Genetics of anorexia nervosa: An overview of genome-wide association studies and emerging biological links

Abstract

Anorexia nervosa (AN) is a complex disorder with a strong genetic component. Comorbidities are frequent and there is substantial overlap with other disorders. The lack of understanding of the molecular and neuroanatomical causes has made it difficult to develop effective treatments and it is often difficult to treat in clinical practice. Recent advances in genetics have changed our understanding of polygenic diseases, increasing the possibility of understanding better how molecular pathways are intertwined. This review synthetizes the current state of genetic research providing an overview of genome-wide association studies (GWAS) findings in AN as well as overlap with other disorders, traits, pathways, and imaging results. This paper also discusses the different putative global pathways that are contributing to the disease including the evidence for metabolic and psychiatric origin of the disease.

Full text

Review
Genetics of anorexia nervosa: An overview of genome-wide
association studies and emerging biological links
Clara de Jorge Martı
´nez
a
, Gull Rukh
a
,
*
, Michael J. Williams
a
, Santino Gaudio
a
,
b
,
Samantha Brooks
a
,
c
,
d
, Helgi B. Schi€
oth
a
,
e
a
Department of Neuroscience, Functional Pharmacology, Uppsala University, Uppsala, Sweden
b
Department of Biomedicine and Prevention, University of Rome Tor Vergata, Rome, Italy
c
School of Psychology, Faculty of Health, Liverpool John Moores University, UK
d
Department of Psychology, School of Human and Community Development, University of the Witwatersrand, Johannesburg, South Africa
e
Institute for Translational Medicine and Biotechnology, Sechenov First Moscow State Medical University, Moscow, Russia
ARTICLE INFO
Article history:
Received 1 March 2021
Received in revised form
28 September 2021
Accepted 28 September 2021
Available online 8 October 2021
Keywords:
Anorexia nervosa
Genome-wide association studies
Genetic variants
Psychiatry
ABSTRACT
Anorexia nervosa (AN) is a complex disorder with a strong genetic component. Comorbidities are frequent
and there is substantial overlap with other disorders. The lack of understanding of the molecular and
neuroanatomical causes has made it difficult to develop effective treatments and it is often difficult to treat
in clinical practice. Recent advances in genetics have changed our understanding of polygenic diseases,
increasing the possibility of understanding better how molecular pathways are intertwined. This review
synthetizes the current state of genetic research providing an overview of genome-wide association studies
(GWAS) findings in AN as well as overlap with other disorders, traits, pathways, and imaging results. This
paper also discusses the different putative global pathways that are contributing to the disease including
the evidence for metabolic and psychiatric origin of the disease.
Copyright ©2022, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, and
Genetics Society of China. Published by Elsevier Limited and Science Press. All rights reserved.
Introduction
Anorexia nervosa (AN) has the highest rate of mortality among all
mental disorders with 0.51% of deaths per year (Smink et al., 2012).
Lifetime prevalence estimation for AN in adults is 0.9%e1.4% for
women and 0.2%e0.3% for men (Hudson et al., 2007;Galmiche et
al., 2019), with large differences across continents (Galmiche et al.,
2019). The DSM-5 (Diagnostic and Statistical Manual of Mental
Disorders, 5th ed.) (American Psychiatric Association, 2013) de-
scribes AN as a restrictive eating disorder normally developed at the
onset of puberty (van Noort et al., 2018;Peterson and Fuller, 2019),
consisting of three main criteria: (1) restriction of energy intake in
relation to nutritional requirements, leading to a significantly low
weight; (2) intense fear of gaining weight or becoming fat, or
persistent behaviour interfering the weight gain; and (3) perceptive
distortions of body shape or weight or lack of recognition of actual
low body weight (American Psychiatric Association, 2013). Relapse
in AN is common even in patients achieving full remission, and is
especially critical during the first 18 months following treatment
(Berends et al., 2016). Comorbidities are also frequently being re-
ported, such as major depressive disorders, anxiety disorders,
obsessive-compulsive disorders, developmental disorders among
autistic spectrum and attention-deficit hyperactivity disorder, per-
sonality disorders, substance abuse and borderline traits (Marucci
et al., 2018). AN has been characterized as multifactorial in nature,
interacting environmental, psychological, cultural and biological
factors (Batista et al., 2018). The genetic basis of AN is supported by
the high rate of familial aggregation and heritability (Steinhausen
et al., 2015) as relatives of AN patients are 11-fold more likely to
develop the illness than relatives of healthy individuals (Strober et al.,
2000). Heritability estimates by some of the twin-based studies are
reported to be 50%e60% (Bulik et al., 2015). It is uncontested that
genetics contribute strongly to the aetiology of AN. Genetic studies
have been in the spotlight for the last three decades and since the
advent of the Human Genome Project (HGP), in an effort to under-
stand the molecular basis for this disorder. Genome-wide associa-
tion studies (GWAS) have provided new possibilities to identify the
genes and the pathways involved in the onset of this illness. GWAS
studies provide an unbiased approach for the discovery of underlying
*Corresponding author.
E-mail address: gull.rukh@neuro.uu.se (G. Rukh).
Contents lists available at ScienceDirect
Journal of Genetics and Genomics
Journal homepage: www.journals.elsevier.com/journal-of-genetics-
and-genomics/
https://doi.org/10.1016/j.jgg.2021.09.005
1673-8527/Copyright ©2022, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, and Genetics Society of China. Published by Elsevier Limited and
Science Press. All rights reserved.
Journal of Genetics and Genomics 49 (2022) 1e12

mechanisms and can provide valuable information such as identifi-
cation of the actionable genetic variants (Giacomini et al., 2016).
Deeper knowledge on genetic basis provides the foundation for new
pharmacological approaches and offers a perspective on the po-
tential pharmacological treatments to the AN patients.
Previous conceptualisations of AN have focused on the behav-
ioural, cognitive, and neurological aspects of the disorder. For
example, the Activity-Based-Anorexia (ABA) model (Schalla and
Stengel, 2019) mimics the core behaviours in most suffers of AN
following starvation, alongside neuronal changes, hormonal abnor-
malities and immune system adaptations. However, hyperactivity is
not observed in all those with AN, limiting the explanatory power of
the ABA model. In addition, AN patients consistently demonstrate
cognitive deficits in non-verbal as opposed to verbal performance,
altered attentional bias to disorder-specific stimuli, impaired body
images perception, weak central coherence, reduced set shifting
(inflexible cognitive style) at low weight status, failure in optimal
decision-making processes and greater neural resources required
for working memory associated with poorer performance (Reville
et al., 2016). Finally, structural and functional neurological deficits
are observed in frontal, parietal, basal ganglia and insular cortices,
some of which return to normal following weight recovery, although
discrepancies between subtypes of AN during chronic illness and
recovery remain (Titova et al., 2013;Steinglass and Walsh, 2016).
With these traditional conceptualisations of AN in mind, there re-
mains a gap in the explanatory power of the current behavioural,
cognitive, and neurological models, which may be bridged with a
consideration of the genetic contributions.
In this review, we discuss the current state of genetics research in
AN focusing on GWAS studies conducted to date and the applica-
tions of GWAS data. Additionally, the possible pathways involved in
the development of AN were discussed.
Historical evidence of genetic research on anorexia nervosa
Initially, linkage studies and candidate gene studies were con-
ducted to identify the genes involved in AN. An exhaustive review of
early molecular techniques can be found elsewhere (Rask-Andersen
et al., 2010;Hinney and Volckmar, 2013). Linkage studies are scarce
(Devlin, 2002), however, abundant literature on candidate gene
studies can be found (Gorwood et al., 1998;Klump and Gobrogge,
2005). The linkage studies estimated the prevalence of a trait of in-
terest within the members of a family to identify genomic regions
containing the responsible genes. In 2002, Grice et al. (2002) per-
formed the first linkage study, detecting a signal from chromosome
1p34.2 replicating results in later studies (Devlin, 2002;Nakabayashi
et al., 2009). These linkage studies required large family pedigrees
and entailed a high cost (Breithaupt et al., 2018). Results have proved
inconsistent and a few replications have been achieved (Breithaupt et
al., 2018). More detailed information of linkage studies on AN and ED
can be found in Bulik et al. (2015) and Klump and Gobrogge (2005).
Candidate gene studies rely on an a priori hypothesis and
search for a specific gene or group of genes involved in the trait of
interest, necessitating even larger samples than linkage studies.
Candidate gene studies for AN mainly focused on neurotrans-
mission systems and systems controlling food intake. In 1997,
Collier et al. (1997) found the first association between 5-HT2A
gene promoter polymorphism and AN, focusing later studies on
feeding regulation proteins and neurotransmitter systems.
Significant findings were observed in the serotoninergic system,
dopaminergic system, neuropeptide and feeding regulation
systems associated with Brain-Derived Neurotrophic Factor
(BDNF) (Pinheiro et al., 2009), although, the majority of these re-
sults have not been replicated in larger samples (Himmerich et al.,
2019). A systematic search for candidate genes summarizes the
association of AN with genes related to neurotransmitters, hunger
regulatory systems, rewards system and feeding motivation, en-
ergy metabolism, neuroendocrine system and immune response
(Rask-Andersen et al., 2010). Overall, despite important efforts,
linkage studies and candidate gene studies did not yield conclu-
sive findings on the genetic component of AN, mainly due to small
sample sizes, low statistical power and lack of replication (Baker et
al., 2017). Moreover, these techniques assume an underlying
Mendelian transmission, making them difficult to interpret due to
the polygenic origin of the disease (Lander and Schork, 1994;Fu et
al., 2013;Breithaupt et al., 2018). Candidate gene meta-analyses
have been carried out on the most promising SNPs in an
attempt to obtain more reliable information. The most significant
association was found between the serotonin receptor 5-
hydroxytryptamine receptor 2A (5HTR2A) and increased risk for
AN (Baker et al., 2017) whereas other meta-analyses conducted
over other polymorphism have failed to replicate results (Solmi et
al., 2016). Many of these studies do not consider other highly
relevant alterations, such as mutations in non-coding sections
where most mutations tend to occur (Shih and Woodside, 2016).
After the indication of family aggregation in AN (Strober et al.,
2000;Steinhausen et al., 2015), all subsequent efforts to identify
AN associated genes proved unfruitful. Former genetic techniques
used in the early days of research in this illness lacked the possibility
of providing particularly relevant data because of a weak baseline
hypothesis that would provide limited information from the outset.
The genetic complexity of those conditions not followed by auto-
somal dominant inheritance patterns condemned early genetic
techniques to failure.
The current state of genetic research on anorexia nervosa
Genome-wide association studies
Cross-disorder risk and comorbidities increase the heterogeneity
within complex disorders, increasing the need for larger samples
(Manchia et al., 2013). Several recent studies require hundreds of
thousands of subjects in order to have adequate power (Risch and
Merikangas, 1996). Research on several psychiatric diseases has
required collaboration between institutions to obtain a high number
of participants to achieve meaningful results. For this purpose, in
2007 the Psychiatric Genomics Consortium (PGC) was created
(https://www.med.unc.edu/pgc/), a global collaboration that brings
together more than 800 scientists from over 40 countries with the aim
of advancing the genetic discovery in the psychiatric field.
To date, 13 GWAS on anorexia nervosa and eating disorders have
been conducted (Tables 1 and 2) and a total number of 25 genes
have been identified either with suggestive or significant results,
which previously have been associated with different traits or dis-
orders (Fig. 1 A and 1B). The first GWAS was carried out by
Nakabayashi et al. (2009), on microsatellite markers in a modest
sample comprised of 456 Japanese female patients with ED (331
cases with AN and 125 cases with BN) and 2 control groups totalling
872 healthy individuals. The most significant association was
observed at SNP rs2048332 on chromosome one, but GWAS sig-
nificance was not reached.
Wang et al. (2011) in a European sample comprising of 1033 AN
cases and 3733 paediatric controls. None of the SNPs reached
genome-wide significance level in this study, and the most significant
signal was observed within ZNF804B. Subsequent GWAS also
remained unsuccessful in identifying genetic variants that reach
GWAS significance. In 2012, Boraska et al. (2012) conducted a study
based on six ED phenotypes namely drive for thinness, body
dissatisfaction, bulimia, weight fluctuation symptom, breakfast
skipping behaviour and childhood obsessive-compulsive personality
C. de Jorge Martı´nez, G. Rukh, M.J. Williams et al. Journal of Genetics and Genomics 49 (2022) 1e12
2

disorder, in a sample comprising of 9391 individuals from the Twins
UK discovery dataset. No significant results were found but a sug-
gestive marker within RUFY1 in association with the body dissatis-
faction phenotype was reported.
In 2013, Wade et al. (2013) performed a study in a sample of 2564
female twins from the volunteer adult Australian Twin Registry (ATR),
for four different phenotypes including AN spectrum, bulimia nervosa
spectrum, purging via substances, and a binary measure of no
disordered eating behaviours versus three or more. No variants
reached genome-wide significance, but six regions were suggestive
within CLEC5A,TSHZ1,LOC136242 and SYTL5 for the AN spectrum
phenotype.
In 2014, Boraska et al. (2014) conducted a GWAS comprising of
2907 AN cases from 14 countries and 14,860 controls from the Ge-
netic Consortium for AN (GCAN) and the Wellcome Trust Case
Control Consortium 3 (WTCCC3), and reported 2 intronic variants in
SOX2OT and PPP3CA suggestively associated with AN as the find-
ings did not reach genome-wide significance. Although results are
suggestive and despite Boraska et al. (2014) utilized a larger sample
than previous GWAS, the power of this study was still insufficient.
However, the replication results within the study pointed in the same
direction as the discovery results, being unlikely to be due to chance.
In 2016, Liu et al. (2016) conducted a GWAS comprising 184
subjects with bipolar disorder (BD) and comorbid ED against 1370
controls and 2006 subjects with bipolar disorder only from the Bi-
polar Genome Study (BiGS). Again, a suggestive but not genome-
wide significant signal was found within SOX2-OT locus. The au-
thors suggested that the lack of signals approaching GWAS
significance could be due to the limited sample size, nevertheless, it
provided replication to the study conducted by Boraska et al. (2014)
and it is considered to be the only replication achieved so far in AN
and ED.
In 2017, Duncan et al. (2017) provided an important breakthrough
by identifying the first genome-wide significant locus associated with
AN (Table 2), emphasizing for the first time the reconceptualization of
AN as a metabo-psychiatric disorder. By utilizing a sample
comprising of 3495 AN cases and 10,982 controls from the 1000
Genomes Project, they observed a genome-wide significant signal
(top SNP rs4622308) on chromosome 12 in a region overlapping 6
genes: IKZF4,RPS26,ERBB3,PA2G4,RPL41, and ZC3H10. Previ-
ous GWAS have found associations between these genes and
several disorders (https://drive.google.com/drive/folders/1euEcoDD
qTAhopAxpUuMHNZuFsvvi3is7?usp¼sharing).
This exciting discovery was followed by two more AN GWAS but
with less success (Li et al., 2017;Huckins et al., 2018). Li et al.
(2017) performed a GWAS using data from their previous study
(Wang et al., 2011) but excluding from the sample 212 patients with
AN experiencing diagnostic crossover during the course of the
illness. By reducing phenotypic variability, the study was under-
taken over a more defined AN phenotype aiming to enhance gene
discovery. However, none of the SNPs reached GWAS significance.
Two variants in EBF1 with value close to GWAS significance were
identified, pointing to dysregulation of leptin signalling implicated in
AN (Li et al., 2017).
The other study carried by Huckins et al. (2018) performed an
exome-chip based GWAS in 2158 female AN cases and 15,485
Table 1
Currently published genome-wide association studies of anorexia nervosa and eating disorders and most relevant findings.
Study Phenotype Sample size Genome-wide significance Top SNP Candidate gene
Nakabayashi et al., 2009 AN/BN 456 cases
872 controls
No rs2048332 SPATA17
Wang et al., 2011 AN 1033 cases
3733 controls
No rs6959888 ZNF804B
Boraska et al., 2012 BOD 1934 (qt) No rs6894268 RUFY1
Wade et al., 2013 ED 2564 (ts) No rs145241704
rs62090893
rs56156506
CLEC5A
LOC136242
TSHZ1
SYTL5
Boraska et al., 2014 AN 2907 cases
14.860 controls
No rs9839776
rs17030795
SOX2OT
PPP3CA
Liu et al., 2016 BD/ED 184 BD/ED cases
2006 BD cases
1370 controls
No rs4854912
rs1805576
SOX2-OT
FXR1
Duncan et al., 2017 AN 3495 cases
10982 controls
Yes rs4622308 ERBB3
Li et al., 2017 AN 692 cases
3570 controls
No rs929626 EBF1
Huckins et al., 2018 AN 2158 cases
15485 controls
No rs10791286
rs7700147
OPCML intergenic variant
Warrier et al., 2018 EM 46861 (qt) No rs4882760 TMEM132C
Yilmaz et al., 2018 AN/OCD 3495 AN cases
2688 OCD cases
18013 controls
No rs75063949 LRRC16A
Hu
¨bel et al., 2019a, 2019b BD% 155961 (qt) Yes 77 significant loci associated with
BF% and 174 with fat-free mass
Walton et al., 2019 AN 16992 cases
55525 controls
Yes rs9821797
rs6589488
rs2287348
rs2008387
rs9874207
rs10747478
rs370838138
rs13100344
NCKIPSD
CADM1
ASB3
ERLEC1
MGMT
FOXP1
PTBP2
CDH10
NSUN3
SNP, single nucleotide polymorphism; AN, anorexia nervosa; BN, bulimia nervosa; BOD, body dissatisfaction; ED, eating disorder; BD, bipolar disorder; EM, empathy; OCD,
obsessive compulsive disorder; BF%, body fat percentage; qt, quantitative trait; ts, twin sample.
C. de Jorge Martı´nez, G. Rukh, M.J. Williams et al. Journal of Genetics and Genomics 49 (2022) 1e12
3

Table 2
Overview of genome wide association studies identified genetic variants and gene regions associated with anorexia nervosa.
Chr Gene (region) SNP A1 A2 P-value OR Gene product function Reference
12 IKZF4
RPS26
ERBB3
PA2G4
RPL41
ZC3H10
rs4622308 T C 4.252 10
9
1.2 IKZF4: expression in lymphocytes and is implicated in the control of lymphoid
development. Reported in type 1 diabetes, alopecia areata, adult asthma, vitiligo,
and autoimmune susceptibility.
RPS26: expression in ovary, adrenal and other tissues. Encodes a ribosomal
protein from the S26E family, component of the 40S subunit. Reported in
Diamond-Blackfan anemia, Non-syndromic cleft lip with palate, Pierre Robin
sequence and Klippel Feil syndrome, and cancer.
ERBB3: expression in small intestine, duodenum and other tissues. Encodes a
member of the epidermal growth factor receptor (EGFR) family of receptor tyrosine
kinases.This membrane-bound protein does not have an active kinase domain
but can form heterodimers with other EGF receptors, activating pathways leading
to cell proliferation or differentiation. Reported in cancer and type 1 diabetes.
PA2G4: expression in skin, oesophagus and other tissues. Encodes an RNA-
binding protein involved in growth regulation. This protein is implicated in growth
inhibition and the induction of differentiation of human cancer cells, reported also
in combined features of Diamond Blackfan anemia, Pierre Robin sequence and
Klippel Feil deformity.
RPL41: expression in ovary, colon and other tissues. Encodes a ribosomal protein,
component of the 60S subunit. Reported in cancer, Alzheimer’s disease,
endometriosis, and combined features of Diamond Blackfan anemia, Pierre Robin
Sequence and Klippel Feil Deformity.
ZC3H10: expression in testis, ovary and other tissues. Implied in mitochondrial
regulation. Reported in increased body mass index, fat mass, fasting glucose, and
triglycerides and combined features of Diamond Blackfan anemia, Pierre Robin
Sequence and Klippel Feil Deformity.
Duncan et al., 2017
3NCKIPSD rs9821797 A T 6.99 10
15
1.17 NCKIPSD: expression in ovary, brain and other tissues. Encodes a protein
containing a nuclear localization signal, involved in the formation and maintenance
of dendritic spines, and modulation synaptic activity in neurons. Reported in
leprosy.
Watson et al., 2019
11 CADM1 rs6589488 A T 6.31 10
11
1.14 CADM1: expression in lung, thyroid and other tissues. Cell adhesion molecule 1 is
a member of the immunoglobulin superfamily (IgSF). Reported in Autism spectrum
disorder, attention-deficit/hyperactivity disorder, suicidality, regulation of leptin
sensitivity and bone mass, type 2 diabetes, venous trombosis, atopic dermatitis,
autoimmune alopecia, nephropathies, chronic inflammatory gastrointestinal
diseases, and cancer.
2ASB3
ERLEC1
rs2287348 T C 5.62 10
9
1.11 ASB3: expression in brain, testis and other tissues. Encodes a protein member of
the ankyrin repeat and SOCS box-containing (ASB) family of proteins. Reported in
cancer and type 2 diabetes.
ERLEC1: expression in thyroid, prostate and other tissues. Encodes a resident
endoplasmic reticulum protein that functions in N-glycan recognition. The protein
is thought to be involved in ER-associated degradation. It also functions as a
regulator of multiple cellular stress-response pathways.
Watson et al., 2019
10 MGMT rs2008387 A G 1.73 10
8
1.08 MGMT: expression in liver, kidney and other tissues. Encodes a DNA repair protein
involved in cellular defence against mutagenesis and toxicity from alkylating
agents. Methylation has been associated with several cancer types.
3FOXP1 rs9874207 C T 2.05 10
8
1.08 FOXP1: expression in lung, ovary and other tissues. This gene belongs to
subfamily P of the forkhead box (FOX) transcription factor family that play
important roles in the regulation of tissue- and cell type-specific gene
transcription. Reported in intellectual disability, autism spectrum disorder and
speech and language impairment, glucose homeostasis, congenital heart defects,
severe obstructive sleep apnea, endometriosis, cancer and tumour suppressor,
vitiligo, schizophrenia, and as a contributing factor in Huntington disease.
Watson et al., 2019
1PTBP2 rs10747478 T G 3.13 10
8
1.08 PTBP2: expression in testis, brain and other tissues. Encodes a protein that binds
to intronic polypyrimidine clusters in pre-mRNA molecules and is implicated in
controlling the assembly of other splicing-regulatory proteins. Reported in obesity,
cancer, cognition and social behaviour, and progressive supranuclear palsy.
5CDH10 rs370838138 G C 3.17 10
8
1.08 CDH10: expression in brain and prostate. Encodes a type II classical cadherin of
the cadherin superfamily thought to be involved in synaptic adhesions, axon
outgrowth and guidance. Reported in congenital heart disease, cancer, autism
spectrum disorder, suicidality, schizophrenia, bipolar and major depressive
disorder.
Watson et al., 2019
3NSUN3 rs13100344 T A 4.21 10
8
1.08 NSUN3: expression in colon, testis and other tissues. It is a S-
adenosylmethionine-dependent 5-methylcytosine methyltransferase that is
required for modification of cytosine in the wobble position (C34) of mitochondrial
tRNA. Implicated in regulation of embryonic stem cell differentiation by promoting
mitochondrial activity. Reported in cancer, and encephalomyopathy and seizures.
Information regarding genes expressionand description has been extracted from OMIM database (https://omim.org/), RefSeq database (https://www.ncbi.nlm.nih.gov/refseq/rsg/)
and the cited publications. Chr and gene region are reported as in the original publications for SNP with Pvalues <510
8
. A1, Allele 1; A2, Allele 2; OR, Odds Ratio.
4
C. de Jorge Martı´nez, G. Rukh, M.J. Williams et al. Journal of Genetics and Genomics 49 (2022) 1e12

controls from the WTCCC3 investigating common, low-frequency
and rare variants. No findings reached genome-wide significance,
but two common variants were identified, an intronic variant in
OPCML and an intergenic variant. Low frequency and rare variants
neither reached GWAS significance nor showed any suggestive re-
sults even though the study was well-powered, suggesting that they
may not be involved in aetiology of AN.
One of the alterations commonly attributed to AN is difficulties in
empathy measures, although these results are not exclusive to AN
but are common to other psychiatric disorders. Warrier et al. (2018)
conducted the first GWAS with the aim of exploring the genetic ar-
chitecture of empathy using a sample of 46,861 participants from
23andMe. The main SNPs were identified and heritability and genetic
correlation with certain psychiatric conditions including AN was
performed using linkage disequilibrium risk score (LDSR). The results
obtained in this GWAS are explained in more detail in the cross-
disorder analysis section.
Hu
¨bel et al. (2019a) conducted sex-specific GWAS using a
healthy and medication-free subsample of the UK Biobank
(n¼155,961) and summary statistics from the Eating Disorders
working group of the Psychiatric Genomics Consortium (PGC-ED).
This sex specific GWAS focused on investigating anthropometric
traits in men and women in an attempt to elucidate the factors
contributing to gender bias in the development of AN. Epidemiology
highlights that women are significantly more affected by this disorder
than men, with nine females affected for each male (Bulik et al.,
2006). This study identified 77 genome-wide significant loci for
body fat percentage (BF%) and 174 for fat-free mass (Fig. 1A). A
further meta-analysis with the largest sample size for AN and genetic
correlations obtained for anthropometric and metabolic traits was
performed. A description of the meta-analysis and correlations can
be found in the cross-disorder section.
Watson et al. (2019) performed a GWAS utilizing data from the
Anorexia Nervosa Genetics Initiative (ANGI), the PGC-ED and the UK
Biobank, with the largest sample size used to date comprised of
16,992 AN cases and 55,525 controls. In this recently published
GWAS, eight loci with genome-wide significance (Table 2) were
identified confirming the metabo-psychiatric origin suggested by
earlier studies. This study implies the complete reconceptualization
of the disorder towards what earlier pointed out by Duncan et al.
(2017) highlighting the metabolic component of AN. No previous
studies had been sufficiently well-powered to identify so many vari-
ants reaching the threshold for genome-wide significance (Boraska
et al., 2012) until the present work. The loci harbouring genes
CADM1,MGMT,FOXP1, and PTBP2 obtained the most robust re-
sults pointing to their implication in AN aetiology. These eight loci
reaching genome wide significance had not been identified as
possible candidates in earlier studies either. Besides, among these
findings no significant signals were obtained from chromosome 12,
failing to replicate the results obtained in a previous study by Duncan
et al. (2017).
Pathway analysis was also performed in order to test whether
genes associated with AN were more enriched in a determined
pathway (Watson et al., 2019). Results were significant for one
pathway, GO: positive_regulation_of_embryonic_development,
pointing to a possible first indication of a specific pathway involved in
the development of AN. A detailed overview of the gene regions and
implications can be found in Table 1.
GWAS are reporting findings of uncontested value for under-
standing the genetics of AN. The identification of the first significant
loci offers a broader scope, including contributing factors that until
now have not been considered in the onset of AN. Although only a
few works have released significant results, GWAS applied to eating
disorders are still at an early stage providing guidelines and a
framework for future GWAS on AN. The need for a significant in-
crease in the sample size in GWAS is understood to obtain more
robust findings.
The genes identified in these GWAS have been associated in
different genetic studies with a variety of traits and disorders. For
these associations, a supplementary extended list of genes with
significant associations and their bibliographic references can be
found here. In addition, these new insights from recent studies are
prompting increased emphasis that metabolic dysregulation may
play an underlying role in this disorder.
Fig. 1. Genetic overlap of anorexia nervosa associated loci with other traits and disorders. A: Genome wide significant loci for anorexia nervosa. B: Suggestive genes for anorexia
nervosa. Associated traits/disorders were identified and classified on a scale according to the number of publications, in order to provide a quantitative overview. The genes with 10
publications are all included in the references, but those with >10 publication s, a selection is shown. All references are included in the extended gene list provided in the supplementary
material.
C. de Jorge Martı´nez, G. Rukh, M.J. Williams et al. Journal of Genetics and Genomics 49 (2022) 1e12
5

Cross-disorder analysis
Cross-disorder analysis uses the data generated by GWAS to
examine shared variants across disorders, allowing the identification
of the genetic overlap occurring between different disorders or traits
of interest. This technique has contributed to a better understanding
of psychiatric comorbidities and diagnostic cross-over, so frequent
in the case of EDs. The frequent comorbidities in psychiatric disor-
ders could be due to genetic overlap of different disorders or it could
indicate a general polygenic predisposition to psychopathology
(Docherty et al., 2016). Traditional GWAS are based on univariate
tests across millions of common variants along the genome, while
newest genetic approaches include aggregation effects and quanti-
fication of the risk (Docherty et al., 2016). The increasing collabora-
tion between institutions and the creation of databases of biological
samples and repositories permit the exploration of variants across
disorders. There are currently several methodologies available to
identify genetic overlap but only two have been used so far in the
study of AN, cross-trait analysis (Hinney et al., 2017) and linkage
disequilibrium score regression (LDSC) (Bulik-Sullivan et al., 2015;
Duncan et al., 2017;Warrier et al., 2018;Yilmaz et al., 2018;Hu
¨bel
et al., 2019a;Watson et al., 2019).
Cross-trait analysis consists of the analysis of the association
between two phenotypes based on a predetermined significance or
threshold, e.g., selecting the 1000 SNPs with the lowest P-values in a
genome-wide association meta-analysis, while LDSC includes a
wide range of phenotypes within the analysis, requiring only GWAS
summary statistics and not presenting any bias by sample overlap. A
brief description of the cross-disorder analysis conducted so far in
the investigation of AN in chronological order is provided and a
summary of the main genetic correlations is presented in Table 3.
Hinney et al. (2017) performed the only cross-trait analysis of risk
alleles for AN. The analysis was comprised of 1000 SNPs with the
lowest P-values from the genome-wide association meta-analysis
(GWAMA) of AN (GCAN), with evidence of association in the largest
published GWAMA for BMI (GIANT). Significant associations were
found for nine SNPs at three independent loci, pointing to variants
involved in risk for AN may also be associated with increased BMI.
This study presents conflicting results with other studies, reflecting
the complexity of the relation between AN and BMI. In this study,
variants associated to increased BMI enhances risk for AN. The au-
thors hypothesized that some specific variants involved in AN and
increased BMI, may lead to increase or decrease in BMI depending
on additional factors, or the variant may predispose to BMI dysre-
gulation in interaction with other genetic or environmental factors.
Bulik-Sullivan et al. (2015) performed the first cross-disorder study
using LDSC. This study estimated 276 genetic correlations among 24
traits, including genetic correlations between AN and schizophrenia,
and AN and obesity. In contrast to findings by Hinney et al. (2017),a
negative genetic correlation between AN and obesity was reported.
Also, results indicated a positive genetic correlation between AN and
schizophrenia. The authors suggested that the same genetic factors
may influence normal variation and dysregulation in BMI occurring in
mental disorders (Bulik-Sullivan et al., 2015).
Duncan et al. (2017) performed a large LDSC including 159
different phenotypes related to genetic risk for AN. Although this
study, tested fewer phenotypes than were tested by Bulik-Sullivan
et al. (2015), associations were found with more metabolic compo-
nents beyond BMI. Significant positive genetic correlations between
AN and other phenotypes were observed including neuroticism,
schizophrenia and cross-disorder psychiatric risk, educational
attainment and HDL cholesterol. Significant negative genetic corre-
lations were observed between AN and metabolic and anthropo-
metric traits including obesity, normal-range body mass index,
insulin, glucose, and lipid phenotypes. The strongest correlations of
all phenotypes were with insulin resistance (r
g
¼0.50) and fasting
insulin (r
g
¼0.41). Genome-wide common variant heritability was
also calculated and was established at 0.20.
While metabolic, anthropometric and psychiatric traits have
gathered nearly all the interest of cross-disorder studies, one study
aimed to explore psychological traits. Warrier et al. (2018) performed
the first GWAS of self-reported empathy over the obtained empathy
quotient (EQ). Studies on empathy in AN patients have yielded con-
flicting data, some studies have reported greater alexithymia and
personal distress (Beadle et al., 2013) and lower cognitive empathy
scores (Kerr-Gaffney et al., 2019) while other report no differences
between AN patients and healthy controls (Hambrook et al., 2010)
The results from the GWAS and its most significant SNPs were
selected for LDSC to identify genetic correlation between the EQ and
different psychiatric conditions including autism, schizophrenia, and
AN. The results obtained reported a significant positive genetic
correlation between the EQ and risk for AN, correlation that was also
shared with schizophrenia. These results would contradict the
studies pointing to diminished empathy in AN, nevertheless, the
authors suggest that this correlation could be mediated by educa-
tional attainment.
Due to the shared similarities and mixed symptomatology of AN
with other psychiatric disorders, more studies have focused on
cross-disorder and psychiatric conditions. Obsessive-compulsive
disorder (OCD) has received special interest in the study of AN due
to the comorbidity between both disorders and their shared
Table 3
Main significant genetic correlations from cross-disorder analysis.
Trait/Disorder r
g
SE P-value Reference
BMI schizophrenia 0.18
0.19
0.04
0.04
310
7
210
5
Bulik et al., 2015
Neuroticism schizophrenia
Psychiatric phenotypes
Years of education
Attending college
Extreme high BMI
Insulin resistance
Fasting insulin
Fasting glucose
Total cholesterol
Total lipids in VLDL
Phospholipids in VLDL
LDL cholesterol
0.39
0.29
0.22
0.34
0.30
0.29
0.50
0.41
0.26
0.39
0.30
0.33
0.20
0.14
0.07
0.07
0.08
0.07
0.08
0.11
0.09
0.07
0.12
0.12
0.13
0.08
4.4 10
3
4.4 10
5
3.4 10
3
5.2 10
6
4.4 10
5
2.0 10
4
1.3 10
5
5.2 10
6
3.0 10
4
1.6 10
3
0.01
4.4 10
3
0.011
Duncan et al., 2017
EQ 0.32 0.09 6 10
4
Warrier et al., 2018
OCD 0.49 0.13 9.07 10
7
Yilmaz et al., 2018
BF%
female
BF%
male
fat-free mass
Fasting insulin
female
Fasting insulin
male
0.44
0.26
0.14
0.36
0.16
0.04
0.04
0.03
0.07
0.05
8.28 10
27
1.04 10
13
5.79 10
6
5.29 10
7
0.003
Hu
¨bel et al., 2019a
OCD
MDD anxiety disorders
Schizophrenia
Physical activity
Educational attainment
Fat mass
Fat-free mass
BMI obesity
Type 2 diabetes
Fasting insulin
Insulin resistance
Leptin
HDL cholesterol
0.45
0.28
0.25
0.25
0.17
0.25
0.33
0.12
0.32
0.22
0.22
0.24
0.29
0.26
0.21
0.08
0.07
0.05
0.03
0.05
0.03
0.03
0.03
0.03
0.03
0.05
0.06
0.07
0.06
0.04
4.97 10
9
8.95 10
5
8.90 10
8
4.61 10
18
1.00 10
4
1.69 10
15
7.23 10
25
4.65 10
5
8.93 10
25
2.96 10
11
3.82 10
5
2.31 10
5
2.83 10
5
4.98 10
5
3.08 10
7
Walton et al., 2019
BMI, body mass index; VLDL, very low density lipoprotein; LDL, low density lipopro-
tein; EQ, empathy quotient; OCD, obsessive-compulsive disorder; BF%, body fat
percentage; MDD, major depressive disorder; HDL, high density lipoprotein.
C. de Jorge Martı´nez, G. Rukh, M.J. Williams et al. Journal of Genetics and Genomics 49 (2022) 1e12
6

symptoms (Błachno and Bry
nska, 2012.). Yilmaz et al. (2018) per-
formed a cross-disorder GWAS meta-analysis of 3495 AN cases,
2688 OCD cases and 18,013 controls, confirming the high genetic
correlation between AN and OCD (r
g
¼0.49). The SNP heritability for
the cross-disorder phenotype was established at h
2
¼0.21 and
strong genetic correlations of the cross-disorder phenotype were
shown with other psychiatric disorders and metabolic traits. A pos-
itive correlation was reported with bipolar disorder and neuroticism
and a negative correlation with BMI and triglycerides. Although an
overlap occurs in both disorders, the follow-up analysis showed that
correlations with metabolic and anthropometric traits resulted
notably stronger for AN phenotype.
As mentioned earlier, Hu
¨bel et al. (2019a), following sex-specific
GWAS on a healthy sample, performed an meta-analysis with the
largest sample size until date comprising data from the Watson et al.
(2019) and additional samples from the GCAN, the WTCCC3
(Boraska et al., 2014), and the UK Biobank (Sudlow et al., 2015)
comprising a total sample of 72,358 females (16,531 AN cases) and
24,454 males (460 AN cases). The GWAS resulted in the same eight
genome-wide significant loci found in Watson et al. (2019) since there
is a major sample overlap with this study. Heritability and correlations
with anthropometric traits, metabolic traits, physical activity, edu-
cation years and other psychiatric disorders were obtained using
BOLT-LMM and LDSC estimates. Genetic correlations for BF% with
AN resulted in significantly different outcomes for male and female,
pointing to sex-specific anthropometric and metabolic traits as fac-
tors increasing liability to develop AN. Genetic correlations of
anthropometric traits across the lifespan were calculated indicating
heterogeneity between females and males in genetic architecture but
suggesting that a proportion of BF%-associated genomic variation
may become active at a later age. Significant correlations were found
between AN and major depressive disorder, anxiety, OCD, neuroti-
cism, schizophrenia and education years, not differing between
sexes except for male with OCD and major depressive disorder. This
study reports replication for several traits observed in previous
studies such as the genetic correlation between AN and OCD (Yilmaz
et al., 2018), schizophrenia (Bulik-Sullivan et al., 2015;Duncan et al.,
2017), neuroticism (Duncan et al., 2017) and educational attainment
(Duncan et al., 2017).
In terms of other traits relevant to the development of eating
disorders, the restraint versus binge eating spectrum has been
touted (Brooks et al., 2012) and variance in eating behaviour has
been associated with the FTO rs9939609 genotype (Wade et al.,
2020). Carriers of this variation are found to consume more food
than “normal”weight population, moreover, A/A FTO homozygotes,
are also found to eat at a faster rate. Strong associations have been
found between obesity and the A allele (Zabena et al., 2009), also
observed an increased expression of FTO mRNA in subcutaneous
adipose tissue. This gene locus has been examined in a recent study
as a modulator of psychopathological features in AN. FTO haplo-
types presented strong association with interoceptive awareness,
bulimia and maturity fears, besides anxiety, depression and phobic
anxiety within distal regions in AN patients, pointing to this locus as a
relevant region (Gonz
alez et al., 2021).
It is most recently argued that a two-dimensional spectrum be-
tween restriction and over-ingest is more applicable to eating dis-
orders, additionally incorporating impulsivity and compulsivity
(Brooks and Schi€
oth, 2019). For example, impulsive traits have been
associated with the lack-of-control underlying binge eating behav-
iours, whereas compulsive traits typify the obsessive cognitions
characteristic of restrained eating and AN (Lavender et al., 2017).
However, a recent review does not demonstrate that AN is associ-
ated with compulsivity and bulimia nervosa with impulsivity, but
rather highlights the heterogeneity of subtypes of disorder and
experimental methods, as well as the transdiagnostic nature of
chronic illness is shrouding definitive conclusions (Howard et al.,
2020). Nevertheless, a wealth of previous research links impulsivity
and compulsivity to variance in eating behaviour (Brooks and
Schi€
oth, 2019), but the lack of definitive conclusions from
cognitive-behavioural studies suggests that genetic analysis could
prove fruitful. Analysis of genetic association to impulsivity and
compulsivity in AN have been scarce, although collectively, genetic
studies have demonstrated that impulsivity is associated with striatal
dopamine function and low levels of D2/D3 receptor availability
(London, 2020), and that D2 receptor expression is associated with
the Disrupted in Schizophrenia 1 (DISC1) gene (Su et al., 2020).
Conversely, compulsivity is a broader construct than impulsivity
(Brooks et al., 2017) and so might be more difficult to pinpoint in
terms of genetic linkage. Nevertheless, compulsivity is associated
with excessive top-down control of appetite and comorbid OCD in
AN, and in a recent genetic study, disrupted metabolic systems
showed some link with the subtypes of OCD and SNPs on the genes
SETD3 and CPE (Alemany-Navarro et al., 2020). Taken together,
while previous literature has linked variance in impulsivity and
compulsivity to eating disorders, the lack of consensus linking these
traits to AN specifically warrants further exploration of underlying
metabolic systems and genetic susceptibility.
Finally, the latest cross-disorder analysis is the recent work
published by Watson et al. (2019). Apart from identifying eight loci in
GWAS analyses, they conducted a series of analyses and meta-
analyses to explore relations with other phenotypes. Heritability
was estimated with LDSC at 11%e17%, providing a lower value than
Duncan et al. (2017). Multi-trait analysis revealed associations be-
tween SNP variants and type 2 diabetes, education years, HDL
cholesterol, neuroticism, and schizophrenia. SNP-based genetic
correlations with external traits using bivariate LDSC reported posi-
tive correlations with obsessive-compulsive disorder, major
depression disorder, anxiety disorders and schizophrenia reflecting
comorbidities common in clinical presentations. In line with previous
studies (Bulik-Sullivan et al., 2015;Duncan et al., 2017;Yilmaz et al.,
2018;Hu
¨bel et al., 2019a), anthropometric and metabolic traits were
explored presenting negative correlation with fat mass, fat-free
mass, BMI, obesity, type 2 diabetes, fasting insulin, insulin resis-
tance, and leptin, and a positive correlation with HDL cholesterol.
Data from cross-disorder studies have shown a high degree of
overlap between AN and other psychiatric disorders, indicating that
risk genes are shared between these disorders. A vulnerability to
general psychopathology may underlie this shared genetic variation.
At the same time, associations reported for variants related to
metabolic and anthropometric traits emphasize its relevance in the
origin of AN.
Polygenic risk score
Polygenic risk score (PRS) elaborates individualized profiles for
genetic risk for a determined disorder using SNPs information con-
tained in GWASs, generating an overall composite score. The three
main applications from these composite scores are within-trait as-
sociation, cross-trait association and prediction (Breithaupt et al.,
2018). This technique potentially enables enhanced screening and
permits early diagnosis and preventive therapies for those people
identified to be at risk. In 2018, a study managed to identify people at
genetic risk for five diseases (Khera et al., 2018). However, the
technique can be hampered by poor GWAS results and requires well-
powered GWASs to make accurate predictions.
As improvements in the implementation of PRS occur, this tech-
nique may potentially elaborate genetic predictors in those diseases
of polygenic origin. Currently, genetic risk identification is only
applied in clinical practice for diseases of monogenic origin such as
Huntington’s disease (Migliore et al., 2019). Only single-gene
C. de Jorge Martı´nez, G. Rukh, M.J. Williams et al. Journal of Genetics and Genomics 49 (2022) 1e12
7

diseases can currently benefit from genetic counselling and, in some
cases, early diagnosis. Although there is not yet a technique that can
generate predictions of individual risk for genetic disease compara-
ble to monogenic mutations, the use of PRS could turn out to be very
useful for polygenic diseases (Zheutlin and Ross, 2018). Obtaining
these risk scores would imply that the carriers of risk variants
could also benefit from the same counselling and interventions
addressed to patients meeting other high-risk criteria (Zheutlin and
Ross, 2018).
The only PRS analysis conducted in AN to date was performed by
Watson et al. Results from the analysis indicated that the PRS
captured ~1.7% of the phenotypic variance on the liability scale for
discovery P¼0.5 in the within-trait analysis. No differences in
polygenic architecture were found between two AN subtypes, those
presenting binge eating and without binge eating, and between
males and females (Watson et al., 2019).
Pathway enrichment analysis applied to significant GWAS
genes
Pathway enrichment analysis is a statistical method highly useful
when working with large gene datasets. This can result in a tedious
and highly complex task and understanding the underlying biolog-
ical processes may prove elusive. Gene ontology (GO) consists of a
taxonomy that statistically determines those sets of genes involved
in the same biological processes, functions, or localizations
(Carbon et al., 2019). These genes are classified into subgroups
providing a nomenclature that describes their gene products. Thus,
the task of gaining insight into the underlying biology of differentially
expressed genes and proteins is simplified (Hung, 2013;Reimand
et al., 2019).
A gene data set was obtained from GWAS, selecting significant
and adjacent genes in order to explore perturbed pathways in AN
(Table S1). Biological processes (GO terms) involved in these genes
were explored with Panther utility (http://pantherdb.org/). Results
revealed that cellular process (GO:0009987), biological regulation
(GO:0065007) and metabolic process (GO:0008152) composed the
most represented categories of biological processes (Fig. 2;
Tables S2 and S3).
The gene set was tested for enrichment analysis using David
(https://david.ncifcrf.gov/). Enrichment analysis resulted initially
significant for 12 predicted pathways, including GO: cellular
response to UV-B, GO: negative regulation of RNA splicing, GO:
ossification involved in bone maturation, GO: hepatocyte growth
factor receptor signalling pathway, GO: GTP biosynthetic process,
GO: modulation of synaptic transmission, GO: response to reactive
oxygen species, GO: cellular response to interleukin-1, GO: retina
development in camera-type eye, GO: response to virus, GO:
negative regulation of cell growth, and GO: intracellular signal
transduction. None of the initially significant GOs survived after
multiple testing correction (Table S4).
The genes conforming these GOs and their biological products
might contribute to AN pathogenesis requiring further studies. Ge-
netic enrichment analysis in the field of eating disorders remains
unexplored and related studies are somewhat anecdotal. These
studies could help to crack genetics underlying eating disorders, and
thereby, improve treatment outcomes.
Future prospects and conclusions
The GWAS results have strengthened the notion that anthro-
pometric and metabolic traits might be playing an important role in
the aetiology of AN. Most of the cross-trait analyses show negative
genetic correlations between AN and anthropometric traits such as
BMI and BF% (Bulik-Sullivan et al., 2015;Duncan et al., 2017;
Yilmaz et al., 2018;Hu
¨bel et al., 2019a) and cross-disorder corre-
lations have also been found with metabolic diseases such as type
1diabetes(Duncan et al., 2017). The results also highlight the
complexity of the genetic component in AN involving range of
systems that at this stage may have an obscure role and but also
that underlying psychiatric disorders may not necessarily be the
dominating component.
GWAS studies are exploratory and based on a hypothesis-free
approach. The current findings, although very revealing and
requiring replication in further studies, are directing us to involve
genes related to metabolism and anthropometric traits, and to
consider the important contribution they have on the development of
AN. Venturing a hypothesis about how a given gene might predis-
pose to the development of the disorder would be somehow spec-
ulative at this stage. However, disruption of genes affecting
metabolism and anthropometric traits should be considered as an
important contributing factor to AN, especially genetic variants
associated with low BMI.
Low body weight is a pathognomonic symptom of AN, and it is
thought as a consequence of the disorder. Carriers of low BMI-
related genotype could entail a higher risk of developing AN, by a
difficulty in maintaining body weight within the normal range. Genetic
predisposition to lower BMI might be key in the onset of this illness
and should not be only considered as an effect of the illness but also
a contributing factor. One of the eating disorder phenotypes which
have not received much attention so far, is atypical anorexia nervosa
(AAN) (Moskowitz and Weiselberg, 2017) that could help exploring
this relationship between predisposition to low BMI and the AN
phenotype. This variant of AN is characterized by same symptom-
atology except for a BMI within the normal range or above (Sawyer
et al., 2016). Thus, by comparing the genetic variation between AN
and AAN, patients could be provided with more information and may
help in devising more personalized approach to treatment. The
Fig. 2. Gene enrichment and pathway analysis. A total of 125 selected genes from
genome-wide association studies were subjected to gene ontology (GO) classification
using the PANTHER gene analysis tool in order to explore biological processes.
C. de Jorge Martı´nez, G. Rukh, M.J. Williams et al. Journal of Genetics and Genomics 49 (2022) 1e12
8

conception of AN as a psychiatric disorder has already been chal-
lenged, considering that symptoms produced in AN are triggered by
an evolutive response consequence of starvation, that assists the
individual in restoring feeding behaviour (S€
odersten et al., 2019).
Under this perspective, psychiatric symptoms are produced by
altered eating behaviours, resulting starvation the trigger for AN
presentation, and rejecting the mental causation hypothesis
(S€
odersten et al., 2019). Anxiety, obsession and depression,
frequently associated with AN, can be triggered by starvation, and
this fact is known long ago in scientific literature. The Minnesota
Starvation Experiment, aclassic clinical study conducted by Ancel
Keys, was pioneering in exploring the physiological and psycholog-
ical effects of prolonged food restriction. Most of the participants
experienced strong emotional distress, depression, anxiety, irrita-
bility, obsession with food, isolation and decreased cognitive per-
formance after undergoing food deprivation (Widdowson, 1952;Kalm
and Semba, 2005). This experiment probed that not only those
affected by mental disorders, but any person would be vulnerable to
experience most of the consequences as those occurring in eating
disorders after undergoing food restriction. However, although the
starvation effects reproduce most of the symptoms present in AN,
there is an important exception in body image distortion, which is a
key symptom from AN, would not be affected. This indicates that,
despite new evidence indicates that we are facing a disorder whose
basis does not seem to be exclusively mental, psychological factors
continue to be very relevant.
Another perspective is that eating disorders could be placed on a
continuum where the opposite side of the AN would be represented
by obesity, which is not classified as a psychiatric disorder in either
the DSM-IV-TR (Diagnostic and Statistical Manual of Mental Disor-
ders, 4th ed., Text Revision) (American Psychiatric Association, 2000)
or the DSM-5 (American Psychiatric Association, 2013). Only the
ICD-10 (International Statistical Classification of Diseases and
Related Health Problems, 10th revision, 2nd ed). World health or-
ganization recognizes a category related to polyphagia in other
mental disorders (World Health Organization, 2004). In this case,
obesity is treated as an endocrine, metabolic and nutritional disease,
and despite obesity is increasingly being related to psychiatric dis-
orders (Berkowitz and Fabricatore, 2005), the psychological
component remains anecdotal. This metabolic and endocrine aspect
present in obesity has not been considered in the case of AN, being
exclusively categorized as a severe mental disorder. The findings
from GWAS in anthropometric traits, reporting negative correlations
between AN and anthropometric traits such as BMI and BF% (Bulik-
Sullivan et al., 2015;Duncan et al., 2017;Yilmaz et al., 2018;Hu
¨bel
et al., 2019a) are in agreement with this continuum in which AN
and obesity would be in the extreme of the presentations of patho-
logical weight regulation. While such models are simplification, this
could be a foundation to build more integrated complex models that
explain better the aetiology of AN.
Watson et al. (2019) defend the reconceptualization of the AN as a
metabo-psychiatric disorder after reporting significance in genes
involved in metabolic and anthropometric traits (see Table 1). New
data released by GWAS confirm the multi-causality underlying to this
disorder and its high complexity, in which genes increasing risk for
AN fit in with genes increasing risk for other mental disorders and
traits (see Table 3). Thus, the heterogeneity of this disorder is re-
flected in its clinical presentations. “Pure syndromes”are not very
frequent within the health care and co-occurrence with symptom-
atology from other psychiatric disorders provide better account for
the clinical presentations (Jagielska and Kacperska, 2017;Marucci
et al., 2018). Mixed symptoms and overlap among diagnoses are
very common for mental disorders and psychiatric comorbidities
have been suggested as an artefact of current diagnostic systems
(Maj, 2005;Øiesvold et al., 2013). According to this, qualitative
categorization might turn out to be problematic to describe patients
accurately and might interfere research in negative manner. Specific
mental disorders traditionally linked to AN do not differ qualitatively
but vary along overlapping dimensions. The transdiagnostic model is
receiving increased interest and might be the solution to reflect more
reliably the overlap between disorders and the continuum in domains
targeted by clinical practice and research (Fairburn et al., 2003;
Curzio et al., 2018;Fusar-Poli et al., 2019). Eating disorders, along
with anxiety disorders, obsessive disorders and depressive disorders
among other, are not specific and well-defined categories, but these
vary along a dimensional overlapping continuum, in which genetics
would be one of the main factors increasing vulnerability for mental
disorder in general. The P factor proposed by Caspi et al. (2014)
might account for such unspecific vulnerability for psychiatric dis-
order, this P factor would correspond to a general psychopathology
dimension gathering three higher-order factors (internalizing, exter-
nalizing, and thought disorder). A better explanation to psychiatric
disorders might be provided by this general factor, for which high
scores were associated with life impairment, greater familiarity,
worse developmental histories, and compromised early-life brain
function (Caspi et al., 2014).
In this framework, it is interesting to acknowledge that AN is
commonly related to both structural (Titova et al., 2013) and func-
tional (Gaudio et al., 2016) brain alterations. It is still unclear whether
brain alterations in AN are related to its aetiology or represent a non-
specific/residual scar of the disease. Some MRI studies on weight-
recovered patients showed a normalization of structural deficits
(Bernardoni et al., 2016). On the other hand, several MRI studies
highlighted slight group differences after AN recovery (Mainz et al.,
2012). However, a meta-analysis on AN structural studies (Titova
et al., 2013) and a systematic review on AN resting-state func-
tional-MRI studies (Gaudio et al., 2016) have pointed out specific
grey matter deficits in reward and somatosensory areas and altered
functional connectivity in regions involved in visuospatial and body-
signal integration and cognitive control in AN patients respectively.
Considering these AN brain alterations, a possible question is
whether they could be linked to a genetic substrate. To date, this field
of research is poorly explored (Walton et al., 2019). Two papers with a
limited sample size have been published on this topic showing that
COMT and 5-HTTLPR genotype might modulate functional con-
nectivity in patients with AN (Favaro et al., 2013;Collantoni et al.,
2016). A recent study (Walton et al., 2019) (with a large sample size
and combining different genetic methods) explored the relationship
between common genetic variants involved in the volumes of
subcortical brain areas and those associated with AN diagnosis.
These authors claimed “weak evidence”for relationships between
common genetic variants involved in AN and those associated with
the volumes of subcortical brain regions. On the other hand, they
found “suggestive evidence”for the effects of single genetic markers
on selected subcortical structures. These preliminary results suggest
a possible genetic influence on brain alterations and risk for AN and
warrant further studies, adopting GWAS and multimodal brain im-
aging studies, to disentangle the possible genetic role on brain
structure/function and risk for AN and to expand our knowledge of
AN pathophysiology and possible link with other psychiatric disor-
ders, commonly associated with AN.
Establishing etiologic factors to specific disorders can be chal-
lenging since isolating pathologies seems unlikely. Thinking of risk
dimensions might help to overcome this obstacle. The data obtained
in the GWAS would reinforce the hypothesis of a common inherited
vulnerability among psychiatric disorders observed by the over-
lapping risk genes (Fig. 1 A and 1B). Thus, the specificity of AN
phenotype would be given by the combination and interaction
C. de Jorge Martı´nez, G. Rukh, M.J. Williams et al. Journal of Genetics and Genomics 49 (2022) 1e12
9

between other risk genes affecting areas such as metabolism,
anthropometric and psychiatric traits, and the degree of expression
within the dimensional continuum. Within these dimensions, the
metabolic component acquires a relevance so far unknown in the
origin of AN. Just as obesity is increasingly considered to be both a
metabolic/endocrine and psychiatric disorder, approaching AN as
both a psychiatric and metabolic condition could ignite interest in
developing or repositioning pharmacologic agents for its treatment
where currently none exist (Duncan et al., 2017) considering that
prescribed drugs for AN proved relatively ineffective.
The genetic data provided by GWAS improves the understanding
of the biological basis of AN, which have uncontested value in
developing new pharmacologic therapies. The advances produced in
genetics through GWAS along with the application of pathways
analysis, open new horizons for the discovery of more specific tar-
gets that may help to improve pharmacological devices.
The findings that are beginning to emerge from these studies,
combined with other current genetic technologies, may help to
obtain a more comprehensive picture of the disorder. It is widely
agreed that in AN genetic research we should not place the spotlight
on a single target, due to the multiple factors that come into play in
the case of complex disorders. Other mechanisms and interactions
such as alterations of gene expression through mRNA examination
must be considered in this equation of ED, especially epigenetic
mechanisms. Due to the important contribution of socio-cultural
factors in AN along with the alteration of the organism physiology
due to extreme malnutrition, the study of epigenetics is strongly
recommended, linking the interaction between genes and environ-
ment, that gives place to the onset of the disorder. Epigenetics
provides gene regulatory information mainly categorized in three
groups, methylation, histone modification and non-coding RNAs
(Hu
¨bel et al., 2019b). Epigenome-wide association studies (EWAS)
could offer valuable information, but sample sizes should match
those used in recent GWAS resorting to currently available data-
bases. Special mention deserves new disciplines such as nutrige-
netics, nutrigenomics and gut microbiome. The genome is affected
by both macronutrients and micronutrients, and changes can take
place as a result of a deficient diet. But this relationship also works
the other way round, i.e., genetic variations can influence dietary
response and metabolism (Shih and Woodside, 2016). At the same
time, the study of gut microbiome is starting to gain attention in a
wide range of fields including eating disorders. These emerging
disciplines could contribute new insights that may converge on a
holistic scope of AN.
A different approach would be necessary in the future of psy-
chiatric research. Diagnostic categories in mental disorders, despite
simplify communication between professionals and are useful to
make prognosis and decide treatments, may lead to reductionism
and illusion of homogeneity between different conditions. Including
dimensional clusters highlighted by genetic data into research might
help to reflect more accurately the complexity/reality of mental dis-
orders. The intricacy of genetic study of AN requires a comprehen-
sive approach integrating the different techniques available to date,
making use of consortia and databases in order to identify causal
variants.
Conflict of interest
All authors declare no conflict of interest.
Acknowledgements
Gull Rukh was supported by the grant from Svenska S€
allskapet
f€
or Medicinsk Forskning (SSMF). Helgi Birgir Schi€
oth was supported
by the grant from Swedish Research Council (VR 2014-02812). The
funding bodies had no role in the design, preparation or approval of
the manuscript.
Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.jgg.2021.09.005.
References
Alemany-Navarro, M., Cruz, R., Real, E., Segal
as, C., Bertolı
´n, S., Rabionet, R.,
Carracedo,

A., Mench
on, J.M., Alonso, P., 2020. Looking into the genetic bases
of OCD dimensions: a pilot genome-wide association study. Transl. Psychiatry
10, 1e16.
American Psychiatric Association, 2000. Diagnostic and statistical manual of mental
disorders, fourth ed., text rev.https://doi.org/10.1176/appi.books.9780890423349.
American Psychiatric Association, 2013. Diagnostic and statistical manual of mental
disorders, fifth ed. https://doi.org/10.1176/appi.books.9780890425596.
Baker, J.H., Schaumberg, K., Munn-Chernoff, M.A., 2017. Genetics of anorexia
nervosa. Curr. Psychiatr. Rep. 19, 84.
Batista, M., Anti
c, L.

Z.,

Zaja, O., Jakovina, T., Begovac, I., 2018. Predictors of eating
disorder risk in anorexia nervosa adolescents. Acta. Clin. Croat. 57, 399e410.
Beadle, J.N., Paradiso, S., Salerno, A., McCormick, L.M., 2013. Alexithymia,
emotional empathy, and self-regulation in anorexia nervosa. Ann. Clin. Psychiatr.
25, 107e120.
Berends, T., van Meijel, B., Nugteren, W., Deen, M., Danner, U.N., Hoek, H.W., van
Elburg, A.A., 2016. Rate, timing and predictors of relapse in patients with anorexia
nervosa following a relapse prevention program: a cohort study. BMC Psychiatr.
16, 316.
Berkowitz, R.I., Fabricatore, A.N., 2005. Obesity, psychiatric status, and psychiatric
medications. Psychiatr. Clin. North. Am. 28, 39e54.
Bernardoni, F., King, J.A., Geisler, D., Stein, E., Jaite, C., N€
atsch, D., Tam, F.I.,
Boehm, I., Seidel, M., Roessner, V., et al., 2016. Weight restoration therapy
rapidly reverses cortical thinning in anorexia nervosa: A longitudinal study. Neu-
roimage 130, 214e222.
Błachno, M., Bry
nska, A., 2012. Comorbidity and characteristic of obsessive-
compulsive symptoms in anorexia nervosa. Psychiatr. Pol. 46, 1019e1028.
Boraska, V., Davis, O.S.P., Cherkas, L.F., Helder, S.G., Harris, J., Krug, I., Pei-Chi
Liao, T., Treasure, J., Ntalla, I., Karhunen, L., et al., 2012. Genome-wide asso-
ciation analysis of eating disorder-related symptoms, behaviors, and personality
traits. Am. J. Med. Genet. Part B Neuropsychiatr. Genet. 159, 803e811.
Boraska, V., Franklin, C.S., Floyd, J.A.B., Thornton, L.M., Huckins, L.M., Southam, L.,
Rayner, N.W., Tachmazidou, I., Klump, K.L., Treasure, J., et al., 2014. A genome-
wide association study of anorexia nervosa. Mol. Psychiatr. 19, 1085e1094.
Breithaupt, L., Hubel, C., Bulik, C.M., 2018. Updates on genome-wide association
findings in eating disorders and future application to precision medicine. Curr.
Neuropharmacol. 16, 1102e1110.
Brooks, S.J., Schi€
oth, H., 2019. Impulsivity and compulsivity in anorexia nervosa:
cognitive systems underlying variation in appetite restraint from an RDoC
perspective. In: Anorexia and Bulimia Nervosa. IntechOpen. https://doi.org/
10.5772/intechopen.83702.
Brooks, S.J., Rask-Andersen, M., Benedict, C., Schi€
oth, H.B., 2012. A debate on
current eating disorder diagnoses in light of neurobiological findings: is it time
for a spectrum model? BMC Psychiatr. 12, 76.
Brooks, S.J., Lochner, C., Shoptaw, S., Stein, D.J., 2017. Using the research domain
criteria (RDoC) to conceptualize impulsivity and compulsivity in relation to
addiction. Prog. Brain Res. 235, 177e218.
Bulik, C.M., Sullivan, P.F., Tozzi, F., Furberg, H., Lichtenstein, P., Pedersen, N.L.,
2006. Prevalence, heritability, and prospective risk factors for anorexia nervosa.
Arch. Gen. Psychiatr. 63, 305e312.
Bulik, C., Yilmaz, Z., HArdaway, A., 2015. Genetics and epigenetics of eating disor-
ders. Adv. Genom. Genet. 5, 131e150.
Bulik-Sullivan, B., Finucane, H.K., Anttila, V., Gusev, A., Day, F.R., Loh, P.R.,
Duncan, L., Perry, J.R.B., Patterson, N., Robinson, E.B., et al., 2015. An atlas of
genetic correlations across human diseases and traits. Nat. Genet. 47,
1236e1241.
Carbon, S., Douglass, E., Dunn, N., Good, B., Harris, N.L., Lewis, S.E., Mungall, C.J.,
Basu, S., Chisholm, R.L., Dodson, R.J., et al., 2019. The gene ontology resource:
20 years and still GOing strong. Nucleic Acids Res. 47, D330eD338.
Caspi, A., Houts, R.M., Belsky, D.W., Goldman-Mellor, S.J., Harrington, H., Israel, S.,
Meier, M.H., Ramrakha, S., Shalev, I., Poulton, R., et al., 2014. The p factor: one
general psychopathology factor in the structure of psychiatric disorders? Clin.
Psychol. Sci. 2, 119e137.
Collantoni, E., Michelon, S., Tenconi, E., Degortes, D., Titton, F., Manara, R.,
Clementi, M., Pinato, C., Forzan, M., Cassina, M., et al., 2016. Functional con-
nectivity correlates of response inhibition impairment in anorexia nervosa. Psy-
chiatry Res. Neuroimaging. 247, 9e16.
Collier, D.A., Arranz, M.J., Li, T., Mupita, D., Brown, N., Treasure, J., 1997. Associ-
ation between 5-HT(2A) gene promoter polymorphism and anorexia nervosa.
Lancet 350, 412.
Curzio, O., Maestro, S., Rossi, G., Calderoni, S., Giombini, L., Scardigli, S.,
Ragione, L.D., Muratori, F., 2018. Transdiagnostic vs. disorder-focused
C. de Jorge Martı´nez, G. Rukh, M.J. Williams et al. Journal of Genetics and Genomics 49 (2022) 1e12
10

perspective in children and adolescents with eating disorders: findings from a
large multisite exploratory study. Eur. Psychiatr. 49, 81e93.
Devlin, B., 2002. Linkage analysis of anorexia nervosa incorporating behavioral
covariates. Hum. Mol. Genet. 11, 689e696.
Docherty, A.R., Moscati, A.A., Fanous, A.H., 2016. Cross-disorder psychiatric ge-
nomics. Curr. Behav. Neurosci. Rep. 3, 256e263.
Duncan, L., Yilmaz, Z., Gaspar, Helena, Walters, R., Goldstein, J., Anttila, V., Bulik-
Sullivan, B., Ripke, S., Thornton, L., Hinney, A., et al., 2017. Significant locus and
metabolic genetic correlations revealed in genome-wide association study of
anorexia nervosa. Am. J. Psychiatr. 174, 850e858.
Fairburn, C.G., Cooper, Z., Shafran, R., 2003. Cognitive behaviour therapy for eating
disorders: a transdiagnostic theory and treatment. Behav. Res. Ther. 41,
509e528.
Favaro, A., Clementi, M., Manara, R., Bosello, R., Forzan, M., Bruson, A., Tenconi, E.,
Degortes, D., Titton, F., Di Salle, F., et al., 2013. Catechol-O-methyltransferase
genotype modifies executive functioning and prefrontal functional connectivity in
women with anorexia nervosa. J. Psychiatry Neurosci. 38, 241e248.
Fu, W., O’Connor, T.D., Akey, J.M., 2013. Genetic architecture of quantitative traits
and complex diseases. Curr. Opin. Genet. Dev. 23, 678e683.
Fusar-Poli, P., Solmi, M., Brondino, N., Davies, C., Chae, C., Politi, P., Borgwardt, S.,
Lawrie, S.M., Parnas, J., McGuire, P., 2019. Transdiagnostic psychiatry: a sys-
tematic review. World Psychiatry 18, 192e207.
Galmiche, M., D
echelotte, P., Lambert, G., Tavolacci, M.P., 2019. Prevalence of
eating disorders over the 2000e2018 period: a systematic literature review. Am.
J. Clin. Nutr. 109, 1402e1413.
Gaudio, S., Wiemerslage, L., Brooks, S.J., Schi€
oth, H.B., 2016. A systematic review of
resting-state functional-MRI studies in anorexia nervosa : evidence for functional
connectivity impairment in cognitive control and visuospatial and body-signal
integration. Neurosci. Biobehav. Rev. 71, 578e589.
Giacomini, K.M., Yee, S.W., Mushiroda, T., Weinshilboum, R.M., Ratain, M.J.,
Kubo, M., 2016. Genome-wide association studies of drug response and toxicity:
an opportunity for genome medicine. Nat. Rev. Drug Discov. 16, 1.
Gonz
alez, L.M., Garcı
´a-Herr
aiz, A., Mota-Zamorano, S., Flores, I., Albuquerque, D.,
Gervasini, G., 2021. Variants in the Obesity-Linked FTO gene locus modulates
psychopathological features of patients with Anorexia Nervosa. Gene 783, 145572.
Gorwood, P., Bouvard, M., Mouren-Sim
eoni, M.C., Kipman, A., Ad
es, J., 1998. Ge-
netics and anorexia nervosa: a review of candidate genes. Psychiatr. Genet. 8,
1e12.
Grice, D.E., Halmi, K.A., Fichter, M.M., Strober, M., Woodside, D.B., Treasure, J.T.,
Kaplan, A.S., Magistretti, P.J., Goldman, D., Bulik, C.M., et al., 2002. Evidence for
a susceptibility gene for anorexia nervosa on chromosome 1. Am. J. Hum. Genet.
70, 787e792.
Hambrook, D., Tchanturia, K., Schmidt, U., Russell, T., Treasure, J., 2010.
Empathy, systemizing, and autistic traits in anorexia nervosa: a pilot study, 47.
335, 339.
Himmerich, H., Bentley, J., Kan, C., Treasure, J., 2019. Genetic risk factors for eating
disorders: an update and insights into pathophysiology. Ther. Adv. Psycho-
pharmacol. 9, 2045125318814734.
Hinney, A., Kesselmeier, M., Jall, S., Volckmar, A.L., F€
ocker, M., Antel, J., Heid, I.M.,
Winkler, T.W., Grant, S.F.A., Guo, Y., et al., 2017. Evidence for three genetic loci
involved in both anorexia nervosa risk and variation of body mass index. Mol.
Psychiatr. 22, 192e201.
Hinney, A., Volckmar, A.L., 2013. Genetics of eating disorders. Curr. Psychiatr. Rep.
15, 423.
Howard, M., Gregertsen, E.C., Hindocha, C., Serpell, L., 2020. Impulsivity and
compulsivity in anorexia and bulimia nervosa: a systematic review. Psychiatr.
Res. 293, 113354.
Hu
¨bel, C., Gaspar, H.A., Coleman, J.R.I., Finucane, H., Purves, K.L.,
Hanscombe, K.B., Prokopenko, I., Graff, M., Ngwa, J.S., Workalemahu, T., et al.,
2019a. Genomics of body fat percentage may contribute to sex bias in anorexia
nervosa. Am. J. Med. Genet. Part B Neuropsychiatr. Genet. 180, 428e438.
Hu
¨bel, C., Marzi, S.J., Breen, G., Bulik, C.M., 2019b. Epigenetics in eating disor-
ders: a systematic review. Mol. Psychiatr. 24, 901e915.
Huckins, L.M., Hatzikotoulas, K., Southam, L., Thornton, L.M., Steinberg, J., Aguilera-
Mckay, F., Treasure, J., Schmidt, U., Gunasinghe, C., Romero, A., et al., 2018.
Investigation of common, low-frequency and rare genome-wide variation in
anorexia nervosa. Mol. Psychiatr. 23, 1169e1180.
Hudson, J.I., Hiripi, E., Pope, H.G., Kessler, R.C., 2007. The prevalence and corre-
lates of eating disorders in the national comorbidity survey replication. Biol.
Psychiatr. 61, 348e358.
Hung, J.-H., 2013. Gene set/pathway enrichment analysis. Methods Mol. Biol. 939,
201e213.
Jagielska, G., Kacperska, I., 2017. Outcome, comorbidity and prognosis in anorexia
nervosa. Psychiatr. Pol. 51, 205e218.
Kalm, L.M., Semba, R.D., 2005. They starved so that others be better fed: remem-
bering Ancel keys and the Minnesota experiment. J. Nutr. 135, 1347e1352.
Kerr-Gaffney, J., Harrison, A., Tchanturia, K., 2019. Cognitive and affective empathy in
eating disorders: a systematic review and meta-analysis. Front. Psychiatr. 10, 102.
Khera, A.V., Chaffin, M., Aragam, K.G., Haas, M.E., Roselli, C., Choi, S.H.,
Natarajan, P., Lander, E.S., Lubitz, S.A., Ellinor, P.T., et al., 2018. Genome-wide
polygenic scores for common diseases identify individuals with risk equivalent to
monogenic mutations. Nat. Genet. 50, 1219e1224.
Klump, K.L., Gobrogge, K.L., 2005. A review and primer of molecular genetic studies
of anorexia nervosa. Int. J. Eat. Disord. 37, S43eS48.
Lander, E.S., Schork, N.J., 1994. Genetic dissection of complex traits. Science 265,
2037e2048.
Lavender, J.M., Goodman, E.L., Culbert, K.M., Wonderlich, S.A., Crosby, R.D.,
Engel, S.G., Mitchell, J.E., Le Grange, D., Crow, S.J., Peterson, C.B., 2017.
Facets of impulsivity and compulsivity in women with anorexia nervosa. Eur. Eat
Disord. Rev. 25, 309e313.
Li, D., Chang, X., Connolly, J.J., Tian, L., Liu, Y., Bhoj, E.J., Robinson, N., Abrams, D.,
Li, Y.R., Bradfield, J.P., et al., 2017. A genome-wide association study of anorexia
nervosa suggests a risk locus implicated in dysregulated leptin signaling. Sci.
Rep. 7, 3847.
Liu, X., Kelsoe, J.R., Greenwood, T.A., Kelsoe, J.R., Green-Wood, T.A.,
Nievergelt, C.M., McKinney, R., Shilling, P.D., Smith, E.N., Schork, N.J., et al.,
2016. A genome-wide association study of bipolar disorder with comorbid eating
disorder replicates the SOX2-OT region. J. Affect. Disord. 189, 141e149.
London, E.D., 2020. Human brain imaging links dopaminergic systems to impulsivity.
Curr. Top Behav. Neurosci. 47, 53e71.
Mainz, V., Schulte-Ru
¨ther, M., Fink, G.R., Herpertz-Dahlmann, B., Konrad, K., 2012.
Structural brain abnormalities in adolescent anorexia nervosa before and after
weight recovery and associated hormonal changes. Psychosom. Med. 74,
574e582.
Maj, M., 2005. “Psychiatric comorbidity”: an artefact of current diagnostic systems?
Br. J. Psychiatry 186, 182e184.
Manchia, M., Cullis, J., Turecki, G., Rouleau, G.A., Uher, R., Alda, M., 2013. The
impact of phenotypic and genetic heterogeneity on results of genome wide as-
sociation studies of complex diseases. PLoS One 8, e76295.
Marucci, S., Ragione, L.D., De Iaco, G., Mococci, T., Vicini, M., Guastamacchia, E.,
Triggiani, V., 2018. Anorexia nervosa and comorbid psychopathology. Endocr.
Metab. Immune Disord. Drug Targets 18, 316e324.
Migliore, S., Jankovic, J., Squitieri, F., 2019. Genetic counseling in Huntington’s
disease: potential new challenges on horizon? Front. Neurol. 10, 453.
Moskowitz, L., Weiselberg, E., 2017. Anorexia nervosa/atypical anorexia nervosa.
Curr. Probl. Pediatr. Adolesc. Health Care 47, 70e84.
Nakabayashi, K., Komaki, G., Tajima, A., Ando, T., Ishikawa, M., Nomoto, J., Hata, K.,
Oka, A., Inoko, H., Sasazuki, T., et al., 2009. Identification of novel candidate loci
for anorexia nervosa at 1q41 and 11q22 in Japanese by a genome-wide asso-
ciation analysis with microsatellite markers. J. Hum. Genet. 54, 531e537.
Øiesvold, T., Nivison, M., Hansen, V., Skre, I., Østensen, L., Sørgaard, K.W., 2013.
Diagnosing comorbidity in psychiatric hospital: challenging the validity of
administrative registers. BMC Psychiatr. 13, 13.
Peterson, K., Fuller, R., 2019. Anorexia nervosa in adolescents. Nursing 49, 24e30.
Pinheiro, A.P., Root, T., Bulik, C.M., 2009. The genetics of anorexia nervosa: current
findings and future perspectives. Int. J. Child Adolesc. Health 2, 153e164.
Rask-Andersen, M., Olszewski, P.K., Levine, A.S., Schi€
oth, H.B., 2010. Molecular
mechanisms underlying anorexia nervosa: focus on human gene association
studies and systems controlling food intake. Brain Res. Rev. 62, 147e164.
Reimand, J., Isserlin, R., Voisin, V., Kucera, M., Tannus-Lopes, C., Rostamianfar, A.,
Wadi, L., Meyer, M., Wong, J., Xu, C., et al., 2019. Pathway enrichment analysis
and visualization of omics data using g:Profiler, GSEA, Cytoscape and Enrich-
mentMap. Nat. Protoc. 14, 482e517.
Reville, M.C., O’Connor, L., Frampton, I., 2016. Literature review of cognitive
neuroscience and anorexia nervosa. Curr. Psychiatr. Rep. 18, 18.
Risch, N., Merikangas, K., 1996. The future of genetic studies of complex human
diseases. Science 273, 1516e1517.
Sawyer, S.M., Whitelaw, M., Le Grange, D., Yeo, M., Hughes, E.K., 2016. Physical
and psychological morbidity in adolescents with atypical anorexia nervosa. Pe-
diatrics 137, e20154080.
Schalla, M.A., Stengel, A., 2019. Activity based anorexia as an animal model for
anorexia nervosaea systematic review. Front. Nutr. 6, 69.
Shih, P.B., Woodside, D.B., 2016. Contemporary views on the genetics of anorexia
nervosa. Eur. Neuropsychopharmacol. 26, 663e673.
Smink, F.R.E., van Hoeken, D., Hoek, H.W., 2012. Epidemiology of eating disorders:
incidence, prevalence and mortality rates. Curr. Psychiatr. Rep. 14, 406e414.
S€
odersten, P., Brodin, U., Zandian, M., Bergh, C.E.K., 2019. Verifying Feighner’s
hypothesis; anorexia nervosa is not a psychiatric disorder. Front. Psychol. 10,
16.
Solmi, M., Gallicchio, D., Collantoni, E., Correll, C.U., Clementi, M., Pinato, C.,
Forzan, M., Cassina, M., Fontana, F., Giannunzio, V., et al., 2016. Serotonin
transporter gene polymorphism in eating disorders: data from a new biobank and
META-analysis of previous studies. World J. Biol. Psychiatr. 17, 244e257.
Steinglass, J.E., Walsh, B.T., 2016. Neurobiological model of the persistence of
anorexia nervosa. J. Eat. Disord. 4, 19.
Steinhausen, H.C., Jakobsen, H., Helenius, D., Munk-Jørgensen, P., Strober, M.,
2015. A nation-wide study of the family aggregation and risk factors in anorexia
nervosa over three generations. Int. J. Eat. Disord. 48, 1e8.
Strober, M., Freeman, R., Lampert, C., Diamond, J., Kaye, W., 2000. Controlled family
study of anorexia nervosa and bulimia nervosa: evidence of shared liability and
transmission of partial syndromes. Am. J. Psychiatr. 157, 393e401.
Su, P., Zhang, H., Wong, A.H.C., Liu, F., 2020. The DISC1 R264Q variant increases
affinity for the dopamine D2 receptor and increases GSK3 activity. Mol. Brain
13, 87.
Sudlow, C., Gallacher, J., Allen, N., Beral, V., Burton, P., Danesh, J., Downey, P.,
Elliott, P., Green, J., Landray, M., et al., 2015. UK biobank: an open access
resource for identifying the causes of a wide range of complex diseases of middle
and old age. PLoS Med. 12, e1001779.
C. de Jorge Martı´nez, G. Rukh, M.J. Williams et al. Journal of Genetics and Genomics 49 (2022) 1e12
11

Titova, O.E., Hjorth, O.C., Schi€
oth, H.B., Brooks, S.J., 2013. Anorexia nervosa is linked
to reduced brain structure in reward and somatosensory regions: a meta-analysis
of VBM studies. BMC Psychiatr. 13, 110.
van Noort, B.M., Lohmar, S.K., Pfeiffer, E., Lehmkuhl, U., Winter, S.M., Kappel, V., 2018.
Clinical characteristics of early onset anorexia nervosa. Eur. Eat Disord. Rev. 26, 519e525.
Wade, T.D., Gordon, S., Medland, S., Bulik, C.M., Heath, A.C., Montgomery, G.W.,
Martin, N.G., 2013. Genetic variants associated with disordered eating. Int. J. Eat.
Disord. 46, 594e608.
Wade,K.H., Clifford,L., Simpkin, A.J.,Beynon, R., Birch,L., Northstone,K., Matthews,S.,
Davey Smith, G., Hamilton-Shield, J., Timpson, N.J., 2020. Piloting the objective
measurement of eating behaviour at a population scale: a nested study within the
Avon Longitudinal Study of Parents and Children. Wellcome Open Res. 5, 185.
Walton, E., Hibar, D., Yilmaz, Z., Jahanshad, N., Cheung, J., Batury, V.L., Seitz, J.,
Bulik, C.M., Thompson, P.M., Ehrlich, S., 2019. Exploration of shared genetic
architecture between subcortical brain volumes and anorexia nervosa. Mol.
Neurobiol. 56, 5146e5156.
Wang, K., Zhang, H., Bloss, C.S., Duvvuri, V., Kaye, W., Schork, N.J., Berrettini, W.,
Hakonarson, H., 2011. A genome-wide association study on common SNPs and
rare CNVs in anorexia nervosa. Mol. Psychiatr. 16, 949e959.
Warrier, V., Toro, R., Chakrabarti, B., Børglum, A.D., Grove, J., Agee, M.,
Alipanahi, B., Auton, A., Bell, R.K., Bryc, K., et al., 2018. Genome-wide analyses
of self-reported empathy: correlations with autism, schizophrenia, and anorexia
nervosa. Transl. Psychiatry 8, 35.
Watson, H.J., Yilmaz, Z., Thornton, L.M., Hu
¨bel, C., Coleman, J.R.I., Gaspar, H.A.,
Bryois, J., Hinney, A., Lepp€
a, V.M., Mattheisen, M., et al., 2019. Genome-wide
association study identifies eight risk loci and implicates metabo-psychiatric or-
igins for anorexia nervosa. Nat. Genet. 51, 1207e1214.
Widdowson, E.M., 1952. Biology of human starvation. Nature 170, 177.
World Health Organization, 2004. ICD-10 : international statistical classification of
diseases and related health problems, tenth revision, 2nd ed. https://apps.who.
int/iris/handle/10665/42980.
Yilmaz, Z., Halvorsen, M., Bryois, J., Yu, D., Thornton, L.M., Zerwas, S., Micali, N.,
Moessner, R., Burton, C.L., Zai, G., et al., 2018. Examination of the shared ge-
netic basis of anorexia nervosa and obsessive-compulsive disorder. Mol. Psy-
chiatr. 25, 2036e2046.
Zabena, C., Gonz
alez-S
anchez, J.L., Martı
´nez-Larrad, M.T., Torres-Garcı
´a, A.,
Alvarez-Fern
andez-Represa, J., Corbat
on-Anchuelo, A., P
erez-Barba, M.,
Serrano-Rı
´os, M., 2009. The FTO obesity gene. Genotyping and gene expression
analysis in morbidly obese patients. Obes. Surg. 19, 87e95.
Zheutlin, A.B., Ross, D.A., 2018. Polygenic risk scores: what are they good for? Biol.
Psychiatr. 83, e51ee53.
C. de Jorge Martı´nez, G. Rukh, M.J. Williams et al. Journal of Genetics and Genomics 49 (2022) 1e12
12