceeds 1,000 g/day. It is not known whether or not psyllium could be tolerated in doses larger than 30 g/day.

Finally, all three of the agents we tested resulted in an increase in stool water, sodium, and potassium output. However, the magnitude of the increases were small compared with the magnitude of daily intake of water and electrolytes. In adults with mild to moderate diarrhea, it seems unlikely that extra fecal losses of 100—200 g of water, 10—20 mEq of sodium, and 5—15 mEq of potassium would be of clinical significance.