rectal manometry measuring anal sphincter function, rectal sensation, expulsion dynamics, and rectal compli- ance. Patients were studied both early (
'r" ~'" '"' c'" '"'·"' "
THE AMERICAN JOURNAL OF GASTROl:J'\TFROI OCiY
NOTICE: This material may be protected By copyright law (Title 17 U.S. Code)
Copyright© 1997 by Am. Coll. of Gastrocntcrology
Vol. 92, No. 10, 1997 Printed in U.S.A.
Colonic and Anorectal Function in Constipated Patients with Anorexia Nervosa Andrew B. Chun, M.D., Mae S. Sokol, M.D., Walter H. Kaye, M.D., William R. Hutson, M.D., and Arnold Wald, M.D. Department ol Medicine, Division ol Gastroenterology and Hepatology, and Department of Psychiatry, Center for Overcoming Problem Eating, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania
1
INTRODUCTION
Objectives: Many patients with eating disorders complain of severe constipation. Previous studies have suggested that constipation in patients with anorexia nervosa may be :issociated with slow colonic transit. However, it is unclear whether a refeeding program will alter colonic transit in these patients. The aim of this study was to investigate colorectal function by measuring colonic transit and anorectal function in anorexic patients with constipation during treatment with a refeeding program. Methods: We prospectively studied 13 female patients with anorexia nervosa who were admitted to an inpatient treatment unit and compared them to 20 previously studied, age-matched, healthy female control subjects. Patients underwent colonic transit studies using a radiopaque marker technique and anorectal manometry measuring anal sphincter function, rectal sensation, expulsion dynamics, and rectal compliance. Patients were studied both early ( 3 wk) in their admission. We restudied two patients who had slow colonic transit. All patients also underwent structured interviews. Results: Four of six patients studied within the first 3 wk of their admission had slow colonic transit, defined as >70 h (108.0 ± 17.0 h, mean± SEM), on initial evaluation. In contrast, none of the seven patients studied later than 3 wk into their admission had slow colonic transit. Two of the four patients with slow transit were restudied later in their admission and were found to have normal transit times. Rectal sensation, internal anal sphincter relaxation threshold, rectal compliance, sphincter pressures, and expulsion pattern were normal in all subjects. Conclusions: Despite complaints of severe constipation, colonic transit is normal or returns to normal in the majority of patients with anorexia nervosa once they are consuming a balanced weight gain or weight maintenance diet for at least 3 wk.
Anorexia nervosa is a disease that has significant morbidity and mortality. More than 5 million Americans are thought to be affected, including 2-5% of high school and college age women (1). Constipation is a common complaint in these patients and has been reported by up to 60% of patients with eating disorders as one of their most debilitating symptoms (2, 3). This may be associated with deranged colorectal physiology or may represent the distorted perceptions of body function that are associated with the illness (2). Making this distinction provides insight into the nature of constipation in these patients and leads to potential strategies for treating them. Anorectal manometry and radiopaque marker studies to evaluate anorectal function and colonic transit times are objective methods for evaluating constipation (4). Colonic transit studies identify patients who have normal or slow colonic passage and have identified three patterns of slow transit: colonic inertia (prolonged transit through the entire colon), hindgut dysfunction (normal transit through the right colon but prolonged transit through the left), and outlet dysfunction (delayed transit through the anorectum) (5, 6). Anorectal manometry identifies several disorders of anorectal sensory and motor function, including increased threshold of rectal sensation, megarectum and rectoanal dyssynergia, and inappropriate expulsion dynamics (4, 6). Any one of these factors could be a cause of constipation. The combination of colonic transit studies and anorectal manometry provides a comprehensive means of evaluating patients with chronic constipation. Earlier studies used radiopaque markers and lactulose breath tests in patients with anorexia and bulimia to demonstrate an overall delay in GI transit upon admission to the hospital (7). It was speculated that delayed transit could contribute to the eating disorder by causing bloating, which could increase a patient's fear of fatness, or by causing rectal distension, which could reflexively inhibit gastric emptying. The implication was that early and aggressive treatment of constipation in these patients might be beneficial. However, other studies have noted that GI symptoms
Received Feb. 6, 1997; accepted June JO, 1997. 1879
1880
AJG - Vol. 92, No. 10, 1997
CHUN et al.
TABLE 1 Characteristics of Patients with Anorexia Nervosa Compared with Control Subjects
Patients
N Age (yr) % IBW BMs* per week Laxative use Onset of anorexia (yr)
13 27 74.4 1.4 II 16.9
(19-40) (59-81) (0.3- 3)
Control subjects 20 31
(21 - 52)
NIA 7.5 (3-21) 0
(I 0- 25)
NIA
Values are expressed as the mean (range) where applicable. = bowel movements.
* BMs
DAV 4
(2) and delayed gastric emptying (8) improve significantly with a renutrition program. Specific aims The aim of this project was to investigate the colorectal physiological characteristics of constipation in patients with eating disorders at various times during a refeeding program. We compared patients with anorexia nervosa to subjects in an age- and sex-matched control group.
MATERIALS AND METHODS Study populations Subjects for this study were an inpatient population of women, ages 18-40 yr, who met the American Psychiatric Association's criteria for anorexia nervosa (9) and who complained of chronic constipation. Chronic constipation was defined as less than two stools per week for at least 6 months and/or the inability to defecate without the regular use (at least once per week) of laxatives, enemas, or suppositories. Thirteen female patients with anorexia nervosa were compared to 20 female control subjects who were matched for age (Table 1). Our decision to study a female population reflects the fact that eating disorders affect primarily women (> 95%) (1). All anorexic patients were admitted to the inpatient unit of the Center for Overcoming Problem Eating at the University of Pittsburgh Medical Center, which specializes in the treatment of eating disorders. Patients were closely monitored to minimize the misrepresentation of eating habits or stool frequency and the surreptitious use of laxatives or purging. In addition, patients were required to be in compliance with a strictly enforced program of eating all of their meals and snacks in a renutrition program. Exclusion criteria included a history of bowel obstruction, active GI infection, previous GI surgery or trauma, pregnancy, and use of medications or the presence of systemic disorders that are possible causes of constipation. All participants gave informed consent to participate in the study, which was approved by the Radiation Safety Committee and Institutional Review Board of the University of Pittsburgh Medical Center.
DAV 7
F10. 1. Colonic transit study. Abdominal film s taken on days 4 and 7 of the study are divided into three sections corresponding to the right, left, and rectosigmoid colon. The sum of the markers were multiplied by 1.2 to calculate regional and total colonic transit.
Procedures Colonic transit studies. These were performed while the patients were consuming a stable weight gain or weight maintenance high residue diet containing at least 25 g of fiber or equivalent daily. Patients did not use laxatives, prokinetic agents, or enemas during the study. Patients and nursing staff maintained diaries to record food consumption and the passage of stools. On day 1, patients ingested 20 ring-shaped markers in the morning. At approximately the same time on the following 2 days, they ingested 20 additional markers. Roentgenograms of the abdomen were obtained on days 4 and 7 of the study (Fig. 1). The total number of x-rays did not exceed two per transit time evaluation and did not exceed four per patient. To minimize exposure, a high penetration technique was used (100-110 kV); whole body exposure per film was 175 mR, and exposure to reproductive organs was 70 mR per film. Segmental and total colonic transit times were ~alculated using a previously described formula (5, 10). Patients were classified as having normal or slow colonic transit. Anorectal manometry. Resting anal sphincter pressure was measured by station pull-through (at 0.5-cm intervals) of five perfused catheters (individual inner diameter, 0.8 mm; total outer diameter, 4.8 mm; sidehole diameter, 0.8 mm) whose openings were at 0.5-cm intervals and oriented at 60° to each other. They were perfused with water at a rate of 0.5 ml/min by a pneumohydraulic pump. Pressures were measured by strain gauge transducers connected to a recorder. Values are expressed as the mean of the five recordings for each patient. Rectal sensation, internal anal sphincter relaxation, and expulsion dynamics were measured as described previously (4, 11). Approximately 2 h after a cleansing phosphate enema, a rectal balloon, 5.5 cm in length and 4 cm in diameter when inflated with 50 ml of air, was positioned 5 cm above the anal verge to obtain the threshold of rectal sensation (defined as the smallest volume of distention sensed by the patient repeatedly) and the threshold of inter-
1
r
AJG - October /997
COLONIC AND ANORECT AL FUNCTION IN ANOREXIA NERVOSA
nal sphincter relaxation (the smallest volume of distention that produced relaxation of the internal anal sphincter). These measurements were determined by rapidly inflating and deflating the rectal balloon with volumes starting at 65 ml and continuing in decrements of 5 ml at 30-s intervals. Expulsion dynamics were obtained as previously reported (11) and were defined as the manometric profile obtained when the patient was asked to simulate defecation, i.e., to attempt to expel the apparatus. It was defined as abnormal if increased pressures on the external balloon were seen on at least two of three determinations. Rectal compliance was measured using a previously described technique ( 12) that consisted of progressive inflation of a cylindrical balloon (I 0 cm in length placed in the rectum 5 cm from the anal verge) with air in 50-ml increments to a maximum of 250 ml. Pressures within the balloon were measured by an external transducer and were corrected by subtracting pressures when inflated with the same volumes while in' ambient air. Anorectal sensory and motor parameters in patients with eating disorders were compared to those previously reported for control subjects from our laboratory ( I 3).
Data analysis
RESULTS
100 80
§,
Colonic transit times were significantly slower in patients with anorexia nervosa tested within 3 wk of their admission to an inpatient treatment program compared with patients tested more than 3 wk after admission (86.6 ::±:: 17.8 h vs 34.1 ::±:: 4.1 h, mean ::±:: SEM, p < 0.03). There was no significant difference between the anorexic patients tested more than 3 wk after admission and control subjects (28.0 ::±:: 8.6 h, mean ::±:: SEM) (Fig. 2). Four of six anorexic patients who were studied within 3 wk of admission had slow colonic transit, whereas none of the seven patients studied 3 wk or more after admission had slow transit. Two of the four patients with slow transit were retested approximately 6 wk after admission and had normal transit times (Fig. 3). Of the remaining two patients with slow transit, one refused retesting, and the other was discharged from the unit shortly after her initial evaluation because of insurance problems. We found no correlation between colonic transit times and percentage of ideal body weight (IBW) (Fig. 4).
Anorectal manometry studies There were no significant differences in any of the anorectal manometry parameters measured between anorexic patients and control subjects. Furthermore, all patients and
60
raAN3 weeks Ill Controls
*
"' =
.... "' r-
