The relationship between serum FSTL1 levels and bone destruction in patients with ankylosing spondylitis

A typical autoimmune disease, ankylosing spondylitis (AS), is characterized primarily by chronic inflammation of the axial skeleton and destruction of osseous structures. Follistatin-like protein 1 (FSTL1) is a glycoprotein that regulates inflammation and bone metabolism and has been implicated in bone destruction in individuals with disorders such as rheumatoid arthritis. However, its role in AS remains unknown. The present cross-sectional observational study included 70 patients diagnosed with AS and 35 sex- and age-matched healthy controls (HCs). The AS patients were stratified into two subgroups on the basis of disease activity: the active phase ( n = 41) and the inactive phase ( n = 29). Serum levels of FSTL1, RANKL, OPG, TNF-α, MMP‑3, and hs-CRP were quantified by ELISA. Disease activity was evaluated using both BASDAI and ASDAS-CRP. Radiological assessments included X-ray, magnetic resonance imaging (MRI), and ultrasonographic examinations, with the number of bone erosions, the degree of enthesitis, and the presence of power Doppler signals recorded. Spearman’s correlation and negative binomial regression were used to evaluate relationships among variables. Serum FSTL1 levels were significantly higher in patients with AS than in healthy controls [12.60 (6.10, 22.33) ng/mL vs. 4.83 (3.25, 7.75) ng/mL; p < 0.001], and FSTL1 levels were higher in patients with active AS than in patients with inactive AS [21.15 ± 14.16 ng/mL vs. 7.45 (3.61, 15.10) ng/mL; p < 0.001]. FSTL1 levels were significantly positively correlated with the levels of inflammatory markers, including hs-CRP ( ρ s = 0.426), ESR ( ρ s = 0.530), and TNF-α (ρ s = 0.374) ( p < 0.001). Furthermore, serum FSTL1 levels were significantly correlated with the disease activity scores BASDAI ( ρ s = 0.500) and ASDAS-CRP (ρ s = 0.614) ( p < 0.001). In terms of bone destruction markers, serum levels of FSTL1 were positively correlated with the number of bone erosions ( ρ s = 0.443, p < 0.001), serum levels of RANKL (ρ s = 0.331, p = 0.005), serum levels of OPG (ρ s = 0.382, p = 0.001), the RANKL/OPG ratio (ρ s = 0.269, p = 0.024), and serum levels of MMP‑3 (ρ s = 0.414, p < 0.001). Negative binomial regression was applied to model bone erosion count. In crude and partially adjusted models, serum FSTL1 was significantly associated with bone erosion count. After full adjustment for age, disease duration, hs‑CRP and ASDAS‑CRP, this association was no longer statistically significant (OR = 1.0060; 95% CI 0.9962‑1.0161; p = 0.231). Serum FSTL1 levels were elevated in AS patients compared with healthy controls. Serum FSTL1 levels were correlated with AS disease activity and inflammatory status. The associations between FSTL1 and bone erosion require further larger‑scale investigation.