The effect of smoking on clinical and radiographic variables, and acute phase reactants in patients with ankylosing spondylitis.

To investigate the association between smoking and clinical, inflammatory and radiographic parameters in patients with ankylosing spondylitis (AS). One hundred and six tumour necrosis factor inhibitor naïve patients with AS were included in the study. The erythrocyte sedimentation rate, C-reactive protein, Bath AS Disease Activity Index (BASDAI), Bath AS Functional Index (BASFI) and modified Stroke AS Spine Score (mSASSS) were assessed cross-sectionally for each patient. Smoking history was obtained, and smoking pack years were calculated. Current smokers had significantly higher BASDAI (p < 0.001) and a trend for higher BASFI (p = 0.059). Ever smokers had significantly higher BASFI (p = 0.035) and a trend for higher mSASSS (p = 0.063) compared to never smokers. Pack years (smoking intensity) were positively correlated with duration of inflammatory back pain (r = 0.628, p < 0.001), BASFI (r = 0.443, p < 0.001) and mSASSS (r = 0.683, p < 0.001). Multivariate regression analyses showed that current smoking was independently associated with a higher BASDAI score [regression coefficient (B) = 14.75, p < 0.001] and increasing pack years were independently associated with higher mSASSS (B = 0.26, p = 0.005). In patients with AS, current smoking was strongly and independently associated with higher disease activity, and cumulative smoking exposure with more radiographic spinal damage. In AS smokers, smoking cessation should be strongly recommended.