Ghrelin and Obestatin in Adolescent Patients with Anorexia Nervosa: Is There an Association with Disordered Eating, Depression, and Obsessive-Compulsive Symptoms?

Abstract

Anorexia nervosa (AN) is an eating disorder characterized by restrictive eating and significant weight loss. In the course of AN, changes are observed in appetite regulation, including orexigenic ghrelin and potentially anorexigenic obestatin. The study aimed to determine if any changes in serum ghrelin and obestatin levels during treatment of AN are observed, while investigating the correlations between these peptides and the severity of disturbed eating attitudes, depression, and anxiety. Thirty adolescent inpatients with AN (examined twice: before hospitalization treatment AN-BT and after treatment AN-AT) and thirty healthy age- and height-matched girls (CG) participated in the study. Anthropometric, serum ghrelin and obestatin concentrations and psychometric evaluations (Eating Attitudes Test 26 Item-EAT-26, Beck Depression Inventory-BDI, Hamilton Depression Rating Scale-HDRS, and Yale Brown Obsessive-Compulsive Scale-Y-BOCS) were performed. The study revealed significantly higher ghrelin and obestatin levels in AN-BT than in AN-AT. A trend toward lower levels during treatment provided partial normalizations. Analyzing correlations in the AN-BT vs. CG group, correlations of peptides with EAT-26, BDI, and HDRS scores were detected. These results suggest a potential role for ghrelin and obestatin in the context of defense mechanisms regulating appetite and body weight in the course of AN and in terms of psychopathological changes co-occurring with this eating disorder.

1. Introduction

Anorexia nervosa (AN) is an eating disorder mainly characterized by dietary restrictions that lead to rapid weight loss, an intense fear of weight gain, and distorted body image perception [1]. As the disease progresses, and therefore as malnutrition increases, patients may develop additional symptoms such as depression or anxiety (most commonly in the form of obsessions, compulsions, or both), which can remarkably complicate the course of treatment [2]. AN is a disease with a long duration, frequent relapses, and a high mortality rate among all mental disorders [3]. The pathomechanism for developing and maintaining AN symptoms is complex, with psychosocial and biological factors playing an important role. Recent Genome-Wide Association Studies (GWAS) [4] indicate significant genetic correlations also with metabolic traits (including, among others: glycemic and lipid). The authors of this study conclude that AN should be reconceptualized as a metabo-psychiatric disorder, and the direction of future research is to integrate the psychiatric and metabolic components. Among the biological (metabolic) factors, the role of enterohormones that exhibit appetite-regulating effects, such as ghrelin and obestatin, is still of interest to researchers dealing with eating disorders [5,6,7]. Due to the common path of synthesis and the opposite effect, both enterohormones may be tested simultaneously [8,9].

Ghrelin is an orexigenic peptide which is mainly secreted by endocrine cells (P/D1) located in the gastric fundus, but its expression has also been described in the duodenum, jejuna, ileum, and colon and at lower concentrations in the pancreas, adipose tissue, kidneys, testes, placenta, hypophysis, and nucleus arcuatus in the hypothalamus, an essential region for appetite regulation [10]. However, in the CNS, it is believed that ghrelin is not produced at levels that meet essential physiological functions (see review [11]). The process of ghrelin synthesis takes place in several stages Figure 1). It is secreted pulsatile and depends on meal intake and other lifestyle factors (e.g., caloric balance, type of diet, nutritional status, stress level, and physical activity) [12], it has also been observed that ghrelin secretion shows a circadian rhythm with the peak of secretion between midnight and two a.m. [13].

The gene encoding human preproghrelin (GHRL) is located on the third chromosome (3p25–26). Preproghrelin (117 amino acids-AA), resulting from the gene’s transcription, is subjected to proteolysis to produce proghrelin (94 AA). Proghrelin is cleaved by the PC1/3 protease, resulting in the formation of non-acylated ghrelin (28 AA) and ghrelin c (precursor of obestatin (23 AA)-an opposing acting peptide). Next, the ghrelin O-acetyltransferase enzyme (GOAT) acylates serine-the third amino acid in the polypeptide chain.

Appetite stimulation is considered the primary function of ghrelin, and it occurs in multiple pathways. Ghrelin interacts with other orexigenic peptides (e.g., neuropeptide Y-NPY) on cells of the arcuate nucleus, leading to the expression of Agouti-related peptide (AgRP). The action of ghrelin on NPY/AgRP neurons is mediated by the downregulation of anorexigenic proopiomelanocortin (POMC) neuronal activity [14]. Based on the initiated changes, the hypothalamic ghrelin receptor type 1 (GHS-R1) is stimulated, resulting in an anabolic response-an increase in food intake and a decrease in energy expenditure. When the NPY/AgRP system is dysregulated, which most likely occurs in the course of starvation, a situation of insufficient ghrelin signaling within the arcuate nucleus arises [14]. At the hypothalamus level, ghrelin affects the ventral nucleus, the paraventricular nucleus, and the vestibular region of the hypothalamus. In addition to the hypothalamus, other brain locations express GHS-R1 ghrelin receptors, such as structures located in the brainstem [14]. In addition to increasing appetite and food intake, the action of ghrelin contributes to the sparing of stored body fat and, in the long term, should increase body weight.

Due to the localization of ghrelin receptors (GHS-R) in the hypothalamus, pituitary, adipose tissue, pancreas, heart, blood vessels, and kidneys, its action is described as pleiotropic. The peptide plays a role in regulating sleep and wakefulness, mood, motivation, and memory. It affects the secretion of neurotransmitters and the regulation of mood [15,16]. In an animal model, ghrelin has been shown to have anxiolytic and antidepressant effects by inhibiting the p38-mitogen-activated protein kinase (MAPK) signaling pathway in mouse hippocampal neurons [17].

As mentioned above, obestatin is formed from preproghrelin. This peptide is involved in appetite regulation by an antagonistic-anorexigenic effect. Like ghrelin, it is secreted in a pulsatile and meal-dependent way. However, such secretion is interpreted as a consequence of a common peptide origin. The initial hypothesis indicated that obestatin probably binds to the GPR39 receptor or GLP-1R receptor of pancreatic β-cells [18], but studies did not confirm this thesis. So far, the expression of obestatin in the CNS has not been established; however, it is believed that it may participate in the activity of the brain–gut axis [19].

The potential role of obestatin is associated with chronic metabolic disorders. Numerous studies have demonstrated a negative correlation between obestatin and serum insulin levels [19]. Obestatin also correlates inversely with Body Mass Index (BMI), leptin, glucose levels, and homeostatic model assessment for insulin resistance (HOMA-IR index); thus, it is considered a protein with an essential role in the pathogenesis of diabetes [20,21]. Nevertheless, the anorexigenic mechanism of obestatin has not been unequivocally described so far. Some studies indicate that it may affect appetite suppression and weight reduction when co-administered with ghrelin, while other studies do not confirm this effect [22,23,24].

Obestatin is often studied with ghrelin since these molecules belong to a common endocrine system and regulate each other’s functions [25]. Due to their involvement in appetite regulation and the dependence of secretion on food intake and BMI, both enterohormones may have a potential role in the persistence of AN symptoms as a disorder proceeding with a drastic change in eating habits and weight loss, often to the level of extreme malnutrition. Furthermore, previous studies have demonstrated protein effects on anxiety-like behaviors and depression in animal models [26,27] and humans [15,28].

A review of the available literature prompted us to perform replication studies of two selected enterohormones in a clinical group of teenage patients with AN compared to a matched control group. Due to the debilitating nature of the disorder, we focused our attention on patients hospitalized for the first time with a disease duration not exceeding one year.

The main objective of this study was to investigate the relationship between the selected anthropometric, psychometric, and enterohormones measurements to discover disease-determining connections, particularly regarding weight restoration based on determinants such the percentage of ideal body weight, severity of disturbed eating attitudes, depression, and anxiety. To achieve this aim, patients with AN were examined in the state of severe malnutrition (anorexia nervosa before treatment, AN-BT group) and after the weight restoration (anorexia nervosa after treatment, AN-AT group).

2. Materials and Methods

2.1. Ethics

This case–control study was carried out at the University of Medical Sciences, Poznan. Approval was obtained from the Bioethics Committee of the Poznan University of Medical Sciences (resolution no. 1029/13). The study followed the rules of the Declaration of Helsinki and complied with Good Clinical Practice guidelines. The course of the study was explained to all participants and their legal guardians, who provided their written informed consent to participate in the study.

2.2. Participants

The study involved adolescent female inpatients admitted in the acute phase of anorexia nervosa (AN). According to the WHO (https://www.who.int/health-topics/adolescent-health#tab=tab_1, accessed on 20 July 2022), adolescence is the phase of life from ages 10 to 19. However, our study included girls up to 18 years old hospitalized in the Department of Child and Adolescent Psychiatry. Thus, adolescents in our study were between 12 and 18 y.o. Patients with AN were hospitalized for the first time and had symptoms of the disease for less than a year with a diagnosis of AN based on International Statistical Classification of Diseases and Related Health Problems version 10 (ICD-10, code F50.0) and The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5,code 307.1) criteria. The control group (CG) consisted of healthy middle school students, matched for age, sex, and height to the AN group.

The exclusion criteria related to general health for the AN group were the same as for the GC group: pregnancy, chronic somatic diseases, other mental or neuro-developmental disorders, medications, dietary supplements, or contraception. Additional exclusion criteria for AN were: failure to obtain target weight (minimal appropriate for particular height and age) and/or increase of BMI less than 2 kg/m², mainly due to discharge on patient (and guardian) request.

During hospitalization all patients were enrolled in a behaviorally oriented nutritional rehabilitation program with the available group, family therapy, or both. The daily caloric intake was 2000–2500 kcal and increased gradually to 3500–4000 kcal, depending on weight gain (1.0–1.5 kg/week). Some patients received medication—mostly hydroxyzine, benzodiazepines, or small doses of atypical antipsychotics—only temporarily (not chronically) to relieve some acute symptoms (severe agitation, anxiety, or insomnia).

The anthropometric, psychometric, and laboratory measurements were performed as a standard at hospital admission (anorexia nervosa before treatment, AN-BT group) and in the last week of hospitalization (anorexia nervosa after treatment, AN-AT group). The CG group was examined once for case-control comparisons.

2.3. Anthropometric Measurements

Anthropometric measurements were carried out by measuring height (cm), fasting morning weight (kg), and on this basis, calculating BMI (kg/m²). The percentage of ideal body weight (%IBW) was calculated according to Lorentz’s formula as a ratio of actual to ideal body weight (IBW) × 100%, where IBW: (height-100) − ((height-150)/2). Interviews were conducted to collect the necessary demographic and disease history data: age, duration of the illness, maximum body weight achieved before the disease.

2.4. Psychometric Measurements

Psychometric measurements were conducted using tests assessing psychopathological symptoms commonly observed in the course of eating disorders such as: Eating Attitudes Test (EAT-26) to measure disordered eating attitudes [29]; Hamilton Depression Rating Scale (HDRS) [30] and Beck Depression Inventory (BDI) [31,32] to measure the degree of depression, and Yale–Brown Obsessive Compulsive Scale (Y-BOCS) to assess the nature and severity of the symptoms of obsessions, compulsions, or both [33].

2.5. Laboratory Testing

All samples were treated according to a consistent protocol to avoid the influence of pre-laboratory errors on the discrepancy of the obtained measurements. The standard operating procedures for serum collection, processing, handling, and storage for biomarker discovery and validation were applied [34]. The 15 mL blood samples were taken in the morning (7:00–8:00 a.m.) after overnight fasting. All the blood samples were allowed to coagulate at room temperature for 30 min and then centrifuged for 15 min at 1000× g. Serum separated by centrifugation was aliquoted and after two-stage freezing, at −20 and then −80 °C, stored until further analyses. The measurements of the total human ghrelin and obestatin were performed using commercially available enzyme-linked immunosorbent assay (ELISA) for research use in serum, plasma, tissue homogenates, cell culture supernates, and other biological fluids. We verified the test based on recommendations for research [35]. The results of both parameters were expressed in pg/mL.

2.6. Ghrelin

A quantitative assay of ghrelin was performed using the Human GHRL/Appetite-regulating hormone ELISA Kit E2142h (EIAab^®, Wuhan, China). The detection range and sensitivity were, respectively, 0.156–10 ng/mL and 0.08 ng/mL. The intra and inter-assay coefficients of variability were, respectively, <6.1% and <9.3% [36,37,38].

2.7. Obestatin

The obestatin concentrations were quantitatively determined using the Human Obestatin ELISA Kit E2039h (EIAab^®, Wuhan, China). The assay ranged between 15.6 and 1000 pg/mL and analytical sensitivity was equal to 10 pg/mL. The intra and the intercoefficient of variation were, respectively, <6.7% and <9.8% [39].

All ELISA tests were performed according to manufacturer’s instructions, without any modifications. All samples and standards were run in duplicate, and the mean value of the two assays was used for statistical evaluation. Optical density was read using a spectrophotometric plate reader (Asys UVM 340 Microplate Reader from Biochrom, Cambridge, UK) for a wavelength of 450 ± 10 nm. A four-parameter algorithm was used to assay concentration in the tested samples.

2.8. Statistical Analysis

Statistical analyses were performed with STATISTICA v13.3 (StatSoft, Cracow, Poland). The data are reported as mean ± SD and median. Normality of distribution was tested using the Shapiro–Wilk test and equality of variances by Levene’s test. Due to the lack of normality for most of the studied variables, more restrictive non-parametric tests were used. The Mann–Whitney U-test was used to compare unpaired groups, while the Wilcoxon test was used for paired groups. Spearman’s rank correlation was used to detect the relationship between variables. All tests were two-sided, and p values ≤ 0.05 were considered significant.

3. Results

The final version of the study presented the results of 30 patients with AN (examined twice: before and after weight restoration) and 30 healthy girls. The mean age in the AN and CG groups was similar-in the AN group; it was 15.80 (±1.69) years, and in the CG group, it was 15.38 (±1.47) years. In terms of height, the groups were also matched, with the AN group having a mean of 162.97 (±5.95) cm and the CG group 164.55 (±4.79) cm (

Table 1. Anthropometric data, fasting serum levels of ghrelin, obestatin and ghrelin:obestatin ratio in adolescents with anorexia before treatment (AN-BT), anorexia after treatment (AN-AT), and in the healthy controls (CG).

	AN-BT (N = 30)			AN-AT (N = 30)			CG (N = 30)			p-Value AN-BT vs. AN-AT	p-Value AN-BT vs. CG	p-Value AN-AT vs. CG
	Mean	SD	Median	Mean	SD	Median	Mean	SD	Median	*	**	**
Treatment duration [days]	58.08	26.27	52.50
Age [years]	15.80	1.69	16.00				15.38	1.47	15.00		0.2400
Height	162.97	5.95	161.00				164.52	4.75	164.00		0.2251
Body mass [kg]	37.34	4.64	39.00	45.72	5.85	45.15	53.73	11.22	53.70	<0.0001	<0.0001	0.0015
BMI [kg/m2]	14.03	1.21	14.24	17.17	1.67	17.42	19.14	5.32	18.67	<0.0001	<0.0001	0.0137
%IBW [%]	66.00	6.10	66.67	80.79	8.25	81.35	93.67	18.44	90.76	<0.0001	<0.0001	0.0021
Ghrelin [pg/mL]	1284.97	1206.58	1040.50	750.62	775.25	505.00	674.52	315.60	660.50	0.0001	0.0002	0.1975
Obestatin [pg/mL]	1319.51	844.86	1027.48	738.64	515.11	555.54	1110.12	959.75	753.61	<0.0001	0.0269	0.0385
Ghrelin:obestatin ratio	1.18	1.11	0.95	1.33	1.39	0.96	2.17	5.94	0.86	0.6420	0.1393	0.1663

Data presented as mean ± SD and median; p-value; bolded—significant p-value < 0.05, BMI—body mass index; IBW—ideal body weight. * Wilcoxon signed-rank test, ** U Mann–Whitney test.

). Subjects from both groups showed similarities regarding education level or other socio-demographic variables.

The results indicate statistically significant differences between data collected before (AN-BT) and after treatment (AN-AT) in the group of female patients. In terms of anthropometric changes, the patients significantly increased their body weight and BMI value, closer to the recommended norms for gender and age. In terms of enterohormone synthesis, significant differences were also observed, showing that after treatment, the patients obtained lower serum ghrelin and obestatin concentrations. Given this unidirectional trend, there were no significant differences in the ghrelin/obestatin ratio between the subgroup of patients before and after treatment (Table 1).

Intergroup comparisons showed statistically significant differences in body weight and BMI values between the subgroup of patients in acute anorexia nervosa (AN-BT) and healthy girls (CG). Despite the minimal assumed weight restoration after treatment, patients with AN-AT still differed significantly from the CG.

Regarding enterohormone synthesis, CG girls achieved significantly lower levels of ghrelin and obestatin than AN-BT patients. The level of ghrelin in the AN-AT group no longer differed significantly from GC, which may indicate a gradual normalization of the appetite regulation system during AN treatment. The obestatin level in the AN-AT group was significantly lower than GC. The observed change may be due to the treatment administered but probably to a different course of normalization for this enterohormone (Table 1). Figure 2 illustrates the differences in both enterohormones concentrations (before and after treatment) compared to the measurements in CG (Figure 2).

Psychometric data indicated improvement in eating attitudes during treatment in patients with AN. Patients after treatment (AN-AT) had significantly lower scores compared to the measurement from the beginning of hospitalization (AN-BT) in total EAT-26 score and subscales. The results also suggest a reduction in the level of depression in the course of treatment—the HDRS total score was significantly lower in the AN-AT subgroup than in AN-BT, as was the BDI total score. The reduction of obsessive-compulsive symptoms during treatment was evidenced by significantly lower Y-BOCS total scores in the AN-AT group compared to AN-BT (

Table 2. Psychometric data in adolescents with anorexia before treatment (AN-BT), anorexia after treatment (AN-AT), and in the healthy controls (CG).

	AN-BT (N = 30)			AN-AT (N = 30)			CG (N = 30)			p-Value AN-BT vs. AN-AT	p-Value AN-BT vs. CG	p-Value AN-AT vs. CG
	Mean	SD	Median	Mean	SD	Median	Mean	SD	Median	*	**	**
EAT-26 total score	29.85	19.77	22.00	17.20	17.88	7.00	5.31	5.64	4.00	0.0004	<0.0001	0.0034
EAT-26 dieting subscale	15.30	12.35	10.00	10.52	11.06	3.00	2.38	3.63	1.00	0.0537	<0.0001	0.0001
EAT-26 bulimia&food preoccupation subscale	4.37	3.60	4.00	2.04	3.14	0.00	0.72	1.41	0.00	0.0045	0.0001	0.3313
EAT-26 oral control subscale	9.19	5.54	8.00	4.64	5.30	4.00	2.28	2.99	2.00	0.0006	<0.0001	0.1335
HDRS total score	12.90	8.09	11.00	7.50	6.82	6.00	1.86	3.14	0.00	0.0004	<0.0001	0.0001
BDI total score	17.21	13.25	12.00	11.81	11.04	9.00	7.33	8.35	4.50	0.0001	0.0007	0.0661
Y-BOCS total score	12.66	9.67	10.00	7.04	8.61	4.00	3.90	4.40	2.50	0.0095	0.0002	0.2931

Data presented as mean ± SD and median; p-value; bolded—significant p value < 0.05; EAT-26-Eating Attitude Test; HDRS—Hamilton Depression Rating Scale; BDI—Beck Depression Inventory; Y-BOCS—Yale–Brown Obsessive Compulsive Scale. * Wilcoxon signed-rank test, ** U Mann–Whitney test.

).

Compared to CG, AN-BT patients differed significantly on psychological test scores, exhibiting more disturbed eating attitudes as measured by EAT-26 total score, dieting subscale, bulimia & food preoccupation subscale, and oral control subscale. After treatment, female patients (AN-AT) had more similar scores to CG. However, significant differences remained in the EAT-26 total score and dieting subscale. AN-BT patients, compared to CG, had higher levels of depression measured by HDRS total score and BDI total score. AN-AT group showed significant differences from CG only in HDRS. After treatment, patients showed a significant reduction in HDRS scores from 12.90 ± 8.09, indicating the presence of depressive symptoms, to 6.82 ± 6.00, indicating the absence of depressive symptoms. AN-BT achieved a significantly higher YBOCS total score compared to CG. Reducing anxiety symptoms during treatment resulted in no differences between AN-AT and CG. After treatment, a significant reduction in the Y-BOCS score was observed from 12.66 ± 9.67 for mild and moderate obsessive-compulsive symptoms to 7.04 ± 8.61 for no or mild symptoms (Table 2).

Correlation analysis (

Table 3. The correlation between fasting serum level of ghrelin and obestatin and %IBW, EAT-25, HDRS, BDI, YBOCS in adolescents with anorexia before treatment (AN-BT), anorexia after treatment (AN-AT), healthy controls (CG), and combined AN-BT + CG group.

	AN-BT				AN-AT				CG				AN-BT + CG
	Ghrelin		Obestatin		Ghrelin		Obestatin		Ghrelin		Obestatin		Ghrelin		Obestatin
	Rs	p-value	Rs	p-value	Rs	p-value	Rs	p-value	Rs	p-value	Rs	p-value	Rs	p-value	Rs	p-value
%IBW	0.44	0.0171	0.21	0.2768	−0.06	0.7680	−0.01	0.9515	0.04	0.8471	−0.16	0.4197	−0.28	0.0355	−0.26	0.0547
EAT-26 total	0.09	0.6692	0.28	0.1541	−0.25	0.2256	−0.24	0.2499	−0.09	0.6254	−0.16	0.4090	0.42	0.0014	0.29	0.0301
EAT-26 dieting	0.14	0.4942	0.37	0.0593	−0.33	0.1105	−0.29	0.1673	0.13	0.5103	−0.04	0.8562	0.43	0.0009	0.29	0.0318
EAT-26 b&f	0.03	0.8953	0.16	0.4173	−0.23	0.2653	−0.11	0.6141	−0.19	0.3234	−0.19	0.3356	0.26	0.0490	0.2	0.1493
EAT-26 oral control	0.14	0.4995	0.2	0.3281	−0.25	0.2354	−0.25	0.2324	−0.13	0.4993	−0.1	0.6214	0.45	0.0006	0.32	0.0181
HDRS total score	−0.22	0.2532	0.05	0.8081	−0.09	0.6769	−0.08	0.6979	0.16	0.4055	0.09	0.6643	0.39	0.0026	0.37	0.0042
BDI total score	−0.06	0.7392	0.16	0.4065	−0.35	0.0752	−0.26	0.1956	0.29	0.1136	0.17	0.3718	0.34	0.0086	0.34	0.0101
Y-BOCS total score	−0.11	0.5754	−0.03	0.8681	−0.27	0.1911	−0.4	0.0438	−0.21	0.2661	0.03	0.8917	0.12	0.3814	0.15	0.2544

Rs-Spearman rank correlation coefficient, p-value; bolded—significant p value < 0.05, %IBW—percent of ideal body weight; EAT-26—Eating Attitude Test; HDRS—Hamilton Depression Rating Scale; BDI—Beck Depression Inventory; Y-BOCS—Yale–Brown Obsessive Compulsive Scale.

) between variables revealed a significant moderate positive correlation (Rs = 0.44; p = 0.0171) between %IBW and ghrelin concentration in the AN-BT subgroup (Figure 3) and a significant moderate negative correlation (Rs = −0.40; p = 0.0438) between Y-BOCS total score and obestatin concentration in the AN-AT subgroup (Figure 4).

The results of the AN-BT and CG groups were combined to investigate the relationships between variables in the broader amplitude of scores obtained. This combination may indicate relationships between a wide range of %IBW (both normative and impaired due to AN) and the aspects of mental functioning and concentrations of the peptides studied. Analyzing the correlation of ghrelin concentration with %IBW in the combined AN-BT + CG group also revealed a significant correlation, but with a negative direction (Rs = −0.28; p = 0.0355). The scatter plots also show the differences in the correlation between ghrelin concentration and %IBW (Figure 5).

Analyzing the combined AN-BT and CG group psychological test scores with ghrelin levels, statistically significant correlations were noted with disordered eating attitudes: EAT-26 total score (Rs = 0.42; p = 0.0014) and subscales: dieting subscale (Rs = 0.43; p < 0.0009), bulimia&food preoccupation (Rs = 0.26; p = 0.0490), oral control (Rs = 0.45; p = 0.0006). Obestatin concentrations also correlated significantly with the results of this test, except for the bulimia&food preoccupation subscale. The strength of the correlation with obestatin levels was slightly weaker than the correlations with ghrelin. Regarding the level of depression, significant positive correlations were observed between HDRS total score and ghrelin (Rs = 0.39; p = 0.0026) and obestatin (Rs = 0.37; p = 0.0042) and between BDI total score and ghrelin (Rs = 0.34; p = 0.0089) and obestatin (Rs = 0.34; p = 0.0101) when analyzing the combined AN-BT + CG scores. No significant correlation with OCD symptoms measured by the Y-BOCS test was found neither for ghrelin nor for obestatin levels in the AN-BT + CG group (Table 3).

4. Discussion

The results indicate that patients with AN have different ghrelin and obestatin secretion than the healthy population. It is known from previous studies that dietary restrictions increase the secretion of ghrelin, which has an orexigenic effect-stimulating appetite, which is a defense mechanism against progressive malnutrition [7,40,41,42]. During AN, several changes occur in releasing essential hormones responsible for regulating appetite, including modification of ghrelin levels [7]. The previous data indicate that acylated fraction and total serum ghrelin level increase during AN. It translates into a higher acylated to desacylated ghrelin ratio than the control group [43]. However, the acylated form of ghrelin, passing through the blood–brain barrier, is responsible for increased appetite and food intake [44].

This mechanism determines the negative correlation between BMI, body fat levels, and ghrelin concentrations [45,46,47,48]. Our study confirmed significantly higher concentrations of ghrelin in a group of patients with AN in a state of severe malnutrition and normalization of ghrelin secretion during nutritional treatment to achieve concentrations comparable to CG. The function of releasing more ghrelin is to increase food intake and motivate eating through CNS stimulation [49]. Moreover, elevated ghrelin levels in patients with AN also have many peripheral functions, including helping to maintain glycemic homeostasis during caloric restriction [50]. Therefore, increased ghrelin levels in cachectic diseases characterized by weight loss have a compensatory function [51].

The involvement of immune mechanisms in the formation of ghrelin resistance is also not excluded-a meta-analysis of studies on immune function among patients with AN showed a significantly higher prevalence of pro-inflammatory cytokines, in particular: TNF-α, IL-6, IL-1β, compared to controls [52]. Neuroimaging studies confirm possible changes in the area of the arcuate nucleus in a group of patients in the advanced stage of AN, which may reflect disturbances in the orexigenic pathway (dysregulation of the NPY/AgRP system, translating into poor reception of ghrelin signals) and lead to resistance to orexigenic effects of the enterohormone [14].

Our study confirmed a negative correlation between %IBW and ghrelin levels but only when we analyzed a broad group of pre-treatment and healthy girls (AN-BT + CG). In the group of severely malnourished patients (AN-BT), this study showed a different correlation between ghrelin concentration and %IBW, with a moderately strong positive correlation. The recent studies on ghrelin indicate that its concentrations may also depend on the level of physical activity [53,54]. Hofmann et al. found a positive correlation between protein concentrations and the level of physical activity [54]. Thus, the positive correlation found in the AN-BT group with %IBW may be indirectly related to the high physical activity of the patients (increased physical activity as a symptom of AN, aimed at reducing energy balance). Patients’ physical activity is deficient during hospitalization, so no significant correlation was detected in the AN-AT group.

Our study reported significantly higher levels of obestatin in the group of patients with AN in the acute phase of the disease and a tendency to reduce the levels of this protein during treatment, assuming an increase in dietary energy supply. These results are consistent with other studies [55,56,57,58]. Analyses of specific factors affecting obestatin metabolism indicate that levels of this peptide are significantly higher in patients with AN of the restrictive type compared with patients with a disorder of the bulimic subtype [55]. Germain et al. [56] compared daily profiles of obestatin in groups of patients with anorexia nervosa, constitutional thinness, and healthy controls. The daily cycle of obestatin in the patient group was significantly higher compared to the group of physiologically thin women and the control group. In AN, total ghrelin to obestatin ratios were substantially lower than in skinny but healthy women. These results suggest a role of obestatin in the pathomechanism of AN independently from BMI or %IBW. Similarly, we did not detect a significant correlation between obestatin levels and %IBW in any study groups. This evidence shows that obestatin levels may depend on other biochemical parameters (e.g., carbohydrate metabolism) rather than strictly on body weight.

We observed a simultaneous decrease in both tested enterohormones in treated patients. The decrease in ghrelin approached the GC values. In contrast, the decline in obestatin reached values lower than those in GC. The reduction in the levels of both enterohormones indicates a normalization process. Nevertheless, both proteins from the proghrelin state undergo different post-translational modifications, which may affect the results obtained. We can assume that continuation therapy in the out-of-hospital setting and further weight gain, which was still lower than GC at the AN-AT stage, will also bring normalization in the case of obestatin.

The present study confirmed significant correlations between enterohormone levels and specific psychopathological symptoms co-occurring with AN. These are positive correlations between protein levels and impaired eating attitudes, as measured by EAT-26 total score, dieting subscale, oral control subscale, and bulimia&food preoccupation (significant correlation only with ghrelin). Noteworthy at this point is that previous studies suggest a lower synthesis of obestatin among patients with bulimic symptoms [55]. In our work, obestatin does not correlate significantly with the EAT-26 subscale measuring bulimic symptoms. Troisi et al. proved a positive correlation between ghrelin levels and scores on the EAT-26 oral control subscale [59]. Still, the authors conducted research on patients with AN and other eating disorders (bulimia nervosa, binge eating disorder). In the same work, the researchers also proved the negative correlation between ghrelin and BULIT-R total score (a test measuring bulimic symptoms). They found it to be a good predictor of ghrelin concentrations. Our study confirmed the correlations of ghrelin with the EAT-26 oral control subscale and the EAT-26 bulimia&food preoccupation subscale, measuring similar variables as the BULIT-R. In a study by Monteleone et al., a correlation was found between ghrelin levels and individual Three-Factor Eating Questionnaire (TFEQ) factor 2 (measures the tendency to lose control over eating) and factor 3 (measures hunger) scores [60]. Monteleone et al. also analyzed correlations with obestatin levels—they found no significant relationships [60]. Our study proves the existence of correlations between protein concentrations and individual variables of impaired eating attitudes. However, it should be emphasized that these correlations are observed when analyzing a broad cross-section of results (combined AN-BT + CG groups).

The present work showed significant correlations between ghrelin and obestatin concentrations and depression levels. Ozsoy et al., in their analysis of ghrelin and leptin concentrations in depressed patients, did not report significant relationships between ghrelin levels and HDRS scores. However, the authors highlighted the higher concentrations of this hormone in depressed patients before treatment than in controls [61]. Associations between ghrelin and depression have been confirmed in patients after suicide attempts and in postmenopausal women [62,63]. Mills et al., in their work, suggest that different ghrelin concentrations may be specific to patients with eating disorders co-occurring with depression rather than depression per se [64]. Our study detected positive correlations between ghrelin and obestatin levels and BDI total score and HDRS total score only when analyzing the combined AN-BT + CG groups.

In contrast, in a study by Emul et al. conducted on OCD patients, no correlation was detected between ghrelin levels and BDI and Y-BOCS scores [65]. We did not observe significant correlations between ghrelin or obestatin levels and OCD symptoms in the subgroup of patients in acute condition (AN-BT) or analyze the combined groups (AN-BT + CG). Only the negative correlation between Y-BOCS and obestatin in the post-treatment group (AN-AT) achieved a significance level. We hypothesized that low levels of this protein might increase OCD symptoms.

The present work has limitations that should be considered when continuing research in presented subject. The first limitation, typical for research conducted on a population of patients with AN, is the size of the group subjected to analyses. In the future, multicenter studies could be considered. Moreover, in future projects, it would be helpful to distinguishing stages of biological development–early and late stage of adolescence and determine the levels of proteins regulating appetite from the perspective of long-term weight restoration. Another limitation of the study may be the metabolism of both molecules, in particular, the rapid degradation processes described by Kosowicz et al. [66] and several factors influencing protein secretion, including the daily cycle, the meal consumed, fluid consumption, body composition parameters, and other biochemical factors such as glucose or insulin levels. It is also reasonable to develop and use objective ways to monitor patients physical activity before and during treatment to assess hyperactivity affecting body weight [67]. In the present study, we attempted to minimize the influence of the above factors as much as possible by collecting biological material according to a standardized procedure.

5. Conclusions

AN’s pathomechanism is still not sufficiently understood, especially in the interaction between metabolic indicators and psychopathological symptoms. Therefore, the analysis of enterohormonal metabolism and its correlation with symptoms of impaired eating, depression, and OCD is an important research issue. The present work confirms the different ghrelin and obestatin metabolism in the course of AN (significantly higher protein levels), differentiating between the acute phase of the disease and the state after nutritional rehabilitation. Specific results—the positive correlation between %IBW and ghrelin levels in the group of patients in the acute phase of the disease— may suggest the existence of a hypothetical threshold of malnutrition, beyond which the defense mechanism of increasing orexigenic ghrelin synthesis disappears. The results of our study confirm that obestatin levels are significantly increased in the acute phase of the disease and suggest that obestatin metabolism is not dependent on %IBW.

Moreover, our results indicate a significant association between proteins and psychopathological symptoms associated with AN. Ghrelin positively correlates with impaired eating attitudes and depression. Obestatin positively correlates with impaired eating (in the context of dietary restriction, eating control, but not bulimic symptoms) and depression. In contrast, it negatively correlates with OCD symptoms, especially in patients with moderately low levels of this protein in response to treatment, including increased dietary energy intake.

Further projects focusing on AN as a metabo-psychiatric disease, exploring the role of specific peptides from the appetite regulation system in the emergence or maintenance of psychopathological symptoms, are necessary to comprehensively understand AN’s pathomechanism. Recent studies have explored the importance of various peptides assayed in serum [68,69,70,71] and saliva of patients with AN [72] looking for optimal assay methods in a state of severe malnutrition. The results of the studies we presented here support and add to the body of knowledge regarding ghrelin as a biomarker in AN. They confirm the potential value of ghrelin as a possible indicator of treatment success or recovery status. The accumulated knowledge supports the development of potential drugs—synthetic ghrelin agonists characterized by greater potency and longer plasma half-life with the ability to penetrate the CNS. Studies of ghrelin agonist use in women with AN are promising because synthetic ghrelin has been shown to accelerate gastric emptying and increase appetite and weight gain. However, due to the pleiotropic nature of ghrelin’s action, the potential for therapeutic effects and activation of a specific biological pathway is unknown [73]. It seems essential to conduct further research in this area and observe somatic changes and effects not only biological but also related to psychopathological symptoms during such treatment.