**4. Conclusions and Further Perspectives**

AN is a devastating and complex disease with a multifactorial origin. Through targeted and untargeted metabolomics mainly based on 1H NMR or MS, the metabolic phenotype of individuals with anorexia nervosa provided has shed light on the metabolic alterations beyond the classical tests. Alterations in pathways such as glycolysis and gluconeogenesis, methionine and cysteine metabolism, serine and glycine metabolism, lipid metabolism, urea cycle, tricarboxylate cycle, phenylalanine and tyrosine metabolism, glutamate, glutamine, proline and histidine metabolism, branched-chain amino acids metabolism, serotonin pathway, kynurenine pathway, indole pathway, and tryptophan metabolism have been revealed. In particular, gut-microbiota metabolites emerge as possible factors contributing to AN development, progression, and maintenance, and deserve further research.

However, validation is a critical point in untargeted metabolomics studies to avoid chance findings. Discrepancies among different studies can be due to the small sample size and the selection of the control group. For this reason, studies with larger sample sizes considering all possible covariates, and validation of results with targeted analytical methods are compulsory to obtain reliable results. Large cohort studies will diminish chance findings, enforcing the statistical significance of the obtained results. Nonetheless, the recruitment of patients with eating disorders is complicated and hinders the capability of obtaining consistent metabolomic findings.

Moreover, given the intestinal dysbiosis in patients with AN and the two-way communication between the gut microbiota and the brain, known as the "gut-brain axis", this relationship could be an interesting aim for new studies on the mechanism and development of AN. In this line, putting together information about different samples such as plasma and feces and integrating different omics as well as psychiatric data will give an understanding of this disease and help to design personalized treatments.

In the articles reviewed, sometimes D- or L- amino acids are referred to in the list of metabolites. After careful reading of the full text, no chiral analysis is described, which makes such chiral assignation doubtful and therefore, D- or L- chiral assignation has not been included in this review. Furthermore, up to now, as a global evaluation, chiral assessment in biological samples from patients is scarcely available, and absent in the case of feces samples; thus, it requires further research and development to unveil the role of chirality in anorexia.

Additionally, longitudinal studies have demonstrated that long term starvation has important metabolic consequences in AN patients. Thus, as discussed throughout this review, whether the described metabolic phenotype is a trait marker of the disease or is a consequence of the food deprivation needs to be further elucidated. Importantly, the application of metabolomics approaches could provide more and better information in longitudinal studies about the development of nutritional disorders, such as the Avon Longitudinal Study [81]. Nevertheless, metabolomics in anorexia nervosa has been almost exclusively employed to study single time points, when the disease is already diagnosed, and comparing versus healthy controls.

Notwithstanding the foregoing, although technological advances in the field of omics sciences are making important progress in the application of precision medicine in most medical diseases, further research in the comprehension of the etiology, pathogenesis, and pathophysiology of eating disorders in general and, in the area of AN in particular, are still required. These studies are necessary to facilitate the future successful application of specific targeted therapies for individuals and thus allowing the groundwork for a new era of precision medicine.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10.339 0/nu13124249/s1, Table S1: Metabolites altered in human samples of individuals with AN.

**Author Contributions:** Funding acquisition, A.G. and J.A.; Project administration, F.J.R. and A.G.; Resources Provision, C.B., J.A. and A.G.; Supervision, F.J.R., A.G., J.A. and C.B.; Visualization, L.M.-M.; Writing—original draft, L.M.-M.; Writing—review & editing, L.M.-M., G.Á.M.-M., M.G., F.J.R., J.A., C.B. and A.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Ministry of Science, Innovation and Universities of Spain (MICINN) and the European Regional Development Fund FEDER, grant number RTI2018-095166-B-I00. supported through, as well as Comunidad de Madrid (B-2017/BMD-3751 "NOVELREN-CM" and Ref. S2017/BMD3684 MOIR-2) and Instituto de Salud Carlos III, European FEDER (grant number FIS-PI19/00166) and CIBEROBN. LMM is recipient of MCINN FPU grant (FPU19/05505).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.
