| |||||||||||||||||||||||
We recently studied two patients with chronic unexplained diarrhea who were found to have diluted their stool samples. Dilution was suspected when the measured stool osmolality was found to be considerably lower than plasma osmolality, suggesting that water (or another dilute fluid) had been added to the stool specimens. The diagnosis was confirmed by finding normal stool osmolality when defecation was supervised or when colonic contents were sampled endoscopically. We report the clinical findings in this variant of factitious disease and discuss the value of measuring stool osmolality as an objective test for the diagnosis.
Case Reports
Two patients were hospitalized for evaluation of chronic diarrhea at the Yale University General Clinical Research Center. They were instructed to collect stool and urine separately, and the stool collections were handled by staff members who were experienced in the clinical assessment of diarrhea. Osmolality was measured by freezing-point depression in supernatants of stool specimens centrifuged at 3000 rpm for five minutes. Colonoscopy was performed without colonic preparation.
Patient 1
Patient 1 was a 25-year-old female nurse in whom watery, nonbloody diarrhea developed while she was on vacation in St. Lucia. She continued to have 5 to 12 stools per day after her return to the United States. No stool pathogens were identified, and the results of colonoscopy and ileal, colonic, and duodenal biopsies were normal. The results of a Schilling test with intrinsic factor were abnormal, and the diagnosis of tropical sprue was entertained. The patient was treated with tetracycline with partial improvement, but her diarrhea persisted. The results of repeated colonic and duodenal biopsies and duodenal aspiration were normal.
The patient was hospitalized seven months after the onset of diarrhea to document its magnitude. During the first 48 hours, 1870 ml of watery stool was collected. The stool osmolality was 16 mOsm per kilogram of water, and the stool sodium, potassium, and magnesium concentrations were <10, 2.0, and 0.98 mmol per liter, respectively (Table 1). Alkalinization did not reveal evidence of phenolphthalein. The results suggesting that water had been added to the stool were discussed with the patient, but she denied adding water to the samples. A stool specimen obtained during supervised defecation was semisolid, and its osmolality was 279 mOsm per kilogram.
|
Patient 2
Patient 2 was a 40-year-old woman who had watery diarrhea, cramping abdominal pain, and fever after her third pregnancy. These symptoms persisted for two years and remained unexplained despite extensive evaluation. The results of stool tests for pathogens, endoscopy and colonoscopy with biopsies, computed tomography of the abdomen and pelvis, small-bowel radiography, gallium scanning, measurement of urinary 5-hydroxyindoleacetic acid excretion, and testing for the human immunodeficiency virus were all normal. Urinary excretion of d-xylose was low. The patient reported increased symptoms, including up to 15 liquid stools per day, when she ate. Trials of sulfasalazine, metronidazole, cholestyramine, and an elemental diet had no benefit. The patient was treated with long-term total parenteral nutrition, but her diarrhea persisted. Home treatment with total parenteral nutrition was complicated by septicemia.
The patient was hospitalized for further evaluation of her diarrhea. While she was eating 80 g of fat daily, her 48-hour stool volume was 5250 ml. Stool osmolality was 19 mOsm per kilogram of water, and the stool sodium and potassium concentrations were <10 and 1.9 mmol per liter, respectively (Table 1). The results of colonoscopy and upper gastrointestinal endoscopy were normal, as were those of small-bowel and colonic biopsies. Fluid aspirated from the colon during colonoscopy had an osmolality of 227 mOsm per kilogram, and the sodium and potassium concentrations were 56 and 33 mmol per liter, respectively. Psychiatric evaluation revealed no evidence of a specific psychiatric disorder. The findings were discussed with the patient, who denied adding water to the stool collections.
The patient returned to her primary internist, to whom she made no subsequent reports of diarrhea.
Discussion
Chronic unexplained diarrhea can be a perplexing problem for both patients and their physicians. The approach to such patients, which has been well described elsewhere, includes evaluation of stools for laxatives5,6,7 and measurements of stool electrolyte concentrations and osmolality. Electrolytes and osmolality are usually measured not to diagnose factitious disease but to distinguish osmotic from secretory diarrhea by calculating the osmotic gap (the difference between stool osmolality and twice the sum of the stool sodium and potassium concentrations)8. In patients with diarrhea due to the intestinal secretion of fluid and electrolytes, the osmotic gap should be less than 50. A larger gap may indicate the presence of an unabsorbed osmotic agent in the stool9,10. The use of some laxatives may be suspected when a large osmotic gap is found, and some of these substances (such as magnesium) can be measured directly in stool.
Why does low stool osmolality suggest that water has been added to the stool? The osmolality of fresh liquid stool is roughly equivalent to that of serum, although the osmolality of stored or transported stool specimens is often higher than serum osmolality (because of the ongoing bacterial metabolism of stool carbohydrate)9,11. Fecal osmolality and serum osmolality remain equivalent in patients who have abused laxatives and in those with various forms of naturally occurring and experimentally induced secretory and osmotic diarrhea9,10,11,12,13,14. The intestine cannot secrete free water, and as a consequence, stool should not be hyposmolar. A low stool osmolality can be explained only by the addition of water (or another dilute fluid) to stool. This possibility has been mentioned in textbooks6,7 and one well-documented case report15. We confirmed the presence of a factitious disorder, rather than a novel pathophysiologic process, by measuring the osmolality of specimens obtained at the time of colonoscopy or supervised defecation in our two patients.
A stool osmolality of 290 to 300 mOsm per kilogram of water is usually assumed in calculating the stool osmotic gap7,10. This approach avoids artifactual widening of the osmotic gap caused by increases in measured stool osmolality that occur during transportation and handling of the specimen. If factitious diarrhea is suspected, stool osmolality should be measured directly to exclude the possibility that the stool has been diluted. This is especially important if a large osmotic gap is calculated with an assumed stool osmolality value.
Our two patients had many clinical characteristics in common, and their condition is a variant of Munchausen's syndrome16. Both had undergone extensive evaluation for chronic watery diarrhea and had received potentially harmful therapies, and both insisted that they had not tampered with their stool collections. One of the patients ultimately told her primary physician that she had diluted her stools, and she later entered psychotherapy. Neither has required further treatment for diarrhea.
In patients with low stool osmolality, simultaneous measurements of urinary and stool osmolality should be performed to exclude the possibility that the stool has been diluted by hyposmolar urine. In addition, the hospital personnel collecting and analyzing the stool specimens should avoid adding water. The containers used for stool collection must be dry and should not be rinsed during the collection process. Patient 2 had a low stool osmolality of 227 mOsm per kilogram in an endoscopically obtained specimen (as compared with a value of 19 mOsm per kilogram in an ordinary stool collection). The former value may reflect the presence of water in the suction channel of the endoscope, or the patient may have given herself tap-water enemas before the colonoscopic procedure.
A search for self-induced or factitious disease is appropriate when the cause of chronic diarrhea remains elusive. Laxative use or abuse can be detected by measuring stool electrolytes, magnesium, and phenolphthalein and by chromatographic analysis of stool and urine specimens2,17. Factitious diarrhea may not be discovered, however, unless stool osmolality is also measured.
Supported in part by a General Clinical Research Center grant (RR00125) from the National Center of Research Resources.
We are indebted to Drs. Zvi Fischer and Kenneth Mauer for referring the two patients and providing follow-up information.
Source Information
From the Department of Internal Medicine, Section of Digestive Diseases, and the General Clinical Research Center, Yale University School of Medicine.
Address reprint requests to Dr. Topazian at Yale University School of Medicine, 1080 LMP, 333 Cedar St., New Haven, CT 06510.
References
| |||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
