*1.2. Anorexia Nervosa*

Anorexia nervosa is a psychiatric disorder characterized by excessive dieting, some compensatory behaviors (excessive exercise, vomiting, and use of laxatives) and, specific psychopathological symptoms (disturbances in the perception of body weight and/or image and fear of becoming fat) that leads to severe and maintained weight loss, which results in progressive malnutrition. The American Psychiatry Association, in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) [14], proposed the existence of other restrictive eating disorders such as avoidant/restrictive food intake disorder (ARFID) that presents no specific eating psychopathological symptoms of anorexia (weight concerns and body image disturbance) but features eating or feeding alteration. ARFID includes individuals who meet criteria for the Feeding Disorder of Infancy and Early Childhood DSM-IV category, but also other individuals with clinically significant eating problems who are not included in former DSM categories or therefore must be assigned a diagnosis of eating disorder not otherwise specified (EDNOS), such as selective eating and/or dysphagia [14]. DSM-5 also introduces a new category "Other Specified Feeding or Eating Disorder" (OSFED) for individuals who do not meet criteria for anorexia nervosa, bulimia nervosa or binge disorder and includes five disorders: atypical anorexia, purging disorder, subthreshold bulimia, subthreshold binge eating disorder, and night eating disorder.

AN can be further classified into two main subtypes, restricting and binge/purging disorder. Restrictive anorexia nervosa (AN-R) courses mainly with reduced food intake and excessive exercise, while binge/purging anorexia nervosa (AN-BP) also presents severe energy intake restriction but is combined with recurrent episodes of binge eating or purging behaviors [14,15]. Regardless of the subtype, AN has become one of the most predominant eating disorders with a lifetime prevalence in the general population of 0.6%, being three times higher among females (0.9%) than males (0.3%) [16,17].

The more inclusive DSM-5 criteria reduce the proportion of EDNOS diagnoses regarding DSM-IV and increase the proportion of anorexia and bulimia nervosa, with the new cases probably tending to have a higher minimum body mass index (BMI) and a more benign course [14]. Moreover, it is the eating disorder with the highest mortality rate, mainly due to cardiac complications or suicide [18–20].

Although the etiology of AN remains unclear, there is evidence for disturbed appetite and behavioral pathways that could suggest the physiological origin. Furthermore, it is well established that neurological and genetic predispositions, as well as biological and psychological traits and early experiences in life, might sensitize the individual to stress and hypothalamic-pituitary-adrenal (HPA) axis dysregulation. This sensibility can be further aggravated with environmental and socio-cultural factors that may favor the onset of an eating disorder [21,22]. Thus, AN is postulated to have a multifactorial etiology, and certain conditions may promote the onset in a predisposed population (Figure 1).

Once the disease starts, the maintained weight loss and the altered eating behaviors of the patients lead to wide metabolic dysfunctions complicating the overall clinical picture of the disorder. Among the metabolic alterations, individuals with anorexia nervosa are commonly found to present mild plasma hyper aminoacidemia [23], increased cholesterol levels [24,25], electrolyte imbalances leading to hyponatremia and hypokalemia [26], and profound endocrine disturbances [27–29]. AN patients are reported to have lower triiodothyronine (T3) and thyroxine (T4), showing an altered hypothalamic-pituitary-thyroid axis [26]. In addition, increased cortisol levels in serum and urine have been found, suggesting hyperstimulation of the HPA axis [30,31]. Moreover, patients also have distorted appetite-regulating mechanisms, characterized by increased levels of peptide PYY and ghrelin and decreased concentrations of leptin [26,31–34]. Finally, their extreme eating behaviors lead to micronutrient deficiencies, including reduced levels of zinc, copper, vitamin C, riboflavin, and vitamin B6 [26,35]. Nutritional deprivation can eventually lead to severe complications, including cardiac problems, which is one of the principal causes of death in this disease.

**Figure 1.** Main factors predisposing to the development of anorexia nervosa [21,22].

Recent findings in the alteration of intestinal microbiota due to eating behavior and diet have led to increasing interest in microbiota in eating disorders. Few studies have analyzed the microbiota composition in AN, and modifications in microbiome composition and their by-products are scarcely described [18,36,37]. Several fecal metabolites are involved in mood modulation, learning, and memory mechanisms. Regarding this, short-chain fatty acids, lipopolysaccharides from gram-negative bacteria, and neurotransmitters are microbiota metabolites that have autocrine or paracrine functions in the human body [18,38,39]. Moreover, there are chiral metabolites that can have different stereoisomers D and L, with very different biological activity. Amongst these, amino acids and hydroxy acids are involved in neuro-immuno endocrine regulation. The sources of D-amino acids and D-hydroxy acids are food, endogenous enzymatic processes, and the microbiome. The D form of some amino acids, mainly D-serine and D-aspartate, is altered in psychiatric diseases such as schizophrenia or bipolar disorder, but also depression, a common feature in AN [40–43]. Lactate, a hydroxy acid, is also modified in schizophrenia, depression, and stress disorders [44,45]. Some of these compounds act on receptors in intestinal endothelial cells, and others reach the systemic circulation and can enter the central nervous system, where they mediate different responses [18,38,39]. Thus, the analysis of these microbiota by-products could be relevant for improving our knowledge of the psychopathology of AN. Whether dysbiosis and altered microbiota metabolites are a consequence of malnutrition or if they are involved in AN onset and progression requires further research [21].

To date, the treatment for AN is based on renourishment therapies, as well as psychotherapy and psychopharmacological interventions, to reduce the core psychopathology and the associated disorders (mainly anxiety, depressive and obsessive–compulsive). Unfortunately, specific treatments that target the origin of the disease are still lacking. Thus, there is an increased probability of relapse and to develop a chronic state of the disease, and around 50% of the patients do not achieve full recovery even during a long follow-up period [46]. Treatment objectives should have a strict priority: prevent the death of the patient, prevent the disease from becoming chronic, and attainment of physical and mental recuperation. An integral treatment program should be carried out by a multidisciplinary and coordinated team, including a pediatrician, endocrinologist, psychiatrist, psychologist, and nutritionist.

As a result, the elucidation of the etiology of the disease is of enormous interest to improve treatment and disease outcomes [21]. The pathophysiology of AN could be better elucidated by combining different "omics" approaches to obtain an accurate characterization of the alterations present. In this context, metabolomics appears to be an excellent tool to characterize the metabolic profile of AN patients, leading to the identification of potential alterations that could be useful for the clinical management of anorexia.

The present review aims to present and discuss the information contained in the metabolomics studies that have been performed to date on AN patients with two main objectives: (1) To clearly define the metabolic phenotype of individuals with anorexia nervosa, which is essential for providing new insight into the etiology and pathophysiology of the disease, and (2) To identify the areas that are still uncovered by metabolomics and need further research in the field, with the final purpose of improving disease management and prognosis.
