Reduced Gastrin Releasing Peptide in Cerebrospinal Fluid After Recovery from Bulimia Nervosa
Total Page:16
File Type:pdf, Size:1020Kb
Appetite (2001) 37, 9±14 doi:10.1006/appe.2001.0407, available online at http://www.idealibrary.com on 1 Original Article Reduced gastrin releasing peptide in cerebrospinal fluid after recovery from bulimia nervosa G.K. Franka,W.H. Kaye a, E.E. Ladenheimb and C. McConahaa aSchoolof Medicine, Department of Psychiatry,Western Psychiatric Institute and Clinic,University of Pittsburgh; bSchoolof Medicine, Department of Psychiatryand Behavioral Sciences,The Johns Hopkins University,Baltimore (Received 20 September 2000, revision11April 2001,accepted in revised form 20 April 2001) People with anorexia (AN) and bulimia nervosa (BN) have altered patterns of eating. It is possible that alterations of the neuropeptide gastrin releasing peptide (GRP), a bombesin (BBS) -like peptide with potent central anorexigenic activity, could contribute to disturbed eating behavior. To avoid the confounding effects of pathologic eating behavior, we measured cerebrospinal fluid (CSF) GRP concentrations in women who were long-term recovered ( >1 year, normal weight, and regular menstrual cycles, no bingeing or purging) from AN (REC AN, N 12) or BN (REC BN, N 21) compared to healthy control women (NC, N 15). CSF GRP was significantly lower (2 9Á41(3), p < 0Á01) in REC BN (9Á6 Æ 3Á1 pg/ml) compared to NC (13Á4 Æ 5Á5 pg/ml) and REC AN (11Á6 Æ 2Á9 pg/ ml). Persistent GRP abnormalities after recovery from BN raise the possibility that this alteration might be trait- related and contribute to episodic hyperphagia in BN. # 2001 Academic Press Introduction of food restriction alternating with regular binge-purge episodes. The etiology of EDs is not known (American The eating disorders (EDs) anorexia (AN) and bulimia Psychiatric Association, 1994). Recent studies suggest nervosa (BN) are characterized by disturbed eating that EDs are familial and that a genetic predisposition behaviors and comorbid disturbances of mood, impulse might contribute to their etiology (Lilenfeld et al., 1998; control, and temperament. Women with AN show a Strober et al., 2000). pattern of food restriction and emaciation; however, A number of neuropeptides contribute to the modu- some AN are pure food restrictors, whereas others also lation of appetite regulation (Kalra et al., 1999). It is show binge-purging type behaviors. Women with BN are theoretically possible that a disturbance of the modu- usually at a normal body weight, and engage in a pattern lation of one or more neuropeptide systems plays a role in disturbed eating behavior in humans. One neuro- peptide system of interest is human gastrin releasing Supported in part by grants from NIMH #2 R01 MH peptide (GRP). GRP is a 27 amino acid peptide that 42984-09 ``The Neurobiology of Feeding Behavior in shares a similar decapeptide with bombesin (BBS; Bulimia''; Children's Hospital Clinical Research Center, Pittsburgh, PA (#5M01 RR00084), and NIMH #2 R01 Brown et al., 1980), and belongs to the family of MH 46001-07A1 ``Serotonin: A Trait Disturbance in ``bombesin-like peptides.'' BBS was first isolated from Anorexia Nervosa''. amphibian skin (Anastasi et al., 1971) and showed Address correspondence to: Walter H. Kaye, M.D., anorexigenic effects when administered peripherally or University of Pittsburgh, Western Psychiatric Institute and centrally in mammals (Gibbs and Smith, 1988). GRP is Clinic, Room E-724, 3811 O'Hara Street, Pittsburgh, PA 15213, U. S. A. Fax: 1 412 624 6618. not as potent as BBS, but otherwise has similar Email: [email protected] effects (Morley et al., 1985). Peripheral and central 0195±6663/01/040009+06 $35.00/0 # 2001 Academic Press 10 G. K. Frank et al. administration of GRP attenuates food intake in menstrual cycles; have not binged, purged (i.e. not mammals and humans (Bray 1995; Flynn, 1994). Per- engaged in self-induced vomiting, the use of laxatives ipherally, BBS-like receptors are distributed throughout or diuretics), excessively exercised, or engaged in the gastrointestinal (GI) tract, with receptors that have restrictive eating patterns; and have not met criteria specifically high affinity for GRP (von Schrenck et al., for substance-use related disorders. Subjects were free 1990). They take part in the regulation of GI hormone of medication for 6 weeks prior to the study and gave and enzyme secretion, cell growth, smooth muscle informed consent. The control women (NC) had no activity and blood glucose levels (Dietrich, 1994; Walsh history of an eating disorder or any psychiatric, et al., 1981). In the central nervous system (CNS), dis- medical, or neurologic illness. They had normal tinct BBS-like receptor subtypes have been identified in menstrual cycles, and had been within a normal weight brain tissue such as the bed nucleus of the stria termi- range since menarche. They also had no first degree nalis, the olfactory tubercle, the putamen and neocortex, relatives with an eating disorder. Methods and with a neuromedin B and a GRP preferring subtype psychological assessment data are described in detail (Ladenheim et al., 1992; Wolf & Moody, 1985). Both elsewhere (Kaye et al., 1998a, 1999). For many of the subtypes have been implicated in the modulation of subjects, CSF 5-hydroxyindole acetic acid had been BBS-like peptide induced food suppression (Ladenheim measured previously. et al., 1996). In addition, central GRP also modulates A lumbar puncture (LP) for obtaining CSF samples temperature regulation and grooming behavior in ani- was performed during the first 10 days of the follicular mals (Cowan, 1988). Disturbances of neuropeptides phase of the menstrual cycle. Subjects were admitted to a could be a consequence of disturbed eating behavior or research laboratory on the Eating Disorders Unit of malnutrition, or pre-morbid traits that contribute to a Western Psychiatric Institute and Clinic at 9 a.m. on the vulnerability to develop an ED. Determining whether day prior to the LP for adaptation to the laboratory and abnormalities are a consequence or a potential ante- for psychological assessments. Subjects fasted over- cedent of pathological feeding behavior is a major night, and the LP was performed the next day between 9 question in the study of eating disorders. It is impractical a.m. and 9:30 a.m. LP samples were frozen in liquid to study people with ED prospectively due to the young nitrogen and stored at 80C. For the GRP assay a age of onset and difficulty in pre-morbid identification commercial kit for radio immuno assay (RAI) from of people who will develop an ED. Therefore, we deci- Peninsula Laboratories, Inc., San Carlos, CA, was used. ded to study women who had recovered for more than a The antibody in that kit shows 100% cross reactivity year from AN or BN. Any persistent psychobiological with human GRP, as well as with BBS. No cross reac- abnormalities might be trait-related and potentially tivity is reported for human gastrin I, human big gastrin- have contributed to the pathogenesis of this disorder. In 1 peptide, gastric inhibitory peptide, or substance P. The this study we measured cerebrospinal fluid (CSF) values variability of the assay between runs was 4 Æ 3% of human GRP, a human BBS-like peptide. We hypo- (mean Æ SD). Plasma -hydroxybutyrate (BHBA), a thesized that AN would have increased and BN ketone body, was assessed as an index of starvation. decreased CSF values of GRP. BHBA was measured by the enzymatic method of Williamson et al. (1962). Statistical analyses were performed using the SPSS software package (SPSS, Chicago, IL; Barcikowski Materials and methods 1984). Data exploration assessed whether data were normally distributed (Shapiro-Wilk test). When the Thirty-three women who had previously met DSM-IV assumption of normality was met, One-Way ANOVA criteria for an ED were recruited: 21 women recovered assessed whether group means were similar, and from normal weight BN (REC BN), and 12 subjects Scheffe's post hoc tests compared individual group dif- recovered from binge eating/purging (eight) or restrict- ferences. All variables aside from age and CSF GRP ing type (four) AN (REC AN). The REC BN women were normally distributed. When the assumption of had never had AN. Recovered ED subjects were normality was not met, Kruskal-Wallis test assessed compared to 15 female normal controls (NC), who three group comparisons, and Mann-Whitney U tests were recruited through advertisements in local news- assessed individual group differences. Correlations were papers and who were without lifetime psychiatric or investigated using Pearson Correlations for normally major medical illnesses. In the year before the study, all distributed and Spearman Correlation Coefficients for participating recovered subjects had to maintain a non-normally distributed variables. All data are body weight of > 90% average body weight (ABW; expressed as mean Æ standard deviation (SD). Group Metropolitan Life Insurance, 1959); have regular differences were considered significant if p < 0Á05. Gastrin releasing peptide and bulimia nervosa 11 Results compared with REC AN. Lifetime high % ABW was higher in REC BN compared with NC and REC AN. Age was similar between groups (Table 1). However, Lifetime low % ABW was lower in REC AN % ABW at the time of the study was higher in REC compared with NC and REC BN. BHBA was similar BN compared with NC and REC AN, and in NC between groups. There was no difference between length of recovery between groups. One CSF GRP value for the NC group was excluded 35 as a non-true outlier (195 pg/ml). There was a significant difference between group CSF GRP levels (Table 1). Figure 1 shows a scatterplot with the distribution of the GRP values across groups. The CSF GRP in REC BN 30 showed some overlap in distribution with NC and AN. However, the majority of CSF GRP data points in REC BN women was below 10 pmol/ml (15/21), whereas the majority of data points for NC (12/15) and REC AN (9/ 25 12) was above 10 pmol/ml.