Table 2

Univariable and multivariable metaregression analysis of characteristics to explain heterogeneity of prevalence of extra-articular manifestations in patients with AS

AAUPsoriasisIBD
UnivariableMultivariable*UnivariableMultivariable†UnivariableMultivariable‡
Variableβ (95% CI)p Valueβ (95% CI)p Valueβ (95% CI)p Valueβ (95% CI)p Valueβ (95% CI)p Valueβ (95% CI)p Value
Clinical variables
 Disease duration, year0.06 (0.05 to 0.08)<0.010.05 (0.03 to 0.06)<0.010.02 (−0.01 to 0.04)0.140.02 (0.00 to 0.04)0.04
 % women0.01 (0.00 to 0.02)0.18−0.01 (−0.03 to 0.02)0.640.02 (0.01 to 0.03)0.020.02 (0.00 to 0.03)0.05
 % HLA-B27+patients0.01 (0.00 to 0.02)0.090.00 (−0.01 to 0.02)0.58−0.01 (−0.02 to 0.01)0.28
Geographical area
 EuropeReferenceReferenceReferenceReferenceReferenceReference
 North America0.27 (−0.04 to 0.58)0.080.17 (−0.10 to 0.44)0.21−0.44 (−0.85 to −0.03)0.04−0.44 (−0.85 to −0.03)0.040.05 (−0.32 to 0.43)0.780.15 (−0.22 to 0.51)0.43
 Latin America−0.51 (−0.85 to −0.15)<0.01−0.65 (−1.00 to −0.30)<0.01−0.06 (−1.33 to 1.21)0.93−0.06 (−1.33 to 1.21)0.930.32 (−0.32 to 0.96)0.330.40 (−0.23 to 1.03)0.21
 Asia−0.42 (−0.64 to −0.19)<0.01−0.22 (−0.45 to −0.00)0.05−1.33 (−2.09 to −0.57)<0.01−1.33 (−2.09 to −0.57)<0.01−0.96 (−1.44 to −0.49)<0.01−0.80 (−1.29 to −0.32)<0.01
 Middle East−0.64 (−0.97 to −0.30)<0.01−0.38 (−0.67 to −0.09)0.01−1.03 (−1.73 to −0.32)<0.01−1.03 (−1.73 to −0.32)<0.01−0.03 (−0.64 to 0.58)0.92−0.01 (−0.61 to 0.59)0.98
 Australia−0.76 (−1.90 to 0.38)0.19–§
Methodological variables
Risk of bias criteria
 Sampling frame¶−0.22 (−0.47 to 0.03)0.09−0.37 (−0.80 to 0.05)0.08−0.21 (−0.64 to 0.23)0.35
 Case definition for AS¶−0.08 (−0.31 to 0.14)0.48−0.20 (−0.58 to 0.19)0.320.02 (−0.31 to 0.36)0.89
 Random selection¶−0.22 (−0.46 to 0.02)0.07−0.24 (−0.43 to −0.04)0.02−0.12 (−0.50 to 0.25)0.530.06 (−0.26 to 0.38)0.70
 Case definition for EAM¶0.00 (−0.20 to 0.20)0.99−0.08 (−0.43 to 0.26)0.64−0.26 (−0.57 to 0.05)0.10
Study design
 Cross-sectionalReferenceReferenceReference
 Prospective cohort0.27 (−0.12 to 0.66)0.110.27 (−0.31 to 0.85)0.370.25 (−0.23 to 0.72)0.31
 Retrospective cohort−0.20 (−0.45 to 0.05)0.120.46 (0.02 to 0.90)0.040.18 (−0.23 to 0.59)0.11
 Case-control study0.75 (−0.35 to 1.86)0.190.02 (−0.81 to 0.84)0.960.36 (−0.42 to 1.14)0.37
 Intervention study0.20 (−0.10 to 0.49)0.190.48 (0.07 to 0.89)0.020.11 (−0.29 to 0.52)0.58
 Sample size0.00 (0.00 to 0.00)0.130.00 (0.00 to 0.00)0.480.00 (0.00 to 0.00)0.62
  • Parameters with a p value >0.30 in univariable analysis were entered in the multivariable model and removed in order of significance if p>0.05 (backward procedure).

  • *Multivariable model of AAU was based on 119 studies.

  • †Multivariable model of psoriasis was based on 56 studies.

  • ‡Multivariable model of IBD was based on 65 studies.

  • §The study on prevalence of uveitis from Australia was excluded from multivariable analysis because data on disease duration were missing.

  • ¶Low risk of bias versus possible or unknown risk of bias. The criterion on response rate was not considered because too many scores were not reported.

  • AAU, acute anterior uveitis; AS, ankylosing spondylitis; EAM, extra-articular manifestation; HLA-B27,Human Leukocyte Antigen-B27; IBD, inflammatory bowel disease.