The effect of polymorphisms in startle-related genes on anxiety symptom severity

Abstract

Given the limited effectiveness of treatments for pathological anxiety, there is a pressing need to identify genetic markers that can aid the precise selection of treatments and optimize treatment response. Anxiety and startle response levels demonstrate a direct relationship, and previous literature suggests that exaggerated startle reactivity may serve as an endophenotype of pathological anxiety. In addition, genetic variants related to startle reactivity may play a role in the etiology of pathological anxiety. In the current study, we selected 22 single nucleotide polymorphisms (SNPs) related to startle reactivity in the literature, and examined their association with anxiety symptom severity across psychiatric disorders (n = 508), and in a subset of patients with an anxiety disorder (n = 298). Overall, none of the SNPs pass correction for multiple independent tests. However, across psychiatric patients, rs6323 from the monoamine oxidase A (MAOA) gene and rs324981 from the neuropeptide S receptor 1 (NPSR1) gene were nominally associated with baseline anxiety symptom severity (p = 0.017, 0.023). These preliminary findings provide support for investigating startle-related genetic variants to identify biomarkers of anxiety symptom severity.

Introduction

Anxiety disorders are the most common class of mental disorders worldwide, with as many as 31% of individuals experiencing an anxiety disorder in their lifetime (Kessler et al., 2007). While anxiety disorders are heterogeneous, the hallmark characteristic linking them together is pathological anxiety symptomatology, which includes excessive apprehensive anticipation (i.e., subjective worries), physical sensations (i.e., heart palpitations, sweating), and avoidance behaviours that lead to impairments in functioning (American Psychiatric Association, 2013; Craske and Stein, 2016; World Health Organization, 1993). These anxiety symptoms are also frequently observed across psychiatric disorders, leading to worsened prognoses (Lysaker and Salyers, 2007; Novick et al., 2016; Rüsch et al., 2008; Soloff and Chiappetta, 2018; Zutshi et al., 2006). Antidepressants are the first-line pharmacological treatment for individuals suffering from anxiety, yet partial response or nonresponse to such treatment is substantial, and sustained remission from anxiety symptoms is lacking (Anderson, 2006; Farach et al., 2012). It is thus imperative that biomarkers, such as genetic polymorphisms, related to increased risk for pathological anxiety are identified, which can be used to establish novel treatment targets.

The identification of genetic factors associated with psychiatric disorders is very complicated due to the polygenic nature and heterogeneity of diagnostic phenotypes and symptomatology, as well as the fact that these disorders are diagnosed using subjective methods rather than objective biological measures (Bas-Hoogendam et al., 2016; Gottesman and Gould, 2003). One method that can aid the identification of genetic markers delineating risk for pathological anxiety is the endophenotype approach. Endophenotypes are measurable and genetically-regulated factors that link genes with disorders (Bas-Hoogendam et al., 2016). Similar to how anxiety symptoms can exist across psychiatric diagnoses, an endophenotype can be seen across different disorders, thus supporting a dimensional framework to studying psychiatric disorders instead of the traditionally used categorical framework. For a characteristic to be deemed an endophenotype of a disorder, it must fulfill four qualifications: 1) be associated with the disorder, 2) be a trait that is stable and state-independent, expressed regardless of disorder activity status, 3) be heritable, and 4) co-segregate with the disorder in families, with the endophenotype existing in non-affected family members more so than in the general population (Bas-Hoogendam et al., 2016; Gottesman and Gould, 2003). Exaggerated startle reactivity has been proposed as an endophenotype of anxiety disorders as it fulfills these four endophenotype qualifications. Specifically, exaggerated baseline startle and anxiety-potentiated startle have been evidenced in patients with anxiety disorders (Cornwell et al., 2011; Gorka et al., 2017; Grillon et al., 2017, 2008; 1994; Larsen et al., 2002; Ludewig et al., 2005; McTeague and Lang, 2012; Melzig et al., 2007; Ray et al., 2009; Shankman et al., 2013). It thus appears that all anxiety disorders are unified by the presence of an exaggerated startle response at baseline and/or when anticipating unpredictable threats that elicit anxiety states. In regard to the second endophenotype qualification, exaggerated startle is stable and tends to be present prior to disorder onset, as shown in studies where children at a high risk for anxiety disorders due to parental anxiety diagnosis display elevated startle reactivity both at baseline and during aversive anticipation (Grillon et al., 1998, 1997; Waters et al., 2008). Furthermore, startle reactivity is 50–67% heritable, which fulfills the third endophenotype qualification (Anokhin et al., 2007; Hasenkamp et al., 2010; Vaidyanathan et al., 2014). Finally, exaggerated startle is seen in unaffected siblings of children with anxiety disorders, fulfilling the last qualification (Bakker et al., 2009). However, startle reactivity is not only an indicator of anxiety disorder, but it also fluctuates depending on one's state of anxiety, where startle is reliably potentiated in healthy individuals upon exposure to fearful or anxiety-inducing stimuli (Grillon, 2008; Poli and Angrilli, 2015). Additionally, among healthy individuals, those with elevated trait anxiety have an increased baseline startle response (De Pascalis et al., 2013). Overall, given the involvement of the startle reflex in response to threats, as well as the fact that anxiety by definition is a reaction to a perceived future threat, we suggest that differences in startle reactivity between individuals can potentially identify those with an overactive threat-response system, which may be predisposing them to excessive anxiety.

Since exaggerated startle reactivity shows promise as an endophenotype related to pathological anxiety, genes associated with variations in startle magnitude should be investigated to discern the development of pathological anxiety symptomatology. Several single nucleotide polymorphisms (SNPs) have been associated with variations in startle reactivity in humans. Many of these SNPs have also been previously associated with anxiety and stress-related disorders, such as rs4680 from catechol-O-methyltransferase (COMT) (Armbruster et al., 2011; Domschke et al., 2015; Klauke et al., 2012; Lonsdorf et al., 2009; Montag et al., 2008; Wendt et al., 2015), rs878886 from corticotropin-releasing hormone receptor 1 (CRHR1) (Heitland et al., 2016, 2013), rs6323 from monoamine oxidase A (MAOA) (Larson et al., 2010), rs6265 from brain-derived neurotrophic factor (BDNF) (Hajcak et al., 2009; Lonsdorf et al., 2010; Mühlberger et al., 2014; Torrents-Rodas et al., 2012), and rs324981 from neuropeptide S receptor 1 (NPSR1) (Glotzbach-Schoon et al., 2013). However, other startle reactivity-related SNPs, such as rs1006737 from calcium voltage-gated channel subunit alpha1 C (CACNA1C) (Pasparakis et al., 2015; Roussos et al., 2011), rs10994336 from ankyrin 3 (ANK3) (Roussos et al., 2011), and rs182455 from stathmin 1 (STMN1) (Brocke et al., 2010) have not been examined for association with anxiety or stress-related disorders. The ability of startle-related SNPs to modulate anxiety symptom severity across psychiatric disorders has not been explored. In this study, we investigated the hypothesis that genetic polymorphisms related to variations in startle reactivity will be associated with anxiety symptom severity in individuals with psychiatric disorders and in a subset of individuals with an anxiety disorder.