The prevailing clinical assumption that eating disorders and obsessive-compulsive disorder are distinct psychiatric conditions may need significant revision. At the brain's molecular level, these two conditions share a surprisingly coherent transcriptional profile — a finding with real implications for how researchers design treatments and how clinicians understand treatment-resistant cases that span both diagnoses.

Using postmortem brain tissue from 170 individuals — 86 controls, 57 with eating disorders, and 27 with OCD — researchers performed large-scale transcriptomic profiling of two key regions: the dorsolateral prefrontal cortex (DLPFC) and the caudate nucleus. Eating disorders produced robust, region-specific gene expression changes: 102 differentially expressed genes in the DLPFC and 222 in the caudate at a stringent 1% false discovery rate, with findings independently replicated. OCD yielded no single-cohort significant genes, but a meta-analysis across three datasets identified 57 caudate-associated genes. Despite this surface-level asymmetry, transcriptome-wide correlations between ED and OCD were striking — r = 0.67 in the DLPFC and r = 0.75 in the caudate — signaling shared molecular pathology. A combined ED+OCD analysis identified 815 differentially expressed genes in the caudate, implicating GABAergic signaling, neuroendocrine regulation, mitochondrial metabolism, and CHD8-associated networks. Five genes — WDR6, NCKIPSD, P4HTM, DALRD3, and SHISA5 — showed convergent genetic risk associations across both disorders and brain regions, all clustering on chromosome 3.

This study is among the most rigorously powered postmortem transcriptomic analyses of psychiatric disorders to date, and its cross-disorder design is genuinely novel. The cortico-striatal axis identified here aligns with decades of neuroimaging literature implicating compulsive behavior in both conditions, but moving from circuit-level observations to gene-level specificity is a meaningful mechanistic leap. Key limitations include the postmortem nature of the tissue, which cannot establish causal directionality, and the modest OCD cohort size. Still, the chromosome 3 gene cluster offers concrete targets for transdiagnostic drug development — an area of growing urgency given how poorly both conditions respond to existing pharmacotherapy.