Irritable bowel syndrome (IBS) is frequently regarded

GASTROENTEROLOGY 1988;95:709-14 Symptoms of Psychologic Distress Associated With Irritable Bowel Syndrome Comparison of Community and Medical Clinic Samples WILLIAM E. WHITEHEAD, LINDA BOSMAJIAN, ALAN B. ZONDERMAN, PAUL T. COSTA, Jr., and MARVIN M. SCHUSTER Departments of Psychiatry and Medicine, Johns Hopkins University School of Medicine, ana Gerontology Research Center, National Institute on Aging, Baltimore. Maryland Women with symptoms indicative of irritable bowel syndrome who had not consulted a physician were compared with female patients at a gastroenterology clinic to investigate whether self-selection for treatment accounts for psychologic abnormalities in clinic patients with irritable bowel syndrome. Two sets of diagnostic criteria were compared: restrictive criteria based on the work of Manning and conventional criteria (abdominal pain plus altered bowel habits). Lactose malabsorbers were included as a control group because they have medically explained bowel symptoms similar to those that define irritable bowel syndrome. Thus they control for the causative effects of chronic bowel symptoms on psychologic distress. Women who met restrictive criteria for irritable bowel syndrome but had not consulted a physician had no more symptoms of psychologic distress on the Hopkins Symptom Checklist than asymptomatic controls. However, medical clinic patients with both irritable bowel syndrome and lactose malabsorption had significantly more psychologic symptoms than asymptomatic controls or nonconsulters with the same diagnoses. Individuals who met only the conventional criteria for irritable bowel syndrome reported more psychologic distress than controls, whether or not they consulted a physician. These results suggest that (a) symptoms of psychologic distress are unrelated to irritable bowel syndrome but influence which patients consult a doctor and (b) conventional diagnostic criteria identify more psychologically distressed individuals than do restrictive criteria. Irritable bowel syndrome (IBS) is frequently regarded as a psychophysiologic disorder because 72%-100% of patients in published series are found to have diagnosable psychiatric disorders (1,2), and studies using standardized psychometric tests consistently demonstrate abnormal test profiles for groups of patients with!bs attending medical clinics (3---,9). However, neither a specific psychiatric diagnosis nor a specific psychometric test profile is found. If one pools studies that have used diagnoses made by psychiatrists based on research diagnostic criteria, the leading diagnoses are hysteria (20%), depression (20%), and anxiety disorder (14%) (10). Similarly, studies using the Hopkins Symptom Checklist find elevations on several subscales: somatization of affect, interpersonal sensitivity, depression, anxiety, and hostility (7,8). These findings do not suggest a specific psychologic mechanism for the symptoms of IBS. They are compatible with three alternative hypotheses: 1. Self-selection hypothesis. Symptoms of psychologic distress may be unrelated to development of the bowel symptoms that define IBS but may influence who among those with symptoms of bowel dysfunction are most likely to visit a medical clinic. It is known, for example, that patients Abbreviations used in this paper: FBD, functional bowel disorder; IBS, irritable bowel syndrome; LMA, lactose malabsorption; MMPI, Minnesota Multiphasic Personality Inventory. 1988 by the American Gastroenterological Association 0016-5085/88/$3.50

710 WHITEHEAD ET AL. GASTROENTEROLOGY Vol. 95, No.3 with essential hypertension who are attending medical clinics have more symptoms of psychologic distress than individuals newly discovered to have hypertension in the community (11,12), and it is inferred that this occurs because psychologically distressed individuals are more likely to seek medical assistance. This hypothesis is tested by comparing clinic patients with IBS and patients with lactose malabsorption (LMA, inability to digest milk products) with individuals in the community who have similar bowel symptoms but have not consulted a physician. The selfselection hypothesis would be supported by showing that clinic attenders have more symptoms of psychologic distress than individuals with the same bowel symptoms who are not consulting a physician. 2. Inappropriate diagnostic criteria. The use of vague, diffuse symptoms as diagnostic criteria for IBS could result in the identification of a heterogeneous population, thus accounting for the nonspecific psychologic findings. Common clinical practice is to diagnose IBS on the basis of abdominal pain plus either constipation or diarrhea in the absence of physical findings adequate to explain these symptoms (7). However, these criteria have been criticized as much too general (10,13). Whitehead and Schuster (10) proposed more restrictive research diagnostic criteria for IBS based on the series of studies by Thompson et al. (13-15) and Drossman et al. (16) and on the findings of a workshop on diagnostic criteria for IBS sponsored by the National Institutes of Health [Iohannes R, personal communication). They (10) suggest that the term functional bowel disorder (FBD) be used to refer to patients who complain of vague abdominal pain plus constipation or diarrhea but who do not satisfy the more restrictive criteria for IBS. The present study tests the hypothesis that symptoms of psychologic distress will be more closely associated with the vague clinical definition (FBD) than with more restrictive diagnostic criteria. 3. Somatopsychic hypothesis. If symptoms of psychologic distress are associated with IBS, they may be consequences of having chronic bowel symptoms rather than causes of bowel symptoms. This hypothesis can be tested by comparing patients with IBS to patients with bowel symptoms that are indistinguishable from IBS (17) but that are due to a different physiologic mechanism, namely deficiencies of the small intestinal enzyme lactase (18). This hypothesis would be supported by showing that both lactose malabsorbers and individuals with IBS have more symptoms of psychologic distress than normals. Materials and Methods Subjects A community sample was recruited by asking the directors of church women's societies and charities to provide subjects in exchange for a contribution to their organization's treasury for each woman completing the study. They were told that the study would involve questionnaires about bowel symptoms and personality traits and a3-h test of milk intolerance. They were also told that participants need not have any bowel symptoms in order to participate. A total of 149 socioeconomically middle class women aged 18-89 yr (average age 46.7 yr) participated. This included 3 black subjects. This method of recruiting subjects resulted in an overrepresentation of women who felt that they had bowel symptoms: 22% of participants had a positive hydrogen breath test suggestive of LMA and 26% reported symptoms suggestive of IBS (criteria given below) within the last 6 mo. This compared with population estimates of 8% for LMA (19) and up to 17% for IBS (16). Only 20% of these women had been to a doctor for bowel symptoms during the last 6 mo. Thus, the recruitment procedure succeeded in identifying a sample of adult women who were demographically similar to our medical clinic population and many of whom had bowel symptoms for which they had not sought medical care. All patients referred to our gastroenterology service for symptoms suggestive of IBS are evaluated for lactose intolerance. The medical clinic sample was drawn from 209 consecutive patients tested for lactose intolerance by the hydrogen breath method (20). Patients with other disorders that might account for their bowel symptoms were identified by history and physical examination, including proctoscopy or endoscopy when indicated, and were excluded. Men, women below age 18 yr, and patients with diabetes and spinal cord injury were also excluded, leaving 121 female patients of whom 54 had complete data for the Hopkins Symptom Checklist (21) and a bowel symptom questionnaire. The average age of these 54 women was 40.9 yr (not significantly different from the community sample), and 3 were black. Socioeconomic status was predominantly middle class. Procedures Women in the community sample were tested in small groups at their organization's normal meeting place. On the first visit, a bowel symptom questionnaire was administered, which contained 29 upper and lower gastrointestinal symptoms, including the six symptoms found by Manning et al. (13) to distinguish patients with IBS from patients with other gastrointestinal disorders. Patients marked the symptoms that they had experienced during the previous 6 mo and also indicated whether they had consulted a doctor for any of these symptoms. Community subjects were given a set of personality tests to complete at home. These included the Hopkins Symptom Checklist (21), a brief research version of the Neuroticism-Extroversion-Openness Inventory (22), the Cornell Medical Index, and the Minnesota Multiphasic Personality

September 1988 PSYCHOLOGIC SYMPTOMS IN IBS 711 Inventory (MMPI). The MMPI was scored for nine content dimensions derived from a full-item factor analysis of the MMPI (23) as well as the standard K-corrected clinical scales. At the second meeting with community subjects, a hydrogen breath test for lactose intolerance was administered (20). Subjects reported to the test site after an overnight fast and were given 50 g of lactose suspended in 250 ml of water to drink. End-expiratory breath samples were collected before the lactose meal and at t-h intervals for 3 h after the meal. A positive hydrogen breath test indicative of lactose malabsorption was defined as an increase in hydrogen of at least 22 parts per million (ppm) above the baseline determination at any time during the 3-h test. Clinic patients participated during two visits to the medical clinic. The bowel symptom questionnaire and the Hopkins Symptom Checklist were completed before the patient's initial medical consultation. The hydrogen breath test was performed on a subsequent visit. This study was approved by the Institutional Review Board on March 19, 1984. Subjects in both samples were classified on the basis of the bowel symptom questionnaire and the breath hydrogen test into the following four groups: 1. Lactose malabsorption, defined as all subjects who showed at least a 22-ppm increase over baseline in breath hydrogen at any point within 3 h after ingesting 50 g of lactose suspended in 250 ml of water. 2. Irritable bowel syndrome, defined by self-reports of relief of abdominal pain after a bowel movement plus at least two of the following five symptoms in the presence of a negative hydrogen breath test: (a) loose stools at the onset of pain, (b) more frequent bowel movements with the onset of pain, (c) distention of abdomen, (d) mucus passed by rectum, (e) frequent feeling of incomplete emptying (10). 3. Functional bowel disorder, defined by (a) self-reports of abdominal pain plus constipation or diarrhea, or both, but (b) failure to satisfy inclusion criteria for IBS above and (c) negative hydrogen breath test (10). 4. Normals, defined as all other subjects in the community sample. In the community sample, IBS and FBD were diagnosed solely on the basis of self-reported bowel symptoms; alternative medical diagnoses that might account for these symptoms, except for LMA, were not investigated. However, 3 subjects were excluded because they volunteered that pregnancy, food poisoning, or appendicitis might have caused their bowel symptoms. Data Analyses The first analysis compared the community sample with the clinic sample. Normal subjects and people in the community sample who reported having consulted a physician for bowel symptoms were excluded. A two-way multivariate analysis of variance was employed in which the nine subs cales of the Hopkins Symptom Checklist were the dependent variables with clinic versus commu- nity sample and diagnostic group (LMA, IBS, or FBD) as grouping factors. The second set of analyses involved only subjects in the community sample who had not consulted a doctor for the treatment of bowel symptoms in the last 6 mo. Separate multivariate analyses of variance were performed for each psychometric test. Each analysis treated scores on the scales of the test as dependent variables and diagnosis as a between-subjects factor. Results Self-Selection Hypothesis Table 1 shows that medical clinic patients had more psychologic symptoms than nonconsulters from the community sample. The multivariate analysis of variance showed the main effect for medical clinic patients vs. nonconsulters to be statistically significant [F(1,109) = 19.78, P < 0.01]' and when two-way analyses of variance were run on each of the Hopkins Symptom Checklist scales, the two samples differed significantly on all nine standard scales and on the global symptom index (Table 1). This comparison is shown for IBS patients in Figure 1. Diagnostic group differences were also evaluated in this multivariate analysis of variance. The main effect for diagnosis was significant [F(2,109) = 3.23, P < 0.05]' but the interaction between sample and diagnosis was not significant. These diagnostic group differences were due to greater numbers of symptoms in the FBD group as compared with the IBS and LMA groups (Table 1). Inappropriate Diagnostic Criteria Hypothesis Separate analyses of all psychometric inventories agreed in showing that subjects in the community sample who satisfied restrictive criteria for IBS reported no more symptoms of psychologic distress than asymptomatic controls or lactose malabsorbers, whereas subjects in the FBD group tended to report more psychologic distress than normals (Table 1 and Figure 2). On the Hopkins Symptom Checklist the main effect for. diagnosis approached statistical significance [F(3,112) = 2.66, P = 0.05], and t-tests showed the FBD group to be higher than the normals on the subscales for somatization, anxiety, hostility, and phobia. Similar results were seen on the Neuroticism-Extroversion-Openness Inventory, the MMPI, and the Cornell Medical Index. These analyses demonstrate that the criteria used to diagnose IBS influence the conclusions one draws regarding the association of IBS with symptoms of psychologic distress. Further post hoc analyses revealed that the psychologic differences between subjects who satis-

712 WHITEHEAD ET AL. GASTROENTEROLOGY Vol. 95, No.3 Table 1. Medical Clinic Patients Compared With Community Nonconsulters on Hopkins Symptom Checklist" Nonconsulters Medical clinic Statistically significant Normal FBD IBS LMA FBD IBS LMA effects" Somatization 0.43 (0.37) 0.73 (0.60) 0.34 (0.29) 0.55 (0.33) 1.27 (0.88) 1.34 (0.75) 1.22 (0.67) S Obsessive-compulsive 0.67 (0.53) 0.92 (0.71) 0.55 (0.55) 0.73 (0.51) 1.08 (0.94) 1.32 (1.07) 1.00 (0.73) S Interpersonalsensitivity 0.57 (0.49) 0.73 (0.66) 0.60 (0.64) 0.49 (0.35) 1.13 (0.99) 0.93 (0.85) 0.66 (0.57) S Depression 0.64 (0.49) 0.94 (0.91) 0.56 (0.63) 0.59 (0.50) 1.85 (1.88) 1.15 (0.76) 1.21 (0.89) SID Anxiety 0.36 (0.33) 0.67 (0.81) 0.46 (0.80) 0.38 (0.31) 1.03 (0.84) 0.95 (1.22) 0.87 (0.64) S Hostility 0.31 (0.35) 0.71 (0.84) 0.30 (0.36) 0.36 (0.27) 1.25 (0.86) 1.05 (0.89) 0.65 (0.67) SID Phobia 0.07 (0.14) 0.23 (0.39) 0.09 (0.17) 0.13 (0.19) 0.69 (0.94) 0.51 (0.87) 0.38 (0.42) S Paranoia 0.36 (0.43) 0.50 (0.39) 0.25 (0.44) 0.32 (0.36) 0.83 (0.79) 0.83 (0.98) 0.52 (0.56) S Psychoticism 0.25 (0.36) 0.35 (0.54) 0.17 (0.26) 0.13 (0.17) 0.78 (0.82) 0.32 (0.45) 0.27 (0.32) SID Global symptoms index 0.45 (0.32) 0.69 (0.60) 0.41 (0.44) 0.46 (0.30) 1.08 (0.76) 1.01 (0.73) 0.87 (0.48) S No. of subjects 46 26 16 28 12 10 23 FBD, functional bowel disorder; IBS, irritable bowel syndrome; LMA, lactose malabsorption. a Group means with standard deviations in parentheses. b S, Significant main effect for medical clinic sample vs. community nonconsulters. D, Significant main effect for diagnosis: FBD vs. IBS vs. LMA. fied restrictive diagnostic criteria for IBS and subjects who satisfied the looser criteria for FBD resulted from the robust correlation of abdominal pain (not otherwise specified) with measures of psychologic distress. When all individuals in the community sample who complained of abdominal pain (regardless of other symptoms) were compared with those who did not, the patients with abdominal pain scored significantly higher (p < 0.05) on the neuroticism scale of the Neuroticism-Extraversion-Openness Inventory (34.24 vs. 31.04), the neuroticism factor extracted from the MMPI (23.68 vs. 19.41), and the Global Symptom Index of the Hopkins Symptom Checklist (0.60 vs. 0.41). However, when people with abdominal pain indicated that their pain was relieved by defecation and was associated with alterations in stool consistency and frequency [i.e., when they met criteria for IBS), they were no more likely than pain-free individuals to exhibit eleva- II! ~ \5 1.25 ~ ~.75 l!l ei.5 :::.25.~., "... ~.A:., r-. ~. ~ Figure 1. Patients with IBS (~) from a medical clinic are compared with nonpatients with symptoms of IBS (0) and with asymptomatic controls (e). Ordinate represents raw scores on the Hopkins Symptom Checklist. SOM, somatization; OBS, obsessive-compulsive; SEN, interpersonal sensitivity; DEP, depression; ANX, anxiety; HOS, hostility; PHO, phobia; PAR, paranoia; PSY, psychoticism. tions on these psychologic symptom scales (Table 1 and Figure 2). Somatopsychic Hypothesis As shown in Table 1, lactose malabsorbers in the community sample who had a lifelong history of bowel symptoms similar to IBS reported no more psychologic symptoms than asymptomatic normals. Thus, the somatopsychic hypothesis was not supported. Discussion When women were identified from the community who had symptoms suggestive of IBS for which they had not sought medical care, and when restrictive diagnostic criteria were used to define IBS, the individuals so identified showed no more symptoms of psychologic distress than asymptomatic controls or individuals with LMA. These findings suggest that symptoms of psychologic distress are unrelated to occurrence of the bowel symptoms II! 0 l;l \5 1.25.A..". ~.' ~.75 l!l /...... ~ a e!.5.-' '\ " '"... :::....25 Figure 2. Nonpatients with symptoms of FBD (~) are compared with nonpatients with IBS (0) and with asymptomatic controls (e) on the Hopkins Symptom Checklist. Abbreviations are the same as for Figure 1.,6.,

September 1988 PSYCHOLOGIC SYMPTOMS IN IBS 713 used to define IBS, and to the extent that our restrictive diagnostic criteria identify individuals with a motility disorder, the findings suggest that symptoms of psychologic distress are unrelated to this motility disorder. The association between bowel symptoms and psychologic symptoms that is often seen in studies of medical clinic patients appears to be due to psychologically distressed patients selecting themselves for inclusion by going to physicians for treatment of bowel symptoms that other people ignore or treat themselves. Particularly compelling evidence for the self-selection hypothesis is the fact that medical clinic patients with LMA reported more symptoms of psychologic distress than individuals with LMA seen in the community, because a physical mechanism for LMA has been identified, and no psychologic mechanism for LMA has been proposed. The criteria used here to define FBD are accepted by many clinicians as adequate to define IBS and have been used in many research studies. However, these vague bowel symptoms appear to reflect primarily a psychologic tendency to complain of bodily sensations. This suggests that the more restrictive criteria that relate abdominal pain to alterations in bowel habits should be used to diagnose IBS. These findings are consistent with the report of Drossman and his colleagues (24), who compared patients with IBS identified through their medical clinic with nonpatients with similar bowel symptoms and with controls identified through a survey of medical center employees. Patients with IBS from the medical clinic showed significantly mote abnormal MMPI test profiles and more illness behaviors than nonpatients with IBS. The latter group were similar to controls. Our study differed from that of Drossman et al. by investigating an older sample (average age of 47 yr for our community sample compared with 26 yr in the study by Drossman et al.), by distinguishing between LMA and IBS, and by investigating the consequences of using different diagnostic criteria for IBS. The similar findings in these two studies increase confidence in the conclusions. These results differ from those of investigators in New Zealand (25), who found no psychologic differences between clinic patients with IBS and nonpatients with similar bowel symptoms. Both groups were psychologically similar to a control group except for higher scores on the somatization scale of the Hopkins Symptom Checklist. Cultural differences between New Zealand and the United States in the way psychologic symptoms influence the decision to seek medical care may explain these different results. The severity of bowel symptoms, especially pain, also contributes to the decision to consult a physician about bowel symptoms (26). As we were unable to compare these two samples with respect to the severity of bowel symptoms, it is possible that such differences existed and that they contributed both to the decision to consult a physician and to psychologic differences between consulters and nonconsulters. The majority of patients with IBS report that psychologic stress causes exacerbations of their bowel symptoms (16,27,28). These observations are not in conflict with our data as psychologic stress may elicit gastrointestinal physiologic arousal in psychologically healthy individuals as well as in those with abnormal personality traits or symptoms of psychologic distress (16,27). These findings suggest that symptoms of psychologic distress are unrelated to the bowel symptoms that define IBS, but they do influence who will come to the medical clinic for treatment. The implication of these findings for the management of patients with IBS is that these patients should be thought of as having two distinct types of symptoms-bowel symptoms possibly due to a motility disorder and psychiatric symptoms. These two types of symptoms may require separate treatments. These findings also underscore the high prevalence of symptoms of psychologic distress in patients presenting with bowel complaints and suggest that routine psychologic screening by a questionnaire such as the Hopkins Symptom Checklist or the Neuroticism Extroversion-Openness Inventory is desirable. References 1. Liss JL, Alpers D, Woodruff RA Jr. The irritable colon syndrome and psychiatric illness. Dis Nerv Syst 1973;34:151-7. 2. Young SJ, Alpers DH, Norland CC, Woodruff RA Jr. Psychiatric illness and the irritable bowel syndrome: practical implications for the primary physician. Gastroenterology 1976;70: 162-6. 3. Esler MD, Goulston KJ. Levels of anxiety in colonic disorders. N Engl J Med 1973;288:16-20. 4. Palmer RL, Stonehill E, Crisp AH, Waller SL, Misiewica JJ. Psychological characteristics of patients with the irritable bowel syndrome. Postgrad Med J 1974;50:416-9. 5. Lancaster-Smith MJ, Prout BJ, Pinto T, Anderson JA, Schiff AA. Influence of drug treatment on the irritable bowel syndrome and its interaction with psychoneurotic morbidity. Acta Psychiatr Scand 1982;66:33-44. 6. Latimer P, Sarna S, Campbell D, Latimer M, Waterfall W, Daniel EE. Colonic motor and myoelectrical activity: a comparative study of normal subjects, psychoneurotic patients, and patients with irritable bowel syndrome. Gastroenterology 1981;80:893-901. 7. Whitehead WE, Engel BT, Schuster MM. Irritable bowel syndrome: physiological and psychological differences between diarrhea-predominant and constipation-predominant patients. Dig Dis Sci 1980;25:404-13. 8. Wise TM, Cooper IN, Ahmed S. The efficacy of group therapy

714 WHITEHEAD ET AL. GASTROENTEROLOGY Vol. 95, No.3 for patients with irritable bowel syndrome. Psychosomatics 1982;23:465-9. 9. West KL. MMPI correlates of ulcerative colitis. J Clin Psychol 1970;26:214-29. 10. Whitehead WE, Schuster MM. Gastrointestinal disorders: behavioral and physiological basis for treatment. New York: Academic, 1985. 11. Davies M. Blood pressure and personality. J Psychosom Res 1970;14:89-104. 12. Cochrane R. Hostility and neuroticism among unselected essential hypertensives. J Psychosom Res 1973;17:215-8. 13. Manning AP, Thompson WG, Heaton KW, Morris AF. Towards positive diagnosis of the irritable bowel. Br Med J 1978;2:653-4. 14. Thompson WG, Heaton KW. Functional bowel disorders in apparently healthy people. Gastroenterology 1980;79:283-8. 15. Thompson WG. Gastrointestinal symptoms in the irritable bowel compared with peptic ulcer and inflammatory bowel disease. Gut 1984;25:1089-92. 16. Drossman DA, Sandler RS, McKee DC, Lovitz A. Bowel patterns among subjects not seeking health care. Gastroenterology 1982;82:529-34. 17. Enck P, Mellibruda L, Wright E, Whitehead WE, Tucker H, Schuster MM. Manning criteria fail to distinguish IBS from lactose intolerance (abstr). Gastroenterology 1984;86:1070. 18. Paige DM, Bayless TM, eds. Lactose digestion: clinical and nutritional implications. Baltimore: Johns Hopkins University Press, 1981. 19. Johnson JD. Regional and ethnic distribution of lactose malabsorption: adaption and genetic hypothesis. In: Paige DM, Bayless TM, eds. Lactose digestion: clinical and nutritional implications. Baltimore: Johns Hopkins University Press, 1981:11-22. 20. Rosado JL, Solomons NW. Sensitivity and specificity of the breath analysis test for detecting malabsorption of physiological doses of lactose. Clin Chern 1983;19:545-8. 21. Derogatis LI. The SCL-90-R. Administration, scoring, and procedures manual II. Towson, Md.: Clinical Psychometric Research, 1983. 22. Costa PT Ir, McCrae RR. The NEO personality inventory manual. Odessa, Fla.: Psychological Assessment Resources, 1985. 23. Costa PT [r, Zonderman AB, McCrae RR, Williams RB Jr. Content and comprehensiveness in the MMPI: an item factor analysis in a normal adult sample. Pers Soc PsychoI1985;48: 925-33. 24. Drossman DA, McKee DC, Sandler RS, et al. Psychosocial factors in irritable bowel syndrome: a multivariate study (abstr). Gastroenterology 1987;92:1374. 25. Welch GW, Hillman LC, Pomare EW. Psychoneurotic symptomatology in the irritable bowel syndrome: a study of reporters and non-reporters. Br Med J 1985;291:1382-4. 26. Sandler RS, Drossman DA, Nathan HP, McKee DC. Symptom complaints and health care seeking behavior in subjects with bowel dysfunction. Gastroenterology 1984;87:314-8. 27. Almy TP, Tulin M. Alterations in colonic function in man under stress. I. Experimental production of changes simulating the "irritable colon." Gastroenterology 1947;8:616-26. 28. Chaudhary NA, Truelove SC. The irritable colon syndrome: a study of the clinical features, predisposing causes, and prognosis in 130 cases. Q J Med 1962;3:307-23. Received August 27, 1987. Accepted April 1, 1988. Address requests for reprints to: William E. Whitehead, Ph.D., Division of Digestive Diseases, Francis Scott Key Medical Center, 4940 Eastern Avenue, Baltimore, Maryland 21224. This work was supported by grant DK31369 from the National Institute for Digestive Diseases and Kidney Disease, by Career Development Award MH00133 from the National Institute of Mental Health, and by the Gerontology Research Center of the National Institute on Aging.