Relationship between diffuse idiopathic skeletal hyperostosis and fragility vertebral fracture: a prospective study in older men.

Abstract

To analyse the risk of incident vertebral and non-vertebral fracture in men with DISH. In 782 men ages 50-85 years, DISH was diagnosed using Resnick's criteria. In men followed prospectively for 7.5 years, a radiographic incident vertebral fracture was defined by a decrease of ≥20% or ≥4mm in any vertebral height vs baseline. Self-reported incident non-vertebral fractures were confirmed by medical records. Men with DISH had higher BMD at the lumbar spine (P < 0.05), but not at other skeletal sites. After adjustment for confounders including disc space narrowing (DSN) and endplate irregularity, the risk of vertebral fracture was higher in men with DISH vs men without DISH [10/164 (6.1%) vs 16/597 (2.7%), P < 0.05; odds ratio (OR) 2.89 (95% CI 1.15, 7.28), P < 0.05]. DISH and low spine BMD were each associated with a higher vertebral fracture risk. The vertebral fracture risk was higher in men who had both DISH and severe DSN. DISH and endplate irregularities (EIs) were each associated with higher vertebral fracture risk. DISH, DSN and EIs define the intervertebral space dysfunction, which was associated with higher vertebral fracture risk [OR 3.99 (95% CI 1.45, 10.98), P < 0.01]. Intervertebral space dysfunction improved the vertebral fracture prediction (ΔAUC = +0.111, P < 0.05), mainly in men with higher spine BMD (>0.9 g/cm2; ΔAUC = +0.189, P < 0.001). DISH was not associated with the risk of non-vertebral fracture. DISH is associated with higher vertebral fracture risk, independently of other risk factors. Assessment of the intervertebral space dysfunction components may improve the vertebral fracture prediction in older men.