Objective  Systematically conduct a screening of the potential targets within circulating inflammatory proteins (CIPs) that have an impact on osteoporosis (OP) and fractures. Validate their association with bone mineral density (BMD) via clinical samples, thereby offering novel evidence for the regulation of bone metabolism.

Methods  The study employs bidirectional Mendelian randomization analysis to investigate the causal relationship between CIPs and OP, as well as fractures. Two groups are included: patients with type 2 diabetes mellitus (T2DM) and healthy individuals. General patient data are collected, and fibroblast growth factor 21 (FGF21) and bone metabolism indicators are measured and compared among the T2DM group with OP (DOP), the T2DM group without OP (NDOP), and the control group.

Results  Among the 91 types of CIPs retrieved from the EBI GWAS Catalog database, the inverse variance weighting (IVW) analysis indicated that FGF21 potentially had a causal relationship with both OP and fractures that transpired within the past five years. An elevation in its level augmented the risk of OP (odds ratio OR = 1.003, 95% CI: 1.001-1.005; P = 0.004) and the risk of fractures in the past five years (OR = 1.004, 95% CI: 1.000-1.008; P = 0.038). Reverse Mendelian randomization demonstrated that OP and fractures in the past five years had no causal impact on the eight CIPs, including FGF21, identified in the forward analysis. Clinical investigations showed that the plasma FGF21 level in the DOP group (396.92 308.98, 523.94 pg/mL) was higher than that in the NDOP group (346.88 283.82, 466.86 pg/mL) and the control group (233.66 169.95, 327.78 pg/mL) (all P < 0.05). In all samples, subsequent to adjusting for age, body mass index, glycated hemoglobin, calcium, and phosphorus, FGF21 was negatively correlated with the T-value of the lumbar spine (β = －0.003, 95% CI: －0.004 to －0.002, P < 0.0001) and the T-value of the total hip (β = －0.002, 95% CI: －0.003 to －0.001, P < 0.0001). In T2DM, FGF21 was negatively correlated with the T-value of the lumbar spine (β = －0.003, 95% CI: －0.004 to －0.0014, P < 0.0001) and the T-value of the total hip (β = －0.001, 95% CI: －0.0022 to －0.0005, P = 0.002), and this negative correlation persisted after adjusting for the aforementioned covariates.

Conclusion FGF21 elevates the risk of OP and fractures that have occurred over the past five years, and a causal relationship exists. It serves as a potential risk factor for impaired bone health in T2DM.