Eating Disorders and Obesity in Children and Adolescents

Johannes Hebebrand, Beate Herpertz-Dahlmann

• Publisher: Elsevier
• Year: 2018
• ISBN:
  9780323548533
• Topic: Médecine
• Subtopic: Pédiatrie

Get a quick, expert overview of best practices for diagnosis and treatment of eating disorders in children and adolescents. This concise resource by Drs. Johannes Hebebrand and Beate Herpertz-Dahlmann provides psychiatrists and pediatricians with current information in this increasingly important field, including practical sections on developmental aspects of eating disorders, symptomology, epidemiology, etiology and pathyphysiology, treatment and outcomes, and prevention.

• Discusses general concepts for feeding, eating, and weight disorders; body weight and composition, appetite regulation, and the emergence of body perception and image.
• Covers genetics of eating and weight disorders, influence of hormones, intergenerational effects, and food addiction.
• Includes information on cognitive behavioral therapy, family-based therapies, early intervention, pharmacotherapy, bariatric surgery, and other treatments.
• Consolidates today's available information on this timely topic into a single convenient resource.

Section VI

Etiology and Pathophysiology

Outline

Chapter 11. Genetics of Eating and Weight Disorders

Chapter 12. Influence of Hormones on the Development of Eating Disorders

Chapter 13. Endocrine Mechanisms in Obesity

Chapter 14. Microbiome and Inflammation in Eating Disorders

Chapter 15. Eating Disordered Mothers and Their Children: Intergenerational Effects

Chapter 16. An Addiction Perspective on Eating Disorders and Obesity

Chapter 17. Stigmatization Associated With Obesity in Children and Adolescents

Chapter 11

Genetics of Eating and Weight Disorders

Anke Hinney, PhD, and Johanna Giuranna, msc

Abstract

Human obesity is only infrequently caused by single gene mutations. Accordingly, the genetic predispositions to obesity and eating disorders (e.g., anorexia nervosa [AN]) are polygenic for most affected individuals. Genome-wide association studies (GWASs) on more than 700,000 population-based individuals have been performed for body mass index (BMI); more than 3000 patients with AN were included in a GWAS for this eating disorder. 716 polygenic loci associated with BMI/obesity have been identified; for AN, one polygenic locus has been described so far. The effect size of each single polygenic locus is small. Cross-trait analyses have identified regions/genes that were not detected by single-trait analyses. Subsequent functional studies of the GWAS-derived polygenes aim to improve the understanding of the biological mechanisms involved in eating and weight disorders.

Keywords

Cross-disorder; Cross-phenotype; Genome-wide association study; Leptin; Melanocortin

Introduction

Monogenic forms of obesity are mostly the result of mutations in genes of the leptinergic-melanocortineric system.¹ The absence or reduced function of a gene product in this system can lead to early-onset extreme obesity. However, these variants are infrequent to rare and do not explain a large part of the high heritability of body mass index (BMI). For eating disorders, monogenic forms have not been described. Heritability estimates derived from twin studies suggest that genetic variation accounts for 50%–70% of the variance in human BMI. Twin studies for anorexia nervosa (AN), bulimia nervosa (BN), and binge eating disorder (BED) have also provided evidence for substantial heritability for the respective eating disorders.² The largest part of the heritability for eating and weight disorders is attributable to polygenic factors, which have become detectable by GWAS.^3–6 Such GWASs or metaanalyses thereof have been conducted for BMI, obesity, fat mass, fat-free mass, and AN. The variants (alleles) predisposing to an elevated BMI or high fat mass are identified more frequently in subjects with obesity. In AN, only a single locus has been detected; it is a matter of time until the analysis of larger sample sizes will entail the identification of further loci.^3,4,6 The respective alleles can be identified and validated as BMI/obesity or AN risk alleles by statistical analyses only with sample sizes being crucial for achievement of significant results.¹

GWASs have proven extremely successful for complex disorders.^3–6 Major technological advances have made single nucleotide polymorphism (SNP)-based GWAS feasible and led to the identification of currently more than 50,000 loci for numerous disorders and phenotypes (http://www.ebi.ac.uk/gwas/). Within a brief period of time, GWAS revolutionized the molecular genetic analyses of complex phenotypes. Assuming an average number of 1,000,000 analyzed SNPs per individual, Bonferroni correction for multiple testing results in the P-value threshold of p ≤ 5 × 10^–8, which has been accepted as the gold standard to circumvent the problem of multiple testing.⁷ Unfortunately, even upon availability of large total sample sizes in the hundreds of thousands, a large number of potentially truly associated SNPs cannot be identified due to this stringent threshold.⁸ On the positive side, almost all SNPs with a p ≤ 5 × 10^–8 continue to be associated with the phenotype in question upon enlargement of the original sample size.

“Polygene” is a term used for an etiologically relevant gene in a chromosomal locus that harbors intragenic or extragenic interindividual sequence variations accounting for a small fraction of the variation of a specific quantitative trait. GWAS based on SNPs are used to detect polygenic loci. Such a polygenic locus usually encompasses more than one gene. As such, the detection of the relevant gene(s) in a locus detected via GWAS remains a complex and time-consuming task. The additive effect of such polygenes has been shown to explain almost 25% of the variance of body height.⁵ For most phenotypes, however, the explained variance is at or below 5%. The explained variance crucially depends on sample size, heritability, and the genetic architecture of the respective phenotype.

A GWAS is followed-up by extensive analyses with the aim to detect the underlying gene. As a first guess the gene closest to the GWAS lead SNP (the SNP with the lowest P-value in the region) is usually regarded as a candidate gene for the analyzed trait. For obesity, only 2 of 97 GWAS hits³ were located in the coding regions of genes. An example for the long road from a GWAS hit to the functionally relevant target gene is the fat mass and obesity-associated (FTO) locus. The first BMI/obesity GWAS in 2007 identified a chromosomal region that harbors the FTO gene.⁹ It is currently still not known whether FTO is the gene underlying the association signal. Recent data provide impressive evidence for a major importance of genes downstream of FTO.¹⁰

Major Findings

Obesity

More than 700 polygenic loci relevant for body weight regulation have been identified in the most recent metaanalysis of GWAS for BMI, the results of which have preliminarily published in a non-peer reviewed article. This recent update of previous GWAS metaanalyses of the Genetic Investigation of ANthropometric Traits consortium (GIANT; http://portals.broadinstitute.org/collaboration/giant/index.php/GIANT_consortium) combined the analysis of N∼700,000 individuals. A total of 716 near-independent SNPs associated with BMI, upon use of a very strict genome-wide significance threshold of p < 1 × 10^–8, were identified; 554 of these were novel. These genome-wide significant SNPs explained ∼5% of the variance of BMI in an independent sample.⁵ The interindividual heterogeneity is pronounced and implies that a specific set of polygenic variants predisposing to obesity in any one individual will rarely overlap with those of an independent obese subject.^1,11

The largest GIANT GWAS for BMI, that was published as a peer-reviewed article, comprised 339,224 individuals and identified 97 BMI-associated loci, 41 of which had been detected in previous GWAS. These loci accounted for ∼2.7% of the BMI variation. Based on the central expression of most of the genes in the respective loci, a major role of the central nervous system was postulated. The mean additive effect of the lead SNP risk alleles at the 97 BMI loci in 8164 individuals of European descent averaged 0.1 BMI units (kg/m²). Therefore, homozygous and heterozygous carriers of a body height in the range of 160–1...