Table 4

Modelling of radiographic progression over time

Univariable regression analysis* β (95% CI)p Value†N‡
Time (years)0.98 (0.84 to 1.12)202
Time×HLAB270.052195
 Progression rate if HLAB27 negative0.70 (0.48 to 0.90)31
 Progression rate if HLAB27 positive1.03 (0.88 to 1.19)164
Time×gender0.005202
 Progression rate in women0.69 (0.57 to 0.80)58
 Progression rate in men1.11 (0.93 to 1.29)144
Time×HLAB27 in women0.95856
Time×HLAB27 in men0.044139
 Progression rate in men HLAB27 negative0.69 (0.34 to 1.03)19
 Progression rate in men HLAB27 positive1.18 (0.98 to 1.38)120
Time×baseline mSASSS<0.001183
 Progression rate in baseline mSASSS <100.69 (0.63 to 0.75)124
 Progression rate in baseline mSASSS ≥101.44 (1.18 to 1.70)59
Time×baseline mSASSS in women<0.00155
 Progression rate in women with baseline mSASSS <100.49 (0.42 to 0.56)44
 Progression rate in women with baseline mSASSS ≥101.41 (0.97 to 1.85)11
Time×baseline mSASSS in men<0.001128
 Progression rate in men with baseline mSASSS <100.81 (0.73 to 0.90)80
 Progression rate in men with baseline mSASSS ≥101.45 (1.16 to 1.74)48
Time×baseline mSASSS in HLAB27 negative0.01328
 Progression rate in HLA-B27 negative with baseline mSASSS <100.75 (0.63 to 0.86)22
 Progression rate in HLA-B27 negative with baseline mSASSS ≥100.54 (−0.20 to 1.28)6
Time×baseline mSASSS in HLAB27 positive<0.001149
 Progression rate in HLA-B27 positive with baseline mSASSS <100.68 (0.61 to 0.75)99
 Progression rate in HLA-B27 positive with baseline mSASSS ≥101.54 (1.27 to 1.81)50
Time×baseline mSASSS in men HLAB27 negative0.01117
 Progression rate in HLA-B27 negative men with baseline mSASSS <100.80 (0.63 to 0.98)12
 Progression rate in HLA-B27 negative men with baseline mSASSS ≥100.45 (−0.40 to 1.29)5
Time×baseline mSASSS in men HLAB27 positive<0.001107
 Progression rate in HLA-B27 positive men with baseline mSASSS <100.82 (0.72 to 0.92)67
 Progression rate in HLA-B27 positive men with baseline mSASSS ≥101.56 (1.25 to 1.87)40
Time×disease duration0.701
Time×symptom duration0.214
  • *Because the dependent variable was the mSASSS and the independent variable was time, the regression coefficient obtained reflects the progression of mSASSS per year.

  • †p values are only shown for the interactions between time and other variables. In the presence of a significant interaction, results are stratified, and the progression in each of the subgroups presented by the regression coefficient.

  • ‡All patients with an mSASSS evaluable at each time point independently of the availability of a 2 year progression score were included in the models.

  • mSASSS, modified Stoke Ankylosing Spondylitis Spine Score.