Genomics of Anxiety Disorders

Anxiety disorders are the most prevalent psychiatric conditions in the general population. Despite the early observation of family aggregation of anxiety disorders with a heritability of 30% to 50%, their exact genetic structure is not yet determined. Evidence suggests a composition of common and rare genetic factors contributing to the etiology of anxiety disorders. Recent hypothesis-free genome-wide association studies in mega cohorts mostly with a broad anxiety phenotype rendered an increasing number of novel genetic loci. Epigenetic research is still in its infancy with first evidence showing dynamic changes in response to environmental influences and during the therapy course.

Key points

•

The genetic contribution to the development of anxiety disorders (ANX) and anxiety-related personality traits ranges from 30% to 50%.
•

SNP heritability derived from recent genome-wide studies ranges between 5% and 30%, suggesting that other factors, such as rare variants or epigenetic processes, additionally contribute to disease pathology.
•

Pharmacogenetic data for specific ANX is rare, thus, general recommendations are applied for ANX treatment in terms of drug safety regarding SSRIs, SNRIs, and tricyclic drugs.
•

DNA-methylation is the most frequently studied epigenetic mechanism in anxiety disorders so far.
•

First epigenome-wide results indicate that DNA-methylation changes can serve as putative markers for stress effects and differential clinical therapy response.

Abbreviations

BTN3A2	Butyrophilin subfamily 3 member A2
AG	Agoraphobia
ANX	Anxiety disorders
BDNF	Brain-derived neurotrophic factor
CBT	Cognitive behavioral therapy
CNTN5	Contactin 5
CNV	Copy number variation
COMT	Catechol-o-methyl-transferase
CpG	Cytosine-phosphate-guanine
CRHR1	Corticotropin-releasing hormone receptor 1
EWAS	Epigenome-wide association studies
GABA	Gamma-aminobutyric acid
GAD	Generalized anxiety disorder
GLRB	Glycine receptor beta
GLT8D1	Glycosyltransferase 8 domain containing 1
GNL3	G protein nucleolar 3
GWAS	Genome-wide association studies
HMGN1	High mobility group nucleosome binding domain 1
KCNK7	Potassium two pore domain channel subfamily K member 7
KHK	Ketohexokinase
MAOA	Monoamine oxidase A
MR	Mendelian randomization
MRPL28	Mitochondrial ribosomal protein L28
MVP	Million veteran program
NCAM1	Neural cell adhesion molecule 1
NEK4	Never in mitosis A-related kinase 4
PBK	PDZ binding kinase
PD	Panic disorder
PDE4B	Phosphodiesterase 4B
PGC-ANX	Anxiety Disorders Working Group of the Psychiatric Genomics Consortium
PRS	Polygenic risk scores
PTCH1	Patched 1
RAB27	Rab-27B
RBFOX1	RNA binding fox-1 homolog a
SAD	Social anxiety disorder
SHANK2	SH3 and multiple ankyrin repeat domains 2
SIAH3	Siah E3 ubiquitin protein ligase family member 3
SLC43A2	Solute carrier family 43 member 2
SNP	Single nucleotide polymorphism
SP	Specific phobias
SRPD	Self-report physician diagnosed
TMEM	Transmembrane protein
TNXB	Tenascin XB
ZNF823	Zinc finger protein 823

Heritability of anxiety disorders

Anxiety disorders (ANX) are the most common psychiatric conditions in the general population, with a lifetime prevalence (at least once during the lifespan) of up to 25%, and leading to high individual burden and disability rates. ^, Women are affected approximately twice as often as men for most of the ANX. The group of ANX consists of specific phobias (SP), social anxiety disorder (SAD), panic disorder (PD) with and without agoraphobia (AG) and generalized anxiety disorder (GAD). Additionally, as the lifetime perspective was considered in the DSM-V and ICD-11, selective mutism, and separation anxiety were recently added to the ANX group. Although the lifetime prevalence rates for specific ANX vary between the epidemiologic studies, SP seem to have the highest prevalence with up to 15%, followed by SAD with 3% to 13%, GAD with 3% to 6%, PD with 2% to 4%, and AG with 2%. The age of the first onset of the anxiety psychopathology occurs early in life with a peak at age 15 years; the prevalence rates usually decrease in the elderly population. ^, Mean age at onset of specific ANX differs throughout the development, starting with SP, selective mutism, and separation anxiety, followed by SAD, PD, and finally GAD. The comorbidity rate of ANX with other anxiety-related psychopathology and other psychiatric conditions is high; for example, for depression, the comorbidity rate ranges up to 70%.

The etiology of ANX is multifactorial, involving genetic and environmental factors. The genetic contribution was first investigated in family and twin studies to assess the familial aggregation and heritability of ANX. Heritabilit y estimates the degree of variation in a trait that is due to genetic variation between individuals in a population. Family and twin studies suggest an increased risk of 4 to 6 times for first degree relatives of patients with ANX to develop any anxiety psychopathology compared to first relatives of healthy individuals. The heritability is estimated to 30% to 50% across anxiety disorders. These heritability estimates for specific ANX vary between studies and show slight differences between individual anxiety disorders, for example, for PD 43%, for GAD 32%, for SP 28%, for SAD 30%, and for AG 30% to 50%. Furthermore, there is a substantial shared genetic component between anxiety disorders leading to increased risk of first-degree relatives of affected individuals to develop a wide range of pathologic anxiety phenotypes. However, existing genetic data suggest a very complex and inter-individually varying structure of genetic contribution to ANX. Additionally, genetic factors of ANX overlap with other major psychiatric conditions, specifically to a high level with depressive syndromes.

Genetics of Anxiety Disorders and Related Traits

The genetic structure of complex disorders, such as ANX, is heterogenous, likely including common and rare variants which contribute with different effect sizes to the anxiety phenotype and are distributed across the entire genome. Most of the molecular genetic studies investigating genetic factors contributing to the liability for ANX use linkage and association approaches. The earliest studies used the linkage approach and have been conducted within families to discern the chromosomal locations of risk genes for ANX transmitted from parents to the offspring. Overall, linkage studies identified promising candidate genes for anxiety phenotypes, such as regulator of G-protein signaling (intracellular signaling mediator), corticotropin-releasing hormone receptor 1 ( CRHR1 , endocrine stress system), neuropeptide y (anxiolytic effects in animal models), catechol-o-methyl-transferase ( COMT, monoamine metabolism) and gamma-aminobutyric acid ( GABA, main inhibitory neurotransmitter) receptors. A meta-analysis of linkage studies for ANX and neuroticism identified additional candidate regions and genes, such as amiloride-sensitive cation channel 1 for PD. ^, However, replication of these linkage loci in other samples was mostly not successful, most likely due to low power and differential effect sizes of familial-based transmission versus the highly polygenic structure of ANX in the general population. Therefore, linkage studies might be more relevant for a limited number of highly familial cases.

Genetic association studies investigate single nucleotide polymorphisms (SNPs) located across the entire genome. An SNP is positively associated with the disease phenotype if a certain allele or genotype differs in its frequency in affected individuals compared to controls subjects ( case-control design ). Candidate gene studies are based on a targeted approach and include variants in selected genes which are hypothesized to be involved in the pathophysiology of ANX. Candidate genes were derived from animal and human studies investigating systems involved in pharmacology or fear-related circuits with a wide spectrum of approaches, for example, serotonin/noradrenalin/GABAergic neurotransmission and its metabolisms, neuroplasticity processes, and stress-related endocrine system function. Candidate SNPs in the genes COMT , monoamine oxidase A ( MAOA, degradation of serotonin and noradrenalin ), promoter length polymorphism ( 5HTLPPR ) in the serotonin transporter gene SLC6A4, and variants in the CRHR1 gene were repeatedly analyzed with inconsistent outcomes in relation to ANX. Using meta-analysis strategies, results of existing candidate gene studies have been integrated to increase the power of individual analyses and clarify the significance of these single associations for ANX. The latter approach provided a significant meta-analytical result for the COMT gene, nominal associations were detected in the genes neuropeptide s receptor 1 (anxiolytic effects of NPS), tryptophan hydroxylase 1 ( serotonin formation) and serotonin receptor 2A. Replication attempts in general were widely unsuccessful, mainly due to inadequate sample sizes and high genetic and phenotypic heterogeneity in anxiety psychopathology. We note that prior candidate genes have not been identified in recent genome-wide association studies (see below), casting doubt on their relevance for human ANX. To improve insufficient knowledge about significant genetic targets of ANX, hypothesis-free association studies have been launched.

Genome-Wide Association Studies

Genome-wide association studies (GWAS) interrogate a large number of single nucleotide variants across the entire genome to detect associations with anxiety phenotypes or ANX. Such studies are aimed to discover new target variations and genes associated with anxiety pathology and need large sample sizes. GWAS use mainly premanufactured microarrays which have been optimized over recent years to improve genome coverage. Generally, GWAS interrogate mainly common variations with a minor allele frequency of greater than 1% in the population. Using the whole information derived from genome-wide associations, polygenic risk scores (PRS) can be derived. PRS represent the sum or collective measurement of genetic risk for a phenotype (including significant and non-significant associations) and can be applied to confer the genetic risk for a specific phenotype in independent populations or on individual level. Lastly, previous genetic results suggest that 1 variant can code for different psychiatric outcomes; this phenomenon is named genetic pleiotropy .

Extant GWAS have been conducted primarily in PD and broad anxiety phenotypes. Compared to other disorders, such as schizophrenia and depression, anxiety GWAS in mega cohorts have been available for a few years only and include mostly broad anxiety disorder phenotypes and patients comorbid with other primary psychiatric disorders, such as depression, bipolar disorder, or PTSD. Results of the larger-scale studies comparable to other psychiatric GWAS are summarized in Table 1 .

Table 1

Summary of the top results from genome-wide association studies in anxiety disorders and anxiety-related phenotypes

Genome-Wide Association Studies Anxiety Disorders and Related Phenotypes (GWAS)
Reference	Sample Size	Ancestry	Phenotype	<SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='hSNP2′>ℎ2SNP h SNP 2	SNP ID/n loci	Top Genes	Gene Function	Replication (Reference)	Phenotype
Erhardt et al, 2011	1.824	European	Panic Disorder (DSM-IV)	–	rs7309727	TMEM132D	Putative involvement neuronal sprouting	Erhardt et al, 2012	Panic Disorder (ICD-10)
Davies et al, 2015	730 twins	European	Anxiety sensitivity (ASI)	44.45%	rs13334105	RBFOX1	Regulates tissue specific alternative splicing	Otowa et al, 2016	Composite anxiety disorders (qFS)
Deckert et al, 2017	1.370	German	Agoraphobia symptoms (ACQ)	–	rs78726293	GLRB	Glycine receptor beta subunit	Deckert et al, 2017	Agoraphobia symptoms (SLC-90)
Otowa et al, 2016	21.761	European	Composite anxiety disorders (DSM-IV)	13.8%	rs1709393	LOC152225	Unknown	–	–
Otowa et al, 2016	18.186	European	Composite anxiety disorders (qFS)	9.5%	rs1067327	CAMKMT	Calcium metabolism	Hettema et al , 2 0 20	Composite anxiety disorders (SAQ, clinical self-report)
Meier et al, 2019	31.880	Denmark	Composite anxiety disorders (DSM-IV, CIDI)	28%	rs7528604	PDE4B	Signal transduction	Nagel et al, 2018	Neuroticism (EPQ-R-S, NEO-PIR)
					rs1458103	Intergenic		–	–
					rs113209956	Intergenic		–	–
Purves et al, 2019	83.566	European	Composite anxiety disorders (DSM-IV, CIDI)	26.0%	rs10809485	Intergenic		Purves et al, 2019	Neuroticism (EPQ-R-S)
					rs1187280	NTRK2	Neuroplasticity	Purves et al, 2019	Neuroticism (EPQ-R-S)
					rs3807866	TMEM106B	Cell toxicity, lysosomal enlargement	Purves et al, 2019	Neuroticism (EPQ-R-S)
					rs2861139	Intergenic		Wray et al, 2018	Depression
					rs4855559	MYH15	Motor proteins	Wray et al, 2018	Depression
Purves et al, 2019	77.125	European	Generalized anxiety symptoms (GAD-7)	31.0%	rs17189482	Intergenic		Purves et al, 2019	Neuroticism (EPQ-R-S)
Purves et al, 2019	114.091	European	Composite anxiety disorders (DSM-IV, CIDI)		rs10959577	Intergenic		Purves et al, 2019	Neuroticism (EPQ-R-S)
Purves et al, 2019	114.091	European	Composite anxiety disorders (DSM-IV, CIDI)		rs7723509	Intergenic		–	–
Levey et al, 2020	17.5163	European	Generalized anxiety symptoms (GAD-2)	5.6%	rs4603973	SATB1-AS1	Gene expression regulation in neuronal development	Nagel et al, 2018	Neuroticism (EPQ-R-S, NEO-PIR)
					rs6557168	ESR1	Estrogen related anxiety regulation	Meier et al, 2019	Composite anxiety disorders (DSM-IV, CIDI)
					rs12023347	LINC01360/LRRIQ3	Unclear	Meier et al, 2019	Composite anxiety disorders (DSM-IV, CIDI)
					rs56226325	MAD1L1	Cell cycle control	Nagel et al, 2018	Neuroticism (EPQ-R-S, NEO-PIR)
					rs6090040	OPRL1	Opioid related neurotransmission	–	–
Levey et al, 2020	24.448	African	Generalized anxiety symptoms (GAD-2)		rs575403075	TRPV6	Homeostasis in kidney, intestine	–	–
Levey et al, 2020	192.256	European	Composite anxiety disorders (clinical self-report)	8.8%	rs35546597	AURKB	Chromosome aggregation during mitosis and meiosis	–	–
Levey et al, 2020	192.256	European	Composite anxiety disorders (clinical self-report)	8.8%	rs10534613	MAD1L1	Cell cycle control	Nagel et al, 2018	Neuroticism (EPQ-R-S, NEO-PIR)
Li et al, 2023	Cases 74,973	European	Composite anxiety disorders and GAD-2 score		14 risk loci	RAB27B	Positive regulation exocytosis, multivesicular body sorting
GWAS meta-analyses	Controls 400,243	UKBB, iPsych, Angst, MVP, FinnGen				BTN3A2	Adaptive immune response
Plus functional genomics	5 European cohorts					PCLO	Synaptic function, vesicle trafficking
Plus functional genomics	5 European cohorts					CNTNND1	Dendritic spine and synapse development
Tesfaye et al, 2023	>200,000 cases	Mostly European	Generalized anxiety symptoms (GAD-2, GAD7)	5.1%	meta-analysis 11 loci CondFDR 119 loci	NPPC	preproprotein for cardiac natriuretic peptides
meta-analysis	MPV, UKBB					SATB1	Chromatin structure and gene expression
Cond FDR						FOXP2	Transcription factor regulation, language development
Cond FDR						PTCH1	embryonic development
Friligkou et al, 2024	>1.1 Mio	Cross-ancestry	Composite anxiety disorders		41 loci	CNTN5	Involved in developmental phase of the nervous system
medRxiv preprint	AD cases 97,383					KHK	Fructose metabolism
		European		2.6%–15.1%	40 loci	LINC01360	Unclear
		African			rs575403075	intergenic
Strom et al, 2024	Cases 122,225	European	Composite anxiety disorders		58 loci	GNL3	Stem cell proliferation	23andMe, African samples
PGC-ANX	Controls 729,960					GLT8D1	Unknown
Personal communication						NEK4	Cell cycle arrest in response to DNA damages, cilium integrity
						BTN3A2	Adaptive immune response
						PTCH1	Embryonic development and tumorigenesis
						NCAM1	Regulation of neurogenesis, neurite outgrowth, and cell migration
						RAB27B	Positive regulation exocytosis, multivesicular body sorting
						HMGN1	Maintain open chromatin configuration around transcribable genes