BACKGROUND: Obsessive-compulsive disorder (OCD) and allergic diseases co-occur far more often than expected, yet genome-wide genetic correlations are typically modest or negative. The objective of this study was to test whether OCD and allergic diseases show convergence at the level of pre-specified biological pathways, especially synaptic pruning and circadian regulation, rather than at the level of broad genome-wide correlation.
METHODS: Using the Multi-marker Analysis of GenoMic Annotation (MAGMA) pipeline, we re-analysed summary statistics from a large allergic-diseases genome-wide association study (GWAS) (N = 360,838). We then used gene set enrichment analysis (GSEA) and cross-tissue transcriptome-wide association study (TWAS) analyses with S-PrediXcan and Genotype-Tissue Expression (GTEx) version 8 prediction models.
RESULTS: The synaptic pruning gene set reached Bonferroni-corrected significance in allergic diseases (p = 6.75 × 10⁻⁵), driven by major histocompatibility complex (MHC) class I genes (HLA-B, HLA-A, HLA-C, TAP2), complement genes (C4A, C4B), and inflammatory regulators (NFKB1, RELA). The glutamatergic signalling set also survived the eight-set Bonferroni threshold (p = 0.005). The REACTOME Circadian Clock set showed supportive enrichment in the MAGMA analysis (raw p = 1.92 × 10⁻³; Bonferroni-adjusted p = 0.015), although this did not meet the pre-specified eight-set Bonferroni threshold of 0.0063. Standard GSEA nevertheless showed REACTOME Circadian Clock enrichment in both allergic diseases and OCD. Cross-tissue TWAS projected allergic risk onto six OCD-relevant brain regions and yielded false discovery rate (FDR)-significant pruning and circadian hits, including TAP2 in amygdala (p = 3.02 × 10⁻¹¹). Projecting OCD risk onto allergy-relevant peripheral tissues produced strong circadian gene-set enrichment, especially for the Kyoto Encyclopedia of Genes and Genomes (KEGG) Circadian Rhythm Mammal set (1.95-fold, p = 1.62 × 10⁻⁸). Integration with published atopic dermatitis transcriptomic studies, rather than new skin transcriptomic profiling in this study, suggested global clock downregulation alongside specific shifts involving ARNTL2, NOCT, and RORC.
CONCLUSIONS: These findings support pathway-level and candidate gene-level convergence between OCD and allergic diseases, centred on pruning biology and circadian regulation. The proposed skin-brain neuroimmune loop is a hypothetical framework, not causal proof, and requires functional and longitudinal validation. The results move beyond broad polygenic correlations and identify immune-circadian pathways that may be worth testing in future mechanistic and clinical studies of comorbid patients.
