W2. THE ANXIETY OF ASSOCIATION: GWAS, PHEWAS, LABWAS AND GENETIC CORRELATIONS IN SOCIAL ANXIETY DISORDER

Background Social Anxiety Disorder (SAD), characterized by an intense fear of social situations, is a prevalent and heritable disorder, yet it remains understudied compared to other anxiety disorders. While previous genome-wide association studies (GWAS) have begun to identify risk loci, much is still unexplained, particularly across diverse populations. Methods We conducted a cross-ancestry GWAS of SAD using the All of Us (AoU) V8 dataset (N=414,830), including African (AFR)[N=74,576], Admixed American (AMR)[N=70,626], East Asian (EAS)[N=9,707], European (EUR)[N=222,500], South Asian (SAS)[N=5,368] and Middle-Eastern (MID)[N=1,467] participants. SAD diagnosis was based on an ICD-10 code of social phobia and/or answering ‘yes’ to three SAD related self-reported questions. We additionally performed a fixed-effects inverse-variance weighted EUR and cross-ancestry meta-analysis using EUR statistics from FinnGen and UK Biobank[N=788,384] and AoU. To identify phenotypes associated with SAD and to explore similarities and differences with other anxiety disorders we conducted a laboratory‑wide association study (LabWAS), a phenome‑wide association study (PheWAS), and estimated genetic correlations with other traits using LD Score Regression. Results In the EUR meta-analysis, we identified 60 suggestive significant loci (P < 1 × 10⁻⁵) (lead SNP:rs7997602, p=4.29 × 10-6, Closest gene: LINC00378, which is characteristic of Phencyclidine Abuse, also linked to insomnia). The cross-ancestry meta-analysis including all datasets yielded 112 additional suggestive loci (most significant SNP: rs1036995, p=4.55 × 10-7, Closest gene: PCDH9, which has been linked to smoking initiation and major depressive disorder). Genetic correlation analyses indicated a significant strong correlation between SAD and other anxiety disorders (rg=0.67, 4 × 10-9). This suggests that while SAD and anxiety overlap significantly there are also differences. PheWAS identified shared associations with depression and mood disorders, while SAD showed additional enrichment for physical comorbidities such as sleep apnea. LabWAS analyses implicated BMI in both conditions, however, lymphocyte counts emerged as a SAD-specific signal absent in broader anxiety data. Both phenotypes showed strongest genetic correlation with PTSD, with general anxiety being more correlated with gastroesophageal disorders. Discussion These findings advance understanding of the genetic basis of SAD and highlight both shared and ancestry-specific influences.