Genetic architecture of lipoma susceptibility implicates telomere biology

Lipomas are common benign adipocytic neoplasms with well-characterized somatic cytogenetic alterations, but the inherited genetic architecture predisposing to their formation remains poorly understood. Recent phenome-wide evidence identified lipoma as strongly associated with genetic predisposition toward longer telomeres. Here, I examined this relationship from the complementary perspective of lipoma genetics. FinnGen R13 genome-wide association and native fine-mapping data resolved nine independent lipoma susceptibility signals across seven genomic regions. Unbiased annotation of these signals implicated telomere-maintenance and genome-stability genes, including telomerase reverse transcriptase (TERT), regulator of telomere elongation helicase 1 (RTEL1), and STN1 subunit of CST complex (STN1), with proximity-expanded mapping additionally encompassing telomerase RNA component (TERC). Using 198 leukocyte telomere length (LTL)-associated genetic variants from All of Us, multiplicative-random-effects inverse-variance-weighted Mendelian randomization showed a strong positive association between genetically proxied LTL and lipoma susceptibility (beta=0.445, 95% CI 0.355-0.535; P=2.49 x 10^-22). The association replicated using independent UK Biobank LTL effects (beta=0.587, 95% CI 0.459-0.715; P=3.09 x 10^-19), remained positive across alternative estimators and sensitivity analyses, persisted after exclusion of 15 variants assigned to canonical telomere/DNA-damage-response genes, and extended across lipomas of the limbs, trunk, and head/face/neck. Regional analyses showed strong evidence for a shared LTL-lipoma association component at TERC (PP.H4=0.989), strong but unresolved regional overlap at TERT (PP.H4=0.924), and distinct association components at STN1 (formerly OBFC1; PP.H3=0.9995), demonstrating heterogeneous local architectures. The study used a human-supervised agentic research workflow in which ChatGPT contributed to analytical planning, interpretation, quality control, and synthesis, while Codex performed substantial computational execution and reproducibility auditing. Together, these findings connect the independently derived genetic architecture of lipoma susceptibility with a broader trade-off between maintenance of cellular replicative capacity and suppression of neoplastic growth.