Abstract
OBJECTIVE: To determine clinical and radiographic determinants of hyperkyphosis in patients with ankylosing spondylitis. METHODS: Spinal hyperkyphosis was assessed by occiput to wall distance (OWD) in 135 patients participating in the OASIS cohort and defined as OWD >0. Disease activity was assessed by the Bath ankylosing spondylitis disease activity index (BASDAI). Wedging of the vertebrae was calculated as the Ha/Hp ratio. Structural damage of the spine was assessed by the modified Stoke ankylosing spondylitis spine score (mSASSS). Hip involvement was assessed by the Bath ankylosing spondylitis radiology index (BASRI) and defined as a score >2. Data were analysed by multiple regression analysis on van der Waerden-normal OWD values, with mean Ha/Hp ratio, mSASSS, hip involvement, and BASDAI as explanatory variables, and age, sex, and disease duration after diagnosis as covariates. RESULTS: 61 patients (45.2%) had an OWD >0 cm. Of these, 81% were male, v 57% in the group with normal OWD (p<0.001). Forty two patients had wedged thoracic vertebrae, and 27 of these (44%) had an increased OWD, compared with 15 of 74 with a normal OWD (20%) (p = 0.005). OWD was correlated with mean wedging of the thoracic spine (r = -0.45, p = 0.01), mSASSS (r = 0.56, p = 0.01), and hip involvement (r = 0.2, p = 0.05). Multivariate analysis showed that mSASSS (standardised beta (stbeta) = 0.52; p<0.001), wedging of the thoracic spine (stbeta = -0.28; p = 0.01), and BASDAI (stbeta = 0.15; p = 0.05) were independent determinants of OWD. CONCLUSIONS: Radiological damage of the cervical and lumbar spine, thoracic wedging, and disease activity are determinants of hyperkyphosis in patients with ankylosing spondylitis. These findings could be important in determining treatment goals in this disease.