RegulatoryPeptides,36 (1991) 141-150

141

© 1991ElsevierSciencePublishersB.V.All rightsreserved0167-0115/91/$03.50 REGPEP 01093

Clinical study

Cholecystokinin, vasoactive intestinal peptide and peptide histidine methionine responses to feeding in anorexia nervosa R i c h a r d F. H a r t y 1, Paul H. P e a r s o n l, Travis E. S o l o m o n 2 and J a m e s E. M c G u i g a n 3 1Departments of Internal Medicine and Pediatrics, University of Nebraska Medical Center and Swanson Centerfor Nutrition, Omaha, NE (U.S.A.), 2Department of Medicine, University of Kansas, and VA Medical Center, Kansas City, MO (U.S.A.) and 3Department of Medicine, University of Florida College of Medicine, Gainesville, FL (U.S.A.)

(Received 12 April 1991;revised version received 1 July 1991; accepted 2 July 1991) K e y words: Anorexia nervosa; Eating disorder; Cholecystokinin; Vasoactive

intestinal peptide; Peptide histidine methionine

Summary Anorexia nervosa (AN) is a syndrome of unknown cause characterized by voluntary starvation. Cholecystokinin has been implicated as a neuroendocrine regulatory factor in control of satiety. Relatively little information is known about gastrointestinal hormone responses to feeding in subjects with anorexia nervosa. In the present studies, we examine fasting and postprandial levels of cholecystokinin (CCK), vasoactive intestinal peptide (VIP) and peptide histidine methionine (PHM) in anorexia nervosa subjects and in control individuals. Results of these studies indicate that plasma CCK response to a liquid meal (Ensure Plus) in untreated AN subjects was distinctly different from that observed in healthy controls, both in terms of temporal pattern of peptide released and the amount of CCK secreted into the circulation. Peak levels of CCK release occurred at 30 min following meal ingestion in AN patients and at 60 min in control subjects. Integrated CCK release in untreated AN patients was approximately twice that measured in control individuals. Renutrition therapy was associated with reversion of the pattern of CCK release to that observed in control subjects. Plasma VIP levels were unchanged following meal ingestion in both control and anorexic subjects. In contrast, P H M levels in AN subjects were significantly greater than that observed in control

Correspondence: R.F. Harry, Research Service (151), Department of Veterans Affairs Medical Center, 4101 Woolworth Avenue, Omaha, NE 68105, U.S.A.

142 individuals. The pattern of PHM release following liquid meal ingestion was similar to that observed with plasma CCK; namely, peak release of peptide was observed at 30 min which was significantly greater than corresponding control values (P < 0.05). In conclusion, these results demonstrate distinctive differences in plasma CCK and PHM levels in response to feeding in AN subjects when compared to control individuals. These findings suggest that the earlier and greater rise in plasma CCK levels in AN subjects following meal ingestion may contribute to the abnormal sensation of satiety in this condition.

Introduction

Anorexia nervosa is a syndrome of unknown cause characterized by voluntary self-starvation. The extreme weight loss often associated with this condition is perpetuated by disturbance in the person's perception of their body image. Patient descriptions of their body image frequently include references to being 'too fat', even in states of severe emaciation. Consequently, individuals with anorexia nervosa exhibit fear of weight gain which may lead to obesity. Subjective sensations of extreme fullness and early satiety reinforce fears of becoming obese and account for the mistrust often encountered by individuals entering re-feeding programs. The concept that gastrointestinal hormones or regulatory peptides may contribute to postprandial satiety was proposed in 1973 by Gibbs and coworkers [1,2]. This hypothesis was based on their observations of inhibition of food intake by exogenous administration of the brain: gut regulatory peptide cholecystokinin (CCK). Experimental studies have shown that CCK, when given by an intraperitoneal or intravascular route, inhibits food intake in a large number of animal models as well as in human subjects [3]. Central (CNS) administration of CCK has also been shown to induce satiety [4-7]. Furthermore, impairment of postprandial cholecystokinin release has been reported in subjects with bulimia nervosa [8]. Relatively little information is known, however, about gastrointestinal hormone responses to feeding in subjects with anorexia nervosa. The aims of the present studies were to examine basal and postprandial plasma levels of cholecystokinin and vasoactive intestinal peptide (VIP)- and peptide histidine methionine (PHM)-related peptides in subjects with anorexia nervosa and in control individuals. Furthermore, the effects of renutrition therapy on circulating levels of CCK and VIP-related peptides were examined.

Materials and Methods

Six female patients meeting DSM-III [9] criteria for anorexia nervosa (AN) gave informed consent to participate in this study. These individuals were admitted to the Eating Disorders Center at the University of Nebraska Medical Center during a 6-month period beginning in August 1986 and extending through January 1987. Four

143 female AN subjects successfully completed the study protocol. Criteria for diagnosis of anorexia nervosa included intense fear of becoming fat despite weight loss and malnutrition, disturbance of body image, weight loss of at least 25 percent of original body weight, refusal to maintain body weight over a minimal normal weight for age, height and sex and no known physical illness which would explain anorexia and weight loss. History and physical examination demonstrated no evidence of gastroduodenal disease or prior gastrointestinal surgery. None of the subjects was taking medication known to alter gastric emptying. Control subjects were female employees at the medical center and were of the same age range as subjects with anorexia nervosa. As with eating disorder subjects, control individuals gave no history of gastrointestinal disease, prior gastrointestinal surgery or taking medication known to effect gastric motility. Anorexia nervosa subjects ranged in age from 16 to 26 years; body weight expressed as percent of ideal body weight for the four study subjects was 69, 68, 70 and 77 percent. The control group (n = 3) ranged in age from 23 to 28 and body weight as percent of ideal body weight was between 95 and 110 percent. The study protocol was approved by the institutional review board at the University of Nebraska Medical Center. Fasting and meal-stimulated regulatory peptide levels in AN subjects were obtained within 48 h of entry into the hospital and, subsequently, at 2, 4 and 6 weeks after initiation of therapy and re-nutrition program. Following an overnight fast, AN and control subjects were allowed to consume an 8 oz. liquid test meal in the form of Ensure Plus (13.0 g protein, 47.3 g carbohydrate and 12.6 g fat). Prior to meal consumption, fasting blood samples were obtained and after ingestion of the meal, blood was drawn at 30, 60 and 90 min. Blood samples were placed on ice and subsequently centrifuged (500g for 10 min at 4 °C) and sera removed, aliquoted and frozen at - 2 0 °C until performance of radioimmunoassay. Serum cholecystokinin levels were determined by radioimmunoassay performed in the laboratory of Dr. Travis E. Solomon according to previously described methods [10]. Vasoactive intestinal peptide (VIP) and peptide histidine methionine (PHM) were measured in the laboratory of Dr. James E. McGuigan employing methods previously reported [11,12]. Data presented represent the mean 1 S.E.M. Univariate analysis of variance was used to examine and compare effects of fasting, meal ingestion and treatment duration on regulatory peptide release between the study subjects and controls. Statistical significance was assigned a value of P < 0.05.

Results

Fig. 1 depicts fasting and postprandial plasma CCK levels in individual control (A) subjects and AN (B) patients at time of entry into treatment. Fasting CCK levels in AN and control subjects were similar: 5.3 + 1.1 and 3.7 + 0.6 pM, respectively (P < 0.1). Peak postcibal CCK response in each control subject was observed at 60 min. In contrast, each AN patient demonstrated maximal CCK release in response to feeding at 30 min. CCK values at 30 min for AN patients were significantly greater than those for control subjects: 13.6 + 3.6 vs. 5.9 + 1.0 pM, P < 0.0001. These data were analyzed further in Fig. 2 where results of postprandial CCK release

144 24

A.

CONTROL

2O

~o. 16v (J

c~

12-

,s

8o.

4-

F'ASTING

3'0

6'0

9'0

TIME (minutes)

24

B.

ANOREXIANERVOSA- ENTRY

2O

16 U

u

12

" vO

~

u ~

1so-

.. ~

10o-

~

AN

5o-

o="

c

~'o

6'0

9'0

TIME (minutes) Fig. 2. Postprandial CCK release in response to liquid test meal (Ensure Plus). Data are expressed as the mean + S.E.M. and have been normalized to percent above fasting values. * Denotes significant difference (P < 0.05) from either corresponding control or AN value at 30 and 60 min following meal ingestion.

145

are expressed as percent above corresponding fasting values. At 30 min following meal, CCK values in AN individuals increased significantly to maximal levels which were 160~ above fasting (P < 0.05). This value was more than 1.5-times greater than control CCK levels at 30min (62.6 + 3.5~o; P < 0.05). 1 h after test meal, serum CCK in control subjects achieved peak levels of 142.3 + 2.1 percent above fasting determinations. This value was significantly greater than corresponding results for AN patients (36.0 + 18.5 percent at 60 rain; P < 0.05). However, peak levels of CCK release in control subjects at 60 rain were not significantly different from measured CCK release observed for AN patients at 30 min: 142.3 + 21.0 vs. 160.3 + 32.2 percent, respectively; P < 0.1. Fig. 3 shows that integrated CCK release for AN patients during the 60 min following meal ingestion was approximately twice that observed for control subjects (283.1 + 84.7 vs. 143.0 + 19.5 pM, 60 min). However, this difference was not statistically significant ( P > 0.l). 2 weeks after initiation o f treatment, fasting and meal-stimulated p l a s m a C C K levels were determined a n d c o m p a r e d to values from control subjects (Fig. 4). A t this testing

interval, postprandial increases in plasma CCK in AN patients were not different from control values. Furthermore, peak CCK release in both AN patients and control subjects was observed at 60 min. Fasting and postprandial CCK values in AN subjects at 4 and 6 weeks after entry into the renutrition program were not different from control values. Vasoactive intestinal peptide measurements in AN patients at entry and in control individuals following liquid meal did not change appreciably from fasting values (Fig. 5). In contrast, fasting and postprandial plasma P H M determinations in AN patients were substantially different from values observed in control subjects (Fig. 6) with serum PHM levels in AN patients significantly greater than those in control at 30 min (P < 0.05). Unlike CCK, fasting and postprandial values in AN subjects did not revert to control levels at subsequent testing intervals (2, 4 and 6 weeks). P)0.1 375

i

~ 300t ,,~ "~

22.5-

U 0

'" ~ _z

"

150-

75,\\\

0 CONTROL

\

ANOREXIA NERVOSA

Fig. 3. Integrated plasma CCK release during the initial hour following liquid meal. Although AN subjects released 98 percent more CCK than control subjects during this postprandial period, the values were not significantly different.

146 250-

cm 200-

~';

15o-

~o

,oo

T

/

5o/. O~

o

~

3'0

6'0

9'0

TIME (minutes) Fig. 4. Plasma CCK release in anorexia nervosa ( A N ) subjects 2 weeks after inidation o f renutrition

program. When compared to levels of CCK released in control (C) subjects, no significant differences were observed.

3025.

v

20.

>