Table 5.

Association of either r22axSpA or r3+axSpA with spinal radiographic progression

Type of analysis	Key variable*	Progression defined as ≥2 mSASSS units in 2 years			Progression defined as ≥1 new syndesmophyte in 2 years
		OR	95% CI	P value	OR	95% CI	P value
Based on multiple imputation	r22axSpA versus nr-axSpA	1.76	0.54 to 5.69	0.35	0.88	0.30 to 2.55	0.81
	r3+axSpA versus nr-axSpA	2.81	1.00 to 7.89	0.05	1.60	0.68 to 3.75	0.28
Complete case analysis	r22axSpA versus nr-axSpA	2.67	0.56 to 12.7	0.22	1.31	0.33 to 5.12	0.70
	r3+axSpA versus nr-axSpA	4.72	1.16 to 19.2	0.03	3.05	0.97 to 9.63	0.06

Analyses performed in 724 radiographic intervals from 505 patients after multiple imputation of missing data and in 532 intervals from 394 patients in the complete case analysis. The full model is presented in online supplemental table S6.

*Adjustment for Ankylosing Spondylitis Disease Activity Score, sex, tumour necrosis factor inhibitor use prior to the radiographic interval, non-steroidal anti-inflammatory drug use at start of the radiographic interval, current smoking, length of the radiographic interval, as well as structural spinal damage at start of each radiographic interval (either mSASSS or presence of syndesmophytes).

mSASSS, modified Stoke Ankylosing Spondylitis Spine Score; nr-axSpA, nonradiographic axial spondyloarthritis; r3+axSpA, unilateral/bilateral grade 3–4 radiographic sacroiliitis; r22axSpA, bilateral grade 2 radiographic sacroiliitis; r22axSpA, bilateral grade 2 sacroiliitis.