Repeated Spinal Mobility Measures and Their Association With Radiographic Damage in Ankylosing Spondylitis

Objective We sought to explore the relationship between changes in repeated mobility measures and spinal structural progression in patients with ankylosing spondylitis (AS) over time. Methods We studied patients with AS from the PSOAS (Prospective Study of Outcomes in AS) cohort and performed longitudinal multivariable regression modeling to assess the relationship of structural damage measured by their regional (cervical or lumbar) modified Stoke AS Spinal Score(mSASSS) and selected cervical (eg, cervical rotation, lateral bending, and occiput-to-wall distance) and lumbar spinal mobility measures (eg, Schober's test and lumbar lateral bending) that were collected at least every 2 years from 2003 to 2019. Results The median length of follow-up for our 518 patients with cervical mSASSS measurements and 573 with lumbar mSASSS measurements was 4.08 (interquartile range [IQR] 2.25-6.67) and 4.17 (IQR 2.25-6.67) years, respectively. Among the mobility measures, based on multivariable regression models adjusting for clinical/demographic variables and C-reactive protein, we did not observe meaningful associations between changes in spinal mobility with their respective regional mSASSS. Baseline mSASSS, male sex, increased C-reactive protein (CRP), and longer disease duration were associated with increased longitudinal mSASSS in all analyses. Conclusion Our study shows that 2-year changes in individual spinal mobility measures are not reliably associated with increased, longitudinal, AS-related spinal structural progression. We also confirmed the relationship of baseline mSASSS, sex, CRP, and disease duration with AS-related structural spinal progression over time.