Osteoporosis is common in postmenopausal women but is often asymptomatic until a fracture occurs, highlighting the importance of early screening and preventive interventions. This study aimed to develop molecular subtype risk stratification of postmenopausal osteoporosis and analyze the immune infiltration microenvironment. Microarray data for osteoporosis were downloaded and analyzed. Logistic and least absolute shrinkage and selection operator (LASSO) regression analyses were used to construct the molecular risk model. Circulating blood samples were collected from 10 enrolled participants to validate the key differentially expressed genes, and consistent clustering based on the expression profiles of candidate genes was performed to obtain molecular subtypes. Three key genes, CTNNB1, MITF, and TNFSF11, were obtained as variables and used to construct the risk model. External experimental validation showed substantial differences in the three key genes between patients with osteoporosis and the controls (p < 0.05). Three subtypes were obtained based on dimensionality reduction clustering results. Cluster 3 had significantly more patients with low bone mineral density (BMD), whereas Cluster 2 had significantly more patients with high BMD (p < 0.05). This study introduced a novel molecular risk model and subtype classification system, which is an evidence-based screening strategy that will guide the active prevention, early diagnosis, and treatment of osteoporosis in high-risk postmenopausal women.
Keywords: bioinformatics; immune infiltration microenvironment; molecular subtype; osteoclast; osteoporosis; risk stratification.