Abstract
Anorexia nervosa (AN) is a severe psychiatric disorder characterized by substantial heritability and a high mortality rate among psychiatric disorders. While cerebrospinal fluid (CSF) metabolomics has emerged as a novel approach to investigating central nervous system pathologies, its specific causal relationship with anorexia nervosa remains to be fully elucidated. Using genome-wide association study (GWAS) summary statistics for human CSF metabolites and AN information from publicly available datasets, we performed a two-sample Mendelian randomization (MR) analysis using the inverse-variance weighted (IVW) method as the primary approach, complemented by sensitivity analyses. Through a comprehensive analysis of 338 CSF metabolites, we identified six metabolites with significant causal relationships with AN risk. 1-stearoyl-2-linoleoyl-gpc (18:0/18:2) (OR = 1.09, 95% CI 1.00-1.18) and alpha-tocopherol (OR = 1.36, 95% CI 1.00-1.83) showed positive associations, increasing AN risk. Conversely, sphingomyelin (d18:1/20:0, d16:1/22:0) (OR = 0.86, 95% CI 0.77-0.95), 2,3-dihydroxy-2-methylbutyrate (OR = 0.92, 95% CI 0.86-0.98), N-acetylhistidine (OR = 0.92, 95% CI 0.86-0.98), and oxalate (ethanedioate) (OR = 0.83, 95% CI 0.73-0.94) had protective effects, reducing AN risk. Sensitivity analyses showed no evidence of horizontal pleiotropy or heterogeneity in the MR results. An MR directionality test and a Steiger filtering test confirmed the absence of reverse causality, thereby substantiating the robustness of our findings. These findings suggest that these CSF metabolites could serve as potential biomarkers for early AN detection and highlight novel therapeutic targets, potentially improving diagnosis and intervention strategies for this challenging disorder.