Relationship between BDNF genotype, hippocampus volume, and anxiety

Anxiety is one of the most prevalent mood disturbances in the United States. The underlying genetic and neural bases of anxiety are complex, but previous research suggests that anxiety may be associated with alterations in brain structures, such as the hippocampus, and with genetic variations within genes like that of the Brain Derived Neurotrophic Factor (BDNF). To try to better understand these factors, we examined hippocampal volume, BDNF genotype, and self-report measures of anxiety together in a single study. We hypothesized that individuals with the Met allele of the BDNF gene would have reduced hippocampal volumes and elevated anxiety scores relative to individuals with the Val/Val genotype. This hypothesis was not well supported by the data. Implications for understanding the genetic and neural bases of anxiety are discussed. BDNF genotype, hippocampus volume, and anxiety 3 Introduction Anxiety disorders are among the most common mood disorders in the United States, affecting approximately 18% of adults aged 18 and older annually, yet the origins of anxiety are still largely a mystery (National Institute of Mental Health, 2002). The Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) recognizes over ten different types of clinically relevant anxiety disorders, the most common of which include generalized anxiety disorder (GAD), simple phobia disorder, obsessivecompulsive disorder (OCD), panic disorder, and post-traumatic stress disorder (APA, 2000). These conditions can differ extensively in their presentation of symptoms, but at the most general level, anxiety is characterized by marked uncertainty, fear and uneasiness, as well as uncontrollable repetitive thoughts that activate feelings of panic and inhibit the ability to stay calm. When left untreated, anxiety disorders can often develop into more persistent conditions with poorer treatment outcomes, or may become comorbid with other mood disorders including major depressive disorder (National Institute of Mental Health, 2002). Additionally, many of the symptoms of anxiety can be experienced in a more mild range within otherwise psychologically healthy individuals, which is classified as subclinical anxiety, but nevertheless can have substantial effects on cognition, sleep and mood (Ng, Chan, & Schlaghecken, 2012; Spira et al., 2008). Although there is still considerable debate as to the specific mechanisms underlying anxiety disorders, a genetic link has been identified that may predispose individuals to higher rates of anxiety. Specifically, many studies have looked at the brainderived neurotrophic factor protein, which is more commonly referred to as BDNF, and is coded for by the BDNF gene. Within the brain, BDNF is a prominent neurotrophinBDNF genotype, hippocampus volume, and anxiety 4 signaling molecule that is vital to the differentiation, growth and survival of neurons. It is found throughout the cortex but is abundant in areas of the limbic system such as the hippocampus and amygdala, which are both involved in executive functions, including memory formation and emotion control (Huang & Reichardt, 2001). BDNF displays a single nucleotide polymorphism at codon 66 that results in a Valine to Methionine substitution known as the Val66Met polymorphism (Appendix A). The Val66Met polymorphism has been shown to influence several attributes in humans including hippocampal volume, memory performance and anxiety-related traits (Hajek, Kopecek, & Höschl, 2012; Jiang et al., 2005; Szeszko et al., 2005). Pertinent to this investigation, individuals with the Met allele of the BDNF gene show a reduced volume of the hippocampal formation compared to Val/Val individuals (Bueller et al., 2006; Hajek et al., 2012; Montag, Weber, Fliessbach, Elger, & Reuter, 2009; Pezawas et al., 2004; Szeszko et al., 2005; Yang et al., 2012). Additionally, the Val66Met polymorphism may be a risk allele for the development of anxiety as some argue that the Met allele is implicated in increased anxiety-related traits (Chen et al., 2006; Hashimoto, 2007; Jiang et al., 2005). However, others have argued the opposite, that the Val allele is more of a risk allele for increased anxiety (Lang et al., 2005) The volume of the hippocampus is an important neurological element of anxiety to investigate because the hippocampus is a major component of the limbic system, and is most well known for its role in emotion and the formation of memory (Appendix B). It is also more subject to stress-related changes than any other area of the brain (Woon, Sood, & Hedges, 2010). This is due to the fact that the hippocampus has a high level of glucocorticoid receptors to which stress hormones, such as cortisol, bind. With abnormal BDNF genotype, hippocampus volume, and anxiety 5 levels of stress or longer durations of stress exposure, the excitability of some hippocampal neurons may be decreased, which can lead to increased cell death and eventual atrophy of the brain region itself (Joëls, 2009). For this reason, patients with mood disorders, such as anxiety and depression, tend to have reduced hippocampal grey matter volume (Bremner & Narayan, 2000; Campbell & Marriott, 2004; Sheline, Gado, & Kraemer, 2004). Some investigations have explored the relationship between hippocampal grey matter volume and anxiety-related traits, with conflicting results. That is, some studies have found a positive correlation between hippocampal volume and anxiety, with higher anxiety being associated with larger hippocampi (Baur, Hänggi, & Jäncke, 2012; Rusch, Abercrombie, Oakes, Schaefer, & Davidson, 2001). Others have found the opposite, with higher anxiety and depression associated with smaller hippocampal structures (Campbell & Marriott, 2004; Kalisch et al., 2006; Kühn, Schubert, & Gallinat, 2011). The evidence of the distinct associations between genes and anxiety, genes and hippocampal volume, and hippocampal volume and anxiety, naturally leads to the investigation of how all three variables interact together. Previous studies have looked at the relationship between all three variables (Montag et al., 2009), but none have focused solely on healthy participants and their subjective experience of anxiety symptoms as they relate to BDNF genotype and hippocampal volume. Such an investigation is essential to understanding the relationship between the genetic, structural, and behavioral components of sub-clinical levels of anxiety. The purpose of this study is to investigate how a genetic polymorphism within the BDNF gene may lead to differences in hippocampal volume and vulnerability to anxiety. BDNF genotype, hippocampus volume, and anxiety 6 Therefore, the present study examines the nature of the relationship between three variables: BDNF polymorphism, subjective anxiety scores, and regional hippocampus volumes, to obtain a more comprehensive understanding of how all three variables may interact and influence one another. This study will address three distinct hypotheses: 1. The Met allele will be associated with a reduction in hippocampal volume, 2. Individuals with this reduction in hippocampal volume will show an increase in anxiety symptoms, and 3. Carriers of the Met allele will have higher anxiety. Methods Participants Sixty-two healthy University of Colorado undergraduate students (33 males; 53.2%) from the department of Psychology and Neuroscience participated for course credit or payment. All participants had no history of previous psychiatric diagnoses or medication. Informed consent was acquired according to the guidelines of the Institutional Review Board of the University of Colorado at Boulder. The final sample included two groups, Met+ (N=21 (Met/Val N=20, Met/Met=1); 52% male; Age: M=19.43; SD=1.5) and Val/Val (N=41; 54% male; Age: M=20.29; SD=6.31). The two groups did not differ in age (p>.5) or gender (p>.9). Procedure Participants first came to the Cognitive Development Center to complete a selfreport questionnaire measuring symptoms related to anxiety, and to provide a saliva sample used to identify their individual genotype. After this data was collected and BDNF genotype, hippocampus volume, and anxiety 7 analyzed, subjects were invited to participate in the second phase of the study in which an fMRI brain scan was used to obtain brain-imaging data for VBM analysis of hippocampal volumes. Genotyping All subjects provided a 3mL saliva sample, which was analyzed for BDNF alleles at The Children’s National Medical Center in Washington D.C. The Met and Val alleles were identified at codon 66 of chromosome 11. Genotype frequencies were in HardyWeinberg equilibrium (X= .688, df= 2, P>.7). Self-Report Anxiety Measurement The State-Trait Anxiety Inventory for Adults (STAI) was used to obtain a measurement of self-reported anxiety (Spielberger, 1970). The inventory consists of two scales with 20 questions each and is intended to measure current (state) anxiety and longer-term (trait) anxiety in adults (Spielberger, 2005). The STAI is used to evaluate feelings such as worry, apprehension, and nervousness. For the “state” score, participants were asked to respond to a particular sentiment in terms of how strongly they feel that emotion in that moment. For example, one statement is “I feel calm” and participants can respond with 1indicating “not at all”, 2 being “somewhat”, 3 being “moderately so”, or 4 being “very much so”. For the “trait” score, they are asked to respond to a statement with how often they feel that statement in general. For example, the statement “I feel pleasant” can have a response of 1 indicating “almost never”, 2 being “sometimes”, 3 being BDNF genotype, hippocampus volume, and anxiety 8 “often”, and 4 being “almost always”. Scores range from 20 to 80 on each scale, with a higher score indicating elevated levels of state or trait anxiety. The raw score on both the state and trait inventories were translated into a percentile score using the guidelines of the STAI manual (Spielberger & Vagg, 1984). Percentile scores indicate the standardized rank of anxiety traits among undergraduate college students as a whole. The percentile scores for both state and trait anxiety were thus the primary source of data used in this study to measure and represent relative anxiety within the undergraduate population. Overall, we were most interested in the trait measure of anxiety as it best represents an individual’s stable anxiety level as well as differences in response intensity to psychological stress among people. Due to copyright laws, we are unable to provide the STAI in an appendix. MRI and VBM analysis Voxel Based Morphometry (VBM) analyses were performed using FSL software and the processing procedure followed that established by Ashburner et al. (2000) and Good et al. (2001). VBM is a technique of neuroimaging in which the volumes of different brain areas are calculated and can then be compared, both within and between groups. This technique relies on structural brain images, which are obtained through 3DMPRAGE scans conducted within a functional MRI scan. After a brain scan from all participants was collected, the images were normalized, and the grey matter was extracted using the FSL default BET brain extraction process, to strip the skull and remove non-brain tissue from the image using the FAST4 tool. This results in grey matter images that are then aligned to MNI152 standard space using the affine registration tool BDNF genotype, hippocampus volume, and anxiety 9 FLIRT (FMRIB’s Linear Image Registration Tool) and then the nonlinear registration tool using FNIRT (FMRIB’s Nonlinear Image Registration Tool). A study-specific template is then made from averaging all of these images and then non-linearly reregistering the original grey matter images using FNIRT. To correct for possible local expansion and contraction, the registered partial volume images were modulated by dividing the Jacobian of the warp field. This results in modulated segmented images that are then smoothed by an isotropic Gaussian kernel with a sigma of 3, resulting in fullwidth half-maximum (FWHM) of 3x2.3mm=6.9mm. Using permutation-based, nonparametric testing with Monte Carlo simulations, voxelwise thresholding is applied which corrects for multiple comparisons. Finally, clusterwise correction is applied using the built-in cluster-based thresholding technique in FSL’s VBM toolbox. (Ashburner & Friston, 2000; Good et al., 2001; Smolker, Depue, Reineberg, & Banich, in progess). Through this process, every brain is compiled and normalized onto a template, which represents an average of all the volumes of each brain while also getting rid of some of the bigger differences in brain anatomy among people. When the brain images are smoothed, each voxel (or 3-D pixel) represents a volumetric average of itself and all of the voxels that surround it. From these measurements, volume can then be compared across brains at each voxel in order to determine small areas where volumes differ between groups. Total volume measurements can also be obtained, but in identifying an overall measure of an area, smaller more specific differences can often be overlooked. Masks may be drawn to narrow down analysis on one specific area of the brain, in this case the hippocampus. In this study, measures of both total hippocampus volume as well as voxel-by-voxel analyses were used to help identify differences in hippocampus BDNF genotype, hippocampus volume, and anxiety 10 volume; however, an emphasis was placed on variations in total hippocampus volume to investigate the more broad-based hypotheses as well as gain a more comprehensive generalization of hippocampus differences among people in relation to anxiety traits. Results 1. BDNF genotype and hippocampus volume There was no significant relationship between BDNF genotype and total hippocampus volume (t62)=1.470, p=0.147; Figure 1). However, in a voxel-by-voxel measurement, the Val/Val group showed significantly larger hippocampal grey matter volumes than the Met+ group in small regions of both right posterior hippocampus (t(62)=3.824, p=0.013; Figure 2) and left anterior hippocampus (t(62)=3.287, p=0.036; Figure 3). Figure 1. Effect of BDNF group on total hippocampus volume: Val/Val participants did not differ when compared to Met+ individuals in mean total hippocampal volumes. Error bars represent standard error. BDNF genotype, hippocampus volume, and anxiety 11 Figure 2. Voxel–wise analysis of grey matter volume of right posterior hippocampus: The highlighted area represents the region that is significantly larger in Val/Val groups than Met+ groups. Figure 3. Voxel–wise analysis of grey matter volume of left anterior hippocampus: The highlighted area represents the region that is significantly larger in Val/Val groups than Met+ groups. 2. Hippocampus volume and anxiety No significant correlation was found between total hippocampus volume and trait anxiety (r(62)=.120, p=0.353; Figure 4) or state anxiety (r(62)=.104, p=.421). Additionally, there was no significance between the right posterior hippocampal region and trait anxiety (r(62)=.159, p=.216, Figure 5) or the left anterior region and trait anxiety (r(62)=.132, p=.306 Figure 6). BDNF genotype, hippocampus volume, and anxiety 12 Figure 4. Relationship between total hippocampus volume and trait anxiety: Hippocampus volume does not significantly predict anxiety score. Figure 5. Relationship between volume of right posterior hippocampus region and trait anxiety: Volume of right posterior hippocampus volume does not significantly predict anxiety. BDNF genotype, hippocampus volume, and anxiety 13 Figure 6. Relationship between volume of left anterior hippocampus region and trait anxiety: Volume of left anterior hippocampus region does not significantly predict anxiety. 3. BDNF genotype and anxiety No significance was found between BDNF genotype and anxiety during the first phase of the study (Trait: t(62)=-1.353, p=0.181, Figure 7; State: t(62)=-.974, p=.334). However, during the second phase of the study at the time of fMRI scanning, the Met+ group showed lower state anxiety than the Val/Val group (t(52)=2.475, p=0.017; Figure 8). BDNF genotype, hippocampus volume, and anxiety 14 Figure 7. Effect of BDNF genotype on trait anxiety: Genotype group does not significantly predict anxiety. Error bars represent standard error. Figure 8. Effect of BDNF genotype on anxiety: The Met+ group showed significantly lower state anxiety than the Val/Val group at the time of fMRI scanning. Error bars represent standard error. BDNF genotype, hippocampus volume, and anxiety 15 4. BDNF genotype by total hippocampus volume interaction Overall, the interaction of BDNF genotype and total hippocampus volume was not significantly related to trait anxiety (r(62)=0.035, p=0.785). Additionally, no significant relationship was found within individual BDNF genotype groups of total hippocampus volume on trait anxiety (Val/Val: r(62)=0.720, p=0.580; Met+: r(62)=0.089, p=.235, Figure 9). Figure 9. Effect of Hippocampus Volume X BDNF interaction on trait anxiety: Interaction of hippocampus volume and BDNF genotype does not significantly predict anxiety. BDNF genotype, hippocampus volume, and anxiety 16 5. BDNF genotype by total hippocampus volume group interaction To further explore this interaction, one other approach was taken whereby each participant’s total hippocampus volume was categorized into either “high” or “low” volume categories depending on if it was above or below the average volume for all participants. The hippocampus volume and genotype classifications were then combined and the effect of group on anxiety was noted. The interaction was not significant for trait anxiety (t(62)= -1.498, p=0.139, Figure 10), although the groups show an interesting, opposing pattern depending on genotype. That is, individuals with the Val/Val genotype and a low hippocampus volume classification exhibited the most anxiety whereas the Met+ carriers with a low hippocampus volume classification exhibited the lowest levels of anxiety. Figure 10. Effect of BDNF genotype X hippocampus volume group designations: BDNF by hippocampus group interaction does not significantly predict anxiety though relationships are opposing depending on genotype group. BDNF genotype, hippocampus volume, and anxiety 17