Abstract
The major anxiety disorders (ANX; including generalized anxiety disorder, panic disorder, and phobias) are highly prevalent, often onset early, persist throughout life, and cause substantial global disability. Although distinct in their clinical presentations, they likely represent differential expressions of a dysregulated threat-response system. Here we present a genome-wide association meta-analysis comprising 122,341 European ancestry ANX cases and 729,881 controls. We identified 58 independent genome-wide significant ANX risk variants and 66 genes with robust biological support. In an independent sample of 1,175,012 self-report ANX cases and 1,956,379 controls, 51 of the 58 associated variants were replicated. As predicted by twin studies, we found substantial genetic correlation between ANX and depression, neuroticism, and other internalizing phenotypes. Follow-up analyses demonstrated enrichment in all major brain regions and highlighted GABAergic signaling as one potential mechanism underlying ANX genetic risk. These results advance our understanding of the genetic architecture of ANX and prioritize genes for functional follow-up studies.
Introduction
Fear and anxiety are characterized by a set of evolutionarily conserved physiological and behavioral responses to threats that are critical for survival. Thus, anxiety disorders (ANX) may represent dysregulation of the complex brain circuits underlying the human threat-response system. While perturbations in various neurotransmitter systems such as serotonin or gamma-aminobutyric acid (GABA) have been proposed as a basis of their etiology, no reliable biomarkers have yet been identified (1). The major ANX, including generalized anxiety disorder, panic disorder, and phobias (specific phobia, social phobia, and agoraphobia), represent different clinical presentations of that underlying common diathesis (2–4). Up to 25% of the population develop an ANX sometime during their lifetime (5–7). These tend to onset early in life, are persistent, and are highly comorbid with other psychiatric conditions for which they often present as a predisposing risk factor, e.g., major depressive disorder (MDD) and substance use disorders (6,8–10). ANX are also associated with other medical conditions such as neurological, cardiovascular, gastrointestinal disorders, and cancers (11–13). They may worsen the course of infections such as COVID-19 (14), thereby lowering quality of life and increasing mortality in affected individuals (15). These features make ANX a leading source of world-wide disability (16).
Each ANX aggregates in families (odds ratio 4-6) primarily due to genetic risk factors (17). Estimates from twin studies indicate that ANX are moderately heritable (h2 = 30-50%, (2,17)), similar to other common psychiatric disorders like MDD but lower than less prevalent disorders like schizophrenia and bipolar disorder. Different ANX exhibit overlapping clinical features and exhibit strong lifetime comorbidity with each other likely due to shared genetic susceptibility (17–19) and other common risk factors such as childhood maltreatment (20–22). Research implicates genetic and environmental risk factors that affect the structure and functional capacity of brain networks/circuits involved in emotion and cognition (23–25). Twin studies report substantial genetic correlations between ANX and other psychiatric conditions, particularly MDD (26), helping to explain their high comorbidity. In addition, ANX and depression both share genetic risk factors with heritable personality traits such as neuroticism (27,28). Anxiety symptoms often precede suicidal behaviors (29) with possible causal implications (30). Thus, examining the genetic relationship between ANX and related phenotypes on the internalizing spectrum is essential.
The combination of high prevalence, extensive comorbidity, and high polygenicity makes it particularly difficult to identify genetic variants underlying risk for ANX. Prior genome-wide association studies (GWAS) have identified a handful of genetic loci, however, with inconsistent results (31–36). A recent meta-analysis using five publicly available datasets reported ten additional novel associations (37). Genome-wide single nucleotide polymorphism (SNP)-based heritability estimates range from 10-28%, supporting that ANX (like other psychiatric disorders) have a polygenic basis. As for other complex genetic phenotypes, sufficiently large samples are required to achieve the necessary power to detect the small effects of common variants. Consistent with twin studies, previous psychiatric GWAS have demonstrated that ANX polygenic risk is highly correlated with that of MDD and neuroticism (38–42).
Here we present the first GWAS meta-analysis from the Anxiety Disorders Working Group of the Psychiatric Genomics Consortium (PGC-ANX) consisting of 122,341 individuals diagnosed with any ANX and 729,881 controls, all of European ancestry (EUR). We analyzed the data at the level of variant, gene, pathway/gene set, and tissue by using both functionally-informed and functionally-agnostic methods. Subsequently, these results were compared with those of other phenotypes and investigated for possible molecular mechanisms and avenues for drug repurposing.
Results
(i) GWAS meta-analysis
Implicates 58 independent ANX susceptibility loci.
We present the results of a GWAS meta-analysis of 36 case-control cohorts from Europe, the United States, Australia, and New Zealand (122,341 ANX cases and 729,881 controls; Supplementary Table S1). Details about phenotype, quality control, and GWAS analysis for each individual cohort are described in Supplementary Note 2. Among the 7.2 million autosomal SNPs analyzed, we identified 58 independent, genome-wide significant (GWS) SNPs associated with ANX. Figure 1 provides the Manhattan plot and Table 1 a list of all significant lead SNPs. We provide further information in Supplementary Table S2 and in Supplementary Figure S1 (quantile-quantile (QQ) plot), Supplementary Figures S2-S56 (regional association plots of each significant SNP and forest plots indicating each individual cohort’s effect size). Estimates of the genomic inflation factor (λ = 1.41, λ1000 = 1. 00) and the LD score regression (LDSC) intercept and attenuation ratio (intercept = 1.05, SE = 0.01; ratio = 0.082, SE = 0.014) suggest that inflation was likely caused by polygenicity and not by cryptic population structure. LDSC estimates a SNP-based heritability of 10.1% (SE = 0.004) assuming a 20% population prevalence.
Index SNP = rs number of variant; CHR = chromosome; BP = base pair position (hg19); OR = odds ratio for allele 1; SE = standard error; A1/A2=allele 1 and allele 2; Freq Cases = Frequency of A1 in cases; Freq Controls = Frequency of A1 in controls; Closest Genes (distance kb): Closest genes to the SNP with distance in kilobytes in parentheses (if the SNP lies within the gene, no distance is given).
Manhattan plot of the main ANX GWAS (122,341 ANX cases and 729,881 unaffected controls) showing 58 genome-wide significant loci. The x-axis shows the position in the genome (chromosome 1 to 22), the y-axis represents –log10 p-values for the association of variants with ANX from the meta-analysis using an inverse-variance weighted fixed effects model. The horizontal red line shows the threshold for genome-wide significance (P = 5×10−08). Dots represent each SNP that was tested in the GWAS with a green diamond indicating the lead SNP of a genome-wide significant locus and green dots below representing SNPs within the locus with high levels of LD with the lead SNP.
A series of sensitivity analyses revealed no substantial genome-wide heterogeneity across the 36 cohorts. This included a heterogeneity GWAS of Cochran’s Q (Supplementary Figure S57) and I2 statistics (in forest plots Supplementary Figures S2-S56). We subdivided our study cohorts into five sub-groups based on their ascertainment or ANX assessment and performed subgroup specific GWASs (Manhattan- and QQ-plots in Supplementary Figures S58-S62). Within these five groups, diagnoses were respectively determined by 1) a health-care professional in a clinical setting (Clinical), 2) health records or biobanks (Biobanks), 3) either Clinical or Biobanks, but the sample was ascertained for another comorbid psychiatric disorder (Comorbid), 4), population-based or volunteer samples (Community), and 5) self-reported professional diagnosis (SRPD; see details in methods and Supplementary Table S1). While some cohorts meet criteria for more than one subgroup, they were grouped according to their main defining characteristic (i.e., if a biobank sample relied on self-reported diagnosis, it was categorized into SRPD rather than Biobanks). The subgroup specific GWASs identified four (SRPD), three (Biobanks), one each (Comorbid, Community), and zero (Clinical) independent significant loci, respectively. Of those nine identified loci, six were significant in the main meta-analysis (Supplementary Table S3). Assuming a population prevalence of 20%, the subgroups provided a range of estimated SNP-based heritabilities from 6.9% (95% CI = 0.034; Community) to 23.7% (95% CI = 0.12; Clinical; see Supplementary Table S4A for heritability estimates with other prevalences). Genetic correlations (rG) among the sub-groups ranged from 0.63 (SE = 0.21, between Clinical and Community) to 1.00 (between Community and both Comorbid and SRPD; see Supplementary Table S4B for all correlation estimates). Conducting a confirmatory factor analysis in GenomicSEM (43) with the five sub-groups as indicators resulted in a one-factor solution with an excellent fit (ChiSq = 1.92, df = 5, Pchisq = 0.86, CFI = 1, SRMR = 0.04). The loadings across all five subgroups were high and similar with the factor explaining 81.8% of the total genomic variance (see Supplementary Table S5 and Supplementary Figure S63). This supports our hypothesis that the genetic association effects were generally consistent across samples and study designs and tapped into a common underlying ANX genetic vulnerability.
(ii) Replication and validation of GWAS SNPs
51 of the ANX-associated loci are replicated in an independent EUR sample; we further found significant polygenic overlap across European and South Asian ancestries.
We conducted two replication analyses of the 58 significant loci: (1) in a large independent European ancestry ANX GWAS from 23andMe, Inc, and (2) in an African-American ancestry ANX GWAS from the VA Million Veteran Program (MVP). The 23andMe sample consisted of 1,175,012 ANX self-report cases and 1,956,379 controls (see Methods for details). Among the 58 SNPs identified in the discovery GWAS, all but one (rs7121169) were available for replication testing in the 23andMe genotype platform. Two additional variants failed QC procedures (rs72704544 and rs11599236). Considering the remaining 55 loci tested, all showed the same direction of effect as the primary GWAS and 51 were significant at a Bonferroni-corrected p-value of P = 0.0009 (0.05/55) (Supplementary Table S6). At the time of this analysis, only MVP had published an ANX GWAS in a reasonably sized non-European ancestry sample (military ascertainment, African Americans, MVP-AFR, 5,664 cases and 26,410 controls) (34). Analyzing those data, we compared the direction of effect and p-values of association for our 58 lead SNPs to examine consistency with our EUR results (Supplementary Table S7). Among the 53 SNPs available in MVP-AFR, only 27 (50.9%) showed the same sign, consistent with chance. Given differences in linkage disequilibrium and allele frequency between European and African genomes, we also searched for the most significant SNP in a 50kb window in MVP-AFR around each lead SNP. Thirty-six of these SNPs were nominally associated but only two significantly after adjustment for multiple testing.
We compared our associations with those reported in previous ANX case/control GWASs (31–34,37) (Supplementary Table S8). A recent GWAS using broader anxiety-related case/control and symptom-based phenotypes reported 40 European-ancestry significant SNPs (44); all but one showed the same direction of effect while ten were also GWS in our analysis. Importantly, most of the associations in our GWAS are novel discoveries with a total of 15 reported in these prior GWASs. We note that some of the previously identified SNPs are in LD with each other, and all previously published ANX GWASs partially overlap with our samples. Therefore, these are not independent replications but demonstrate the consistency of results when additional samples are incorporated.
To study the generalizability of our results across different ancestral groups, we tested the extent to which polygenic risk scores (PRS) derived from our GWAS (excluding any UK datasets) predicted ANX in the UK Biobank for participants of European, African, and South Asian ancestry (see Supplementary Table S9). The PRS predicted 2.27% of the variance (P < 2.0 × 10−16) in ANX liability for those of European ancestry assuming a prevalence of 20%. The variance explained for those of South Asian and African ancestries was 1.94% (P = 6.37 × 10−5) and 0.54% (P = 0.051), respectively, revealing significant polygenic overlap across European and South Asian ancestries.
(iii) Characterization and functional annotation of GWAS SNPs
Genetic variation associated with ANX is enriched only in brain and linked to GABAergic neurons
To identify potential causal variants, we conducted statistical fine mapping of our GWS loci using FINEMAP v.1.3.1 (45). This identified six credible SNP sets defined as having a posterior probability > 0.95 which was set to avoid excessive false positive rates (Supplementary Table S10). The lead SNPs of these credible sets were located at the following chromosomal positions: 3:67895104 (within SUCLG2-GT gene), 10:104654873 (within SORCS3), 17:8187590 (near TRI-AAT-5), and 20:20876379 (near KIZ), and two within the MHC region 6:28329086 (within ZSCAN31), 6:30170699 (within TRIM15).
To examine the biological relevance of our GWS SNPs, we performed functional annotation in FUMA v1.6.1 to link our GWS SNPs with expression quantitative trait loci (eQTL) and brain chromatin interaction (HiC) data. The results suggest that most of our associated loci are supported by some degree of functional significance in the form of gene regulatory mechanisms such as eQTLs and chromatin interactions (circos plots, Supplementary Figures S64-S83).
We conducted stratified LDSC to partition the heritability by different functional genetic annotations and cell types. As noted in Supplementary Table S11, the association signal is highly conserved across species and significantly enriched for introns, monomethylated and polyacetylated histone marks (H3K4me1 and H3K4ac), and DNase I hypersensitivity sites in both adult and fetal tissues. Similar to other GWASs, our findings are also enriched in certain non-coding regions rather than coding regions. Cell-type specific enrichment was observed for CNS structures only including multiple cortical and subcortical areas as well as cervical spine.
We also examined whether genetic associations with ANX were enriched among transcriptomic profiles of human tissues and/or individual cell types in FUMA v1.6.1. Tissue enrichment analyses for general tissue types using data from the GTEx (v8) consortium suggested that the expression patterns related to brain and pituitary tissues were significantly associated with the genetic risk of ANX (P = 1.18 × 10−13 and P = 6.50 × 10−05, respectively; Supplementary Table S12A, Supplemental Figure S84). All individual brain tissues showed significant enrichment (Supplementary Table S12B and Supplementary Figure S85) with cortex overall (P = 2.62 × 10−12) as well as frontal and anterior cingulate cortices and nucleus accumbens as most significant. At the level of individual cell-types, we found consistent association of GABAergic neurons with genetic variation associated with ANX (Supplementary Figure S86). Our strongest association was found with GABAergic neuroblasts (via GSE 76381 (46); P = 3.24 × 10−08).
(iv) Gene-based association and enrichment
66 genes with multiple sources of support
To identify putative genes associated with ANX, we applied MAGMA v1.08 (47) implemented in the FUMA v1.6.1 (48) pipeline with standard settings. We identified 91 significantly associated genes (adjusted P < 0.05/18,490 = 2.7 × 10−6; Supplementary Table S13). Historically interesting candidates include CLOCK, GABBR1, PCLO, NCAM1, and DRD2.
To test whether our loci produce significant functional changes, we used summary-data based mendelian randomization (SMR) (49) to conduct transcriptome-, proteome-, and methylome-wide analyses (T-SMR, P-SMR, M-SMR) (49). We used the largest available eQTL, pQTL, mQTL reference data sets, respectively, for both brain and blood tissues (Supplementary Table S14). By using the conservative p-values adjusted for HEIDI test (see Methods), we detected 27 Bonferroni-corrected significant genes/isoforms in the brain to be associated with changes in the methylome, 16 in the transcriptome, and seven in the proteome (Supplementary Tables S15 to S17). Increased ANX risk was associated with increased expression of BTN3A2, NEK4, and SLC12A5 and decreased expression of PSMG1. To improve signal detection in brain transcriptome and methylome data, we used Primo (50) to jointly analyze blood and brain statistics (see Gedik et al. (51)). We did not jointly analyze proteome data due to their rather low number of brain probes. These between-tissue concordance analyses yielded 22 significant ANX signals (posterior probability > 0.95) for the transcriptome and 133 for the methylome (Supplementary Tables S18 and S19). BTN3A2 remains a leading signal in both analyses, and interesting subthreshold genes from single tissue analyses become very strong findings in the joint T-SMR (ZDHHC5, FURIN, and NEGR1).
To highlight genes for which there was strongest support, we summarize the findings across multiple analyses within Supplementary Table S20. This table includes an expanded set of 151 genes associated with ANX susceptibility. Starting with the 91 significant associations from MAGMA, we added genes supported by joint T-SMR or joint M-SMR with posterior probability > 0.95. We annotated these using additional support from P-SMR, eQTL, and HiC data. Figure 2 lists the 66 genes with three or more sources of support (score ≥ 3). Most of these have prior reported associations with one or more psychiatric phenotypes suggesting gene-based pleiotropy, while a small proportion appear specific to ANX risk (reviewed in Discussion).
List of 66 of the most highly-supported ANX genes that were implicated in at least three of the five SNP-(eQTL, HiC) or gene-based (MAGMA, M-SMR, P-SMR, T-SMR) tests. The left side indicates the position of the gene in the genome. Significance is indicated by a coloured dot. eQTL (blue dots) comparing results from brain-related eQTL studies for overlap in significance between our GWAS and the eQTL studies. HiC (green dots) used brain-related HiC information available through FUMA to functionally annotate our results. MAGMA (gray dots) tested genetic associations at the gene level for the combined effect of SNPs in or near protein coding genes. M/P/T-SMR (yellow, red, and pink dots, respectively) refers to transcriptome/proteome/methylome-wide analyses that assessed likely causal association between traits and gene/proteins/genomic regions by inferring the association between trait and gene expression/protein concentration/methylation, as predicted from genomic data.
To test whether pre-existing gene sets are enriched for our ANX risk loci, we examined 10,894 gene sets obtained from MsigDB v5.2 (Curated gene sets: 4728, GO terms: 6166). Specifically, we used MAGMA to test for enrichment of our ANX signals (see Supplementary Table S21). Overall, one gene set was significant after correction for multiple testing: dawson_methylated_in_lymphoma_tcl1 (P = 1.71×10−06) including 57 genes which are hypermethylated in at least one of the lymphoma tumors in transgenic mice overexpressing TCL1 in germinal center B lymphocytes; the top three genes were also supported by T-SMR or M-SMR (NCAM1, HMGN1, and ZDHHC5). Among the most highly associated sets were genes related to commissural neuron axon guidance (P = 5.24E-5) and GABAergic synapse (P = 9.67E-5), the latter with 66 genes including GABBR1, DRD2, CDH13, and LRFN5.
(v) Gene-drug associations
Existing anxiolytic agents interact with ANX risk genes
To reveal possible drug repurposing opportunities for ANX, we used DrugTargetor (52) (v.1.3) with our main ANX summary statistics. Among the 161 drug classes analyzed, several already successfully used for ANX treatment demonstrated significant associations (q-valueBF < 0.05, Supplementary Table S22): psycholeptics (drugs with a calming effect) and psychoanaleptics (mostly antidepressants) as well as other sedating drugs like antihistamines, antipsychotics, general anesthetics, and opioids. However, none of the more than 1500 individual compounds cataloged in ChEMBL (53) and DgiDB (54) yielded a significant signal (Supplementary Table S23) possibly due to the moderate power of this GWAS.
(vi) Genetic overlap between ANX and other phenotypes
43 of the significant SNPs are pleiotropic; ANX is significantly genetically correlated with 82/112 phenotypes; bidirectional effects between ANX and MDD; posttraumatic stress disorder (PTSD) and neuroticism more strongly affect ANX than vice versa
To examine overlap between our ANX association signals with other phenotypes at the locus level, we conducted a phenome-wide association study (PheWAS). Of the 58 SNPs significantly associated with ANX, 15 were deemed ANX specific from these data (red diamonds; Figure 3A), i.e., variants not reported as GWS in other extant, published GWAS. A total of 43 pleiotropic variants were associated with at least one other phenotype. We note that the higher number of pleiotropic associations with cardiometabolic, hematological, and immunological outcomes reflect both the robust genetic architectures of these phenotypes and the number of GWASs that have been published in these domains. ANX-related pleiotropy with cardiometabolic and hematological traits was heavily skewed towards a subset of the variants (rs2710323, rs58825580, and rs174560). Figure 3B depicts a dendrogram-based heatmap showing the association with psychiatric or personality traits among 24 pleiotropic SNPs (other heatmaps for Cognitive and Behavioral domains are found in Supplementary Figure S87 and S88). Not surprisingly, there are far more ANX SNPs overlapping with internalizing phenotypes (neuroticism, depression) than with psychotic disorders (schizophrenia, bipolar disorder).
The (rotated) Manhattan plot of the -log10 p-values of the ANX meta-analysis (left) and PheWAS alluvial plot of the pleiotropic variants (right). The colored ribbons depict variants that are pleiotropically associated with at least one other published GWAS finding and correspond with the color of the ribbon in the alluvial plot. The red diamonds in the Manhattan plot depict the most significant variant in the region corresponding with potentially ANX specific SNPs, i.e., a variant that reached the genome-wide significance threshold for ANX but not in any other published GWAS.
Dendrogram-based heatmap indicating the numbers of unduplicated reports of genome-wide psychiatric or personality associations among 24 pleiotropic SNPs. Shading indicates the number of GWAS reporting associations between a specific SNP and the outcomes. Symptom Dimensions (mood disturbance, mania, psychosis) and Self-Report Professional Diagnoses (depression, anxiety, distress) are from the UK Biobank.
To estimate the genetic correlations between ANX and a wide variety of other traits, we employed bivariate LDSC. We included 112 previously published GWASs on various traits including psychiatric, substance use, cognition/SES, personality, psychological, neurological, autoimmune, cardiovascular, anthropomorphic, dietary, and fertility phenotypes. After a false discovery rate (FDR) correction for multiple testing, we found that 82 traits showed a significant genetic correlation with ANX (Figure 4 and Supplementary Table S24). Among the psychiatric disorders and traits, ANX showed the strongest correlations with MDD (rG = 0. 91) followed by childhood internalizing symptoms (rG = 0. 76), mood disturbance (rG = 0. 76), symptoms of depression (rG = 0. 71), PTSD (rG = 0. 71), psychosis (rG = 0. 68), mania (rG = 0. 66), suicide attempt (rG = 0. 58), and obsessive-compulsive disorder (rG = 0. 41). Genetic correlations were also high with total neuroticism score (rG = 0. 70) and its various clusters and items. P-values of these correlations were all highly significant (see Supplementary Table S24). We found somewhat lower correlations with other psychiatric and substance use disorders. ANX genetic risk was also modestly correlated with that of several neurological disorders as well as adult-onset asthma and heart disease (positive) and inflammatory bowel diseases (negative). As shown in Supplementary Figure S89 and Supplementary Table S24, the different ANX data sub-groups show a variable but overall similar pattern of correlations.
Genetic correlations (rg) between the main ANX GWAS and 112 psychiatric, substance use, cognition/socioeconomic status (SES), personality, psychological, neurological, autoimmune, cardiovascular, anthropomorphic/diet, fertility, and other phenotypes. References of the corresponding summary statistics of the GWAS studies can be found in Supplementary Table S24. Bars represent 95% confidence intervals, red circles indicate significant associations after FDR correction for multiple testing. Black circles indicate associations that are not significant after FDR correction.
These results highlight the complex interrelations between the three internalizing phenotypes that also have the highest genetic correlations with ANX: MDD (55), PTSD (56) and neuroticism (39). To examine potential directional effects underlying these genetic correlations, we applied bi-directional GSMR (57) with the latest available GWAS summary statistics. These results (Supplementary Table S25) indicate a highly significant bi-directional effect between ANX and each of these phenotypes. Based on beta-values, the strength of reverse (MDD → ANX = 0.657) and forward (ANX → MDD = 0.545) effects are similar between ANX and MDD. However, both PTSD (PTSD → ANX = 0.891 vs. ANX → PTSD = 0.239) and neuroticism (neuroticism → ANX = 1.25 vs. ANX → neuroticism = 0.17) effect on ANX is stronger than the reverse.
Discussion
In this ANX GWAS meta-analysis, we identified 58 independent genome-wide associated loci. We integrated data from 36 cohorts with 122,341 ANX cases and 729,881 controls (Neffective= 390,560). Three-quarters of the identified variants are novel discoveries with only fifteen reported in prior anxiety GWASs. Fifty-one of these SNP associations were replicated in an independent EUR-ancestry sample from 23andMe, providing strong confidence in their relevance. These results represent a major advance in identifying validated susceptibility loci for anxiety disorders.
The SNP-based heritability estimated at 10.1% captures approximately one quarter of the broad-sense heritability reported in twin studies of adult ANX (17) consistent with findings across many similar complex traits like MDD (40). To further probe the heritability of ANX, we divided our cohorts into several sub-groups based upon similarities in ascertainment and conducted separate GWAS as a sensitivity test. We observed moderate to high genetic correlations between these subgroups which were represented by a single latent genetic factor with very good fit. This supported our decision to combine all the samples into a comprehensive, overall meta-analysis. Notably but consistent with expectations based on power, the GWAS of the sub-groups each identified a minimal number of GWS associations, verifying that no individual sub-group was responsible for the overall findings. SNP-based heritability varied across these subgroups from 23.7% in the Clinical subgroup to 6.9% in the Community subgroup; this is consistent with the hypothesis that more severe syndromes like those presenting in treatment settings tend towards higher heritability (58–60). We note that the majority of the Clinical subgroup derived from a prior panic disorder GWAS that also reported high heritability (41).
In addition to replication in an independent EUR cohort from 23andMe, we conducted other analyses to test the transferability of our results. First, we tested whether our SNP associations replicate in the MVP African ancestry sample. We could identify nominally significant proxy loci for all 58 lead SNPs in their immediate regions; however, only two showed significant association with ANX case-control status after Bonferroni adjustment. This is not surprising due to both ancestry and ascertainment differences. Second, we applied PRS to estimate the variance explained in liability to ANX case-control status in the UK Biobank after removing all UK datasets to avoid any potential sample overlap. The PRS explained 2.27% of the variance in those individuals of European ancestry, which is comparable with PRS reports of MDD (40). We then tested whether our findings would generalize to non-European ancestry samples. The EUR-ANX PRS explained 1.94% of the variance in the South Asian subsample of UKBB (significant) but only 0.54% for the African-ancestry subsample (non-significant), in line with the low replication in the MVP African-ancestry cohort. This shows that for anxiety, as for other phenotypes, genetic liability estimated from European samples more closely reflects that of South Asian than African ancestry (31). These findings underscore the importance of explicitly including data from African and other ancestry groups in future ANX GWAS.
We applied LDSC to estimate the degree of genome-wide correlation of ANX with a wide range of psychiatric, behavioral, and medical phenotypes. Consistent with prior psychiatric twin studies and extant GWASs, ANX shares the largest genetic overlap with MDD (rG = 0.91) with which it has the highest lifetime comorbidity. The next highest SNP-based genetic correlation with ANX was PTSD (rG = 0.71). This is not unexpected given their high comorbidity and the prior classification of PTSD amongst the anxiety disorders (61), although this correlation is over twice that estimated in an early twin study (28). The genetic correlation with total neuroticism (rG = 0.7) as well as some of its items and sub-clusters was similarly high. Neuroticism is an important predisposing personality trait for both ANX and MDD; prior twin studies suggest this is due to partially shared genetic factors (28). In addition, ANX has moderate, but still noteworthy, genetic correlations with other lifetime comorbid disorders including ADHD (rG = 0.42), OCD (rG = 0.41), schizophrenia (rG = 0.41), bipolar disorder (rG = 0.34), and anorexia nervosa (rG = 0.33). ANX also correlates highly with childhood internalizing symptoms (62) in keeping with the genetic continuity seen across development (63,64). Among behavioral phenotypes, we highlight the substantial genetic correlation with suicide attempt (rG = 0.58). While some of this is likely driven by comorbid depression, anxiety disorders have been shown to independently increase risk for suicide (65).
Follow-up MR analyses suggest bidirectional genetic effects between ANX and the most highly correlated outcomes of MDD, PTSD, and neuroticism. Although most studies report that ANX tends to onset in childhood and adolescence prior to MDD (66)(67), some analyses find that these disorders mutually predict each other over time (68,69). Our MR analyses support a higher genetic causation from neuroticism to ANX consistent with the stability of this personality trait (70) and its persistent relationship with psychiatric disorders (71) over time. Somewhat unexpectedly, MR suggests that PTSD is more likely to cause ANX than the reverse. This could be due to confounding (e.g., diagnostic misclassification), ascertainment bias (PTSD presents with more severe symptoms), or because trauma can impact both disorders. Our findings support clinical experience that comorbid internalizing disorders exacerbate each other over the lifespan.
The PheWAS provides a more nuanced, variant-level story of pleiotropy. While 15 SNPs are unique to ANX, 43 SNPs are notably pleiotropic. As illustrated in Figure 3A and 3B, many of the pleiotropic SNPs are implicated in either neuroticism traits or other psychiatric (primarily depressive) disorders. Overall, the PheWAS depicts a complex hierarchical structure of genetic overlap of ANX with loci generally falling into one of three categories. The largest set are highly pleiotropic variants that contribute to many medical, physiological and behavioral outcomes. Next are variants that are primarily relevant for psychiatric and behavioral phenotypes. Finally there is a small group of variants that are, to varying degrees, anxiety specific. We note that identifying significant pleiotropy depends upon the availability of a large set of sufficiently powered GWAS results. Accordingly, the effects of immunological, cardiometabolic, and hematological traits are likely over-stated in our PheWAS due to their relative over-representation in the complex genetic trait literature.
Post-GWAS gene-set and single-cell RNA expression analyses support GABAergic signaling as one potential mechanism underlying ANX genetic risk. This might not be surprising to clinicians, since the drug classes of barbiturates and benzodiazepines produce anxiolysis by binding to the GABAA receptor and enhancing GABA neurotransmission. Indeed, results of our gene-drug analysis included several classes of drugs already successfully used to relieve anxiety.
Given the high comorbidity and genetic correlation of ANX with other psychiatric disorders like MDD or related traits like neuroticism, it is not surprising that many of the loci identified in our analyses have been reported in prior GWASs for other phenotypes. We note that most prior psychiatric GWASs did not explicitly exclude ANX in their phenotyping or analyses, making this common comorbidity and their earlier publication dates likely to influence reported associations. Nonetheless, looking across both our SNP- and gene-based analyses, several of our associated loci stand out as not previously associated with other psychiatric phenotypes. We focus our discussion of specific loci on a few of these novel candidate ANX susceptibility loci.
Figure 2. lists 66 protein-coding genes associated with ANX risk with at least three sources of biological support. Among the narrower set of 29 genes with at least four indicators of additional support, several thus far have not been reported in prior GWAS of psychiatric disorders. Due to the complexity of the MHC region, we will not discuss the three genes in chromosomal band 6p21.32 (TAPBP, ZBTB22, DAXX). Seven others (ZNF502, ZNF501, STAB1, NT5DC2, GNL3, GLT8D1, NEK4) are located within an extended region on chromosome 3p21. Other nearby genes include prior reports of association with depression (55), schizophrenia (72), bipolar disorder (73), and suicide (74), making this region a “hot spot” for overall psychiatric susceptibility. A prior GWAS of neuroticism (39) reported association with STAB1 suggesting this gene’s dual role in anxiety-related phenotypes. Little is known about most of these seven genes beyond their basic cellular functions. Except for the role of GLT8D1 in familial ALS (75), there is scant data regarding their function in human disease. Immediately adjacent to NEK4 are ITIH1 and ITIH3, both previously associated with neuroticism and schizophrenia. A study in rats implicated their murine analogs in anxiety-like behaviors (76). This leaves four genes with strong evidence in the present study and no prior GWAS-based reports: PAX6, PROX2, VAMP2, and HMGN1. We discuss these in greater detail in Supplementary Note 4.
Given similarly high lifetime prevalence, moderate twin-based heritability, and extensive comorbidity, our ANX genetic results should be most comparable to those for MDD among all psychiatric diagnoses. In fact, Wray et al. (40) describe results from their PGC-MDD2 analyses that are highly similar to ours regarding (1) number of GWS SNPs identified per effective sample size, (2) SNP-based heritability, (3) enrichment of non-exonic classes of variants, and (4) proportion of variance explained by PRS. These highly polygenic internalizing disorders require massive sample sizes to detect association signals from the small effects of many common SNPs. From what we have learned about MDD and other complex psychiatric phenotypes, the 58 loci we report herein are likely “the tip of the iceberg” among the many hundreds of loci presumed to underlie individual differences in ANX risk. Thus, further genomic discovery efforts for ANX will demand even larger sample sizes.
This study has several potential limitations. First, heterogeneity in ANX case phenotype assessments - from structured clinical interviews to ICD clinical assignments to self-report diagnoses - limit the validity and resultant power to detect susceptibility variants. As the field of psychiatric GWAS has learned, there is often a trade-off between clinical validity and sample size (60,86). Consistent with this, our largest samples provided the lowest depth of phenotyping. Second, we created the largest number of ANX cases by collapsing across all five of the adult anxiety diagnoses. Although prior twin studies and our current results support this procedure to detect shared ANX risk, this further increases underlying phenotypic heterogeneity. Sufficiently sized samples phenotyped for individual ANX diagnoses will be necessary to disaggregate this heterogeneity and investigate disorder-specific risk. Additionally, a life course perspective indicates that different anxiety disorders may manifest at various developmental stages and exhibit intra-anxiety comorbidity. Genetic contributions to ANX may therefore change over time, making longitudinal analyses particularly informative. Finally, we limited our first large-scale meta-analysis to data from EUR ancestry individuals which limits generalizability. We are developing pipelines to aggregate data across ancestries in order to reach sufficiently powered sample sizes and perform future trans-ancestry GWAS for identifying variants generalizable across populations. In summary, these results represent a major advance in our understanding of the genetic basis of ANX. They provide an initial foundation for future studies to investigate the biological mechanisms leading from genes to anxiety syndromes. It is our sincere hope that this opens new lines of investigation for expanding the clinical armamentarium of the next generation of clinicians who treat individuals affected by these conditions.
Methods
Samples
To maximize sample size and power, we assigned the composite Any Anxiety case status if a participant had at least one of five core adult ANX across their lifetime: GAD, panic disorder, social phobia, agoraphobia, and specific phobias. This amounts to seeking to identify common (pleiotropic) genetic effects shared across this cluster of disorders. We did not exclude comorbid mood or other anxiety-related disorders in the cases. Controls had no lifetime anxiety disorder. Due to the genetic overlap between ANX and depression (87,88), we excluded controls if they had a lifetime comorbid mood disorder like MDD or bipolar disorder. We also excluded from all analyses individuals identified with any diagnosis of severe mental health conditions such as schizophrenia, autism, or intellectual disability. As much as possible, we uniformly applied these criteria across the 36 samples included in this study (Supplementary Table S1). However, like most large-scale psychiatric GWAS, these samples were ascertained and assessed with variable approaches that introduce known and cryptic sources of heterogeneity – see Supplementary Note 2 for details of each study. In an attempt to address such heterogeneity, we classified each of the 36 cohorts into five overall sub-groups based on ascertainment and/or assessment similarities– see Supplementary Table S1. While some cohorts meet criteria for more than one sub-group, cohorts were grouped according to their most defining characteristic (i.e. if a biobank sample relied on self-reported diagnosis, it was categorized into SRPD rather than Biobanks). Cases from the two largest cohorts (MVP and UKBB) were primarily identified by self-reported professional diagnosis (SRPD) (2 cohorts, 53,978 cases and 221,844 controls). UK Biobank also included likely DSM-IV GAD cases assigned from responses to the anxiety section of the Composite International Diagnostic Interview (CIDI) short-form survey included in the Mental Health Questionnaire as described in Purves et al. (32). Community cases were ascertained from either a certain geographic area or from a birth cohort or twin registry (19 cohorts, 14,044 cases and 48,283 controls). Biobank cases were identified in large-scale biobanks or registries using ICD or DSM ANX codes (6 cohorts, 37,714 cases and 420,412 controls). Comorbid cases were primarily recruited for another comorbid psychiatric disorder (3 cohorts, 12,858 cases and 27,514 screened controls). Clinical cases were diagnosed by a healthcare professional in a clinical setting (6 cohorts totaling 3,631 cases and 11,907 screened controls); these were mostly identified for panic disorder.
Our analyses fall into six categories which are described in detail below. These include (i) core GWAS, SNP heritability, and sensitivity analyses including differences between ascertainment groups; (ii) replication and validation of the GWAS SNPs; (iii) characterization and functional annotation of the significant SNPs, (iv) gene-based associations and enrichment; (v) gene-drug associations; and (vi) genetic associations and pleiotropy shared with other traits.
(i) GWAS, SNP-based heritability and sensitivity analyses
Genetic data processing and individual GWAS analyses
Each data set was imputed using either the Haplotype Reference Consortium (89) or the 1000 Genomes Project Phase 3 (90) reference panels, and a GWAS was conducted for each (Supplementary Note 2 for details). The results from the individual GWASs were then harmonized and transformed to ‘daner’ file format following Rapid Imputation and COmpuational PIpeLIne for GWAS (RICOPILI) (91) specifications. All datasets were transformed to genome-build GRCh37 assembly (b37/Hg19). Variants were removed if they had: minor allele frequency (MAF) of less than 1% in cases or controls, imputation quality score (INFO) of less than 0.8 or out-of-bounds (above 1.2), or if the effect measures, p-value or standard error (SE) was missing or out of bounds. The data were aligned to the HRC reference panel. If necessary, variants were flipped to match the orientation in the reference panel, and marker names were standardized to those present in the HRC reference. Any variant that did not overlap with the HRC reference was removed. Strand-ambiguous A/T and C/G SNPs with MAF over 0.4 were removed. For A/T and C/G SNPs with a MAF below 0.4, their allele frequencies were compared to the frequencies in the HRC reference. If the allele frequency matched, meaning it was also the minor allele in the reference, the same strand orientation was maintained. If the frequency didn’t match, meaning it had a frequency greater than 0.5 in the reference, the alleles were assumed to be on different strands and were flipped. As a last step, we applied the software-package DENTIST (Detecting Errors in aNalyses of summary staTISTics) (92) as an additional quality control step and removed SNPs with errors and/or heterogeneity between the summary data and the LD reference by testing the difference between the observed Z-score of each variant and its predicted value from surrounding variants. The removal of SNPs due to DENTIST filters were between 83 and 206,037 SNPs per cohort dataset.
GWAS meta-analysis
The GWAS meta-analysis was performed on the 36 cohorts using an inverse variance weighting method in METAL (93) within RICOPILI. This resulted in over 7.2 million SNPs on all autosomes. The level of heterogeneity between the studies was evaluated through two measures: Cochran’s Q and I2 statistics. Cochran’s Q is a weighted sum of squared differences between the individual study effects and the combined effect across studies with the weights coming from the pooling method. The I2-statistic measures the proportion of variation among studies that is due to heterogeneity and not due to chance. While Q depends on the number of included studies, I2 is independent of the number of cohorts. The genomic control factor lambda (λ) was calculated for each GWAS and for the overall meta-analysis to detect residual population stratification or systematic technical issues. The summary statistics from the GWAS underwent linkage disequilibrium (LD) score regression (LDSC) (94) analysis of high-quality common SNPs (INFO score > 0.9) to calculate the LDSC intercept to distinguish polygenicity from other causes of inflation. The GWS threshold for association was set at P < 5x10 −08. Automated LD-based ‘clumping’ of genome-wide significant single nucleotide polymorphisms was conducted in RICOPILI using plink to facilitate identification of independently associated loci. We defined LD-independent SNPs as those with low LD (r2 < 0.1) to a more significantly associated SNP within a 500 kb window. When loci contained several significant SNPs, the SNP with the lowest p-value in each locus was selected as the lead SNP reported here. In addition to the main meta-analysis, we meta-analyzed similar datasets together according to the sub-group assignments described above.
Internal consistency of the ANX phenotype - sensitivity analyses of ANX ascertainment subgroups
SNP-based heritability estimation and genetic correlations
We used LDSC to calculate the SNP-based heritability of the overall meta-analysis and the five sub-group meta-analyses. Additionally, we employed cross-trait LDSC, a method unaffected by ancestry variations or sample overlap, to compute pairwise genetic correlations among the subgroups. SNP-based heritability is defined as the amount of variation in the phenotype that can be accounted for by all the included SNPs. SNP-based heritability was estimated from the slope of the LD score regression with liability scale heritability calculated based on the assumption of a 20% population prevalence of ANX. To avoid a downward bias in our liability scale heritability estimates (95), the effective sample size across the contributing cohorts was calculated and used as the input sample size for LDSC. The sample prevalence was then specified as 0.5 for the conversion to the liability scale. Genetic correlation is calculated by estimating the slope from regressing the product of the Z-scores from two separate GWASs onto the LD score. It reflects the genetic covariation between two traits that is captured by all SNPs included in the GWAS. For both heritability estimation and genetic correlation analysis, we used pre-calculated LD scores from samples of European descent in the 1000 Genomes Project which were filtered for SNPs present in the HapMap3 reference panel.
GenomicSEM 1-factor model
As an extension of the genetic correlation analysis, we used genomic structural equation modeling GenomicSEM (43) to model the joint genetic architecture of the five subgroups. We first conducted an exploratory factor analysis (EFA outside of GenomicSEM) followed by a confirmatory factor analysis (CFA with GenomicSEM). EFA suggests the best factor structure that fits the data and estimates the variance explained by the model. CFA identifies specific loadings for each ascertainment subgroup and model-fit parameters. First, the summary statistics were harmonized and filtered (with the munge-function) using HapMap3 as the reference file, using the effective sample size as the input sample size and filtering SNPs to INFO > 0.9 and MAF > 0.01. In a second step, multivariable LDSC was run to obtain the genetic covariance matrix and corresponding sampling covariance matrix using pre-computed European-ancestry LD scores, a sample prevalence of 0.5 and a population prevalence of 0.2. In a third step, we conducted a EFA followed by a CFA using the pre-packaged common factor model in GenomicSEM using diagonally weighted least squares (DWLS) estimation.
(ii) Replication and validation of GWAS SNPs
Replications
After identifying GWS associations in the primary GWAS, lead SNPs were tested for replication in the commercial database of 23andMe. Self-reported ANX cases were individuals who checked “anxiety” in response to either of the following survey questions: “Have you ever been diagnosed with any of the following…” or “What mental health problems have you had? Please check all that apply.” This GWAS includes data from 1,188,813 European ancestry cases and 1,977,833 controls filtered to remove close relatives. We excluded 35,255 samples (1.1%) based on consent as of 2023-06-09, leaving an analyzed sample of 1,175,012 ANX self-report cases and 1,956,379 controls. Participants provided informed consent and participated in the research online under a protocol approved by the external AAHRPP-accredited IRB, Ethical & Independent Review Services (E&I Review). We performed logistic regression assuming an additive model for allelic effects after covarying for age, sex, five PCs, and genotyping platform. Only the first five PCs were included as covariates, as previous work has demonstrated the first five PCs in the 23andMe dataset explain more variance than the first ten PCs from the UK BioBank (96). The p-values are adjusted using the standard genomic control procedure (97) in which the chi-square test statistic is divided by the genome-wide estimated lambda inflation factor λ = 1.491 (SE = 0.024). The estimated SNP heritability is h2 = 0.088 (SE = 0.002) consistent with the estimate from our discovery GWAS.
Further, we conducted a replication analysis of our 58 ANX-associated SNPs in an independent African-ancestry (AFR) sample from MVP comprising 5,664 ANX cases and 26,410 controls. Initially, we assessed the association results of the same 58 SNPs that reached significance in our main EUR-ancestry GWAS. Recognizing that the lead SNP might not necessarily be the causal SNP in this region and considering the differing LD structures between EUR and AFR ancestry groups, we anticipated that the same SNP might not exhibit significant association. However, the genomic region might still be associated in AFR samples. Therefore, we performed a second lookup to identify the most significant SNP within a 50kb window (+/-25kb) to accommodate potential differences in LD across EUR and AFR ancestries (proxy loci). LD between African and European populations was evaluated using R2 and D’ metrics (as reported on https://ldlink.nih.gov/). We considered replication significant at a Bonferroni-corrected significance threshold of 8.62×10−04 (0.05 / 58).
To evaluate the consistency of previous reported ANX-associated SNPs, we performed a look-up of those SNPs in our main GWAS meta-analysis. We restricted the look-up to prior findings from case-control GWAS (as opposed to dimensional, symptom-based GWASs). Of note is that none of the previously published ANX GWAS are independent of our sample but are partially overlapping.
PRS analyses
We further conducted validation of our results with PRS analyses in independent UKBB samples after removing all UK-based samples (UK Biobank and Generation Scotland) from the primary GWAS. We defined ANX cases as meeting one of the following three criteria. The first criterion involved meeting a likely lifetime diagnosis of DSM-IV GAD determined by assessing anxiety-related questions from the Composite International Diagnostic Interview (CIDI) short-form questionnaire (98). The second was SRPD of one of the five core anxiety disorders (GAD, panic disorder, social phobia, agoraphobia, specific phobia). The third was having a score of ≥ 10 on the GAD-7 (99), a screening measure for anxiety symptoms over the past two weeks. This is a standard threshold for recent moderate generalized anxiety symptoms as measured by the GAD-7; we and others have shown previously that it has a genetic correlation of around 0.9 with lifetime diagnosis of any anxiety disorder (32,34). Data were available for the first criteria from the first UK Biobank Mental Health Questionnaire (100) and for the second and third from both the first and second UK Biobank Mental Health Questionnaires. Controls were defined as for our GWAS, such that we excluded individuals who met criteria for or reported being given diagnoses of any anxiety disorder or major depression or for any major psychiatric condition (e.g. bipolar disorder, schizophrenia). We grouped individuals into three ancestry groups: European, African, and South Asian.
We generated polygenic risk scores using genomic data from the UK Biobank and summary statistics of the GWAS after removing data from all UK samples. MegaPRS (101) implements polygenic scoring approaches using the LDAK heritability model, where the variance explained by each SNP depends on its allele frequency, linkage disequilibrium and functional annotations. MegaPRS for ANX was calculated using the GenoPred (102) pipeline for participants of European, African and South Asian ancestry. Furthermore, logistic regression was run to estimate the PRS prediction effect for ANX adjusting for genotyping batch, assessment center, and 10 genetic principal components.
(iii) Characterization and functional annotation of GWAS SNPs
Variant fine mapping
We conducted statistical fine mapping using FINEMAP v.1.3.1 (45). Only variants located in a region of 1 Mb around index variants were included in the analyses. We used the default k = 5 maximum number of SNP in credible sets and the significant/suggestive threshold for signals was set at 95%/50% total posterior probability for the variants in credible sets (see Supplementary Table S10).
FUMA/Functional Annotation (eQTL/HiC)
We used FUMA v1.6.1 to examine the functional significance of our GWS loci. We compared results from brain-related eQTL studies to identify overlap in significance between our GWAS SNPs and the eQTL results. Furthermore, we used brain-related HiC information available through FUMA to functionally annotate our results. Standard settings were applied and results visualized using FUMA’s inbuilt circos plot routine. More information about the individual third-party datasets (available through the FUMA website) included in the analyses can be found in Supplementary Note 3 or online in FUMA’s tutorial (https://fuma.ctglab.nl/tutorial).
Stratified LDSC
The overall SNP heritability was partitioned first into 53 overlapping functional genomic categories (103) and then into 220 cell-type-specific regulatory elements based on GTEx data and data from the Franke Lab (104). In both partitioned heritability analyses, we regressed the χ2 from the meta-analysis summary statistics onto LD scores downloaded from https://console.cloud.google.com/storage/browser/broad-alkesgroup-public-requester-pays. European ancestry allele frequencies derived from the 1000 Genome Project data were used as the reference genomes in both analyses. The enrichment of a functional or cell-type specific category was defined as the proportion of SNP heritability in the category divided by the proportion of SNPs in that category.
FUMA/Cell type/tissue enrichment
We used MAGMA v1.08 (47) as implemented in the FUMA v1.6.1 (48) to perform tissue enrichment and cell type enrichment analyses. For tissue enrichment analyses, we considered a set of 30 tissue groupings (average enrichment across all tissues in these groups) and 54 individual tissues (with 13 individual tissues from the ‘Brain’ group). Default settings were applied for all above-mentioned analyses. More information about the individual third-party datasets (available through the FUMA website) included in the analyses can be found in Supplementary Note 3 or online in FUMA’s tutorial (https://fuma.ctglab.nl/tutorial).
(iv) Gene-based associations and enrichment
MAGMA – gene-based GWAS and gene-set analysis
We performed gene-based analysis and gene-set analysis using MAGMA (47) v1.08 as implemented in FUMA (48) v1.6.1. To test genetic associations at the gene level for the combined effect of SNPs in or near protein coding genes, we applied default settings (SNP-wise model for gene-based analysis and competitive model for gene-set analysis). Gene-based p-values were computed by mapping SNPs to their corresponding gene(s) based on their position in the genome. Positional mapping was based on ANNOVAR annotations, and the maximum distance between SNPs and genes was set to 10 kb (default). A multiple regression model was employed while accounting for linkage disequilibrium (LD) between the markers. The 1000 Genomes phase 3 reference panel (105), excluding the MHC region, was used to adjust for gene size and LD across SNPs. Using the result of the gene-based analysis (gene level p-values), competitive gene-set analysis was performed with default parameters: 15,496 gene sets were tested for association. Gene sets were obtained from MSigDB v7.0 (see www.gsea-msigdb.org for details) including ‘Curated gene sets’ consisting of nine data resources including KEGG, Reactome, and BioCarta, and ‘GO terms’ consisting of three categories (biological processes, cellular components, and molecular functions).
T-SMR, P-SMR, M-SMR with summary-data based mendelian randomization (SMR)
This type of analysis is used to assess likely causal association between traits and gene/proteins/genomic regions by inferring the association between trait and gene expression/protein concentration/methylation as predicted from genomic data. We performed transcriptome/proteome/methylome -wide association studies (T/P/M-SMR – denoted collectively as X-SMR) using SMR (v.1.03) (49) in conjunction with the largest available external blood and brain xQTL reference data sets (Supplementary Table S16). When pQTL summary statistics from reference data were not available (blood and brain pQTL) in the SMR-required input binary file format (i.e., .besd), we processed them into the required format. One advantage of SMR over competing tools is their inclusion of HEterogeneity In Dependent Instruments (HEIDI) test which can be used as a proxy for likely causality.
SMR analyses were based on cis-xQTLs (SNPs with P < 5×10−8 within 2 MBP of the probe). We also used the default maximum (20) and minimum (3) number of xQTLs selected for the HEIDI test. We set the significance threshold as P < 1.57×10−3 for xQTL and the mismatch of minimum allele frequency among input files as < 15%. For the HEIDI test, SNPs with LD > 0.9 and < 0.05 with top associated xQTL SNP were pruned.
To prioritize genes and perform pathway analyses, we adjusted probe (RNA/protein/CpG) SMR p-value (PSMR) for HEIDI test p-value (PHEIDI) by combining the two p-values into a single one by requiring that i) P was not penalized when P was above 0.01 and ii) P was penalized by the amount P fell below 0.01. Consequently, we adjusted PSMR to 
. We used this approach instead of filtering by PHEIDI < 0. 01, because a misalignment between the GWAS cohort population and the European LD reference panel used by SMR might yield very low PHEIDI. We previously arrived at this compromise between the two types of SMR p-values when applying this approach to many psychiatric disorders (51), e.g., the well-known SCZ C4A signal yielded a T-SMR PHEIDI = 5. 94x10 ™4 but a much lower PSMR. However, for researchers who prefer to use the more conservative approach based on strict PHEIDI thresholds described in the SMR paper (49), we also provide gene PHEIDI values for all SMR analyses, as documented in Supplemental Tables S15-S17.
(v) Gene-drug associations
To uncover potential repurposing of existing drugs to ANX, we conducted gene-drug interaction analyses by applying the DrugTargetor (52), method (v.1.3) to ANX summary statistics. DrugTargetor assesses the association of individual drugs or small molecule related gene sets and drug class enrichment. The method employed two drug-gene interaction databases ChEMB (53,106) and DgiDB (54). The analysis used the following settings: 1) hypothesized action for nervous system, 2) both drug class and single drug, and 3) 1500 max number of unique drugs and 200 max classes of drugs. Please see Supplementary Tables S22-S23 and the README tab for the source databases used to accumulate the gene-sets. Analyses were run using MAGMA version 1.10 (47)using gene flanks of −35kb 5’ and +10kb 3’ (107). Drug class enrichment was calculated using area under the curve defined by the % of drug class gene-sets vs their rank in all the gene-sets (108).
(vi) Genetic overlap between ANX and other phenotypes
PheWAS
Using the identified 58 GWS SNPs, we conducted a phenome-wide association study (PheWAS) to identify the variants that have been significantly associated with other psychiatric, physiological, medical, and behavioral traits in prior GWAS, using the phewas function from the R packages ieugwasr (109). The R package uses publicly available GWAS data from over 10,000 studies compiled by the IEU Open GWAS Project (109,110). The PheWAS used the following databases:
ebi-a: Datasets that satisfy minimum requirements imported from the EBI database of complete GWAS summary data
finn-b: FinnGen study Data Freeze 5
ieu-a: GWAS summary datasets generated by many different consortia that have been manually collected and curated, initially developed for MR-Base
ieu-b: GWAS summary datasets generated by many different consortia that have been manually collected and curated, initially developed for MR-Base (round 2)
ubm-a: Complete GWAS summary data on brain region volumes as described by Elliott et al 2018
ukb-d: Neale lab analysis of UK Biobank phenotypes, round 2
This combination of databases provides the maximum coverage of published GWAS summary statistics that could be used for the PheWAS while minimizing duplication. To increase the accuracy of the PheWAS and consistency of the results across analyses for psychiatric disorders and related behavioral phenotypes, we supplemented the default GWAS summary statistics from the IEU Open GWAS Project for the traits we curated for the genetic correlation analyses. Curating the primary psychiatric and behavioral studies removed duplication from sequential GWAS analyses of the key disorders. We required that a SNP’s p-value was GWS in both the current ANX GWAS and the alternative GWAS. Figure 2A was constructed using edited combinations of the following packages in R: alluvial (111), qqman (112), and pheatmap (113).
Cross-trait genetic correlations
We used cross-trait LDSC to compute genetic correlations between the ANX meta-analysis and 112 selected disorders and traits with publicly available summary statistics. The sources of GWAS summary statistics can be found in the Supplementary Table S24. Details of cross-trait LDSC can be found in the methods section “SNP heritability estimation and genetic correlations’’. As a follow-up, we also calculated genetic correlations between the 112 phenotypes and each ascertainment-specific sub-cohort and compared the genetic correlation patterns between the four groups.
Generalized summary-data based mendelian randomization (GSMR)
We performed bi-directional GSMR (57) analyses for trait pairs (ANX with MDD (114), PTSD (56) and neuroticism (39)) using GSMR v1.1.1 available in the GSMR R package. We employed commonly used parameters: (a) 5×10−8 threshold for significant GWAS signals, (b) the original HEIDI outlier method, (c) single and multi-SNP HEIDI outlier P = 0.01, (d) LD threshold for selecting MR SNP instruments = 0.05, and (e) FDR threshold = 0.05. LD between SNPs with significant signals in at least one trait were computed using GCTA (115) v.1.94.1 based on the 1000 Genome (105) European genetic data.
Data Availability
Summary statistics excluding 23andMe will be made available on the PGC data-download page (https://pgc.unc.edu/for-researchers/download-results/). The replication GWAS summary statistics for the 23andMe data set will be made available through 23andMe to qualified researchers under an agreement with 23andMe that protects the privacy of the 23andMe participants. Datasets will be made available at no cost for academic use. Please visit https://research.23andme.com/research-innovation-collaborations/ for more information and to apply to access the data.
Conflicts of interest
Per Hoffmann receives Salary from the Life & Brain GmbH, Bonn, Germany. James L. Kennedy is a member of the Scientific Advisory Board for Myriad Neuroscience Inc. Ian B. Hickie was an inaugural Commissioner on Australia’s National Mental Health Commission (2012-18). He is the Co-Director, Health and Policy at the Brain and Mind Centre (BMC) University of Sydney. The BMC operates an early-intervention youth services at Camperdown under contract to headspace. He is the Chief Scientific Advisor to, and a 5% equity shareholder in, InnoWell Pty Ltd. InnoWell was formed by the University of Sydney (45% equity) and PwC (Australia; 45% equity) to deliver the $30 M Australian Government-funded Project Synergy (2017-20; a three-year program for the transformation of mental health services) and to lead transformation of mental health services internationally through the use of innovative technologies. Andrew M. Mcintosh has received research support from Eli Lilly, Janssen, and The Sackler Trust. AMM has also received speaker fees from Illumina and Janssen. Murray B. Stein has in the past 3 years received consulting income from Acadia Pharmaceuticals, Aptinyx, atai Life Sciences, Boehringer Ingelheim, Bionomics, BioXcel Therapeutics, Clexio, Eisai, EmpowerPharm, Engrail Therapeutics, Janssen, Jazz Pharmaceuticals, and Roche/Genentech. Dr. Stein has stock options in Oxeia Biopharmaceuticals and EpiVario. He is paid for his editorial work on Depression and Anxiety (Editor-in-Chief), Biological Psychiatry (Deputy Editor), and UpToDate (Co-Editor-in-Chief for Psychiatry). He has also received research support from NIH, Department of Veterans Affairs, and the Department of Defense. He is on the scientific advisory board for the Brain and Behavior Research Foundation and the Anxiety and Depression Association of America. Joel Gelernter is named as an inventor on PCT patent application #15/878,640 entitled: “Genotype-guided dosing of opioid agonists,” filed January 24, 2018 and issued on January 26, 2021 as U.S. Patent No. 10,900,082; and is paid for editorial work for the journal “Complex Psychiatry.” Iiris Hovatta received speaker’s honoraria from Lundbeck. Ole A. Andreassen received speaker’s honorarium from Lundbeck and Sunovion, consultant for Cortechs.ai and Precision Health AS. Katharina Domschke has been a member of the Steering Committee Neurosciences, Janssen, Inc. until 2022 and is currently a member of the Board of the German National Society of Psychiatry (DGPPN) and the Neurotorium Editorial Board of the Lundbeck Foundation. Jordan W. Smoller is a member of the Scientific Advisory Board of Sensorium Therapeutics (with equity) and has received an honorarium for an internal seminar Tempus Labs. He is PI of a collaborative study of the genetics of depression and bipolar disorder sponsored by 23andMe for which 23andMe provides analysis time as in-kind support but no payments. Eduard Maron has received research support and has also received speaker fees from Lundbeck. Hans J. Grabe has received travel grants and speakers honoraria from Indorsia, Neuraxpharm, Servier and Janssen Cilag. Henrik Larsson has served as a speaker for Evolan Pharma, Medici and Shire/Takeda and has received research grants from Shire/Takeda; all outside the submitted work. Gerome Breen is an advisory board member for Compass Pathways. Jürgen Deckert is a member of the board of the German Society of Biological Psychiatry and is on the scientific advisory boards of non-profit organizations and foundations. Volker Arolt worked as an advisor for Sanofi-Adventis Germany. Zach Fuller and Xin Wang are employees of 23andMe and hold stock or stock options in 23andMe. All other authors have no competing interests to declare.
Funding
This work was supported by grant NIH R01MH113665 (PI: John M. Hettema) and the PGC4 grant 5R01MH124847 (MPI: PF Sullivan). Michelle K. Lupton is supported by a Boosting Dementia Leadership Fellowship [APP1140441]. Silviu-Alin Bacanu is funded by the following grants: MH113665, MH118239. Rosa Cheesman is funded by the Jacobs Foundation (no. 2023-1510-00) and the Research Council of Norway (288083). Annika B. Faucon is funded by the following grants: T32GM080178. Sarah Medland NHMRC APP1172917. Stephan Ripke 1U01MH109528 01. Swapnil Awashi has received money from the Stanley Center Gift 2020. Roseann E. Peterson is supported by NIMH R01MH125938 and The Brain & Behavior Research Foundation NARSAD grant 28632 P&S Fund. Matthew H. Iveson was supported by a Wellcome Trust Mental Health award (226770/Z/22/Z), a Medical Research Council Mental Health Data Pathfinder award (MRC-MC_PC_17209) and by the UK Research and Innovation-funded DATAMIND project (MR/W014386/1). Bendik S. Winsvold: This work was supported by the South-Eastern Norway Regional Health Authority (grant no. 2020034). Abigail R. ter Kuile was funded by the NIHR Maudsley Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London (The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care). Daniel F. Levey was supported by an NARSAD Young Investigator Grant from the Brain & Behavior Research Foundation and a Career Development Award CDA-2 from the Veterans Affairs Office of Research and Development (1IK2BX005058-01A2). Baptise Couvy-Duchesne is supported by a CJ Martin Fellowship, awarded by the NHMRC (app. 1161356). Sandra van der Auwera: This project was supported by the Federal Ministry of Education and Research (BMBF, gr. no 01KU2004) under the frame of ERA PerMed (TRAJECTOME project, ERAPERMED2019-108). Giovanni A. Salum is supported by the National Institute of Developmental Psychiatry for Children and Adolescents, São Paulo, with grants from the São Paulo Research Foundation (Fapesp 2014/50917-0; 2013/08531-5) and the Brazilian National Council for Scientific and Technological Development (CNPq 465550/2014-2). James L. Kennedy is supported by the Tanenbaum Family Foundation. Clement C. Zai is supported by the Brain and Behavior Research Foundation (NARSAD), CAMH Foundation. Gwyneth Zai is supported by the Brain and Behavior Research Foundation (BBRF), the Physicians’ Services Incorporated (PSI) Foundation, the International OCD Foundation (IOCDF), the University of Toronto Department of Psychiatry Academic Scholars Award, and the CAMH AFP Innovation Fund.. Arun K. Tiwari is supported by the Ministry of Research and Innovation of Ontario, CAMH foundation and Canadian Institutes of Health Research. Jaakko Kaprio is supported by the Academy of Finland (grant 312073). Joseph Boden is supported by the Health Research Council of New Zealand Programme Grant 16/600. Christel M. Middeldorp: This study was funded by grants received from the National Health and Medical Research Council (NHMRC) and Australian Research Council (ARC). Fabiana L. Lopes is supported by the National Institute of Mental Health R25 MH101076. Francis J. McMahon and Nirmala Akula are supported in part by the Intramural Research Program of the NIMH (ZIAMH002843). Henning Tiemeier was supported by grant 016.VICI.170.200 from Netherlands Organization for Health Research and Development. Naomi R. Wray was supported by grant numbers NHMRC 1113400, 1173790. Enda M. Byrne was supported by the National Health and Medical Research Council grant 1145645. Glyn Lewis was supported by the Wellcome Trust 084268/Z/07/Z. UCLH BRC. Nicholas J. Timpson is a Wellcome Trust Investigator (202802/Z/16/Z), is the PI of the Avon Longitudinal Study of Parents and Children (MRC & WT 217065/Z/19/Z), is supported by the University of Bristol NIHR Biomedical Research Centre (BRC-1215-2001), the MRC Integrative Epidemiology Unit (MC_UU_00011/1) and works within the CRUK Integrative Cancer Epidemiology Programme (C18281/A29019). Lea K. Davis CTSA (SD, Vanderbilt Resources): The project described was supported by the National Center for Research Resources, Grant UL1 RR024975-01, and is now at the National Center for Advancing Translational Sciences, Grant 2 UL1 TR000445-06. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Nathan A Gillepsie was supported by grant R00DA023549. Work by Kristi Krebs, Lili Milani and Kelli Lehto was funded by the European Union through the European Regional Development Fund Project No. 2014-2020.4.01.15-0012 GENTRANSMED, the Estonian Research Council (PRG184, PSG615). Data analysis was carried out in part in the High-Performance Computing Center of University of Tartu. Andreas J. Forstner received funding from the Else Kröner-Fresenius-Stiftung (2019_A127). Markus M. Nöthen is a member of the DFG-funded Excellence Cluster ImmunoSensation2 (EXC 2151 – 390873048). Iiris Hovetta has received support from the Sigrid Jusélius Foundation. Elisa M. Tasanko has received support from the Sigrid Jusélius Foundation. Hermine H. Maes: This research was supported by the National Institute on Drug Abuse (U01DA024413, R01DA025109, R01DA054313), the National Institute of Mental Health (R01MH045268, R01MH068521). Andrew M. McIntosh is supported by the Wellcome Trust (104036/Z/14/Z, 216767/Z/19/Z, 220857/Z/20/Z, 223165/Z/21/Z, 226770/Z/22/Z) and UKRI MRC (MC_PC_17209, MR/S035818/1). This work is part of a project that has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 847776. Ole A. Andreassen has received support from KG Jebsen Stiftelsen (SKGJ-MED-021), Research Council of Norway (223273, 324252, 324499), UiO LifeScience Program, NordForsk #164218. Ted Reichborn-Kjennerud is supported by the Research Council of Norway (274611, PI: Reichborn-Kjennerud). Jacob M. Najman is supported by the National Health and Medical Research Council. Murray B. Stein is supported by the USA Department of Defense, USA Army, and NIMH. Joel Gelernter is supported by the Department of Veterans Affairs Office of Research and Development, USVA, grant I01CX001849. Dorret I. Boomsma is supported through the Royal Netherlands Academy of Science Award (KNAW PAH/6635). NWO 480-15-001/674: Netherlands Twin Registry Repository; Biobanking and Biomolecular Research Infrastructure (NWO BBMRI–NL, 184.033.111); NWO/SPI 56-464-14192, Genetic Association Information Network (GAIN) of the Foundation for the National Institutes of Health, Rutgers University Cell and DNA Repository (NIMH U24 MH 068457-06), the Avera Institute, Sioux Falls (USA) and the National Institutes of Health (NIH R01 HD042157-01A1, MH081802, Grand Opportunity grants 1RC2 MH089951 and 1RC2 MH089995) and European Research Council (ERC-230374). Christian Rück is supported by the Swedish Research Council (2018-02487), Swedish Research Council for Health, Working Life and Welfare ((2018-00221 and 2021-00132). Tilo T.J. Kircher & Georg W. Alpers: The study was funded by the BMBF. Volker Arolt was supported by the German Ministry of Education and Research and was funded by EU Horizon 2020 (Project MOODSTRATIFICATION). Jordan W. Smoller was supported by grant NIMH R01 MH085542. Andreas Reif is supported by DFG TR CRC 58, BMBF Panic-Net. Gerome Breen and Thalia Eley are supported by MR/V012878/1. Anna R. Docherty is supported by R01 MH123619, R01 MH123489. Hilary Coon is supported by R01MH122412, R01MH123489. Kelli Lehto is supported by the Estonian Research Council grant no PSG615. Jürgen Deckert has received support by the EU, BMBF, BMEARDE and DFG. Tan-Hoang Nguyen is supported by NIAAA K25AA030072 and The Brain & Behavior Research Foundation NARSAD grant 28599. Brittany Mitchell was supported by an Australian NHMRC Investigator Grant (APP2017176). Alex S. F. Kwong was supported by an Economics and Social Research Council (ESRC) Postdoctoral Fellowship (ES/V011650/1). Brenda W. Penninx is supported by the research project ‘Stress in Action’, financially supported by the Dutch Research Council and the Dutch Ministry of Education, Culture and Science (NWO gravitation grant number 024.005.010). The UK Medical Research Council and Wellcome (Grant Ref: 217065/Z/19/Z) and the University of Bristol provide core support for ALSPAC. GWAS data was generated by Sample Logistics and Genotyping Facilities at Wellcome Sanger Institute and LabCorp (Laboratory Corporation of America) using support from 23andMe. A comprehensive list of grant funding is available on the ALSPAC website. Vanderbilt University Medical Center’s BioVU is supported by numerous sources: institutional funding, private agencies, and federal grants. These include the NIH funded Shared Instrumentation Grant S10RR025141; and CTSA grants UL1TR002243, UL1TR000445, and UL1RR024975. Genomic data are also supported by investigator-led projects that include U01HG004798, R01NS032830, RC2GM092618, P50GM115305, U01HG006378, U19HL065962, R01HD074711; and additional funding sources listed at https://victr.vumc.org/biovu-funding/. Generation Scotland received core funding from the Chief Scientist Office of the Scottish Government Health Directorate CZD/16/6 and the Scottish Funding Council HR03006. Genotyping of the Generation Scotland samples was carried out by the Genetics Core Laboratory at the Wellcome Trust Clinical Research Facility, Edinburgh, Scotland and was funded by the Medical Research Council and the Wellcome Trust (Reference 104036/Z/14/Z, 220857/Z/20/Z). This work has made use of the resources provided by the Edinburgh Compute and Data Facility (ECDF). (http://www.ecdf.ed.ac.uk/). MoBa thanks the Norwegian Institute of Public Health for generating high-quality genomic data. This research is part of the HARVEST collaboration, supported by the Research Council of Norway (RCN) (#229624). We also thank the NORMENT Centre for providing genotype data, funded by the RCN (#223273), South East Norway Health Authority (SENHA) and KG Jebsen Stiftelsen. We further thank the Center for Diabetes Research, the University of Bergen for providing genotype data and performing QC and imputation of the data funded by the ERC AdG project SELECTionPREDISPOSED, Stiftelsen Kristian Gerhard Jebsen, Trond Mohn Foundation, the RCN, the Novo Nordisk Foundation, the University of Bergen, and the Western Norway Health Authorities. The RCN supported H. Ask, E. Corfield and T. Reichborn-Kjennerud (#274611, #324620). A. Havdahl L. Hannigan and E. Corfield were supported by SENHA (#2020022, #2018058, #2021045). Partial support for all datasets housed within the Utah Population Data Base is provided by the Huntsman Cancer Institute (HCI), http://www.huntsmancancer.org/, and the HCI Cancer Center Support grant, P30CA42014 from the National Cancer Institute. DNA extraction was performed by the University of Utah Center for Clinical and Translational Science supported by the National Center for Advancing Translational Sciences of the NIH (grant number UL1TR002538). The Swedish Twin Registry is managed by Karolinska Institutet and receives funding through the Swedish Research Council under the grant 2017-00641. The Gedi study was supported by the National Institute on Drug Abuse (U01DA024413, R01DA11301), the National Institute of Mental Health (R01MH063970, R01MH063671, R01MH048085, K01MH093731 and K23MH080230), NARSAD, and the William T. Grant Foundation. We are grateful to all the GSMS and CCC study participants who contributed to this work. The GEDI-VTSABD research was supported by the National Institute on Drug Abuse (U01DA024413, R01DA025109), the National Institute of Mental Health (R01MH045268, R01MH068521). We are grateful to all the VTSABD study participants who contributed to this work. The Christchurch Health and Development Study (CHDS) has been supported by funding from the Health Research Council of New Zealand, the National Child Health Research Foundation (Cure Kids), the Canterbury Medical Research Foundation, the New Zealand Lottery Grants Board, the University of Otago, the Carney Centre for Pharmacogenomics, the James Hume Bequest Fund, US National Institutes of Health grant MH077874 and National Institute on Drug Abuse grant R01DA024413. SHIP-START is part of the Community Medicine Research net of the University of Greifswald, Germany, which is funded by the Federal Ministry of Education and Research (grants no. 01ZZ9603, 01ZZ0103, and 01ZZ0403), the Ministry of Cultural Affairs as well as the Social Ministry of the Federal State of Mecklenburg-West Pomerania, and the network ‘Greifswald Approach to Individualized Medicine (GANI_MED)’ funded by the Federal Ministry of Education and Research (grant 03IS2061A). Genome-wide data have been supported by the Federal Ministry of Education and Research (grant no. 03ZIK012) and a joint grant from Siemens Healthineers, Erlangen, Germany and the Federal State of Mecklenburg-West Pomerania. The University of Greifswald is a member of the Caché Campus program of the InterSystems GmbH. The Trøndelag Health Study (HUNT) is a collaboration between HUNT Research Centre (Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology NTNU), Trøndelag County Council, Central Norway Regional Health Authority, and the Norwegian Institute of Public Health. The genotyping was financed by the National Institute of health (NIH), University of Michigan, The Norwegian Research council, and Central Norway Regional Health Authority and the Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology (NTNU). The genotype quality control and imputation has been conducted by the K.G. Jebsen center for genetic epidemiology, Department of public health and nursing, Faculty of medicine and health sciences, Norwegian University of Science and Technology (NTNU). The iPSYCH team was supported by grants from the Lundbeck Foundation (R102-A9118, R155-2014-1724, and R248-2017-2003), NIH/NIMH (1U01MH109514-01 and 1R01MH124851-01 to ADB) and the Universities and University Hospitals of Aarhus and Copenhagen. The Danish National Biobank resource was supported by the Novo Nordisk Foundation. High-performance computer capacity for handling and statistical analysis of iPSYCH data on the GenomeDK HPC facility was provided by the Center for Genomics and Personalized Medicine and the Centre for Integrative Sequencing, iSEQ, Aarhus University, Denmark (grant to ADB). The infrastructure for the NESDA study (www.nesda.nl) is funded through the Geestkracht program of Netherlands Organization for Health Research and Development (ZonMw, grant number: 10-000-1002) and financial contributions by participating universities and mental health care organizations (VU University Medical Center, GGZ inGeest, Leiden University Medical Center, Leiden University, GGZ Rivierduinen, University Medical Center Groningen, University of Groningen, Lentis, GGZ Friesland, GGZ Drenthe, Rob Giel Onderzoekscentrum). The CoLaus|PsyCoLaus study was supported by unrestricted research grants from GlaxoSmithKline, the Faculty of Biology and Medicine of Lausanne, the Swiss National Science Foundation (grants 3200B0–105993, 3200B0-118308, 33CSCO-122661, 33CS30-139468, 33CS30-148401, 33CS30_177535, 3247730_204523 and 320030_220190) and the Swiss Personalized Health Network (grant 2018DRI01). The BLTS and QIMR adults samples were collected using grant funding awarded from many grant funding bodies including the Australian National Health and Medical Research Council (241944, 339462, 389927, 389875, 389891, 389892, 389938, 442915, 442981, 496675, 496739, 552485, 552498, 613608), the FP-5 GenomEUtwin Project (QLG2-CT-2002-01254), the US National Institutes of Health (NIH grants AA07535, AA10248, AA13320, AA13321, AA13326, AA14041, MH66206, DA12854, DA019951) and the Center for Inherited Disease Research, Baltimore. This also incorporated updated data collected as a part of the PISA study. PISA is funded by a National Health and Medical Research Council (NHMRC) Boosting Dementia Research Initiative - Team Grant [APP1095227]. The AGDS was primarily funded by the National Health and Medical Research Council (NHMRC) of Australia grant 1086683. This work was further supported by NHMRC grants 1145645, 1078901, 1113400 and 1087889 and NIMH. The QSkin study was funded by the NHMRC (grant numbers 1073898, 1058522 and 1123248). NGM is supported through NHMRC investigator grants (1172917, 1173790 and 1172990). PISA is funded by a National Health and Medical Research Council (NHMRC) Boosting Dementia Research Initiative - Team Grant [APP1095227]. The Rotterdam Study is funded by Erasmus Medical Center and Erasmus University, Rotterdam, Netherlands Organization for the Health Research and Development (ZonMw), the Research Institute for Diseases in the Elderly (RIDE), the Ministry of Education, Culture and Science, the Ministry for Health, Welfare and Sports, the European Commission (DG XII), and the Municipality of Rotterdam. TRAILS has been financially supported by various grants from Netherlands Organization for Scientific Research NWO (Medical Research Council program grant GB-MW 940-38-011; ZonMW Brainpower grant 100-001-004; ZonMw Risk Behavior and Dependence grant 60-60600-97-118; ZonMw Culture and Health grant 261-98-710; Social Sciences Council medium-sized investment grants GB-MaGW 480-01-006 and GB-MaGW 480-07-001; Social Sciences Council project grants GB-MaGW 452-04-314 and GB-MaGW 452-06-004; NWO large-sized investment grant 175.010.2003.005; NWO Longitudinal Survey and Panel Funding 481-08-013 and 481-11-001; NWO Vici 016.130.002 and 453-16-007/2735; NWO Gravitation 024.001.003), the Dutch Ministry of Justice (WODC), the European Science Foundation (EuroSTRESS project FP-006), the European Research Council (ERC-2017-STG-757364 en ERC-CoG-2015-681466), Biobanking and Biomolecular Resources Research Infrastructure BBMRI-NL (CP 32), the Gratama foundation, the Jan Dekker foundation, the participating universities, and Accare Centre for Child and Adolescent Psychiatry. MUSP: This study was funded by grants received from the National Health and Medical Research Council (NHMRC) and Australian Research Council (ARC). Thalia Eley: This study presents independent research part-funded by the National Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health. The PanicNet I and II studies have been funded by the BMBF. The FinnGen project is funded by two grants from Business Finland (HUS 4685/31/2016 and UH 4386/31/2016) and the following industry partners: AbbVie Inc., AstraZeneca UK Ltd, Biogen MA Inc., Bristol Myers Squibb (and Celgene Corporation & Celgene International II Sàrl), Genentech Inc., Merck Sharp & Dohme LCC, Pfizer Inc., GlaxoSmithKline Intellectual Property Development Ltd., Sanofi US Services Inc., Maze Therapeutics Inc., Janssen Biotech Inc, Novartis AG, and Boehringer Ingelheim International GmbH.
Acknowledgements
We are extremely grateful to all the families who took part in the ALSPAC study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists and nurses. Part of this data was collected using REDCap, see the REDCap website for details (https://projectredcap.org/resources/citations/). This study includes data from the Norwegian Mother, Father and Child Cohort Study (MoBa) conducted by the Norwegian Institute of Public Health. MoBa is supported by the Norwegian Ministry of Health and Care Services and the Ministry of Education and Research. We are grateful to all the participating families in Norway who take part in this on-going cohort study. We thank the Norwegian Institute of Public Health (NIPH) for generating high-quality genomic data. We acknowledge The Swedish Twin Registry for access to data. We thank the veterans who participate in the VA Million Veteran Program. The Trøndelag Health Study (HUNT) is a collaboration between HUNT Research Centre (Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology NTNU), Trøndelag County Council, Central Norway Regional Health Authority, and the Norwegian Institute of Public Health. The genotyping was financed by the National Institute of health (NIH), University of Michigan, The Norwegian Research council, and Central Norway Regional Health Authority and the Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology (NTNU). The genotype quality control and imputation has been conducted by the K.G. Jebsen center for genetic epidemiology, Department of public health and nursing, Faculty of medicine and health sciences, Norwegian University of Science and Technology (NTNU). We are grateful to the Ministry of Research and Innovation of Ontario, for funding the IMPACT project. We would also like to thank Larry and Judy Tanenbaum for their generous support in creating the Tanenbaum Centre for Pharmacogenetics, which is advancing research for the CAMH Pharmacogenetic Program. We are grateful to all the VTSABD study participants who contributed to this work, and to leadership and guidance offered by the former PI Dr. Lindon J. Eaves. The QIMR twin samples are made available through the generous and willing participation of twins and their families registered at the Australian Twin Registry. We thank Dixie Statham (sample collection); Leanne Wallace, Anthony Caracella and staff of the Molecular Epidemiology Laboratory (DNA processing); David Smyth, Harry Beeby and Daniel Park (IT support). We are also indebted to all of the participants of the Australian Genetics of Depression Study for giving their time to contribute to this study. We wish to thank all the people who helped in the conception, implementation, media campaign and data cleaning. We thank Richard Parker, Simone Cross and Lenore Sullivan for their valuable work coordinating all the administrative and operational aspects of the AGDS project. The generation and management of GWAS genotype data for the Rotterdam Study (RS I, RS II, RS III) was executed by the Human Genotyping Facility of the Genetic Laboratory of the Department of Internal Medicine, Erasmus MC, Rotterdam, Netherlands. The GWAS datasets are supported by Netherlands Organisation of Scientific Research NWO Investments (nr. 175.010.2005.011, 911-03-012), the Genetic Laboratory of the Department of Internal Medicine, Erasmus MC, the Research Institute for Diseases in the Elderly (014-93-015; RIDE2), Netherlands Genomics Initiative (NGI)/Netherlands Organisation for Scientific Research (NWO) Netherlands Consortium for Healthy Aging (NCHA), project nr. 050-060-810. We thank Pascal Arp, Mila Jhamai, Marijn Verkerk, Lizbeth Herrera and Marjolein Peters, MSc, and Carolina Medina-Gomez, MSc, for their help in creating the GWAS database, and Karol Estrada, PhD, Yurii Aulchenko, PhD, and Carolina Medina-Gomez, MSc, for the creation and analysis of imputed data. The authors are grateful to the study participants, the staff from the Rotterdam Study and the participating general practitioners and pharmacists. This research is part of the TRacking Adolescents’ Individual Lives Survey (TRAILS). Participating centers of TRAILS include the University Medical Center and University of Groningen, the University of Utrecht, the Radboud Medical Center Nijmegen, and the Parnassia Bavo group, all in The Netherlands. We are grateful to everyone who participated in this research or worked on this project to make it possible. We thank UK Biobank volunteers for their participation. We thank the National Institute for Health Research, NHS Blood and Transplant, and Health Data Research UK as part of the Digital Innovation Hub Programme. This work was part-funded by the National Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London. Patient and public involvement groups and services were provided by the NIHR KCL-Maudsley Biomedical Research Centre. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care. Gwyneth Zai: We thank all research participants and staff for their valuable time and involvement in our research studies. Matthew H. Iveson: We are grateful to the individuals and families who took part in Generation Scotland and have so generously donated their time and data, as well as the general practitioners and the Scottish School of Primary care for their help in recruiting them. Mark J. Adams: This work has made use of the resources provided by the Edinburgh Compute and Data Facility (ECDF) (http://www.ecdf.ed.ac.uk/). Thanks to the study participants who have continued to participate in the study and to the MUSP study team. Francis J. McMahon: This work utilized the computational resources of the NIH HPC Biowulf cluster. We thank the Mass General Brigham Biobank for providing samples, genomic data, and health information data. We are thankful to the participants of the PanicNet I and II studies. MUSP Thanks to Shelby Marrington (Project Manager) and Greg Shuttlewood (Data Manager) who have supervised the day-to-day management of the study. We also extend our thanks to the mothers and children who have continued to participate in the study. FinnGen: We would like to acknowledge the participants and investigators of the FinnGen study. Following biobanks are acknowledged for delivering biobank samples to FinnGen: Auria Biobank (www.auria.fi/biopankki), THL Biobank (www.thl.fi/biobank), Helsinki Biobank (www.helsinginbiopankki.fi), Biobank Borealis of Northern Finland (https://www.ppshp.fi/Tutkimus-ja-opetus/Biopankki/Pages/Biobank-Borealis-briefly-in-English.aspx), Finnish Clinical Biobank Tampere (www.tays.fi/en-US/Research_and_development/Finnish_Clinical_Biobank_Tampere), Biobank of Eastern Finland (www.ita-suomenbiopankki.fi/en), Central Finland Biobank (www.ksshp.fi/fi-FI/Potilaalle/Biopankki), Finnish Red Cross Blood Service Biobank (www.veripalvelu.fi/verenluovutus/biopankkitoiminta), Terveystalo Biobank (www.terveystalo.com/fi/Yritystietoa/Terveystalo-Biopankki/Biopankki/) and Arctic Biobank (https://www.oulu.fi/en/university/faculties-and-units/faculty-medicine/northern-finland-birth-cohorts-and-arctic-biobank). All Finnish Biobanks are members of BBMRI.fi infrastructure (https://www.bbmri-eric.eu/national-nodes/finland/). Finnish Biobank Cooperative-FINBB (https://finbb.fi/) is the coordinator of BBMRI-ERIC operations in Finland. The Finnish biobank data can be accessed through the Fingenious® services (https://site.fingenious.fi/en/) managed by FINBB. We are thankful to all Estonian Biobank participants for their contribution and to the Estonian Biobank Research Team (Andres Metspalu, Tõnu Esko, Reedik Mägi, Mari Nelis and Georgi Hudjashov) for data collection, genotyping, QC and imputation. 23andMe: We would like to thank the research participants and employees of 23andMe for making this work possible.
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.
- 10.↵
- 11.↵
- 12.
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.
- 19.↵
- 20.↵
- 21.
- 22.↵
- 23.↵
- 24.
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.
- 34.↵
- 35.
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.
- 78.
- 79.
- 80.
- 81.
- 82.
- 83.
- 84.
- 85.
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
- 101.↵
- 102.↵
- 103.↵
- 104.↵
- 105.↵
- 106.↵
- 107.↵
- 108.↵
- 109.↵
- 110.↵
- 111.↵
- 112.↵
- 113.↵
- 114.↵
- 115.↵
