Association of the c.385C>A (p.Pro129Thr) polymorphism of the fatty acid amide hydrolase gene with anorexia nervosa in the Japanese population.

The functional c.385C>A single-nucleotide polymorphism (SNP) in the fatty acid amide hydrolase (FAAH) gene, one of the major degrading enzymes of endocannabinoids, is reportedly associated with anorexia nervosa (AN). We genotyped the c.385C>A SNP (rs324420) in 762 lifetime AN and 605 control participants in Japan. There were significant differences in the genotype and allele frequencies of c.385C>A between the AN and control groups. The minor 385A allele was less frequent in the AN participants than in the controls (allele-wise, odds ratio = 0.799, 95% confidence interval [CI] 0.653-0.976, P = 0.028). When the cases were subdivided into lifetime restricting subtype AN and AN with a history of binge eating or purging, only the restricting AN group exhibited a significant association (allele-wise, odds ratio = 0.717, 95% CI 0.557-0.922, P = 0.0094). Our results suggest that having the minor 385A allele of the FAAH gene may be protective against AN, especially restricting AN. This finding supports the possible role of the endocannabinoid system in susceptibility to AN.

No association of brain-derived neurotrophic factor Val66Met polymorphism with anorexia nervosa in Japanese.

The Met66 allele of the Val66Met polymorphism in the brain-derived neurotrophic factor (BDNF) gene has been reported to be associated with anorexia nervosa (AN), and also lower minimum body mass index (BMI) and higher harm avoidance in AN. We genotyped the Val66Met polymorphism (rs6265) in 689 AN cases and 573 control subjects. There were no significant differences in the genotype or allele frequencies of the Val66Met between AN and control subjects (allele wise, odds ratio = 0.920, 95% CI 0.785-1.079, P = 0.305). No difference was found in minimum BMIs related to Val66Met in AN (one-way ANOVA, P > 0.05). Harm avoidance scores on the Temperament and Character Inventory were lower in the Met66 allele carriers (P = 0.0074) contrary to the previous report. Thus we were unable to replicate the previous findings that the Met66 allele of the BDNF is associated with AN and that the minimum BMI is lower or the harm avoidance score is higher in AN patients with the Met66 allele.

A ghrelin gene variant may predict crossover rate from restricting-type anorexia nervosa to other phenotypes of eating disorders: a retrospective survival analysis.

BACKGROUND: Patients with anorexia nervosa restricting type (AN-R) often develop bulimic symptoms and crossover to AN-binge eating/purging type (AN-BP), or to bulimia nervosa (BN). We have reported earlier that genetic variants of an orexigenic peptide ghrelin are associated with BN. Here, the relationship between a ghrelin gene variant and the rate of change from AN-R to other phenotypes of eating disorders (EDs) was investigated. METHODS: Participants were 165 patients with ED, initially diagnosed as AN-R. The dates of their AN-R onset and changes in diagnosis to other subtypes of ED were investigated retrospectively. Ghrelin gene 3056 T-->C SNP (single nucleotide polymorphism) was genotyped. Probability and hazard ratios were analyzed using life table analysis and Cox's proportional hazard regression model, in which the starting point was the time of AN-R onset and the outcome events were the time of (i) onset of binge eating, that is, when patients changed to binge eating AN and BN and (ii) recovery of normal weight, that is, when patients changed to BN or remission. RESULTS: Patients with the TT genotype at 3056 T-->C had a higher probability and hazard ratio for recovery of normal weight. The ghrelin SNP was not related with the onset of binge eating. CONCLUSION: The 3056 T-->C SNP of the ghrelin gene is related to the probability and the rate of recovery of normal body weight from restricting-type AN.

Psychological and weight-related characteristics of patients with anorexia nervosa-restricting type who later develop bulimia nervosa.

BACKGROUND: Patients with anorexia nervosa-restricting type (AN-R) sometimes develop accompanying bulimic symptoms or the full syndrome of bulimia nervosa (BN). If clinicians could predict who might change into the bulimic sub-type or BN, preventative steps could be taken. Therefore, we investigated anthropometric and psychological factors possibly associated with such changes. METHOD: All participants were from a study by the Japanese Genetic Research Group for Eating Disorders. Of 80 patients initially diagnosed with AN-R, 22 changed to the AN-Binge Eating/Purging Type (AN-BP) and 14 to BN for some period of time. The remaining 44 patients remained AN-R only from the onset to the investigation period. Variables compared by ANOVA included anthropometric measures, personality traits such as Multiple Perfectionism Scale scores and Temperament and Character Inventory scores, and Beck Depression Inventory-II scores. RESULTS: In comparison with AN-R only patients, those who developed BN had significantly higher current BMI (p < 0.05) and maximum BMI in the past (p < 0.05). They also scored significantly higher for the psychological characteristic of parental criticism (p < 0.05) and lower in self-directedness (p < 0.05), which confirms previous reports, but these differences disappeared when the depression score was used as a co-variant. No significant differences were obtained for personality traits or depression among the AN-R only patients irrespective of their duration of illness. CONCLUSION: The present findings suggest a tendency toward obesity among patients who cross over from AN-R to BN. Low self-directedness and high parental criticism may be associated with the development of BN by patients with AN-R, although the differences may also be associated with depression.

Obestatin, acyl ghrelin, and des-acyl ghrelin responses to an oral glucose tolerance test in the restricting type of anorexia nervosa.

BACKGROUND: Obestatin is a recently identified peptide encoded by the same ghrelin gene. It has been reported that obestatin has anorexigenic and antigastroprokinetic activities as opposed to ghrelin. We investigated simultaneously obestatin, acyl ghrelin, and des-acyl ghrelin in the restricting type of anorexia nervosa (AN-R) patients. METHODS: Three hormonal responses to the oral glucose tolerance test (OGTT) were measured in 10 AN-R patients and 10 healthy women. RESULTS: Plasma obestatin, acyl ghrelin, and des-acyl ghrelin levels were significantly higher in AN-R patients than in control subjects throughout the OGTT. All of the three hormones decreased after the OGTT in both groups. CONCLUSIONS: We found that AN-R patients exhibited increased plasma levels of obestatin, acyl ghrelin, and des-acyl ghrelin throughout the OGTT compared with control subjects. The hormonal differences between groups are statistically most significant in obestatin, suggesting obestatin may serve as a marker reflecting both acute and chronic changes of the nutritional state in AN-R patients.

A role for pancreatic polypeptide in feeding and body weight regulation.

PP administration induces negative energy balance by suppressing food intake and gastric emptying while increasing energy expenditure in rodents. The mechanism of PP actions involves the changes in the expression of hypothalamic feeding-regulatory peptides and the activity of the vago-vagal and vago-sympathetic reflex arc. PP-overexpressing mice we developed exhibited the thin phenotype with decreased food intake and gastric emptying rate. Plasma cholecystokinin (CCK) concentrations were increased in the transgenic mice and CCK-1 receptor antagonist improved the anorexia of the animals. These results, together with the previous notion of PP as an anti-CCK hormone in pancreatic exocrine secretion and gallbladder contraction, indicate that PP-CCK interactions may be either antagonistic or synergistic and the transgenic mice may exhibit the mixed phenotype by overproduction of PP and CCK.

Incomplete restoration of the secretion of ghrelin and PYY compared to insulin after food ingestion following weight gain in anorexia nervosa.

BACKGROUND: In humans, ghrelin has been found to stimulate appetite while PYY3-36 to reduce it; these orexigenic and anorexigenic peptides play significant roles in appetite control. We investigated pre- and postprandial responses of ghrelin and PYY in anorexia nervosa (AN) and the influence of weight gain. METHODS: Plasma ghrelin, PYY3-36, glucose and insulin responses after ingestion of a 400 kcal standard meal were measured in 14 patients with restricting type of AN and 12 controls. The AN patients were evaluated before therapy and after inpatient therapy. Psychometry was performed by the use of Eating Disorders Inventory. RESULTS: Ghrelin was suppressed during the meal test, while PYY3-36 was increased in all of the groups. Before therapy, AN patients had significantly increased levels of ghrelin and PYY3-36 compared to the control (P<0.01). After therapeutic intervention, as the nutritional status of AN patients improved, the secretion of these hormones were increased (P<0.05), but not normalized as in psychological testing. In contrast, insulin and glucose responses were normalized after inpatient therapy. CONCLUSIONS: We found that both ghrelin and PYY3-36 increased in AN patients and these changes were not normalized in contrast to insulin after treatment. The increase in both orexigenic ghrelin and anorexigenic PYY3-36 may have a role in pathological eating behavior in AN.

Possible role of preproghrelin gene polymorphisms in susceptibility to bulimia nervosa.

Previous investigations have suggested that ghrelin, an endogenous orexigenic peptide, is involved in the pathology of eating disorders. We conducted a study to determine whether any preproghrelin gene polymorphisms are associated with eating disorders. Three hundred thirty-six eating disorder patients, including 131 anorexia nervosa (AN)-restricting types (AN-R), 97 AN-binge eating/purging types (AN-BP) and 108 bulimia nervosa (BN)-purging types (BN-P), and 300 healthy control subjects participated in the study. Genotyping was performed to determine the polymorphisms present, and with this information, linkage disequilibrium (LD) between the markers was analyzed and the distributions of the genotypes, the allele frequencies, and the haplotype frequencies were compared between the groups. The Leu72Met (408 C > A) (rs696217) polymorphism in exon 2 and the 3056 T > C (rs2075356) polymorphism in intron 2 were in LD (D' = 0.902, r2 = 0.454). Both polymorphisms were significantly associated with BN-P (allele-wise: P = 0.0410, odds ratio (OR) = 1.48; P = 0.0035, OR = 1.63, for Leu72Met and 3056 T > C, respectively). In addition, we observed a significant increase in the frequency of the haplotype Met72-3056C in BN-P patients (P = 0.0059, OR = 1.71). Our findings suggest that the Leu72Met (408 C > A) and the 3056 T > C polymorphisms of the preproghrelin gene are associated with susceptibility to BN-P.

Relationship between pretreatment laboratory-measured episodes of reactive hypoglycemia and short-term weight restoration in anorexia nervosa: a preliminary study.

Refeeding outcome is difficult to predict in anorexia nervosa (AN). Because reactive hypoglycemia (RH) during an oral glucose tolerance test (OGTT) correlates with rapid increases of energy intake just before the OGTT in AN patients, this study investigated whether pretreatment laboratory-measured RH episodes might be associated with refeeding progress in this disorder. Forty-six female patients with AN (25 restrictors and 21 binge/purgers) and 11 controls underwent an OGTT before treatment. The patients were divided into groups according to the presence of RH. Thereafter, AN patients underwent nutritional rehabilitation, and weight gain and daily energy intake were evaluated. In both AN subtypes, the RH groups showed more daily energy intake and gained more weight compared with the non-RH groups. The present study found a close relationship between pretreatment laboratory-measured RH episodes and refeeding progress, suggesting that pretreatment laboratory-measured RH episodes may be an important predictor of short-term refeeding outcome in AN.

The effects of bone therapy on tibial bone loss in young women with anorexia nervosa.

OBJECTIVE: Osteoporosis is recognized as a common medical complication of anorexia nervosa (AN). The purpose of the current study was to investigate the recovery mechanism of osteoporosis in AN and the effect of medical treatment on the skeletal system. METHOD: We conducted a randomized placebo-controlled study of the effects of etidronate and calcium and vitamin D on bone loss in 41 outpatients with the restricting type of AN (AN-R). We measured the tibial speed of sound (SOS) before and after 3 months of treatment. RESULTS: The bone mineral density (BMD) of the tibial SOS change in both the etidronate group and the calcium and vitamin D Group was significantly greater (p < .001) than in the control group. Urine-N-telopeptide cross-links of type I collagen (NTx) before and after treatment decreased significantly (p < .01) in the etidronate group. CONCLUSION: These findings suggest that both etidronate and calcium and vitamin D are equally efficacious for reversing the degree of osteoporosis in patients with AN.

Comparison of regional cerebral blood flow in patients with anorexia nervosa before and after weight gain.

We investigated changes in regional cerebral blood flow (rCBF) before and after weight gain in patients with restrictive anorexia nervosa (AN-R) in comparison with findings in normal subjects. We assessed resting rCBF using single photon emission computed tomography with technetium-99m hexamethylpropylene amine oxime in 12 AN-R patients and 11 controls. Each patient was examined at two time points, at the beginning of treatment and after weight gain (average examination interval=88+/-26 days). Control subjects were examined only once. Before treatment, the AN-R group had lower rCBF in the bilateral anterior lobes, including the anterior cingulate cortex (ACC), and in the right parietal lobe, the insula, and the occipital lobes. After weight gain, the patients showed significant increases in the right parietal lobe and decreases in the basal ganglia and cerebellum in accordance with significant improvement in body weight and eating attitudes. However, they showed persistent decreases in the ACC area even after weight gain compared with findings in the controls. A significant positive correlation was observed between body mass index and rCBF in the occipital lobes in the patients. These results suggest that weight gain is associated with a normalization of rCBF in a number of brain areas, but that the low level of rCBF in the ACC at baseline is unaffected by treatment in AN-R.

"End-stage kidney" in longstanding bulimia nervosa.

OBJECTIVE: The extent of renal damage over long-term binge/purges has not been well documented in bulimia nervosa (BN). METHOD: We describe a 52-year-old woman with longstanding BN subsequent to an 8-year history of anorexia nervosa (AN). RESULTS: The patient showed chaotic binge/purges and chronic severe hypokalemia after recovery from AN at age 26 years, and renal biopsy showed juxtaglomerular hyperplasia, which was diagnosed as pseudo-Bartter's syndrome. DISCUSSION: Over the following 26 years, the patient's eating behaviors remained chaotic, and her renal function gradually deteriorated. After the patient died of pneumonia and sepsis at age 52 years, autopsy of her kidney showed chronic interstitial nephritis, proximal tubular swelling, and diffuse glomerular sclerosis, suggesting chronic glomerular injury associated with long-term binge/purges. To our knowledge, this is the first case report of a patient with BN with long-term binge/purges who developed an eventual "end-stage kidney" characterized by hypokalemic nephropathy and diffuse glomerulosclerosis.

Glucose tolerance predicts short-term refeeding outcome in females with anorexia nervosa.

OBJECTIVE: Little is known about biologic predictors of refeeding outcome in anorexia nervosa (AN). Because nutritional status mirrors glucose metabolism during an oral glucose tolerance test (OGTT) in AN, this study investigated whether pretreatment glucose response patterns during the OGTT might be associated with refeeding progress in patients with AN. METHODS: Sixty-four female patients with anorexia (33 restrictors and 31 binge/purgers) and 13 healthy control subjects underwent an OGTT before nutritional rehabilitation, including desensitization to fear of energy intake of 1000 to 1600 kcal/day. Patients were divided into flat-type responders, impaired glucose tolerance (IGT)-type responders, and normal-type glucose responders. Daily energy intake, weekly weight gain, and the duration of desensitization period were evaluated until the 12th week. RESULTS: The patients with anorexia consisted of 20 flat-type, 21 IGT-type, and 23 normal- type responders. Normal-type responders required a shorter time to complete the desensitization period than other responders (p = .003 for restrictors, p < .001 for binge/purgers). In terms of refeeding progress, significant group effects for daily energy intake and weekly weight gain were evident in restrictors (p = .006, p = .028, respectively) and binge/purgers (p < .001, p = .003, respectively); normal-type responders showed good refeeding progress compared with other responders in both AN subtypes. CONCLUSIONS: The present study found a close relationship between pretreatment glucose responses, therapeutic progress of desensitization to fear of energy intake, and refeeding progress in both AN subtypes. Our findings suggest that glucose tolerance may be a useful predictor of short-term refeeding outcome in this disorder.

Psychologic management of brace therapy for patients with idiopathic scoliosis.

STUDY DESIGN: A trial of brace therapy modified by a measured personality pattern of patients with idiopathic scoliosis was performed. OBJECTIVE: To evaluate the effectiveness of performing personality tests for patients with idiopathic scoliosis who undergo brace therapy. SUMMARY OF BACKGROUND DATA: Brace therapy has often been used for the treatment of scoliosis. However, emotional distress can result from this therapy. Few attempts have been made to reduce such stress. METHODS: A test using the Maudsley Personality Inventory was performed on 145 adolescent females with idiopathic scoliosis, treated with brace therapy alone, before the start of brace therapy and 1 month after the start of brace therapy. On the basis of test results, the patients were rated as normal type and four abnormal types. Brace therapy was continued considering the personality pattern of patients. For all patients, changes in psychologic test results, compliance with braces wearing instructions, and correction of scoliosis were analyzed. RESULT: Of the 134 patients rated as normal before the start of therapy, 108 patients were rated as abnormal pattern when tested 1 month after the start of therapy. After performing autogenic training for patients with E-N+ and E-N- personalities, and giving advice to school teachers to decrease the emotional stress for patients with E+N+ personality, 47 patients were finally rated as abnormal pattern. In total, 12 (8%) of the 145 patients dropped out. In dropouts, the average pretreatment deformity of 29 degrees (range: 21 degrees -37 degrees ) had increased to an average of 37 degrees (range, 31 degrees -48 degrees ). CONCLUSION: Psychologic tests may be useful and provide a means of modifying brace therapy tailored to the psychologic conditions of individual patients.

Altered ghrelin and peptide YY responses to meals in bulimia nervosa.

OBJECTIVE: In recent years great advances have been made in our understanding of the peripheral signals produced within the gastrointestinal tract that regulate appetite, such as ghrelin and peptide YY (PYY). While ghrelin elicites hunger signals, PYY elicites satiety. Therefore, alterations in hormone physiology may play a role in the pathogenesis of bulimia nervosa (BN). In this study, we investigated the postprandial profile of ghrelin and PYY levels in patients with BN. DESIGN AND PATIENTS: Postprandial plasma ghrelin and PYY levels and insulin and glucose responses were measured in 10 patients with BN and 12 control patients in response to a standard 400 kcal meal. RESULTS: Basal ghrelin levels present in BN subjects (265.0 +/- 25.5 pmol/l) were significantly higher than those in healthy controls (199.3 +/- 18.4 pmol/l, P < 0.05), while basal PYY levels were equivalent in BN (14.6 +/- 1.3 pmol/l) and control (12.8 +/- 1.1 pmol/l, P = 0.30) subjects. Postprandial ghrelin suppression (decremental ghrelin area under the curve) was significantly attenuated in BN patients, compared to controls (-96.3 +/- 26.8 pmol/l x 3 h vs. -178.2 +/- 25.7 pmol/l x 3 h, P < 0.05). After a meal, the incremental PYY area under the curve in BN patients was significantly blunted from that observed in controls (9.2 +/- 2.6 pmol/l x 3 h vs. 26.8 +/- 3.2 pmol/l x 3 h, P < 0.01). Glucose and insulin responses to meals were similar between the two groups. CONCLUSIONS: BN patients exhibit elevated ghrelin levels before meals with reduced ghrelin suppression after eating. In bulimia nervosa subjects, the rise in PYY levels after meals is also blunted. A gut-hypothalamic pathway involving peripheral signals, such as ghrelin and PYY, may be involved in the pathophysiology of BN.

Uncoupling protein-2/uncoupling protein-3 gene polymorphism is not associated with anorexia nervosa.

Energy expenditure abnormalities have been observed in anorexia nervosa (AN). The uncoupling proteins (UCPs) have been implicated as having a role in energy metabolism and thermogenesis, and an association between a marker flanking the UCP-2/UCP-3 gene cluster and AN has been reported. Also known are insertion/deletion and -866G/A polymorphisms in the UCP-2 gene, and the -55C/T polymorphism in the UCP-3 gene. Differences in these alleles are reportedly related to changes in energy expenditure, body mass index, fat tissue accumulation and obesity. Therefore, this case-control association analysis was done to determine whether any of these UCP-2/3 gene polymorphisms are related to a predisposition to AN. In analysis of a cohort of 106 female Japanese AN sufferers and 126 normal female controls, we found no between-group differences in the polymorphism frequencies of these groups. The hypothesis that differences in the UCP-2/3 gene influence the susceptibility to AN was not supported.

Effect of nutritional rehabilitation on circulating ghrelin and growth hormone levels in patients with anorexia nervosa.

Circulating ghrelin and growth hormone (GH) are up-regulated in anorexia nervosa (AN) as a consequence of prolonged starvation. The current study examines the effect of nutritional rehabilitation with improvement of eating behavior on ghrelin and GH levels in AN patients during the course of inpatient treatment. The subjects included 34 female AN patients and 9 age-matched female controls. Fasting blood samples were collected before, during and after treatment. For data analysis, AN subjects were divided into three subtypes. The first group included seven patients with emergent hospitalization (E-AN), who were accompanied by severe emaciation due to their inability for food intake for more than a month. The other two groups included 14 AN with restricting (AN-R) and 13 AN with binge-eating/purging (AN-BP) patients. There were significant correlations between ghrelin, GH and body mass index (BMI) before treatment in all subjects. However, ghrelin levels were not significantly correlated with BMI and GH although there was a relationship between GH and BMI after treatment. Before treatment, E-AN patients had the highest levels of ghrelin and GH with the lowest glucose levels and liver dysfunction. The AN-BP group had a higher level of ghrelin than the AN-R group. During treatment, comparing with the controls group only the AN-R group showed higher level of ghrelin. Contrarily, the ghrelin levels in the E-AN group, who showed improved glucose levels, and the AN-BP group, who stopped vomiting behavior due to our treatment, decreased ghrelin levels. After treatment, only the AN-BP group showed a higher ghrelin level as compared to the controls. Although GH levels of the three AN groups decreased gradually according to our treatment progress, it still showed the higher value than the control group at the end of the treatment because every AN patients could not reach to more than 80% of their ideal body weight at discharge. These findings suggest that (1) severe emaciation with abnormal fasting hypoglycemia in AN patients may cause very high levels of GH and ghrelin, (2) that GH levels in AN patients may relate to nutritional status and (3) that ghrelin may be influenced by not only nutritional status but also the eating behavior of the patients.

Insulinogenic index at 15 min as a marker of stable eating behavior in bulimia nervosa.

BACKGROUND & AIMS: The aim of this study was to determine the relationship between insulinogenic index at 15 min (II15 min), body weight maintenance, and the presence of vomiting in patients with bulimia nervosa. METHODS: Forty-eight bulimic inpatients and 14 controls underwent an oral glucose tolerance test on the seventh hospital day. We calculated II15 min and other biological markers, including serum amylase concentrations. During the first week after admission, we monitored the frequency of vomiting and calculated changes in body weight. Patients were divided into 4 subgroups according to the presence of vomiting and weight loss. RESULTS: Two-factor analysis of variance of the II15 min value revealed significant main effects of vomiting and body weight change (P < 0.001 for both). The II15 min values for controls and bulimic patients with weight loss and no vomiting were lower than those of other bulimic groups. The II15 min values were positively correlated with serum amylase concentrations (r = 0.37, P < 0.01), body weight change (r = 0.35, P < 0.05), and frequencies of vomiting (r = 0.49, P < 0.05). CONCLUSIONS: These findings suggest that II15 min values may be a useful marker for assessing the stability of eating behavior in patients with bulimia nervosa.

Eating pattern and the effect of oral glucose on ghrelin and insulin secretion in patients with anorexia nervosa.

OBJECTIVE: Ghrelin is thought to be involved in the regulation of eating behaviour and energy metabolism in acute and chronic feeding states. Circulating plasma ghrelin levels in healthy humans have been found to decrease significantly after oral glucose administration. Because it is suggested that eating behaviour may influence the secretion of ghrelin and insulin in anorexia nervosa (AN), we examined the effect of oral glucose on ghrelin and insulin secretion in subtypes of AN patients. DESIGN AND PATIENTS: Twenty female AN patients and 10 age-matched female controls were subjects. The patients were subdivided into two subtypes based on eating behaviour as follows: 11 restricting type (AN-R), nine binge-eating and purging type (AN-BP). Subjects underwent an oral glucose tolerance test at 08.00 h. Blood was collected 0, 30, 60, 120 and 180 min after the glucose load. RESULTS: Both AN-R and AN-BP had a significant increased basal ghrelin level (P < 0.01) and a significantly decreased basal insulin level (P < 0.05) as compared to controls. The time of the nadir of mean ghrelin in AN-BP (120 min, 58.1% of basal level, 204.9 +/- 34.3 pmol/l, mean +/- SEM) was delayed compared to controls (60 min, 60.2%, 74.3 +/- 7.9 pmol/l), and in the AN-R group it kept decreasing for 180 min (80.0%, 182.4 +/- 31.5 pmol/l). The peaks insulin levels in AN-BP (120 min, 319.3 +/- 88.8 pmol/l) and AN-R (180 min, 418.9 +/- 68.4 pmol/l) were also delayed as compared to controls (60 min, 509.2 +/- 88.8 pmol/l). The glucose level at 180 min in AN-R was significantly (P < 0.05) higher than in controls. CONCLUSIONS: These findings suggest that differences in eating behaviour in AN may induce alterations in both ghrelin and insulin metabolism in the acute feeding state. Furthermore, metabolic changes in the restrictive eating pattern may be related to the pathophysiology of small quantitative meal intake in AN-R patients.