Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Anorexia nervosa From single SNP studies, through biomarkers, to genome-wide association
Marek K. Brandys
ISBN: 978-94-6182-640-4
Printed by: Offpage, Amsterdam
Layout: Marek K. Brandys
Cover design: Marek K. Brandys, based on Effect of Butterfly by Anastasiya
Markovich (Picture Labberté K.J.) via Wikimedia Commons
© Marek K. Brandys
Anorexia nervosa
From single SNP studies, through biomarkers, to genome-wide association
Anorexia nervosa
Van SNP studies via biomarkers naar genoomwijde associatie
(met een samenvatting in het Nederlands)
Anorexia nervosa
Od badań polimorfizmów pojednynczego nukleotydu, przez biomarkery, po
badania asocjacyjne całego genomu
(ze streszczeniem w języku polskim)
Proefschrift
ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag
van de rector magnificus, prof.dr. G.J. van der Zwaan, ingevolge het besluit
van het college voor promoties in het openbaar te verdedigen op
dinsdag 19 januari 2016 des ochtends te 10.30 uur
door
Marek Kajetan Brandys
geboren op 27 november 1983 te Kraków, Polska (Polen)
Promotoren: Prof. dr. R.A.H. Adan
Prof. dr. A. van Elburg
Copromotoren: Dr. M.J.H. Kas
Dr. C. de Kovel
This thesis was partly accomplished with financial support from the Marie
Curie Research Training Network INTACT (Individually tailored stepped care
for women with eating disorders; reference number: MRTN-CT-2006-
035988)
Table of contents CHAPTER 1 ................................................................................................................................... 6
Introduction Scope and outline of the thesis
CHAPTER 2 ................................................................................................................................. 35
Are recently identified genetic variants regulating BMI in the general population associated with anorexia nervosa?
CHAPTER 3 ................................................................................................................................. 46
Association study of POMC variants with body composition measures and nutrient choice CHAPTER 4 ................................................................................................................................. 69
Anorexia nervosa and the Val158Met polymorphism of the COMT gene: meta-analysis and new data
CHAPTER 5 ................................................................................................................................. 90
A meta-analysis of circulating BDNF concentrations in anorexia nervosa CHAPTER 6 ............................................................................................................................... 129
The Val66Met polymorphism of the BDNF gene in anorexia nervosa: new data and a meta-analysis
CHAPTER 7 ............................................................................................................................... 164
No evidence for involvement of CNVs associated with neurodevelopmental disorders in anorexia nervosa
APPENDIX ................................................................................................................................ 193
A genome-wide association study of anorexia nervosa CHAPTER 8 Discussion and conclusions................................................................................... 226
Overview of genetic research in anorexia nervosa: the past, the present and the future Concluding remarks
ADDENDUM ............................................................................................................................ 264
English summary Nederlandse samenvatting Streszczenie w języku polskim Curriculum Vitae List of publications Acknowledgements
Chapter 1
6
Chapter 1
Introduction
The main focus of the present thesis is to describe the scientific undertaking
of exploring the genetic underpinnings and biomarkers of anorexia nervosa
(AN). We begin by introducing the history, intricate phenotypic
manifestations, as well as the clinical and epidemiological characteristics of
this intriguing disease.
According to DSM-5 AN belongs to the category of feeding and
eating disorders (ED), under the code 307.1 (F50.01) for AN restricting type
and (F50.02) for the binge-eating/purging type. Other classes in this category
include bulimia nervosa (BN; 307.51 (F50.2)), binge eating disorder (BED;
307.51 (F50.8)), other specified feeding or eating disorder (OSFED; 307.59
(F50.8)) and unspecified feeding or eating disorder (307.50 (F50.9)).
Diagnostic criteria of AN, according to DSM-5, are listed in a later section.
History
It was a British royal physician, Sir William Gull, who in 1873 established the
term ‘anorexia’ (derived from Greek ‘an-’, meaning negation, and ‘orexis’,
signifying appetite) 1. The first medical descriptions of cases with AN are
dated earlier than that, and ascribed to Richard Morton, also a British
physician. Looking even further back, there exist historical accounts of
people who appear to have suffered from this disorder. In the ancient
Hellenistic culture fasting and self-starvation were seen as expressions of
religious zealousness. While only a few reports are available from the
medieval ages, a larger number of descriptions of the possible cases of AN
comes from the times of the Renaissance. Religious ascetics would forge
their way to sanity via starvation, self-mutilation and social isolation 2. A
Chapter 1
7
number of historical figures are suspected to have suffered from AN, such as
Saint Catherine of Siena, Mary, Queen of Scotts or Elisabeth, Empress of
Austria (source: http://divainternational.ch/spip.php?article97). In the
modern times, a general interest in AN surged after the death of a famous
musician, Karen Carpenter (4 February 1983).
Somatic health risks
The most striking feature of the patients suffering from AN is their low body-
weight (85% or less than the weight expected). This symptom is
accompanied by a refusal to consume sufficient amount of calories – an
amount necessary to prevent further emaciation and restoration of the body
weight to the normative levels. This continuous undernourishment damages
body systems and, in extreme cases, leads to death. Serious medical
complications associated with malnutrition in AN include:
� Reduced heart rate and low blood pressure, entailing increased risk
of heart failure
� Amenorrhea in post-pubertal females (lack of menstruation)
� Osteoporosis (decreased bone density)
� Loss of muscles
� Dehydration, possibly leading to kidney failure
� Fainting, fatigue, general weakness
� Hair loss, changes of complexion, growth of lanugo – a thin hair layer
covering the body
Furthermore, health risks associated with the purging behaviors present in
the purging subtype of AN include:
� Electrolyte imbalance (caused by dehydration, loss of potassium,
chloride and sodium), possibly leading to a heart failure
� Inflammation and possible rupture of esophagus (as a result of
vomiting)
� Tooth decay
� Constipation and chronic irregular bowel movements, coming from
abuse of laxatives (source:
Chapter 1
8
http://www.nationaleatingdisorders.org/nedaDir/files/documents/h
andouts/HlthCons.pdf)
Criteria and symptoms
These serious somatic complications are paralleled by the devastation which
the disease incurs to the psyche. Suicide is the most frequent cause of death
in EDs 3. There is a 57-fold increase in risk of death from suicide among
patients with AN, compared to the age-matched cohort 4. Individuals with AN
are characterized by the immense fear of weight (fat) gain and disturbed
body image. The criteria for diagnosis of AN established in the 5-th edition of
The Diagnostic and Statistical Manual of Mental Disorders (DSM-5) are
presented below:
A. Restriction of energy intake relative to requirements leading to a
significantly low body weight in the context of age, sex, developmental
trajectory, and physical health.
B. Intense fear of gaining weight or becoming fat, even though underweight.
C. Disturbance in the way in which one's body weight or shape is
experienced, undue influence of body weight or shape on self-evaluation, or
denial of the seriousness of the current low body weight.
Symptoms which might suggest AN (warning signs) include (after
http://www.allianceforeatingdisorders.com/):
• Significant weight loss
• Distorted body image
• Intense fear/anxiety about gaining weight
• Preoccupation with weight, calories, food, etc.
• Feelings of guilt after eating
• Denial of low weight
• High levels of anxiety and/or depression
Chapter 1
9
• Low self-esteem
• Self-injury
• Withdrawal from friends and activities
• Excuses for not eating/denial of hunger
• Food rituals
• Intense, dramatic mood swings
• Pale appearance/yellowish skin-tone
• Thin, dull, and dry hair, skin, and nails
• Cold intolerance/hypothermia
• Fatigue/fainting
• Abuse of laxatives, diet pills, or diuretics
• Excessive and compulsive exercise
Subtypes and diagnostic cross-over
There are two subtypes within the category of AN: the restricting (restrictive)
type (ANR) and the bingeing-purging type (ANBP). The main difference is that
the individuals with the latter one experience periods of binge eating
(consumption of excessive amounts of food coupled with a subjective feeling
of loss of control over eating) followed by purging behaviors, which are
means to compensate for the calories consumed. Purging can take forms of
self- induced vomiting or/and use of laxatives, diuretics or diet pills. Patients
exhibiting the restricting type do not purge, but maintain low body weight
solely by reduced food intake and increased energy expenditure via
exercising and hyperactivity.
In the clinical reality, the observed phenotypes are often not clear-
cut. This is exemplified by the fact that EDNOS used to be the most
frequently assigned diagnosis among EDs (from 40 to 60% of all intakes in ED
units, 5 and about 75% of cases of EDs detected among adolescent female
population 6), according to DSM-IV criteria. Furthermore, although the
category of EDs is relatively stable, moving across the diagnoses within this
category is quite common 7. Cases with AN often turn into BN or EDNOS,
whereas cases with BN evolve into EDNOS and, much less frequently, into AN
Chapter 1
10
8. Moving between AN subtypes is also frequent – over 7 years, nearly 50% of
women diagnosed with AN crossed over from one subtype into another and
34% evolved from AN into BN (with a high chance of relapse into AN) 9.
Agras et al. observed that 80% of patients with EDNOS diagnosis had
a lifetime history of AN, BN or BED. Additional 10% developed AN, BN or BED
during a 4-year follow-up. In this study, EDNOS was a way station between a
fully-blown ED and recovery 10.
In the most recent, 5-th edition of the DSM the criteria for diagnosis
of AN became more inclusive. The amenorrhea criterion has been removed
and criterion A became more general. These changes resulted in a better
classification of patients with AN and reduction of the vague EDNOS
category. A study of 309 patients with ED found that almost all of the 60% of
patients with EDNOS according to DSM-IV were re-assigned to the specific
diagnoses within the DSM-5 framework 11.
Empirical classifications
The outward symptoms ground the division of AN into the ANR and ANBP
subtypes. There is a growing body of literature, however, indicating that this
classification has limited usefulness for aethiological research and treatment
improvement 12. Most of the patients with ANR are going to develop binge-
purge behavior at some point, suggesting that these subtypes may in fact
represent alternate stages of the same condition, rather than the subtypes 13. Finally, the differences in treatment utilization, relapse and mortality rates
are very slight if any 13,14. The observations of the limited usefulness of the
clinically-derived subphenotypes drove a number of studies which applied
formal statistical procedures to estabilish the experimental classification of
eating disorders. Traditional approaches, such as the cluster analysis or the
taxometric analysis are currently being replaced by the latent class or latent
profile analyses (LPA) 15. In short, LPA employs a maximum likelihood
estimation to assign participants to mutually exclusive (unobserved) latent
classes. Classes are inferred by the pattern of inter-correlations between
indicators (the variables used to infer the classes, e.g. personality traits). It
Chapter 1
11
uses general probability model which allows for inequality of variances in
groups and enables determination of the optimal number of classes via
formal statistical procedures 15. These analyses were performed in a number
of studies on eating disorders (see review 16), and although the particular
results will depend on the selected indicators and parameters’ set-up 17, the
picture which emerged from these studies is quite consistent 18. In general,
the empirical approaches to classification result in the division of the
patients with eating disorders into three distinct classes 16,19:
1. the over-regulated and over-controlled class, characterized by constraint
and inhibition,
2. the under-regulated and disinhibited subtype, with impulsivity and
dysregulated emotional functioning,
3. low psychopathology group, characterized by normative levels of
personality functioning and perfectionism.
The relevance of these classes has been confirmed in several studies.
For instance, Wildes et al. 18 have shown that the empirically derived
classification of patients with AN proves useful clinically. The subtypes
differed in terms of multiple baseline characteristics, initial response to
treatment, readmission rates and outcome at discharge (the undercontrolled
patients had worse outcome than the overcontrolled (OR=3.56, p=.01), who,
in turn, were worse than the low psychopathology class (OR=11.23, p
Chapter 1
12
on a strict definition of AN (according to DSM-IV), and they increase
substantially when some of the criteria are relaxed 25. On the whole,
although EDs are rare in the general population, they are quite common in
young females 26. The incidence of AN has been increasing in the past
century until the 1970s and remained stable henceforth 27.
AN occurs more often in women (men are affected in only 5% to 10%
of cases) 28.
AN is notorious for having the highest standardized mortality ratio
among psychiatric illnesses (mortality rate being 5 to 10 times higher than in
a reference population 29,30. Stratification of patients according to body mass
index (height in cm divided by weight in kg squared; BMI) or age of onset
shows that SMR is highest in a group of lower BMI and a group of onset later
than 17 years of age 31. 20% of individuals with AN who died had committed
suicide 32.
Risk factors
Although the list of putative risk factors for AN is long 33, the studies
examining them are most frequently of a cross-sectional design and hindered
by the low frequency of AN in the general population. The focus on the
psychosocial factors, rather than on the biological ones, results from the fact
that the former are better established in the field and are more easily
measurable. A last decade has observed a surge in the number of studies of
genetic risk factors. These, however, will be discussed in the other sections.
A study with a longitudinal design is preferable when it comes to
establishing or verifying putative risk factors. In a prospective study on a
birth cohort, Nicholls et al. (2009) tested 22 childhood risk factors proposed
in the literature and found that only six were independently associated with
development of AN at older age. As expected, female sex turned out to be
the most potent risk factor (OR=22.1), followed by history of undereating
(OR 2.7), infant feeding problems (OR=2.6), and maternal depressive
symptoms (OR 1.8). Conversely, higher self-esteem and higher maternal BMI
were found to be protective (OR=0.3 and 0.91, respectively)34. Other studies
Chapter 1
13
add childhood sleeping problems, excessive physical exercise, anxious
parenting and perfectionism to this list 35. Another longitudinal study of risk
factors in EDs investigated 88 putative factors in a high risk group and found
that 7 were independently associated with a chance of developing an ED
(critical comments about eating from teacher/coach/siblings and a history of
depression had strongest effects on ED risk) 33. Distinguishing a causative risk
factor from proxies remains a problematic issue in all studies.
There is some evidence supporting the effect of the season of birth
on the risk of developing AN, although the effect sizes are small 36. An excess
of patients with AN was found among those born in spring (March to June;
OR=1.15) 37. The mechanisms underlying the association are not clear.
Interestingly, in utero exposure to virus infections (higher incidence of
chickenpox and rubella infections) was also related to AN risk (OR=1.6 and
1.5, respectively) 38.
Some studies suggested that in utero exposure to male or female
steroids may alter the risk of disordered eating in the future 39, but others
could not replicate this finding 40.
Comorbidity
87% of patients with EDs 31 had some kind of lifetime psychiatric
comorbidity, such as (in order of frequency) depressive disorders, anxiety
disorders, suicide attempts, substance abuse disorders, personality disorders
(predominantly the borderline personality disorder), obsessive compulsive
disorders and others. Suicide attempts were more frequent among ANBP
(34%) than in ANR (20%). In about 20% of patients with AN, developmental
disorders (autistic spectrum disorder, attention deficit-hyperactivity
disorder) are also observed 41. From the range of somatic disorders, which
can be comorbid with AN, diabetes mellitus, thyroid disorders and renal
calculus are seen most often. Psychiatric and somatic comorbidities are
negatively associated with the outcome of AN 30.
Chapter 1
14
Treatment and outcome
AN is a disease of a serious social significance. It often runs a chronic course
and mainly affects young people 42. Its treatment is expensive 43. Therefore,
AN generates substantial direct and indirect costs (for example, the cost of
an inpatient treatment for AN in Germany was estimated to be 4647 EUR per
case 44).
Studies of efficacy and effectiveness of treatment for AN offer only a
moderate or low level of evidence 28. Psychotherapeutical approaches which
were studied in the context of AN include cognitive-behavioral therapy (CBT),
interpersonal therapy, dialectical behavior therapy, psychodynamic therapy,
family therapy, adolescent-focused therapy and several others. CBT is the
most often recommended modality of treatment (with specific, disease-
tailored approaches preferred over non-specific approaches), although the
evidence for superiority of any particular approach is far from being
conclusive. The main treatment goals include normalization of body weight
and eating behaviors and alleviation of psychological problems related to EDs 28. Both outpatient and inpatient settings are used. In cases of extreme
emaciation and resistance to treatment, a forced treatment may be used.
However, it should be avoided whenever possible.
There is little evidence for justification of pharmacotherapy use in
AN. Initially promising findings with regards to Olanzapine 45,46 were 47 or
were not confirmed in more recent studies 48,49. All these studies were based
on small samples and their results are not conclusive. Antidepressive
medications have no effect on the course of AN, but they might be used to
treat co-morbid depression 28,50. Nonetheless, pharmacological treatment of
patients with AN is performed by means of antidepressants (tricyclic and
selective serotonin reuptake inhibitors), antipsychotics (typical and atypical),
Lithium, naltrexone, antihistamines, clonidine, human growth hormone or
cannabis 51.
About 57% of patients whose original diagnosis was AN were fully-
recovered at a follow-up measurement (the mean duration of the follow-up
of 4.8 years) in the Netherlands 52. Another conclusion of this study is that
Chapter 1
15
early detection is associated with a more positive outcome. A study in
Germany found that at the 12-year follow-up measurement 27.5% of the
patients initially diagnosed with AN had a good outcome, 25.3% an
intermediate outcome, 39.6% had a poor outcome, and 7 (7.7%) were
deceased 53. Factors associated most strongly with an unfavourable outcome
were sexual problems, impulsivity, long duration of inpatient treatment, and
long duration of an eating disorder. A review by Steinhausen (2002) adds
vomiting, bulimia, and purgative abuse, chronicity of illness, and obsessive-
compulsive personality symptoms to the list of unfavourable prognostic
features, and notes that other psychiatric disorders at follow-up
measurements are very common 42.
One of the reasons why therapy of AN is particularly challenging and
treatment drop-out rates are high (30-50% 54) is the fact that at least some of
the AN symptoms are ego-syntonic. This means that they are in harmony
with patients' goals and desires, hence, it is difficult to illicit motivation for
treatment.
Some of the aspects of AN which might be experienced as rewarding
by patients include:
• physiological sensations associated with starvation (e.g. stress
response leading to endogenous opioid secretion)
• gratification from exerting control over body weight and bodily
drives (appetite, hunger)
• positive feedback from the society or societal groups of reference
(e.g. pro-ana groups)
• positive feedback from the internalized social mirror (satisfaction of
own standards)
• hyperactivity might be rewarding (hypothetically, an evolutionary
conserved reaction to food scarcity which is supposed to promote
foraging)
• excessive exercising might be rewarding in several ways
• initially rewarding stimuli might lead to adaptation and possible
withdrawal effects
Chapter 1
16
Cultural context
The modern societies no longer struggle with scarcity of food and the times
of famine fade away in the memory of the Western countries. Food has
become easily obtainable, both in terms of financial resources and time.
Being one of the greatest achievements of the Western civilization, this
availability appears to entail increased rates of obesity and, presumably, EDs.
There is a stark contrast between ubiquitous food advertisements, which are
encouraging overindulging (especially in food that is of low nutritional
quality), and the societal pressure to be thin, exercised implicitly or explicitly
by our culture. The ambiguous attitude towards food and bodies permeates
the modern societies. Of note, some believe that the efforts focused on
combating obesity may unintentionally lead to an increase in incidence of
EDs. Solid evidence for or against this conviction is lacking. It should be kept
in mind that in spite of these conflicting societal pressures and common
dissatisfaction with own body (body dissatisfaction in both men and women
in western societies is so common that it is considered to be normative 55),
only a small fraction of individuals develop an eating disorder.
It is a matter of a debate to what extent culture determines EDs.
Historical studies and studies on Western and non-Western populations
report occurrences of AN without body image concerns or fear of gaining
weight (non-fat-phobic AN) 56,57. This means that the sociocultural pressures
are neither necessary, nor sufficient for the development of AN 58. Keel &
Klump (2003) in their systematic review of the historical and epidemiological
data as well as the data coming from studies on non-Western cultures
concluded that BN is a more culturally bound condition than AN 59. On the
other hand, the incidence of AN was much lower in the Netherlands Antilles
than in the Netherlands, but it was found to be similarly common among
Netherlands Antilleans living in the Netherlands as among native Dutch 60.
The Western idealization of thinness (pressure to be thin) appears to be a
risk factor for the development of AN (possibly in interaction with migration-
related stress and increased drive to conform in order to counteract
alienation) and dieting is a possible triggering factor for the onset of an ED.
Chapter 1
17
Selected candidate molecules for association with AN
The alterations of biological functioning in patients with AN are quite
dramatic. The difficulty in investigating those lies in determining the
difference between premorbid effects (predisposing factors) and effects
elicited by starvation and hyperactivity (biomarkers). Thorough discussion of
those alterations is beyond the scope of the present thesis (see for example 61,62), but three molecules which are plausible candidates to play a role in AN
will be briefly introduced.
Brain-derived neurotrophic factor (BDNF) is the most ubiquitous
member of the family of neurotrophins. It plays a role in neurodevelopment 63, neural plasticity, connectivity and synaptogenesis 64. It also has been
implicated in the regulation of body weight and eating behavior in humans 65
and animals 66. Genome-wide association studies (GWASs) found the BDNF
gene locus to be strongly associated with body mass index 67,68. Furthermore,
mice with reduced expression of BDNF display increased locomotor activity
and aberrant eating behavior leading to obesity 69,70. A hyperphagic
phenotype has also been observed in mice with reduced hypothalamic
expression of the TrkB – high affinity BDNF receptor 71. Similarly,
hyperphagia, obesity and hyperactivity are present in humans who have a
functional loss of one copy of the BDNF gene 72. BDNF operates downstream
of the melanocortin pathway to regulate energy balance 71. Finally, there is
evidence showing BDNF's involvement in reward and stress functioning 73.
Catechol-O-methyl transferase (COMT) is an enzyme which degrades
catecholamines, such as dopamine and noradrenaline 74. It has been
implicated in the pathogenesis of several mental disorders 75. One allele of a
functional variant on the COMT gene (rs4680) has been associated with a
less stable product and, therefore, lower enzymatic activity 76, which in turn
has been hypothesized to lead to higher dopamine availability 77. Rs4680 was
studied in mental disorders such as schizophrenia 78, autism 79, depression 80
and eating disorders 81,82.
Pro-opiomelanocortin (POMC) is a precursor peptide in the
melanocortin system (the melanocortin system is involved in body weight
Chapter 1
18
regulation via effects on appetite and energy expenditure)61. POMC can be
cleaved into several important peptides, such as α, β, and γ-MSH and β-
Endorphin. Among many other functions, it plays a role in regulation of
feeding behaviour 83.
BDNF, COMT and POMC and their genetic loci are viable candidates
to study in the context of body-weight related phenotypes, especially AN.
The next five sections are based on:
Brandys MK, de Kovel CG, Kas MJ, van Elburg AA, Adan RA. Overview of
genetic research in anorexia nervosa: The past, the present and the future.
Int J Eat Disord 2015. 84
Rationale for gene-association studies
Several lines of evidence suggest that there is a substantial genetic
component in the aetiology of AN. AN has been observed across many
cultures 59. Strong familiar aggregation of AN has been documented (relative
risk of 11.3 in first-degree relatives of cases with AN, as compared to the
general population 85,86), and the heritability (h2) has been estimated in
several twin studies and one adoption study of disordered eating symptoms 87. These estimates range from 0.56 (95% CI, 0.00-0.87) 88 to 0.74 (95% CI:
0.35-0.95) 89, depending on the studied population, definition of AN and
applied methodology.
The evidence coming from several lines of research demonstrates
that the genetic factors are pivotal in the aetiology of AN. No monogenic
forms of AN have been found and the data suggest that the genetic
underpinning of AN is multifactorial (i.e. multiple genetic variants with small
effects, rather than one or a few potent variants, working in concert with
environmental factors) 90. Two main types of studies have been employed in
a search for those genetic factors.
The linkage approach, which investigates co-segregation of the
genetic regions with the disease status in large families, has been successful
in detecting rare and very potent genetic variants involved in the aetiology of
Chapter 1
19
single-gene disorders (Mendelian), e.g. cystic fibrosis or Huntington’s disease 91,92. However, its usefulness in unravelling common variants of small effects
in complex, polygenic diseases or traits remains very limited.
The second category is a population-based genetic-association study,
which investigates whether frequencies of certain genotypes or alleles are
different between cases and controls (significant difference implies
association) or if they are correlated with a quantitative trait. This approach
focuses on variants with small or medium effects, in a multifactorial model.
Within this category, candidate-gene studies (CGSs) look into the single-
nucleotide polymorphisms (SNPs) in biologically plausible genes, whereas
GWASs test the common SNPs distributed throughout the whole genome.
Candidate gene approach
The candidate-gene approach in AN, much like in other psychiatric disorders,
turned out to be a primarily futile effort. The scarcity of successful
replications can be explained by several reasons, such as genetic differences
between the discovery population and the populations in the replication
attempts, or by the errors and biases leading to false positive results.
Retrospectively, given the complexity and redundancy of biological
pathways, and in light of what is now known about the genetic architecture
of psychiatric diseases, the hypotheses about which genes could potentially
harbor causative mutations had small chances to be proven right. Out of the
hundreds of associations indicated by CGSs in the biomedical research only a
few were replicated in GWASs 93. This ratio is even less favourable in the field
of psychiatry. One study found a lack of enrichment of the association signal
in a large genome-wide dataset of cases with schizophrenia and controls
after the analysis of 732 autosomal genes indicated in 1374 CGSs
(investigation of signal enrichment involves collective testing of a selected
group of variants in an independent dataset; it has much greater power,
compared to testing of individual variants) 94.
Chapter 1
20
Candidate gene studies in anorexia nervosa
Comprehensive reviews of CGSs in AN are available elsewhere 95,96. Although
the selection of candidate genes for studies of AN was based on interesting
hypotheses 97, and more than 200 gene-association studies were performed
in the context of EDs, up to date none of the initially promising findings have
been convincingly replicated in the subsequent candidate or genome-wide
studies. Meta-analyses, which summarized and weighted the evidence from
multiple studies, were also disillusioning 98-101. Also the relatively recent CGS
which used the modern standards of design, quality control and statistical
significance was negative 102. Still, there are a few findings which await
replication attempts, such as rs1473473 of TPH2 103, the 5-HTTLPR
polymorphism on SLC6A4 104, rs7180942 in NTRK3 105 and Ala67Thr variant in
AGRP 106 (these polymorphisms were not tested in two recent GWASs of AN,
because they were not present on the genotyping arrays used in those
studies).
In parallel to the growing disillusionment about the candidate-gene
method, a new approach towards investigation of genetic associations
emerged. GWAS technology is relatively recent (first GWAS dates back to
2005 107), but it already has had significant impact on the landscape of
biomedical research and resulted in progression of the aetiological
knowledge about diseases and traits 108.
Genome-wide association approach
GWAS is a hypothesis-free approach. It uses microarray platforms to
examine the genotypic data from a large number of SNPs (from hundreds of
thousands up to millions), which cover most of the human common SNP
variation (a SNP is considered common if the frequency of its minor allele is
larger than 1%). This coverage is increased via imputation - a procedure
which uses statistical algorithms to infer the genotypes of the ungenotyped
SNPs by employing the reference data coming from e.g. HapMap or 1000
Genomes Project populations. Genome-wide data also allows for
Chapter 1
21
investigation of copy number variants (CNVs; deleted or duplicated stretches
of the genome).
Below is a list of the main goals of GWASs:
• Furthering the understanding of the biological mechanisms of the
disease, by finding the genes and pathways involved in the aetiology.
This is the foremost goal of GWASs.
• Learning about the genetic architecture. This includes the expected
range of effect sizes, allelic frequencies of the associated variants,
underlying genetic models (additive, dominant, recessive,
overdominant, multiplicative) and the possibility of gene x
environment and gene x gene interactions.
• Understanding of the genetic overlap between diseases and traits.
This has a potential of enhancing the nosological system and
treatment.
• Genetic screening to identify populations at risk (risk prediction) or
individual genotyping of a patient to inform diagnosis and treatment
(personalized medicine). As exciting as these prospects are, they are
distant goals, and in view of a highly polygenic nature of psychiatric
diseases, they are unlikely to be achievable in the near future 109.
What needs to be remembered when interpreting a GWAS is that its
results inform about association but do not determine causality, and that a
statistical strength of association at a given locus should not be confused
with its biological relevance (the most significant finding in GWASs might not
be the most informative).
Scope and outline of the thesis
The overarching theme of this thesis is the effort to shed light on the genetic
background of AN. That undertaking predominantly involves searching for
Chapter 1
22
the genetic associations via the candidate-gene and genome-wide studies.
Thus, we aimed to find the genetic variants which change the risk of
developing AN, and to increase the understanding of the mechanisms of the
disease. Beyond investigating the genetic associations, in one chapter we
also take interest in studies examining the serum levels of the BDNF
neurotrophin in patients with AN. BDNF is a product of the BDNF gene which
was also investigated in this thesis.
We begin by studying a possible genetic relation of AN and obesity
by testing several genetic variants associated with the latter in a sample of
patients with AN110. This work adds to the discussion about the genetic
nature of AN and its hypothesized relation to the opposite extreme of the
weight spectrum - obesity (chapter 2). The next chapter (3) presents the only
study which does not directly involve patients with AN. In that publication
we describe the search for association of variants from the POMC locus with
detailed measures of body composition and nutrient choice in the general
population111. POMC molecule, due to its effects on the appetite regulation,
was a plausible candidate for having a role in AN diathesis.
We also explore the alterations of biological functioning in AN, in the
context of possible candidates for genetic associations. Chapter 4 is a meta-
analysis of several studies which compared the serum BDNF levels in patients
with AN and healthy controls112. A discussion of whether the observed
effects are state or trait dependent is included. Thereafter, in chapter 5, we
apply the meta-analytical methodology to a promising candidate for genetic
association with AN, the Val66Met polymorphism on the BDNF gene101.
Meta-analyses combine evidence from multiple studies on a single subject,
which (together with our own, novel data) allows us to draw stronger
conclusions than any of these studies alone.
The further investigations (chapter 6) consider a genetic
polymorphism long thought to play a role in several mental disorders,
including ED. Val158Met polymorphism of the COMT gene was tested in the
Chapter 1
23
new data and the results were merged in a meta-analytical framework with
the results of previous studies on this subject100.
Chapter 7 presents the work which used the genome-wide data of
patients with AN and controls to test for association of AN with selected
structural variants (rather than SNPs). The nature of the available data was
not sufficient to render the results fully credible; hence this study remains a
preliminary investigation.
The appendix describes a large collaborative study which uses
thousands of DNA samples from all over the world and analyses them in a
hypothesis-free, genome-wide approach113. This remains the largest genetic
study in AN up to date.
The final publication included in this dissertation (chapter 8) is an
opinion paper which reviews and discusses the past approaches to the
studies of gene-association in AN, shows how they evolved over the years (a
process well reflected in the present thesis), and tries to outline the
directions for future research.
References
1. Gull WW. Anorexia nervosa (apepsia hysterica, anorexia hysterica). 1868.
Obes Res 1997;5:498-502.
2. Pearce JMS. Richard morton: Origins of anorexia nervosa. Eur Neurol
2004;52:191-192.
3. Sullivan P. Mortality in anorexia nervosa. Am J Psychiatry 1995;152:1073-
1074.
4. Keel PK, Dorer DJ, Eddy KT, Franko D, Charatan DL, Herzog DB. Predictors
of mortality in eating disorders. Arch Gen Psychiatry 2003;60:179-183.
5. Thomas JJ, Vartanian LR, Brownell KD. The relationship between eating
disorder not otherwise specified (EDNOS) and officially recognized eating
disorders: Meta-analysis and implications for DSM. Psychol Bull
2009;135:407-433.
6. Machado PPP, Machado BC, Gonçalves S, Hoek HW. The prevalence of
eating disorders not otherwise specified. Int J Eat Disord 2007;40:212-217.
Chapter 1
24
7. Milos G, Spindler A, Schnyder U, Fairburn CG. Instability of eating disorder
diagnoses: Prospective study. The British Journal of Psychiatry 2005;187:573-
578.
8. Tozzi F, Thornton LM, Klump KL, Fichter MM, Halmi KA, Kaplan AS, et al.
Symptom fluctuation in eating disorders: Correlates of diagnostic crossover.
Am J Psychiatry 2005;162:732-740.
9. Eddy KT, Dorer DJ, Franko DL, Tahilani K, Thompson-Brenner H, Herzog DB.
Diagnostic crossover in anorexia nervosa and bulimia nervosa: Implications
for DSM-V. Am J Psychiatry 2008;165:245-250.
10. Agras WS, Crow S, Mitchell JE, Halmi KA, Bryson S. A 4-year prospective
study of eating disorder NOS compared with full eating disorder syndromes.
Int J Eat Disord 2009;42:565-570.
11. Fisher MM, Gonzalez M, Malizio J. Eating disorders in adolescents: How
does the DSM-5 change the diagnosis? Journal of Adolescent Health
2014;54:S9.
12. Peat C, Mitchell JE, Hoek H, Wonderlich S. Validity and utility of subtyping
anorexia nervosa. Int J Eat Disord 2009;9999:NA.
13. Eddy KT, Keel PK, Dorer DJ, Delinsky SS, Franko DL, Herzog DB.
Longitudinal comparison of anorexia nervosa subtypes. Int J Eat Disord
2002;31:191-201.
14. Keel PK, Dorer DJ, Eddy KT, Delinsky SS, Franko DL, Blais MA, et al.
Predictors of treatment utilization among women with anorexia and bulimia
nervosa. Am J Psychiatry 2002;159:140-142.
15. Magidson J, Vermunt JK. Latent class models for clustering: A comparison
with K-means. Canadian Journal of Marketing Research 2002;20:36-36-43.
16. Wonderlich SA, Joiner TE, Keel PK, Williamson DA, Crosby RD. Eating
disorder diagnoses: Empirical approaches to classification. Am Psychol
2007;62:167-180.
17. Swanson SA, Lindenberg K, Bauer S, Crosby RD. A monte carlo
investigation of factors influencing latent class analysis: An application to
eating disorder research. Int J Eat Disord 2011:n/a-n/a.
Chapter 1
25
18. Wildes JE, Marcus MD, Crosby RD, Ringham RM, Dapelo MM, Gaskill JA,
et al. The clinical utility of personality subtypes in patients with anorexia
nervosa. J Consult Clin Psychol 2011;79:665-674.
19. Steiger H, Richardson J, Schmitz N, Israel M, Bruce KR, Gauvin L. Trait-
defined eating-disorder subtypes and history of childhood abuse. Int J Eat
Disord 2010;43:428-432.
20. Miller MW, Kaloupek DG, Dillon AL, Keane TM. Externalizing and
internalizing subtypes of combat-related PTSD: A replication and extension
using the PSY-5 scales. J Abnorm Psychol 2004;113:636-645.
21. Hudson JI, Hiripi E, Pope Jr. HG, Kessler RC. The prevalence and correlates
of eating disorders in the national comorbidity survey replication. Biological
Psychiatry 2007;61:348-358.
22. Keski-Rahkonen A, Hoek HW, Susser ES, Linna MS, Sihvola E, Raevuori A,
et al. Epidemiology and course of anorexia nervosa in the community. Am J
Psychiatry 2007;164:1259-1265.
23. Lucas A, Beard C, O'Fallon W, Kurland L. 50-year trends in the incidence
of anorexia nervosa in rochester, minn.: A population-based study. Am J
Psychiatry 1991;148:917-922.
24. Hoek HW, van Hoeken D. Review of the prevalence and incidence of
eating disorders. International Journal of Eating Disorders 2003;34:383-396.
25. Isomaa R, Isomaa A, Marttunen M, Kaltiala-Heino R, Björkqvist K. The
prevalence, incidence and development of eating disorders in finnish
adolescents - a two-step 3-year follow-up study. European Eating Disorders
Review 2009;17:199-207.
26. Smink FR, van Hoeken D, Hoek HW. Epidemiology of eating disorders:
Incidence, prevalence and mortality rates. Curr Psychiatry Rep 2012;14:406-
414.
27. Currin L, Schmidt U, Treasure J, Jick H. Time trends in eating disorder
incidence. The British Journal of Psychiatry 2005;186:132-135.
28. Herpertz S, Hagenah U, Vocks S, von Wietersheim J, Cuntz U, Zeeck A.
The diagnosis and treatment of eating disorders. Dtsch Arztebl Int
2011;108:678-685.
Chapter 1
26
29. Birmingham CL, Su J, Hlynsky JA, Goldner EM, Gao M. The mortality rate
from anorexia nervosa. Int J Eat Disord 2005;38:143-146.
30. Papadopoulos FC, Ekbom A, Brandt L, Ekselius L. Excess mortality, causes
of death and prognostic factors in anorexia nervosa. The British Journal of
Psychiatry 2009;194:10-17.
31. Rosling AM, Sparén P, Norring C, von Knorring A. Mortality of eating
disorders: A follow-up study of treatment in a specialist unit 1974?2000. Int J
Eat Disord 2011;44:304-310.
32. Arcelus J, Mitchell AJ, Wales J, Nielsen S. Mortality rates in patients with
anorexia nervosa and other eating disorders: A meta-analysis of 36 studies.
Arch Gen Psychiatry 2011;68:724-731.
33. Jacobi C, Fittig E, Bryson SW, Wilfley D, Kraemer HC, Taylor CB. Who is
really at risk? identifying risk factors for subthreshold and full syndrome
eating disorders in a high-risk sample. Psychol Med 2011;41:1939.
34. Nicholls DE, Viner RM. Childhood risk factors for lifetime anorexia
nervosa by age 30 years in a national birth cohort. J Am Acad Child Adolesc
Psychiatry 2009;48:791-799.
35. Pike KM, Hilbert A, Wilfley DE, Fairburn CG, Dohm FA, Walsh BT, et al.
Toward an understanding of risk factors for anorexia nervosa: A case-control
study. Psychol Med 2008;38:1443-1453.
36. Winje E, Willoughby K, Lask B. Season of birth bias in eating disorders--
fact or fiction? Int J Eat Disord 2008;41:479-490.
37. Disanto G, Handel AE, Para AE, Ramagopalan SV, Handunnetthi L. Season
of birth and anorexia nervosa. The British Journal of Psychiatry 2011.
38. Favaro A, Tenconi E, Ceschin L, Zanetti T, Bosello R, Santonastaso P. In
utero exposure to virus infections and the risk of developing anorexia
nervosa. Psychol Med 2011:1-7.
39. Culbert KM, Breedlove SM, Burt SA, Klump KL. Prenatal hormone
exposure and risk for eating disorders: A comparison of opposite-sex and
same-sex twins. Arch Gen Psychiatry 2008;65:329-336.
Chapter 1
27
40. Baker JH, Lichtenstein P, Kendler KS. Intrauterine testosterone exposure
and risk for disordered eating. The British Journal of Psychiatry
2009;194:375-376.
41. Wentz E, Lacey JH, Waller G, Rastam M, Turk J, Gillberg C. Childhood
onset neuropsychiatric disorders in adult eating disorder patients. A pilot
study. Eur Child Adolesc Psychiatry 2005;14:431-437.
42. Steinhausen H. The outcome of anorexia nervosa in the 20th century. Am
J Psychiatry 2002;159:1284-1293.
43. Striegel-Moore RH, Leslie D, Petrill SA, Garvin V, Rosenheck RA. One-year
use and cost of inpatient and outpatient services among female and male
patients with an eating disorder: Evidence from a national database of health
insurance claims. Int J Eat Disord 2000;27:381-389.
44. Haas L, Stargardt T, Schreyoegg J, Schlösser R, Danzer G, Klapp BF.
Inpatient costs and predictors of costs in the psychosomatic treatment of
anorexia nervosa. Int J Eat Disord 2011:n/a-n/a.
45. Brambilla F, Garcia CS, Fassino S, Daga GA, Favaro A, Santonastaso P, et
al. Olanzapine therapy in anorexia nervosa: Psychobiological effects. Int Clin
Psychopharmacol 2007;22:197-204.
46. Leggero C, Masi G, Brunori E, Calderoni S, Carissimo R, Maestro S, et al.
Low-dose olanzapine monotherapy in girls with anorexia nervosa, restricting
subtype: Focus on hyperactivity. J Child Adolesc Psychopharmacol
2010;20:127-133.
47. Attia E, Kaplan AS, Walsh BT, Gershkovich M, Yilmaz Z, Musante D, et al.
Olanzapine versus placebo for out-patients with anorexia nervosa. Psychol
Med 2011:1-6.
48. Kafantaris V, Leigh E, Hertz S, Berest A, Schebendach J, Sterling WM, et al.
A placebo-controlled pilot study of adjunctive olanzapine for adolescents
with anorexia nervosa. J Child Adolesc Psychopharmacol 2011;21:207-212.
49. Norris ML, Spettigue W, Buchholz A, Henderson KA, Gomez R, Maras D,
et al. Olanzapine use for the adjunctive treatment of adolescents with
anorexia nervosa. J Child Adolesc Psychopharmacol 2011.
Chapter 1
28
50. Suarez-Pinilla P, Pena-Perez C, Arbaizar-Barrenechea B, Crespo-Facorro B,
Del Barrio JA, Treasure J, et al. Inpatient treatment for anorexia nervosa: A
systematic review of randomized controlled trials. J Psychiatr Pract
2015;21:49-59.
51. Aigner M, Treasure J, Kaye W, Kasper S. World federation of societies of
biological psychiatry (WFSBP) guidelines for the pharmacological treatment
of eating disorders. World Journal of Biological Psychiatry 2011;12:400-443.
52. van Son GE, van Hoeken D, van Furth EF, Donker GA, Hoek HW. Course
and outcome of eating disorders in a primary care-based cohort. Int J Eat
Disord 2009;9999:NA.
53. Fichter MM, Quadflieg N, Hedlund S. Twelve-year course and outcome
predictors of anorexia nervosa. Int J Eat Disord 2006;39:87-100.
54. Guarda AS. Treatment of anorexia nervosa: Insights and obstacles.
Physiology & Behavior, 2008;94:113-120.
55. Tantleff-Dunn S, Barnes RD, Larose JG. It's not just a "woman thing:" The
current state of normative discontent. Eat Disord 2011;19:392-402.
56. Santonastaso P, Bosello R, Schiavone P, Tenconi E, Degortes D, Favaro A.
Typical and atypical restrictive anorexia nervosa: Weight history, body image,
psychiatric symptoms, and response to outpatient treatment. Int J Eat Disord
2009;42:464-470.
57. Becker AE, Thomas JJ, Pike KM. Should non-fat-phobic anorexia nervosa
be included in DSM-V? Int J Eat Disord 2009;42:620-635.
58. Schmidt U. Epidemiology and aetiology of eating disorders. Psychiatry,
2005;4:5-9.
59. Keel PK, Klump KL. Are eating disorders culture-bound syndromes?
implications for conceptualizing their etiology. Psychol Bull 2003;129:747-
769.
60. van Hoeken D, Veling W, Smink FRE, Hoek HW. The incidence of anorexia
nervosa in netherlands antilles immigrants in the netherlands. Eur Eat Disord
Rev 2010;18:399-403.
61. Miller K. Endocrine dysregulation in anorexia nervosa update. The Journal
of Clinical Endocrinology & Metabolism 2011;96:2939-2949.
Chapter 1
29
62. Adan RAH, Hillebrand JJG, Danner UN, Cano SC, Kas MJH, Verhagen LAW.
Neurobiology driving hyperactivity in activity-based anorexia. Curr Top Behav
Neurosci 2011;6:229-250.
63. Hyman C, Hofer M, Barde Y, Juhasz M, Yancopoulos GD, Squinto SP, et al.
BDNF is a neurotrophic factor for dopaminergic neurons of the substantia
nigra. Nature 1991;350:230-232.
64. Greenberg ME, Xu B, Lu B, Hempstead BL. New insights in the biology of
BDNF synthesis and release: Implications in CNS function. J Neurosci
2009;29:12764-12767.
65. Lommatzsch M, Zingler D, Schuhbaeck K, Schloetcke K, Zingler C, Schuff-
Werner P, et al. The impact of age, weight and gender on BDNF levels in
human platelets and plasma. Neurobiol Aging 2005;26:115-123.
66. King BM. The rise, fall, and resurrection of the ventromedial
hypothalamus in the regulation of feeding behavior and body weight. Physiol
Behav 2006;87:221-244.
67. Willer CJ, Speliotes EK, Loos RJ, Li S, Lindgren CM, Heid IM, et al. Six new
loci associated with body mass index highlight a neuronal influence on body
weight regulation. Nat Genet 2009;41:25-34.
68. Locke AE, Kahali B, Berndt SI, Justice AE, Pers TH, Day FR, et al. Genetic
studies of body mass index yield new insights for obesity biology. Nature
2015;518:197-206.
69. Kernie SG, Liebl DJ, Parada LF. BDNF regulates eating behavior and
locomotor activity in mice. EMBO J 2000;19:1290-1300.
70. Rios M, Fan G, Fekete C, Kelly J, Bates B, Kuehn R, et al. Conditional
deletion of brain-derived neurotrophic factor in the postnatal brain leads to
obesity and hyperactivity. Mol Endocrinol 2001;15:1748-1757.
71. Xu B, Goulding EH, Zang K, Cepoi D, Cone RD, Jones KR, et al. Brain-
derived neurotrophic factor regulates energy balance downstream of
melanocortin-4 receptor. Nat Neurosci 2003;6:736-742.
72. Gray J, Yeo GSH, Cox JJ, Morton J, Adlam AR, Keogh JM, et al.
Hyperphagia, severe obesity, impaired cognitive function, and hyperactivity
Chapter 1
30
associated with functional loss of one copy of the brain-derived neurotrophic
factor (BDNF) gene. Diabetes 2006;55:3366-3371.
73. Narita M, Aoki K, Takagi M, Yajima Y, Suzuki T. Implication of brain-
derived neurotrophic factor in the release of dopamine and dopamine-
related behaviors induced by methamphetamine. Neuroscience
2003;119:767-775.
74. Chen J, Lipska BK, Halim N, Ma QD, Matsumoto M, Melhem S, et al.
Functional analysis of genetic variation in catechol-O-methyltransferase
(COMT): Effects on mRNA, protein, and enzyme activity in postmortem
human brain. Am J Hum Genet 2004;75:807-821.
75. Mier D, Kirsch P, Meyer-Lindenberg A. Neural substrates of pleiotropic
action of genetic variation in COMT: A meta-analysis. Mol Psychiatry
2010;15:918-927.
76. Lachman HM, Papolos DF, Saito T, Yu YM, Szumlanski CL, Weinshilboum
RM. Human catechol-O-methyltransferase pharmacogenetics: Description of
a functional polymorphism and its potential application to neuropsychiatric
disorders. Pharmacogenetics 1996;6:243-250.
77. Shield AJ, Thomae BA, Eckloff BW, Wieben ED, Weinshilboum RM.
Human catechol O-methyltransferase genetic variation: Gene resequencing
and functional characterization of variant allozymes. Mol Psychiatry
2004;9:151-160.
78. Costas J, Sanjuán J, Ramos-Ríos R, Paz E, Agra S, Ivorra JL, et al.
Heterozygosity at catechol-O-methyltransferase Val158Met and
schizophrenia: New data and meta-analysis. J Psychiatr Res 2011;45:7-14.
79. Stergiakouli E, Thapar A. Fitting the pieces together: Current research on
the genetic basis of attention-deficit/hyperactivity disorder (ADHD).
Neuropsychiatr Dis Treat 2010;6:551-560.
80. Åberg E, Fandiño-Losada A, Sjöholm LK, Forsell Y, Lavebratt C. The
functional Val158Met polymorphism in catechol-O-methyltransferase
(COMT) is associated with depression and motivation in men from a swedish
population-based study. J Affect Disord 2011;129:158-166.
Chapter 1
31
81. Mikolajczyk E, Smiarowska M, Grzywacz A, Samochowiec J. Association of
eating disorders with catechol- O -methyltransferase gene functional
polymorphism. Neuropsychobiology 2006;54:82-86.
82. Frieling H, Romer KD, Wilhelm J, Hillemacher T, Kornhuber J, de Zwaan
M, et al. Association of catecholamine-O-methyltransferase and 5-HTTLPR
genotype with eating disorder-related behavior and attitudes in females with
eating disorders. Psychiatr Genet 2006;16:205-208.
83. Millington G. The role of proopiomelanocortin (POMC) neurones in
feeding behaviour. Nutr Metab (Lond) 2007;4:18.
84. Brandys MK, de Kovel CG, Kas MJ, van Elburg AA, Adan RA. Overview of
genetic research in anorexia nervosa: The past, the present and the future.
Int J Eat Disord 2015.
85. Lilenfeld LR, Kaye WH, Greeno CG, Merikangas KR, Plotnicov K, Pollice C,
et al. A controlled family study of anorexia nervosa and bulimia nervosa:
Psychiatric disorders in first-degree relatives and effects of proband
comorbidity. Arch Gen Psychiatry 1998;55:603-610.
86. Strober M, Freeman R, Lampert C, Diamond J, Kaye W. Controlled family
study of anorexia nervosa and bulimia nervosa: Evidence of shared liability
and transmission of partial syndromes. Am J Psychiatry 2000;157:393-401.
87. Klump KL, Suisman JL, Burt SA, McGue M, Iacono WG. Genetic and
environmental influences on disordered eating: An adoption study. J Abnorm
Psychol 2009;118:797.
88. Bulik CM, Sullivan PF, Tozzi F, Furberg H, Lichtenstein P, Pedersen NL.
Prevalence, heritability, and prospective risk factors for anorexia nervosa.
Arch Gen Psychiatry 2006;63:305-312.
89. Klump KL, Miller K, Keel P, McGue M, Iacono W. Genetic and
environmental influences on anorexia nervosa syndromes in a
population–based twin sample. Psychol Med 2001;31:737-740.
90. Steinhausen H, Jakobsen H, Helenius D, Munk-Jorgensen P, Strober M. A
nationwide study of the family aggregation and risk factors in anorexia
nervosa over three generations. Int J Eat Disord 2014;48:1-8.
Chapter 1
32
91. Knowlton RG, Cohen-Haguenauer O, Van Cong N, Frezal J, Brown VA,
Barker D, et al. A polymorphic DNA marker linked to cystic fibrosis is located
on chromosome 7. Nature 1985;318:380-382.
92. Gusella JF, Wexler NS, Conneally PM, Naylor SL, Anderson MA, Tanzi RE,
et al. A polymorphic DNA marker genetically linked to huntington's disease.
Nature 1983;306:234-238.
93. Siontis KC, Patsopoulos NA, Ioannidis JP. Replication of past candidate
loci for common diseases and phenotypes in 100 genome-wide association
studies. European Journal of Human Genetics 2010;18:832-837.
94. Collins AL, Kim Y, Sklar P, International Schizophrenia Consortium,
O'Donovan MC, Sullivan PF. Hypothesis-driven candidate genes for
schizophrenia compared to genome-wide association results. Psychol Med
2012;42:607-616.
95. Clarke TK, Weiss ARD, Berrettini WH. The genetics of anorexia nervosa.
Clin Pharmacol Ther 2011;91:181-188.
96. Rask-Andersen M, Olszewski PK, Levine AS, Schioth HB. Molecular
mechanisms underlying anorexia nervosa: Focus on human gene association
studies and systems controlling food intake. Brain Res Rev 2010;62:147-164.
97. Hinney A, Volckmar A. Genetics of eating disorders. Curr Psychiatry Rep
2013;15:1-9.
98. Slof-Op't Landt MC, Furth EF, Meulenbelt I, Bartels M, Hottenga JJ,
Slagboom PE, et al. Association study of the estrogen receptor I gene (ESR1)
in anorexia nervosa and eating disorders: No replication found. Int J Eat
Disord 2013;47:211-214.
99. Zhang C, Chen J, Jia X, Yu S, Jiang W, Zhang R, et al. Estrogen receptor 1
gene rs2295193 polymorphism and anorexia nervosa: New data and
meta†analysis. Asia⠀ Pacific Psychiatry 2013;5:331-335.
100. Brandys MK, Slof-Op't Landt MC, van Elburg AA, Ophoff R, Verduijn W,
Meulenbelt I, et al. Anorexia nervosa and the Val158Met polymorphism of
the COMT gene: Meta-analysis and new data. Psychiatr Genet 2012;22:130-
136.
Chapter 1
33
101. Brandys MK, Kas MJ, van Elburg AA, Ophoff R, Slof-Op't Landt MC,
Middeldorp CM, et al. The Val66Met polymorphism of the BDNF gene in
anorexia nervosa: New data and a meta-analysis. The World Journal of
Biological Psychiatry 2013;14:441-451.
102. Pinheiro AP, Bulik CM, Thornton LM, Sullivan PF, Root TL, Bloss CS, et al.
Association study of 182 candidate genes in anorexia nervosa. Am J Med
Genet B Neuropsychiatr Genet 2010;153B:1070-1080.
103. Slof-O't Landt M, Meulenbelt I, Bartels M, Suchiman E, Middeldorp C,
Houwing†Duistermaat J, et al. Association study in eating disorders:
TPH2 associates with anorexia nervosa and self-â induced vomiting. Genes,
Brain and Behavior 2011;10:236-243.
104. Calati R, De Ronchi D, Bellini M, Serretti A. The 5-HTTLPR polymorphism
and eating disorders: A meta- analysis. Int J Eat Disord 2011;44:191-199.
105. Mercader JM, Saus E, Aguera Z, Bayes M, Boni C, Carreras A, et al.
Association of NTRK3 and its interaction with NGF suggest an altered cross-
regulation of the neurotrophin signaling pathway in eating disorders. Hum
Mol Genet 2008;17:1234-1244.
106. Vink T, Hinney A, Van Elburg A, van Goozen SH, Sandkuijl L, Sinke R, et
al. Association between an agouti-related protein gene polymorphism and
anorexia nervosa. Mol Psychiatry 2001;6:325-328.
107. Klein RJ, Zeiss C, Chew EY, Tsai JY, Sackler RS, Haynes C, et al.
Complement factor H polymorphism in age-related macular degeneration.
Science 2005;308:385-389.
108. Kim Y, Zerwas S, Trace SE, Sullivan PF. Schizophrenia genetics: Where
next? Schizophr Bull 2011;37:456-463.
109. Wray NR, Yang J, Hayes BJ, Price AL, Goddard ME, Visscher PM. Pitfalls
of predicting complex traits from SNPs. Nature Reviews Genetics
2013;14:507-515.
110. Brandys MK, van Elburg AA, Loos RJF, Bauer F, Hendriks J, van der
Schouw YT, et al. Are recently identified genetic variants regulating BMI in
the general population associated with anorexia nervosa? American Journal
of Medical Genetics Part B: Neuropsychiatric Genetics 2010;153B:695-699.
Chapter 1
34
111. Ternouth A, Brandys MK, van der Schouw YT, Hendriks J, Jansson J,
Collier D, et al. Association study of POMC variants with body composition
measures and nutrient choice. Eur J Pharmacol 2011.
112. Brandys MK, Kas MJH, van Elburg AA, Campbell IC, Adan RAH. A meta-
analysis of circulating BDNF concentrations in anorexia nervosa. World J Biol
Psychiatry 2011.
113. Boraska V, Franklin C, Floyd J, Thornton L, Huckins L, Southam L, et al. A
genome-wide association study of anorexia nervosa. Mol Psychiatry
2014;19:1085-1094.
Chapter 2
35
Chapter 2
Are recently identified genetic variants regulating BMI in the
general population associated with anorexia nervosa?
Marek K. Brandys
Annemarie A. van Elburg
Ruth J.F. Loos
Florianne Bauer
Judith Hendriks
Yvonne T. van der Schouw
Roger A.H. Adan
American Journal of Medical Genetics Part B: Neuropsychiatric Genetics
2010; 153B(2): 695-699.
Chapter 2
36
Abstract
The influence of body mass index (BMI) on susceptibility to anorexia nervosa
(AN) is not clear. Recently published genome-wide association (GWA) studies
of the general population identified several variants influencing BMI. We
genotyped these variants in an AN sample to test for association and to
investigate a combined effect of BMI-increasing alleles (as determined in the
original GWA studies) on the risk of developing the disease. Individual single
nucleotide polymorphisms (SNPs) were tested for association with AN in a
sample of 267 AN patients and 1636 population controls. A logistic
regression for the combined effect of BMI-increasing alleles included 225
cases and 1351 controls. We found no significant association between
individual SNPs and AN. The analysis of a combined effect of BMI-increasing
alleles showed absence of association with the investigated condition. The
percentages of BMI-increasing alleles were equal between cases and
controls. This study found no evidence that genetic variants regulating BMI in
the general population are significantly associated with susceptibility to AN.
Chapter 2
37
Introduction
Two extensive population-based genome-wide association studies (GWA) of
BMI and obesity have been published recently 1,2. Both studies revealed
associations of new loci and confirmed already known roles of FTO and
MC4R 3,4. All together, these loci harbor 10 genes, most of them
predominantly expressed in the central nervous system (Table I).
Although little is known about the function of the genes linked to the
newly identified genetic variants for body mass index (BMI) and common
obesity, preliminary evidence suggests that they might affect BMI via
involvement in the neuronal regulation of food intake 1,2. Additionally, it has
been proposed that eating disorders and obesity may be considered on the
same continuum of psychopathology (as opposed to discrete models; 5). We,
therefore, hypothesized that variants affecting BMI and obesity may
potentially alter the risk of developing an eating disorder such as anorexia
nervosa (AN).
It is under debate whether high or low BMI before the onset of the
disease has an impact upon susceptibility to AN 6,7. AN patients with lower
premorbid BMI (self-reported) tend to present with lower BMI at first
referral for the treatment 8. A recent study found a correlation between
premorbid BMI and BMI at discharge from the treatment and at follow-up in
AN 6,8. Patients with lower BMI before the onset of the disease and at
admission had poorer general indices of functioning 6. Elevated BMI could
either protect against the disease – since AN is a disorder of low body weight
– or increase the risk, via a tendency to a general eating pathology such as
e.g. restrained eating, excessive dieting or a persistent desire to lose weight.
In the present study we aimed to test whether the genetic variants
increasing BMI in the general population escalate the susceptibility to AN or
diminish it (or have no effect upon it).
Chapte
r 2
38
TA
BL
E I
. C
ha
ract
eris
tics
of
inv
esti
ga
ted
SN
Ps
an
d a
na
lysi
s o
f a
sso
cia
tio
n w
ith
AN
(a
llel
ic t
est,
1d
f)
SN
P
Nea
rby
gene
BM
I
incr.
alle
le
Eff
ect
all
ele
freq
. re
port
ed
in r
efe
rred
GW
AS
1,2
Eff
ect
alle
le
freq
. in
case
s
Eff
ect
all
ele
freq
. in
contr
ols
HW
E i
n
cont.
Eff
ect
size
s det
ecta
ble
at
80%
pow
er
OR
fo
r
the
asso
c.
(95%
CI)
P-v
alu
e, a
lleli
c te
st
OR
fo
r hete
ro-
zygo
tes
OR
fo
r ho
mo
-
zygo
tes
rs1121980
*
FT
O
A
41%
41.3
%
42.1
%
.10
1.3
1.7
.9
6
(.80-1
.16)
.73
rs17700633
M
C4R
A
32%
29.5
%
30.1
%
.99
1.3
1.7
.9
7
(.79-1
.18)
.78
rs17782313
M
C4R
C
21%
26.0
%
26.1
%
1.0
0
1.4
1.7
.9
9
(.80-1
.22)
.96
rs6548238
T
ME
M1
8
C
84%
83.3
%
83.7
%
.86
1.5
2.3
1.0
3
(.80-1
.32)
.81
rs10938397
G
NP
DA
2
G
45%
42.2
%
42.0
%
.98
1.3
1.7
1.0
0
(.83-1
.21)
.94
rs7498665
S
H2
B1
G
41%
38.5
%
37.8
%
.86
1.3
1.7
1.0
2
(.84-1
.23)
.79
rs368794
*
KC
TD
15
T
68%
68.5
%
67.5
%
1.0
0
1.3
1.8
.9
5
(.78-1
.17)
.67
rs10838738
M
TC
H2
G
34%
34.0
%
32.5
%
.45
1.3
1.7
1.0
6
(.87-1
.29)
.51
rs2568958
*
NE
GR
1
A
62%
60.2
%
57.9
%
.30
1.3
1.8
.9
0
(.75-1
.09)
.32
rs1488830
*
BD
NF
T
79%
75.9
%
78.0
%
.69
1.5
2.1
1.1
3
(.91-1
.40)
.26
rs925946
B
DN
F
T
30%
28.1
%
29.2
%
.57
1.3
1.8
.9
5
(.77-1
.16)
.62
rs7647305
E
TV
5
C
80%
79.9
%
80.1
%
.93
1.5
2.2
1.0
1
(.80-1
.27)
.93
SN
P,
Sin
gle
Nucle
oti
de
Po
lym
orp
his
m;
BM
I, b
ody m
ass
index;
Fre
q.,
fre
quency;
Co
nt.
, co
ntr
ols
; O
R,
odds
rati
o;
CI,
confi
dence
inte
rvals
; H
WE
, H
ard
y-W
ein
ber
g e
quil
ibri
um
; χ
2
test
wit
h 1
df
for
HW
E;
assu
mp
tio
ns
for
pow
er c
alcula
tio
n:
all
eli
c te
st (
1d
f), α
=.0
5,
pre
vale
nce
=.0
2.
* S
NP
s in
LD
wit
h S
NP
s id
enti
fied i
n G
WA
stu
die
s o
f B
MI
(pro
xie
s)
Chapter 2
39
Methods and materials
A total of 13 BMI-associated single nucleotide polymorphisms (SNP), selected
from the recent GWA studies of BMI 1-4, were genotyped in 267 AN patients
and 1636 control individuals. Nine SNPs were the same as those identified in
the GWA studies and four SNPs were in perfect or high linkage disequilibrium
(LD; r2 > 0.84) with the SNPs of interest (Table I).
The patients’ group consisted of female AN cases with ascertained
Dutch descent (patients are asked whether all of their grandparents were of
Dutch origin). There were 182 AN restrictive type and 99 AN purging type
cases. Subjects were recruited for the study after referral to Eating Disorders
treatment center (in- and outpatients, at various stages of the disease).
Diagnosis was established by experienced clinicians according to DSM-IV
criteria, with use of a semi-structured interview (Eating Disorder
Examination; 9). Cases in which AN was not the primary diagnosis or with
physical illnesses such as diabetes mellitus were excluded.
The control group consisted of a random sample of Dutch female
participants in the Utrecht contribution to the European Prospective
Investigation into Cancer and Nutrition, also known as Prospect-EPIC 10. Lack
of selection criteria is balanced by a relatively large size of this random
population sample.
Fourteen (5%) out of 281 cases and twenty (1%) out of 1656 controls
were excluded because of more than two missing genotypes. In the
remaining 267 cases and 1636 controls the mean age (SD) was 22.4 (4.3)
years and 49.0 (6.0) years, and the mean BMI was 16.4 (2.1) kg/m2 and 25.9
(4.0) kg/m2, respectively.
Genotyping call rate among successfully genotyped individuals was
98.4 %. Apart from SNP rs2844479, which was excluded from further analysis
because of significant difference in missing calls between cases and controls,
all SNPs passed the quality control requirements (more than 95% successful
calls per SNP, Hardy-Weinberg Equilibrium test (χ2; 1 degree of freedom (df))
Chapter 2
40
p > .01, minor allele frequency > .05, difference in missing calls between
groups at p>.01). To further ensure quality, blind duplicates were included on
plates (100% concordance of duplicates, excluding missed calls). Genotyping
was performed on a commercial platform (KBiosciences; Hertsfordshire,
U.K.). Statistical analyses were conducted with PLINK 11, UNPHASED 12 and
SPSS 15.0 (SPSS, Chicago, Illinois).
In the first step of analysis we performed case-control tests for
individual SNPs using a standard allelic test with 1 df. In the next step, after
having ascertained which alleles from the SNPs identified in GWA studies 1-4
were carrying the risk for higher BMI, we combined the information from 12
SNPs by counting the number of BMI-increasing alleles present in each
subject. Only subjects with 12 complete genotypes were included, i.e. 225
cases and 1351 controls. This score was entered into a logistic regression
model with case-control status as an outcome.
Figure 1. Distribution of BMI-increasing alleles in cases and controls.
Results
The analysis of individual SNPs showed an absence of association between
any of the studied markers and AN (allelic test, 1 df). Assuming a power of
80%, an α-level of .05 and a disease prevalence of 2.2 % 13, we would be able
to detect an association with odds ratio of at least 1.3 or 0.77 for a
heterozygote and 1.7 or 0.59 for a risk homozygote for most of the SNPs
0
2
4
6
8
10
12
14
16
18
20
=18
Number of BMI increasing alleles
Perc
enta
ge o
f in
div
iduals
_
cases
controls
Chapter 2
41
individually, except for rs6548238, rs1488830, rs7647305 for which the
effect sizes would have to be larger (Table I).
Performing the same analysis solely on the ANR subset yielded
nearly identical results, but with diminished power. For this reason both
subsets are taken together in the study.
To make sure that obese individuals in the control group were not relevantly
influencing results we conducted a separate single SNP analysis with obese
controls (BMI>30) excluded. Results were not materially different (data not
shown).
To test whether variants increasing BMI in the general population
play a role in AN, we entered the combined number of effect alleles (i.e.
BMI-increasing alleles from GWAS 1-4) into a logistic regression model.
Frequencies of cases and controls per number of effect alleles are shown in
Fig. 1 and the results of the logistic regression analysis are presented in Table
II.
The logistic regression analysis, assuming an α-level of .05, 80%
power and a two-sided hypothesis, would be able to detect a change from
the baseline probability (prevalence of the disease) of .02 to .05 with an
increase of one SD (SD=2.47) in a number of effect alleles.
TABLE II. Logistic regression: the number of effect alleles is not associated with
probability of being a case
Independent variable Β-coefficient df p-
value OR
95% CI
Lower Upper
Number of effect alleles .00 1 .84 1.00 .95 1.06
Β-coefficient represents a change in probability of being a case with each additional risk-
allele; OR, represents an increase in the odds of being a case with each additional risk-allele.
The analysis included only individuals with 12 complete genotypes; n cases = 225; n controls
= 1351. Mean (SD) number of effect alleles in cases = 12.20(2.32) and controls =
12.17(2.50). CI, confidence intervals.
The number of BMI-increasing alleles was not associated with the risk of AN
(p = 0.84; Table II).
Chapter 2
42
Accordingly, mean numbers of effect alleles were similar between groups
(p=.48 in a t-test) with a mean (SD) of 12.3 (2.4) alleles in cases and 12.2 (2.5)
in controls (with α=.05 and at 80% power the test would detect difference in
means of at least .18).
FIG 2. A plot showing OR for being a case along increasing number of
effect alleles. The vertical bars represent 95% CIs.
Discussion
In the present study 12 SNPs found to be associated with BMI in the general
population 1-4 were successfully genotyped in AN cases (n=267) and
population controls (n=1636). We found no evidence for association with the
risk of AN. Furthermore, by calculating the combined score of effect alleles
(i.e. BMI-increasing alleles from GWA studies of BMI 1-4) we tested whether
genetic variants increasing BMI in the general population play a role in AN.
Logistic regression model revealed absence of association between the
number of effect alleles and the risk of AN.
These results contribute to the discussion about a supposed
continuum of eating disorders, normal weight range and obesity 14,15. In our
study, on the level of genetic aetiology, AN appeared to be a discrete entity
rather than a part of this continuum.
0
0.25
0.5
0.75
1
1.25
1.5
1.75
2
2.25
=16
Number of effect alleles
Odds r
atio _
Chapter 2
43
A limitation of the current study is its relatively small sample size and
thus limited power to identify small effect sizes. Our sample has sufficient
power (80%) to identify effect sizes of at least 1.3 OR at a 5% α-level, which is
substantially larger than the effect sizes (OR 1.07 – 1.67) reported for obesity
or extreme childhood obesity in the original GWA studies 1-4. With the same
assumptions, a combined analysis of BMI-increasing alleles could detect a
change in risk of the disease from baseline of .02 to .05, with an increase of
one SD (2.47) in a number of effect alleles. To reduce phenotypic
heterogeneity, we focused solely on AN because this subtype is distinct from
the other types of eating disorders 15,16. Our main conclusions are based on
the analysis of the combined score of effect alleles which, along with the fact
that mean numbers of effect alleles between cases and controls were
remarkably similar (power in the t-test sufficient to detect a difference of at
least .18), shows that the investigated SNPs had no significant impact on
susceptibility to AN. In this study no support was found for the hypothesis
that the common genetic variants influencing BMI in the general population
are substantial risk factors of AN, suggesting that effects of those variants
may be overridden by other genetic factors of susceptibility to the disease.
However, we cannot exclude that some association might be found with a
considerable increase in sample size and refinement of phenotypes.
In conclusion, this study found no evidence that SNPs which were
previously proven to be robustly associated with BMI in the general
population protect against or contribute to the risk of AN.
Acknowledgements
We are thankful to The GIANT (Genetic Investigation of ANtropometric
Traits) Consortium for sharing the data on SNPs associated with BMI.
This work was supported by funding from the Marie Curie Research Training
Network INTACT (Individually tailored stepped care for women with eating
disorders; reference number: MRTN-CT-2006-035988).
Chapter 2
44
We thank Bobby Koeleman, Behrooz Alizadeh and Caroline de Kovel for
helpful comments.
Financial disclosures
The authors reported no potential conflicts of interests.
References
1. Willer CJ, Speliotes EK, Loos RJ, Li S, Lindgren CM, Heid IM, et al. Six new
loci associated with body mass index highlight a neuronal influence on body
weight regulation. Nat Genet 2009;41:25-34.
2. Thorleifsson G, Walters GB, Gudbjartsson DF, Steinthorsdottir V, Sulem P,
Helgadottir A, et al. Genome-wide association yields new sequence variants
at seven loci that associate with measures of obesity. Nat Genet 2009;41:18-
24.
3. Frayling TM, Timpson NJ, Weedon MN, Zeggini E, Freathy RM, Lindgren
CM, et al. A common variant in the FTO gene is associated with body mass
index and predisposes to childhood and adult obesity. Science 2007;316:889-
894.
4. Loos RJF, Lindgren CM, Li S, Wheeler E, Zhao JH, Prokopenko I, et al.
Common variants near MC4R are associated with fat mass, weight and risk of
obesity. Nat Genet 2008;40:768-775.
5. Stice E, Killen JD, Hayward C, Taylor CB. Support for the continuity
hypothesis of bulimic pathology, J Consult Clin Psychol 1998;66:784-790.
6. Steinhausen H, Grigoroiu-Serbanescu M, Boyadjieva S, Neumärker K,
Metzke CW. The relevance of body weight in the medium-term to long-term
course of adolescent anorexia nervosa. findings from a multisite study. Int J
Eat Disord 2008;9999:NA.
Chapter 2
45
7. Hebebrand J, Remschmidt H. Anorexia nervosa viewed as an extreme
weight condition: Genetic implications. Hum Genet 1995;95:1-11.
8. Coners H, Remschmidt H, Hebebrand J. The relationship between
premorbid body weight, weight loss, and weight at referral in adolescent
patients with anorexia nervosa. Int J Eat Disord 1999;26:171-178.
9. Cooper Z, Fairburn C. The eating disorder examination: A semi-structured
interview for the assessment of the specific psychopathology of eating
disorders. Int J Eat Disord 1987;6:1-8.
10. Boker LK, van Noord PA, van der Schouw YT, Koot NV, Bueno de
Mesquita HB, Riboli E, et al. Prospect-EPIC utrecht: Study design and
characteristics of the cohort population. european prospective investigation
into cancer and nutrition. Eur J Epidemiol 2001;17:1047-1053.
11. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, et
al. PLINK: A tool set for whole-genome association and population-based
linkage analyses. The American Journal of Human Genetics 2007;81:559-575.
12. Dudbridge F. Likelihood-based association analysis for nuclear families
and unrelated subjects with missing genotype data. Hum Hered 2008;66:87-
98.
13. Keski-Rahkonen A, Hoek HW, Susser ES, Linna MS, Sihvola E, Raevuori A,
et al. Epidemiology and course of anorexia nervosa in the community. Am J
Psychiatry 2007;164:1259-1265.
14. Collier DA, Treasure JL. The aetiology of eating disorders. The British
Journal of Psychiatry 2004;185:363-365.
15. Gleaves DH, Brown JD, Warren CS. The continuity/discontinuity models of
eating disorders: A review of the literature and implications for assessment,
treatment, and prevention. Behav Modif 2004;28:739-762.
16. Keel PK, Fichter M, Quadflieg N, Bulik CM, Baxter MG, Thornton L, et al.
Application of a latent class analysis to empirically define eating disorder
phenotypes. Arch Gen Psychiatry 2004;61:192-200.
Chapter 3
46
Chapter 3
Association study of POMC variants with body composition
measures and nutrient choice
Andrew Ternouth*
Marek K. Brandys*
Yvonne T. van der Schouw
Judith Hendriks
John-Olov Jansson
David Collier
Roger A. Adan
*These authors contributed equally
European Journal of Pharmacology 2011; 660(1): 220-5.
Chapter 3
47
Abstract
Genome linkage scans and candidate gene studies have implicated the pro-
opiomelanocortin (POMC) locus in traits related to food intake, metabolic
function, and body mass index. Here we investigate single nucleotide
polymorphisms at the POMC locus in order to evaluate the influence of its
genetic variance on body fat distribution and diet in a sample of middle-aged
men from the Netherlands. 366 Dutch males from the Hamlet cohort were
asked detailed questions about food choice, nutrient intake and exercise.
Furthermore, their weight and body fat composition were measured. Each
cohort member was genotyped for a set of single nucleotide polymorphisms
(SNPs) at the POMC locus. Regression analysis, adjusted for several
covariates, was used to test for association between genetic variants and the
phenotypes measured. POMC variation was associated with waist:hip ratio,
visceral fat and abdominal fat (rs6713532, P=0.020, 0.019, 0.021,
respectively), and nutrient choice (rs1042571, P=0.034), but in light of
limited power and multiple testing these results should be taken with
caution. POMC is a strong candidate for involvement in appetite regulation
as supported by animal, physiological, and genetic studies and variation at
the POMC locus may affect an individual’s energy intake which in turn leads
to variation in body composition and