Background: Hoarding disorder (HD) has been classed alongside obsessive-compulsive disorder (OCD) for decades, yet its later age of onset, ego-syntonic saving, and limited response to OCD treatments imply a separate biology.Methods: We re-analysed the 2022 genome-wide association meta-analysis of hoarding symptoms with the same three-step pipeline recently applied on a larger 2025 OCD GWAS. The approach combined (1) MAGMA gene-based tests, (2) partitioned heritability by stratified LD-score regression and custom χ² enrichment, and (3) S-PrediXcan transcriptome-wide association in six brain regions. Identical annotation panels—two glutamatergic sets, two pruning sets, a monoaminergic control, and a housekeeping control—were applied to both disorders to allow direct comparison.Results: HD showed no single-variant genome-wide hits but did reveal pathway-level patterns distinct from OCD. Hoarding heritability concentrated in genes supporting adult synaptic plasticity and cellular metabolism, most notably the BDNF → TrkB → mTOR → CREB cascade and sigma-1/CYP homeostatic modules. The strongest nominal signals included predicted down-regulation of NTRK2 and enrichment of several mTOR components. Pruning pathways displayed modest, secondary enrichment. By contrast, OCD heritability was dominated by immune-mediated synaptic elimination, adhesion, and astrocytic support genes, with glutamatergic panels contributing little.Conclusions: The data argue against a single "pruning-driven" mechanism for compulsive disorders. Instead, they support a model in which HD arises mainly from impaired adult synaptic remodeling and metabolic resilience within reward and decision circuits, producing enduring attachment to possessions rather than ritualistic neutralisation. This plasticity framework matches the clinical picture of HD and suggests new treatment directions that enhance circuit flexibility—such as BDNF or mTOR agonism—rather than attempting to reverse developmental pruning defects. Replication in larger, deeply phenotyped hoarding cohorts is needed to confirm these findings and to refine therapeutic targets.